Issue Highlights Editors' Note Editorials

Volume 116, Issue 1; July 3, 2007,PP. 1-124

Issue Highlights

Circulation 2007 116: 1, doi:10.1161/CIRCULATIONAHA.107.183534 Editors' Note

Gary J. Balady and Ravin Davidoff Circulation 2007 116: 2, doi:10.1161/CIRCULATIONAHA.107.184813

Editorials

Cardiovascular Biomarkers: Added Value With an Integrated Approach? Wolfgang Koenig Circulation 2007 116: 3 - 5, doi:10.1161/CIRCULATIONAHA.107.707984

The ST-Segment–Elevation Myocardial Infarction Chain of Survival Joseph P. Ornato Circulation 2007 116: 6 - 9, doi:10.1161/CIRCULATIONAHA.107.710970

Original Articles

Arrhythmia/Electrophysiology

Common NOS1AP Variants Are Associated With a Prolonged QTc Interval in the Rotterdam Study Albert-Jan L.H.J. Aarnoudse, Christopher Newton-Cheh, Paul I.W. de Bakker, Sabine M.J.M. Straus, Jan A. Kors, Albert Hofman, André G. Uitterlinden, Jacqueline C.M. Witteman, and Bruno H.C. Stricker Circulation 2007 116: 10 - 16; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.106.676783 Nonsense Mutations in hERG Cause a Decrease in Mutant mRNA Transcripts by Nonsense-Mediated mRNA Decay in Human Long-QT Syndrome Qiuming Gong, Li Zhang, G. Michael Vincent, Benjamin D. Horne, and Zhengfeng Zhou Circulation 2007 116: 17 - 24; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.107.708818

Coronary Heart Disease

Coronary Artery Calcification Progression Is Heritable Andrea E. Cassidy-Bushrow, Lawrence F. Bielak, Patrick F. Sheedy, II, Stephen T. Turner, Iftikhar J. Kullo, Xihong Lin, and Patricia A. Peyser Circulation 2007 116: 25 - 31; published online before print June 11 2007, doi:10.1161/CIRCULATIONAHA.106.658583

Epidemiology

Association of Carotid Artery Intima-Media Thickness, Plaques, and C-Reactive Protein With Future Cardiovascular Disease and All-Cause Mortality: The Cardiovascular Health Study Jie J. Cao, Alice M. Arnold, Teri A. Manolio, Joseph F. Polak, Bruce M. Psaty, Calvin H. Hirsch, Lewis H. Kuller, and Mary Cushman Circulation 2007 116: 32 - 38; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.106.645606 Abdominal Visceral and Subcutaneous Adipose Tissue Compartments: Association With Metabolic Risk Factors in the Framingham Heart Study Caroline S. Fox, Joseph M. Massaro, Udo Hoffmann, Karla M. Pou, Pal Maurovich-Horvat, Chun-Yu Liu, Ramachandran S. Vasan, Joanne M. Murabito, James B. Meigs, L. Adrienne Cupples, Ralph B. D’Agostino, Sr, and Christopher J. O’Donnell Circulation 2007 116: 39 - 48; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.106.675355 Heart Failure

Metoprolol Reverses Left Ventricular Remodeling in Patients With Asymptomatic Systolic Dysfunction: The REversal of VEntricular Remodeling with Toprol-XL (REVERT) Trial Wilson S. Colucci, Theodore J. Kolias, Kirkwood F. Adams, William F. Armstrong, Jalal K. Ghali, Stephen S. Gottlieb, Barry Greenberg, Michael I. Klibaner, Marrick L. Kukin, Jennifer E. Sugg on behalf of the REVERT Study Group Circulation 2007 116: 49 - 56; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.106.666016

Negative Inotropy of the Gastric Proton Pump Inhibitor Pantoprazole in Myocardium From Humans and Rabbits: Evaluation of Mechanisms Wolfgang Schillinger, Nils Teucher, Samuel Sossalla, Sarah Kettlewell, Carola Werner, Dirk Raddatz, Andreas Elgner, Gero Tenderich, Burkert Pieske, Giuliano Ramadori, Friedrich A. Schöndube, Harald Kögler, Jens Kockskämper, Lars S. Maier, Harald Schwörer, Godfrey L. Smith, and Gerd Hasenfuss Circulation 2007 116: 57 - 66; published online before print June 18 2007, doi:10.1161/CIRCULATIONAHA.106.666008

Interventional Cardiology

Emergency Department Physician Activation of the Catheterization Laboratory and Immediate Transfer to an Immediately Available Catheterization Laboratory Reduce Door-to-Balloon Time in ST-Elevation Myocardial Infarction Umesh N. Khot, Michele L. Johnson, Curtis Ramsey, Monica B. Khot, Randall Todd, Saeed R. Shaikh, and William J. Berg Circulation 2007 116: 67 - 76; published online before print June 11 2007, doi:10.1161/CIRCULATIONAHA.106.677401 Contemporary Reviews in Cardiovascular Medicine

The Brain–Heart Connection Martin A. Samuels Circulation 2007 116: 77 - 84, doi:10.1161/CIRCULATIONAHA.106.678995

Cardiovascular Involvement in General Medical Conditions

Chronic Kidney Disease: Effects on the Cardiovascular System Ernesto L. Schiffrin, Mark L. Lipman, and Johannes F.E. Mann Circulation 2007 116: 85 - 97, doi:10.1161/CIRCULATIONAHA.106.678342

ACCF/AHA/SCAI Clinical Competence Statement

ACCF/AHA/SCAI 2007 Update of the Clinical Competence Statement on Cardiac Interventional Procedures: A Report of the American College of Cardiology Foundation/American Heart Association/American College of Physicians Task Force on Clinical Competence and Training (Writing Committee to Update the 1998 Clinical Competence Statement on Recommendations for the Assessment and Maintenance of Proficiency in Coronary Interventional Procedures)

Circulation 2007 116: 98 - 124; published online before print June 25 2007, doi:10.1161/CIRCULATIONAHA.107.185159

Images in Cardiovascular Medicine

An Unusual Site for a Common Disease Maysaa Alzetani, Joseph J. Boyle, David Lefroy, and Petros Nihoyannopoulos Circulation 2007 116: e1, doi:10.1161/CIRCULATIONAHA.106.677120

Sine-Wave Pattern Arrhythmia and Sudden Paralysis That Result From Severe Hyperkalemia Maurice J.H.M. Pluijmen and Ferry M.R.J. Hersbach Circulation 2007 116: e2 - e4, doi:10.1161/CIRCULATIONAHA.106.687202

Lipomatous Metaplasia in Ischemic Cardiomyopathy: A Common but Unappreciated Entity Matthias Schmitt, Nilesh Samani, and Gerry McCann Circulation 2007 116: e5 - e6, doi:10.1161/CIRCULATIONAHA.107.690800

Correspondence

Letter by Brewster and van Montfrans Regarding Article, "Risks Associated With Statin Therapy: A Systematic Overview of Randomized Clinical Trials" Lizzy M. Brewster and Gert A. van Montfrans Circulation 2007 116: e7, doi:10.1161/CIRCULATIONAHA.107.689497

Letter by Rosenberg and Uretsky Regarding Article, "Risks Associated With Statin Therapy: A Systematic Overview of Randomized Clinical Trials" Lauren Rosenberg and Seth Uretsky Circulation 2007 116: e8, doi:10.1161/CIRCULATIONAHA.107.690867

Response to Letters Regarding Article, "Risks Associated With Statin Therapy: A Systematic Overview of Randomized Clinical Trials" Amir Kashani, JoAnne M. Foody, Yongfei Wang, Harlan M. Krumholz, Christopher O. Phillips, Sandeep Mangalmurti, and Dennis T. Ko Circulation 2007 116: e9, doi:10.1161/CIRCULATIONAHA.107.697227

Acknowledgment of Reviewers

Circulation 2007 116: e10 - e21, doi:10.1161/CIRCULATIONAHA.107.184814

News From the American Heart Association

Circulation 2007 116: 1B - 2B, doi:10.1161/CIRCULATIONAHA.107.184673

Meetings Calendar

Circulation 2007 116: 3B - 4B, doi:10.1161/CIRCULATIONAHA.107.184672

American Heart Association Newly Elected Fellows, Spring 2007

Circulation 2007 116: 5B - 6B, doi:10.1161/CIRCULATIONAHA.107.184674

European Perspectives

Circulation 2007 116: 1F - 6F, doi:10.1161/CIRCULATIONAHA.107.185417

IIssssuuee HHiigghhlliigghhttss Vol 116, No 1, July 3, 2007

ASSOCIATION OF CAROTID ARTERY INTIMA-MEDIA EMERGENCY DEPARTMENT PHYSICIAN ACTIVATION THICKNESS, PLAQUES, AND C-REACTIVE PROTEIN OF THE CATHETERIZATION LABORATORY AND WITH FUTURE CARDIOVASCULAR DISEASE AND IMMEDIATE TRANSFER TO AN IMMEDIATELY ALL-CAUSE MORTALITY: THE CARDIOVASCULAR AVAILABLE CATHETERIZATION LABORATORY HEALTH STUDY, by Cao et al. REDUCE DOOR-TO-BALLOON TIME IN ST-ELEVATION There is increasing interest in methods to risk-stratify individuals’ risk for MYOCARDIAL INFARCTION, by Khot et al. cardiovascular disease. Cao et al examined the ability of C-reactive protein Guidelines recommend that hospitals strive to achieve a door-to-balloon concentrations with or without carotid intima-media thickness and carotid time within 90 minutes based upon considerable observational data. plaques to predict incident cardiovascular events and death in about 5000 Currently, most hospitals are not achieving this goal. National efforts are elderly participants in the Cardiovascular Health Study. The investigators now under way to improve the door-to-balloon times, but the impact of report that C-reactive protein was not prognostically useful without these efforts has not been prospectively evaluated. In this prospective evidence of carotid atherosclerosis. However, they observed an interaction observational study, the impact of a protocol mandating that the between C-reactive protein and carotid disease; increasing C-reactive emergency department physician activate the cardiac catheterization protein concentrations were associated with a 72% and 52% increased risk laboratory and transfer the patient immediately to the laboratory was of cardiovascular death and all-cause mortality, respectively, in the setting evaluated and compared with the door-to-balloon times achieved prior of carotid atheroslerosis. Similar to other studies, as assessed by the c to the institution of the protocol. This study by Khot et al showed that statistic, both C-reactive protein and carotid atherosclerosis added only door-to-balloon times decreased significantly from 113.5 minutes to modest incremental information to standard cardiovascular disease risk 75.5 minutes after adoption of the protocol. Treatment within 90 factors. The study underscores the need for further statistical and clinical minutes rose from 28% to 71%. As a result, mean infarct size tools to enhance clinical risk prediction. See p 32 (editorial p 3). decreased, as did hospital length of stay and total hospital costs per admission. The findings suggest that emergency department physician activation of the catheterization laboratory with immediate transfer to the laboratory is highly effective in reducing door-to-balloon times and also appears to improve outcomes and reduce cost. See p 67 (editorial METOPROLOL REVERSES LEFT VENTRICULAR p 6). REMODELING IN PATIENTS WITH ASYMPTOMATIC SYSTOLIC DYSFUNCTION: THE REVERSAL OF VENTRICULAR REMODELING WITH TOPROL-XL (REVERT) TRIAL, by Colucci et al. Until now, there has been no randomized, controlled trial data to Visit http://circ.ahajournals.org: support the benefit of ␤-blockers in patients with asymptomatic left Images in Cardiovascular Medicine ventricular systolic dysfunction. Colucci and colleagues investigate An Unusual Site for a Common Disease. See p e1. this question with the REversal of VEntricular Remodeling with Toprol-XL (REVERT) trial by randomly assigning patients, with a left ventricular ejection fraction Ͻ40%, mild left ventricular dilation, and no symptoms of heart failure (New York Heart Associa- tion class I), to 3 treatment groups: extended-release metoprolol succinate 200 mg or 50 mg and placebo. Echocardiographic assessment of left ventricular end-systolic volume, end-diastolic volume, mass, and ejection fraction were performed at baseline. After 12 months, in the 200-mg group, there was a decrease in Sine-Wave Pattern Arrhythmia and Sudden Paralysis That end-systolic volume index and an increase in left ventricular Result From Severe Hyperkalemia. See p e2. ejection fraction. In the 50-mg group, similar effects of a lesser Lipomatous Metaplasia in Ischemic Cardiomyopathy: A Com- magnitude were observed. These results demonstrate that the mon but Unappreciated Entity. See p e5. antiremodeling benefits of ␤-blocker therapy with metoprolol succinate extend to patients with asymptomatic left ventricular Correspondence dysfunction. See p 49. See p e7. Editors’ Note

Medical students often learn about the cardiovascular system as an isolated entity; in many cases a focused approach to anatomy, physiology, and pathophysiology is presented without consider- ation of other biological systems. While this may be a reasonable way to learn the fundamentals of cardiovascular science, it becomes quite clear to these students during their clinical training that the cardiovascular system functions in a remarkably complex milieu in concert with other organ systems. The development of perturbations in one system often leads to responses in other systems in an attempt to maintain functional homeostasis. Accordingly, the cardiovascular system is subject to the complex interplay among organ systems and a multitude of other factors, including those from the environment and the individual’s lifestyle. When problems with other organ systems develop, the initial clinical manifestations may be cardiovascular in nature (eg, abnormalities in heart rate, rhythm, and blood pressure). Understanding that today’s busy practitioner is regularly faced with patients who have many complex medical problems, the Editors of Circulation have commissioned this special series that focuses on the cardiovascular consequences of other medical disorders. Articles in this series, Cardiovascular Involvement in General Medical Conditions, will be published monthly over the next 7 months. Each article, which is written by highly respected experts in the field, will provide a comprehensive and insightful overview of the pathophysiology, clinical manifestations, and treatment options for a specific condition. Topics will cover thyroid diseases, rheumatological disorders, sepsis, pulmonary diseases, cancer and chemotherapy, and alcohol use and abuse. We anticipate that this series will provide a valuable resource for the clinician, who can readily bring this information to the bedside. We also hope that the gaps in the knowledge base that are highlighted in each article of this series will inspire the researcher to move the field forward. Gary J. Balady, MD Ravin Davidoff, MD Series Editors, Cardiovascular Involvement in General Medical Conditions, Circulation

Cardiovascular Biomarkers Added Value With an Integrated Approach?

Wolfgang Koenig, MD, FRCP, FESC

n primary prevention, traditional risk factors are a useful ment procedure is well standardized and automated, and high- first step in the determination of who could be at risk for sensitive assays with sufficient precision are available. On the Icardiovascular events. In the era of “global risk assessment” basis of substantial evidence of a contribution of inflammation to scores such as the Framingham score, the Prospective Cardio- atherothrombogenesis, a recent American Heart Association/ vascular Münster (PROCAM) score, or the European Society of Centers for Disease Control and Prevention consensus report has Cardiology Systematic Coronary Risk Evaluation (SCORE), recommended the measurement of CRP in asymptomatic sub- which are derived from multivariable statistical models, should jects at intermediate risk for future coronary events (10-year risk, be used.1 However, it has been noted that a considerable number 10% to 20%).9 However, there are other emerging biomarkers of at-risk patients cannot be identified on the basis of traditional like lipoprotein-associated phospholipase A2 (Lp-PLA2), an risk factors alone.2 This has prompted the search for novel enzyme that is produced by monocytes/macrophages, T-cells, markers of cardiovascular risk to help improve risk prediction.3 and mast cells and has been found to generate proinflammatory 10 Such markers could either represent various blood biomarkers and proatherogenic molecules. Because Lp-PLA2, in contrast relevant to the pathophysiology of atherothrombosis (eg, mark- to CRP, does not correlate with most other risk factors, there is 11,12 ers of the inflammatory response, coagulation markers, markers an additive effect of CRP and Lp-PLA2 in risk prediction. of platelet aggregation, lipoproteins, or lipid-related variables), This may also apply to combinations of other biomarkers, genetic markers, or markers of subclinical disease, which may though evidence so far is limited. In the future, we might see a also aid in improved risk prediction. Determination of global risk biomarker profile that covers various aspects of the complex on the basis of traditional risk factors allows categorization into pathophysiology of the atherothrombotic process, and poten- high (10-year risk, Ͼ20%), low (10-year risk, Ͻ10%), or tially, we would be able to focus on biological patterns or intermediate risk (10-year risk, 10% to 20%). Subjects at high systems rather than on single biomarkers. To date, however, risk should be recommended lifestyle changes or prescribed a there is no sound evidence to suggest such a procedure for statin. Subjects at low risk would be reevaluated 3 to 5 years clinical practice, and there is even an ongoing discussion of later. Those at intermediate risk, however, who comprise up to whether any of the emerging blood biomarkers alone contributes 40% of the population at risk,4 would be candidates for addi- incremental information over and above the information gained tional testing to increase or decrease their actual risk. A large from available “global risk” scores.13,14 panel of blood biomarkers are available for this purpose, but most of them are not yet applicable in clinical practice for Markers of Subclinical Atherosclerotic Disease various reasons5,6 There is mounting evidence that markers of subclinical Article p 32 disease (eg, intima-media thickness as assessed by high- resolution carotid ultrasound;15 coronary calcium determined with multislice computed tomography;16 or ankle-brachial Emerging Blood Biomarkers index, a strong marker of atherosclerotic burden17) may also Atherosclerosis is characterized by a nonspecific local inflam- contribute to improved risk prediction. However, the clinical matory process7 that is accompanied by a systemic response. utility of multislice computed tomography needs to be further Thus a number of prospective studies in initially healthy subjects tested, and measurement of carotid intima-media thickness have convincingly demonstrated an independent association may be burdened by considerable interobserver variability between even slightly elevated concentrations of various sys- when it is used in routine clinical practice. Thus, similar to temic markers of inflammation and important cardiovascular blood biomarkers, the potential incremental value of such end points. At this time, the largest database exists for C-reactive surrogate markers of clinical atherosclerotic complications is protein (CRP), the classic acute-phase protein.8 The measure- not unequivocally evident. Still, from a theoretical viewpoint the combination of blood biomarkers and markers of subclin- The opinions expressed in this article are not necessarily those of the ical disease seems an attractive approach because this may editors or of the American Heart Association. integrate information on structural or functional vascular wall From the Department of Internal Medicine II, Cardiology, University of Ulm Medical Center, Ulm, Germany. pathology and systemic “activity” of the disease (Figure). Correspondence to Wolfgang Koenig, MD, Department of Internal However, for markers of subclinical disease as well as for Medicine II, Cardiology, University of Ulm Medical Center, Robert-Koch blood biomarkers, controversy exists with regard to which Str 8, D-89081 Ulm, Germany. E-mail [email protected] (Circulation. 2007;116:3-5.) parameter represents the most useful one and for which time © 2007 American Heart Association, Inc. period of the atherosclerotic process, and which combination of Circulation is available at http://www.circulationaha.org markers may be most appropriate for decision making. Finally, DOI: 10.1161/CIRCULATIONAHA.107.707984 analytical and cost considerations deserve further study. 3

4 Circulation July 3, 2007

Screening for subjects at risk for cardiovascular complications: blood biomarkers/risk factors and/or markers of subclinical disease. Apo indicates apolipoprotein; BP, blood pressure; CT, computed tomography; HDL, high-density lipoprotein; IMT, intima-media thickness; LDL, low- density lipoprotein; Lp(a), lipoprotein a; Lp-PLA2, lipoprotein-associated phospholipase A2; MRI, magnetic resonance imaging; and Syn, syndrome. Reprinted from Naghavi M et al. Am J Cardiol. 2006;98(suppl):2H–15H, with permission from Elsevier. Copyright 2006.

Statistical Methodology: Limitations has recently been discussed in detail by Cook,18 and alternative in Assessment of Incremental statistical approaches have been suggested, such as clinical risk Diagnostic Information reclassification.19 This procedure attempts to improve risk pre- A great deal of such uncertainty is based on the limited diction by development and validation of algorithms that more availability of adequate statistical tools to demonstrate the precisely allocate an individual to a risk category by use of a incremental value of an emerging biomarker in addition to model that has incorporated a new risk variable in addition to global risk scoring. We have realized that evidence of just some conventional risk factors, compared with a basic model that moderately strong association in epidemiological studies is contains conventional risk factors alone. Such approach focuses insufficient to assess the true clinical utility of a new candidate particularly on those subjects at intermediate risk to either marker. Most frequently, c statistics and area under the receiver- reclassify an individual into the low- or high-risk category. operating characteristic curve have been used. Risk estimates that would be needed here to show a clinically important Integration of Biochemical and Bioimaging increase in the area under the curve are usually not seen in Markers: The Solution? cardiovascular medicine.18 Thus, disappointingly, only a few In the presence of such complex background, Cao and studies have shown a statistically significant improvement in the colleagues20 present important data from the Cardiovascular area under the curve, which, however, in most cases was too Health Study in this issue of Circulation. The investigators small to be considered clinically relevant. The aggregate expe- simultaneously measured carotid intima-media thickness, rience from a number of such studies demonstrates that once plaque characteristics, and CRP, and related all 3 variables to there is a single strong predictor of risk in the model, which may the 12-year incidence of cardiovascular disease (CVD) events be even age alone, it is extremely difficult to show a relevant and all-cause mortality in 5888 elderly subjects. Main results contribution of any additional variable to model prediction. This showed that all parameters were correlated with one another, Koenig Cardiovascular Biomarkers 5 yet each parameter independently predicted risk of CVD 3. Morrow DA, ed. Cardiovascular Biomarkers. Pathophysiology and events and mortality in multivariable models, which included Disease Management. Totowa, New Jersey: Humana Press Inc.; 2006. 4. Greenland P, Smith SC Jr, Grundy SM. Improving coronary heart disease all 3 measures and traditional risk factors. Being in the top risk assessment in asymptomatic people: role of traditional risk factors tertile of the carotid intima-media thickness distribution was and noninvasive cardiovascular tests. Circulation. 2001;104:1863–1867. more predictive for various events than having CRP Ͼ3 mg/L 5. Vasan RS. Biomarkers of cardiovascular disease: molecular basis and or than being in the high-risk group on the basis of carotid practical considerations. Circulation. 2006;113:2335–2362. 6. Koenig W, Khuseyinova N. Biomarkers of atherosclerotic plaque insta- plaque characteristics. Elevated CRP was a particularly useful bility and rupture. Arterioscler Thromb Vasc Biol. 2007;27:15–26. predictor in the presence of subclinical atherosclerosis with a 7. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. 72% increase in risk for CVD and 52% increase in total N Engl J Med. 2005;352:1685–1695. mortality. Cumulative event rates suggested a possible addi- 8. Ridker PM, Rifai N, eds. C-Reactive Protein and Cardiovascular Disease. St-Laurent, Canada: MediEdition Inc.; 2006. tive interaction for composite CVD and all-cause mortality 9. Pearson TA, Mensah GA, Alexander RW, Anderson JL, Cannon RO 3rd, with an excess risk attributable to the interaction of CRP and Criqui M, Fadl YY, Fortmann SP, Hong Y, Myers GL, Rifai N, Smith SC Jr, subclinical atherosclerosis of 54% for CVD death and 79% Taubert K, Tracy RP, Vinicor F; Centers for Disease Control and Prevention; for all-cause mortality. By contrast, CRP did not add predic- American Heart Association. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for tive power in the absence of carotid atherosclerosis. Finally, healthcare professionals from the Centers for Disease Control and Prevention both CRP and subclinical atherosclerosis added only modest and the American Heart Association. Circulation. 2003;107:499–511. incremental information to risk prediction when adjusted for 10. Zalewski A, Macphee C. Role of lipoprotein-associated phospholipase A2 the effect of conventional risk factors with either c statistics in atherosclerosis: biology, epidemiology, and possible therapeutic target. Arterioscler Thromb Vasc Biol. 2005;25:923–931. or area under the curve derived from receiver-operating 11. Koenig W, Khuseyinova N, Lowel H, Trischler G, Meisinger C. Lipoprotein- characteristic analysis. associated phospholipase A2 adds to risk prediction of incident coronary events by C-reactive protein in apparently healthy middle-aged men from the Conclusions general population: results from the 14-year follow-up of a large cohort from First, global risk assessment, with traditional risk factors, still southern Germany. Circulation. 2004;110:1903–1908. represents the rational basis for cardiovascular risk stratification. 12. Ballantyne CM, Hoogeveen RC, Bang H, Coresh J, Folsom AR, Second, although theoretically attractive, currently available Chambless LE, Myerson M, Wu KK, Sharrett AR, Boerwinkle E. Lipoprotein-associated phospholipase A2, high-sensitivity C-reactive biomarkers, even the combination of a robust systemic marker of protein, and risk for incident ischemic stroke in middle-aged men and “disease activity” with a marker that provides information on women in the Atherosclerosis Risk in Communities (ARIC) study. Arch structural changes of the arterial vasculature, which must be seen Intern Med. 2005;165:2479–2484. as a surrogate/precursor of clinical disease, does not appreciably 13. Folsom AR, Chambless LE, Ballantyne CM, Coresh J, Heiss G, Wu KK, Boerwinkle E, Mosley TH Jr, Sorlie P, Diao G, Sharrett AR. An improve risk prediction. However, the Cardiovascular Health assessment of incremental coronary risk prediction using C-reactive Study cohort was an elderly population and results may not be protein and other novel risk markers: the Atherosclerosis Risk in Com- generalizable to younger individuals with low risk, in whom munities study. Arch Intern Med. 2006;166:1368–1373. CRP may work in the absence of significant atherosclerotic 14. Wang TJ, Gona P, Larson MG, Tofler GH, Levy D, Newton-Chen C, Jacques PF, Rifai N, Selhub J, Robins SJ, Benjamin EJ, D’Agostino RB, burden. Also, the statistical tools used, as mentioned earlier, may Vasan RS. Multiple biomarkers for the prediction of first cardiovascular be debatable. Third, in the future, despite such somewhat events and death. N Engl J Med. 2006;355:2631–2639. disappointing information regarding single markers, the clinical 15. Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M. Prediction of application of multimarker panels, for which the possibilities of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis. Circulation. 2007;115:459–467. model improvement are greater, may still prove to be a prom- 16. Budoff MJ, Achenbach S, Blumenthal RS, Carr JJ, Goldin JG, Greenland ising approach, provided that such variables show low correla- P, Guerci AD, Lima JA, Rader DJ, Rubin GD, Shaw LJ, Wiegers SE; tions with conventional risk factors and with each other but American Heart Association Committee on Cardiovascular Imaging and provide strong associations with clinical events. Such emerging Intervention; American Heart Association Council on Cardiovascular Radiology and Intervention; American Heart Association Committee on markers will have to be rigorously evaluated in large cohorts for Cardiac Imaging, Council on Clinical Cardiology. Assessment of their clinical efficacy and effectiveness with innovative statisti- coronary artery disease by cardiac computed tomography: a scientific cal analytical tools. The world of proteomics and metabolomics, statement from the American Heart Association Committee on Cardio- together with advanced imaging modalities such as functional vascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on molecular imaging, may offer such promising candidates. Clinical Cardiology. Circulation. 2006;114:1761–1791. 17. Heald CL, Fowkes FG, Murray GD, Price JF; Ankle Brachial Index Collab- Disclosures oration. Risk of mortality and cardiovascular disease associated with the None. ankle-brachial index: systematic review. Atherosclerosis. 2006;189:61–69. References 18. Cook NR. Use and misuse of the receiver-operating characteristic curve in risk prediction. Circulation. 2007;115:928–935. 1. De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R, 19. Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, Cats VM, Orth-Gomer K, Perk J, Pyorala K, Rodicio JL, Sans S, Sansoy V, Sechtem improved algorithms for the assessment of global cardiovascular risk in U, Silber S, Thomsen T, Wood D; European Society of Cardiology; women: the Reynolds Risk Score. JAMA. 2007;297:611–619. American Heart Association; American College of Cardiology. European 20. Cao JJ, Arnold AM, Manolio TA, Polak JF, Psaty BM, Hirsch CH, Kuller guidelines on cardiovascular disease prevention in clinical practice. Third LH, Cushman M. Association of carotid artery intima-media thickness, Joint Task Force of European and other Societies on Cardiovascular Disease plaques, and C-reactive protein with future cardiovascular disease and Prevention in Clinical Practice (constituted by representatives of eight all-cause mortality: the Cardiovascular Heath Study. Circulation. 2007; societies and by invited experts). Atherosclerosis. 2004;173:381–391. 116:32–38. 2. Khot UN, Khot MB, Bajzer CT, Sapp SK, Ohman EM, Brener SJ, Ellis SG, Lincoff AM, Topol EJ. Prevalence of conventional risk factors in KEY WORDS: Editorials Ⅲ atherosclerosis Ⅲ epidemiology Ⅲ imaging Ⅲ patients with coronary heart disease. JAMA. 2003;290:898–904. inflammation Ⅲ risk factors

Editorial

The ST-Segment–Elevation Myocardial Infarction Chain of Survival

Joseph P. Ornato, MD

he benefit of expertly performed, timely, primary the ACC’s Guidelines Applied in Practice Door-to-Balloon percutaneous coronary intervention (PCI) over fibri- (GAP-D2B) campaign goal of a door-to-balloon time interval Tnolysis is clear for patients with ST-segment–eleva- of Յ90 minutes in 75% of PCI-treated STEMI patients at tion myocardial infarction (STEMI). Primary PCI is superior participating hospitals nationwide.6 to fibrinolysis for reduction of overall short-term mortality, On the ACC President’s Page, Nissen et al6 described the nonfatal reinfarction, stroke, and the combined end point of new GAP-D2B campaign and stated: death, nonfatal reinfarction, and stroke.1 These results remain In successful hospitals, the arrival of a STEMI valid during long-term follow-up and are independent of both patient initiates a chain of well-orchestrated events, the type of fibrinolytic used and whether the patient is including activation of the catheterization laboratory transferred for primary PCI. directly by an emergency department physician with a single phone call to the interventional cardiologist on Article p 67 call. The catheterization laboratory team is expected to arrive within 20 to 30 minutes. Programs with the best Although the relationship between time delay from hospi- outcomes employ a multidisciplinary team-based ap- tal emergency department arrival to fibrinolytic treatment and proach that includes committed administrators, physi- increasing mortality has been firmly established,2 a similar cian champions, and nurse champions, along with relationship for primary PCI treatment has been proven only mechanisms for rapid data feedback. This collabora- recently. De Luca et al3 assessed the relationship between tion can extend to the local and regional emergency medical systems (EMS). In some successful hospitals, ischemic time and 1-year mortality in 1791 primary PCI- the catheterization laboratory is activated based on a treated STEMI patients. After adjustment for age, gender, prehospital electrocardiography. diabetes, and previous revascularization, these investigators In this issue of Circulation, Khot et al7 report on their showed that every 30 minutes of primary PCI treatment delay experience before and after implementation of STEMI GAP- is associated with a 7.5% (95% CI, 1.008 to 1.15; Pϭ0.041) D2B–like strategies in a 591-bed, tertiary care, Indianapolis- relative increase in 1-year mortality. With use of hierarchical area, community hospital that consists of 2 separate campuses models adjusted for patient characteristics to evaluate the 7 miles apart (a 13-minute drive). Although they began their effect of door-to-balloon time on in-hospital mortality on program long before the recently announced ACC initiative, 29 222 PCI-treated STEMI patients treated in Յ6 hours of Khot et al7 instituted most of the GAP-D2B recommendations presentation at 395 hospitals that participated in the National on the basis of characteristics of best-performing National Registry of Myocardial Infarction–3 and –4 from 1999 to Registry of Myocardial Infarction hospitals.8–10 Critical ele- 2002, McNamara et al4 found that a longer door-to-balloon ments of their new system included empowerment of emer- time interval is associated with increased in-hospital mortal- gency physicians to activate the catheterization laboratory ity. Adjusted for patient characteristics, patients with a and team without cardiology consultation as well as imple- door-to-balloon time interval Ͼ90 minutes were more likely mentation of an in-house transfer team. Their new strategy to die (odds ratio, 1.42; 95% CI, 1.24 to 1.62) compared with reduced the median door-to-balloon time interval signifi- patients who had a door-to-balloon time interval Յ90 min- cantly during normal and off-duty work hours, even for utes. These findings provide evidence-based support for the patients who had to be transferred physically from 1 facility goal of a door-to-balloon time interval “within 90 minutes” to another, which thus increased the percentage of patients cited in the 2004 American College of Cardiology/American treated within the Յ90-minute door-to-balloon goal from Heart Association (ACC/AHA) guidelines for the manage- 28% to 71%. And, as predicted by the relationship between ment of patients with STEMI5 and serve as a foundation for time to treatment and outcome, there were significant improvements in mean infarct size, hospital length of stay, and The opinions expressed in this article are not necessarily those of the total hospital cost per admission. editors or of the American Heart Association. The ACC’s GAP-D2B initiative stands on even more solid From the Department of Emergency Medicine, Virginia Common- ground as a result of the Indiana Heart Physicians/St. Francis wealth University, Richmond. Correspondence to Joseph P. Ornato, MD, Virginia Commonwealth Heart Center experience reported by Khot et al.7 The common University, Department of Emergency Medicine, 1201 East Marshall St, element shared by both is choreographed multidisciplinary AD Williams 2nd Floor, Richmond, VA 23298᎑0401. E-mail teamwork with effective communication among different [email protected] (Circulation. 2007;116:6-9.) disciplines of healthcare providers, rather than the traditional © 2007 American Heart Association, Inc. linear progression of care most patients experience as they Circulation is available at http://www.circulationaha.org pass through a series of hospital units that operate as DOI: 10.1161/CIRCULATIONAHA.107.710970 individual silos. Both initiatives are focused on the portion of

6 Ornato STEMI Chain of Survival 7

STEMI patient treatment delay that is potentially most guidelines.15 Khot et al7 have shown us that this challenging changeable by hospital-based healthcare providers—that time interval also can be decreased dramatically by an which occurs after a patient presents to the hospital. This is organized transfer and PCI treatment team that can be clearly the right place to start, but it represents only part of a activated by emergency physicians. broader community-based opportunity to improve STEMI Our national challenge to provide optimal STEMI care patient care. needs to be solved at 2 levels: the community and the In 1991, the AHA adopted a metaphor—the “Chain of hospital. We must continue to educate the public on the signs Survival”—to describe the sequence of events that must and symptoms of a myocardial infarction and reinforce the occur for the best likelihood of successful resuscitation from National Heart Attack Alert Program and AHA message to hospital cardiopulmonary arrest.11 This educational construct, “Call 911, Call Fast.”16 The community needs to be organized which consists of early access, early cardiopulmonary resus- into a system of care that directs STEMI patients quickly and citation, early defibrillation, and early advanced cardiac life efficiently to primary PCI centers whenever possible, and all support, now serves as the structural foundation for improve- hospitals, whether primary PCI-capable or not, need to have ments in the community approach to sudden cardiac death a system in place to avoid unnecessary delays, just like that worldwide. It may now be appropriate for the AHA to which has been implemented by Khot et al.7 consider adoption of a similar metaphor—the “STEMI Chain The AHA’s Acute Myocardial Infarction Advisory Work- of Survival” (Figure)—that can be used to target global ing Group recently released recommendations on how to improvements in STEMI patient care. This approach is very increase the number of STEMI patients who have timely similar to that which has been followed for Ͼ25 years by the access to primary PCI.17 The group commissioned Pricewa- American College of Surgeons Committee on Trauma, as it terhouseCoopers to conduct national market research, and the has led our nation to develop one of the finest and most Working Group interviewed a wide variety of key stakehold- effective trauma care systems in the world. The new STEMI ers (such as patients, physicians, nurses, EMS representa- chain begins with an emphasis on the role of the patient in the tives, community hospitals, primary PCI facilities, payers, recognition of early or prodromal heart attack symptoms and and evaluation/outcomes organizations such as the Agency immediate request for help, preferably by calling 911 and for Healthcare Research and Quality, the Food and Drug accessing the EMS system12,13; the chain works its way Administration, and the Joint Commission on Accreditation through the critical elements of prehospital, emergency de- of Healthcare Organizations) to determine the desirability, partment, and reperfusion care. feasibility, and potential effectiveness of establishment of There is presently no uniform national STEMI triage and regional systems and/or centers of care for STEMI patients treatment system equivalent to the system that directs major with a focus on whether and how this might improve patient trauma victims to verified trauma centers in the United States. access to quality care and outcomes. The researchers found Because the majority of US hospitals do not have primary that key stakeholders would support a national primary PCI PCI capability, many communities struggle to decide whether certification program with the understanding that some non- they should direct EMS-transported, prehospital, 12-lead primary PCI hospitals would experience a modest decline in ECG–identified STEMI patients to only primary PCI facili- revenue. On the basis of these findings, the AHA hosted a ties to bypass nonprimary PCI hospitals. Unfortunately, the national stakeholder meeting in 2006 to continue develop- majority of STEMI patients do not use the 911 ambulance ment of a more detailed plan for a national system of STEMI system for transport to the hospital, the national paramedic patient care. As has been suggested, this is an idea whose time training curriculum considers 12-lead ECG training as an may have come.18 enhanced rather than core skill, and not all EMS ambulances Many communities are currently developing organized currently have 12-lead ECG capability.14 Many US hospitals STEMI care plans. Three sites—a major city, a large region continue to use fibrinolysis as their primary reperfusion of a state, and an entire state—already have model commu- strategy with transfer to an interventional facility for rescue nity STEMI systems in place based on the trauma care system when needed. Other hospitals transfer patients more regularly model. In 2003, Boston, Mass., implemented a comprehen- for primary PCI, but even the most recently published sive system of care in which STEMI patients identified by National Registry of Myocardial Infarction data on 4278 paramedics with the use of prehospital 12-lead ECGs were patients transferred to 419 hospitals for primary PCI show a transported only to designated PCI centers.19 Participating median initial hospital door-to-balloon time of 180 minutes, PCI centers agreed to collect and submit performance mea- with only 4.2% of patients treated with reperfusion in Ͻ90 sures data to a Central Data Coordinating Center on all EMS- minutes, the benchmark recommended by national quality and non-EMS–transported STEMI patients. System oversight

Figure 1. The STEMI chain of survival. 8 Circulation July 3, 2007 was provided by a Steering Committee with representatives only on 1 hospital. A third model is the Reperfusion of Acute from 9 area hospitals and the Boston EMS, which developed Myocardial Infarction in Carolina Emergency Departments their performance indicators and minimum standards on the (RACE) project, which is a collaborative effort to increase the basis of nationally accepted guidelines. A central Data rate and speed of coronary reperfusion through systematic Coordinating Center at Tufts–New England Medical Center changes in emergency care. The project is based on the received and tabulated data from EMS and area hospitals to collaborative efforts of EMS personnel, physicians, nurses, provide aggregated data reports with receiving hospitals administrators, and payers from 5 regions and 68 hospitals designated only A, B, C, D, etc, rather than by name. The throughout North Carolina. The recommendations of this reports were reviewed by an independent Data and Safety project are based on established national guidelines, pub- Monitoring Board composed of highly respected cardiolo- lished data, and the knowledge and experience of numerous gists and a statistician. After discussion between the hospital individuals who specialize in STEMI patient care. Detailed and Data and Safety Monitoring Board and review by the information about the program, such as transfer criteria, Steering Committee, any hospital that did not meet preestab- protocols, training materials, and educational posters, are lished quality treatment, door-to-balloon, and outcome goals available on the North Carolina ACC Chapter Web site for 2 successive 6-month periods could be excluded from (http://www.nccacc.org/race.html). receiving EMS-transported STEMI patients for the next In summary, cardiologists (and interventionalists), emer- 6-month period. gency physicians, nurses, and EMS providers must work From March 2003 to May 2005, 448 STEMI patients were together to establish effective regional community systems of transferred from 31 community hospitals by paramedic- STEMI patient care similar to the well-developed and highly staffed ambulance (nϭ149) or paramedic/critical care nurse– successful systems that direct major trauma victims to veri- staffed helicopter (nϭ299) to the Minneapolis Heart Institute fied trauma centers in the United States. All hospitals, in Minneapolis, Minn., for primary PCI. A standardized whether primary PCI–capable or not, should develop a protocol with accompanying checklists was developed on the STEMI protocol that includes procedures to expedite time to basis of national guidelines. Community hospitals were reperfusion treatment modeled after concepts inherent in the required to transfer all patients with STEMI or new left GAP-D2B program and the Indiana Heart Physicians/St. bundle–branch block within 12 hours of symptom onset to Francis Heart Center experience. A multidisciplinary com- the regional interventional center. A level 1 myocardial mittee should oversee the process and provide performance infarction protocol was developed and used to specify the improvement suggestions based on continuous data collection sequence of events, diagnostic tests, and treatments. Patients and analysis. are preregistered by admitting personnel prior to arrival by use of a demographic form faxed from the referring hospital. Disclosures On arrival at the primary PCI center, patients are admitted Dr Ornato has served on the science advisory board for the National directly to the cardiac catheterization laboratory and thus Registry of Myocardial Infarction, which is funded by Genentech. bypass the emergency department except in rare circumstances, such as when 2 STEMI patients arrive simulta- References 1. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous neously. Prompt verbal and written feedback (which includes thrombolytic therapy for acute myocardial infarction: a quantitative 1-month and 1-year follow-up phone calls) is provided to the review of 23 randomised trials. Lancet. 2003;361:13–20. referring hospital physician and nursing staff, and the time 2. Cannon CP, Gibson CM, Lambrew CT, Shoultz DA, Levy D, French WJ, intervals, clinical and angiographic data, and clinical out- Gore JM, Weaver WD, Rogers WJ, Tiefenbrunn AJ. Relationship of symptom-onset-to-balloon time and door-to-balloon time with mortality comes are entered into a database. Time and outcome in patients undergoing angioplasty for acute myocardial infarction. summary reports meeting Joint Commission on Accreditation JAMA. 2000;283:2941–2947. of Healthcare Organizations requirements are sent to each 3. De Luca G, Suryapranata H, Ottervanger JP, Antman EM. Time delay to treatment and mortality in primary angioplasty for acute myocardial community hospital on a quarterly basis. infarction: every minute of delay counts. Circulation. 2004;109: Patient treatment times and outcomes have been superb 1223–1225. with this regional STEMI care system. No STEMI patients 4. McNamara RL, Wang Y, Herrin J, Curtis JP, Bradley EH, Magid DJ, were excluded from transfer, even those with cardiogenic Peterson ED, Blaney M, Frederick PD, Krumholz HM. Effect of door- to-balloon time on mortality in patients with ST-segment elevation myo- shock (13.7%), cardiac arrest (9.9%), and the elderly (17%, cardial infarction. J Am Coll Cardiol. 2006;47:2180–2186. Ͼ80 years of age). No patient died during transport. The 5. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, median total door-to-balloon time was reduced from Ͼ3 Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, hours before implementation of the regional system to 97 Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL, Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka minutes for 11 participating hospitals located Յ70 miles LF, Hunt SA, Jacobs AK. ACC/AHA guidelines for the management of (zone 1) after implementation.18,20 The median total door-to- patients with ST-elevation myocardial infarction: a report of the balloon time was 117 minutes with use of a facilitated PCI American College of Cardiology/American Heart Association Task Force Յ on Practice Guidelines (Committee to Revise the 1999 Guidelines for the protocol in 17 participating hospitals located 210 miles Management of Patients with Acute Myocardial Infarction). Circulation. (zone 2) from the interventional center. The improvements in 2004;110:e82–e292. time to treatment were accompanied by low 30-day mortality 6. Nissen SE, Brush JE Jr, Krumholz HM. President’s page: GAP-D2B: an rates of 4.3% in zone 1 and 3.4% in zone 2. alliance for quality. J Am Coll Cardiol. 2006;48:1911–1912. 7. Khot UN, Johnson ML, Ramsey C, Khot MB, Todd R, Shaikh SR, Berg The common denominator of these 2 models is that they WJ. Emergency department physician activation of the catheterization are based on a community system of care rather than centered laboratory and immediate transfer to an immediately available catheter-

Ornato STEMI Chain of Survival 9

ization laboratory reduce door-to-balloon time in ST-elevation myo- 15. Nallamothu BK, Bates ER, Herrin J, Wang Y, Bradley EH, Krumholz cardial infarction. Circulation. 2007;116:67–76. HM. Times to treatment in transfer patients undergoing primary percu- 8. Bradley EH, Curry LA, Webster TR, Mattera JA, Roumanis SA, Radford taneous coronary intervention in the United States: National Registry of MJ, McNamara RL, Barton BA, Berg DN, Krumholz HM. Achieving Myocardial Infarction (NRMI)-3/4 analysis. Circulation. 2005;111: rapid door-to-balloon times: how top hospitals improve complex clinical 761–767. systems. Circulation. 2006;113:1079–1085. 16. Faxon D, Lenfant C. Timing is everything: motivating patients to call 9. Bradley EH, Herrin J, Wang Y, McNamara RL, Radford MJ, Magid DJ, 9-1-1 at onset of acute myocardial infarction. Circulation. 2001;104: Canto JG, Blaney M, Krumholz HM. Door-to-drug and door-to-balloon 1210–1211. times: where can we improve? Time to reperfusion therapy in patients 17. Jacobs AK, Antman EM, Ellrodt G, Faxon DP, Gregory T, Mensah GA, with ST-segment elevation myocardial infarction (STEMI). Am Heart J. Moyer P, Ornato J, Peterson ED, Sadwin L, Smith SC. Recommendation 2006;151:1281–1287. to develop strategies to increase the number of ST-segment-elevation myocardial infarction patients with timely access to primary percutaneous 10. Bradley EH, Roumanis SA, Radford MJ, Webster TR, McNamara RL, coronary intervention. Circulation. 2006;113:2152–2163. Mattera JA, Barton BA, Berg DN, Portnay EL, Moscovitz H, Parko- 18. Henry TD, Atkins JM, Cunningham MS, Francis GS, Groh WJ, Hong sewich J, Holmboe ES, Blaney M, Krumholz HM. Achieving door-to- RA, Kern KB, Larson DM, Ohman EM, Ornato JP, Peberdy MA, balloon times that meet quality guidelines: how do successful hospitals do Rosenberg MJ, Weaver WD. ST-segment elevation myocardial it? J Am Coll Cardiol. 2005;46:1236–1241. infarction: recommendations on triage of patients to heart attack centers: 11. Cummins RO, Ornato JP, Thies WH, Pepe PE. Improving survival from is it time for a national policy for the treatment of ST-segment elevation sudden cardiac arrest: the “chain of survival” concept. A statement for myocardial infarction? J Am Coll Cardiol. 2006;47:1339–1345. health professionals from the Advanced Cardiac Life Support Subcom- 19. Moyer P, Feldman J, Levine J, Beshansky J, Selker HP, Barnewolt B, mittee and the Emergency Cardiac Care Committee, American Heart Brown D, Cardoza J, Grossman S, Jacobs A, Kerman B, Kimmelstiel C, Association. Circulation. 1991;83:1832–1847. Larson R, Losordo D, Pearlmutter M, Pozner C, Ramirez A, Rosenfield 12. Bahr RD. Access to early cardiac care: chest pain as a risk factor for heart K, Ryan TJ, Zane RD, Cannon CP. Implications of the mechanical (PCI) attacks, and the emergence of early cardiac care centers. Md Med J. vs thrombolytic controversy for ST segment elevation myocardial 1992;41:133–137. infarction on the organization of emergency medical services: the Boston 13. Ornato JP, Hand MM. Warning signs of a heart attack. Circulation. EMS experience. Crit Pathways Cardiol. 2004;3:53–61. 2001;104:1212–1213. 20. Henry TD, Sharkey SW, Graham KJ, Pedersen WR, Lips DL, Wang YL, 14. Garvey JL, MacLeod BA, Sopko G, Hand MM. Pre-hospital 12-lead Unger BT, Henry CR, Larson DM. Transfer for direct percutaneous electrocardiography programs: a call for implementation by emergency coronary intervention for ST-elevation myocardial infarction: the Minne- medical services systems providing advanced life support: National Heart apolis Heart Institute level 1 myocardial infarction program. Circulation. Attack Alert Program (NHAAP) Coordinating Committee; National 2005;112:II᎑620. Heart, Lung, and Blood Institute (NHLBI); National Institutes of Health. J Am Coll Cardiol. 2006;47:485–491. KEY WORDS: Editorials Ⅲ infarction Ⅲ myocardium Common NOS1AP Variants Are Associated With a Prolonged QTc Interval in the Rotterdam Study

Albert-Jan L.H.J. Aarnoudse, MD*; Christopher Newton-Cheh, MD, MPH*; Paul I.W. de Bakker, PhD; Sabine M.J.M. Straus, MD, PhD; Jan A. Kors, PhD; Albert Hofman, MD, PhD; André G. Uitterlinden, PhD; Jacqueline C.M. Witteman, PhD; Bruno H.C. Stricker, PhD

Background—QT prolongation is an important risk factor for sudden cardiac death. About 35% of QT-interval variation is heritable. In a recent genome-wide association study, a common variant (rs10494366) in the nitric oxide synthase 1 adaptor protein (NOS1AP) gene was found to be associated with QT-interval variation. We tested for association of 2 NOS1AP variants with QT duration and sudden cardiac death. Methods and Results—The Rotterdam Study is a population-based, prospective cohort study of individuals Ն55 years of age. The NOS1AP variants rs10494366 TϾG and rs10918594 CϾG were genotyped in 6571 individuals. Heart rate–corrected QT interval (QTc) was determined with ECG analysis software on up to 3 digital ECGs per individual (total, 11 108 ECGs from 5374 individuals). The association with QTc duration was estimated with repeated-measures analyses, and the association with sudden cardiac death was estimated by Cox proportional-hazards analyses. The rs10494366 G allele (36% frequency) was associated with a 3.8-ms (95% confidence interval, 3.0 to 4.6; Pϭ7.8ϫ10Ϫ20) increase in QTc interval duration for each additional allele copy, and the rs10918594 G allele (31% frequency) was associated with a 3.6-ms (95% confidence interval, 2.7 to 4.4; Pϭ6.9ϫ10Ϫ17) increase per additional allele copy. None of the inferred NOS1AP haplotypes showed a stronger effect than the individual single-nucleotide polymorphisms. There were 233 sudden cardiac deaths over 11.9 median years of follow-up. No significant association was observed with sudden cardiac death risk. Conclusions—Common variants in NOS1AP are strongly associated with QT-interval duration in an elderly population. Larger sample sizes are needed to confirm or exclude an effect on sudden cardiac death risk. (Circulation. 2007;116: 10-16.) Key Words: arrhythmia Ⅲ death, sudden Ⅲ electrocardiography Ⅲ genetics Ⅲ long-QT syndrome

udden cardiac death (SCD) claims 300 000 lives annually great accuracy and genome-wide coverage of common vari- Sin the United States.1 Although certain high-risk groups ation,12 together with analytical methods,13 has enabled un- have been identified,2 most SCD occurs in individuals unrec- biased surveys of most of the common variation in the human ognized to be at risk.3 genome. Still, the relatively small size of existing SCD collections and etiologic heterogeneity limit the statistical Clinical Perspective p 16 power to detect causal variants; therefore, initial attention has Familial aggregation of SCD suggests a substantial contri- focused on quantitative SCD risk factors in large cohorts. bution of genetic variation to SCD risk,4–7 but mendelian The electrocardiographic QT interval is a noninvasive mutations identified to date individually explain little of the measure of ventricular repolarization. About 35% of the population burden of SCD.8,9 Until recently, the search for variation in QT-interval duration in unselected community- sequence variants contributing to SCD risk has been re- based samples is heritable.14,15 Mendelian congenital long- stricted to candidate genes known for their role in arrhyth- and short-QT syndromes are both characterized by SCD from mogenesis.10 The recent development of large single-nucle- ventricular arrhythmias. Moreover, nonsyndromal long QT otide polymorphism (SNP) databases,11 genotyping arrays of interval16–19 and short QT interval20 impart increased risk of

Received November 16, 2006; accepted May 1, 2007. From the Departments of Epidemiology and Biostatistics (A.L.H.J.A., S.M.J.M.S, A.H., A.G.U., J.C.M.W., B.H.C.S.), Internal Medicine (A.G.U., B.H.C.S.), and Medical Informatics (J.A.K.), Erasmus Medical Center, Rotterdam, the Netherlands; Cardiology Division (C.N.-C.), Department of Molecular Biology (P.I.W.d.B.), and Center for Human Genetics Research (P.I.W.d.B.), Massachusetts General Hospital, Boston; Program in Medical and Population Genetics (C.N.-C., P.I.W.d.B.), Broad Institute of Harvard and MIT, Cambridge, Mass; National Heart, Lung, and Blood Institute’s Framingham Heart Study (C.N.-C.), Framingham, Mass; Department of Genetics, Harvard Medical School (P.I.W.d.B.), Boston, Mass; Inspectorate for Health Care (A.L.H.J.A., B.H.C.S.), the Hague, the Netherlands; and Dutch Medicines Evaluation Board (S.M.J.M.S), the Hague, the Netherlands. *The first 2 authors contributed equally to this work. Correspondence to Bruno H.C. Stricker, PhD, Department of Epidemiology and Biostatistics, Erasmus Medical Center, PO Box 2040, 3000 CA, Rotterdam, The Netherlands. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.676783

D Aarnoudse et al NOS1AP Variants Are Associated With QTc Duration 11

160,200kb 160,300kb 160,400kb 160,500kb 160,600kb Chr. 1q23

OLFML2B NOS1AP

rs10918594 rs10494366

NOS1AP and location of rs10494366 and rs10918594. The ruler indicates the physical position on chromosome 1. Thick horizontal lines indicate genes in the region; thick vertical lines, NOS1AP exons; and arrows, the direction of transcription. Thick vertical lines on the ruler indicate the positions of rs10918594 and rs10494366, which are Ϸ55 kb apart. The 2 SNPs were in linkage disequilibrium, with an r2 of 0.63 and DЈ of 0.89.

SCD in unselected populations. In addition, medication- infarction is conducted continuously between exams. In addition, induced prolonged QT interval and ventricular arrhythmias exposure of study participants to medications has been gathered have led to the withdrawal of many noncardiac medications,21 continuously from January 1, 1991, to the present through computerized pharmacy records of the pharmacies in the study area. making the QT interval an important phenotype to study. Previously, we identified a locus on chromosome 3 with Genotyping suggestive evidence of linkage to QT-interval duration, but All participants were genotyped for the NOS1AP SNP rs10494366 the genomic interval was large, and the finding has yet to be TϾG, previously shown to be associated with QT interval in 3 confirmed.15 More recently, Arking et al22 reported the independent samples.22 The correlated SNP rs10918594 CϾG, finding from a genome-wide association study that a common which had evidence of association with QT interval in one of the 22 variant (rs10494366; minor allele frequency, 38%) in the original samples, also was genotyped (see the Figure). Both were genotyped with Taqman assays C_1777074_10 and C_1777009_10 NOS1AP gene was reproducibly associated with QT-interval (Applied Biosystems, Foster City, Calif) in 1 ng genomic DNA variation in several large population samples. The NOS1AP extracted from leukocytes, as previously reported.37 gene, encoding the nitric oxide synthase 1 adaptor protein, has been found to regulate neuronal nitric oxide synthase Assessment of QTc Interval and Other activation23 and to enhance Dexras1 activation by neuronal Electrocardiographic Measurements nitric oxide synthase through a ternary complex.24 Neuronal The electrocardiography (ECG) phenotype studied was the heart nitric oxide synthase–knockout mice have been found to have rate–corrected QT interval (QTc) in milliseconds using Bazett’s ϭ ͌ 38 altered cardiac contractility, which suggests a role for formula (QTc QT/ RR). As in previous studies of QTc in the Rotterdam Study19 we used a 10-second resting 12-lead ECG 25–27 NOS1AP in cardiac depolarization. Furthermore, (average of 8 to 10 beats), which was recorded on an ACTA ECG NOS1AP is capable of interaction with ion channels through (ESAOTE, Florence, Italy) at a sampling frequency of 500 Hz and its PDZ domain.28–30 Nevertheless, the involvement of stored digitally. All ECGs were processed by the Modular ECG NOS1AP in myocardial repolarization was not known until Analysis System (MEANS) to obtain ECG measurements.39–41 the initial report of the association. MEANS determines the QT interval from the start of the QRS complex until the end of the T wave. MEANS also determines the The impact of NOS1AP variants on QT-interval duration in presence of right or left bundle-branch block and left ventricular older populations, in whom nongenetic factors might play a hypertrophy. To study the association between NOS1AP variants and stronger role than heritable factors, is unknown. QTc duration, all eligible ECGs from subjects with DNA available The goal of the present study was to test for association of were used. ECGs with right or left bundle-branch block were the NOS1AP variant with QT duration and to test for its excluded from analyses. In addition, to minimize confounding by nongenetic influences on QT duration, all ECGs taken while the association with SCD in the Rotterdam Study. subject was on any QT-altering drugs were excluded from analyses. Drugs were considered possibly QT prolonging if they appeared on Methods any of lists 1 through 4 at www.qtdrugs.org.42 We also excluded Study Population ECGs if subjects were on flupentixol, levomepromazine, meflo- quine, olanzapine, or sertindole, which may prolong QT interval, or The Rotterdam Study is a prospective population-based cohort study digoxin, which shortens the QT interval. Up to 3 QTc measurements of chronic diseases in the elderly. All inhabitants of Ommoord, a were recorded across the 3 examination cycles. Rotterdam suburb, Ն55 years of age (nϭ10 278), were ascertained Finally, in additional analyses, the mean QTc interval per individ- from the municipal register and invited to participate. Of them, 78% ual was divided into 3 gender-specific categories as previously (nϭ7983; 58% female, 98% white) took part in the baseline described. For women, the cut points were Յ450 ms (normal), 451 examination from March 1990 through July 1993. Second and third to 470 ms (borderline), and Ͼ470 ms (prolonged); for men, the cut examinations were conducted from September 1993 to August 1996 points were Յ430 ms (normal), 431 to 450 ms (borderline), and and from April 1997 to December 1999, respectively. Objectives and Ͼ450 ms (prolonged).19,43 methods of the Rotterdam Study have been described in detail.31 The medical ethics committee of Erasmus Medical Center (Rotterdam, the Netherlands) approved the study, and all participants provided Adjudication of SCD signed informed consent for participation, including retrieval of For the SCD analyses, all genotyped subjects were included. The medical records, use of blood and DNA for scientific purposes, and ascertainment of SCD cases in the Rotterdam Study has been publication of data. DNA for genotyping is available for 6571 described previously.19 SCDs were defined operationally as a wit- participants (82%) from the baseline visit. nessed natural death attributable to cardiac causes, heralded by Clinical characteristics, including smoking, body mass index, abrupt loss of consciousness, within 1 hour of onset of acute hypertension, diabetes mellitus, heart failure, and myocardial infarc- symptoms, or as an unwitnessed, unexpected death of someone seen tion, were ascertained as previously described.19,32–36 Active surveil- in a stable medical condition Ͻ24 hours previously with no evidence lance for incident diabetes mellitus, heart failure, and myocardial of a noncardiac cause.44,45

12 Circulation July 3, 2007

TABLE 1. Baseline Characteristics

Genotyped Sample QTc Sample SCD Cases

Men Women Men Women Men Women Characteristic (nϭ2666, 40.6%) (nϭ3905, 59.4%) (nϭ2191, 40.8%) (nϭ3183, 59.2%) (nϭ116, 49.8%) (nϭ117, 50.2%) Age at baseline, y, meanϮSD 68.2Ϯ8.2 70.4Ϯ9.6 67.0Ϯ7.7 69.0Ϯ9.1 71.8Ϯ7.8 74.4Ϯ7.7 Follow-up time, y, meanϮSD 10.0Ϯ3.8 10.5Ϯ3.7 10.6Ϯ3.4 11.1Ϯ3.2 6.4Ϯ3.8 7.3Ϯ3.8 Current smoking, n (%) 774 (29.0) 680 (17.4) 634 (28.9) 582 (18.3%) 32 (27.6) 15 (12.8%) Past smoking, n (%) 1635 (61.3) 1040 (26.6) 1343 (61.3) 872 (27.4) 75 (64.7) 38 (32.5%) Body mass index, kg/m2, meanϮSD 25.7Ϯ3.0 26.7Ϯ4.1 25.7Ϯ3.0 26.7Ϯ4.0 25.3Ϯ3.0 27.3Ϯ3.9 Systolic blood pressure, mm Hg, meanϮSD 138.7Ϯ21.7 139.8Ϯ22.6 138.3Ϯ21.5 139.2Ϯ22.2 144.6Ϯ24.2 152.8Ϯ27.7 Diastolic blood pressure, mm Hg, meanϮSD 74.6Ϯ11.5 73.2Ϯ11.4 74.9Ϯ11.3 73.2Ϯ11.1 74.0Ϯ12.5 77.0Ϯ14.1 Hypertension, n (%) 780 (29.3) 1415 (36.2) 621 (28.3) 1102 (34.6) 53 (45.7) 65 (55.6) Diabetes mellitus, n (%) 281 (10.5) 422 (10.8) 213 (9.7) 309 (9.7) 14 (12.1) 27 (23.1) Myocardial infarction, n (%) 447 (16.8) 320 (8.2) 345 (15.7) 243 (7.6) 44 (37.9) 19 (16.2) Heart failure, n (%) 81 (3.0) 131 (3.4) 34 (1.6) 75 (2.4) 17 (14.7) 7 (6.0) Shown are characteristics of all individuals with DNA available for genotyping (genotyped sample), of the subset of genotyped subjects with ECGs without bundle-branch block or use of a QT-prolonging drug or digoxin (QTc sample), and of the SCD cases. The SCD source sample includes all genotyped subjects.

Statistical Analysis performed with SPSS for Windows, version 11.0 (SPSS Inc, Chi- Genotype frequencies were tested for Hardy-Weinberg equilibrium cago, Ill). with a ␹2 test. The authors had full access to and take full responsibility for the Because the QTc in subsequent ECGs of the same subject are integrity of the data. All authors have read and agree to the correlated, we used repeated-measures analyses implemented in manuscript as written. PROC MIXED (SAS 8.2, SAS Institute, Cary, NC). Both allelic and general genotype models were tested for the 2 polymorphisms, Results although the allelic model was considered primary because of the Study Subjects previously reported rs10494366–QT relationship.22 Haplotypes were estimated with the expectation-maximization algorithm implemented Baseline characteristics for the total study population, con- in PHASE 2.0 (University of Washington, Seattle),46,47 and only sisting of all genotyped subjects from the Rotterdam Study individuals with successful genotyping for both SNPs and a posterior (nϭ6571), are summarized in Table 1. Within the study probability of PϾ0.95 for assigned haplotypes were included in population, 12 967 ECGs were available from 6052 subjects haplotype analyses. In total, we identified 2245 double heterozy- across up to 3 examination cycles. After exclusion of ECGs gotes, all of whom were phased as heterozygous haplotype TC-GG with right or left bundle-branch block (nϭ640) and those because these are the major haplotypes, with posterior probabilities in excess of 0.95. In haplotype analyses, the haplotype with major performed in individuals taking QT-prolonging or alleles for both SNPs was considered the reference to which the other -shortening drugs (nϭ1334) or both, a total of 11 108 ECGs 3 haplotypes were compared individually. QTc was tested for from 5374 subjects remained (on average, 2.1 ECGs per association with genotype as the sole predictor (crude) and with individual). The 5374 subjects included in the QTc analyses adjustment for age and gender (multivariable). To compare the were 1.3 years younger at baseline, reflecting exclusions outcomes of haplotype analysis with individual SNP analysis, the latter analyses also were performed restricting the analysis to among older participants (Table 1). Women had an 8.9-ms- subjects in whom genotyping was successful for both SNPs. Finally, longer age-adjusted QTc interval (431.4 versus 422.5 ms; a sensitivity analysis was carried out, excluding ECGs with an PϽ0.0001), as previously shown,38,48 and were 2.2 years abnormally prolonged QTc and using gender-specific cutoff points older than men (70.4 versus 68.2 years at baseline; of Ͼ450 ms for men and Ͼ470 ms for women. Jonckheere-Terpstra PϽ0.0001). The numbers of abnormally prolonged QTc in tests were used to test whether individuals carrying NOS1AP minor men and women of our study population were slightly higher alleles had an increased frequency of borderline and abnormal mean QTc. than expected on the basis of numbers from reference Hazard ratios for time to SCD from baseline were estimated with populations.48,49 However, our study population was on Cox proportional-hazards models. Again, both allelic and general average already considerably older at baseline (69.5 versus 53 genotype models were tested for the 2 polymorphisms. In addition to and 61 years, respectively), and this mean further increased NOS1AP genotype, known SCD risk factors—including age, gender, when follow-up ECGs were taken. body mass index, smoking, hypertension, diabetes mellitus, heart failure, and myocardial infarction at baseline and time-dependent incident diabetes mellitus, heart failure, and myocardial infarction— Genotyping Ͼ were included as predictors. To minimize misclassification of SCD, The G-allele (minor) frequency of rs10494366 T G was we additionally performed a subanalysis restricting the case defini- 36.4% and of rs10918594 CϾG was 31.4%. Successful tion to witnessed deaths only. As we have previously shown, the risk genotype calls were made in 95.8% and 95.9% of subjects, of SCD for increasing QTc is stronger in the younger than in the respectively. Both SNPs were in Hardy-Weinberg equilib- 19 older age group, so we determined the hazard ratios for time to rium (Pϭ0.32 for rs10494366 and Pϭ0.89 for rs10918594). SCD separately in groups stratified by age above and below the 2 median age at baseline. Finally, we performed a sensitivity analysis, The 2 SNPs were in linkage disequilibrium, with an r of 0.63 excluding subjects with a history of myocardial infarction at baseline and DЈ of 0.89. On phasing, we observed 2 common 2-SNP from the analysis. All Cox proportional hazards analyses were haplotypes, TC (61.4%) and GG (29.1%), consisting of the 2 Aarnoudse et al NOS1AP Variants Are Associated With QTc Duration 13

TABLE 2. Difference in QTc by NOS1AP Genotype

Genotypic Model* Allelic Model†

Genotype Genotype Genotype P Per G allele P rs10494366 (36.4% MAF) TT TG GG Subjects, n‡ 2100 2334 704 5138 Crude, ms Ref 4.2 (3.0–5.5) 7.1 (5.3–8.9) 2.2ϫ10Ϫ17 3.7 (2.9–4.6) 3.3ϫ10Ϫ18 Age and gender adjusted, ms Ref 4.2 (3.0–5.4) 7.2 (5.5–9.0) 5.9ϫ10Ϫ19 3.8 (3.0–4.6) 7.8ϫ10Ϫ20 rs10918594 (31.4% MAF) CC CG GG Subjects, n‡ 2456 2217 530 5203 Crude, ms Ref 4.3 (3.1–5.5) 6.4 (4.4–8.3) 1.7ϫ10Ϫ15 3.6 (2.7–4.5) 6.9ϫ10Ϫ16 Age and gender adjusted, ms Ref 4.3 (3.2–5.5) 6.3 (4.4–8.2) 1.5ϫ10Ϫ16 3.6 (2.7–4.4) 6.9ϫ10Ϫ17 MAF indicates minor allele frequency; Ref, reference. Values are difference from reference group (95% CI) in milliseconds. *Linear regression model using dummy variables per genotype. †Linear regression model entering genotype as an ordinal variable. ‡Because of failures in genotyping for the individual SNPs, genotype counts do not add up to the total of 5374 individuals. major and 2 minor alleles, respectively, and 2 remaining multivariable-adjusted QTc per additional GG haplotype haplotypes containing 1 major and 1 minor allele each, GC copy (Pϭ2.0ϫ10Ϫ18) using the TC haplotype as reference. (7.2%) and TG (2.3%). Genotype distributions did not differ The GC and TG haplotypes were associated with a 3.2-ms- between men and women and between quartiles of age at longer (Pϭ7.0ϫ10Ϫ4) and 4.1-ms-longer (Pϭ0.01) multivari- baseline. able-adjusted QT interval per additional copy, respectively. None of the haplotypes had a more significant effect than the NOS1AP Polymorphisms and QTc individual SNPs. Minor alleles of both NOS1AP SNPs were significantly Furthermore, rs10494366 and rs10918594 were associated associated with an increase in QTc duration. SNP rs10494366 with a larger proportion of borderline and prolonged QTc TϾG was associated with a 3.8-ms increase in multivariable- intervals using gender-specific cut points19 (test for trend, adjusted QTc interval for each additional G allele, and SNP both PϽ0.0001; Table 3). rs10918594 CϾG was associated with a 3.6-ms increase per additional G allele (Table 2). Additional adjustment for ECG NOS1AP Polymorphisms and SCD left ventricular hypertrophy did not alter the results (data not Within the study population (nϭ6571), we identified 233 shown). We observed no difference in effect of the SNPs sudden cardiac deaths, 121 of which were witnessed. Base- between men and women. A sensitivity analysis excluding line characteristics of all adjudicated SCD cases are shown in ECGs with an abnormally prolonged QTc (using gender- Table 1. After adjustment for known risk factors, the specific cut points) resulted in slightly lower estimates (2.9 NOS1AP polymorphisms rs10494366 TϾG and rs10918594 and 2.7 ms for the allelic models); however, the association of CϾG showed nonsignificant trends in the direction of in- NOS1AP genotypes with QTc duration remained highly creased hazard of SCD, with hazard ratios per additional significant (all PϽ10Ϫ11). minor allele for time to SCD of 1.09 (95% confidence All 3 haplotypes containing 1 (GC and TG) or 2 (GG) interval, 0.90 to 1.33) and 1.10 (95% confidence interval, minor alleles for the 2 SNPs were associated with increased 0.90 to 1.34), respectively. In the subset of 121 adjudicated QTc compared with the homozygous TC reference haplotype. SCD cases that were witnessed, a similar nonsignificant trend The GG haplotype was associated with a 4.1-ms-longer toward increased SCD risk was found (Table 4). Stratification

TABLE 3. Number of Individuals With Normal, Borderline, and Abnormal Mean QTc per Genotype Group Using Gender-Specific Cut Points

Genotype Normal Borderline Prolonged P, Test for Trend rs10494366, n (% within genotype) ...... Ͻ0.0001 TT 1679 (80.0) 329 (15.7) 92 (4.4) TG 1715 (73.5) 447 (19.2) 172 (7.4) GG 498 (70.7) 144 (20.5) 62 (8.8) rs10918594, n (% within genotype) ...... Ͻ0.0001 CC 1945 (79.2) 390 (15.9) 121 (4.9) CG 1609 (72.6) 448 (20.2) 160 (7.2) GG 385 (72.6) 96 (18.1) 49 (9.2) QTc interval divided into 3 gender-specific categories. For women, the cut points were Յ450 ms (normal), 451 to 470 ms (borderline) and Ͼ470 ms (prolonged); for men, they were Յ430 ms (normal), 431 to 450 ms (borderline), and Ͼ450 ms (prolonged).19,43 14 Circulation July 3, 2007

TABLE 4. Hazard Ratio of All Adjudicated SCD and Witnessed SCD per NOS1AP Genotype or Allele

Genotypic Model* Allelic Model†

Genotype Genotype Genotype P Per G Allele P All SCD, HR (95% CI) rs10494366 TT (nϭ90) TG (nϭ95) GG (nϭ36) Crude Ref 0.97 (0.72–1.30) 1.26 (0.85–1.87) 0.41 1.08 (0.89–1.32) 0.42 Full model Ref 0.99 (0.74–1.33) 1.27 (0.85–1.89) 0.44 1.09 (0.90–1.33) 0.37 rs10918594 CC (nϭ101) CG (nϭ103) GG (nϭ24) Crude Ref 1.13 (0.85–1.50) 1.11 (0.70–1.76) 0.69 1.08 (0.88–1.32) 0.46 Full model Ref 1.16 (0.88–1.54) 1.13 (0.71–1.80) 0.58 1.10 (0.90–1.34) 0.37 Witnessed SCD, HR (95% CI) rs10494366 TT (nϭ47) TG (nϭ43) GG (nϭ26) Crude Ref 0.82 (0.54–1.24) 1.66 (1.02–2.70) 0.02 1.20 (0.93–1.56) 0.17 Full model Ref 0.84 (0.55–1.28) 1.68 (1.04–2.74) 0.02 1.22 (0.94–1.58) 0.14 rs10918594 CC (nϭ52) CG (nϭ51) GG (nϭ16) Crude Ref 1.11 (0.75–1.64) 1.43 (0.80–2.54) 0.47 1.17 (0.89–1.53) 0.25 Full model Ref 1.14 (0.77–1.68) 1.45 (0.81–2.59) 0.44 1.18 (0.91–1.55) 0.22 Cox proportional hazards model. HR indicates hazard ratio; Ref, reference; and n, number of cases. Crude model was age and gender adjusted. Full model included age, gender, body mass index, smoking, hypertension, diabetes mellitus, heart failure, and myocardial infarction. *Genotype-specific HR. †HR entering genotype as an ordinal variable under an allelic model. for baseline age above and below the median showed no proximates the effect of medications that delay myocardial difference between age groups (data not shown). Finally, a repolarization and increase liability to ventricular sensitivity analysis excluding 767 subjects with a history of arrhythmias. myocardial infarction at baseline did not result in a substan- The mechanism by which a common variation in NOS1AP tial change of the effect estimates or confidence intervals affects QTc interval duration is unknown at present. How- (data not shown). ever, the statistical evidence supporting the association with QTc interval of rs10494366 (PϽ10Ϫ19) and rs10918594 Discussion (PϽ10Ϫ16) in 5374 individuals confirms that this is a genuine We observed strong replication in the Rotterdam Study, a association, consistent with evidence from 4 independent large, well-phenotyped cohort of European ancestry, of the cohorts totaling Ͼ13 000 individuals of European ancestry. finding from a prior genome-wide association study22 that Our study examined the relationship of genetic variation, common NOS1AP variants are associated with increased present at birth, in an elderly cohort in whom one might age-, gender-, and heart rate–adjusted QT-interval duration. assume that genetic factors play a smaller role than in None of the haplotypes showed a more significant effect than younger cohorts. However, these results demonstrate that the individual SNPs, which were not specifically selected to genetic factors continue to play a role even at older age. characterize haplotype variation at the locus. The 2 SNPs, One major advantage of our study was the availability of which are 55 kb apart, are not known to be functional, nor are up to 3 ECGs per subject at regular intervals during follow- they highly correlated with any known functional SNP. These up, resulting in more precise long-term ECG measures. results support the existence of a causal untyped SNP that is Furthermore, the use of digital ECG recordings all measured correlated with both rs10494366 and rs10918594. with the MEANS system likely reduced systematic differ- The association with SCD was not statistically significant. ences in assessment of the QTc interval. In addition, the Although we cannot fully exclude survival bias because of intersection of the Rotterdam Study with detailed pharmacy the older age of our study population, we did not find that the exposure data allowed us to exclude ECGs recorded in genotype distribution differed between different age groups at individuals on QT-prolonging or -shortening drugs, which baseline, making this less likely. The modest QTc prolonga- could have attenuated the power to detect the association. tion associated with NOS1AP variation, despite the strong Although no information on long-QT syndrome cases was effect of prolonged QTc on SCD risk, suggests that a much available, the number of relatives in the Rotterdam Study is larger study is needed to definitively confirm or rule out an low, and the sensitivity analysis excluding abnormally pro- increased risk of SCD by NOS1AP variants. At least 510 longed QTc further minimized influence of potential familial cases would be needed to detect an odds ratio of 1.2 per minor long-QT syndrome cases. Another advantage of the Rotter- allele with 80% power. Even if no association with SCD is dam Study is the prospective ascertainment of risk factors and ultimately identified, the 7.2-ms increase in QTc interval in the active surveillance for SCD events over a relatively long minor homozygotes compared with major homozygotes ap- period of follow-up. Thus, extensive information surrounding

Aarnoudse et al NOS1AP Variants Are Associated With QTc Duration 15

SCD events was available, including the time between start of 3. Myerburg RJ, Castellanos A. Emerging paradigms of the epidemiology symptoms and death, which enabled rigorous adjudication of and demographics of sudden cardiac arrest. Heart Rhythm. 2006;3: 235–239. SCD events. 4. Jouven X, Desnos M, Guerot C, Ducimetiere P. Predicting sudden death One limitation of the study resides in the variety of in the population: the Paris Prospective Study I. Circulation. 1999;99: competing causes of abrupt death at increasing age, which 1978–1983. 5. Friedlander Y, Siscovick DS, Arbogast P, Psaty BM, Weinmann S, may have led to misclassification, especially in cases in Lemaitre RN, Raghunathan TE, Cobb LA. Sudden death and myocardial which death was unwitnessed. Because SCD coding was infarction in first degree relatives as predictors of primary cardiac arrest. blinded to NOS1AP genotype, this would likely have biased Atherosclerosis. 2002;162:211–216. our study to detect no effect. This might explain our finding 6. Dekker LR, Bezzina CR, Henriques JP, Tanck MW, Koch KT, Alings MW, Arnold AE, de Boer MJ, Gorgels AP, Michels HR, Verkerk A, of a slightly increased, but still nonsignificant, hazard ratio Verheugt FW, Zijlstra F, Wilde AA. Familial sudden death is an when the analyses were restricted to witnessed sudden car- important risk factor for primary ventricular fibrillation: a case-control diac deaths. Our results and those of the prior study by Arking study in acute myocardial infarction patients. Circulation. 2006;114: 1140–1145. 22 et al were restricted to population samples of European 7. Kaikkonen KS, Kortelainen ML, Linna E, Huikuri HV. Family history ancestry. Further testing in samples of African and Asian and the risk of sudden cardiac death as a manifestation of an acute ancestry is needed to establish the role of genetic variation at coronary event. Circulation. 2006;114:1462–1467. NOS1AP 8. Splawski I, Shen J, Timothy KW, Lehmann MH, Priori S, Robinson JL, the locus in myocardial repolarization in these Moss AJ, Schwartz PJ, Towbin JA, Vincent GM, Keating MT. Spectrum population groups. Moreover, substantial frequency differ- of mutations in long-QT syndrome genes: KVLQT1, HERG, SCN5A, ences are observed among European, African, and Asian KCNE1, and KCNE2. Circulation. 2000;102:1178–1185. HapMap samples, which raises the possibility of natural 9. Mohler PJ, Schott JJ, Gramolini AO, Dilly KW, Guatimosim S, duBell WH, Song LS, Haurogne K, Kyndt F, Ali ME, Rogers TB, Lederer WJ, 50 selection in the region. Attempts to validate the NOS1AP Escande D, Le Marec H, Bennett V. Ankyrin-B mutation causes type 4 association in recently admixed populations, such as African long-QT cardiac arrhythmia and sudden cardiac death. Nature. 2003;421: Americans, will need to account for global and local chro- 634–639. 10. Splawski I, Timothy KW, Tateyama M, Clancy CE, Malhotra A, Beggs mosomal differences in ancestry because of the strong asso- AH, Cappuccio FP, Sagnella GA, Kass RS, Keating MT. Variant of ciation with continental ancestry and the risk of population SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science. stratification. 2002;297:1333–1336. 11. International HapMap Consortium. A haplotype map of the human genome. Nature. 2005;437:1299–1320. Conclusions 12. de Bakker PI, Yelensky R, Pe’er I, Gabriel SB, Daly MJ, Altshuler D. We have strongly confirmed the association of NOS1AP Efficiency and power in genetic association studies. Nat Genet. 2005;37: variants with QT-interval duration. With the limited number 1217–1223. 13. Skol AD, Scott LJ, Abecasis GR, Boehnke M. Joint analysis is more of SCD cases in our cohort, it was not possible to demonstrate efficient than replication-based analysis for two-stage genome-wide asso- that this association translates into an influence on risk of ciation studies. Nat Genet. 2006;38:209–213. SCD, although the point estimates suggest that such a risk 14. Hong Y, Rautaharju PM, Hopkins PN, Arnett DK, Djousse L, Pankow JS, Sholinsky P, Rao DC, Province MA. Familial aggregation of QT-interval increase may truly exist. Additional larger studies are re- variability in a general population: results from the NHLBI Family Heart quired to determine whether NOS1AP genotype is associated Study. Clin Genet. 2001;59:171–177. with SCD. 15. Newton-Cheh C, Larson MG, Corey DC, Benjamin EJ, Herbert AG, Levy D, D’Agostino RB, O’Donnell CJ. QT interval is a heritable quantitative trait with evidence of linkage to chromosome 3 in a genome-wide linkage Acknowledgments analysis: the Framingham Heart Study. Heart Rhythm. 2005;2:277–284. We thank Rowena Utberg for genotyping and Cornelis van der Hooft 16. Schouten EG, Dekker JM, Meppelink P, Kok FJ, Vandenbroucke JP, Pool and Charlotte van Noord for their help with adjudication of SCD J. QT interval prolongation predicts cardiovascular mortality in an cases. apparently healthy population. Circulation. 1991;84:1516–1523. 17. Siscovick DS, Raghunathan TE, Rautaharju P, Psaty BM, Cobb LA, Wagner EH. Clinically silent electrocardiographic abnormalities and risk Sources of Funding of primary cardiac arrest among hypertensive patients. Circulation. 1996; Dr Newton-Cheh and the genotyping were supported by a Doris 94:1329–1333. Duke Charitable Foundation Clinical Scientist Development Award 18. Karjalainen J, Reunanen A, Ristola P, Viitasalo M. QT interval as a and NIH K23 (HL80025). cardiac risk factor in a middle aged population. Heart. 1997;77:543–548. 19. Straus SM, Kors JA, De Bruin ML, van der Hooft CS, Hofman A, Heeringa J, Deckers JW, Kingma JH, Sturkenboom MC, Stricker BH, Disclosures Witteman JC. Prolonged QTc interval and risk of sudden cardiac death in None. a population of older adults. J Am Coll Cardiol. 2006;47:362–367. 20. Algra A, Tijssen JG, Roelandt JR, Pool J, Lubsen J. QT interval variables from 24 hour electrocardiography and the two year risk of sudden death. References Br Heart J. 1993;70:43–48. 1. Thom T, Haase N, Rosamond W, Howard VJ, Rumsfeld J, Manolio T, 21. Yap YG, Camm AJ. Drug induced QT prolongation and torsades de Zheng ZJ, Flegal K, O’Donnell C, Kittner S, Lloyd-Jones D, Goff DC Jr, pointes. Heart. 2003;89:1363–1372. Hong Y, Adams R, Friday G, Furie K, Gorelick P, Kissela B, Marler J, 22. Arking DE, Pfeufer A, Post W, Kao WH, Newton-Cheh C, Ikeda M, West Meigs J, Roger V, Sidney S, Sorlie P, Steinberger J, Wasserthiel-Smoller K, Kashuk C, Akyol M, Perz S, Jalilzadeh S, Illig T, Gieger C, Guo CY, S, Wilson M, Wolf P. Heart disease and stroke statistics: 2006 update: a Larson MG, Wichmann HE, Marban E, O’Donnell CJ, Hirschhorn JN, report from the American Heart Association Statistics Committee and Kaab S, Spooner PM, Meitinger T, Chakravarti A. A common genetic Stroke Statistics Subcommittee. Circulation. 2006;113:e85–e151. variant in the NOS1 regulator NOS1AP modulates cardiac repolarization. 2. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert Nat Genet. 2006;38:644–651. JP, Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of 23. Jaffrey SR, Snowman AM, Eliasson MJ, Cohen NA, Snyder SH. a defibrillator in patients with myocardial infarction and reduced ejection CAPON: a protein associated with neuronal nitric oxide synthase that fraction. N Engl J Med. 2002;346:877–883. regulates its interactions with PSD95. Neuron. 1998;20:115–124. 16 Circulation July 3, 2007

24. Fang M, Jaffrey SR, Sawa A, Ye K, Luo X, Snyder SH. Dexras1: a G 38. Bazett H. An analysis of time relations of the electrocardiogram. Heart. protein specifically coupled to neuronal nitric oxide synthase via 1920;7:353–370. CAPON. Neuron. 2000;28:183–193. 39. van Bemmel JH, Kors JA, van Herpen G. Methodology of the modular 25. Massion PB, Pelat M, Belge C, Balligand JL. Regulation of the mam- ECG analysis system MEANS. Methods Inf Med. 1990;29:346–353. malian heart function by nitric oxide. Comp Biochem Physiol A Mol 40. Willems JL, Abreu-Lima C, Arnaud P, van Bemmel JH, Brohet C, Degani Integr Physiol. 2005;142:144–150. R, Denis B, Gehring J, Graham I, van Herpen G, Machado H, Macfarlane 26. Ashley EA, Sears CE, Bryant SM, Watkins HC, Casadei B. Cardiac nitric PW, Michaelis J, Moulopoulos SD, Rubel P, Zywietz C. The diagnostic oxide synthase 1 regulates basal and beta-adrenergic contractility in performance of computer programs for the interpretation of electrocar- murine ventricular myocytes. Circulation. 2002;105:3011–3016. diograms. N Engl J Med. 1991;325:1767–1773. 27. Barouch LA, Harrison RW, Skaf MW, Rosas GO, Cappola TP, Kobeissi 41. de Bruyne MC, Kors JA, Hoes AW, Kruijssen DA, Deckers JW, Grosfeld ZA, Hobai IA, Lemmon CA, Burnett AL, O’Rourke B, Rodriguez ER, M, van Herpen G, Grobbee DE, van Bemmel JH. Diagnostic interpre- Huang PL, Lima JA, Berkowitz DE, Hare JM. Nitric oxide regulates the tation of electrocardiograms in population-based research: computer heart by spatial confinement of nitric oxide synthase isoforms. Nature. program research physicians, or cardiologists? J Clin Epidemiol. 1997; 2002;416:337–339. 50:947–952. 28. Murata M, Buckett PD, Zhou J, Brunner M, Folco E, Koren G. SAP97 42. Drugs That Prolong the QT Interval and/or Induce Torsades de Pointes interacts with Kv1.5 in heterologous expression systems. Am J Physiol Ventricular Arrhythmia. Tucson, Ariz: University of Arizona Center for Heart Circ Physiol. 2001;281:H2575–H2584. Education and Research on Therapeutics, Arizona Health Sciences 29. Leonoudakis D, Mailliard W, Wingerd K, Clegg D, Vandenberg C. Center. Available at: http://www.qtdrugs.org/medical-pros/drug-lists/ Inward rectifier potassium channel Kir2.2 is associated with synapse- drug-lists.htm. Accessed November 8, 2006. associated protein SAP97. J Cell Sci. 2001;114:987–998. 43. The Assessment of the Potential for QT Interval Prolongation by Non- 30. Kim E, Sheng M. Differential Kϩ channel clustering activity of PSD-95 Cardiovascular Medicinal Products. London, UK: Committee for Pro- and SAP97, two related membrane-associated putative guanylate kinases. prietary Medicinal Products; 1997. Neuropharmacology. 1996;35:993–1000. 44. Myerburg RJ, Castellanos A. Cardiac arrest and sudden cardiac death. In: 31. Hofman A, Grobbee DE, de Jong PT, van den Ouweland FA. Deter- Zipes DP, Libby P, Bonow RO, Braunwald E, eds. Heart Disease: A Text minants of disease and disability in the elderly: the Rotterdam Elderly Book of Cardiovascular Medicine. 7th ed. Philadelphia, Pa: Elsevier Study. Eur J Epidemiol. 1991;7:403–422. Saunders Co; 2004:865–908. 32. Diabetes mellitus. In: Technical Reports Series 894. Geneva, Switzerland: 45. Priori SG, Aliot E, Blomstrom-Lundqvist C, Bossaert L, Breithardt G, World Health Organization; 1985. Brugada P, Camm AJ, Cappato R, Cobbe SM, Di Mario C, Maron BJ, 33. Bots ML, Hoes AW, Koudstaal PJ, Hofman A, Grobbee DE. Common McKenna WJ, Pedersen AK, Ravens U, Schwartz PJ, Trusz-Gluza M, carotid intima-media thickness and risk of stroke and myocardial Vardas P, Wellens HJ, Zipes DP. Task Force on Sudden Cardiac Death of infarction: the Rotterdam Study. Circulation. 1997;96:1432–1437. the European Society of Cardiology. Eur Heart J. 2001;22:1374–1450. 34. Mosterd A, Hoes AW, de Bruyne MC, Deckers JW, Linker DT, Hofman 46. Stephens M, Smith NJ, Donnelly P. A new statistical method for hap- A, Grobbee DE. Prevalence of heart failure and left ventricular dys- lotype reconstruction from population data. Am J Hum Genet. 2001;68: function in the general population: the Rotterdam Study. Eur Heart J. 978–989. 1999;20:447–455. 47. Stephens M, Donnelly P. A comparison of bayesian methods for hap- 35. 1999 World Health Organization–International Society of Hypertension lotype reconstruction from population genotype data. Am J Hum Genet. guidelines for the management of hypertension: Guidelines Subcom- 2003;73:1162–1169. mittee. J Hypertens. 1999;17:151–183. 48. Vitelli LL, Crow RS, Shahar E, Hutchinson RG, Rautaharju PM, Folsom 36. Bleumink GS, Knetsch AM, Sturkenboom MC, Straus SM, Hofman A, AR. Electrocardiographic findings in a healthy biracial population: Ath- Deckers JW, Witteman JC, Stricker BH. Quantifying the heart failure erosclerosis Risk in Communities (ARIC) Study Investigators. epidemic: prevalence, incidence rate, lifetime risk and prognosis of heart Am J Cardiol. 1998;81:453–459. failure: the Rotterdam Study. Eur Heart J. 2004;25:1614–1619. 49. Rautaharju PM, Prineas RJ, Kadish A, Larson JC, Hsia J, Lund B. Normal 37. Fang Y, van Meurs JB, d’Alesio A, Jhamai M, Zhao H, Rivadeneira F, standards for QT and QT subintervals derived from a large ethnically Hofman A, van Leeuwen JP, Jehan F, Pols HA, Uitterlinden AG. diverse population of women aged 50 to 79 years (the Women’s Health Promoter and 3Ј-untranslated-region haplotypes in the vitamin D receptor Initiative [WHI]). Am J Cardiol. 2006;97:730–737. gene predispose to osteoporotic fracture: the Rotterdam Study. Am J Hum 50. International HapMap Project. Available at: www.hapmap.org/cgi-perl/ Genet. 2005;77:807–823. gbrowse/hapmap21_B35. Accessed November 6, 2006.

CLINICAL PERSPECTIVE Sudden cardiac death (SCD) claims 300 000 lives annually in the United States. The ECG QT interval is a noninvasive measure of ventricular repolarization, and prolongation of the QT interval is an important risk factor for SCD and drug-induced arrhythmias. Approximately 35% of the variation in QT-interval duration is attributable to heritable factors. Until recently, the search for sequence variants contributing to QT-interval duration and SCD risk has been restricted to candidate genes known for their role in arrhythmogenesis. However, in a recent genome-wide association study, a common variant in the nitric oxide synthase 1 adaptor protein (NOS1AP) gene was found to be associated with QT-interval variation. NOS1AP was not previously known to play a role in repolarization. In the present study, we have strongly confirmed the association of NOS1AP variants and QT-interval duration with a difference between minor homozygotes and major homozygotes of 7.2 ms (PϽ10Ϫ19). We did not find an association of NOS1AP variants with SCD cases in our cohort, but with 228 SCD events, our study was underpowered to demonstrate such an effect. Even if no association with SCD is ultimately identified, the 7.2-ms increase in QTc interval in minor allele homozygotes compared with major homozygotes approximates the effect of medications that delay myocardial repolarization and increase liability to ventricular arrhythmias. The study underscores the power of association methods to identify novel genes and pathways involved in myocardial repolarization and to identify genetic variants that could contribute to the risk of cardiac arrhythmias. Nonsense Mutations in hERG Cause a Decrease in Mutant mRNA Transcripts by Nonsense-Mediated mRNA Decay in Human Long-QT Syndrome

Qiuming Gong, MD, PhD; Li Zhang, MD; G. Michael Vincent, MD; Benjamin D. Horne, PhD, MPH; Zhengfeng Zhou, MD, PhD

Background—Long-QT syndrome type 2 (LQT2) is caused by mutations in the human ether-a-go-go-related gene (hERG). More than 30% of the LQT2 mutations result in premature termination codons. Degradation of premature termination codon–containing mRNA transcripts by nonsense-mediated mRNA decay is increasingly recognized as a mechanism for reducing mRNA levels in a variety of human diseases. However, the role of nonsense-mediated mRNA decay in LQT2 mutations has not been explored. Methods and Results—We examined the expression of hERG mRNA in lymphocytes from patients carrying the R1014X mutation using a technique of allele-specific transcript quantification. The R1014X mutation led to a reduced level of mutant mRNA compared with that of the wild-type allele. The decrease in mutant mRNA also was observed in the LQT2 nonsense mutations W1001X and R1014X using hERG minigenes expressed in HEK293 cells or neonatal rat ventricular myocytes. Treatment with the protein synthesis inhibitor cycloheximide or RNA interference–mediated knockdown of the Upf1 protein resulted in the restoration of mutant mRNA to levels comparable to that of the wild-type minigene, suggesting that hERG nonsense mutations are subject to nonsense-mediated mRNA decay. Conclusions—These results indicate that LQT2 nonsense mutations cause a decrease in mutant mRNA levels by nonsense-mediated mRNA decay rather than production of truncated proteins. Our findings suggest that the degradation of hERG mutant mRNA by nonsense-mediated mRNA decay is an important mechanism in LQT2 patients with nonsense or frameshift mutations. (Circulation. 2007;116:17-24.) Key Words: arrhythmia Ⅲ ion channels Ⅲ long-QT syndrome Ⅲ myocytes

ong-QT syndrome is a disease associated with delayed becoming clear that nonsense and frameshift mutations bear- Lcardiac repolarization and prolonged QT intervals on the ing PTCs can destabilize mRNA transcripts via a mechanism ECG, which can lead to ventricular arrhythmias and sudden known as nonsense-mediated mRNA decay (NMD) in many death.1 The inherited long-QT syndrome type 2 (LQT2) is human diseases, resulting in decreased abundance of mutant caused by mutations in the human ether-a-go-go-related gene mRNA transcripts rather than in production of truncated (hERG), which encodes the pore-forming subunit of the proteins.19,20 ϩ rapidly activating delayed rectifier K channel (IKr)inthe Clinical Perspective p 24 heart.2,3 More than 250 hERG mutations have been identified NMD is an RNA surveillance mechanism that selectively in patients with LQT2.4–7 The mechanisms of hERG channel degrades mRNA transcripts containing PTCs resulting from dysfunction in LQT2 mutations have been studied exten- nonsense or frameshift mutations. The role of NMD as a 8–11 sively in the last 10 years. Most previous studies, how- disease-causing mechanism of PTC mutations is becoming ever, have focused on the analysis of mutant proteins and increasingly evident.19,20 According to the proposed rule, channel function. More than 30% of LQT2 mutations are NMD occurs when translation terminates Ͼ50 to 55 nt nonsense or frameshift mutations that introduce premature upstream of the 3Ј-most exon-exon junction.21,22 The molec- termination codons (PTCs).4–7 These PTC mutations gener- ular mechanisms of NMD have been studied extensively. ally are assumed to result in truncated dysfunctional channel These studies have shown that pre-mRNA splicing deposits proteins, and several nonsense and frameshift mutations have the exon junction complex Ϸ20 to 40 nt upstream of the been studied at the protein level.8,12–18 However, it is now exon-exon junction in spliced mRNA. The exon junction

Received October 26, 2006; accepted May 8, 2007. From the Division of Cardiovascular Medicine, Department of Medicine, Oregon Health and Science University, Portland (Q.G., Z.Z.); and Departments of Medicine and Cardiology (L.Z., G.M.V.) and Genetic Epidemiology Division (B.D.H.), LDS Hospital, Intermountain Healthcare and University of Utah, Salt Lake City. Correspondence to Dr Zhengfeng Zhou, Division of Cardiovascular Medicine, Oregon Health and Science University, 3181 SW Sam Jackson Park Rd, Portland, OR 97239. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.107.708818 17 18 Circulation July 3, 2007

Figure 1. Pedigree of the family with the R1014X mutation.

complex can recruit Upf proteins, which are required for was treated with RNase-free DNase to remove genomic DNA. NMD.22 Several Upf proteins (Upf1, Upf2, Upf3a, Upf3b) Genomic DNA was isolated from lymphocytes or Epstein Barr have been identified.23 The Upf1 protein appears to play a key virus–transformed lymphoblastoid cells with the DNeasy tissue kit (Qiagen, Valencia, Calif). role in the distinction between proper and improper translation termination. Upf1 is a group 1 helicase that has RNA- Allele-Specific Quantification of RNA Transcripts Ј Ј dependent ATPase and ATP-dependent 5 to 3 helicase and Genomic DNA activities. Knockdown of Upf1 by RNA interference (RNAi) The relative abundance of RNA transcripts from WT and R1014X has been shown to inhibit NMD.24,25 alleles was determined by a modified “hot-stop” polymerase chain The objective of this work was to determine whether NMD reaction (PCR) method.27,28 In this assay, the regular reverse- occurs in hERG mutations that contain PTCs. We investi- transcription PCR was carried out using the primers in exon 13 gated 2 nonsense mutations, W1001X and R1014X, in the (E13-F, forward 5Ј-GCCTTCTCAGGAGTGTCCAA-3Ј) and exon Ј Ј C-terminal region of the hERG channel. The W1001X and 14 (E14-R, reverse 5 -GAAAGCGAGTCCAAGGTGAG-3 ). After 35 cycles, [32P]-dCTP was added and subjected to a single cycle of R1014X mutations have previously been studied at the PCR.28 With hot-stop PCR, only homoduplexes incorporated 32P- protein level using hERG cDNAs.14,15 It was found that both labels and any heteroduplexes formed during previous cycles were mutations produced truncated hERG channel proteins and unlabeled. Thus, hot-stop PCR will prevent the detection of WT/ reduced hERG current amplitude. The R1014X mutation also mutant heteroduplexes, which are resistant to restriction enzyme caused a dominant-negative effect on the wild-type (WT) digestion. Because hot-stop PCR analysis yields a relative measure of transcripts from 2 alleles, normalization to a reference housekeep- hERG current, which is expected to result in a severe clinical ing gene is unnecessary. The hERG genomic DNA was analyzed by phenotype. However, the R1014X carriers have presented hot-stop PCR with the same forward primer as used in reverse- with a mild phenotype. In the present study, we demonstrate transcription PCR and a reverse primer in intron 13 (I13-R, 5Ј- that rather than the production of truncated proteins, the CTCCGCGCTAGAGGTGTG-3Ј). For analysis of allelic variation in primary defect of the W1001X and R1014X mutations is the hERG mRNA expression in normal subjects, the ratio of a common degradation of mutant mRNA by NMD. polymorphism 1692A/G was determined by hot-stop PCR using the primers in exon 6 (E6-F, forward 5Ј-ATCAACTTCCGCACCCCTA- 3Ј) and exon 7 (E7-R, reverse 5Ј-TGTGTGGCTGCTCCATGT-3Ј). The Methods labeled PCR products were treated with TaqIorNheI restriction enzyme Subjects and analyzed by 5% PAGE and autoradiography. For quantitative The study was approved by the institutional review board and carried analysis, the intensity of each band was quantified with Scion Image out on receipt of informed consent. The participants were blood- software (Scion Corp, Frederick, Md). The ratio of 2 alleles was related members of a large family previously identified as having the calculated, and a correction factor according to the respective GC 28 R1014X mutation.4 The pedigree included 25 blood-related family content of each digested product was applied to the ratio. members in 4 generations (Figure 1). Phenotyping was performed on the basis of the history of LQTS-related cardiac events, the assess- Construction of Minigenes ment of QT intervals and T-wave morphology, and pedigree analy- Human genomic DNA was used as a template for PCR amplification sis.26 Genotyping was conducted by sequencing of DNA samples of fragments spanning from hERG exons 12 to 15. The PCR collected from buccal swabs. Normal control subjects were unrelated products were cloned into pCRII vector with the TA cloning kit individuals. (Invitrogen, Carlsbad, Calif) and verified by DNA sequencing. The minigenes were then subcloned into a mammalian expression vector RNA and DNA Preparations From Blood Samples pcDNA5/FRT (Invitrogen). The N-terminus of the minigene was Total RNA was isolated from peripheral blood lymphocytes using tagged by Myc epitope, which is in frame with the hERG translation the RiboPure-Blood kit (Ambion, Austin, Tex). The isolated RNA sequence. The W1001X and R1014X mutations in the minigenes Gong et al Nonsense-Mediated mRNA Decay in LQT2 19

were generated with the pAlter in vitro site-directed mutagenesis penicillin (100 U/mL), and streptomycin (100 ␮g/mL). After 1 day system (Promega, Madison, Wis) and verified by DNA sequencing. in culture, myocytes were infected with the recombinant adenoviruses. Stable Expression of Minigene Constructs in HEK293 Cells Statistical Analysis The minigenes in pcDNA5/FRT vector were stably transfected into Data are presented as meanϮSD for QTc intervals or meanϮSEM HEK293 cells by using the Flp-In method (Invitrogen). In this for PCR and RPA analyses. Statistical comparison of QTc intervals approach, an FRT site sequence is integrated into the genome of between R1014X mutation carriers and noncarriers was performed HEK293 cells and recombined by Flp recombinase with the FRT site with a family-based analysis approach using the software package of the pcDNA5/FRT vector. The pcDNA5/FRT vector carries the PedGenie, a Monte Carlo simulation–based program.31 ANOVA hygromycin resistance gene, which is used for the selection of stable with Bonferroni correction for multiple pairwise comparisons be- cell lines. tween mutation/treatment groups was used for statistical analysis of RPA data. Values of PϽ0.05 were considered statistically RNase Protection Assay of mRNA Transcripts significant. From Minigene Transfected Cells The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the RNA isolation and RNase protection assay (RPA) were performed as manuscript as written. previously described.29 Briefly, cytoplasmic RNA was isolated from HEK293 cells or neonatal rat ventricular myocytes expressing hERG minigenes with the RNeasy kit (Qiagen). The antisense RNA Results riboprobes were transcribed in vitro in the presence of biotin-16-UTP Patient Description ␮ (Roche, Indianapolis, Ind). RNA (30 g) was analyzed with the A total of 22 family members were tested for the presence of riboprobes using the RPAIII and BrightStart BioDetect kits (Am- bion). Yeast RNA was used as control for the complete digestion of the R1014X mutation. Nine family members were identified the probes by RNase. The expression level of the hygromycin as R1014X mutation carriers (Figure 1). The ECG data were resistance gene from the pcDNA5/FRT vector or the E2 gene from available for 7 of the mutation carriers, all of which showed adenovirus was used as a loading control for normalization. The a prolonged QTc interval and typical LQT2 ECG pattern with intensity of each band was quantified with Scion Image software. the subtle bifid T waves. The mean initial QTc interval in the RNA Interference mutation carriers was 461Ϯ7ms(nϭ7) versus 420Ϯ13 ms Two plasmids, pSUPERpuro-hUpf1/I and pSUPERpuro-hUpf1/II (nϭ8) in noncarriers (PϽ0.001). Four mutation carriers had (kindly provided by Dr Oliver Mühlemann), were used to inhibit exercise tests, with maximum QTc value of 510Ϯ10 ms. In expression of Upf1 as described by Paillusson et al.25 These plasmids this family, 89% (8 of 9) of the R1014X mutation carriers Ј contain short hairpin RNAs targeting 2 sequences of hUpf1 (5 - were asymptomatic. The only person with a history of cardiac GAGAATCGCCTACTTCACT-3Ј for pSUPERpuro-hUpf1/I and 5Ј-GATGCAGTTCCGCTCCATT-3Ј for pSUPERpuro-hUpf1/II). events is the 71-year-old proband. From 32 to 42 years of age, The HEK293 cells stably expressing WT or R1014X minigenes were she had multiple syncopal episodes and 1 cardiac arrest that transfected with a mixture of 1 ␮g pSUPERpuro-hUpf1/I and 1 ␮g were associated with the presence of hypokalemia (serum Kϩ, pSUPERpuro-hUpf1/II or 2 ␮g pSUPERpuro with scrambled se- 2.7 mEq/L) caused by taking a dietary supplement containing quence of hUpf1/I using LipofectAmine 2000 (Invitrogen). At 24 potassium-wasting diuretics or taking QT-prolonging antihis- hours after transfection, puromycin was added to the final concentration of 1.5 ␮g/mL for 48 hours to eliminate the untransfected cells. tamines. Since then, she has remained asymptomatic by Before analysis, the cells were cultured without puromycin for at stopping the potassium-wasting diet, avoiding QT-prolonging least 24 hours to avoid potential effects of this translation inhibitor drugs, and taking ␤-blockers. on NMD. The knockdown of the Upf1 protein was analyzed by Western blot as described.9,25 Analysis of mRNA Isolated From Blood Samples Construction and Use of Recombinant Adenovirus The R1014X mutation causes premature termination of the The AdEasy vector kit was used to generate WT and R1014X hERG channel protein. This mutation has previously been minigene recombinant adenoviruses (Stratagene, La Jolla, Calif). studied at the protein level.15 However, it has been known First, the WT and R1014X minigenes were subcloned into pShuttle- that nonsense and frameshift mutations that contain PTCs can CMV vector and recombined with the pAdEasy plasmid in Esche- lead to the degradation of mRNA transcripts by NMD in richia coli strain BJ5183. The pAdEasy/minigene plasmids were many diseases.19,20 To determine the underlying pathogenic transfected into HEK293 cells. After 2 days, the transfected cells were cultured in growth medium containing 1.25% Seaplaque- mechanism of the R1014X mutation, it is important to study agarose to promote the formation of recombinant viral plaques. this mutation at the mRNA level. Because the affected heart Approximately 2 to 3 weeks later, individual plaques were picked, tissue from the mutation carriers was not available for this amplified in HEK293 cells, and purified over a discontinuous CsCl study, we analyzed hERG mRNA transcripts isolated from gradient. the lymphocytes of patients carrying the R1014X mutation. Primary Culture of Neonatal Rat To distinguish between WT and R1014X alleles, we per- Ventricular Myocytes formed allele-specific quantification analysis using the hot- Neonatal rat ventricular myocytes were prepared as described.30 stop PCR assay. The WT allele contains a TaqI restriction Briefly, 1- to 3-day-old Sprague-Dawley rat pups were killed under site, which is destroyed by the R1014X mutation. After ether anesthesia by decapitation, and hearts were removed through a reverse transcription of mRNA, cDNA was amplified by sternotomy. The ventricles were trimmed free of atria, fat, and hot-stop PCR. After digestion of the PCR products with TaqI, connective tissues. Myocytes were dissociated by several 20-minute cycles of collagenase/pancreatin treatment and serum neutralization. the WT allele should yield 2 fragments of 287 and 72 bp, and Myocytes were cultured in Dulbecco’s modified Eagle’s medium the R1014X allele should give a fragment of 359 bp. As with 17% Media 199, 10% horse serum, 5% fetal bovine serum, shown in Figure 2A, cDNA from a normal subject showed a 20 Circulation July 3, 2007

hERG mRNA expression. To rule out possible allelic variation in hERG expression in the general population, we examined the allele-specific expression of hERG mRNA in normal subjects. We analyzed 3 normal subjects who are heterozygous for a common polymorphism, 1692A/G. To distinguish between 1692A and 1692G alleles, the relative levels of mRNA transcripts from 1692A and 1692G alleles were measured by the hot-stop PCR assay. The 1692A allele contains an NheI restriction site, which is absent in the 1692G allele. Thus, digestion with NheI should allow us to determine the relative ratio of the 2 WT alleles. After digestion of the PCR products with NheI, the 1692A allele should be cut into 2 fragments of 286 and 46 bp, and the 1692G alleles should remain uncut (332 bp). As shown in Figure 2B, in subjects 3, 4, and 5 (lanes 3 to 5), there are 2 bands of 268 and 332 bp, suggesting that they are heterozygous for the 1692A/G polymorphism. In these 3 normal subjects, the average ratio of 1692G to 1692A was 0.97Ϯ0.07. This result suggests that there is no significant allelic variation in hERG mRNA expression in normal subjects. Subjects 1 and 2 (lanes 1 and 2) are homozygous for 1692A and 1692G, respectively.

Minigene Analysis of the R1014X and W1001X Mutations To study whether the decrease in the abundance of mRNA Figure 2. Hot-stop PCR analysis of mRNA and genomic DNA levels in the R1014X mutation is due to NMD, we con- isolated from lymphocytes. A, Analysis of WT and R1014X mutant alleles in a normal subject (NS) and the proband. Sche- structed minigenes containing the hERG genomic sequence matic diagrams are shown for hot-stop PCR analysis of cDNA spanning from exon 12 to 15 and expressed the minigenes in (left) and genomic DNA (right). In cDNA analysis, a forward HEK293 cells. In the minigene experiments, the 2 LQT2 primer in exon 13 (E13-F) and a reverse primer in exon 14 nonsense mutations R1014X and W1001X were analyzed by (E14-R) were used, and in genomic DNA analysis, the same forward primer and a reverse primer in intron 13 (I13-R) were used. RPA. Figure 3A shows the structure of the minigene and the The position of TaqI and the size of the fragments from WT and mRNAs after splicing. The R1014X and W1001X mutations mutant PCR products are shown. After digestion with TaqI, the lead to a PTC in exon 13, which is expected to trigger NMD. 32P-labeled hot-stop PCR products were analyzed by PAGE and autoradiography. In cDNA analysis (left), the bands from WT and As shown in Figure 3B, the mRNA level of the R1014X R1014X alleles are 287 and 359 bp, respectively; in genomic minigene was significantly lower than that of the WT DNA analysis (right), the bands from WT and R1014X alleles are minigene. Because degradation of mRNA by NMD depends 275 and 347 bp, respectively (the 72-bp band ran off the gel). on protein synthesis, we examined whether inhibition of Similar results were obtained in 2 additional R1014X carriers, and 3 to 5 independent experiments were performed for each protein synthesis by cycloheximide (CHX) abrogates NMD patient. B, Analysis of allelic variation of hERG expression by of the mutant mRNA, as has been shown for other PTC- analyzing 1692A/G polymorphism in 5 normal subjects. A sche- containing transcripts.32 The cells expressing WT and matic diagram is shown for the position of NheI and the size of the fragments from A and G alleles. The 32P-labeled hot-stop R1014X minigenes were treated with CHX for 3 hours before PCR products were digested with NheI. The 286- and 332-bp RNA isolation. Treatment with CHX had no effect on the bands represent 1692A and 1692G alleles, respectively (the level of WT mRNA but significantly increased the level of 46-bp band ran off the gel). Results shown are representative of R1014X mutant mRNA, suggesting that the mutant mRNA is 2 independent experiments. degraded by NMD. Similar results were observed in the W1001X minigene (Figure 3C), suggesting that the degrada- single band at 287 bp, corresponding to the WT alleles, tion of PTC-containing mRNAs by NMD may represent a whereas in cDNA from the proband, in addition to a WT common mechanism in LQT2 patients with nonsense 287-bp band, a weak 359-bp band from the R1014X mutant mutations. allele was observed. Quantitative analysis of the samples from 3 patients carrying the R1014X mutation revealed that Effect of Suppression of Upf1 on NMD of the the level of the R1014X mutant was reduced to 23Ϯ1% of the R1014X Mutation WT level, suggesting that the mRNA derived from the Recently, the Upf1 protein has been identified as a key factor R1014X mutant allele is decreased. As a control, we also for NMD. Reducing Upf1 expression by RNAi has been used analyzed genomic DNA from these 3 patients and showed as a functional assay to assess the NMD sensitivity of that the ratio of R1014X to WT alleles was 1.03Ϯ0.03, very PTC-containing mRNA transcripts.24,25 To study the role of close to the expected ratio of 1 (Figure 2A). Upf1 in the reduced mRNA level of the R1014X mutation, The allele-specific quantification analysis depends on the we used the RNAi method to knock down Upf1 protein assumption that there is no significant allelic variation in expression. In these experiments, HEK293 cells stably ex-

Gong et al Nonsense-Mediated mRNA Decay in LQT2 21

Figure 3. Analysis of the R1014X and W1001X mutations using minigene constructs. A, The structure of the Myc-tagged minigene and spliced mRNAs. The positions of WT termination codon (TER) and mutation-induced PTCs are indicated. B, C, Analysis of mRNA by RPA. HEK293 cells were stably transfected with WT, R1014X (B), or W1001X (C) minigenes, and the expressed mRNA was analyzed by RPA. Cells expressing WT and mutant minigenes were treated (ϩ) or not treated (Ϫ) with 100 ␮g/mL CHX for 3 hours before RNA isolation. The level of hygromycin resistance gene transcripts (Hygro) served as a loading control. The quantitative data after normalization using protected hygromycin resistance gene mRNA are plotted as percentage of WT control from 4 (B) or 3 (C) independent experiments. Probability values are Bonferroni corrected. pressing the WT and R1014X minigenes were transfected that PTC-containing mRNA transcripts in LQT2 are subject with pSUPERpuro-hUpf1/I and pSUPERpuro-hUpf1/II.25 to NMD. NMD is an evolutionarily conserved mRNA sur- The Upf1 knockdown in the transfected cells was confirmed veillance pathway that detects and eliminates PTC-containing by Western blot analysis using anti-Upf1 antibody (a gift mRNA transcripts, thereby preventing the synthesis of trun- from Dr Jens Lykke-Andersen) (Figure 4A).23 Detection of tubulin with anti-tubulin antibody served as a loading control. In the RPA analysis of hERG minigene mRNA transcripts, the level of R1014X mutant mRNA was significantly increased in Upf1-siRNA–transfected cells (Figure 4B). These results suggest that mRNA transcripts of the R1014X mutation undergo NMD.

Analysis of NMD in Neonatal Rat Myocytes Using R1014X Adenovirus Minigene The above experiments indicate that mRNA transcripts of the R1014X mutation are subject to NMD in lymphocytes and HEK293 cells. The noncardiac cells may behave differently from cardiac cells in the degradation of mutant mRNA by NMD. Therefore, it is important to evaluate whether the defects observed in noncardiac systems are present in cardiac myocytes. To test whether NMD of the R1014X mutation occurs in cardiac myocytes, we infected neonatal rat ventricular myocytes with WT or R1014X minigene adenovirus and performed RPA analysis. As shown in Figure 5, the mRNA level of the R1014X mutant was significantly lower than that of WT. Treatment with CHX had no effect on the level of WT mRNA but significantly increased the level of R1014X mutant mRNA, suggesting that the R1014X mutant mRNA is degraded by NMD in cardiac myocytes. No protected bands in the control lane indicate that the riboprobe is specific for Figure 4. Effect of suppression of Upf1 by RNAi on NMD of the R1014X mutation. HEK293 cells stably expressing the WT and exogenous hERG transcripts. R1014X minigenes were transfected with pSUPERpuro-hUpf1/I and pSUPERpuro-hUpf1/II (Upf1) or pSUPERpuro-scrambled Discussion (CON) constructs. A, Western blot analysis of Upf1 protein. B, The present results demonstrate that the W1001X and Analysis of mRNA by RPA. The quantitative data after normalization using protected hygromycin-resistant gene mRNA are R1014X mutations lead to a reduction of mutant mRNA plotted as percentage of WT control from 4 independent experi- transcripts by NMD. Our findings provide the first evidence ments. Probability values are Bonferroni corrected. 22 Circulation July 3, 2007

classification scheme (shown in Figure 6) illustrates the mechanisms underlying LQT2 mutations. Class 1 mutations cause abnormal protein synthesis by defective transcription or translation. Class 2 mutations lead to defective protein trafficking. Class 3 mutations result in abnormal gating and/or kinetics, and class 4 mutations result in altered or absent channel selectivity or permeability.11 In the present study, we show that LQT2 nonsense mutations cause a decrease in mutant mRNAs by NMD, thereby altering the amount of mRNA available for subsequent hERG protein generation. We propose that the degradation of PTC- containing mRNA transcripts by NMD represents a new class of LQT2 pathogenic mechanism (class 5). The mutations that undergo NMD will result in the Figure 5. Analysis of NMD in neonatal rat ventricular myocytes using R1014X minigene adenovirus constructs. Myocytes degradation of mutant mRNAs before they produce large infected by WT and R1014X mutant minigene adenoviruses quantities of truncated proteins. By eliminating abnormal were treated (ϩ) or not treated (Ϫ) with 100 ␮g/mL CHX for 3 mRNA transcripts carrying PTCs, NMD prevents the produc- hours, and expressed mRNA was analyzed by RPA. The mRNA from uninfected myocytes was used as control (CON). The level tion of truncated proteins that could act in a dominant- of E2 transcripts (E2) from adenovirus served as a loading con- negative manner, leading to deleterious effects on the cells. trol. The quantitative data after normalization using protected E2 One of the physiological roles of NMD is to protect against mRNA are plotted as percentage of WT control from 4 indepen- severe disease phenotypes by converting the dominant- dent experiments. Probability values are Bonferroni corrected. negative effect to haploinsufficiency.32 NMD as a modifier of phenotypic severity has been reported in many human dis- cated and potentially harmful proteins.33 NMD occurs when eases.19,20,32,34 For example, in Marfan syndrome, an translation terminates Ͼ50 to 55 nt upstream of the 3Ј-most autosomal-dominant connective tissue disorder caused by exon-exon junction.21,22 According to this rule, Ͼ90 LQT2 mutations in the fibrillin 1 gene, nonsense mutations that nonsense and frameshift mutations are potential targets for result in reduced levels of mutant mRNA are associated with NMD. Several LQT2 nonsense and frameshift mutations have a mild phenotype. In contrast, patients with nonsense alleles been studied at the functional and protein levels with cDNA that escape NMD develop a severe phenotype as a result of 8,12–18 constructs. All previous studies, however, have been the dominant-negative effect.19,34 carried out under the assumption that these nonsense and Most R1014X mutation carriers in this family have pre- frameshift mutations lead to the production of truncated sented with a mild LQT2 phenotype. In contrast to patients proteins. Because NMD requires introns, the absence of with pore-region mutations, who usually present with a introns in cDNA constructs would preclude the degradation longer QT interval and more frequent cardiac events,35 the of PTC-containing transcripts by NMD. As a result, NMD QTc interval in the R1014X mutation carriers is only mildly effects could not be observed when cDNAs were used in prolonged (461Ϯ7 ms), and only the proband experienced these studies. In the present study, we used minigene con- arrhythmia-related cardiac events that were always associated structs that contain the hERG genomic DNA with both exons with hypokalemia or the use of QT-prolonging drugs. We and introns and showed that the W1001X and R1014X have previously shown that the R1014X mutation causes mutations cause a marked decrease in mutant mRNA tran- hERG channel dysfunction by defective trafficking of the scripts. Inhibition of protein synthesis by CHX or knockdown mutant protein.15 In addition, the truncated mutant protein of Upf1 by RNAi results in the restoration of mutant mRNA exhibits a dominant-negative effect on the WT hERG. This to levels comparable to the WT minigene. These results implies that a severe phenotype would be expected in the strongly suggest that the degradation of mutant mRNA by R1014X mutation carriers. However, our present study re- NMD is an important mechanism in LQT2 mutations carry- veals that the R1014X mutant mRNA transcripts are mark- ing PTCs. edly decreased by NMD, and as a result, the dominant- Previous studies have shown that different LQT2 muta- negative effect caused by the production of truncated proteins tions cause hERG channel dysfunction by different mecha- would be minimized. Therefore, haploinsufficiency rather nisms. This led to a proposed classification of LQT2 muta- than a dominant-negative effect is probably the underlying tions according to their underlying mechanisms.11 The mechanism for the R1014X mutation, which is consistent

Figure 6. Classiﬁcation scheme for LQT2 mutations.

Gong et al Nonsense-Mediated mRNA Decay in LQT2 23 with the observed clinical presentation of this family. It is 3. Sanguinetti MC, Jiang C, Curran ME, Keating MT. A mechanistic link interesting to note that the W1001X mutation carriers also between an inherited and an acquired cardiac arrhythmia: HERG encodes 35 35 the IKr potassium channel. Cell. 1995;81:299–307. present with a mild LQT2 phenotype. Moss et al reported 4. Splawski I, Shen J, Timothy KW, Lehmann MH, Priori S, Robinson JL, that LQT2 patients with mutations in the pore region of Moss AJ, Schwartz PJ, Towbin JA, Vincent GM, Keating MT. Spectrum hERG have a significantly higher risk of arrhythmia-related of mutations in long-QT syndrome genes: KVLQT1, HERG, SCN5A, cardiac events than patients with nonpore mutations. Al- KCNE1, and KCNE2. Circulation. 2000;102:1178–1185. 5. Napolitano C, Priori SG, Schwartz PJ, Bloise R, Ronchetti E, Nastoli J, though the difference may be explained by in vitro electro- Bottelli G, Cerrone M, Leonardi S. Genetic testing in the long QT physiological effects of reported hERG mutations, with pore syndrome: development and validation of an efficient approach to geno- mutations having a greater negative effect on hERG current typing in clinical practice. JAMA. 2005;294:2975–2980. than nonpore mutations,35 it also is possible that NMD may 6. Tester DJ, Will ML, Haglund CM, Ackerman MJ. Compendium of cardiac channel mutations in 541 consecutive unrelated patients play a role. It is noted that only 6% of LQT2 mutations in the referred for long QT syndrome genetic testing. Heart Rhythm. 2005; pore region are nonsense or frameshift mutations, whereas 2:507–517. Ͼ40% of the mutations in nonpore regions are nonsense or 7. Millat G, Chevalier P, Restier-Miron L, Da Costa A, Bouvagnet P, frameshift mutations. Clearly, further genotype-phenotype Kugener B, Fayol L, Gonzalez Armengod C, Oddou B, Chanavat V, Froidefond E, Perraudin R, Rousson R, Rodriguez-Lafrasse C. Spectrum correlation studies are required to test whether NMD contrib- of pathogenic mutations and associated polymorphisms in a cohort of 44 utes to the observed differences in clinical presentations of unrelated patients with long QT syndrome. Clin Genet. 2006;70: pore and nonpore LQT2 mutations. 214–227. There are potential limitations to the present study. Our 8. Sanguinetti MC, Curran ME, Spector PS, Keating MT. Spectrum of HERG Kϩ-channel dysfunction in an inherited cardiac arrhythmia. Proc present experiments analyzed endogenously expressed Natl Acad Sci U S A. 1996;93:2208–2212. mRNA from patients carrying the R1014X mutation, but the 9. Zhou Z, Gong Q, Epstein ML, January CT. HERG channel dysfunction in RNA was isolated from lymphocytes rather than the affected human long QT syndrome: Intracellular transport and functional defects. heart tissue. Although we have shown that the R1014X J Biol Chem. 1998;273:21061–21066. mutant minigene expressed in neonatal rat ventricular myo- 10. Thomas D, Kiehn J, Katus HA, Karle CA. Defective protein trafficking in hERG-associated hereditary long QT syndrome (LQT2): molecular cytes leads to reduced mRNA levels by NMD, further studies mechanisms and restoration of intracellular protein processing. Car- are required to determine whether the endogenous PTC- diovasc Res. 2003;60:235–232. containing mRNA in human heart tissue is subject to NMD. 11. Delisle BP, Anson BD, Rajamani S, January CT. Biology of cardiac Verification of our findings in human heart would strengthen arrhythmias: ion channel protein trafficking. Circ Res. 2004;94: 1418–1428. the conclusion that hERG mutations that contain PTCs can 12. Li X, Xu J, Li M. The human delta1261 mutation of the HERG potassium lead to degradation of the mutant mRNA by NMD. channel results in a truncated protein that contains a subunit interaction In summary, our findings that nonsense mutations in hERG domain and decreases the channel expression. J Biol Chem. 1997;272: lead to a reduced level of mutant mRNA by NMD add to our 705–708. 13. Paulussen A, Yang P, Pangalos M, Verhasselt P, Marrannes R, Ver- understanding of the disease-causing mechanisms of hERG faille C, Vandenberk I, Crabbe R, Konings F, Luyten W, Armstrong mutations in LQT2. Thus, in studies of hERG nonsense and M. Analysis of the human KCNH2 (HERG) gene: identification and frameshift mutations, it is important to first analyze the characterization of a novel mutation Y667X associated with long QT abundance of mRNA to determine whether these PTC muta- syndrome and a non-pathological 9 bp insertion. Hum Mutat. 2000; 15:483. tions are targeted by NMD. Obviously, this important point 14. Kupershmidt S, Yang T, Chanthaphaychith S, Wang Z, Towbin JA, had been overlooked in previous studies that analyzed hERG Roden DM. Defective human ether-a-go-go-related gene trafficking PTC mutations only at the protein and functional levels. linked to an endoplasmic reticulum retention signal in the C terminus. Because PTC mutations account for Ͼ30% of LQT2 muta- J Biol Chem. 2002;277:27442–27448. 15. Gong Q, Keeney DR, Robinson JC, Zhou Z. Defective assembly and tions, the RNA surveillance imposed by NMD is of funda- trafficking of mutant HERG channels with C-terminal truncations in long mental importance in the pathogenesis of LQT2. QT syndrome. J Mol Cell Cardiol. 2004;37:1225–1233. 16. Teng S, Ma L, Dong Y, Lin C, Ye J, Bahring R, Vardanyan V, Yang Y, Acknowledgment Lin Z, Pongs O, Hui R. Clinical and electrophysiological characterization of a novel mutation R863X in HERG C-terminus associated with long QT We thank Dr Kent Thornburg for helpful comments on the syndrome. J Mol Med. 2004;82:189–196. manuscript. 17. Paulussen AD, Raes A, Jongbloed RJ, Gilissen RA, Wilde AA, Snyders DJ, Smeets HJ, Aerssens J. HERG mutation predicts short QT based on Sources of Funding channel kinetics but causes long QT by heterotetrameric trafficking defi- The present study was supported in part by NIH grant HL68854 (Dr ciency. Cardiovasc Res. 2005;67:467–475. Zhou), Deseret Foundation grant DF400 (Dr Vincent), and NIH grant 18. Choe CU, Schulze-Bahr E, Neu A, Xu J, Zhu ZI, Sauter K, Bahring R, 1UL1RRO24140–01 from the National Center for Research Priori S, Guicheney P, Monnig G, Neapolitano C, Heidemann J, Clancy Resources. CE, Pongs O, Isbrandt D. C-terminal HERG (LQT2) mutations disrupt IKr channel regulation through 14-3-3⑀. Hum Mol Genet. 2006;15: 2888–2902. Disclosures 19. Frischmeyer PA, Dietz HC. Nonsense-mediated mRNA decay in health None. and disease. Hum Mol Genet. 1999;8:1893–1900. 20. Holbrook JA, Neu-Yilik G, Hentze MW, Kulozik AE. Nonsense- References mediated decay approaches the clinic. Nat Genet. 2004;36:801–808. 1. Schwartz PJ, Periti M, Malliani A. Fundamentals of clinical cardiology: 21. Nagy E, Maquat LE. A rule for termination-codon position within intron- the long QT syndrome. Am Heart J. 1975;89:378–390. containing genes: when nonsense affects RNA abundance. Trends 2. Curran ME, Splawski I, Timothy KW, Vincent GM, Green ED, Keating Biochem Sci. 1998;23:198–199. MT. A molecular basis for cardiac arrhythmia: HERG mutations cause 22. Maquat LE. Nonsense-mediated mRNA decay: splicing, translation and long QT syndrome. Cell. 1995;80:795–803. mRNP dynamics. Nat Rev Mol Cell Biol. 2004;5:89–99. 24 Circulation July 3, 2007

23. Lykke-Andersen J, Shu MD, Steitz JA. Human Upf proteins target an 30. Kapiloff MS, Schillace RV, Westphal AM, Scott JD. mAKAP: an mRNA for nonsense-mediated decay when bound downstream of a ter- A-kinase anchoring protein targeted to the nuclear membrane of differ- mination codon. Cell. 2000;103:1121–1131. entiated myocytes. J Cell Sci. 1999;112:2725–2736. 24. Mendell JT, ap Rhys CM, Dietz HC. Related separable roles for 31. Allen-Brady K, Wong J, Camp NJ. PedGenie: an analysis approach for rent1/hUpf1 in altered splicing and decay of nonsense transcripts. genetic association testing in extended pedigrees and genealogies of Science. 2002;298:419–422. arbitrary size. BMC Bioinformatics. 2006;7:209–220. 25. Paillusson A, Hirschi N, Vallan C, Azzalin CM, Muhlemann O. A 32. Inoue K, Khajavi M, Ohyama T, Hirabayashi S, Wilson J, Reggin JD, GFP-based reporter system to monitor nonsense-mediated mRNA decay. Mancias P, Butler IJ, Wilkinson MF, Wegner M, Lupski JR. Molecular Nucleic Acids Res. 2005;33:e54. mechanism for distinct neurological phenotypes conveyed by allelic trun- 26. Zhang L, Timothy KW, Vincent GM, Lehmann MH, Fox J, Giuli LC, cating mutations. Nat Genet. 2004;36:361–369. Shen J, Splawski I, Priori S, Compton SJ, Yanowitz F, Benhorin J, Moss 33. Conti E, Izaurralde E. Nonsense-mediated mRNA decay: molecular AJ, Schwartz PJ, Robinson J, Wang Q, Zareba W, Keating M, Towbin insights and mechanistic variations across species. Curr Opin Cell Biol. JA, Napolitano C, Medina A. Spectrum of ST-T wave patterns and 2005;17:316–325. repolarization parameters in congenital long QT syndrome: ECG findings identify genotype. Circulation. 2000;102:2849–2855. 34. Dietz HC, McIntosh I, Sakai LY, Corson GM, Chalberg SC, Pyeritz RE, 27. Uejima H, Lee MP, Cui H, Feinberg AP. Hot-stop PCR: a simple and Francomano CA. Four novel FBN1 mutations: significance for mutant general assay for linear quantitation of allele ratios. Nat Genet. 2000;25: transcript level and EGF-like domain calcium binding in the pathogenesis 375–376. of Marfan syndrome. Genomics. 1993;17:468–475. 28. Kurreeman FA, Schonkeren JJ, Heijmans BT, Toes RE, Huizinga TW. 35. Moss AJ, Zareba W, Kaufman ES, Gartman E, Peterson DR, Benhorin J, Transcription of the IL10 gene reveals allele-specific regulation at the Towbin JA, Keating MT, Priori SG, Schwartz PJ, Vincent GM, Robinson mRNA level. Hum Mol Genet. 2004;13:1755–1762. JL, Andrews ML, Feng C, Hall WJ, Medina A, Zhang L, Wang Z. 29. Gong Q, Keeney DR, Molinari M, Zhou Z. Degradation of trafficking- Increased risk of arrhythmic events in long-QT syndrome with mutations defective long QT syndrome type II mutant channels by the ubiquitin- in the pore region of the human ether-a-go-go-related gene potassium proteasome pathway. J Biol Chem. 2005;280:19419–19425. channel. Circulation. 2002;105:794–799.

CLINICAL PERSPECTIVE Congenital long-QT syndrome type 2 (LQT2) is caused by mutations in human ether-a-go-go related gene (hERG), which

encodes a voltage-gated potassium channel (IKr) in the heart. The present work demonstrates that LQT2 nonsense mutations show a decrease in mutant mRNA transcripts via nonsense-mediated mRNA decay (NMD), an RNA surveillance mechanism that selectively eliminates the mRNA transcripts that contain premature termination codons. These results indicate that, contrary to intuition, the predominant consequence of hERG nonsense mutations is not the production of truncated proteins but rather the degradation of mutant mRNA by NMD. Given that nonsense and frameshift mutations account for Ͼ30% of LQT2 mutations, the RNA surveillance imposed by NMD is of fundamental importance in the pathogenesis of LQT2. Our findings have important implications for genotype-phenotype correlation investigations in LQT2. By eliminating abnormal mRNA transcripts carrying premature termination codons, NMD prevents the production of truncated proteins that could act in a dominant-negative manner. The clinical significance of NMD is the protection against severe disease phenotypes by converting the dominant-negative effect to haploinsufficiency. Thus, NMD appears to be an important factor in modifying phenotypic severity in LQT2.

Coronary Heart Disease

Coronary Artery Calcification Progression Is Heritable

Andrea E. Cassidy-Bushrow, PhD, MPH; Lawrence F. Bielak, DDS, MPH; Patrick F. Sheedy II, MD; Stephen T. Turner, MD; Iftikhar J. Kullo, MD; Xihong Lin, PhD; Patricia A. Peyser, PhD

Background—Coronary artery calcification (CAC), a marker of coronary artery atherosclerosis, can be measured accurately and noninvasively with the use of electron beam computed tomography. Serial measures of CAC quantify progression of calcified coronary artery plaque. Little is known about the role of genetic factors in progression of CAC quantity. Methods and Results—We quantified the relative contributions of measured risk factors and unmeasured genes to CAC progression measured by 2 electron beam computed tomography examinations an average of 7.3 years apart in 877 asymptomatic white adults (46% men) from 625 families in a community-based sample. After adjustment for baseline risk factors and CAC quantity, the estimated heritability of CAC progression was 0.40 (PϽ0.001). Baseline risk factors and CAC quantity explained 64% of the variation in CAC progression. Thus, genetic factors explained 14% of the variation [(100Ϫ64)ϫ(0.40)] in CAC progression. After adjustment for risk factors, the estimated genetic correlation (pleiotropy) between baseline CAC quantity and CAC progression was 0.80 and was significantly different than 0 (PϽ0.001) and 1 (Pϭ0.037). The environmental correlation between baseline CAC quantity and CAC progression was 0.42 and was significantly different than 0 (Pϭ0.006). Conclusions—Evidence was found that many but not all genetic factors influencing baseline CAC quantity also influence CAC progression. The identification of common and unique genetic influences on these traits will provide important insights into the genetic architecture of coronary artery atherosclerosis. (Circulation. 2007;116:25-31.) Key Words: atherosclerosis Ⅲ calcium Ⅲ genetics Ⅲ imaging Ⅲ population

oronary heart disease (CHD) is the leading cause of CAC quantity measured at a single time point across studies. Cdeath and disability in the United States. Despite recog- Estimated heritability (ϮSE) was 0.42Ϯ0.13 among asymp- nition of numerous factors contributing to development of tomatic white individuals,8 0.40Ϯ0.08 among sibships en- CHD, the ability to predict individuals at risk of CHD events hanced for hypertension,9 and 0.40Ϯ0.23 among individuals remains suboptimal. More than one half of CHD deaths occur from families enriched for type 2 diabetes.10 in individuals without previous symptoms.1 Traditional risk No studies have focused on estimating the genetic contri- factors (high cholesterol, high blood pressure, cigarette smok- bution to CAC progression, although the complex biology of ing, diabetes) are highly prevalent among individuals with progression of calcium appears to be “genetically directed.”11 CHD but are also prevalent in individuals without CHD The purpose of the present investigation was to estimate the 2 events. genetic contribution to variation in noninvasively measured Clinical Perspective p 31 CAC progression among an asymptomatic community-based Atherosclerosis is the primary cause of CHD. Coronary sample. Additionally, evidence for pleiotropy, or shared artery calcification (CAC), a measure of coronary atheroscle- genetic influences, between CAC quantity at baseline and rosis presence and quantity, can be detected noninvasively CAC progression was examined. and reliably with electron beam computed tomography (EBCT). CAC predicts CHD events in asymptomatic individ- Methods uals at intermediate risk on the basis of their CHD risk Study Participants factors.3,4 EBCT can be used to serially measure the progres- The Epidemiology of Coronary Artery Calcification (ECAC) study, 5,6 sion of CAC. CAC progression is associated with CHD. conducted between 1991 and 1998, examined 1240 participants aged Family history of premature CHD is associated with CAC.7 Ն20 years from the Rochester Family Heart Study12,13 and 496 Unmeasured genes contribute to interindividual variation in individuals living in the vicinity of Rochester, Minn, who were not

Received August 22, 2006; accepted May 8, 2007. From the Department of Epidemiology, University of Michigan, Ann Arbor (A.E.C.-B., L.F.B., P.A.P.); Department of Biostatistics and Research Epidemiology, Henry Ford Health System, Detroit, Mich (A.E.C.-B.); Department of Diagnostic Radiology (P.F.S.), Division of Hypertension, Department of Internal Medicine (S.T.T.), and Division of Cardiovascular Diseases (I.J.K.), Mayo Clinic and Foundation, Rochester, Minn; and Department of Biostatistics, Harvard University, Boston, Mass (X.L.). Correspondence to Patricia A. Peyser, PhD, Department of Epidemiology, University of Michigan, 611 Church St, Ann Arbor, MI 48104-3028. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.658583 25 26 Circulation July 3, 2007

pregnant or lactating and who never had coronary or noncoronary CAC quantity. CAC progression was defined as the log annual heart surgery.14,15 A total of 1155 ECAC study participants had a change in CAC area, calculated as follows: log [(difference between follow-up examination between December 2000 and February 2005. follow-up and baseline CAC areaϩ1)/time (in years) between In general, participants were invited to return for a follow-up baseline and follow-up examinations].20 If the difference between examination on the basis of age (older age first) and longer time follow-up and baseline CAC area was Ͻ0, the difference was set to since baseline examination. Study protocols were approved by the 0 (to avoid taking the log of a negative number). Mayo Clinic and University of Michigan institutional review boards, Heritability estimates (h2) were calculated for log baseline CAC and participants gave written informed consent. quantity and CAC progression with the use of a variance components One thousand fifty-five white ECAC participants had complete approach described previously8 and implemented in SOLAR.22 For CAC data at baseline and follow-up and no history of myocardial trait y, the value of y for individual i is modeled as: infarction, stroke, or positive angiogram at baseline or follow-up. ϭ␮ϩ͸␤ ϩ ϩ Individuals with missing baseline or follow-up risk factor data (1) yi jXij gi ei (nϭ68), 79 individuals aged Ͻ45 years at follow-up, and 31 ␮ individuals with outlier values (exceeding Ϯ4 SDs from sample where is the mean of y,Xij is the j-th covariate with associated ␤ mean) for risk factor data were excluded. Individuals were restricted regression coefficient j,gi is an additive genetic effect normally ␴ 2 to being aged Ն45 years at follow-up for comparability with other distributed with mean 0 and variance g , and ei is a random residual ␴ 2 CAC heritability studies8 and because CAC prevalence in younger effect normally distributed with mean 0 and variance e .Itis ␴ 2ϩ␴ 2ϭ individuals, especially women, is very low.16 The final sample size assumed that g e 1. Any nonadditive genetic and unmeasured consisted of 877 individuals (402 men). nongenetic effects (as well as measurement and random error) are ␴ 2 incorporated into ei. Heritability is estimated by g . Likelihood ratio Risk Factor Assessment tests are used to assess significance of a parameter of interest by comparing the log-likelihood of the model in which the parameter is During baseline and follow-up examination interviews, participants estimated with that of the model in which the parameter is fixed to reported current medication use, educational attainment, history of 23 smoking, physician-diagnosed hypertension, myocardial infarction, 0. angiographic evidence of a blocked coronary artery, stroke, or Heritability estimates for CAC progression were calculated as fol- diabetes. Family history of CHD was defined as self-reported lows: (1) unadjusted; (2) adjusted for age and sex; (3) adjusted for age, myocardial infarction or coronary artery revascularization in a parent sex, and the best subset of the following baseline CHD risk factors: body and/or sibling that occurred before age 60 years. Age 60 years was mass index, waist-to-hip ratio, triglycerides, LDL-C, HDL-C, fasting chosen to represent premature disease.17 Height was measured by a glucose level, SBP, DBP, presence of diabetes, presence of hyperten- wall stadiometer, weight was measured by electronic balance, and sion, college education (ie, any education beyond high school), smoking ϩ body mass index (kg/m2) was calculated. Waist circumference was history, log (pack-years smoking 1), and family history of CHD; and measured at the umbilicus, hips were measured at the level of (4) adjusted for age, sex, log baseline CAC quantity, and the best subset maximal circumference, and waist-to-hip ratio was calculated. of the CHD risk factors listed in step 3. Heritability estimates for log Standard enzymatic methods were used to measure total choles- baseline CAC quantity were calculated similarly (steps 1 to 3). Covari- 2 terol, high-density lipoprotein cholesterol (HDL-C), plasma glucose, ates were chosen for similarity to previous h studies.8 All 2-way and triglycerides after overnight fasting.13 Low-density lipoprotein interaction terms between covariates significantly associated with either 2 cholesterol (LDL-C) was calculated by the Friedewald equation.18 outcome were evaluated. The estimates of h and covariate variance Systolic blood pressure (SBP) and diastolic blood pressure (DBP) obtained were used to estimate the percentage of total variation levels were measured in the right arm with a random-zero sphyg- explained by genetic factors: [(1Ϫproportion of variance explained by momanometer (Hawksley and Sons). Three measures at least 2 covariates)ϫh2]ϫ100. ⌿ minutes apart were taken; the average of the second and third The genetic correlation ( g) between log baseline CAC quantity measurements was used. Individuals were considered hypertensive if (trait 1) and CAC progression (trait 2) was estimated to assess they reported a prior diagnosis of hypertension and use of prescrip- pleiotropic genetic effects with the use of maximum-likelihood tion antihypertensive medication or if the average SBP or DBP was estimation in SOLAR.24–26 The phenotypic correlation between the ⌿ ⌿ Ն140 mm Hg or Ն90 mm Hg, respectively. Participants were 2 traits is derived from the g, the environmental correlation ( e), considered diabetic if they reported using insulin or oral hypoglyce- and the heritabilities of the 2 traits, as follows: mic agents or if they reported a physician diagnosis of diabetes but ͓ͱ 2 2ϫ⌿ ͔ϩ͓ͱ Ϫ 2ϫͱ Ϫ 2ϫ⌿ ͔ were not currently taking a pharmacological agent to control glucose (2) h1h2 g 1 h1 1 h2 e levels. The Framingham risk equation was used to estimate the All hypothesis tests were performed with the use of likelihood- 19 10-year probability of CHD (10-year CHD risk) at baseline. 23 ⌿ ratio test statistics. The hypothesis tests of interest are whether g ⌿ ⌿ is different from 0, whether g is different from 1, and whether e Measurement of CAC ⌿ ⌿ is different from 0. If g is different from 0, the estimate of g, its ⌿ ϭ CAC was measured with an Imatron C-150 EBCT scanner (Imatron SE, and test of the hypothesis g 1 determine the magnitude of the Inc, South San Francisco, Calif). Protocols at baseline and follow-up shared genetic effects (ie, pleiotropy).27,28 If the hypothesis that were identical.20 A dual-scan approach was used beginning in 1993. ⌿ ϭ g 1 is not rejected, then all genes influencing 1 trait are assumed A scan run consisted of 40 contiguous 3-mm-thick tomographic to also influence the other trait. Rejection of the null hypothesis that slices from the root of the aorta to the apex of the heart. Scan time ⌿ ϭ e 0 indicates shared environmental components. Covariates sig- was 100 ms per tomogram. ECG gating was used, and all images nificantly associated with both traits were used to adjust both traits, were triggered at end-diastole during 2 to 4 breath-holds. A radio- whereas covariates only associated with a single trait were used to logical technologist scored the tomograms with an automated scoring adjust for that trait alone. Covariates for CAC progression were system without knowledge of other EBCT examination results for chosen from the model in which log baseline CAC quantity was not 21 the same participant. CAC was defined as a hyperattenuating focus included as a covariate. Ն within 5 mm of the midline of a coronary artery, 4 contiguous The authors had full access to and take responsibility for the Ͼ pixels in size, and having CT numbers 130 Hounsfield units integrity of the data. All authors have read and agree to the Ն 2 throughout. Areas 1mm for all CAC foci were summed to provide manuscript as written. a measure of CAC quantity. When 2 scan runs at a single examination were available, CAC quantity was based on the average. Results Statistical Analysis Mean baseline age of women was 56.4 years (range, 36.0 to Baseline CAC quantity was natural logarithm (log) transformed after 82.1 years), and that of men was 54.7 years (range, 35.7 to adding 1 to reduce nonnormality and is referred to as log baseline 79.0 years) (Table 1). Mean time between examinations was

Cassidy-Bushrow et al CAC Progression Is Heritable 27

TABLE 1. Baseline Characteristics of Study Participants estimated h2 of log baseline CAC quantity was 0.376 (Table Women Men 4). Approximately 21% of the total variation in log baseline Characteristic (nϭ475) (nϭ402) P* CAC quantity was explained by genetic factors not acting through model covariates. Age, y 56.4 (10.4) 54.7 (9.8) 0.016 2 Body mass index, kg/m 27.4 (5.5) 27.8 (3.9) 0.196 Risk Factor Associations With CAC Progression Waist-to-hip ratio 0.8 (0.09) 0.9 (0.06) Ͻ0.001 In the best-fitting model of CAC progression, baseline age Triglycerides, mmol/L 1.6 (0.8) 1.6 (0.7) 0.269 (PϽ0.001), waist-to-hip ratio (Pϭ0.024), LDL-C (PϽ0.001), LDL-C, mmol/L 3.1 (0.8) 3.3 (0.8) Ͻ0.001 log pack-years of smoking (Pϭ0.093), hypertension HDL-C, mmol/L 1.4 (0.4) 1.1 (0.3) Ͻ0.001 (PϽ0.001), and log baseline CAC quantity (PϽ0.001) were ϭ SBP, mm Hg 122.1 (17.7) 121.6 (15.7) 0.700 positively significantly associated and female sex (P 0.025) DBP, mm Hg 75.3 (8.9) 79.3 (9.7) Ͻ0.001 was negatively significantly associated with CAC progression (Table 3). These risk factors together explained Ϸ64% of Fasting glucose, mmol/L 5.0 (0.7) 5.1 (0.6) 0.037 the variation in CAC progression. The rate of change at any Log (pack-years of smokingϩ1) 0.8 (1.3) 1.6 (1.6) Ͻ0.001 given baseline age depended on CAC quantity at baseline Ͻ 10-Year CHD risk, %† 5.5 (4.5) 11.4 (7.1) 0.001 (PϽ0.001). Among those with no detectable baseline CAC, History of smoking, % 35.2 57.7 Ͻ0.001 the rate of CAC progression appears slightly higher for older Diabetes, % 1.9 2.0 0.919 individuals; at higher CAC quantities, however, the rate of Hypertension, % 34.7 34.8 0.978 CAC progression appears higher for younger individuals Statin use, % 4.6 6.7 0.180 (Figure 2). College education, % 60.0 63.9 0.233 Heritability of CAC Progression Family history of CHD, % 35.8 31.6 0.191 The estimate of CAC progression h2 was 0.782 (PϽ0.001) Data are mean (SD) unless indicated otherwise. and remained significant after adjustment for baseline age *Sex differences in participant characteristics tested by t test or ␹2 test. 2ϭ Ͻ †One man missing 10-year CHD risk because of missing smoking history. and sex (h 0.671; P 0.001) as well as after adjustment for baseline CHD risk factors significant at an ␣ Ͻ0.1 2ϭ Ͻ longer for women (7.6Ϯ3.5 years [range, 1.8 to 13.7 years]) (h 0.592; P 0.001) (Table 4). After adjustment for base- than for men (6.7Ϯ3.2 years [range, 1.8 to 13.01 years]) line age, sex, log baseline CAC quantity, waist-to-hip ratio, (Pϭ0.008). The 877 participants belonged to 625 families: LDL-C, log pack-years of smoking, hypertension, and a 453 singletons and 125 families of size 2, 28 of size 3, 10 of baseline age–by–baseline CAC quantity interaction term, the 2 Ͻ size 4, 5 of size 5, 3 of size 6, and 1 of size 7. Relationships h estimate was 0.396 (P 0.001). Baseline risk factors and consisted of sibships (384 sib pairs), 25 parent-offspring CAC quantity explained 64% of the variation in CAC pairs, and 34 avuncular pairs. progression. Thus, genetic factors explained 14% of the Table 2 presents baseline data, follow-up data, and annual variation [(100Ϫ64)ϫ(0.40)] in CAC progression. change in CAC quantity, by sex. Among women, baseline CAC prevalence was 38%, and follow-up prevalence was Evidence for Pleiotropy 58%; among men, baseline CAC prevalence was 67%, and Log baseline CAC quantity and CAC progression were signifi- ϭ follow-up prevalence was 83%. cantly correlated (Spearman correlation coefficient 0.74, PϽ0.001; Figure 3). The estimated ⌿g between log baseline Heritability of Baseline CAC Quantity CAC quantity and CAC progression was 0.80 and was statisti- The best model of log baseline CAC quantity included age cally significantly different from 0 (PϽ0.001) and 1 (Pϭ0.037) ⌿ (PϽ0.001), sex (PϽ0.001), LDL-C (Pϭ0.107), SBP (Table 5). The estimated e between log baseline CAC quantity (PϽ0.001), DBP (Pϭ0.016), log pack-years of smoking and CAC progression was 0.42 and was statistically significantly (Pϭ0.002), presence of diabetes (PϽ0.001), a positive family different than 0 (Pϭ0.006). Thus, there was evidence for shared history of CHD (Pϭ0.029), and a sex-by–LDL-C interaction environmental factors and genes for variation in log baseline term (Pϭ0.020) (Table 3). Higher values of LDL-C were CAC quantity and CAC progression; however, there also was associated with higher baseline CAC quantity among men but evidence for some nonoverlapping genes involved in each of not women (Figure 1). After adjustment for risk factors, these measures of atherosclerosis.

TABLE 2. Distribution of CAC Quantity at Baseline and Follow-Up and CAC Progression, by Sex

Women Men

CAC Measure Baseline Follow-Up Annual Change per Year* Baseline Follow-Up Annual Change per Year* CAC quantity, mm2 21.7 (79.7) ͓0, 957.6͔ 41.3 (118.5) ͓0, 1107.2͔ 3.7 (9.6) ͓Ϫ3.1, 100.8͔ 45.0 (98.3) ͓0, 877.6͔ 93.0 (153.6) ͓0, 980.3͔ 8.1 (13.3) ͓Ϫ17.5, 100.3͔ Log (CAC quantityϩ1) 1.1 (1.7) ͓0, 6.9͔ 1.8 (1.9) ͓0, 7.0͔Ϫ0.4 (1.8) ͓Ϫ2.6, 4.6͔ 2.2 (1.9) ͓0, 6.7͔ 3.1 (2.0) ͓0, 6.9͔ 0.8 (1.9) ͓Ϫ2.6, 4.6͔ Presence of any detectable CAC, % 38.1 58.3 NA 67.2 83.1 NA Data are mean (SD) ͓range͔ or percentage. NA indicates not applicable. *On scale of mm2/y, defined as (follow-upϪbaseline CAC quantity/time) (mm2/y); on log scale, defined as CAC progression: ͓(log(follow-upϪbaseline CAC quantityϩ1)/time)͔, where (follow-upϪbaseline CAC quantity)ϭ0 if (follow-upϪbaseline CAC quantity) Ͻ0.

28 Circulation July 3, 2007

TABLE 3. Baseline Risk Factors Associated With Log Baseline CAC Quantity and/or With CAC Progression

Log Baseline CAC Quantity CAC Progression

Baseline Covariate Parameter Estimate (SE) P Parameter Estimate (SE) P Age 0.075 (0.006) Ͻ0.001 0.022 (0.005) Ͻ0.001 Female sex Ϫ1.115 (0.102) Ͻ0.001 Ϫ0.225 (0.117) 0.025 Waist-to-hip ratio ⅐⅐⅐ ⅐⅐⅐ 2.089 (0.613) 0.024 LDL-C, mmol/L 0.145 (0.087) 0.107 0.226 (0.051) Ͻ0.001 SBP, mm Hg 0.023 (0.004) Ͻ0.001 ⅐⅐⅐ ⅐⅐⅐ DBP, mm Hg Ϫ0.016 (0.007) 0.016 ⅐⅐⅐ ⅐⅐⅐ Log (pack-years of smokingϩ1) 0.202 (0.033) 0.002 0.034 (0.028) 0.093 Diabetes 1.984 (0.345) Ͻ0.001 ⅐⅐⅐ ⅐⅐⅐ Hypertension ⅐⅐⅐ ⅐⅐⅐ 0.349 (0.098) Ͻ0.001 Family history of CHD 0.262 (0.109) 0.029 ⅐⅐⅐ ⅐⅐⅐ SexϫLDL-C Ϫ0.255 (0.120) 0.020 ⅐⅐⅐ ⅐⅐⅐ Log baseline CAC quantity NA NA 0.651 (0.030) Ͻ0.001 Ageϫlog baseline CAC quantity NA NA Ϫ0.009 (0.002) Ͻ0.001 Ellipses refer to variable not selected in stepwise regression procedure in SOLAR. NA indicates not applicable.

Discussion period of time (Յ3 years) in study populations with specific The present study is the first to estimate the genetic contribution characteristics (hyperlipidemic and postmenopausal women32; to CAC progression. There is evidence to suggest a strong, patients with Ն2 CAD risk factors plus moderate calcification33; shared genetic component to both CAC quantity at a single time patients with calcific aortic stenosis34). Despite a reduction in point and CAC progression, but there is also evidence suggest- LDL-C, there was no evidence of a slowing of CAC progression. ing that unique genes are involved in each of these measures of In the present study, however, baseline LDL-C was positively subclinical coronary artery atherosclerosis. Although no one has associated with increased CAC progression over a much longer identified candidate genes associated with the rate of progression follow-up period in a community-based sample. This suggests of CAC, others have identified candidate genes associated with that LDL-C levels may be important early in the development CAC progression when defined as a qualitative trait (ie, progres- and progression of atherosclerosis; our finding is consistent with sors versus nonprogressors29) in individuals with type 1 diabe- that of Kuller et al35 (1999), who showed that premenopausal tes.30,31 It would be important to investigate whether any LDL-C levels were powerful predictors of CAC measured 8 identified genes are unique for CAC progression or whether they years after menopause (11 years after LDL-C measurement). also are associated with cross-sectional measures of CAC Future work examining the effect of LDL-C reduction on CAC prevalence or quantity. progression over an extended follow-up period may be war- Several clinical trials32–34 examining LDL-C reduction ranted. Additionally, studies examining LDL-C reduction in through statin therapy and CAC progression have recently been preventing detectable CAC development among those without published. These studies evaluated change in CAC over a short detectable CAC may reveal additional insight into the pathogenesis of LDL-C–mediated CAC development and/or progression. It may also be of use to examine age- and sex-specific effects of LDL-C reduction on CAC progression.

Limitations Approximately one half of individuals did not belong to a sibship. Although these individuals contributed information to estimation of the mean and variance of the traits being investigated, as well as to relationships between covariates and traits of interest, they did not contribute information to the heritability estimation. However, our baseline h2 estimates and their SEs closely resemble those obtained by others,8–10 suggesting that our sample is sufficient for estimating h2 of CAC progression. 2 Figure 1. Relationship between LDL-C and baseline CAC quan- In the present study, h estimates may overestimate the tity depends on sex. Sex-speciﬁc predicted baseline CAC quan- genetic contribution because we have not estimated shared tities were calculated for hypothetical participants over varying environments. All siblings reported living in separate households baseline LDL-C levels, with population mean values of baseline age, SBP, and DBP, 0 pack-years of smoking, without diabetes, from one another and their parents at the time of the study. and without a family history of CHD. However, shared environments early in life may contribute to Cassidy-Bushrow et al CAC Progression Is Heritable 29

TABLE 4. Heritability Estimates for Log Baseline CAC Quantity and CAC Progression

% of Variance Covariate Explained by Trait h2 (SE) Variance* Covariates Adjusted for: Genetic Factors† Log baseline CAC quantity 0.488 (0.104) 0.00 None 48.8 0.391 (0.097) 0.35 Age, sex 25.4 0.376 (0.096) 0.43 Age, sex, LDL-C, SBP, DBP, log 21.4 (pack-years of smokingϩ1), diabetes, family history of CHD, sexϫLDL-C CAC progression 0.782 (0.101) 0.00 None 78.2 0.671 (0.108) 0.35 Age, sex 43.6 0.592 (0.109) 0.44 Age, sex, waist-to-hip ratio, LDL-C, log 33.2 (pack-years of smokingϩ1), diabetes, hypertension, family history of CHD 0.396 (0.133) 0.64 Age, sex, waist-to-hip ratio, LDL-C, log 14.3 (pack-years of smokingϩ1), hypertension, baseline CAC quantity, ageϫbaseline CAC quantity All h2 estimates were significant (PՅ0.001). *Proportion of variance explained by covariates. †Calculated as ͓(1Ϫproportion of variance explained by covariates)ϫh2͔ϫ100. the correlations for CAC quantity8 and CAC progression seen Pϭ0.011) than the remainder of the study sample. Only 28 among adult relatives. (46.2%) of these 52 participants had any detectable CAC at Our study sample was restricted to white individuals; how- follow-up examination; these 28 individuals had small quantities ever, CAC burden36 and progression37 vary across different of detectable CAC at baseline (mean, 2.7Ϯ3.1 mm2; range, 0.7 ethnic populations. Thus, future studies examining the genetic to 12.2 mm2). The negative differences between baseline and contribution to CAC progression in other ethnic groups are follow-up are likely attributable to measurement errors rather warranted. than being true regression of CAC because larger body size Participants whose follow-up CAC quantity was less than creates additional noise in CAC measurement,38,39 and Ն40% of CAC quantity at baseline (nϭ52; 5.9%) were treated as having those with less detectable CAC at follow-up compared with no change in the definition of CAC progression. The mean baseline had small CAC quantity detected at baseline and no change in this group was Ϫ1.3 mm2/y. Individuals with less detectable CAC at follow-up. Furthermore, after we repeated our detectable CAC at follow-up compared with baseline examina- analyses removing these 52 participants from the sample, our tion were younger (mean age, 52.8Ϯ11.7 versus 55.8Ϯ10.1 inferences remained the same. Thus, treatment of these partici- years; Pϭ0.042), had larger mean body mass index (30.1Ϯ5.3 pants as having no change between baseline and follow-up is versus 27.4Ϯ4.8 kg/m2; PϽ0.001), had larger mean waist-to-hip reasonable, particularly because evidence from animal studies ratio (0.89Ϯ0.09 versus 0.85Ϯ0.10; Pϭ0.018), and were less indicates that although calcium progression itself may be slowed likely to report a family history of CHD (13.5% versus 35.2%; or stopped (eg, through dietary intervention), there is no evidence suggesting that calcium deposits will exhibit a true regression in the absence of aggressive intervention.40 Although a direct relationship exists between CAC and both histological and in vivo measures of atherosclerotic plaque on a

Figure 2. Relationship between baseline age and annual change in CAC quantity depends on baseline CAC quantity. Predicted annual changes in CAC quantity were calculated over varying Figure 3. Distribution of CAC progression as a function of log baseline ages and CAC quantities for hypothetical women with baseline CAC quantity. Linear regression equation is as follows: population mean values of waist-to-hip ratio, LDL-C, 0 pack- CAC progressionϭϪ1.08ϩ0.79ϫ(log baseline CAC quantity). years of smoking, and without hypertension. PϽ0.0001, R2ϭ0.57.

30 Circulation July 3, 2007

TABLE 5. Evidence of Pleiotropy Between Log Baseline CAC 2. Greenland P, Knoll MD, Stamler J, Neaton JD, Dyer AR, Garside DB, Quantity and CAC Progression Wilson PW. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. JAMA. 2003;290:891–897. Correlation (⌿) Estimate (SE) P for ⌿ϭ0 P for ⌿ϭ1 3. Bielak LF, Rumberger JA, Sheedy PF II, Schwartz RS, Peyser PA. Probabilistic model for prediction of angiographically defined obstructive Genetic correlation (⌿ ) 0.80 (0.11) Ͻ0.001 0.037 g coronary artery disease using electron beam computed tomography Environmental correlation (⌿e) 0.42 (0.11) 0.006 NA calcium score strata. Circulation. 2000;102:380–385. 4. Budoff MJ. Atherosclerosis imaging and calcified plaque: coronary artery Log baseline CAC quantity and CAC progression were both adjusted for age, disease risk assessment. Prog Cardiovasc Dis. 2003;46:135–148. ϩ sex, LDL-C, log (pack-years of smoking 1), diabetes, and family history of 5. Raggi P, Callister TQ, Shaw LJ. Progression of coronary artery calcium ϫ CHD; log baseline CAC quantity for SBP, DBP, and sex LDL-C; and CAC and risk of first myocardial infarction in patients receiving cholesterol- progression for waist-to-hip ratio and hypertension. NA indicates not applica- lowering therapy. Arterioscler Thromb Vasc Biol. 2004;24:1272–1277. ble. 6. Raggi P, Cooil B, Shaw LJ, Aboulhson J, Takasu J, Budoff M, Callister TQ. Progression of coronary calcium on serial electron beam tomographic scanning is greater in patients with future myocardial infarction. heart-by-heart, vessel-by-vessel, and segment-by-segment ba- Am J Cardiol. 2003;92:827–829. sis,41–45 absence of detectable CAC with EBCT does not 7. Nasir K, Michos ED, Rumberger JA, Braunstein JB, Post WS, Budoff necessarily indicate an absence of coronary artery atherosclero- MJ, Blumenthal RS. Coronary artery calcification and family history of premature coronary heart disease: sibling history is more strongly asso- sis. This measure likely underestimates total atherosclerosis ciated than parental history. Circulation. 2004;110:2150–2156. quantity and progression in some individuals because CAC 8. Peyser PA, Bielak LF, Chu JS, Turner ST, Ellsworth DL, Boerwinkle E, quantity more closely represents calcified plaque burden rather Sheedy PF II. Heritability of coronary artery calcium quantity measured by electron beam computed tomography in asymptomatic adults. Circu- than atherosclerosis. lation. 2002;106:304–308. Finally, we restricted our analyses to account for baseline 9. Turner ST, Peyser PA, Kardia SL, Bielak LF, Sheedy PF II, Boerwinkle measures of risk factors only; however, change in risk factor E, de Andrade M. Genomic loci with pleiotropic effects on coronary artery calcification. Atherosclerosis. 2006;185:340–346. status over time may retard or accelerate CAC progression with 10. Wagenknecht LE, Bowden DW, Carr JJ, Langefeld CD, Freedman BI, unknown effects on estimation of the role of genetic factors. Rich SS. Familial aggregation of coronary artery calcium in families with Future work should examine time-varying covariates in CAC type 2 diabetes. Diabetes. 2001;50:861–866. progression. 11. Greenland P, Bonow RO, Brundage BH, Budoff MJ, Eisenberg MJ, Grundy SM, Lauer MS, Post WS, Raggi P, Redberg RF, Rodgers GP, Conclusion Shaw LJ, Taylor AJ, Weintraub WS, Harrington RA, Abrams J, Anderson JL, Bates ER, Grines CL, Hlatky MA, Lichtenberg RC, Lindner JR, Both individual and familial characteristics (eg, genes) are Pohost GM, Schofield RS, Shubrooks SJ Jr, Stein JH, Tracy CM, Vogel important factors in CAC progression. Importantly, there is a RA, Wesley DJ. ACCF/AHA 2007 clinical expert consensus document genetic component to CAC progression beyond that captured by on coronary artery calcium scoring by computed tomography in global baseline risk factors (including family history of CHD) and cardiovascular risk assessment and in evaluation of patients with chest pain: a report of the American College of Cardiology Foundation Clinical baseline CAC. Baseline risk factors (including family history of Expert Consensus Task Force (ACCF/AHA Writing Committee to CHD) and baseline CAC may provide useful tools for identify- Update the 2000 Expert Consensus Document on Electron Beam ing individuals at otherwise low to moderate risk of a CHD event Computed Tomography). Circulation. 2007;115:402–426. 12. Turner ST, Weidman WH, Michels VV, Reed TJ, Ormson CL, Fuller T, who may benefit from serial CAC screening for additional risk Sing CF. Distribution of sodium-lithium countertransport and blood stratification and/or primary prevention of disease. pressure in Caucasians five to eighty-nine years of age. Hypertension. Identification of specific genes associated with increased 1989;13:378–391. 13. Kottke BA, Moll PP, Michels VV, Weidman WH. Levels of lipids, CAC progression may provide insights into molecular mecha- lipoproteins, and apolipoproteins in a defined population. Mayo Clin nisms of atherosclerosis, identify new targets for therapy, and Proc. 1991;66:1198–1208. lead to blood tests for early detection of susceptible individuals 14. Maher JE, Raz JA, Bielak LF, Sheedy PF II, Schwartz RS, Peyser PA. who would benefit from early, individualized therapeutic or Potential of quantity of coronary artery calcification to identify new risk factors for asymptomatic atherosclerosis. Am J Epidemiol. 1996;144: lifestyle interventions for halting or slowing their CAC progression. 943–953. 15. Bielak LF, Sheedy PF II, Peyser PA. Coronary artery calcification measured at electron-beam CT: agreement in dual scan runs and change Sources of Funding over time. Radiology. 2001;218:224–229. This research was supported by grant R01 HL46292 from the 16. Kaufmann RB, Peyser PA, Sheedy PF II, Rumberger JA, Schwartz RS. National Institutes of Health, by a General Clinic Research Center Quantification of coronary artery calcium by electron beam computed grant from the National Institutes of Health (MO1-RR00585) tomography for determination of severity of angiographic coronary artery awarded to Mayo Clinic Rochester, and by National Human Genome disease in younger patients. J Am Coll Cardiol. 1995;25:626–632. Research Institute grant T32 HG00040. 17. Shemesh J, Koren-Morag N, Apter S, Rozenman J, Kirwan BA, Itzchak Y, Motro M. Accelerated progression of coronary calcification: four-year follow-up in patients with stable coronary artery disease. Radiology. Disclosures 2004;233:201–209. None. 18. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499–502. References 19. Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular 1. Rosamond W, Flegal K, Friday G, Furie K, Go A, Greenlund K, Haase N, disease risk profiles. Am Heart J. 1991;121:293–298. Ho M, Howard V, Kissela B, Kittner S, Lloyd-Jones D, McDermott M, 20. Cassidy AE, Bielak LF, Zhou Y, Sheedy PF II, Turner ST, Breen JF, Meigs J, Moy C, Nichol G, O’Donnell CJ, Roger V, Rumsfeld J, Sorlie Araoz PA, Kullo IJ, Lin X, Peyser PA. Progression of subclinical P, Steinberger J, Thom T, Wasserthiel-Smoller S, Hong Y. Heart disease coronary atherosclerosis: does obesity make a difference? Circulation. and stroke statistics—2007 update: a report from the American Heart 2005;111:1877–1882. Association Statistics Committee and Stroke Statistics Subcommittee. 21. Reed JE, Rumberger JA, Davitt PJ. System for quantitative analysis of Circulation. 2007;115:e69–e171. coronary calcification via electron-beam computed tomography. In: Cassidy-Bushrow et al CAC Progression Is Heritable 31

Hoffman EA, Acharya RS, eds. Medical Imaging 1994: Physiology and standard lipid-lowering treatment with atorvastatin on the progression of Function From Multidimensional Images. Proc SPIE. 1994;2168:43–53. calcified coronary atherosclerosis over 12 months: a multicenter, ran- 22. Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in domized, double-blind trial. Circulation. 2006;113:427–437. general pedigrees. Am J Hum Genet. 1998;62:1198–1211. 34. Houslay ES, Cowell SJ, Prescott RJ, Reid J, Burton J, Northridge DB, 23. Self SA, Liang KY. Asymptotic properties of maximum likelihood Boon NA, Newby DE. Progressive coronary calcification despite estimates and likelihood ratio tests under nonstandard conditions. JAm intensive lipid-lowering treatment: a randomised controlled trial. Heart. Stat Assoc. 1987;82:605–610. 2006;92:1207–1212. 24. Lange K, Boehnke M. Extensions to pedigree analysis, IV: covariance 35. Kuller LH, Matthews KA, Sutton-Tyrrell K, Edmundowicz D, Bunker components models for multivariate traits. Am J Med Genet. 1983;14: CH. Coronary and aortic calcification among women 8 years after 513–524. menopause and their premenopausal risk factors: the Healthy Women 25. Boehnke M, Moll PP, Lange K, Weidman WH, Kottke BA. Univariate Study. Arterioscler Thromb Vasc Biol. 1999;19:2189–2198. and bivariate analyses of cholesterol and triglyceride levels in pedigrees. 36. McClelland RL, Chung H, Detrano R, Post W, Kronmal RA. Distribution of Am J Med Genet. 1986;23:775–792. coronary artery calcium by race, gender, and age: results from the Multi-Ethnic 26. Hokanson JE, Langefeld CD, Mitchell BD, Lange LA, Goff DC Jr, Study of Atherosclerosis (MESA). Circulation. 2006;113:30–37. Haffner SM, Saad MF, Rotter JI. Pleiotropy and heterogeneity in the 37. Kawakubo M, LaBree L, Xiang M, Doherty TM, Wong ND, Azen S, expression of atherogenic lipoproteins: the IRAS Family Study. Hum Detrano R. Race-ethnic differences in the extent, prevalence, and pro- Hered. 2003;55:46–50. gression of coronary calcium. Ethn Dis. 2005;15:198–204. 27. Kent JW Jr., Comuzzie AG, Mahaney MC, Almasy L, Rainwater DL, 38. Hall EF. Use of EBCT in epidemiological studies: the effect of noise Int J Epidemiol VandeBerg JL, MacCluer JW, Blangero J. Intercellular adhesion and body size on coronary calcium scores. . 2005;34: 179–180. molecule-1 concentration is genetically correlated with insulin resistance, 39. Wang TJ, Larson MG, Levy D, Benjamin EJ, Kupka MJ, Manning WJ, Clouse obesity, and HDL concentration in Mexican Americans. Diabetes. 2004; ME, D’Agostino RB, Wilson PW, O’Donnell CJ. C-reactive protein is associated 53:2691–2695. with subclinical epicardial coronary calcification in men and women: the Fra- 28. Comuzzie AG, Rainwater DL, Blangero J, Mahaney MC, VandeBerg JL, mingham Heart Study. Circulation. 2002;106:1189–1191. MacCluer JW. Shared and unique genetic effects among seven HDL 40. Stary HC. The development of calcium deposits in atherosclerotic lesions phenotypes. Arterioscler Thromb Vasc Biol. 1997;17:859–864. and their persistence after lipid regression. Am J Cardiol. 2001;88: 29. Hokanson JE, MacKenzie T, Kinney G, Snell-Bergeon JK, Dabelea D, 16E–19E. Ehrlich J, Eckel RH, Rewers M. Evaluating changes in coronary artery 41. Baumgart D, Schmermund A, Goerge G, Haude M, Ge J, Adamzik M, calcium: an analytic method that accounts for interscan variability. Am J Sehnert C, Altmaier K, Groenemeyer D, Seibel R, Erbel R. Comparison Roentgenol. 2004;182:1327–1332. of electron beam computed tomography with intracoronary ultrasound 30. Kretowski A, Hokanson JE, McFann K, Kinney GL, Snell-Bergeon JK, and coronary angiography for detection of coronary atherosclerosis. JAm Maahs DM, Wadwa RP, Eckel RH, Ogden LG, Garg SK, Li J, Cheng S, Coll Cardiol. 1997;30:57–64. Erlich HA, Rewers M. The apolipoprotein A-IV Gln360His poly- 42. Kajinami K, Seki H, Takekoshi N, Mabuchi H. Coronary calcification morphism predicts progression of coronary artery calcification in patients and coronary atherosclerosis: site by site comparative morphologic study with type 1 diabetes. Diabetologia. 2006;49:1946–1954. of electron beam computed tomography and coronary angiography. JAm 31. Kretowski A, McFann K, Hokanson JE, Maahs D, Kinney G, Snell- Coll Cardiol. 1997;29:1549–1556. Bergeon JK, Wadwa RP, Eckel RH, Ogden L, Garg S, Li J, Cheng S, 43. Mautner GC, Mautner SL, Froehlich J, Feuerstein IM, Proschan MA, Erlich HA, Rewers M. Polymorphisms of the renin-angiotensin system Roberts WC, Doppman JL. Coronary artery calcification: assessment genes predict progression of subclinical coronary atherosclerosis. with electron beam CT and histomorphometric correlation. Radiology. Diabetes. 2007;56:863–871. 1994;192:619–623. 32. Raggi P, Davidson M, Callister TQ, Welty FK, Bachmann GA, Hecht H, 44. Rumberger JA, Simons DB, Fitzpatrick LA, Sheedy PF II, Schwartz RS. Rumberger JA. Aggressive versus moderate lipid-lowering therapy in hypercho- Coronary artery calcium area by electron-beam computed tomography lesterolemic postmenopausal women: Beyond Endorsed Lipid Lowering with and coronary atherosclerotic plaque area: a histopathologic correlative EBT Scanning (BELLES). Circulation. 2005;112:563–571. study. Circulation. 1995;92:2157–2162. 33. Schmermund A, Achenbach S, Budde T, Buziashvili Y, Forster A, 45. Schmermund A, Rumberger JA, Colter JF, Sheedy PF II, Schwartz RS. Friedrich G, Henein M, Kerkhoff G, Knollmann F, Kukharchuk V, Lahiri Angiographic correlates of “spotty” coronary artery calcium detected by A, Leischik R, Moshage W, Schartl M, Siffert W, Steinhagen-Thiessen E, electron-beam computed tomography in patients with normal or near- Sinitsyn V, Vogt A, Wiedeking B, Erbel R. Effect of intensive versus normal coronary angiograms. Am J Cardiol. 1998;82:508–511.

CLINICAL PERSPECTIVE Noninvasively measured progression of quantity of coronary artery calcification (CAC) provides independent information, in addition to traditional coronary heart disease risk factors, for prediction of risk of future coronary events. Little is known about factors that influence progression of CAC quantity in a community-based sample of asymptomatic adults. CAC progression over Ϸ7 years was influenced by the CAC quantity at the baseline examination as well as older age, male sex, and other traditional coronary heart disease risk factors (presence of hypertension, higher low-density lipoprotein cholesterol levels, higher waist-to-hip ratio, family history of coronary heart disease, and smoking more cigarettes). Importantly, there was evidence for a genetic component unique to CAC progression beyond genes for baseline risk factors and baseline CAC quantity. Identification of specific genes associated with increased CAC progression may provide insights into molecular mechanisms of coronary atherosclerosis, identify new targets for therapy, and lead to blood tests for early detection of susceptible individuals who would benefit from early, individualized therapeutic or lifestyle interventions for halting or slowing their CAC progression. This study identified measurable factors at a baseline examination that can be used immediately to identify asymptomatic adults likely to have faster progression of subclinical coronary atherosclerosis. Epidemiology

Association of Carotid Artery Intima-Media Thickness, Plaques, and C-Reactive Protein With Future Cardiovascular Disease and All-Cause Mortality The Cardiovascular Health Study

Jie J. Cao, MD, MPH; Alice M. Arnold, PhD; Teri A. Manolio, MD, PhD; Joseph F. Polak, MD, MPH; Bruce M. Psaty, MD, PhD; Calvin H. Hirsch, MD; Lewis H. Kuller, MD, PhD; Mary Cushman, MD, MSc

Background—Carotid atherosclerosis, measured as carotid intima-media thickness or as characteristics of plaques, has been linked to cardiovascular disease (CVD) and to C-reactive protein (CRP) levels. We investigated the relationship between carotid atherosclerosis and CRP and their joint roles in CVD prediction. Methods and Results—Of 5888 participants in the Cardiovascular Health Study, an observational study of adults aged Ն65 years, 5020 without baseline CVD were included in the analysis. They were followed up for as long as 12 years for CVD incidence and all-cause mortality after baseline ultrasound and CRP measurement. When CRP was elevated (Ͼ3 mg/L) among those with detectable atherosclerosis on ultrasound, there was a 72% (95% CI, 1.46 to 2.01) increased risk for CVD death and a 52% (95% CI, 1.37 to 1.68) increased risk for all-cause mortality. Elevated CRP in the absence of atherosclerosis did not increase CVD or all-cause mortality risk. The proportion of excess risk attributable to the interaction of high CRP and atherosclerosis was 54% for CVD death and 79% for all-cause mortality. Addition of CRP or carotid atherosclerosis to conventional risk factors modestly increased in the ability to predict CVD, as measured by the c statistic. Conclusions—In older adults, elevated CRP was associated with increased risk for CVD and all-cause mortality only in those with detectable atherosclerosis based on carotid ultrasound. Despite the significant associations of CRP and carotid atherosclerosis with CVD, these measures modestly improve the prediction of CVD outcomes after one accounts for the conventional risk factors. (Circulation. 2007;116:32-38.) Key Words: aging Ⅲ arteriosclerosis Ⅲ atherosclerosis Ⅲ cardiovascular diseases Ⅲ carotid arteries Ⅲ inflammation

oth carotid intima-media thickness (IMT) and plaques Editorial p 3 B are measures of carotid atherosclerosis. Carotid IMT Clinical Perspective p 38 has been linked to many cardiovascular outcomes, includ- 10,11 ing cerebral and coronary events.1,2 Characteristics of atherosclerosis measures such as higher carotid IMT and complex plaque,12,13 we have shown that the association of carotid plaque have been associated with stroke risk3–5 and CRP with stroke is more apparent in the presence of a higher coronary events6 in prospective studies. With the growing carotid IMT.10 Whether the association of CRP with CVD interest in cardiovascular disease (CVD) risk stratification risk is modified by the presence of carotid atherosclerosis has by combining vascular imaging with conventional risk not been explored fully. factors, it is essential to understand the relationship be- In the present study, we evaluated the hypothesis that CRP tween carotid IMT and plaque and their independent and is less predictive of CVD outcomes in the absence of combined contribution to the risk of coronary as well atherosclerosis by investigating the associations of carotid cerebrovascular events. IMT, carotid plaque, and CRP, alone and in combination, In addition to ultrasonographic measures of atherosclero- with incident myocardial infarction, stroke, CVD death, and sis, C-reactive protein (CRP) has been shown to be a risk all-cause mortality. We also examined the roles of CRP and factor for CVD.7–9 Although higher CRP is associated with carotid atherosclerosis in CVD prediction.

Received June 22, 2006; accepted April 9, 2007. From the National Heart, Lung, and Blood Institute (J.J.C.), and the National Human Genome Research Institute (T.A.M.), National Institutes of Health, Bethesda, Md; University of Washington, Seattle (A.M.A., B.M.P.); University of California at Davis (C.H.H.); New England Medical Center, Tufts University, Boston, Mass (J.F.P.); University of Pittsburgh, Pittsburgh, Pa (L.H.K.); and University of Vermont, Burlington (M.C.). Correspondence to Jie J. Cao, MD, MPH, National Heart, Lung, and Blood Institute, National Institutes of Health, 10 Center Dr, MSC 1061, Bldg 10, Room B1D-416, Bethesda, MD 20892. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.645606

32 Cao et al Carotid Atherosclerosis, CRP, and CVD 33

Methods Minimal atherosclerosis was defined as having the lowest tertile of We studied participants of the Cardiovascular Health Study (CHS), IMT and no plaque. a population-based, prospective study of men and women aged Ն65 The methods of ascertainment and classification of incident stroke years. Between 1989 and 1990, 5201 participants were enrolled from and myocardial infarction have been reported.22 Participants were Medicare eligibility lists in 4 counties: Forsyth County, North examined annually at each clinical site. In addition, telephone Carolina; Washington County, Maryland; Sacramento County, Cal- interviews were alternated with clinic visits so that contacts were ifornia; and Allegheny County, Pennsylvania. To increase their every 6 months. Follow-up was complete through June 30, 2001. representation, a second cohort of 687 black participants was Potential vascular events were validated through medical record enrolled between 1992 and 1993 with the use of similar methods. review by committees. Myocardial infarction and stroke included Details of the study design have been published.14,15 The study was incident fatal and nonfatal events. Composite CVD was defined to approved by the institutional review boards at each participating include any incident myocardial infarction, stroke, or CVD death. center. All participants gave informed consent. Of the 5888 CHS participants, 868 were excluded from analysis ϭ All participants underwent baseline clinical examinations, which because of prebaseline myocardial infarction or stroke (n 765), ϭ ϭ included medical history, physical examination, and carotid ultra- missing CRP value (n 72), or missing carotid ultrasound (n 31). sound. Blood was drawn in the morning after an overnight fast. The authors had full access to and take full responsibility for the Samples were promptly centrifuged at 3000g for 10 minutes at 4°C. integrity of the data. All authors have read and agree to the Aliquots of plasma were stored in a central laboratory at Ϫ70°C. manuscript as written. CRP was measured in all stored baseline plasma samples by a high-sensitivity immunoassay, with an interassay coefficient of Statistical Analysis variation of 6.25%.16 The diagnosis of diabetes mellitus was made Analyses were done with the use of SPSS for Windows, version following American Diabetes Association criteria as fasting glucose 11.0.1 (SPSS, Inc, Chicago, Ill) and STATA, version 9.2 (College Ն126 mg/dL or use of insulin or oral glucose-lowering agents. Station, Tex). Incidence rates of CVD were calculated by dividing Impaired fasting glucose was defined as fasting glucose Ͼ110 and the total number of events by the total person-years at risk over the Ͻ126 mg/dL. follow-up time within groups defined by the carotid IMT tertile, The carotid arteries were evaluated at baseline with high- plaque risk group, and CRP level. Pearson correlation coefficients resolution B-mode ultrasonography (model SSA-270A; Toshiba were computed to assess the linear relationships among carotid IMT, America Medical Systems, Tustin, Calif). One longitudinal image of CRP, and plaque. CRP was log-transformed when modeled contin- the common carotid artery and 3 longitudinal images of the internal uously. Hazard ratios (HRs) from multivariable Cox proportional carotid artery were acquired. The maximal IMT of the common hazards models were used to estimate the relative risks (RRs) carotid artery and of the internal carotid artery was defined as the associated with high CRP, carotid IMT tertile, and plaque charac- mean of the maximal IMT of the near and far walls on both the left teristics for CVD outcomes and all-cause mortality. The proportional and right sides. Focal plaques, when present, were included in the hazards assumption was assessed for the 3 measures of interest maximum IMT measurement. Carotid IMT was defined as a com- (IMT, plaque risk group, and high CRP) by testing each with an posite measure that combined the maximum common and internal interaction for time in a Cox model. No significant interactions with carotid wall thickness of the left and right carotid arteries after time were found. In addition, we examined Kaplan-Meier plots standardization (subtraction of the mean and division by the standard visually to look for inconsistent effects over time. Participants who deviation).17 The ultrasound reading center located in Boston, Mass, died or were lost to follow-up before the event of interest or June 30, was responsible for developing standardized protocols for both 2001, were censored at the time of death or last follow-up. Multivari- scanning and interpretation of carotid sonographic images. The able models were adjusted for age and sex and then further adjusted ultrasound protocol, including measurement and reading methods, for race, systolic and diastolic blood pressure, use of antihyperten- has been published.18 The interreader variability defined by Spear- sive medications, body mass index, smoking (never, former, cur- man correlation coefficients on maximum wall thickness of the rent), and amount smoked [ln(pack-years)], high-density lipoprotein common carotid artery was 0.91 and of the internal carotid artery and low-density lipoprotein cholesterol, and diabetes (none, im- was 0.81.18 As for the detection of any carotid lesions, including the paired fasting glucose, diabetes). All conventional risk factors were wall thickness and plaque, the ␬ statistics for intrareader and measured at the baseline examination and were imputed if missing, interreader agreement were 0.69 and 0.58 for the common carotid as previously described.23 The maximum percentage missing and artery and 0.73 and 0.65 for the internal carotid artery, respectively.19 imputed for any variable was 3.2% for pack-years of smoking. All Carotid plaque, defined by the appearance of the largest focal other variables were imputed for Ͻ0.3% of participants. Categorical lesion, was classified by surface characteristics, echogenicity, and measures were modeled with the use of indicator variables for each texture. Surface characteristics were classified as smooth, mildly level compared with the lowest level, and continuous measures were irregular (height variations of Յ0.4 mm), markedly irregular (height modeled linearly, per unit. We examined multiplicative and additive variations of Ͼ0.4 mm), and ulcerated (a discrete depression of interactions of CRP with measures of atherosclerosis from ultrasound. Ͼ2 mm in width extended into the media). Lesion echogenicity was We tested our hypothesis that CRP confers excess risk only in the characterized as hypoechoic, isoechoic, hyperechoic, or calcified. presence of atherosclerosis by stratifying on presence of atheroscle- Lesion texture was classified as homogeneous or heterogeneous. In rosis and by computing the relative excess risk, based on an additive case of multiple focal lesions, the largest lesion on each side was model,24 using the following 4-level variable: minimal atherosclero- measured.20 Participants were then classified as having no plaque, sis and CRP Յ3 mg/L, detectable atherosclerosis and CRP Յ3 mg/L, intermediate-risk plaque, and high-risk plaque. Those with no plaque minimal atherosclerosis and CRP Ͼ3 mg/L, and detectable athero- were defined as having a smooth intimal surface with no focal sclerosis and CRP Ͼ3 mL/L. On the basis of an additive model, thickening. High-risk plaque was defined as presence of markedly Rothman24 defined no interaction if the difference in risk between irregular or ulcerated surface or hypodense or heterogeneous plaques having both risk factors and having neither is equal to the sum of the that occupied Ͼ50% of the total plaque volume, those features differences in risk between each risk factor alone and neither, reportedly associated with clinical CVD.3–5,20,21 The remaining arguing that this presents the interaction in terms of the number of plaques, including hyperdense, calcified, or homogeneous plaques or excess cases, which is an appropriate scale for epidemiological those with mildly irregular surface, were defined as intermediate studies. Dividing by the risk when both risk factors are absent risk. When Ͼ1 type of plaque was detected in an individual, the produces an equality in terms of RR when there is no relative excess plaque risk was determined by the more severe type. In some risk due to CRP and atherosclerosis: RR(both)Ϫ1ϭRR(high analyses, we grouped carotid findings into binary variables: detect- CRP)Ϫ1ϩRR(atherosclerosis)Ϫ1. We used Cox proportional haz- able and minimal atherosclerosis. Detectable atherosclerosis was ards models to estimate the RR due to both risk factors and each one defined as present for participants in the upper 2 tertiles of carotid singly and computed the relative excess risk due to interaction wall thickness or in the intermediate- or high-risk plaque groups. (RERI), defined by RR(both)ϪRR(high CRP)ϪRR(atherosclero-

34 Circulation July 3, 2007

TABLE 1. Descriptive Statistics for the Cohort

Age, y 72.6 (5.5) Female, % 60 Black race, % 15 Body mass index, kg/m2 26.7 (4.7) Current smokers, % 12 Pack-years (among ever-smokers)* 27 (12 to 49) Diabetes status, % Normal 73 Impaired fasting glucose 12 Diabetic 15 Figure 1. Distribution of carotid plaque groups in carotid artery Systolic blood pressure, mm Hg 136 (21.7) IMT category with more complex plaque characteristics in the thicker carotid wall. Diastolic blood pressure, mm Hg 70.9 (11.4) Total cholesterol, mg/dL 212 (38.8) tertile (Figure 1). The majority (80.9%) of persons in the High-density lipoprotein cholesterol, mg/dL 55.1 (15.8) highest IMT tertile had high-risk plaques, with only 1.5% Low-density lipoprotein cholesterol, mg/dL 130 (35.7) having no plaque. In contrast, among those in the lowest third CRP, mg/L* 1.86 (0.94 to 3.31) of IMT, 27.4% had high-risk plaque, and 54.3% had no Common carotid IMT, mm 1.06 (0.21) plaques. The Pearson correlation coefficient between plaque Internal carotid IMT, mm 1.40 (0.55) risk group and carotid IMT was 0.51 (PϽ0.001). Plaque risk group, % The linear correlation between (ln)CRP level and carotid IMT was 0.12 and between (ln)CRP and plaque group was 0.08 (both Low 23 PϽ0.001). Within each plaque group, higher CRP was corre- Intermediate 21 lated with higher IMT (Figure 2). For example, in the High 56 intermediate-risk plaque group, the geometric mean CRP ranged Values expressed as mean (SD) unless otherwise indicated. from 1.58 to 1.85 to 2.20 mg/L across increasing tertiles of IMT. *Median (interquartile range). When the comparison was made across the plaque groups, the difference in CRP level again varied by IMT tertile. sis)ϩ1 for each outcome. Probability values and 95% CIs were computed by the delta method.25 The proportion of the disease Risk of CVD Related to Carotid IMT, Plaque related to high CRP and atherosclerosis, either singly or in combination, attributable to their interaction was calculated24 as Group, and CRP HRs increased from the lowest to the highest tertile of carotid RERI (1) RERI%ϭ ϫ100. IMT for every CVD outcome and for all-cause mortality (Table RR(both)Ϫ1 2) after adjustment for the conventional risk factors, carotid We assessed the ability of carotid atherosclerosis and CRP to plaque groups, and CRP. The highest tertile was associated with predict CVD and all-cause mortality by receiver-operating charac- an 84% increased risk of composite CVD events, 54% increased teristic (ROC) curves and by the c statistic,26 a measure equivalent to risk of all-cause mortality, and doubling of risk for CVD death. the area under the ROC curve, but allowing for time to event Compared with those with no plaque, participants with analysis. The Hosmer-Lemeshow goodness-of-fit test, which compares observed and predicted probabilities,27 was used to assess intermediate- or high-risk plaque were at increased risk of every model fit. Because the probabilities were derived from a logistic CVD outcome and of all-cause mortality after adjustment for regression analysis, we used occurrence of events through 8 years of follow-up, which was available for both cohorts. Results Among 5020 individuals in the analysis, the mean age was 72.6Ϯ5.5 years, and 60.2% were women. Other demographic data are shown in Table 1. A total of 593 myocardial infarctions, 613 strokes, 696 CVD deaths, and 1844 all-cause deaths occurred during a median follow-up time of 11 years (range, 5 days to 12 years).

Correlation of Carotid IMT Category, Plaque Groups, and CRP The frequencies of no plaque, intermediate-risk plaques, and high-risk plaques were 23%, 21.4%, and 55.6%, respectively. Carotid IMT category was related to plaque risk group, with no plaque being more frequent in persons in the lowest IMT Figure 2. Relation of geometric mean CRP (mg/L) with carotid tertile and high-risk plaque more frequent in the highest IMT IMT in tertile and plaque risk group. Cao et al Carotid Atherosclerosis, CRP, and CVD 35

TABLE 2. Hazard Ratios (95% CIs)* of CVD and All-Cause Mortality by CRP, Carotid IMT, and Plaque Groups

Carotid IMT in Tertiles Plaque Groups

CRP Ͼ3 mg/L† 1 2 3 No Plaque Intermediate Risk High Risk Event No. (nϭ1433) (nϭ1673) (nϭ1674) (nϭ1673) (nϭ1157) (nϭ1074) (nϭ2789) Myocardial infarction 595 1.33 (1.11 to 1.60) 1.00 1.41 (1.08 to 1.83) 1.80 (1.37 to 2.38) 1.00 1.41 (1.02 to 1.94) 1.46 (1.08 to 1.98) Stroke 613 1.26 (1.05 to 1.51) 1.00 1.18 (0.92 to 1.51) 1.77 (1.36 to 2.30) 1.00 1.38 (1.03 to 1.86) 1.31 (0.99 to 1.73) CVD death 696 1.50 (1.28 to 1.77) 1.00 1.23 (0.95 to 1.59) 2.15 (1.65 to 2.80) 1.00 1.50 (1.10 to 2.05) 1.43 (1.06 to 1.91) Composite CVD 1904 1.33 (1.18 to 1.50) 1.00 1.28 (1.08 to 1.52) 1.84 (1.54 to 2.20) 1.00 1.41 (1.15 to 1.72) 1.38 (1.14 to 1.67) All-cause mortality 1844 1.38 (1.25 to 1.53) 1.00 1.16 (1.01 to 1.35) 1.54 (1.32 to 1.79) 1.00 1.28 (1.08 to 1.52) 1.23 (1.04 to 1.44) *From a model that included age, sex, race, systolic and diastolic blood pressure, use of antihypertensive medications, body mass index, smoking (never, former, current), and amount smoked (in pack-years), high-density lipoprotein and low-density lipoprotein cholesterol, diabetes (none, impaired fasting glucose, diabetes), CRP, plaque risk group, and carotid wall thickness. †Compared with CRP Յ3 mg/L.

CVD risk factors, carotid IMT, and CRP (Table 2). Compared plus 19.9 for the combination of atherosclerosis and high CRP). with those with no plaque, the HRs (95% CI) of composite CVD The observed incidence rate for participants with both risk were 1.86 (1.55 to 2.23) and 2.09 (1.78 to 2.46) for intermediate- factors was 46.5/1000 person-years, suggestive of an excess and high-risk plaques, respectively, after adjustment for age and additive risk due to the interaction of CRP and atherosclerosis. gender only but fell to 1.46 (1.21 to 1.77) and 1.42 (1.20 to 1.70), The total excess additive risk due to CRP, atherosclerosis, and respectively, after further adjustment for carotid IMT. Additional their interaction was (46.5Ϫ13.7)ϭ32.8/1000 person-years, and adjustment for conventional risk factors and CRP only slightly 39% of that excess risk [(32.8Ϫ19.9)/32.8] was due to the attenuated the association, with HRs of 1.41 (1.15 to 1.72) and interaction of CRP and atherosclerosis. This excess risk rose to 1.38 (1.14 to 1.67), respectively. Results were similar for individual CVD outcomes and for all-cause mortality (Table 2). Elevated CRP (Ͼ3 mg/L) was associated with increased risk of every outcome compared with CRP Յ3 mg/L in the multivariable-adjusted model. The magnitude of association ranged from a 26% to a 50% increased risk (Table 2).

Cardiovascular Risk Assessment by Detectable Carotid Atherosclerosis and CRP In multivariable Cox models stratified by amount of atherosclerosis (detectable versus minimally detectable), elevated CRP conferred no increased hazard of composite CVD events, CVD death, or all-cause mortality in individuals with minimally detectable atherosclerosis, with HRs of 1.05 (95% CI, 0.70 to 1.56), 1.14 (0.60 to 2.14), and 0.87 (0.62 to 1.23), respectively. In contrast, the HRs for elevated CRP were significant in those with detectable atherosclerosis: 1.45 (1.29 to 1.62) for composite CVD events, 1.72 (1.46 to 2.01) for CVD death, and 1.52 (1.37 to 1.68) for all-cause mortality. A significant multiplicative interaction between CRP and presence of atherosclerosis was observed for all-cause mortality. The cumulative event rates for composite CVD and all-cause mortality are shown in Figure 3. The increased rates associated with CRP in the presence of atherosclerosis indicated the possibility of an additive interaction. This finding was consistent in individual CVD outcome (data not shown). For example, the incidence of composite CVD in participants with minimal atherosclerosis and CRP Յ3 mg/L was 13.7/1000 person-years. Among participants with detectable atherosclerosis and CRP Յ3 mg/L, it was 32.9/1000 person-years (19.2/1000 person-years higher than the baseline rate), and with minimal atherosclerosis and CRP Ͼ3 mg/L, it was 14.4/1000 person-years (0.7 person- years higher than the baseline rate). If an additive model held, we Figure 3. Kaplan-Meier plots of cumulative cardiovascular events (A) and all-cause mortality (B) over 12-year follow-up would expect the incidence rate for participants with both risk stratiﬁed by carotid atherosclerosis and CRP level (low level Յ3 factors to be 33.6/1000 person-years (the baseline rate of 13.7 mg/L vs high level Ͼ3 mg/L). athero indicates atherosclerosis. 36 Circulation July 3, 2007

TABLE 3. Hazard Ratios and Relative Excess Risk of Cardiovascular Outcomes With and Without Detectable Atherosclerosis and Elevated CRP

Minimal Atherosclerosis Detectable Atherosclerosis Relative Excess Risk

Adjusted Relative CRP Յ3 mg/L CRP Ͼ3 mg/L CRP Յ3 mg/L CRP Ͼ3 mg/L Excess Risk % Attributable Event No. (nϭ686) (nϭ223) (nϭ2901) (nϭ1210) (95% CI)* to Interaction† P Myocardial infarction 595 1.00 1.16 2.24 3.52 0.65 (Ϫ0.14 to 1.43) 39% 0.11 Stroke 613 1.00 1.04 1.80 2.43 0.51 (Ϫ0.10 to 1.12) 49% 0.10 CVD death 696 1.00 1.16 2.22 4.06 1.14 (0.42 to 1.86) 54% 0.002 Composite CVD 1904 1.00 1.08 1.99 3.06 0.70 (0.26 to 1.14) 50% 0.002 All-cause mortality 1844 1.00 0.88 1.47 2.36 0.79 (0.47 to 1.12) 79% Ͻ0.001 *From a model that included age, sex, race, systolic and diastolic blood pressure, use of antihypertensive medications, body mass index, smoking (never, former, current), and amount smoked (in pack-years), high-density lipoprotein and low-density lipoprotein cholesterol, diabetes (none, impaired fasting glucose, diabetes), CRP, plaque risk group, and carotid wall thickness. †Proportion of disease related to high CRP and atherosclerosis, either singly or together, that is attributable to their interaction, from the multivariable models.

50% after adjustment for CVD risk factors (Table 3). The conventional risk factors resulted in a modest increase in the adjusted excess risk attributable to interaction was 54% for CVD ability to predict CVD, as measured by the c statistic. death and 79% for all-cause mortality. Carotid IMT and plaques are both measures of atherosclerosis, perhaps having different attributes or risk associations but still CVD Risk Prediction by Carotid Atherosclerosis closely related.28,29 Common and internal carotid IMT can be and CRP viewed as an estimate of atherosclerosis quantity. Sonographic CVD risk prediction was compared with the use of c statistics characterization of carotid plaque can be considered a measure of based on models with conventional risk factors alone and atherosclerosis quality. Both indices are associated with CVD risk with the sequential addition of CRP Ͼ3 mg/L, carotid IMT, factors and outcomes. High-risk plaques as defined here were more and carotid plaque (Table 4). c Statistics increased only common in those with thicker IMT, thus linking atherosclerosis modestly with each additional risk factor. This observation quality with quantity. The high-risk plaque group had a higher risk was consistent for every CVD outcome and for all-cause of CVD outcomes in age- and gender-adjusted analyses than the mortality. The final models had excellent fit on the basis of intermediate-risk plaque group, but this was significantly attenuated the Hosmer-Lemeshow goodness-of-fit test (PՆ0.28) except after carotid IMT was taken into account. The definition of high-risk for the stroke outcome (Pϭ0.001). As shown in Figure 4, plaque in the present study was based on features previously ROC curves for composite CVD (Figure 4A) and all-cause demonstrated to be associated with stroke risk, and high-risk plaque mortality (Figure 4B) overlapped for models with conven- was common in this older cohort. That the RR of CVD associated tional risk factors alone and with the addition of CRP Ͼ3 with high-risk plaque was comparable to that of intermediate-risk mg/L and the further addition of detectable atherosclerosis. plaque after accounting for wall thickness suggests that ultrasound definition of high-risk or vulnerable plaque can be challenging. We Discussion suggest that future research on plaque quality should evaluate In the present large cohort study, we demonstrated that elevated atherosclerosis quantity when assessing risk of CVD. CRP, carotid IMT, and carotid plaque were all correlated with CRP-related risk of CVD and all-cause mortality differed by one another, yet each remained a significant risk factor for CVD the severity of atherosclerosis in this cohort of older adults. outcomes and all-cause mortality in the presence of the others. Elevated CRP was not associated with increased risk of CVD or Furthermore, elevated CRP was associated with increased CVD all-cause mortality in the group with minimal atherosclerosis, an and all-cause mortality risk only in those with detectable observation that was consistent with our previous report on atherosclerosis. Addition of CRP or carotid atherosclerosis to stroke risk from the present study.10 However, there was signif-

TABLE 4. c Statistics for Models of Cardiovascular Outcomes and All-Cause Mortality With Conventional Risk Factors and Additionally With Elevated CRP, Carotid IMT, and Carotid Plaque

Outcome Covariates* With CRP Ն3 mg/L With Carotid Tertile With Plaque Group Myocardial infarction 0.6799 0.6829 0.6971 0.6981 Stroke 0.6856 0.6869 0.6984 0.6994 CVD death 0.7424 0.7485 0.7626 0.7632 Composite CVD 0.6840 0.6867 0.7009 0.7017 All-cause mortality 0.7151 0.7188 0.7247 0.7252 *Covariates included age, gender, race, body mass index, smoking status, pack-years of smoking, diabetes, systolic and diastolic blood pressure, total cholesterol, and high-density lipoprotein and low-density lipoprotein cholesterol. Cao et al Carotid Atherosclerosis, CRP, and CVD 37

A prediction by adding CRP to the conventional risk factors, as shown recently by Bos et al34 and Wang et al.35 We expanded

1.00 our observation to the detection of carotid atherosclerosis that identifies a population at risk for CVD outcomes but does not seem to significantly increase the ability to predict a CVD event 0.75 for an individual patient, as demonstrated by the modest incre- ment in c statistics over conventional CVD risk factors. Similar

0.50 findings have been demonstrated previously,36,37 although it is Sensitivity still debatable whether ROC curve or c statistics is the best way to assess the power of risk prediction for a given risk factor.38 0.25 We recognize the limitations of the present study. The definition of high-risk plaque was based on published data 0.00 0.00 0.25 0.50 0.75 1.00 linking certain plaque characteristics with clinical events, and by 1-Specificity this definition 53% of participants had high-risk plaques. This B classification was designed to provide a model to study the interaction of carotid IMT and plaque characteristics, and therefore we caution against the clinical use of this approach. In the 1.00 CHS, reproducibility of assessing plaque characteristics by ultrasound was only moderate3 and would need improvement for

0.75 routine clinical application. Finally, a single measure of CRP was used, which may be subject to error. To summarize, carotid IMT, plaque, and elevated CRP each 0.50

Sensitivity independently contributed to the risk of CVD and all-cause mortality in models that included all 3 measures. However, 0.25 elevated CRP was associated with CVD events and all-cause mortality only in those with detectable atherosclerosis. Addition

0.00 of CRP or carotid atherosclerosis to conventional risk factors 0.00 0.25 0.50 0.75 1.00 1-Specificity resulted in a modest increase in the ability to predict CVD on the basis of ROC analysis. Figure 4. ROC curves for composite cardiovascular outcomes (A) and for all-cause mortality (B) during 12-year follow-up. The curves are based on models of the risk prediction with conven- Acknowledgments tional risk factors with or without CRP Ͼ3 mg/L and with or A full list of participating CHS investigators and institutions can be without detectable carotid atherosclerosis. In A, the areas under found at http://www.chs-nhlbi.org. the ROC curves are 0.6942, 0.6963, and 0.7086 for models with cardiovascular risk factors only, with the addition of CRP Ͼ3 mg/L, and with the further addition of carotid atherosclerosis, Sources of Funding respectively. In B, the areas under the ROC curves are 0.7508, This research was supported by contracts N01-HC-85079 through 0.7543, and 0.7582 for the same 3 models as in A, respectively. N01-HC-85086, N01-HC-35129, and N01 HC-15103 from the Dotted lines indicate CVD risk factors; dashed lines, plus CRP; National Heart, Lung, and Blood Institute. and solid lines, plus atherosclerosis. Disclosures icant excess additive risk when CRP was elevated in individuals None. with detectable atherosclerosis. This finding supports a complex relationship among inflammation, subclinical atherosclerosis, References and clinical CVD.30 Determining a patient’s risk for CVD events 1. O’Leary DH, Polak JF, Kronmal RA, Manolio TA, Burke GL, Wolfson or all-cause mortality on the basis of the level of CRP may thus SK Jr. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. N Engl J Med. 1999; be clinically challenging if CRP is used in low-risk populations 340:14–22. in whom atherosclerosis burden might be small. This conclusion 2. Bots ML, Hoes AW, Koudstaal PJ, Hofman A, Grobbee DE. Common is in accord with recent findings in a population of young carotid intima-media thickness and risk of stroke and myocardial women.31 Further research is needed in this area. infarction: the Rotterdam Study. Circulation. 1997;96:1432–1437. 3. Polak JF, Shemanski L, O’Leary DH, Lefkowitz D, Price TR, Savage PJ, In the present study, the increased rates associated with CRP Brant WE, Reid C. Hypoechoic plaque at US of the carotid artery: an only in the presence of atherosclerosis indicated the possibility independent risk factor for incident stroke in adults aged 65 years or of an additive interaction. We demonstrated an excess risk of older: Cardiovascular Health Study. Radiology. 1998;208:649–654. CVD and all-cause mortality as a result of the additive interac- 4. Mathiesen EB, Bonaa KH, Joakimsen O. Echolucent plaques are associated with high risk of ischemic cerebrovascular events in carotid ste- tion of elevated CRP and detectable atherosclerosis. Although nosis: the Tromso Study. Circulation. 2001;103:2171–2175. most atherosclerosis epidemiology studies use multiplicative 5. Gronholdt ML, Nordestgaard BG, Schroeder TV, Vorstrup S, Sillesen H. interaction to test effect modification, there are times when Ultrasonic echolucent carotid plaques predict future strokes. Circulation. additive effects may reflect the underlying mechanism, as 2001;104:68–73. 6. Belcaro G, Nicolaides AN, Ramaswami G, Cesarone MR, De SM, evident by our data and the data of others.32,33 Incandela L, Ferrari P, Geroulakos G, Barsotti A, Griffin M, Dhanjil S, Despite the significant association between CRP and CVD Sabetai M, Bucci M, Martines G. Carotid and femoral ultrasound mor- outcomes, only modest improvement is made in CVD risk phology screening and cardiovascular events in low risk subjects: a 38 Circulation July 3, 2007

10-year follow-up study (the CAFES-CAVE Study(1)). Atherosclerosis. carotid artery atherosclerosis in the elderly: relationship of disease 2001;156:379–387. severity to stroke and transient ischemic attack. Radiology. 1993;188: 7. Haverkate F, Thompson SG, Pyke SD, Gallimore JR, Pepys MB; 363–370. European Concerted Action on Thrombosis and Disabilities Angina 21. Kitamura A, Iso H, Imano H, Ohira T, Okada T, Sato S, Kiyama M, Pectoris Study Group. Production of C-reactive protein and risk of Tanigawa T, Yamagishi K, Shimamoto T. Carotid intima-media thickness coronary events in stable and unstable angina. Lancet. 1997;349: and plaque characteristics as a risk factor for stroke in Japanese elderly 462–466. men. Stroke. 2004;35:2788–2794. 8. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. 22. Psaty BM, Kuller LH, Bild D, Burke GL, Kittner SJ, Mittelmark M, Price Inflammation, aspirin, and the risk of cardiovascular disease in apparently TR, Rautaharju PM, Robbins J. Methods of assessing prevalent cardio- healthy men. N Engl J Med. 1997;336:973–979. vascular disease in the Cardiovascular Health Study. Ann Epidemiol. 9. Ridker PM. Role of inflammatory biomarkers in prediction of coronary 1995;5:270–277. heart disease. Lancet. 2001;358:946–948. 23. Arnold AM, Kronmal RA. Multiple imputation of baseline data in the 10. Cao JJ, Thach C, Manolio TA, Psaty BM, Kuller LH, Chaves PH, Polak Cardiovascular Health Study. Am J Epidemiol. 2003;157:74–84. JF, Sutton-Tyrrell K, Herrington DM, Price TR, Cushman M. C-reactive 24. Rothman KJ. Modern Epidemiology. Boston, Mass: Little, Brown; 1986. protein, carotid intima-media thickness, and incidence of ischemic stroke 25. Hosmer D, Lemeshow S. Confidence interval estimation of interaction. in the elderly: the Cardiovascular Health Study. Circulation. 2005;108: Epidemiology. 1992;3:452–456. 166–170. 26. Harrell FE Jr. Regression Modeling Strategies. New York, NY: Springer; 11. Tracy RP, Psaty BM, Macy EM, Bovill EG, Cushman M, Cornell ES, 2001. Kuller LH. Lifetime smoking exposure affects the association of 27. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: C-reactive protein with cardiovascular disease risk factors and subclinical Wiley & Sons Inc; 2000. disease in healthy elderly subjects. Arterioscler Thromb Vasc Biol. 1997; 28. Bonithon-Kopp C, Touboul PJ, Berr C, Leroux C, Mainard F, Courbon D, Ducimetiere P. Relation of intima-media thickness to atherosclerotic 17:2167–2176. plaques in carotid arteries: the Vascular Aging (EVA) Study. Arterioscler 12. Lombardo A, Biasucci LM, Lanza GA, Coli S, Silvestri P, Cianflone D, Thromb Vasc Biol. 1996;16:310–316. Liuzzo G, Burzotta F, Crea F, Maseri A. Inflammation as a possible link 29. Homma S, Hirose N, Ishida H, Ishii T, Araki G. Carotid plaque and between coronary and carotid plaque instability. Circulation. 2004;109: intima-media thickness assessed by B-mode ultrasonography in subjects 3158–3163. ranging from young adults to centenarians. Stroke. 2001;32:830–835. 13. Avanzas P, Arroyo-Espliguero R, Cosin-Sales J, Quiles J, Zouridakis E, 30. Fuster V, Moreno PR, Fayad ZA, Corti R, Badimon JJ. Atherothrombosis Kaski JC. Multiple complex stenoses, high neutrophil count and and high-risk plaque, part I: evolving concepts. J Am Coll Cardiol. C-reactive protein levels in patients with chronic stable angina. Athero- 2005;46:937–954. sclerosis. 2004;175:151–157. 31. Cook NR, Buring JE, Ridker PM. The effect of including C-reactive 14. Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, protein in cardiovascular risk prediction models for women. Ann Intern Kuller LH, Manolio TA, Mittelmark MB, Newman A, O’Leary D, Psaty Med. 2006;145:21–29. B, Rautaharju P, Tracy R. The Cardiovascular Health Study: design and 32. Li R, Chambless L. Test for additive interaction in proportional hazards rationale. Ann Epidemiol. 1991;1:263–276. models. Ann Epidemiol. 2007;17:227–236. 15. Tell GS, Fried LP, Hermanson BH, Manolio TA, Newman AB, Borhani 33. Schwartz SW, Carlucci C, Chambless LE, Rosamond WD. Synergism NO. Recruitment of adults 65 years and older as participants in the between smoking and vital exhaustion in the risk of ischemic stroke: Cardiovascular Health Study. Ann Epidemiol. 1993;3:358–366. evidence from the ARIC study. Ann Epidemiol. 2004;14:416–424. 16. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of 34. Bos MJ, Schipper CM, Koudstaal PJ, Witteman JC, Hofman A, Breteler C-reactive protein in healthy subjects: implication for reference intervals MM. High serum C-reactive protein level is not an independent predictor and epidemiological applications. Clin Chem. 1997;43:52–58. for stroke: the Rotterdam Study. Circulation. 2006;114:1591–1598. 17. O’Leary DH, Polak JF, Kronmal RA, Savage PJ, Borhani NO, Kittner SJ, 35. Wang TJ, Gona P, Larson MG, Tofler GH, Levy D, Newton-Cheh C, Tracy RP, Gardin JM, Price TR, Furberg CD; Cardiovascular Health Jacques PF, Rifai N, Selhub J, Robins SJ, Benjamin EJ, D’Agostino RB, Study Collaborative Research Group. Thickening of the carotid wall: a Vasan RS. Multiple biomarkers for the prediction of first major cardio- marker for atherosclerosis in the elderly? Stroke. 1996;27:224–231. vascular events and death. N Engl J Med. 2006;355:2631–2639. 18. O’Leary DH, Polak JF, Wolfson SK, Bond MG, Bommer W, Sheth S, 36. Greenland P, O’Malley PG. What is a new prediction marker useful? Psaty BM, Sharrett AR, Manolio TA. Use of sonography to evaluate Arch Intern Med. 2005;165:2454–2456. carotid atherosclerosis in the elderly. Stroke. 1991;22:1155–1163. 37. Folsom AR, Chambless LE, Ballantyne CM, Coresh J, Heiss G, Wu KK, 19. O’Leary DH, Bryan FA, Goodison MW, Rifkin MD, Gramiak R, Ball M, Boerwinkle E, Mosley TH Jr, Sorlie P, Diao G, Sharrett AR. An Bond MG, Dunn RA, Goldberg BB, Toole JF. Measurement variability of assessment of incremental coronary risk prediction using C-reactive carotid atherosclerosis: real-time (B-mode) ultrasonography and protein and other novel risk markers: the Atherosclerosis Risk in Com- angiography. Stroke. 1987;18:1011–1017. munities Study. Arch Intern Med. 2006;166:1368–1373. 20. Polak JF, O’Leary DH, Kronmal RA, Wolfson SK, Bond MG, Tracy RP, 38. Cook NR. Use and misuse of the receiver operating characteristic curve Gardin JM, Kittner SJ, Price TR, Savage PJ. Sonographic evaluation of in risk prediction. Circulation. 2007;115:928–935.

CLINICAL PERSPECTIVE Ultrasound-determined carotid intima-media thickness and presence of plaques are measures of carotid atherosclerosis. Higher carotid intima-media thickness is a risk marker for future stroke and coronary heart disease. Characteristics of carotid plaque have also been associated with vascular risk. C-reactive protein (CRP) is also a risk factor for cardiovascular disease. In the present study we examined the relationship of carotid atherosclerosis and CRP and their joint roles in cardiovascular risk prediction in 5022 elderly individuals (Ն65 years) who were the participants of the Cardiovascular Health Study. After up to 12 years of follow-up, we found that when CRP was elevated (Ͼ3 mg/L) among those with detectable atherosclerosis on ultrasound, there was a significantly increased risk for cardiovascular disease and all-cause mortality. Elevated CRP in the absence of atherosclerosis did not increase cardiovascular or all-cause mortality risk. We concluded that determining a patient’s risk for cardiovascular events or all-cause mortality on the basis of the level of CRP may be clinically challenging if CRP is used in low-risk populations in whom atherosclerosis burden might be small.

Abdominal Visceral and Subcutaneous Adipose Tissue Compartments Association With Metabolic Risk Factors in the Framingham Heart Study

Caroline S. Fox, MD, MPH; Joseph M. Massaro, PhD; Udo Hoffmann, MD, MPH; Karla M. Pou, MD; Pal Maurovich-Horvat, MD; Chun-Yu Liu, PhD; Ramachandran S. Vasan, MD; Joanne M. Murabito, MD, ScM; James B. Meigs, MD, MPH; L. Adrienne Cupples, PhD; Ralph B. D’Agostino, Sr, PhD; Christopher J. O’Donnell, MD, MPH

Background—Visceral adipose tissue (VAT) compartments may confer increased metabolic risk. The incremental utility of measuring both visceral and subcutaneous abdominal adipose tissue (SAT) in association with metabolic risk factors and underlying heritability has not been well described in a population-based setting. Methods and Results—Participants (nϭ3001) were drawn from the Framingham Heart Study (48% women; mean age, 50 years), were free of clinical cardiovascular disease, and underwent multidetector computed tomography assessment of SAT and VAT volumes between 2002 and 2005. Metabolic risk factors were examined in relation to increments of SAT and VAT after multivariable adjustment. Heritability was calculated using variance-components analysis. Among both women and men, SAT and VAT were significantly associated with blood pressure, fasting plasma glucose, triglycerides, and high-density lipoprotein cholesterol and with increased odds of hypertension, impaired fasting glucose, diabetes mellitus, and metabolic syndrome (P range Ͻ0.01). In women, relations between VAT and risk factors were consistently stronger than in men. However, VAT was more strongly correlated with most metabolic risk factors than was SAT. For example, among women and men, both SAT and VAT were associated with increased odds of metabolic syndrome. In women, the odds ratio (OR) of metabolic syndrome per 1–standard deviation increase in VAT (OR, 4.7) was stronger than that for SAT (OR, 3.0; P for difference between SAT and VAT Ͻ0.0001); similar differences were noted for men (OR for VAT, 4.2; OR for SAT, 2.5). Furthermore, VAT but not SAT contributed significantly to risk factor variation after adjustment for body mass index and waist circumference (P Յ0.01). Among overweight and obese individuals, the prevalence of hypertension, impaired fasting glucose, and metabolic syndrome increased linearly and significantly across increasing VAT quartiles. Heritability values for SAT and VAT were 57% and 36%, respectively. Conclusions—Although both SAT and VAT are correlated with metabolic risk factors, VAT remains more strongly associated with an adverse metabolic risk profile even after accounting for standard anthropometric indexes. Our findings are consistent with the hypothesized role of visceral fat as a unique, pathogenic fat depot. Measurement of VAT may provide a more complete understanding of metabolic risk associated with variation in fat distribution. (Circulation. 2007;116:39-48.) Key Words: abdominal fat Ⅲ diabetes mellitus Ⅲ epidemiology Ⅲ hypertension Ⅲ intra-abdominal fat Ⅲ metabolic syndrome X Ⅲ obesity

ardiovascular disease (CVD) is the leading cause of Clinical Perspective p 48 Cmorbidity and mortality in the United States, affecting roughly 70 million people and accounting for nearly 1 million past 50 years,2 but recent data suggest that the increasing deaths per year.1 Improvements in CVD risk factor profiles prevalence of obesity may have slowed this rate of decline.3 have led to significant reductions in death from CVD over the Rates of overweight and obesity continue to increase,4–6

Received November 9, 2006; accepted April 18, 2007. From the National Heart, Lung and Blood Institute’s Framingham Heart Study (C.S.F., C.J.O.), Framingham, Mass; Division of Endocrinology, Metabolism, and Diabetes, Department of Medicine (C.S.F., K.M.P.), Brigham and Women’s Hospital and Harvard Medical School, Boston, Mass; Boston University, Department of Mathematics (J.M.M.; R.B.D.) and School of Public Health, Division of Biostatistics (C.-Y.L., L.A.C.), Boston, Mass; Radiology Department (U.H.) and Department of Medicine (J.B.M., C.J.O.), Massachusetts General Hospital, Harvard Medical School, Boston; Semmelweis University (P.M.-H.), Budapest, Hungary; and Boston University School of Medicine (R.S.V.), Boston, Mass. Guest Editor for this article was Robert H. Eckel, MD. The online-only Data Supplement, which consists of a table, is available with this article at http://circ.ahajournals.org/cgi/ content/full/CIRCULATIONAHA.106.675355/DC1. Correspondence to Caroline S. Fox, MD, MPH, 73 Mt Wayte Ave, Ste 2, Framingham, MA 01702. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.675355 40 Circulation July 3, 2007 which suggests that the full impact of the obesity epidemic have been described elsewhere.40,41 Beginning in 2002, 4095 Third has yet to be realized. Generation Study participants, who had at least 1 parent in the Obesity, defined by a body mass index (BMI) of at least 30 offspring cohort, were enrolled in the Framingham Heart Study and 2 underwent standard clinic examinations. The standard clinic exam- kg/m , is a risk factor for multiple CVD risk factors, including ination included a physician interview, a physical examination, and hypertension, dyslipidemia, diabetes, and the metabolic syn- laboratory tests. For the present study, the study sample consisted of drome (MetS). BMI is a useful indicator of overall adiposity, Offspring and Third Generation Study participants who were part of and recent National Health and Nutrition Examination Survey the multidetector CT substudy. data demonstrate that two thirds of the US population is either Between June 2002 and April 2005, 3529 participants (2111 third generation, 1418 offspring participants) underwent multidetector CT 6 overweight or obese. However, different fat compartments assessment of coronary and aortic calcium. Inclusion in this study may be associated with differential metabolic risk.7 In par- was weighted toward participants from larger Framingham Heart ticular, the visceral adipose tissue (VAT) compartment may Study families and those who resided in the greater New England be a unique pathogenic fat depot.8–10 VAT has been termed area. Overall, 755 families were included in our analysis. Men had to Ն Ն an endocrine organ, in part because it secretes adipocytokines be 35 years of age; women had to be 40 years of age and not pregnant; and all participants had to weigh Ͻ350 pounds. Of the and other vasoactive substances that can influence the risk of participants, 433 (222 offspring and 211 third generation) were developing metabolic traits.10–16 imaged as participants in an ancillary study using an identical Waist circumference (WC) is an easily obtainable but imaging protocol, the National Heart, Lung, and Blood’s Family imprecise measure of abdominal adiposity17 because it is a Heart Study.42 Of the total 3529 subjects imaged, 3394 had inter- function of both the subcutaneous adipose tissue (SAT) and pretable CT measures; of those, 3329 had both SAT and VAT measured; of those, 3124 of them were free of CVD; of those, 3102 VAT compartments. Therefore, assessment of VAT requires attended a contemporaneous examination; and of those, 3001 had a imaging with radiographic techniques such as computed complete covariate profile. Thus, the overall sample size for analysis tomography (CT) or magnetic resonance imaging (MRI). is 3001. Available studies report relations of greater SAT and VAT The study was approved by the institutional review boards of the with a higher prevalence of impaired fasting glucose,9,18 Boston University Medical Center and Massachusetts General Hos- pital. All subjects provided written informed consent. diabetes,9,10,19 insulin resistance,9,20,21 hypertension,22–24 lipids,25–29 MetS,30–32 and risk factor clustering.26 However, Volumetric Adipose Tissue Imaging current imaging studies evaluating SAT and VAT are limited Subjects underwent 8-slice multidetector CT imaging of the abdo- to small, referral-based samples often enriched for adiposity- men in a supine position as previously described (LightSpeed Ultra, related traits.23,25–27,32–34 Furthermore, study samples have General Electric, Milwaukee, Wis).39 Briefly, 25 contiguous 5-mm- often been limited to either women or men, precluding the study thick slices (120 kVp; 400 mA; gantry rotation time, 500 ms; table feed, 3:1) were acquired covering 125 mm above the level of S1. of sex differences.22,23,25–28,34–36 Some studies have focused on Japanese Americans or Southeast Asians,19,22,26,29,35,37 ethnic Abdominal Adipose Tissue Measurements groups with more visceral fat than expected for a given overall SAT and VAT volumes were assessed (Aquarius 3D Workstation, BMI.38 We recently demonstrated that multidetector CT permits TeraRecon Inc, San Mateo, Calif). To identify pixels containing fat, highly reproducible volumetric measurements of both SAT and an image display window width of Ϫ195 to Ϫ45 Hounsfield units Ϫ VAT.39 In addition, this initial observation suggested that volu- (HU) and a window center of 120 HU were used. The abdominal muscular wall separating the visceral from the subcutaneous com- metric fat measurements, as opposed to previous studies using partment was traced manually. Interreader reproducibility was as- single-slice methodology, can accurately characterize the hetero- sessed by 2 independent readers measuring VAT and SAT on a geneity of abdominal fat distribution between individuals and subset of 100 randomly selected participants.39 Interclass correla- the differences in fat distribution with age and between women tions for interreader comparisons were 0.992 for VAT and 0.997 for and men. SAT. Similarly high correlations were noted for intrareader comparisons. Thus, our aims in a community-based sample of women and men free of CVD across the spectrum of BMI were to Risk Factor and Covariate Assessment assess whether the volume of SAT and VAT are associated Risk factors and covariates were measured at the contemporaneous with metabolic risk factors cross-sectionally, to assess examination. BMI, defined as weight (in kilograms) divided by the whether VAT is more strongly associated with metabolic risk square of height (in meters), was measured at each index examina- factors than is SAT, and to determine whether sophisticated tion. WC was measured at the level of the umbilicus. Fasting plasma glucose, total and high-density lipoprotein (HDL) cholesterol, and volumetric imaging methods of SAT and VAT provide triglycerides were measured on fasting morning samples. Diabetes information about metabolic risk other than that offered by was defined as a fasting plasma glucose level Ն126 mg/dL at a simpler measures such as BMI and WC. Framingham examination or treatment with either insulin or a hypoglycemic agent. Participants were considered current smokers if Methods they had smoked at least 1 cigarette per day for the previous year. Assessed through a series of physician-administered questions, Study Sample alcohol use was dichotomized on the basis of consumption of Ͼ14 Participants for this study were drawn from the Framingham Heart drinks per week (in men) or 7 drinks per week (in women). Physical Study Multidetector Computed Tomography Study, a population- activity, assessed with a questionnaire, is a score based on the based substudy of the community-based Framingham Heart Study average daily number of hours of sleep and sedentary, slight, Offspring and Third-Generation Study cohorts. Beginning in 1948, moderate, and heavy activity of the participant. Women were 5209 men and women 28 to 62 years of age were enrolled in the considered menopausal if their periods had stopped for Ն1 year. original cohort of the Framingham Heart Study. The offspring and Impaired fasting glucose was defined as a fasting plasma glucose spouses of the offspring of the original cohort were enrolled in the level of 100 to 125 mg/dL among those not treated for diabetes. Offspring Study starting in 1971. Selection criteria and study design Hypertension was defined as systolic blood pressure Ն140 mm Hg, Fox et al CT Adipose Tissue and Cardiometabolic Risk 41 diastolic blood pressure Ն90 mm Hg, or on treatment. MetS was TABLE 1. Clinical Characteristics of Study Participants Free of defined from modified Adult Treatment Panel criteria.43 Clinical CVD Who Underwent Radiographic Assessment of SAT and VAT Volumes Statistical Analysis Women SAT and VAT were normally distributed. Sex-specific age-adjusted (nϭ1452) Men (nϭ1549) Pearson correlation coefficients were used to assess simple correlations between SAT and VAT and metabolic risk factors. Multivari- Age, y 51Ϯ949Ϯ10 able linear and logistic regression was used to assess the significance BMI, kg/m2 26.7Ϯ5.4 28.3Ϯ4.4 of covariate-adjusted cross-sectional relations between continuous WC, cm 92Ϯ14 103Ϯ11 and dichotomous metabolic risk factors and SAT and VAT. For continuous risk factors, the covariate-adjusted average change in risk Triglycerides,* mg/dL 92 (65, 134) 113 (76, 171) factor per 1–standard deviation (SD) increase in adipose tissue was HDL cholesterol, mg/dL 62Ϯ17 46Ϯ12 estimated; for dichotomous risk factors, the change in odds of the Total cholesterol, mg/dL 197Ϯ36 196Ϯ34 risk factor prevalence per 1-SD increase in adipose tissue was estimated. All models were sex specific to account for the strong sex Systolic blood pressure, mm Hg 119Ϯ17 123Ϯ14 interactions observed. Covariates in all models included age, smok- Diastolic blood pressure, mm Hg 73Ϯ978Ϯ9 ing (3-level variable: current/former/never smoker), physical activ- Hypertension, % 24 28 ity, alcohol use (dichotomized at Ͼ14 drinks per week in men or Ͼ7 drinks per week in women), menopausal status, and hormone Fasting plasma glucose, mg/dL 95Ϯ16 101Ϯ20 replacement therapy. In addition, lipid treatment, hypertension treat- Impaired fasting glucose,† % 18 38 ment, and diabetes treatment were included as covariates in models Diabetes mellitus, % 4.2 6.1 for HDL cholesterol, log triglycerides, systolic and diastolic blood pressures, and fasting plasma glucose, respectively. R2 values were MetS, % 25 35 computed for continuous models and c statistics were computed for Smoking, % dichotomous models to assess the relative contribution of SAT and Current 12 13 VAT to explain the outcomes (risk factors). For each risk factor, tests for the significance of the difference between the SAT and VAT Former 42 34 regression coefficients were carried out within a multivariate stan- Never 46 53 dardized regression (in which variables were first standardized to a Postmenopausal, % 48 mean of 0 and an SD of 1) to assess the relative importance of each ⅐⅐⅐ adipose tissue measurement in predicting the risk factor. To assess Hormone replacement therapy, % 23 ⅐⅐⅐ the incremental utility of adding VAT to models that contain BMI or Alcohol use,‡ % 15 16 WC, the above multivariate analyses were repeated for VAT with SAT, cm3 3071Ϯ1444 2603Ϯ1187 BMI and WC added as covariates in the multivariate regression models. Similar models were not examined for SAT because models VAT, cm3 1306Ϯ807 2159Ϯ967 2 with SAT alone did not yield higher R or c statistics than models Data are presented as meanϮSD when appropriate. that included BMI and WC alone. As a secondary analysis, the above *Median (25th, 75th percentiles). multivariate regressions were rerun using the general estimating †Fasting plasma glucose of 100 to 125 mg/dL; percentage is based on those 44 equation linear and logistic regression to account for correlations without diabetes. among related individuals (siblings) in the study sample. SAS ‡Defined as Ͼ14 drinks per week (men) or Ͼ7 drinks per week (women). version 8.0 was used to perform all computations; a 2-tailed value of PϽ0.05 was considered significant.44 The mean SAT volume was 3071Ϯ1444 cm3 in women Heritability Analysis and 2603Ϯ1187 cm3 in men. The mean VAT volume in Heritability quantifies the proportion of trait variability resulting women was 1306Ϯ807 cm3 and in men was 2159Ϯ967 cm3. from the additive effect of genes; the contributions of both genes and early common environment cannot be differentiated. Heritability Correlations With SAT and VAT calculations rely on the assumption that trait variation can be Correlations of SAT and VAT with metabolic risk factors are partitioned into genetic, known covariates and environmental (unknown) components. It is further assumed that the genetic compo- shown in Table 2. SAT was positively correlated with age in nent is polygenic; there is no variation attributable to dominance women (rϭ0.13, PϽ0.001) but not men, and VAT was components. To determine the heritability of SAT and VAT, we positively correlated with age in both sexes (rϭ0.36 in created sex-specific and cohort-specific residuals from multivariable women and men, PϽ0.001). SAT and VAT were highly regression after adjusting for age, age squared, smoking, and correlated, with an age-adjusted correlation coefficient be- menopausal status using the overall sample with interpretable CT Ͻ scans. Residuals were then pooled, and SOLAR45 was used to tween SAT and VAT of 0.71 (P 0.0001) in women and 0.58 calculate heritabilities using variance-components analysis. (PϽ0.0001) in men. Both BMI and WC were strongly The authors had full access to and take full responsibility for the correlated with SAT and VAT after adjustment for age (Table integrity of the data. All authors have read and agree to the 2). All risk factors were highly correlated with both SAT and manuscript as written. VAT, except for serum total cholesterol with SAT in men and physical activity index with VAT in men. Results Overall, 1452 women and 1549 men were available for Multivariable-Adjusted Regressions With SAT, analysis. The mean age was 50 years (Table 1); approxi- VAT, and Metabolic Risk Factor Variables mately one quarter of the sample was hypertensive, 5% had Results of multivariable-adjusted general linear regression diabetes, and approximately one third had MetS. Approxi- analyses for SAT and VAT for both continuous and dichot- mately half of the women were postmenopausal. omous metabolic risk factors are shown in Table 3. In 42 Circulation July 3, 2007

TABLE 2. Age-Adjusted Pearson Correlation Coefficients and WC. Because models with BMI and WC routinely Between Metabolic Risk Factors and SAT and VAT Volumes yielded higher R2 or c statistic than models with SAT (Table Women Men I in the online-only Data Supplement), the addition of all 3 variables into one model was not pursued. For example, in SAT VAT SAT VAT women, SAT plus covariates were associated with 21% of the Age 0.13† 0.36† 0.03 0.36† variation in log triglycerides (R2ϭ0.21), VAT plus covariates BMI 0.88† 0.75† 0.83† 0.71† were associated with 30% of the variation in log triglycerides, WC 0.87† 0.78† 0.88† 0.73† and both BMI and WC plus the covariates were associated Log triglycerides 0.31† 0.46† 0.18† 0.37† with 26% of the variation in triglycerides. Models with VAT, BMI, and WC demonstrated significant additional contribu- HDL cholesterol Ϫ0.25† Ϫ0.35† Ϫ0.17† Ϫ0.33† tion of VAT for all variables except diabetes in men. Total cholesterol 0.11† 0.15† 0.02 0.08* Statistically significant residual effect sizes for VAT were Systolic blood pressure 0.26† 0.30† 0.18† 0.24† observed for all metabolic risk factors except diabetes in men Diastolic blood pressure 0.26† 0.28† 0.21† 0.27† (Table 3). Blood glucose 0.23† 0.34† 0.12† 0.19† Physical activity index Ϫ0.14† Ϫ0.09* Ϫ0.08* Ϫ0.03 Risk Factor Distribution Based on Quartiles of VAT *PϽ0.01; †PϽ0.001. Because VAT adds to risk factor variation above and beyond BMI and WC, we assessed the impact of stratifying individuals by VAT quartile within clinically defined categories of women, per 1-SD increase in SAT, systolic blood pressure BMI (normal weight, BMI Ͻ25 kg/m2; overweight, BMI of Ϯ Ϯ increased on average 3.9 0.4 mm Hg ( 1 SE), whereas 25 to 29.9 kg/m2; and obese, BMI Ն30 kg/m2). Thirty-three Ϯ VAT was 4.8 0.4 mm Hg higher. For systolic blood pressure percent of the sample was normal weight, 41% was over- in women, the difference between the magnitude of effect of weight, and 26% was obese. Among normal-weight, over- ϭ the SAT versus VAT was not significant (P 0.10; Table 3). weight, and obese individuals, there was a highly statistically In men, the magnitude of the association of the average significant stepwise linear increase in the prevalence of the systolic blood pressure increase per 1-SD increase in VAT MetS across quartiles of VAT in both women and men (see was larger than for SAT (3.3 versus 2.3 mm Hg, respectively; the Figure) after adjustment for age and BMI; similar rela- ϭ P 0.01 for difference in the regression coefficients between tions were noted for additional risk factors, including hyper- SAT and VAT). Similar results were obtained for diastolic tension and impaired fasting glucose. blood pressure. In women and men, the association of both SAT and VAT Secondary Analysis with continuous measures of metabolic risk factors was When education status was included as an additional covari- highly significant. For fasting plasma glucose, the effect of ate, relations between SAT and VAT and the continuous and VAT was stronger than that of SAT (PϽ0.0001 for difference dichotomous metabolic risk factors were not materially dif- in women, Pϭ0.001 in men). Strong and significant results ferent (data not shown). When analyses were rerun using the for log triglycerides and HDL cholesterol followed similar general estimating equation procedure, the resulting proba- patterns (Table 3). bility values were not substantively changed from those Highly significant associations with SAT and VAT also discussed above (data not shown). were noted for dichotomous risk factor variables. Among women and men, both SAT and VAT were associated with an Heritability Analysis increased odds of hypertension (Table 3). In women, the odds Heritability for SAT was 57%, whereas heritability for VAT ratio of hypertension per 1-SD increase in VAT (odds ratio, was 36%. 2.1) was stronger than that for SAT (odds ratio, 1.7; Pϭ0.001 for difference between SAT and VAT); similar differences Discussion were noted for men. Similar highly significant differences In our community-based sample, volumetric CT measures of also were noted for impaired fasting glucose, diabetes, and both SAT and VAT were correlated with multiple metabolic MetS and are presented in Table 3. risk factors, although risk factor correlations with VAT were The magnitude of association between VAT and all risk consistently significantly stronger than those for SAT. VAT, factors examined was consistently greater for women than for not SAT, provided information above and beyond simple men (Table 3). Weaker sex differences were observed for clinical anthropometrics, including BMI and WC, and con- SAT. sistently provided differences in risk factor stratification among individuals who were overweight and obese. VAT Residual Effect of VAT in Multivariable Models was more strongly associated with metabolic risk factors in That Contain BMI and WC women than in men. Finally, both SAT and VAT are heritable To address whether radiographic imaging of abdominal traits. adipose tissue explains variation in metabolic risk factors VAT has traditionally been considered the more patho- over and above the contribution of BMI and WC, we genic adipose tissue compartment compared with SAT, but examined the residual effect size of each metabolic risk factor data confirming these relations using high-resolution volu- from multivariable models that additionally contained BMI metric imaging assessments in large, community-based sam-

Fox et al CT Adipose Tissue and Cardiometabolic Risk 43

TABLE 3. Sex-Specific Multivariable-Adjusted* Regressions for SAT and VAT With Continuous Metabolic Risk Factors (Top) and Dichotomous Risk Factors (Bottom)

Women Men

Residual Effect Residual Effect MV-Adjusted P for Size After MV-Adjusted Size After Residual Either SAT P for MV/BMI/WC Residual P for Either P for MV/BMI/WC P for Sex Effect Size or VAT† SAT vs VAT‡ Adjustment Effect Size SAT or VAT† SAT vs VAT‡ Adjustment Interaction SBP SAT 3.9Ϯ0.4 Ͻ0.0001 ⅐⅐⅐ 2.3Ϯ0.3 Ͻ0.0001 ⅐⅐⅐ 0.01 VAT 4.8Ϯ0.4 Ͻ0.0001 0.10 2.5Ϯ0.7 3.3Ϯ0.4 Ͻ0.0001 0.01 1.8Ϯ0.05 Ͻ0.0001 DBP SAT 2.2Ϯ0.2 Ͻ0.0001 ⅐⅐⅐ 1.9Ϯ0.2 Ͻ0.0001 ⅐⅐⅐ 0.77 VAT 2.6Ϯ0.3 Ͻ0.0001 0.33 1.4Ϯ0.4 2.6Ϯ0.2 Ͻ0.0001 0.008 1.5Ϯ0.3 0.01 FPG SAT 3.4Ϯ0.3 Ͻ0.0001 ⅐⅐⅐ 1.6Ϯ0.4 0.0002 ⅐⅐⅐ 0.03 VAT 4.8Ϯ0.4 Ͻ0.0001 Ͻ0.0001 3.4Ϯ0.6 3.1Ϯ0.5 Ͻ0.0001 0.001 1.8Ϯ0.7 Ͻ0.0001 Log TG SAT 0.14Ϯ0.01 Ͻ0.0001 ⅐⅐⅐ 0.10Ϯ0.01 0.003 ⅐⅐⅐ 0.16 VAT 0.23Ϯ0.01 Ͻ0.0001 Ͻ0.0001 0.19Ϯ0.02 0.22Ϯ0.01 Ͻ0.0001 Ͻ0.0001 0.22Ϯ0.02 0.0002 HDL SAT Ϫ3.9Ϯ0.4 Ͻ0.0001 ⅐⅐⅐ Ϫ2.0Ϯ0.3 Ͻ0.0001 ⅐⅐⅐ 0.006 VAT Ϫ5.9Ϯ0.4 Ͻ0.0001 Ͻ0.0001 Ϫ4.5Ϯ0.7 Ϫ4.5Ϯ0.3 Ͻ0.0001 Ͻ0.0001 Ϫ3.8Ϯ0.5 Ͻ0.0001 HTN SAT 1.7 (1.5–2.0) Ͻ0.0001 ⅐⅐⅐ 1.5 (1.4–1.7) Ͻ0.0001 ⅐⅐⅐ 0.89 VAT 2.1 (1.8–2.4) Ͻ0.0001 Ͻ0.0001 1.6 (1.3–2.0) 1.9 (1.6–2.1) Ͻ0.0001 0.006 1.6 (1.3–1.9) 0.01 IFG SAT 2.0 (1.7–2.3) Ͻ0.0001 ⅐⅐⅐ 1.5 (1.3–1.7) Ͻ0.0001 ⅐⅐⅐ 0.04 VAT 2.5 (2.1–2.9) Ͻ0.0001 0.001 2.1 (1.7–2.6) 1.8 (1.6–2.0) Ͻ0.0001 0.005 1.5 (1.2–1.8) Ͻ0.0001 DM SAT 1.6 (1.2–2.0) 0.007 ⅐⅐⅐ 1.6 (1.3–1.9) Ͻ0.0001 ⅐⅐⅐ 0.27 VAT 2.1 (1.6–2.6) Ͻ0.0001 0.0003 1.9 (1.3–2.7) 1.6 (1.3–2.0) Ͻ0.0001 0.91 0.9 (0.7–1.3) 0.03 MetS SAT 3.0 (2.6–3.5) Ͻ0.0001 ⅐⅐⅐ 2.5 (2.2–2.8) Ͻ0.0001 ⅐⅐⅐ 0.77 VAT 4.7 (3.9–5.7) Ͻ0.0001 Ͻ0.0001 1.9 (1.3–2.7) 4.2 (3.5–5.0) Ͻ0.0001 Ͻ0.0001 2.6 (2.1–3.2) 0.002 MV indicates multivariable; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; TG, triglycerides; HTN, hypertension; IFG, impaired fasting glucose; and DM, diastolic mellitus. Data presented include effect size (the average change in risk factorϮSE) per 1 SD in adipose tissue for continuous data, and the change in odds of the condition per 1 SD of adipose tissue with 95% CIs for dichotomous data. *Adjusted for age, smoking, alcohol use, physical activity, and menopausal status (women only), hormone replacement therapy (women only); for blood pressure, FPG, HDL cholesterol, and log triglycerides, an additional covariate of treatment for HTN, diabetes, or lipid disorders, respectively, was included. †For SAT or VAT in the model. ‡For SAT vs VAT difference. ples of women and men have been lacking. The mechanism community-based sample of men and women in that we show of increased metabolic risk is hypothesized to be related to that all cardiometabolic risk factors examined were more the metabolically active adipose tissue found in the visceral strongly associated with VAT than SAT. We also extend the region,7–16 in addition to the drainage of these substances current literature to note statistically significant, albeit directly into the portal circulation.46 Multiple small studies weaker, correlations with SAT. Although SAT and VAT are have demonstrated that the visceral fat compartment is highly correlated with each other, we used the R2 (for metabolically active, secreting such vasoactive substances as continuous variables) and c statistic (for dichotomous vari- inflammatory markers,15,47 adipocytokines,10,13,48,49 markers ables) to determine the total variance explained by SAT and of hemostasis and fibrinolysis (including plasminogen acti- VAT. We also performed a formal test of the difference in the vator inhibitor-1),50,51 and growth factors (including vascular ␤-coefficients for SAT compared with VAT in relation to the endothelial growth factor),52 which may contribute to its role outcome variables, and in nearly every situation, the in cardiometabolic risk factor manifestation.53–55 ␤-coefficient from the regression model was stronger for Our results are consistent with these prior findings in small VAT than for SAT. Of note, SAT volume is greater than VAT studies and extend these findings to a well-characterized, volume in both women and men. Therefore, although the 44 Circulation July 3, 2007

Women Men A 40 40 30 30 ** 20 ** *** 20 ** 10 10

Prevalence (%) 0 0 HTN IFG DM MetS HTN IFG DM MetS

100 B 100 80 80 60 *** 60 *** *** *** 40 *** 40 *

Prevalence (%) 20 20 0 0 HTN IFG DM MetS HTN IFG DM MetS

C 100 100 80 *** 80 *** 60 ** ** 60 ** 40 40 20 Prevalence (%) 20 0 0 HTN IFG DM MetS HTN IFG DM MetS

*p<0.05 **p<0.01 ***p<0.001 Q1 Q2 Q3 Q4

Prevalence of hypertension (HTN), impaired fasting glucose (IFG), diabetes (DM), and MetS among normal-weight (A), overweight (B), and obese (C) individuals. Probability values represent those for linear trend across VAT quartiles and are adjusted for age and BMI. effect sizes between VAT and risk factors may be higher, it is We also found that both SAT and VAT were associated possible that SAT volume actually contributes to more with triglycerides and HDL cholesterol in women and men. absolute risk. Our results build on those of others, which confirm the known Of interest are recent findings from the Dallas Heart Study, association between VAT and lipids.25,27–29 However, we which examined metabolic risk factors relations in 1934 extend these findings to include strong and significant rela- black and white women and men with SAT and VAT as tions of SAT with HDL cholesterol and triglycerides. The assessed by MRI.56 An important difference between our primary difference with prior studies may be our large sample study and the Dallas Heart Study is the inclusion of percent size in a community-based cohort compared with the few body fat in regression models. Given that we do not have a other studies that were adequately powered to compare the 25,29 measure of percent body fat, our findings are not directly difference between SAT and VAT. Similarly, for impaired fasting glucose and diabetes, mul- comparable. Nonetheless, the results of Vega et al56 also tiple prior studies have demonstrated relations between VAT show considerably higher R2 for models that include VAT and prediabetic hyperglycemia and diabetes,9,10,18,19 but few than for SAT, particularly among white participants, and have yielded significant relations for SAT. However, SAT increased R2 for models that include VAT above and beyond has been shown in multiple studies to be more strongly percent body fat and WC. associated with insulin resistance than is VAT; this has been Our results show that both SAT and VAT are associated reviewed previously.58 Some studies of insulin resistance positively with prevalence of hypertension, but only VAT have demonstrated stronger correlations with SAT than with provides significant information above and beyond BMI and VAT,20 especially in women. In the Insulin Resistance WC. Other studies have demonstrated relations between VAT Atherosclerosis Study (IRAS), both SAT and VAT were and hypertension.22–24,57 Among Japanese Americans and important correlates of insulin resistance.21 One small study whites, VAT but not SAT was associated with hypertension, of 47 women and men demonstrated equivalent correlations even after adjustment for BMI and WC,22,57 whereas among of deep SAT and VAT with respect to cardiometabolic risk blacks, both SAT and VAT were associated with hyperten- factors.59 sion in men and women,24 underscoring the relative impor- Although our results show that VAT is more highly tance of fat depots among different ethnic groups.38 correlated with MetS than is SAT, SAT was an important Fox et al CT Adipose Tissue and Cardiometabolic Risk 45 correlate of the MetS. These findings are in contrast to prior variability in these traits is familial. Two prior studies that studies, which have reported that SAT is only weakly investigated the heritability of intra-abdominal fat depots68,69 associated with MetS. For example, in the Health, Aging, and have found estimates for VAT ranging from 42% to 56% and Body Composition (Health ABC) Study, SAT was associated estimates for SAT of 42%. Differences between our findings only with MetS in normal-weight and overweight men31; and prior published work may be due to the inclusion of however, unlike our study, the Health ABC Study focused younger participants with lower mean BMI, exclusion of primarily on older individuals.32 Therefore, SAT may be an certain metabolic conditions, and inclusion in a fitness study, important adipose tissue compartment that should not be which may have biased estimates. Overall, these findings overlooked. Only 1 very small intervention study has been suggest that a significant proportion of variability in VAT and conducted to date to examine the relation of SAT reduction SAT may be due to genetic causes. Further research is with metabolic variables: In a small study of 15 women who warranted to explore this further, including uncovering underwent large-volume liposuction, despite the loss of genomic regions of linkage and novel candidate genes. nearly 10 kg subcutaneous fat, improvements in cardiometabolic risk factors were not observed.60 However, the small Implications sample size, associated low power, and inclusion of morbidly The relation of MetS and its components with increasing obese study participants make it difficult to rule out a VAT quartiles, particularly in overweight and obese individ- beneficial effect. uals, suggests that VAT in particular may confer increasing The strong correlation between SAT and cardiometabolic risk within clinically defined categories of body weight. Two risk factors may be driven by the results from some20,21,58 but thirds of our study sample were either overweight or obese, not all37,61 studies that have shown that insulin sensitivity is statistics that mirror national data.6 Our data suggest that related to SAT and VAT. In addition to insulin resistance, further observational and possibly interventional studies are relations between specific fat depots and adipocytokines may warranted to test the impact of weight reduction and, in be responsible for mediating the relations with risk factors. In particular, the reduction of VAT on metabolic risk factors and particular, leptin has been shown to be equally, if not more, overall CVD risk. Additionally, our work demonstrates 62 correlated to SAT compared with VAT. Leptin also has strong and significant results for SAT and VAT in relation to 63–66 been implicated in vascular dysfunction, which suggests cardiometabolic risk factors, suggesting that SAT should not another potential mechanism whereby SAT may be associ- be overlooked in regional adipose tissue research. Nonethe- ated with cardiometabolic risk factors. less, we note that the proportion of overall variation of VAT Despite the statistically significant results observed with and of SAT with metabolic risk factors is moderate at best. both SAT and VAT, only VAT provided information above This finding, which has been observed previously,56 suggests and beyond BMI and WC, suggesting that VAT may be a that other factors not accounted for in this study may be unique pathogenic fat depot. Similar findings have been noted responsible for the variation in metabolic risk factors. In fact, among Japanese Americans, for whom VAT but not SAT was many of these traits have a substantial heritable component, associated with hypertension, even after adjustment for BMI with reported heritabilities for glucose being 34% to 51%70; and WC.22,57 Unfortunately, we were unable to analyze SAT for systolic blood pressure, 53%71; and for total cholesterol, in the same models as BMI and WC because of the high 40%.72 Therefore, the potential role of gene–adiposity inter- collinearity between the variables. In fact, the R2 of SAT actions should be considered in future research. versus BMI/WC is much higher for SAT than for VAT (0.76 versus 0.54 for men, 0.81 versus 0.64 for women). Strengths and Limitations Sex Differences Strengths of our study include the use of a community-based sample with participants not enriched for adiposity-related In our study, we found evidence for sex interactions in that increasing volumes of SAT and of VAT were consistently traits. Routine screening of metabolic risk factors was per- and more strongly associated with more adverse risk factors formed, and adjustment was made for several potential levels in women than in men. To the best of our knowledge, confounders. We used a highly reproducible volumetric our findings are the most comprehensive examination of sex method of SAT and VAT assessment, which accounts for differences reported to date. In the Health ABC Study, a heterogeneity of fat distribution throughout the abdomen. We significant sex interaction was observed between VAT and were able to assess the role of SAT and VAT above and diabetes.10 Among women and men from the Quebec Family beyond clinical anthropometry. Our study participants were Study and the Health, Risk factors, Exercise Training, and primarily middle-aged, allowing assessment of relations be- Genetics (HERITAGE) Family Study, only in women were tween fat compartments and risk factors in the absence of higher amounts of VAT associated with adverse CVD risk significant comorbidity. Lastly, we have a large sample with factor profiles.67 The cause of these sex differences is adequate power to detect potentially smaller but significant uncertain but may be related to the higher amount of hepatic relations with SAT. Limitations include the cross-sectional free fatty acid delivery derived from lipolysis from VAT that design; because the associations are not prospective, causality has been observed in women than in men.16 cannot be inferred. Because the Framingham Offspring Study is primarily a white sample, generalizability to other ethnic Heritability groups is uncertain. For example, Japanese Americans and We demonstrate moderate to high heritability for VAT and Southeast Asians have groups with more visceral fat than SAT, respectively, indicating that a significant portion of the expected for a given overall BMI,38 whereas blacks have less 46 Circulation July 3, 2007 visceral fat than do whites for a given BMI.73 Although we 8. Klein S. The case of visceral fat: argument for the defense. J Clin Invest. accounted for sibling–sibling correlations, current analytical 2004;113:1530–1532. 9. Goodpaster BH, Krishnaswami S, Resnick H, Kelley DE, Haggerty C, methods did not allow us to account for all familial relations. Harris TB, Schwartz AV, Kritchevsky S, Newman AB. Association Because we did not subdivide subcutaneous fat into superfi- between regional adipose tissue distribution and both type 2 diabetes and cial and deep compartments, we cannot comment on the impaired glucose tolerance in elderly men and women. Diabetes Care. relative importance of these compartments. Furthermore, we 2003;26:372–379. 10. Kanaya AM, Harris T, Goodpaster BH, Tylavsky F, Cummings SR. measured only abdominal, not truncal, SAT. Truncal SAT has Adipocytokines attenuate the association between visceral adiposity and been shown to be more correlated to insulin resistance than is diabetes in older adults. Diabetes Care. 2004;27:1375–1380. abdominal SAT in men.58,74 Finally, we had only radiograph- 11. Matsuzawa Y. Therapy insight: adipocytokines in metabolic syndrome and related cardiovascular disease. Nat Clin Pract Cardiovasc Med. ic CT measures of intra-abdominal fat, not other techniques 2006;3:35–42. such as MRI or ultrasound. MRI may provide a better 12. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation resolution of fat depots, and ultrasound may be a reasonable to the metabolic syndrome. Endocr Rev. 2000;21:697–738. alternative to CT75 and is less invasive. Neither MRI nor 13. Yatagai T, Nagasaka S, Taniguchi A, Fukushima M, Nakamura T, Kuroe A, Nakai Y, Ishibashi S. Hypoadiponectinemia is associated with visceral ultrasound places patients at risk of radiation exposure, but fat accumulation and insulin resistance in Japanese men with type 2 MRI is limited by its expense and amount of time needed to diabetes mellitus. Metabolism. 2003;52:1274–1278. perform the actual scan. 14. Bacha F, Saad R, Gungor N, Arslanian SA. Adiponectin in youth: relationship to visceral adiposity, insulin sensitivity, and beta-cell function. Conclusions Diabetes Care. 2004;27:547–552. Both SAT and VAT are associated with an adverse metabolic 15. Saijo Y, Kiyota N, Kawasaki Y, Miyazaki Y, Kashimura J, Fukuda M, risk profile. However, only VAT provides information above Kishi R. Relationship between C-reactive protein and visceral adipose tissue in healthy Japanese subjects. Diabetes Obes Metab. 2004;6: and beyond easily obtainable clinical anthropometric mea- 249–258. surements. Measurement of VAT may provide a more com- 16. Nielsen S, Guo Z, Johnson CM, Hensrud DD, Jensen MD. Splanchnic plete understanding of metabolic risk, and further studies are lipolysis in human obesity. J Clin Invest. 2004;113:1582–1588. 17. Kuk JL, Lee S, Heymsfield SB, Ross R. Waist circumference and warranted to prospectively assess the impact of the lowering abdominal adipose tissue distribution: influence of age and sex. Am J Clin of VAT and SAT on the incidence of MetS and CVD. Nutr. 2005;81:1330–1334. 18. Hayashi T, Boyko EJ, Leonetti DL, McNeely MJ, Newell-Morris L, Kahn Sources of Funding SE, Fujimoto WY. Visceral adiposity and the risk of impaired glucose tolerance: a prospective study among Japanese Americans. Diabetes This work was supported by the National Heart, Lung and Blood Care. 2003;26:650–655. Institute’s Framingham Heart Study (N01-HC-25195). Dr Meigs is 19. Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. Visceral adi- supported by an American Diabetes Association Career Develop- posity and risk of type 2 diabetes: a prospective study among Japanese ment Award. Dr Vasan is supported in part by 2K24HL04334 Americans. Diabetes Care. 2000;23:465–471. (National Heart, Lung, and Blood Institute, National Institutes of 20. Tulloch-Reid MK, Hanson RL, Sebring NG, Reynolds JC, Premkumar A, Health). Genovese DJ, Sumner AE. Both subcutaneous and visceral adipose tissue correlate highly with insulin resistance in African Americans. Obes Res. Disclosures 2004;12:1352–1359. Dr Meigs has been the recipient of research grants from GlaxoSmith- 21. Wagenknecht LE, Langefeld CD, Scherzinger AL, Norris JM, Haffner Kline, Pfizer, and Wyeth and has served on Advisory Boards for SM, Saad MF, Bergman RN. Insulin sensitivity, insulin secretion, and GlaxoSmithKline, Merck, Pfizer, and Lilly. The other authors report abdominal fat: the Insulin Resistance Atherosclerosis Study (IRAS) Family Study. Diabetes. 2003;52:2490–2496. no conflicts. 22. Hayashi T, Boyko EJ, Leonetti DL, McNeely MJ, Newell-Morris L, Kahn SE, Fujimoto WY. Visceral adiposity is an independent predictor of References incident hypertension in Japanese Americans. Ann Intern Med. 2004;140: 1. Thom T, Haase N, Rosamond W, Howard VJ, Rumsfeld J, Manolio T, 992–1000. Zheng ZJ, Flegal K, O’Donnell C, Kittner S, Lloyd-Jones D, Goff DC Jr, 23. Sironi AM, Gastaldelli A, Mari A, Ciociaro D, Positano V, Buzzigoli E, Hong Y, Adams R, Friday G, Furie K, Gorelick P, Kissela B, Marler J, Ghione S, Turchi S, Lombardi M, Ferrannini E. Visceral fat in hyper- Meigs J, Roger V, Sidney S, Sorlie P, Steinberger J, Wasserthiel-Smoller tension: influence on insulin resistance and beta-cell function. Hyper- S, Wilson M, Wolf P; American Heart Association Statistics Committee tension. 2004;44:127–133. and Stroke Statistics Subcommittee. Heart disease and stroke statis- 24. Ding J, Visser M, Kritchevsky SB, Nevitt M, Newman A, Sutton-Tyrrell tics–2006 update: a report from the American Heart Association Statistics K, Harris TB. The association of regional fat depots with hypertension in Committee and Stroke Statistics Subcommittee. Circulation. 2006;113: older persons of white and African American ethnicity. Am J Hypertens. e85–e151. 2004;17:971–976. 2. Fox CS, Evans JC, Larson MG, Kannel WB, Levy D. Temporal trends in 25. Pascot A, Lemieux S, Lemieux I, Prud’homme D, Tremblay A, Bouchard coronary heart disease mortality and sudden cardiac death from 1950 to C, Nadeau A, Couillard C, Tchernof A, Bergeron J, Despres JP. Age- 1999: the Framingham Heart Study. Circulation. 2004;110:522–527. related increase in visceral adipose tissue and body fat and the metabolic 3. Hu FB, Stampfer MJ, Manson JE, Grodstein F, Colditz GA, Speizer FE, risk profile of premenopausal women. Diabetes Care. 1999;22: Willett WC. Trends in the incidence of coronary heart disease and 1471–1478. changes in diet and lifestyle in women. N Engl J Med. 2000;343: 26. Nagaretani H, Nakamura T, Funahashi T, Kotani K, Miyanaga M, 530–537. Tokunaga K, Takahashi M, Nishizawa H, Kishida K, Kuriyama H, Hotta 4. Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and K, Yamashita S, Matsuzawa Y. Visceral fat is a major contributor for obesity in the United States: prevalence and trends, 1960–1994. Int J multiple risk factor clustering in Japanese men with impaired glucose Obes Relat Metab Disord. 1998;22:39–47. tolerance. Diabetes Care. 2001;24:2127–2133. 5. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends 27. Nicklas BJ, Penninx BW, Ryan AS, Berman DM, Lynch NA, Dennis KE. in obesity among US adults, 1999–2000. JAMA. 2002;288:1723–1727. Visceral adipose tissue cutoffs associated with metabolic risk factors for 6. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal coronary heart disease in women. Diabetes Care. 2003;26:1413–1420. KM. Prevalence of overweight and obesity in the United States, 28. Lemieux S, Prud’homme D, Moorjani S, Tremblay A, Bouchard C, 1999–2004. JAMA. 2006;295:1549–1555. Lupien PJ, Despres JP. Do elevated levels of abdominal visceral adipose 7. Poirier P, Despres JP. Waist circumference, visceral obesity, and cardio- tissue contribute to age-related differences in plasma lipoprotein concen- vascular risk. J Cardiopulm Rehabil. 2003;23:161–169. trations in men? Atherosclerosis. 1995;118:155–164. Fox et al CT Adipose Tissue and Cardiometabolic Risk 47

29. Kobayashi H, Nakamura T, Miyaoka K, Nishida M, Funahashi T, 48. Azuma K, Katsukawa F, Oguchi S, Murata M, Yamazaki H, Shimada A, Yamashita S, Matsuzawa Y. Visceral fat accumulation contributes to Saruta T. Correlation between serum resistin level and adiposity in obese insulin resistance, small-sized low-density lipoprotein, and progression of individuals. Obes Res. 2003;11:997–1001. coronary artery disease in middle-aged non-obese Japanese men. Jpn Circ 49. Fukuhara A, Matsuda M, Nishizawa M, Segawa K, Tanaka M, Kishimoto J. 2001;65:193–199. K, Matsuki Y, Murakami M, Ichisaka T, Murakami H, Watanabe E, 30. Mori Y, Hoshino K, Yokota K, Yokose T, Tajima N. Increased visceral Takagi T, Akiyoshi M, Ohtsubo T, Kihara S, Yamashita S, Makishima M, fat and impaired glucose tolerance predict the increased risk of metabolic Funahashi T, Yamanaka S, Hiramatsu R, Matsuzawa Y, Shimomura I. syndrome in Japanese middle-aged men. Exp Clin Endocrinol Diabetes. Visfatin: a protein secreted by visceral fat that mimics the effects of 2005;113:334–339. insulin. Science. 2005;307:426–430. 31. Goodpaster BH, Krishnaswami S, Harris TB, Katsiaras A, Kritchevsky 50. Cigolini M, Targher G, Bergamo Andreis I, Tonoli M, Agostino G, De SB, Simonsick EM, Nevitt M, Holvoet P, Newman AB. Obesity, regional Sandre G. Visceral fat accumulation and its relation to plasma hemostatic body fat distribution, and the metabolic syndrome in older men and factors in healthy men. Arterioscler Thromb Vasc Biol. 1996;16: women. Arch Intern Med. 2005;165:777–783. 368–374. 32. Carr DB, Utzschneider KM, Hull RL, Kodama K, Retzlaff BM, Brunzell 51. Mertens I, Van Gaal LF. Visceral fat as a determinant of fibrinolysis and JD, Shofer JB, Fish BE, Knopp RH, Kahn SE. Intra-abdominal fat is a hemostasis. Semin Vasc Med. 2005;5:48–55. 52. Miyazawa-Hoshimoto S, Takahashi K, Bujo H, Hashimoto N, Saito Y. major determinant of the National Cholesterol Education Program Adult Elevated serum vascular endothelial growth factor is associated with Treatment Panel III criteria for the metabolic syndrome. Diabetes. 2004; visceral fat accumulation in human obese subjects. Diabetologia. 2003; 53:2087–2094. 46:1483–1488. 33. von Eyben FE, Mouritsen E, Holm J, Montvilas P, Dimcevski G, Suciu 53. Giusti V, Suter M, Verdumo C, Gaillard RC, Burckhardt P, Pralong FP. G, Helleberg I, Kristensen L, von Eyben R. Intra-abdominal obesity and Molecular determinants of human adipose tissue: differences between metabolic risk factors: a study of young adults. Int J Obes Relat Metab visceral and subcutaneous compartments in obese women. J Clin Endo- Disord. 2003;27:941–949. crinol Metab. 2004;89:1379–1384. 34. Ross R, Freeman J, Hudson R, Janssen I. Abdominal obesity, muscle 54. Arner P. Regional differences in protein production by human adipose composition, and insulin resistance in premenopausal women. J Clin tissue. Biochem Soc Trans. 2001;29:72–75. Endocrinol Metab. 2002;87:5044–5051. 55. Dusserre E, Moulin P, Vidal H. Differences in mRNA expression of the 35. Fujimoto WY, Bergstrom RW, Boyko EJ, Chen KW, Leonetti DL, proteins secreted by the adipocytes in human subcutaneous and visceral Newell-Morris L, Shofer JB, Wahl PW. Visceral adiposity and incident adipose tissues. Biochim Biophys Acta. 2000;1500:88–96. coronary heart disease in Japanese-American men: the 10-year follow-up 56. Vega GL, Adams-Huet B, Peshock R, Willett D, Shah B, Grundy SM. results of the Seattle Japanese-American Community Diabetes Study. Influence of body fat content and distribution on variation in metabolic Diabetes Care. 1999;22:1808–1812. risk. J Clin Endocrinol Metab. 2006;91:4459–4466. 36. Hernandez-Ono A, Monter-Carreola G, Zamora-Gonzalez J, Cardoso- 57. Hayashi T, Boyko EJ, Leonetti DL, McNeely MJ, Newell-Morris L, Kahn Saldana G, Posadas-Sanchez R, Torres-Tamayo M, Posadas-Romero C. SE, Fujimoto WY. Visceral adiposity and the prevalence of hypertension Association of visceral fat with coronary risk factors in a in Japanese Americans. Circulation. 2003;108:1718–1723. population-based sample of postmenopausal women. Int J Obes Relat 58. Garg A. Regional adiposity and insulin resistance. J Clin Endocrinol Metab Disord. 2002;26:33–39. Metab. 2004;89:4206–4210. 37. Tong J, Fujimoto WY, Kahn SE, Weigle DS, McNeely MJ, Leonetti DL, 59. Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions Shofer JB, Boyko EJ. Insulin, C-peptide, and leptin concentrations predict of subcutaneous abdominal adipose tissue and insulin resistance. increased visceral adiposity at 5- and 10-year follow-ups in nondiabetic Am J Physiol Endocrinol Metab. 2000;278:E941–E948. Japanese Americans. Diabetes. 2005;54:985–990. 60. Klein S, Fontana L, Young VL, Coggan AR, Kilo C, Patterson BW, 38. Park YW, Allison DB, Heymsfield SB, Gallagher D. Larger amounts of Mohammed BS. Absence of an effect of liposuction on insulin action and visceral adipose tissue in Asian Americans. Obes Res. 2001;9:381–387. risk factors for coronary heart disease. N Engl J Med. 2004;350: 39. Maurovich-Horvat P, Massaro J, Fox CS, Moselewski F, O’Donnell CJ, 2549–2557. Hoffmann U. Comparison of anthropometric, area- and volume-based 61. Goodpaster BH, Kelley DE, Wing RR, Meier A, Thaete FL. Effects of assessment of abdominal subcutaneous and visceral adipose tissue weight loss on regional fat distribution and insulin sensitivity in obesity. volumes using multi-detector computed tomography. Int J Obes (Lond). Diabetes. 1999;48:839–847. 2007;31:500–506. 62. Ruhl CE, Everhart JE, Ding J, Goodpaster BH, Kanaya AM, Simonsick 40. Dawber TR, Kannel WB, Lyell LP. An approach to longitudinal studies EM, Tylavsky FA, Harris TB. Serum leptin concentrations and body in a community: the Framingham Heart Study. Ann N Y Acad Sci. 1963; adipose measures in older black and white adults. Am J Clin Nutr. 107:539–556. 2004;80:576–583. 41. Shurtleff D. Some characteristics related to the incidence of cardiovas- 63. Sundell J, Huupponen R, Raitakari OT, Nuutila P, Knuuti J. High serum cular disease and death: Framingham study, 18-year follow-up. In: leptin is associated with attenuated coronary vasoreactivity. Obes Res. Kannel WB, Fordon T, eds. The Framingham Study: An Epidemiological 2003;11:776–782. 64. Yamagishi S, Inagaki Y, Amano S, Okamoto T, Takeuchi M. Investigation of Cardiovascular Disease. Washington, DC: Department Up-regulation of vascular endothelial growth factor and down-regulation of Health, Education, and Welfare; 1973. DHEW publication No. NIH of pigment epithelium-derived factor messenger ribonucleic acid levels in 74–599, section 30. leptin-exposed cultured retinal pericytes. Int J Tissue React. 2002;24: 42. Hopkins PN, Ellison RC, Province MA, Pankow JS, Carr JJ, Arnett DK, 137–142. Lewis CE, Heiss G, Hunt SC. Association of coronary artery calcified 65. Cooke JP, Oka RK. Does leptin cause vascular disease? Circulation. plaque with clinical coronary heart disease in the National Heart, Lung, 2002;106:1904–1905. and Blood Institute’s Family Heart Study. Am J Cardiol. 2006;97: 66. Singhal A, Farooqi IS, Cole TJ, O’Rahilly S, Fewtrell M, Kattenhorn M, 1564–1569. Lucas A, Deanfield J. Influence of leptin on arterial distensibility: a novel 43. Executive Summary of the third report of the National Cholesterol Edu- link between obesity and cardiovascular disease? Circulation. 2002;106: cation Program (NCEP) Expert Panel on Detection, Evaluation, and 1919–1924. Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel 67. Tanaka S, Togashi K, Rankinen T, Perusse L, Leon AS, Rao DC, Skinner III). JAMA. 2001;285:2486–2497. JS, Wilmore JH, Despres JP, Bouchard C. Sex differences in the rela- 44. SAS/STAT User’s Guide, Version 8. Cary, NC: SAS Institute Inc; 2000. tionships of abdominal fat to cardiovascular disease risk among normal- 45. Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in weight white subjects. Int J Obes Relat Metab Disord. 2004;28:320–323. general pedigrees. Am J Hum Genet. 1998;62:1198–1211. 68. Hong Y, Rice T, Gagnon J, Despres JP, Nadeau A, Perusse L, Bouchard 46. Bjorntorp P. “Portal” adipose tissue as a generator of risk factors for C, Leon AS, Skinner JS, Wilmore JH, Rao DC. Familial clustering of cardiovascular disease and diabetes. Arteriosclerosis. 1990;10:493–496. insulin and abdominal visceral fat: the HERITAGE Family Study. J Clin 47. Lemieux I, Pascot A, Prud’homme D, Almeras N, Bogaty P, Nadeau A, Endocrinol Metab. 1998;83:4239–4245. Bergeron J, Despres JP. Elevated C-reactive protein: another component 69. Perusse L, Despres JP, Lemieux S, Rice T, Rao DC, Bouchard C. Familial of the atherothrombotic profile of abdominal obesity. Arterioscler aggregation of abdominal visceral fat level: results from the Quebec Thromb Vasc Biol. 2001;21:961–967. Family Study. Metabolism. 1996;45:378–382.

48 Circulation July 3, 2007

70. Meigs JB, Panhuysen CI, Myers RH, Wilson PW, Cupples LA. A estimates of heritability in the Framingham Heart Study data. BMC genome-wide scan for loci linked to plasma levels of glucose and Genet. 2003;4(suppl 1):S36. HbA(1c) in a community-based sample of Caucasian pedigrees: the 73. Tittelbach TJ, Berman DM, Nicklas BJ, Ryan AS, Goldberg AP. Racial Framingham Offspring Study. Diabetes. 2002;51:833–840. differences in adipocyte size and relationship to the metabolic syndrome 71. Levy D, DeStefano AL, Larson MG, O’Donnell CJ, Lifton RP, Gavras H, in obese women. Obes Res. 2004;12:990–998. Cupples LA, Myers RH. Evidence for a gene influencing blood pressure 74. Abate N, Garg A, Peshock RM, Stray-Gundersen J, Grundy SM. Rela- on chromosome 17: genome scan linkage results for longitudinal blood tionships of generalized and regional adiposity to insulin sensitivity in pressure phenotypes in subjects from the Framingham Heart Study. men. J Clin Invest. 1995;96:88–98. Hypertension. 2000;36:477–483. 75. Stolk RP, Wink O, Zelissen PM, Meijer R, van Gils AP, Grobbee DE. Validity 72. Mathias RA, Roy-Gagnon MH, Justice CM, Papanicolaou GJ, Fan YT, and reproducibility of ultrasonography for the measurement of intra-abdominal Pugh EW, Wilson AF. Comparison of year-of-exam- and age-matched adipose tissue. Int J Obes Relat Metab Disord. 2001;25:1346–1351.

CLINICAL PERSPECTIVE Visceral adipose tissue (VAT) compartments may confer increased metabolic risk. The incremental utility of measuring both VAT and subcutaneous abdominal adipose tissue (SAT) in association with metabolic risk factors and underlying heritability has not been well described in a population-based setting. Participants from the Framingham Heart Study underwent multidetector computed tomography assessment of SAT and VAT volumes. SAT and VAT were significantly associated with increased odds of hypertension, impaired fasting glucose, diabetes, and metabolic syndrome. In women, relations between VAT and risk factors were consistently stronger than in men. VAT was more strongly correlated with most metabolic risk factors than was SAT. Furthermore, VAT but not SAT contributed significantly to risk factor variation after adjustment for body mass index and waist circumference. Among overweight and obese individuals, the prevalence of hypertension, impaired fasting glucose, and metabolic syndrome increased linearly and significantly across increasing VAT quartiles. Heritability values for SAT and VAT were 57% and 36%, respectively. Although both SAT and VAT are correlated with metabolic risk factors, VAT remains more strongly associated with an adverse metabolic risk profile even after accounting for standard anthropometric indices. Our findings are consistent with the hypothesized role of visceral fat as a unique, pathogenic fat depot. Measurement of VAT may provide a more complete understanding of metabolic risk associated with variation in fat distribution. Metoprolol Reverses Left Ventricular Remodeling in Patients With Asymptomatic Systolic Dysfunction The REversal of VEntricular Remodeling with Toprol-XL (REVERT) Trial

Wilson S. Colucci, MD; Theodore J. Kolias, MD; Kirkwood F. Adams, MD; William F. Armstrong, MD; Jalal K. Ghali, MD; Stephen S. Gottlieb, MD; Barry Greenberg, MD; Michael I. Klibaner, MD, PhD; Marrick L. Kukin, MD; Jennifer E. Sugg, MS; on behalf of the REVERT Study Group*

Background—There are no randomized, controlled trial data to support the benefit of ␤-blockers in patients with asymptomatic left ventricular systolic dysfunction. We investigated whether ␤-blocker therapy ameliorates left ventricular remodeling in asymptomatic patients with left ventricular systolic dysfunction. Method and Results—Patients with left ventricular ejection fraction Ͻ40%, mild left ventricular dilation, and no symptoms of heart failure (New York Heart Association class I) were randomly assigned to receive extended-release metoprolol succinate (Toprol-XL, AstraZeneca) 200 mg or 50 mg or placebo for 12 months. Echocardiographic assessments of left ventricular end-systolic volume, end-diastolic volume, mass, and ejection fraction were performed at baseline and at 6 and 12 months. The 149 patients randomized to the 3 treatment groups (200 mg, nϭ48; 50 mg, nϭ48; and placebo, nϭ53) were similar with regard to all baseline characteristics including age (mean, 66 years), gender (74% male), plasma brain natriuretic peptide (79 pg/mL), left ventricular end-diastolic volume index (110 mL/m2), and left ventricular ejection fraction (27%). At 12 months in the 200-mg group, there was a 14Ϯ3 mL/m2 decrease (least square meanϮSE) in end-systolic volume index and a 6Ϯ1% increase in left ventricular ejection fraction (PϽ0.05 versus baseline and placebo for both). The decrease in end-diastolic volume index (14Ϯ3) was different from that seen at baseline (PϽ0.05) but not with placebo. In the 50-mg group, end-systolic and end-diastolic volume indexes decreased relative to baseline but were not different from what was seen with placebo, whereas ejection fraction increased by 4Ϯ1% (PϽ0.05 versus baseline and placebo). Conclusion—␤-Blocker therapy can ameliorate left ventricular remodeling in asymptomatic patients with left ventricular systolic dysfunction. (Circulation. 2007;116:49-56.) Key Words: heart failure Ⅲ receptors, adrenergic, beta Ⅲ remodeling Ⅲ ventricles

symptomatic left ventricular (LV) systolic dysfunction ham Heart Study population, the annual incidence of symp- Ais common in the general population, with a prevalence tomatic HF was Ϸ6%, and the annual mortality rate was on the order of 3%,1,2 constituting a high percentage of all Ϸ8%.1 Despite the important consequences of asymptomatic patients with LV dysfunction. For example, in the Olmstead LV systolic dysfunction, very few clinical trials of therapeutic County population, less than half of all patients with an LV agents have been conducted in this population. ejection fraction (EF) Ͻ40% had been diagnosed with con- In patients with symptoms of HF due to systolic LV gestive heart failure (HF).2 Although asymptomatic, these dysfunction, extensive clinical trial data have demonstrated patients are at high risk for developing clinical HF. In that ␤-blockers improve survival, decrease hospitalizations asymptomatic patients with a LVEF Ͻ40% in the Framing- related to HF, and alleviate symptoms.3–6 The improved

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz. Received October 5, 2006; accepted April 20, 2007. From Boston University Medical Center, Boston, Mass (W.S.C.); University of Michigan Medical Center, Ann Arbor (T.J.K., W.F.A.); University of North Carolina School of Medicine, Chapel Hill (K.F.A.); Wayne State University, Detroit, Mich (J.K.G.); University of Maryland Hospital, Baltimore (S.S.G.); University of California at San Diego (B.G.); AstraZeneca LP, Wilmington, Del (M.I.K., J.E.S.); and St Luke’s–Roosevelt Hospitals, Columbia University College of Physicians and Surgeons, New York, NY (M.L.K.). *All members of the REVERT Study Group are listed in the Appendix. Clinical trial registration information—URL: http://www.clinicaltrials.gov. Unique identifier: NCT00038077. Guest Editor for this article was Martin M. LeWinter, MD. Correspondence to Wilson S. Colucci, MD, Cardiovascular Medicine, Boston University Medical Center, 88 E Newton St, Boston, MA 02118. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.666016 50 Circulation July 3, 2007 outcomes are associated with amelioration of LV remodeling ACE inhibitors, angiotensin receptor blockers, diuretics, digoxin, characterized by decreases in end-diastolic and end-systolic and/or vasodilators for at least 3 months before randomization. Patients were excluded if they had indications for or contraindi- volumes and an increase in EF.7–10 The studies that have cations to ␤-blocker therapy or took ␤-blockers (including ophthal- demonstrated these beneficial effects on outcomes and re- mic preparations) for Ͼ1 week during the 6 months preceding modeling have been performed almost exclusively in symp- randomization. Also excluded were patients who, during the 6 tomatic patients in New York Heart Association (NYHA) months before randomization, had myocardial infarction, unstable functional classes II, III, or IV. angina, coronary intervention, or hospitalization for cardiovascular- Ͻ Although an exception would be the Australia/New Zea- related causes, as well as patients with heart rate 60 bpm, sitting blood pressure Ͼ140/90 mm Hg, heart block greater than first degree land Carvedilol Trial, in which approximately one third of not treated with a permanent pacemaker, history of ventricular or subjects were asymptomatic, the survival benefits of carve- atrial fibrillation, or serum creatinine Ͼ3 mg/dL. The institutional dilol in that trial were not analyzed with regard to functional review boards of the participating institutions approved the protocol, class.11 Likewise, although some patients in the Carvedilol and all participants provided written informed consent. The study was conducted in accordance with the Declaration of Helsinki. Prospective Randomized Cumulative Survival (CAPRI- CORN) trial had asymptomatic LV dysfunction,12 the impact Echocardiographic Measurements of therapy was not analyzed with regard to functional class Two-dimensional echocardiograms with Doppler were recorded at and would not be directly applicable to patients with chronic screening (baseline) and at 6 and 12 months, with the same scanner LV dysfunction. used for each patient. The videotapes were evaluated in a blinded There have been no randomized controlled trials of the fashion by the core echocardiography laboratory at the University of ␤ Michigan. effects of -blockers on clinical outcomes or remodeling in The echocardiograms were analyzed with the use of a dedicated patients who have chronic LV systolic dysfunction but are offline echocardiography analysis system (TomTec Imaging Sys- free of symptoms. Theoretically, patients with asymptomatic tems, Munich, Germany). LV end-systolic volume (LVESV), LV dysfunction may be less responsive to ␤-blockers because LVEDV, and LVEF were measured by Simpson’s method in the the degree of sympathetic nervous system activation is apical 4-chamber view, which was used for the main analyses, as well as the apical 2-chamber view when possible. LV mass was 13–15 less. Alternatively, there is reason to believe that calculated from the 2-dimensional parasternal long-axis view with ␤-blockers would be effective in such patients because some the use of the Penn cubed formula.19 LVESV index (LVESVI), studies have shown clinical benefit in patients with mild (eg, LVEDVI, and LV mass index (LVMI) were determined by dividing predominantly NYHA class II) symptoms of HF.16 volume or mass by body surface area. Because of the lack of direct evidence from randomized Treatment Regimen controlled trials, the current recommendation for the use of a ␤ Patients who met inclusion/exclusion criteria were randomized in a -blocker in patients with a chronic reduction in LVEF but no 1:1:1 ratio to a 52-week treatment with metoprolol succinate HF symptoms is based only on expert opinion.17 Furthermore, extended-release tablets (metoprolol succinate, TOPROL-XL, Astra- it is unlikely that a placebo-controlled study can be performed Zeneca, Wilmington, Del) or placebo as follows: (1) metoprolol to test the effects of ␤-blockade on clinical outcomes in this succinate 200 mg/d (high-dose group); (2) metoprolol succinate 50 population. mg/d (low-dose group); or (3) placebo. The study drug was force- ␤ titrated to the assigned dose over the first 2 months on the basis of There is evidence that the outcome benefits of -blockers tolerability, and the achieved dose was maintained as tolerated until in patients with systolic LV dysfunction are related to the the end of the study. Drug blinding was achieved with the use of a antiremodeling effect, which might therefore be used as a double-blind, double-dummy technique. Treatment compliance was reasonable surrogate for clinical benefit.18 Accordingly, we verified by pill count of returned study medication at each visit. designed a placebo-controlled, randomized trial, REversal of Brain Natriuretic Peptide VEntricular Remodeling with Toprol-XL (REVERT), to test ␤ In a substudy that enrolled 82 patients, venous blood was collected the hypothesis that -blocker therapy would ameliorate LV at baseline and at 6 and 12 months for determination of plasma brain remodeling in asymptomatic (NYHA functional class I) natriuretic peptide, which was measured by radioimmunoassay patients with LV systolic dysfunction. (Quest Diagnostics, Van Nuys, Calif).

Methods Statistical Analyses The change in LVESVI from baseline to 12 months was chosen as Entry Criteria the prespecified primary end point because it reflects changes in both To be eligible, patients must have had a LVEF Ͻ40% and mild LV LV size and systolic function and has been shown to be a sensitive dilatation (LV end-diastolic volume index [LVEDVI] Ͼ75 mL/m2) index of LV remodeling.7 The power calculation was based on a SD due to idiopathic, ischemic, or hypertensive cardiomyopathy and for LVESVI at 12 months of 15 mL/m2 and a dropout rate of 20%. must have had no symptoms of HF for at least 2 months. The With an ␣ of 0.050 for a 2-sided test, 150 patients would provide determination of LV systolic dysfunction and dilation was based on 94% power to detect a change of 12 mL/m2 and 84% power to detect a screening echocardiogram that was performed within 14 days of a change of 10 mL/m2. Prespecified key secondary end points randomization and interpreted by the core laboratory. Asymptomatic included the change from baseline in LVESVI at month 6 and LV systolic dysfunction was defined as no limitation of ordinary changes from baseline in LVEDVI, LVMI, and LVEF at months 6 physical activity because of fatigue or dyspnea (NYHA functional and 12. Efficacy was analyzed by a modified intention-to-treat class I). Patients previously treated for symptomatic HF were population (nϭ149 patients) who took at least 1 dose of study allowed to participate if they met the inclusion and exclusion criteria medication after randomization and had at least 1 follow-up echo- for the study. The use of an angiotensin-converting enzyme (ACE) cardiogram. All available data were analyzed, and no substitutions inhibitor or angiotensin receptor blocker for at least 3 months, as were made for missing values. Pairwise comparisons were made tolerated, was required before enrollment. Patients must have had no between the high-dose group and placebo (primary comparison) and changes in the doses of their cardiovascular medications, including between the low-dose group and placebo (secondary comparison) for

Colucci et al Metoprolol for Asymptomatic LV Dysfunction 51

TABLE 1. Demographic and Clinical Characteristics

Placebo Metoprolol Succinate Metoprolol Succinate All Patients (nϭ53) 50 mg (nϭ48) 200 mg (nϭ48) (nϭ149) Age, mean (SD), y 67 (10) 64 (14) 66 (14) 66 (13) Men, n (%) 37 (69.8) 38 (79.2) 35 (72.9) 110 (73.8) Race, n (%) White 41 (77.4) 38 (79.2) 36 (75.0) 115 (77.2) Black 11 (20.8) 9 (18.8) 12 (25.0) 32 (21.5) Other 1 (1.9) 1 (2.1) 0 2 (1.3) BMI, mean (SD), kg/m2 27.2 (4.4) 27.8 (6.0) 28.6 (5.6) 27.8 (5.4) BSA, mean (SD), m2 1.95 (0.25) 2.02 (0.27) 1.96 (0.31) 1.97 (0.28) NYHA class I, n (%) 53 (100.0) 48 (100.0) 48 (100.0) 149 (100.0) Cause of HF, n (%) Ischemic 27 (50.9) 27 (56.3) 26 (54.2) 80 (53.7) Idiopathic 17 (32.1) 12 (25.0) 14 (29.2) 43 (28.9) Hypertension 7 (13.2) 5 (10.4) 6 (12.5) 18 (12.1) Other 2 (3.8) 4 (8.3) 2 (4.2) 8 (5.4) Prior PTCA or CABG, n (%) 21 (39.6) 21 (43.8) 21 (43.8) 63 (42.3) Diabetes, n (%) 16 (30.2) 12 (25.0) 17 (35.4) 45 (30.2) Medications, n (%) ACE inhibitor or ARB 49 (92.5) 44 (91.7) 47 (97.9) 140 (94.0) Diuretics 31 (58.5) 30 (62.5) 35 (72.9) 96 (64.4) Digoxin 20 (37.7) 18 (37.5) 24 (50.0) 62 (41.6) SBP/DBP, mean, mm Hg 125.1/73.9 123.4/73.1 127.1/73.2 125.2/73.4 Heart rate, mean (SD), bpm 78.0 (11.8) 75.4 (10.3) 76.2 (9.7) 76.6 (10.7) BNP,* mean (SD), pg/mL 88.9 (54.3) 73.9 (65.6) 75.1 (87.9) 79.3 (70.9) BMI indicates body mass index; BSA, body surface area; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting; ARB, angiotensin receptor blocker; SBP, systolic blood pressure; DBP, diastolic blood pressure; and BNP, brain natriuretic peptide. *BNP samples were collected in a subset of 25 to 30 patients in each group.

all LV end points. Changes from baseline in LV dimensions were treatment was 357 (9 to 391), 358 (4 to 376), and 358 (15 to analyzed with a repeated-measures ANOVA, with terms for treat 383) days in the placebo, low-dose, and high-dose groups, ment group, time (6 and 12 months), and the interaction between respectively. Mean compliance during the 52-week double- treatment group and time. Least square means of the interaction term blind treatment period ranged from 97% to 100% in the 3 were used to estimate treatment group effects at each time, to make pairwise comparisons of the metoprolol groups versus placebo, and treatment groups. In the low-dose group, 87% of patients to test whether changes within each treatment group were different achieved the 50-mg/d dose (mean dose, 47Ϯ9 mg/d). In the from zero (ie, different from baseline values). All tests were 2-sided high-dose group, 68% of patients achieved the 200-mg/d dose and were performed at a significance level of 0.050. All values are (mean dose, 155Ϯ69 mg/d). reported as meanϮSD unless otherwise specified. All analyses were conducted with SAS software, version 8.2 (SAS Institute Inc, Cary, Baseline Characteristics NC). The 149 patients in the intention-to-treat efficacy analysis had The authors had full access to the data and take responsibility for Ϯ its integrity. All authors have read and agree to the manuscript as a mean age of 66 13 years; 74% were male, and 77% were written. white (Table 1). The cause of HF was attributed to ischemia (54%), idiopathic dilated cardiomyopathy (29%), hyperten- Results sion (12%), or other causes (5%). Ninety-four percent of patients were receiving an ACE inhibitor and/or an angioten- Study Drug Exposure sin receptor blocker, 64% were receiving a diuretic, and 42% Of the 164 patients randomized, 149 patients who had at least were receiving digitalis. 1 dose of drug and 1 follow-up echocardiogram were used for At baseline the mean heart rate was 77Ϯ11 bpm, and the the modified intention-to-treat analysis. Fifteen patients (pla- mean blood pressure was 125/73 mm Hg. Plasma brain cebo, nϭ4; low dose, nϭ6; high dose, nϭ5) were excluded natriuretic peptide averaged 79Ϯ71 pg/mL. Baseline heart from the intention-to-treat analysis because they did not have rate, blood pressure, and brain natriuretic peptide levels were a follow-up echocardiogram. The median (range) duration of similar among the 3 treatment groups.

52 Circulation July 3, 2007

5 icant only versus baseline. Similar effects were seen at 6 months. There was a similar directional pattern for changes in LVEDVI with regard to both dose and treatment duration,

) 0 although the changes were not statistically different from m

p placebo (Table 3, Figure 2B). b ( At 12 months, LVEF was unchanged in the placebo group e t -5 a Ϯ Ϯ r and increased by 4 1% in the low-dose group and 6 1% in

t r

a the high-dose group (Table 3, Figure 2C). There were similar

e **†

H directional changes at 6 months. LVMI tended to increase in -10 *† the placebo group and to decrease in both the high-dose and *† low-dose groups at both 6 and 12 months, although none of -15 these differences reached statistical significance (Table 3, Baseline 6 Months 12 Months Figure 2D).

Figure 1. Effect of metoprolol succinate on heart rate. Shown are the mean changes (SE) in heart rate compared with baseline Adverse Events and Tolerability for patients receiving metoprolol succinate 200 mg (triangles), The most common adverse events were fatigue, dizziness, 50 mg (squares), or placebo (diamonds). *PϽ0.05 vs baseline; dyspnea, peripheral edema, and HF. The percentage of Ͻ †P 0.05 vs placebo. patients discontinuing the study because of adverse events was 18%, 13%, and 11% in the placebo, low-dose, and Effects of Treatment on Heart Rate and high-dose groups, respectively. There were 2 deaths in the Blood Pressure placebo group, 2 in the low-dose group, and 4 in the Heart rate decreased by 3Ϯ10, 8Ϯ10, and 12Ϯ10 bpm in the high-dose group. The cause of death was cardiovascular in 7 placebo, low-dose, and high-dose groups at 12 months, patients. respectively (Figure 1). There were no significant changes in systolic or diastolic blood pressure from baseline to 12 Discussion months in any group. The REVERT trial shows that treatment with metoprolol succinate reduces LVESV and improves LVEF in patients Effects of Treatment on LV Dimensions with asymptomatic LV systolic dysfunction. These results At baseline, LV dimensions and EF were similar in the 3 suggest that metoprolol succinate therapy ameliorates and groups (Table 2). At 12 months, LVESVI, the primary end reverses pathological cardiac remodeling in asymptomatic point, was decreased by 4Ϯ3, 8Ϯ3, and 14Ϯ3 mL/m2 (least patients with LV systolic dysfunction. square meanϮSE) in the placebo, low-dose, and high-dose Although prior controlled studies have demonstrated that groups, respectively (Table 3, Figure 2A). The decrease with ␤-blockers can reverse LV remodeling in patients with HF, the high dose was significant versus both baseline and these studies have been conducted entirely or predominantly placebo, whereas the decrease with the low dose was signif- in symptomatic patients. Metoprolol succinate was shown to

TABLE 2. LV Echocardiographic Measurements at Baseline, 6 Months, and 12 Months

Baseline 6 Months 12 Months

No. Mean 95% CI No. Mean 95% CI No. Mean 95% CI LVESVI, mL/m2 Placebo 53 82.5 74.6 to 90.3 52 78.9 69.0 to 88.9 46 77.5 68.7 to 86.3 Metoprolol succinate 50 mg 48 82.5 74.3 to 90.7 46 79.1 69.4 to 88.8 44 75.3 64.7 to 86.0 Metoprolol succinate 200 mg 48 79.8 72.2 to 87.5 45 66.7 58.0 to 75.3 43 66.5 56.7 to 76.3 LVEDVI, mL/m2 Placebo 53 110.7 103.0 to 118.4 52 106.0 95.7 to 116.4 46 104.8 95.5 to 114.1 Metoprolol succinate 50 mg 48 110.9 102.8 to 119.0 47 108.8 98.8 to 118.8 44 104.8 93.5 to 116.1 Metoprolol succinate 200 mg 48 109.0 99.7 to 118.4 45 96.6 86.9 to 106.3 43 96.8 85.7 to 108.0 LVEF, % Placebo 53 26.8 24.6 to 28.9 52 27.6 25.1 to 30.1 46 27.5 24.7 to 30.4 Metoprolol succinate 50 mg 48 26.6 24.2 to 29.0 46 29.5 26.3 to 32.7 44 30.4 27.1 to 33.7 Metoprolol succinate 200 mg 48 27.2 25.1 to 29.4 45 32.6 29.9 to 35.2 43 33.2 30.3 to 36.2 LVMI, g/m2 Placebo 51 161.5 148.5 to 174.5 51 162.2 148.7 to 175.6 45 169.9 155.8 to 183.9 Metoprolol succinate 50 mg 48 164.3 148.0 to 180.6 46 161.1 144.5 to 177.7 44 160.0 144.3 to 175.7 Metoprolol succinate 200 mg 47 159.9 146.2 to 173.6 44 150.9 137.8 to 164.1 44 151.8 138.6 to 165.1

Colucci et al Metoprolol for Asymptomatic LV Dysfunction 53

TABLE 3. Changes in LV Echocardiographic Measurements From Baseline to 6 Months or 12 Months

Baseline to 6 Months Baseline to 12 Months

LSM 95% P vs P vs LSM 95% P vs P vs Change CI Placebo Baseline Change CI Placebo Baseline LVESVI, mL/m2 Placebo Ϫ4.5 Ϫ9.9 to 0.9 ⅐⅐⅐ 0.10 Ϫ3.7 Ϫ9.2 to 1.9 ⅐⅐⅐ 0.20 Metoprolol succinate 50 mg Ϫ3.9 Ϫ9.6 to 1.9 0.87 0.18 Ϫ7.6 Ϫ13.4 to Ϫ1.8 0.34 0.011 Metoprolol succinate 200 mg Ϫ13.1 Ϫ18.8 to Ϫ7.4 0.032 Ͻ0.001 Ϫ14.5 Ϫ20.3 to Ϫ8.6 0.009 Ͻ0.001 LVEDVI, mL/m2 Placebo Ϫ5.6 Ϫ11.8 to 0.5 ⅐⅐⅐ 0.072 Ϫ5.4 Ϫ11.7 to 1.0 ⅐⅐⅐ 0.10 Metoprolol succinate 50 mg Ϫ2.4 Ϫ9.1 to 3.9 0.50 0.43 Ϫ6.7 Ϫ13.3 to Ϫ0.1 0.77 0.047 Metoprolol succinate 200 mg Ϫ11.9 Ϫ18.5 to Ϫ5.4 0.17 Ͻ0.001 Ϫ13.5 Ϫ20.2 to Ϫ6.8 0.083 Ͻ0.001 LVEF, % Placebo 1.2 Ϫ1.1 to 3.6 ⅐⅐⅐ 0.31 0.0 Ϫ2.5 to 2.5 ⅐⅐⅐ 0.99 Metoprolol succinate 50 mg 2.9 0.4 to 5.5 0.32 0.022 3.9 1.4 to 6.5 0.032 0.003 Metoprolol succinate 200 mg 5.6 3.0 to 8.1 0.014 Ͻ0.001 6.2 3.6 to 8.7 Ͻ0.001 Ͻ0.001 LVMI, g/m2 Placebo 0.5 Ϫ11.6 to 12.7 ⅐⅐⅐ 0.93 8.0 Ϫ4.8 to 20.7 ⅐⅐⅐ 0.22 Metoprolol succinate 50 mg Ϫ4.9 Ϫ17.6 to 7.7 0.54 0.44 Ϫ7.3 Ϫ20.1 to 5.5 0.097 0.26 Metoprolol succinate 200 mg Ϫ8.1 Ϫ20.9 to 4.7 0.34 0.21 Ϫ8.4 Ϫ21.2 to 4.5 0.076 0.20 LSM indicates least square mean. improve survival and decrease hospitalizations in the Meto- ACE inhibitor4,21,23 and similar to the rates observed in prolol CR/XL Randomised Intervention Trial in Congestive asymptomatic patients in the treatment arm of the Studies of Heart Failure (MERIT-HF), a trial that studied predominantly Left Ventricular Dysfunction (SOLVD) Prevention Trial24 or NYHA class II and III patients and excluded patients in a general community population.1 An important indicator of NYHA class I.4 In the magnetic resonance imaging substudy milder LV dysfunction is the lesser extent of LV remodeling of MERIT-HF, treatment with metoprolol succinate for 6 at baseline. For example, the baseline LVEDVI in REVERT months decreased LVEDVI by 24 mL/m2, decreased LVESVI (110Ϯ29 mL/m2) is substantially smaller than the LVEDVI by 26 mL/m2, and increased LVEF by 8%.9 Qualitatively and of 153Ϯ64 mL/m2 in the MERIT-HF substudy.9 Other quantitatively similar effects of metoprolol succinate on LV indicators of mild disease in this study population are the volumes and EF were seen in the echocardiographic substudy relatively low pretreatment heart rate of 77 compared with 82 of MERIT-HF10 and the Randomized Evaluation of Strategies bpm in MERIT-HF, the relatively preserved systolic blood for Left Ventricular Dysfunction (RESOLVD) pilot study.20 pressure of 126 mm Hg, and the need for diuretics of any type Likewise, carvedilol has been shown to improve both survival in only Ϸ64% of patients. It should be noted that although the and remodeling in patients with symptomatic HF. Although patients in REVERT had to be asymptomatic for at least 2 LV dimensions were not reported in the US Carvedilol Trials months before randomization, before that time they may have Program, treatment with carvedilol for a mean of 213 days had symptoms that responded to therapy with diuretics and/or increased LVEF from 22% to 32%.21 In the Australia/New renin-angiotensin system blockade. Zealand Trial, carvedilol decreased LV end-diastolic and In REVERT, the effects of metoprolol on LVESVI and end-systolic dimensions and increased EF.7 Approximately LVEF appear to be dose dependent. Although the study was 75% of patients in the Australia/New Zealand Trial were in not powered to detect a dose-effect relationship, in the class II or III, and the results of that study were not reported high-dose group the decrease in LVESVI and the increase in with regard to NYHA class.11 LVEF were significantly different from those seen in the Although the determination of symptoms is subjective, placebo group at both 6 and 12 months, whereas these several observations suggest that the patients in REVERT changes in the low-dose group were intermediate in magni- differed markedly from those with class II and III symptoms tude and, for the most part, not significantly different from that are typical of prior ␤-blocker trials. Perhaps the most those seen with placebo. Of note, the mean dose achieved in objective measure of clinical severity is the average plasma the high-dose group (155 mg/d) of REVERT is very similar brain natriuretic peptide level of 75 pg/mL, which is below to the mean dose achieved in MERIT-HF (159 mg/d), which, the cutoff of 100 pg/mL that has a high level of selectivity for like REVERT, had a target dose of 200 mg/d. The vast excluding a diagnosis of HF.22 Another important indicator of majority (94%) of patients in REVERT were receiving an disease severity is mortality rate, which was 5% per year in ACE inhibitor or an angiotensin receptor blocker before this study, a rate well below the rate of Ϸ10% to 15% per enrollment, indicating that the antiremodeling effect of year that is typical of symptomatic patients treated with an ␤-blocker therapy in asymptomatic patients is in addition to 54 Circulation July 3, 2007

AB 5 5

0 0 ) ) 2 2 -5 -5

* * -10 -10 LVESVI (mL/m LVESVI LVEDVI (mL/m LVEDVI * *† -15 *† -15 *

-20 -20 Baseline 6 Months 12 Months Baseline 6 Months 12 Months

CD 10 10

*† 5 *† 5 *† ) 0 * 2 -5 LVEF (%) LVEF

0 (g/m LVMI -10

-15

-5 -20 Baseline 6 Months 12 Months Baseline 6 Months 12 Months

Figure 2. Effect of metoprolol succinate on LV volumes. Shown are the least square mean changes (SE) in LVESVI (A), LVEDVI (B), LVEF (C), and LVMI (D) compared with baseline for patients receiving metoprolol succinate 200 mg (triangles), 50 mg (squares), or placebo (diamonds). *PϽ0.05 vs baseline; †PϽ0.05 vs placebo. the benefits afforded by blockade of the renin-angiotensin American College of Cardiology/American Heart Associa- system. tion guidelines, published in 2004, recommended that pa- A decrease in heart rate, as occurs with ␤-blocker therapy, tients who have a history of a myocardial infarction and LV may allow more complete LV filling, thereby leading to dysfunction should be treated with ␤-blockers.26 The findings increases in stroke volume and EF. A limitation of this study of REVERT support this recommendation. is that LV dimensions were not repeated after withdrawal of REVERT was not designed to test the effect of ␤-blockade therapy, which would have allowed the exclusion of a on morbidity or mortality rates. There are very few outcomes transient effect due to the decrease in heart rate. However, the trials in patients with asymptomatic HF. The SOLVD Pre- decrease in LVESVI and the increase in LVEF observed with vention Trial demonstrated an improvement in symptoms and metoprolol at both 6 and 12 months were associated with a a decrease in hospitalization for HF but did not achieve decrease in LVEDVI, indicating that the observed improve- statistical significance with regard to survival.24 Of note, a ments cannot be attributed to the slowing of heart rate, per se. post hoc analysis of the SOLVD Prevention Trial found that In 2001, when REVERT began enrollment, the existing survival was improved significantly in patients who were American College of Cardiology/American Heart Associa- randomized to enalapril and were also taking a ␤-blocker.27 tion guidelines, published in 1999, concluded that the benefit It is unlikely that an outcomes trial of ␤-blockers in of ␤-blocker therapy beyond the first 3 months after acute patients with asymptomatic LV systolic dysfunction could be myocardial infarction had not been established conclusively, conducted. However, it has been suggested that a beneficial and as a result the use of ␤-blockers in patients with moderate effect on remodeling may be used as a surrogate for clinical or severe LV failure received only a class IIb recommenda- outcomes in patients with HF due to LV systolic dysfunction.25 In this setting, REVERT allowed the enrollment of tion.18 It is reasonable to expect that, by ameliorating LV patients who had a history of a remote myocardial infarction, remodeling, ␤-blocker therapy will delay the emergence or defined as Ͼ6 months before randomization, if they had not reemergence of symptoms in asymptomatic patients. This been treated with ␤-blockers. In practice, no patients were premise is further supported by the established positive enrolled in the study who had a myocardial infarction within relationship between improvements in LV remodeling and the year before randomization. Subsequently, the current survival with ␤-blocker therapy in symptomatic patients.8,18 Colucci et al Metoprolol for Asymptomatic LV Dysfunction 55

The results of REVERT suggest that the survival benefit with heart failure: the Metoprolol CR/XL Randomized Intervention Trial observed in symptomatic patients treated with metoprolol in Congestive Heart Failure (MERIT-HF). JAMA. 2000;283:1295–1302. 7. Doughty RN, Whalley GA, Gamble G, MacMahon S, Sharpe N; succinate may extend to asymptomatic patients with LV Australia-New Zealand Heart Failure Research Collaborative Group. Left systolic dysfunction as well. ventricular remodeling with carvedilol in patients with congestive heart failure due to ischemic heart disease. J Am Coll Cardiol. 1997;29: Appendix 1060–1066. 8. Udelson JE. Ventricular remodeling in heart failure and the effect of REVERT Study Group beta-blockade. Am J Cardiol. 2004;93:43B–48B. Alan Camp, Albert A. Carr, Barry Harold Greenberg, Bruce 9. Groenning BA, Nilsson JC, Sondergaard L, Fritz-Hansen T, Larsson HB, K. Jackson, Bruce Kowalski, Chang-seng Liang, Chiayu Hildebrandt PR. Antiremodeling effects on the left ventricle during beta- blockade with metoprolol in the treatment of chronic heart failure. JAm Chen, Chris Boylan, D. Marty Denny, Douglas Chapman, Coll Cardiol. 2000;36:2072–2080. Freny Mody, Garo Garibian, Garrie J. Haas, David R. 10. Hole T, Froland G, Gullestad L, Offstad J, Skjaerpe T. Metoprolol CR/XL Richard, Inder Anand, Jalal K. Ghali, James Zebrack, Jerome improves systolic and diastolic left ventricular function in patients with Lyman Anderson, John Murphy, Jose de Jesus Ortiz, Joseph chronic heart failure. Echocardiography. 2004;21:215–223. 11. Australia-New Zealand Heart Failure Research Collaborative Group. L. Gelormini, Larry Baruch, Lisa Mendes, Flora Sam, Marc Effects of carvedilol, a vasodilator-beta-blocker, in patients with con- Jay Kozinn, Mark J. Geller, Marrick L. Kukin, Michael gestive heart failure due to ischemic heart disease. Circulation. 1995;92: Benjamin Honan, Michael Lesser, Nancy R. Cho, Ralph M. 212–218. Vicari, Raymond Rodriguez, Robert Davidson, Robert E. 12. Dargie HJ. Effect of carvedilol on outcome after myocardial infarction in patients with left-ventricular dysfunction: the CAPRICORN randomised Foster, Robert Michael Kipperman, Robert Weiss, Russell trial. Lancet. 2001;357:1385–1390. Silverman, Stephen Gottlieb, Steven Hutchins, Thomas D. 13. Francis GS, Benedict C, Johnstone DE, Kirlin PC, Nicklas J, Liang CS, Giles, Thomas J. Knutson, Uri Elkayam, W. David Hager, Kubo SH, Rudin-Toretsky E, Yusuf S. Comparison of neuroendocrine Wayne David Old, Vince Figueredo. activation in patients with left ventricular dysfunction with and without congestive heart failure: a substudy of the Studies of Left Ventricular Dysfunction (SOLVD). Circulation. 1990;82:1724–1729. REVERT Steering Committee 14. Benedict CR, Johnstone DE, Weiner DH, Bourassa MG, Bittner V, Kay Wilson S. Colucci (Chairman), Kirkwood F. Adams, William R, Kirlin P, Greenberg B, Kohn RM, Nicklas JM; SOLVD Investigators. F. Armstrong, Stephen S. Gottlieb, Barry Greenberg, Marrick Relation of neurohumoral activation to clinical variables and degree of ventricular dysfunction: a report from the registry of Studies of Left L. Kukin. Ventricular Dysfunction. J Am Coll Cardiol. 1994;23:1410–1420. 15. Benedict CR, Francis GS, Shelton B, Johnstone DE, Kubo SH, Kirlin P, Sources of Funding Nicklas J, Liang CS, Konstam MA, Greenberg B; SOLVD Investigators. The REVERT study was funded by AstraZeneca, which was respon- Effect of long-term enalapril therapy on neurohormones in patients with sible for data collection and analysis, which was conducted accord- left ventricular dysfunction. Am J Cardiol. 1995;75:1151–1157. ing to a prespecified analysis plan. Academic leadership was 16. Colucci WS, Packer M, Bristow MR, Gilbert EM, Cohn JN, Fowler MB, provided by the Steering Committee, which supervised the manage- Krueger SK, Hershberger R, Uretsky BF, Bowers JA, Sackner-Bernstein ment of the study and was responsible for interpretation of the data, JD, Young ST, Holcslaw TL, Lukas MA; US Carvedilol Heart Failure preparation, review, and approval of the manuscript. Study Group. Carvedilol inhibits clinical progression in patients with mild symptoms of heart failure. Circulation. 1996;94:2800–2806. 17. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats Disclosures TG, Jessup M, Konstam MA, Mancini DM, Michl K, Oates JA, Rahko Authors who are employees of AstraZeneca are identified as such. PS, Silver MA, Stevenson LW, Yancy CW, Antman EM, Smith SC Jr, All other authors have received research grants, consultant fees, Adams CD, Anderson JL, Faxon DP, Fuster V, Halperin JL, Hiratzka LF, and/or honoraria from AstraZeneca. Jacobs AK, Nishimura R, Ornato JP, Page RL, Riegel B. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart References failure in the adult: a report of the American College of Cardiology/ 1. Wang TJ, Evans JC, Benjamin EJ, Levy D, Leroy EC, Vasan RS. Natural American Heart Association Task Force on Practice Guidelines (Writing history of asymptomatic left ventricular systolic dysfunction in the com- Committee to Update the 2001 Guidelines for the Evaluation and Man- munity. Circulation. 2003;108:977–982. agement of Heart Failure): developed in collaboration with the American 2. Redfield MM, Jacobsen SJ, Burnett JC Jr, Mahoney DW, Bailey KR, College of Chest Physicians and the International Society for Heart and Rodeheffer RJ. Burden of systolic and diastolic ventricular dysfunction in Lung Transplantation: endorsed by the Heart Rhythm Society. Circu- the community: appreciating the scope of the heart failure epidemic. lation. 2005;112:e154–e235. JAMA. 2003;289:194–202. 18. Konstam MA, Udelson JE, Anand IS, Cohn JN. Ventricular remodeling 3. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised in heart failure: a credible surrogate endpoint. J Card Fail. 2003;9: trial [see comments]. Lancet. 1999;353:9–13. 350–353. 4. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL 19. Devereux RB, Reichek N. Echocardiographic determination of left ven- Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). tricular mass in man: anatomic validation of the method. Circulation. Lancet. 1999;353:2001–2007. 1977;55:613–618. 5. Packer M, Fowler MB, Roecker EB, Coats AJ, Katus HA, Krum H, 20. RESOLVD Investigators. Effects of metoprolol CR in patients with Mohacsi P, Rouleau JL, Tendera M, Staiger C, Holcslaw TL, ischemic and dilated cardiomyopathy: the Randomized Evaluation of Amann-Zalan I, DeMets DL. Effect of carvedilol on the morbidity of Strategies for Left Ventricular Dysfunction Pilot Study. Circulation. patients with severe chronic heart failure: results of the Carvedilol Pro- 2000;101:378–384. spective Randomized Cumulative Survival (COPERNICUS) study. Cir- 21. Packer M, Bristow MR, Cohn JN, Colucci WS, Fowler MB, Gilbert EM, culation. 2002;106:2194–2199. Shusterman NH; U.S. Carvedilol Heart Failure Study Group. The effect 6. Hjalmarson A, Goldstein S, Fagerberg B, Wedel H, Waagstein F, of carvedilol on morbidity and mortality in patients with chronic heart Kjekshus J, Wikstrand J, El Allaf D, Vitovec J, Aldershvile J, Halinen M, failure. N Engl J Med. 1996;334:1349–1355. Dietz R, Neuhaus KL, Janosi A, Thorgeirsson G, Dunselman PH, 22. Maisel AS, McCord J, Nowak RM, Hollander JE, Wu AH, Duc P, Gullestad L, Kuch J, Herlitz J, Rickenbacher P, Ball S, Gottlieb S, Omland T, Storrow AB, Krishnaswamy P, Abraham WT, Clopton P, Steg Deedwania P; MERIT-HF Study Group. Effects of controlled-release G, Aumont MC, Westheim A, Knudsen CW, Perez A, Kamin R, metoprolol on total mortality, hospitalizations, and well-being in patients Kazanegra R, Herrmann HC, McCullough PA. Bedside B-type natriuretic 56 Circulation July 3, 2007

peptide in the emergency diagnosis of heart failure with reduced or agement of Acute Myocardial Infarction). Circulation. 1999;100: preserved ejection fraction: results from the Breathing Not Properly 1016–1030. Multinational Study. J Am Coll Cardiol. 2003;41:2010–2017. 26. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, 23. The SOLVD Investigators. Effect of enalapril on survival in patients with Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, reduced left ventricular ejection fractions and congestive heart failure. Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL, N Engl J Med. 1991;325:293–302. Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka 24. The SOLVD Investigators. Effect of enalapril on mortality and the devel- LF, Hunt SA, Jacobs AK. ACC/AHA guidelines for the management of opment of heart failure in asymptomatic patients with reduced left ven- patients with ST-elevation myocardial infarction: a report of the tricular ejection fractions [published erratum appears in N Engl J Med. American College of Cardiology/American Heart Association Task Force 1992;327:1768]. N Engl J Med. 1992;327:685–691. on Practice Guidelines (Committee to Revise the 1999 Guidelines for the 25. Ryan TJ, Antman EM, Brooks NH, Califf RM, Hillis LD, Hiratzka LF, Management of Patients with Acute Myocardial Infarction). Circulation. Rapaport E, Riegel B, Russell RO, Smith EE III, Weaver WD, Gibbons 2004;110:e82–e292. RJ, Alpert JS, Eagle KA, Gardner TJ, Garson A Jr, Gregoratos G, Smith 27. Exner DV, Dries DL, Waclawiw MA, Shelton B, Domanski MJ. Beta- SC Jr. 1999 update: ACC/AHA guidelines for the management of patients adrenergic blocking agent use and mortality in patients with asymptom- with acute myocardial infarction: executive summary and recommen- atic and symptomatic left ventricular systolic dysfunction: a post hoc dations: a report of the American College of Cardiology/American Heart analysis of the Studies of Left Ventricular Dysfunction. J Am Coll Association Task Force on Practice Guidelines (Committee on Man- Cardiol. 1999;33:916–923.

Go to http://cme.ahajournals.org to take the CME quiz for this article.

Negative Inotropy of the Gastric Proton Pump Inhibitor Pantoprazole in Myocardium From Humans and Rabbits Evaluation of Mechanisms

Wolfgang Schillinger, MD*; Nils Teucher, MD*; Samuel Sossalla, MS; Sarah Kettlewell, PhD; Carola Werner, PhD; Dirk Raddatz, MD; Andreas Elgner, MS; Gero Tenderich, MD; Burkert Pieske, MD; Giuliano Ramadori, MD; Friedrich A. Schöndube, MD; Harald Kögler, MD; Jens Kockskämper, PhD; Lars S. Maier, MD; Harald Schwörer, MD; Godfrey L. Smith, PhD; Gerd Hasenfuss, MD

Background—Proton pump inhibitors are used extensively for acid-related gastrointestinal diseases. Their effect on cardiac contractility has not been assessed directly. Methods and Results—Under physiological conditions (37°C, pH 7.35, 1.25 mmol/L Ca2ϩ), there was a dose-dependent decrease in contractile force in ventricular trabeculae isolated from end-stage failing human hearts superfused with pantoprazole. The concentration leading to 50% maximal response was 17.3Ϯ1.3 ␮g/mL. Similar observations were made in trabeculae from human atria, normal rabbit ventricles, and isolated rabbit ventricular myocytes. Real-time polymerase chain reaction demonstrated the expression of gastric Hϩ/Kϩ–adenosine triphosphatase in human and rabbit myocardium. However, measurements with BCECF-loaded rabbit trabeculae did not reveal any significant 2ϩ pantoprazole-dependent changes of pHi.Ca transients recorded from field-stimulated fluo 3–loaded myocytes (F/F0) were significantly depressed by 10.4Ϯ2.1% at 40 ␮g/mL. Intracellular Ca2ϩ fluxes were assessed in fura 2–loaded, ␮ 2ϩ voltage-clamped rabbit ventricular myocytes. Pantoprazole (40 g/mL) caused an increase in diastolic [Ca ]i by Ϯ 2ϩ 2ϩ Ϯ 33 12%, but peak systolic [Ca ]i was unchanged, resulting in a decreased Ca transient amplitude by 25 8%. The 2ϩ Ϯ 2ϩ amplitude of the L-type Ca current (ICa,L) was reduced by 35 5%, and sarcoplasmic reticulum Ca content was reduced by 18Ϯ6%. Measurements of oxalate-supported sarcoplasmic reticulum Ca2ϩ uptake in permeabilized 2ϩ 2ϩ cardiomyocytes indicated that pantoprazole decreased Ca sensitivity (Kd) of sarcoplasmic reticulum Ca adenosine ϭ Ϯ ␮ ϭ Ϯ Ͻ triphosphatase: control, Kd 358 15 nmol/L; 40 g/mL pantoprazole, Kd 395 12 nmol/L (P 0.05). Pantoprazole also acted on cardiac myofilaments to reduced Ca2ϩ-activated force. Conclusions—Pantoprazole depresses cardiac contractility in vitro by depression of Ca2ϩ signaling and myofilament activity. In view of the extensive use of this agent, the effects should be evaluated in vivo. (Circulation. 2007;116:57- 66.) Key Words: calcium Ⅲ contractility Ⅲ heart failure Ⅲ inotropic agents Ⅲ pharmacology

roton pump inhibitors (PPIs) like pantoprazole, omepra- Clinical Perspective p 66 Pzole, esomeprazole, lansoprazole, and rabeprazole are the most effective pharmacological means of reducing gastric agents for patients in a variety of settings. In a US Medicaid acid secretion by blocking the final step of proton secretion, population, PPIs accounted for 5.6% of the net pharmacy ie, the gastric acid pump, Hϩ/Kϩ–adenosine triphosphatase expenditures and ranked first in expenditures among all drug (ATPase). The efficacy of this class of drugs has been therapy classes during the 12 months before the implemen- demonstrated in the treatment of a number of acid-related tation of the PPI prior-authorization policy.2 Drug shortage gastrointestinal diseases, in particular, gastroesophageal re- bulletins have been issued repeatedly for intravenous PPIs in flux and ulcer disease.1 Hence, there is extensive use of these the United States.3

Received September 25, 2006; accepted May 1, 2007. From the Herzzentrum, Kardiologie und Pneumologie, Universitaet Goettingen, Goettingen, Germany (W.S., S.S., A.E., B.P., H.K., J.K., L.S.M., G.H.); Herzzentrum, Thorax-, Herz-, und Gefaesschirurgie, Universitaet Goettingen, Goettingen, Germany (N.T., F.A.S.); Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, UK (S.K., G.L.S.); Medizinische Statistik, Universitaet Goettingen, Goettingen, Germany (C.W.); Gastroenterologie und Endokrinologie, Universitaet Goettingen, Goettingen, Germany (D.R., G.R., H.S.); and Herz- und Diabeteszentrum Nordrhein- Westfalen, Klinik fuer Thorax- und Kardiovaskularchirurgie, Bad Oeynhausen, Germany (G.T.). *The first 2 authors contributed equally to this work. Correspondence to Wolfgang Schillinger, MD, Herzzentrum, Kardiologie und Pneumologie, Georg-August Universitaet Goettingen, Robert-Koch Strasse 40, 37099 Goettingen, Germany. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.666008

57 58 Circulation July 3, 2007

It has been shown recently that the expression of Hϩ/Kϩ- performed in accordance with institutional guidelines for the care ATPase is not limited strictly to gastric tissue. It has also been and use of laboratory animals. identified in renal4 and colonic epithelial cells,5 vascular 6 smooth muscle cells, and other tissues. In myocardium from pHi Measurements ϩ ϩ rats, the expression of H /K -ATPase has been demonstrated Intracellular pH (pHi) was measured in isolated trabeculae from at the transcriptional and protein levels.7 Biochemical evi- rabbit hearts as described previously.13 Briefly, trabeculae were dence and physiological evidence for a myocardial Hϩ/Kϩ- mounted in a cylindrical glass cuvette, connected to an isometric force transducer, and loaded with 2’,7’-bis(carboxyethyl)-5(6)- ATPase have also been found.8 Furthermore, Beisvag and ␮ ϩ ϩ carboxyfluorescein (BCECF)-AM (15 mol/L) in Tyrode’s solution coworkers7 showed a contribution from the H /K -ATPase of by 45-minute incubation at room temperature. Excitation light from Ϸ25% of total 86Rbϩ uptake in rat hearts and suggested that a mercury lamp was passed alternately through 2 bandpass filters the enzyme could contribute significantly to the regulation of (450 nm/495 nm) and focused on the muscle strip. Fluorescence ϩ ϩ ϩ ϩ emission was collected by a photomultiplier (Scientific Instruments) myocardial K and H homeostasis. Inhibition of H /K - Ϯ after passage through a bandpass filter (535 5 nm). Values of pHi ATPase might therefore induce cellular acidosis, which is were estimated from the ratio of the BCECF fluorescence signals known to depress myocardial contractility mainly at the level (F495/F450) after subtraction of background fluorescence. At the end of 9 of myofilament responsiveness to [Ca]i. each experiment, the BCECF fluorescence ratio was calibrated in vivo ϩ Orally applied PPIs are considered safe1 and have been by means of the high K -nigericin method as previously reported.15 found advantageous in regard to cardiovascular side effects 2؉ compared with histamine type 2 (H2) receptor antagonists Voltage Clamp and Intracellular [Ca ] because of lack of chronotropic and inotropic effects. In Measurements in Rabbit Cardiomyocytes contrast to famotidine, omeprazole did not show any changes The cardiomyocytes were superfused with a solution consisting of in cardiac performance in healthy volunteers as measured by (mmol/L) 144.0 NaCl, 5.4 KCl, 0.3 NaH2PO4, 1.0 MgCl2, 5.0 impedance cardiography and mechanocardiography after HEPES, 11.1 glucose, 1.8 CaCl2, 0.1 niflumic acid, 5.0 4-AP (pH 7.4) at room temperature in a chamber mounted on the stage of an 10 1-week oral treatment with therapeutic doses. Pantoprazole, inverted microscope. Voltage clamp was achieved with the use of lansoprazole, and esomeprazole are currently available as whole-cell patch-clamp technique with an Axoclamp 2A amplifier intravenous formulations in the United States. The rationale (Axon Instruments, Foster City, Calif) in switch clamp (discontinu- for use has come primarily with the suggested efficacy in ous) mode. Pipettes were filled with an intracellular solution of the reducing rebleeding after endoscopic hemostasis of bleeding following composition (mmol/L): 20.0 KCl, 100.0 K aspartate (DL), 20.0 TEA Cl, 10.0 HEPES, 4.5 MgCl2, 4.0 Na2ATP, 1.0 Na2CrP, 2.5 peptic ulcers (eg, References 11 and 12). The target goal for EGTA (pH 7.25 with KOH) and were of resistance 3 to 6 mol/L⍀. gastric pH in these patients has been suggested to be Ͼ6in Intracellular [Ca2ϩ] was measured from fura 2 fluorescence signals order to promote hemostasis and minimize clot lysis, in by a dual-wavelength spectrophotometer method as previously contradistinction to the target pH Ͼ4 for treating patients to described.16 Cytosolic loading of fura 2 was achieved by incubating cardiomyocytes with 5 ␮mol/L fura 2-AM at room temperature for prevent stress ulcer or heal ulcers or reflux esophagitis.1,11,12 12 minutes. As such, the dosing amounts of intravenous PPI have been higher than that of oral PPI. Recently, omeprazole and pantoprazole have been dosed at 80 mg followed by 8 mg/h for 72 Electrophysiological Protocols Ϫ 11,12 Isolated rabbit cardiomyocytes were held at 80 mV, and the hours. However, information regarding the cardiac effects of ϩ voltage was stepped to Ϫ40 mV (50 ms) to inactivate the inward Na high doses is lacking. Moreover, few data are available regard- current before stepping to 0 mV (150 ms). Tetrodotoxin 3ϫ10Ϫ5 ing the direct effect of such agents on the myocardium. mol/L was also used to block INa. This protocol was repeated 40 Hence, because of the abundant use of this drug class and times at a rate of 1 Hz to achieve steady state Ca2ϩ transients. 2ϩ ϩ 2ϩ because of the presence of a Hϩ/Kϩ-ATPase in myocardium, Sarcoplasmic reticulum (SR) Ca content and Na /Ca exchanger (NCX) activity were then estimated by rapidly switching to we sought to investigate the effects of PPI on contractility of ϩ 10 mmol/L caffeine to cause SR Ca2 release. In the continued isolated human myocardium using pantoprazole, which is the presence of caffeine (20 seconds), the SR is unable to reaccumulate most commonly used intravenous formulation. In addition, Ca2ϩ, and therefore Ca2ϩ removal is mainly via NCX. The time the mode of action in cardiac tissue was further analyzed in course of the decay of [Ca2ϩ] and the NCX-mediated inward current 2ϩ isolated cardiac tissue from rabbits. (INCX) represent rates of extrusion of Ca from the cell predominantly by NCX.17 These signals were fitted to exponential decays over Ͼ80% of their amplitude. The magnitude of non-NCX Ca2ϩ- ϩ Methods removal mechanisms was estimated from the Ca2 decay obtained by rapidly switching to 10 mmol/L caffeine in the presence of

Myocardial Tissue 10 mmol/L NiCl2 (which blocks NCX), in which the absence of a Experiments were performed in left ventricular muscle strips from 8 current indicated that the current obtained without NiCl2 was solely end-stage failing human hearts obtained from patients undergoing due to NCX. cardiac transplantation and right atrial trabeculae obtained from 16 patients who underwent cardiac surgery. Additional right ventricular trabeculae were obtained from adult female Chinchilla Bastard Simultaneous Measurements of Myocyte 2؉ rabbits (weight, 2.0 to 2.5 kg; Charles River Deutschland, Kisslegg, Shortening and Intracellular [Ca ] 13 2ϩ Germany). Myocardial trabeculae and adult rabbit ventricular Shortening and [Ca ]i were measured simultaneously as reported cardiac myocytes14 were isolated and forces of electrically stimu- previously.18 Cells were loaded with 10 ␮mol/L fluo 3-AM (Molec- lated preparations were investigated as described previously. Exper- ular Probes, Carlsbad, Calif) for 15 minutes. Fluo 3 was excited at imental procedures with human tissue were reviewed and approved 480 nm, and fluorescence was measured at 535 nm. The field- by the ethical committee of the University Clinics of Goettingen, and stimulation frequency was 1 Hz (37°C, pH 7.35). Normalized the subjects gave informed consent. Procedures with rabbits were amplitude of calcium transients (F/F0) was calculated by dividing Schillinger et al Negative Inotropy of Pantoprazole 59

TABLE 1. Primer Pairs Used in Real-Time PCR

Gene Sequence Gene Bank Accession No. Human Sense: 5Ј-CCACATCCACACAGCTGACAC-3Ј M63962 Hϩ/Kϩ-ATPase Antisense: 5Ј-TCAAACGTCTGCCCTGACTG-3Ј Rabbit Hϩ/Kϩ-ATPase Sense: 5Ј-ACTCTGCACCGACATTTTCCC-3Ј X64694 Antisense: 5Ј-TCAGCCTTCTCATACGCCAAG-3Ј ␤-Actin Sense: 5Ј-CTGGCACCCAGCACAATG-3Ј M10277 Antisense: 5Ј-CCGATCCACACGGAGTACTTG-3Ј

fluorescence F by the baseline fluorescence F0 after subtraction of ments of calcium sensitivity and force development of the myofila- the background fluorescence (IonWizard, IonOptix Corp). Cells ments were performed with the use of 1 relaxation and 10 activation were treated with 0/10/40 ␮g/mL pantoprazole followed by a solutions containing pantoprazole (10/40 ␮g/mL) or NaCl as control washout. Fluorescence and shortening were analyzed at steady state with increasing Ca2ϩ concentrations (from 1.66ϫ10Ϫ7 mol/L to conditions. 5.15ϫ10Ϫ5 mol/L Ca2ϩ). Active tension was measured via force 2؉ transducer and analyzed at steady state conditions. To exclude that Measurements of SR Ca Uptake Characteristics changes in the active tension were due to rundown of the preparation, 2ϩ Measurements of SR Ca uptake were performed as described each exposure to pantoprazole was preceded and followed by 1 14 previously. Ventricular myocytes freshly dissociated from adult control step with NaCl 0.9%. rabbit hearts were permeabilized with 0.1 mg/mL ␤-escin and kept in mock intracellular solution of the following composition (mmol/L): ؉ ؉ 100 Kϩ,40Naϩ, 25 HEPES, 100 ClϪ, 0.05 EGTA, 5 ATP, 10 CrP Transcription of H /K -ATPase in ϩ (pH 7.0). To test the influence of pantoprazole on SR Ca2 uptake, Human Myocardium myocytes were divided into 3 groups, and 0 (control), 10 ␮g/mL, and RNA was isolated from myocardial tissue and gastric corpus mucosa 40 ␮g/mL pantoprazole were added into the cuvette. Oxalate (10 mmol/L) was included to maintain low and constant intra-SR from humans and rabbits with the use of RNeasy Kits (Qiagen). [Ca2ϩ], and ruthenium red (2.7 ␮mol/L) was included to block SR Quantitative reverse transcription polymerase chain reaction (RT- Ca2ϩ efflux. Myocyte suspensions were stirred and [Ca2ϩ] was PCR) was performed by real-time PCR as described earlier.19 Primer ϩ ϩ monitored with the use of fura 2 (10 ␮mol/L). The decline of Ca2ϩ pairs for H /K -ATPase and ␤-actin are summarized in the Table. signal was used to calculate SR Ca2ϩ uptake characteristics. Each sample was analyzed in duplicate. The investigated number of cDNA molecules was related to ␤-actin cDNA molecules detected in Skinned Fibers the same sample to normalize for the quantity of RNA extracted and Fibers dissected from rabbit right ventricles were skinned by the efficiency of cDNA synthesis. The relative expression was then incubation with Triton (1%) for 24 hours at 4°C. Paired measure- calculated as described.19

8 6 A Control B Control (n = 15/12) 5 (n = 12/5)

6 4

3 * 4 Pantoprazole * (n = 16/12) * 2 Pantoprazole (n = 11/5) 2 1 Human atrium Human ventricle 0 0 0.625 1.25 3.125 6.25 12.5 25 50 Wash-out 0 0.625 1.25 3.125 6.25 12.5 25 50 Wash-out

C 30 D Control 6 25 (n = 11/11) 5 20 Control 4 15 Developed Tension (mN/mm²) * Esomeprazole (n = 14/12) * (n = 5/5) Pantoprazole 3 10 (n = 11/11) 5 2 Rabbit ventricle Human atrium 0 1 0 0.625 1.25 2.5 5 10 20 40 Wash-out 0 0.625 1.25 3.125 6.25 12.5 25 50 Wash-out Concentration (µg/mL)

Figure 1. Dose dependence of isometric twitch force from pantoprazole (A to C) and esomeprazole (D) in different types of myocardium as indicated and partial reversibility after washout of the drug. The number of trabeculae and hearts used for each experiment is indicated in parentheses. *PϽ0.025 (only concentrations with potential clinical relevance have been tested).

60 Circulation July 3, 2007

Pantoprazole (µg/mL): A B 0 10 20 40 160 )

11 11 11 11 11 2 10 10 10 10 10 9 9 9 9 9 6 8 8 8 8 8 7 7 7 7 7 6 6 6 6 6 4 5 5 5 5 5 4 4 4 4 4 3 3 3 3 3 2 Force (mN/mm²) 2 2 2 2 2 1 1 1 1 1 0 0 0 0 0 0 500 1000 0 500 1000 0 500 1000 0 500 1000 0 500 1000 1 10 100 Developed Tension (mN/mm Developed Time (ms) Pantoprazole [µg/mL] ) C D 2 6 6 6 6 6 6 5 5 5 5 5

4 4 4 4 4 4 3 3 3 3 3

2 2 2 2 2 2

Force (mN/mm²) 1 1 1 1 1 0 0 0 0 0 0 0 500 1000 0 500 1000 0 500 1000 0 500 1000 0 500 1000 1 10 100 Developed Tension (mN/mm Developed Time (ms) Pantoprazole [µg/mL]

Figure 2. Determination of EC50 for contractile depression of pantoprazole in human myocardium. Single twitches of original recordings of typical experiments in human atrial (A) and ventricular (C) trabeculae are shown. Mean force values in atrial (B) (nϭ8 trabeculae from 4 hearts) and ventricular (D) (nϭ6/3) trabeculae are shown. Curves have been ﬁtted to yield EC50 values as detailed in Results.

Statistical Analysis with esomeprazole in human atrial myocardium (Figure 1D). Data are expressed as meanϮSEM. Student paired t test (Figures 3 The effect was highly reproducible, and the maximum effect and 5) or repeated-measures ANOVA (Figures 1, 4, 6, and 7) were usually occurred within a few minutes after exposure to performed to test for statistically significant differences between different interventions. In general, a value of PϽ0.05 was accepted pantoprazole. To yield EC50 values of the negative inotropic as statistically significant; for post hoc analysis, probability values effect, additional dose-response experiments with maximum were adjusted according to Bonferroni (Figures 1 and 4), or the pantoprazole concentrations of 160 ␮g/mL have been per- Tukey test was used (Figure 6). formed that induced nearly complete suppression of contrac- The authors had full access to and take full responsibility for the tile force. Data points have been fitted with the use of logistic integrity of the data. All authors have read and agree to the manuscript as written. curve fit with OriginPro 7 scientific software (OriginLab Corp). The EC50 was 30.6Ϯ1.8 ␮g/mL in atrial human Results myocardium (Figure 2B) and 17.3Ϯ1.3 ␮g/mL in ventricular Dose-Response Relationship of PPIs in Isolated human myocardium (Figure 2D). Compared with control, at Trabeculae From Human and Rabbit Myocardium doses of 6.25 and 12.5 ␮g/mL of pantoprazole, the contractile Figure 1A and 1B demonstrates a dose-dependent reduction forces in human ventricular myocardium were 3.0Ϯ0.4 ver- of isometric twitch force of electrically stimulated trabeculae sus 4.2Ϯ0.7 mN/mm2 and 2.4Ϯ0.3 versus 4.0Ϯ0.6 mN/mm2 from nonfailing human atrial and failing ventricular myocar- (PϽ0.025, each). This corresponds to a depression of con- dium under the influence of pantoprazole. Negative inotropy tractile forces by 27Ϯ9% and 42Ϯ8%, respectively. In was at least partially reversible after washout of the drug. The human atrial myocardium, forces in the presence of 12.5 pantoprazole-dependent negative inotropy was also present in ␮g/mL pantoprazole were 5.1Ϯ0.8 versus 5.7Ϯ0.4 mN/mm2 ventricular myocardium from healthy adult rabbits (Figure compared with control (PϽ0.025, depression by 12Ϯ14%). 1C). Moreover, a similar dose-dependent effect was found Preincubation of trabeculae with ouabain (0.2 ␮mol/L) in-

Figure 3. pHi measurements in ventricular trabeculae from rabbits. A, pHi changes after application of pantoprazole (40 ␮g/mL) in an individual rabbit ventricular muscle strip. B, Average changes of pHi (left) and developed force (right) induced by 20-minute treatment with 40 ␮g/mL pantoprazole. Data were obtained from 5 ventricular trabeculae isolated from 5 rabbit hearts. **PϽ0.01 vs initial control. Schillinger et al Negative Inotropy of Pantoprazole 61

A 10 µg/mL 40 µg/mL Wash-out 1.90

1.85

1.80 Length (µm)

1.75 0 100 200 300 B Time (s) 1.90

1.85

1.80 Length (µm)

Figure 4. Dependence of myocyte shorten- 1.75 2ϩ ing and [Ca ]i from pantoprazole in rabbit 0 100 200 300 myocytes. Original chart ﬁles of myocyte Time (s) shortening with (A) and without (B) addition of pantoprazole are shown. C, Original re- C cordings of Ca2ϩ transients in an isotoni- cally contracting myocyte. D, Mean values PP (µg/mL): 0 10 40 of fractional shortening. E, Ca2ϩ transient measurements (nϭ16 myocytes from 10 animals for pantoprazole, and nϭ15/8 for control). *PϽ0.025 vs control. ; 400 rel. i ] + [Ca² Fluorescence Units

1s D E

4 * * 1.6 * 3 1.4 (F/F0)

2 i ] 2+ 1.2

1 [Ca Control Pantoprazole Control Pantoprazole

Fractional Shortening (%) 0 1.0 0 10 40 Wash-out 0 10 40 Wash-out Concentration (µg/mL) Concentration (µg/mL)

duced an increase in contractile force by 31.1%. After with ventricular myocardium was 1.2ϫ106 times higher in exposure to pantoprazole (40 ␮g/mL), force decreased by humans (1.3Ϯ0.8 versus 1.1ϫ10Ϫ6Ϯ0.7ϫ10Ϫ6;nϭ3) and 65.2% compared with ouabain treatment and by 54.4% 2.2ϫ106 times higher in rabbits (1262Ϯ673 versus compared with baseline values (PϽ0.05 each). 580ϫ10Ϫ6Ϯ217ϫ10Ϫ6;nϭ3), respectively. Expression of H؉/K؉-ATPase in Myocardium and Gastric Mucosa From Humans and Rabbits Pantoprazole Does Not Affect pHi ϩ ϩ H /K -ATPase mRNA expression was detectable in both pHi is an important modulator of contractility. Therefore, we gastric corpus mucosa and ventricular myocardium from tested whether pantoprazole induced changes in pHi that humans and rabbits. However, the expression was very low in might explain the observed negative inotropic effect. Figure ventricular myocardium from both species. The relative 3A shows pHi changes of a BCECF-loaded rabbit ventricular expression of Hϩ/Kϩ-ATPase in corpus mucosa compared muscle strip before and in the presence of pantoprazole (40

62 Circulation July 3, 2007

A B

(i) Control (ii) Pantoprazole Average changes

** Diastolic [Ca2+] 0 i 40 2+ Systolic [Ca ] -40 i (mV) Transient amplitude m

E -80 20

800 0

-20

] (nM) 600 2+ % change (+/-SEM) % 400 ** [Ca 0 400ms 1 0 -1 -20

Current (nA) **

I amplitude ** Ca,L -40 Ca2+ influx via I 0 % change (+/-SEM) Ca,L

-1 50ms Current (nA)

C D (i) Control (ii) Pantoprazole Average changes

Caffeine (10mM) Caffeine (10mM) 1000 I .time integral NCX 20 Ca2+ transient amplitude Rate of Ca2+ transient decay

(nM) 500 i ] 0 2+

[Ca 0 -20 0.0 * * % change (+/-SEM)

(nA) -0.1 2s m I

Figure 5. Depolarization-induced Ca2ϩ transients recorded from rabbit ventricular cardiomyocytes. A, Records of membrane voltage 2ϩ (Em), membrane current, and [Ca ]i from single cardiomyocytes (average of 4 sequential signals) during control conditions (i) and after 2ϩ 2ϩ 2ϩ perfusion with pantoprazole (40 ␮g/mL) (ii). B, MeanϮSEM of the percent change in diastolic [Ca ]I, systolic [Ca ]i, and Ca transient 2ϩ amplitude (**PϽ0.01). C, Recordings of [Ca ]i and membrane current recorded on rapid application of 10 mmol/L caffeine (as indicated above the records) during control conditions (i) and after perfusion with pantoprazole (40 ␮g/mL) (ii). D, MeanϮSEM of the percent 2ϩ 2ϩ 2ϩ change in INCX · time integral diastolic [Ca ]I,Ca transient amplitude, and rate constant for the decay of the Ca transient. *PϽ0.05.

2ϩ ␮g/mL). In this example, pHi was increased slightly by Ϸ0.1 zole on intracellular Ca cycling were further investigated in pH units. Simultaneously, developed force declined from voltage-clamped and fura 2–loaded rabbit myocytes (Figure 12.0 to 5.6 mN/mm2 or by 53% within the 20-minute 5A and 5B). On addition of 40 ␮g/mL pantoprazole, diastolic 2ϩ recording period (not shown). Average results from a total of [Ca ]i was increased by 33Ϯ12% (PϽ0.05; nϭ13), with no 2ϩ 5 muscle strips (Figure 3B) revealed that pantoprazole did not significant change in peak systolic [Ca ]i (4Ϯ7%; nϭ14). As 2ϩ induce any significant alterations in pHi (⌬pHiϭ0.07Ϯ0.10; a result of these changes, Ca transient amplitude was PϭNS), whereas developed force was reduced by 55Ϯ4% reduced by 25Ϯ8% (nϭ14; PϽ0.05) compared with control

(PϽ0.01). Thus, the negative inotropic effect of pantoprazole cells. These changes were paralleled by a reduction in ICa,L Ϯ Ͻ ϭ was not accompanied by changes in pHi. amplitude (by 35 5%; P 0.05; n 14). To measure SR and NCX function, the intracellular Ca2ϩ signals and the associ- 2؉ Ca Homeostasis in Isolated Rabbit Myocytes ated INCX were analyzed on rapid application of 10 mmol/L Similar to the findings in isolated trabeculae, pantoprazole caffeine (Figure 5C and 5D). A reduction in both the induced a dose-dependent reduction of isotonic shortening of amplitude of the caffeine induced Ca2ϩ release (by 18Ϯ6%; Ϯ ␮ isolated rabbit myocytes by 32.3 5.8% at 10 g/mL and nϭ10; PϽ0.05) and the associated time integral of INCX (by 65.4Ϯ3.4% at 40 ␮g/mL. This was paralleled by a depression 21Ϯ6%; nϭ10; PϽ0.05) on exposure to pantoprazole (40 of Ca2ϩ transient amplitude measured by fluorescence of fluo ␮g/mL) indicated that pantoprazole decreased SR Ca2ϩ con-

3–loaded myocytes (F/F0)by9.0Ϯ2.6% and 10.4Ϯ2.1% at 10 tent. No changes in the rate of decay of the caffeine induced 2ϩ and 40 ␮g/mL, respectively (Figure 4). Effects of pantopra- Ca transient (1Ϯ8%; nϭ10) and INCX (2Ϯ7%; nϭ10) were

Downloaded from circ.ahajournals.org at Mohammed Mahboob on July 15, 2007 Schillinger et al Negative Inotropy of Pantoprazole 63

2+ K of SR Ca2+ uptake Vmax of SR Ca uptake d

Control Control Low Panto 450 low Panto 0.4 High Panto High Panto * cells) *

6 0.3 400 0.2

(nmol/L) 350 0.1 d (fmol/s/10 K max V

A B

2ϩ 2ϩ Figure 6. Characteristics of SR Ca -ATPase activity as determined by use of a cuvette assay. Vmax (A) and Kd (B) of Ca transport activity in the presence of either 0 (control), 10 (low), or 40 (high) ␮g/mL of pantoprazole (Panto) are shown. *PϽ0.05 vs control. observed after pantoprazole administration, indicating that ␮g/mL (PϭNS) and 26.5Ϯ2.8% at 40 ␮g/mL (PϽ0.05). In the sarcolemmal extrusion processes (particularly NCX) were addition, a small but significant rightward shift of the Ca2ϩ unaffected by the drug. response curve indicating a decrease of Ca2ϩ sensitivity was 2ϩ found. The EC50 was 1.44 mol/L Ca at 0 ␮g/mL and 1.59 2ϩ ␮ Ͻ .(SR Ca2؉ Uptake Data mol/L Ca at 40 g/mL pantoprazole (P 0.05 Figure 6 demonstrates the effects of pantoprazole on oxalate- ϩ Discussion supported Ca2 uptake in permeabilized rabbit ventricular car- The present study shows a negative inotropic effect of diomyocytes. In the presence of either carrier solution or low (10 pantoprazole in isolated myocardium. This was dose depen- ␮g/mL) or high (40 ␮g/mL) concentrations of pantoprazole, the dent, induced nearly complete inhibition of twitch force at [Ca2ϩ] at which half maximal Ca2ϩ uptake occurred (K ) and the d very high doses, and was partially reversible. Negative value of the maximum rate of Ca2ϩ uptake (V ) were estimated max inotropy of pantoprazole was present in myocardium from with a fura 2–based SR Ca2ϩ uptake assay. Neither low nor high different species (human and rabbit) and in myocardium from concentration of pantoprazole influenced the V value. How- max different origins (atrial and ventricular), and it was found in ever, the K values of SR Ca2ϩ uptake were increased signifi- d different myocardial preparations (multicellular and single cantly at both concentrations of pantoprazole (nmol/L: control, cells). The EC for contractile force depression was 358Ϯ15; low pantoprazole, 406Ϯ15; high pantoprazole, 50 30.6Ϯ1.8 ␮g/mL in nonfailing human atrial and 17.3Ϯ1.3 395Ϯ12; PϽ0.05). ␮g/mL in failing human ventricular myocardium, respectively. Moreover, similar results could be obtained with Effects of Pantoprazole in Skinned esomeprazole, which is suggestive of a class effect of PPIs. Fiber Preparations Furthermore, we could reveal 2 underlying mechanisms for The effect of pantoprazole in skinned fibers is shown in the pantoprazole-dependent inhibition of contractile force: (1) Figure 7. There was a reduction of the maximum active reduction in the amplitude of Ca2ϩ transients as a conse- tension at saturating Ca2ϩ concentration by 7.8Ϯ0.8% at 10 quence of impaired SR Ca2ϩ uptake and reduced Ca2ϩ influx

A B

30 30

Figure 7. Measurements in skinned fibers. Dose 20 20 dependence of active tension at increasing Ca2ϩ * concentrations in skinned fiber preparations at 10 Control Control 10 10 ␮g/mL (A) and 40 ␮g/mL (B) of pantoprazole is 10 µg/mL 40 µg/mL shown. Maximal active tension at saturating Ca2ϩ Pantoprazole Pantoprazole concentration and Ca2ϩ sensitivity was significantly 0 0 lower at 40 ␮g/mL. *PϽ0.05. Active Tension (mN/mm²) Active Tension Active Tension (mN/mm²) Active Tension

-8 -7 -6 -5 -4 -8 -7 -6 -5 -4 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2+ [Ca 2+ ] (mol/L) [Ca ]o (mol/L) o 64 Circulation July 3, 2007

2ϩ 2ϩ via ICa,L and (2) reduced Ca responsiveness of the myofila- minimal changes in systolic [Ca ]i, the net effect being a ments as a result of a reduced maximal active tension and a reduced Ca2ϩ transient amplitude. Independent evidence for slightly lower Ca2ϩ sensitivity. In contrast, despite the expres- pantoprazole-dependent inhibition of SERCA as an underly- sion of the Hϩ/Kϩ-ATPase at the transcriptional level in ing mechanism was obtained with a fura 2–based SR calcium human and rabbit myocardium, no significant changes in pHi uptake assay in permeabilized cardiomyocytes. Moreover, 2ϩ could be detected in the presence of pantoprazole. Ca influx via ICa,L was also depressed by pantoprazole. The Note that the mechanisms underlying the effects of pantopra- combination of reduced SR calcium uptake and reduced Ca2ϩ zole in myocardium are completely different from the mecha- influx would explain the reduction in SR Ca2ϩ content 2ϩ nisms of the drug in gastric parietal cells and probably do not observed and raised diastolic [Ca ]i. ϩ ϩ involve inhibition of H /K -ATPase. In regard to gastric proton We also investigated the influence of pantoprazole in pump inhibition, all PPIs are prodrugs and require acid to skinned fibers. These preparations allow for the investigation become protonated and converted into the active form.1 After of drug effects directly at the myofilament level under intravenous administration or intestinal absorption, when orally well-defined in vitro conditions. Hence, it can be excluded administered the lipophilic unprotonated form readily penetrates 2ϩ that the observed effects are mediated by changes in [Ca ]i or cell membranes, including that of gastric parietal cells and pHi. We were able to show a significant reduction in maximal myocytes. In parietal cells, as it transverses the cell it is exposed active tension of the contractile proteins at saturating Ca2ϩ to the acidic environment in the secretory canaliculus of the concentrations by 26.5Ϯ2.8% at 40 ␮g/mL. Moreover, at 40 gastric site and becomes protonated, converting it to a hydro- ␮g/mL there was a slight but significant reduction in myo- philic drug that can no longer permeate cell membranes. The filament Ca2ϩ sensitivity. However, when the marked nega- drug becomes trapped in the canaliculus of the parietal cell. For tive inotropy in multicellular trabeculae is considered, these this reason, in parietal cells PPIs exhibit a substantial accumu- effects were too faint to explain entirely the mode of action of lation versus plasma at low pH, eg, 1000-fold for omeprazole pantoprazole in myocardium. Therefore, the negative inot- and 10000-fold for rabeprazole at a pH of 1. Moreover, proto- ropy of pantoprazole observed at 10 ␮g/mL mainly results nation of the drug initiates a series of chemical reactions that from decreased intracellular [Ca2ϩ]. At the higher dose (40 culminates in covalent binding of the drug with selected cysteine ␮g/mL), depression of myofilament responsiveness appears residues of the Hϩ/Kϩ-ATPase.1 to contribute to the negative inotropic effect. We also suggest The present study is the first one reporting on the expres- that the effects of pantoprazole in myocardium depend on the sion of Hϩ/Kϩ-ATPase in human and rabbit ventricular native unprotonated form of the drug and do not require myocardium. Recently, it has been suggested that Hϩ/Kϩ- activation by low pH because all effects occurred at pH 7.3 ATPase may contribute to pH homeostasis in rat myocardi- to 7.4. um.7 We therefore investigated the hypothesis that cellular In contrast to our data, Yenisehirli and Onur23 found a positive acidosis subsequent to proton pump inhibition might explain inotropic effect of 3 different PPIs in rat atria that was potenti- the negative inotropy of PPIs. However, our measurements ϩ ϩ ϩ ated by pretreatment with the Na /K -ATPase inhibitor ouabain. with the H -sensitive fluorescence dye BCECF did not reveal Moreover, lansoprazole induced a prolongation of the action any significant influence of pantoprazole on pH . Moreover, i potential. The authors speculated that this could be mediated by the absence of acidic compartments in cardiac tissue with pH ϩ ϩ Ͻ inhibition of H /K -ATPase promoting altered intracellular 1 precludes significant accumulation of pantoprazole in the 2ϩ myocardium. In moderately acidic compartments like lyso- Ca handling. In our study with human and rabbit myocardium, 2ϩ somes, slow activation of pantoprazole theoretically may pantoprazole decreased tension and reduced shortening and Ca occur with an activation half-life of 4.7 hours.20 However, transient amplitude in “unclamped” trabeculae and single cells. negative inotropy in our experiments usually occurred When single rabbit myocytes were clamped with a fixed voltage 2ϩ quickly within several minutes. Moreover, the effect was at clamp duration, pantoprazole decreased the Ca transient am- least partially reversible after washout of the drug, whereas plitude to a degree similar to that seen in “unclamped” prepara- the recovery of the gastric proton pump from pantoprazole tions. This suggests that in humans and rabbits, action potential 2ϩ has a half-life of Ϸ46 hours21 and was suggested to depend duration changes are not instrumental in the decreased Ca mainly on the synthesis of new pump protein.1,22 In addition, transient and subsequent mechanical response caused by panto- the expression of Hϩ/Kϩ-ATPase in myocardium was very prazole application. Moreover, a significant contribution of ϩ ϩ low compared with gastric mucosa. Therefore, it is unlikely H /K -ATPase to the action potential is unlikely because of the that the negative inotropy of pantoprazole in myocardium is low degree of expression. The different responses to PPIs may the consequence of Hϩ/Kϩ-ATPase inhibition. therefore be related to species. Compared with human and rabbit We investigated the effects of pantoprazole on intracellular myocardium, in rats the action potential is short and intracellular ϩ Ca2ϩ homeostasis and myofilament Ca2ϩ responsiveness, sodium is high. In particular, the latter condition favors Ca2 ϩ ϩ which are the 2 principal physiological mechanisms of the entry by reverse-mode Na -Ca2 exchange, which contributes to ϩ myocardium to alter contractility. We found a reduction in the SR Ca2 load and contractility when action potential duration Ca2ϩ transient amplitude in field-stimulated ventricular myo- increases.9 In contrast to the study by Yenisehirli and Onur, the cytes. Interpretation of this result in terms of Ca2ϩ signaling is negative inotropic effect of pantoprazole in rabbit myocardium complicated by possible effects of pantoprazole on the action observed by us was not influenced by blockade of Naϩ/Kϩ- potential. With fixed depolarizing pulses in voltage-clamped ATPase, and the magnitude of the effect was similar to that in 2ϩ myocytes, pantoprazole caused a rise in diastolic [Ca ]i and experiments without ouabain.

Schillinger et al Negative Inotropy of Pantoprazole 65

Finally, although this was beyond the scope of the present References study, we would like to speculate whether our findings might be 1. Robinson M, Horn J. Clinical pharmacology of proton pump inhibitors: of clinical relevance. Because the negative inotropic effect was what the practising physician needs to know. Drugs. 2003;63:2739–2754. 2. Delate T, Mager DE, Sheth J, Motheral BR. Clinical and financial partially reversible after washout of the drug and significant outcomes associated with a proton pump inhibitor prior-authorization accumulation in the myocardium is unlikely, the effect of program in a Medicaid population. Am J Manag Care. 2005;11:29–36. pantoprazole in vivo probably depends on plasma concentra- 3. American Society of Health-system Pharmacists (ASHP). Drug Product ␮ Shortages Management Resource Center. April 13, 2006. Available at tions. Maximal pantoprazole plasma concentrations of 4.6 g/ http://www.ashp.org/shortage/. Accessed May 16, 2006. mL24 and 10.4 ␮g/mL (Altana Pharma AG, written communi- 4. Wingo CS, Cain BD. The renal H-K-ATPase: physiological significance cation, July 23, 2004) have been found after administration of and role in potassium homeostasis. Annu Rev Physiol. 1993;55:323–347. 5. Del CJ, Rajendran VM, Binder HJ. Apical membrane localization of common oral (40 mg) and intravenous (80 mg) doses. In the ouabain-sensitive Kϩ-activated ATPase activities in rat distal colon. present work, similar concentrations induced a reduction of Am J Physiol. 1991;261:1005–1011. ϩ ϩ contractile force of isolated trabeculae by 27Ϯ9% (6.25 ␮g/mL) 6. McCabe RD, Young DB. Evidence of a K -H -ATPase in vascular Ϯ ␮ smooth muscle cells. Am J Physiol. 1992;262:1955–1958. and 42 8% (12.5 g/mL), which might indicate a potential 7. Beisvag V, Falck G, Loenechen JP, Qvigstad G, Jynge P, Skomedal T, Osnes clinical relevance of our findings. However, the duration of J-B, Sandvik AK, Ellingsen Ö. Identification and regulation of the gastric ϩ ϩ cardiac side effects is temporally limited because all PPIs are H /K -ATPase in rat heart. Acta Physiol Scand. 2003;179:251–262. 8. Nagashima R, Tsuda Y, Maruyama T, Kanaya S, Fujino Y. Possible quickly eliminated from blood with plasma elimination half-life evidence for transmembrane Kϩ-Hϩ exchange system in guinea pig myo- periods of Ϸ1 to 2 hours.20,24 Moreover, cardiac side effects may cardium. Jpn Heart J. 1999;40:351–364. be attenuated in vivo because the activity of the active free 9. Bers DM. Excitation-Contraction Coupling and Cardiac Contractile Force. 2nd ed. Dordrecht, Netherlands: Kluwer Academic Publishers; 2001. compound may be substantially lower because of high plasma 10. Halabi A, Kirch W. Cardiovascular effects of omeprazole and famotidine. protein binding.20 On the other hand, particular conditions may Scand J Gastroenterol. 1992;27:753–756. be associated with increased duration and intensity of side 11. Lau JYW, Sung JJY, Lee KKC, Yung MY, Wong SK, Wu JC, Chan FK, Ng EK, You JH, Lee CW, Chan AC, Chung SC. Effect of intravenous effects. Patients with heart failure must be investigated for PPI omeprazole on recurrent bleeding after endoscopic treatment of bleeding tolerance. These patients are much more susceptible to negative peptic ulcers. N Engl J Med. 2000;343:310–316. inotropic drugs because of blunted contractile reserve subse- 12. van Rensburg CJ, Hartmann M, Thorpe A, Venter L, Theron I, Lühmann R, Wurst W. Intragastric pH during continuous infusion with panto- 25 quent to decreased sympathetic sensitivity or negative force- prazole in patients with bleeding peptic ulcer. Am J Gastroenterol. 2003; ϩ frequency relationship.26 In addition, the dependence of H 98:2635–2641. ϩ ϩ 13. Luers C, Fialka F, Elgner A, Zhu D, Kockskamper J, von Lewinski D, elimination from H /K -ATPase may be increased in heart ϩ 2ϩ ϩ ϩ Pieske B. Stretch-dependent modulation of [Na ]i, [Ca ]i, and pHi in failure because of the impaired function of the Na /H exchange rabbit myocardium: a mechanism for the slow force response. Cardiovasc ϩ 27 subsequent to increased [Na ]i. Moreover, all PPIs undergo Res. 2005;68:454–463. extensive hepatic biotransformation before elimination. In 14. Teucher N, Prestle J, Seidler T, Currie S, Elliott EB, Reynolds DF, Schott Ϸ P, Wagner S, Kogler H, Inesi G, Bers DM, Hasenfuss G, Smith GL. CYP2C19-poor metabolizers that represent 3% to 5% of Excessive sarcoplasmic/endoplasmic reticulum Ca2ϩ-ATPase expression whites, a similar percentage of blacks, and 12% to 25% of causes increased sarcoplasmic reticulum Ca2ϩ uptake but decreases different Asian populations, much higher plasma concentrations myocyte shortening. Circulation. 2004;110:3553–3559. 15. Hasenfuss G, Maier LS, Hermann HP, Luers C, Hunlich M, Zeitz O, 1 and longer elimination half-life periods have been found. The Janssen PM, Pieske B. Influence of pyruvate on contractile performance same holds true for patients with severe liver impairment.28 and Ca2ϩ cycling in isolated failing human myocardium. Circulation. Recent American29 and Canadian30 studies have shown that 2002;105:194–199. 16. Eisner DA, Nichols CG, O’Neill SC, Smith GL, Valdeolmillos M. The appropriate use of intravenous PPIs was seen in less than half effects of metabolic inhibition on intracellular calcium and pH in isolated of the patients. In view of our data, PPIs should be prescribed rat ventricular cells. J Physiol. 1989;411:393–418. carefully. We have recently initiated a clinical study for the 17. Diaz ME, Trafford AW, O’Neill SC, Eisner DA. Measurement of sarcoplasmic reticulum Ca2ϩ content and sarcolemmal Ca2ϩ fluxes in isolated rat ventricular investigation of cardiac side effects of pantoprazole in myocytes during spontaneous Ca2ϩ release. J Physiol. 1997;501:3–16. healthy volunteers. Moreover, we encourage clinical studies 18. DeSantiago J, Maier LS, Bers DM. Frequency-dependent acceleration of to identify individuals with increased intrinsic risk for cardiac relaxation (FDAR) in heart depends on CaMKII, but not phospholamban. J Mol Cell Cardiol. 2002;34:975–984. side effects of PPIs. 19. Raddatz D, Middel P, Bockemuhl M, Benohr P, Wissmann C, Schworer H, Ramadori G. Glucocorticoid receptor expression in inflammatory bowel disease: evidence for a mucosal down-regulation in steroid- Acknowledgments unresponsive ulcerative colitis. Aliment Pharmacol Ther. 2004;19:47–61. We gratefully acknowledge the expert assistance of Hanna Schotola, 20. Kromer W, Kruger U, Huber R, Hartmann M, Steinijans VW. Differences Astrid Steen, Aileen Rankin, Gudrun Muller, Elke Neumann, and in pH-dependent activation rates of substituted benzimidazoles and biological in vitro correlates. Pharmacology. 1998;56:57–70. Michael Kothe. 21. Katashima M, Yamamoto K, Tokuma Y, Hata T, Sawada Y, Iga T. Comparative pharmacokinetic/pharmacodynamic analysis of protein pump inhibitors omeprazole, lansoprazole, and pantoprazole, in humans. Sources of Funding Eur J Drug Metab Pharmacokinet. 1998;23:19–26. This work was supported by grants from the Wellcome Trust (to Dr 22. Gedda K, Scott D, Besancon M, Lorentzon P, Sachs G. Turnover of the gastric ϩ ϩ Kettlewell), the British Heart Foundation (to Dr Smith), and the H ,K -adenosine triphosphatase a subunit and its effect on inhibition of rat Deutsche Forschungsgemeinschaft (to Dr Maier). gastric acid secretion. Gastroenterology. 1995;109:1134–1141. 23. Yenisehirli A, Onur R. Positive inotropic and negative chronotropic effects of proton pump inhibitors in isolated rat atrium. Eur J Pharmacol. 2005;519:259–266. Disclosures 24. Schulz M, Schmoldt. Therapeutic and toxic blood concentrations of more None. than 800 drugs and other xenobiotics. Pharmazie. 2003;58:447–474. 66 Circulation July 3, 2007

25. Bristow MR, Ginsburg R, Minobe W, Cubicciotti RS, Sageman WS, 28. Ferron GM, Preston RA, Noveck RJ, Pockros P, Mayer P, Getsy J, Lurie K, Billingham ME, Harrison DC, Stinson EB. Decreased catechol- Turner M, Abell M, Paul J. Pharmacokinetics of pantoprazole in amine sensitivity and beta-adrenergic receptor density in failing human patients with moderate and severe hepatic dysfunction. Clin Ther. hearts. N Engl J Med. 1982;307:205–211. 2001;23:1180–1192. 26. Schillinger W, Lehnart SE, Prestle J, Preuss M, Pieske B, Maier LS, 29. Guda NM, Noonan M, Kreiner MJ, Partington S, Vakil N. Use of Meyer M, Just H, Hasenfuss G. Influence of SR Ca2ϩ-ATPase and intravenous proton pump inhibitors in community practice: an expla- Naϩ-Ca2ϩ-exchanger on the force-frequency relation. Basic Res Cardiol. nation for the shortage? Am J Gastroenterol. 2004;99:1233–1237. 1998;93(suppl 1):38–45. 30. Kaplan GG, Bates D, McDonald D, Panaccione R, Romagnuolo 27. Pieske B, Maier LS, Piacentino V III, Weisser J, Hasenfuss G, Houser S. J. Inappropriate use of intravenous pantoprazole: extent of the ϩ Rate dependence of [Na ]i and contractility in nonfailing and failing problem and successful solutions. Clin Gastroenterol Hepatol. 2005; human. Circulation. 2002;106:447–453. 3:1207–1214.

CLINICAL PERSPECTIVE The proton pump inhibitor pantoprazole was evaluated for its effects on cardiac contractility in isolated myocardium from humans and rabbits. We found a dose-dependent negative inotropic effect mainly resulting from alterations in intracellular Ca2ϩ handling. At higher concentrations of the drug, depression of myofilament Ca2ϩ responsiveness was also observed. However, despite the expression of the gastric proton pump in human and rabbit hearts, no relevant changes in pH homeostasis could be detected. Moreover, expression of the pump was very low in myocardium compared with gastric tissue. A similar effect was observed with esomeprazole. Thus, proton pump inhibitors affect cardiac contractility by intracellular mechanisms that are distinct from their effects in gastric parietal cells. The effects in isolated myocardium were observed at concentrations that might be of potential clinical relevance. Thus, cardiac effects of proton pump inhibitors should be evaluated in vivo because of their extensive use for acid-related gastrointestinal diseases.

Interventional Cardiology

Umesh N. Khot, MD; Michele L. Johnson, RN; Curtis Ramsey, MS; Monica B. Khot, MD; Randall Todd, MD; Saeed R. Shaikh, MD; William J. Berg, MD

Background—Consensus guidelines and hospital quality-of-care programs recommend that ST-elevation myocardial infarction patients achieve a door-to-balloon time of Յ90 minutes. However, there are limited prospective data on specific measures to significantly reduce door-to-balloon time. Methods and Results—We prospectively determined the impact on median door-to-balloon time of a protocol mandating (1) emergency department physician activation of the catheterization laboratory and (2) immediate transfer of the patient to an immediately available catheterization laboratory by an in-house transfer team consisting of an emergency department nurse, a critical care unit nurse, and a chest pain unit nurse. We collected door-to-balloon time for 60 consecutive ST-elevation myocardial infarction patients undergoing emergency percutaneous intervention within 24 hours of presentation from October 1, 2004, through August 31, 2005, and compared this group with 86 consecutive ST-elevation myocardial infarction patients from September 1, 2005, through June 26, 2006, after protocol implementation. Median door-to-balloon time decreased overall (113.5 versus 75.5 minutes; PϽ0.0001), during regular hours (83.5 versus 64.5 minutes; Pϭ0.005), during off-hours (123.5 versus 77.5 minutes; PϽ0.0001), and with transfer from an outside affiliated emergency department (147 versus 85 minutes; Pϭ0.0006). Treatment within 90 minutes increased from 28% to 71% (PϽ0.0001). Mean infarct size decreased (peak creatinine kinase, 2623Ϯ3329 versus 1517Ϯ1556 IU/L; Pϭ0.0089), as did hospital length of stay (5Ϯ7 versus 3Ϯ2 days; Pϭ0.0097) and total hospital costs per admission ($26 826Ϯ29 497 versus $18 280Ϯ8943; Pϭ0.0125). Conclusions—Emergency department physician activation of the catheterization laboratory and immediate transfer of the patient to an immediately available catheterization laboratory reduce door-to-balloon time, leading to a reduction in myocardial infarct size, hospital length of stay, and total hospital costs. (Circulation. 2007;116:67-76.) Key Words: angioplasty Ⅲ myocardial infarction Ⅲ quality of health care Ⅲ stents Ⅲ quality indicators, health care

mergency percutaneous intervention (PCI) is increas- Editorial p 6 ingly used in the management of ST-elevation myocar- E Clinical Perspective p 76 dial infarction (STEMI). The benefits of emergency PCI are time dependent, with door-to-balloon time delays associated with balloon time,8 leading some to suggest that future improve- increasing mortality.1 Therefore, consensus guidelines recom- ments in door-to-balloon time are unlikely.9 Interestingly, mend that STEMI patients achieve a door-to-balloon time of select hospitals have achieved improvements in door-to- Յ90 minutes.2 More recently, the American College of Cardi- balloon times, and recent studies have highlighted qualitative ology, American Heart Association, the Centers for Medicare characteristics unique to these institutions.10 More recently, a and Medicaid Services, and the Joint Commission on Accredi- survey of hospital strategies revealed that activation of the tation of Healthcare Organizations have all included door-to- catheterization laboratory by the emergency department phy- balloon time as a core hospital quality-of-care indicator.3–5 sician rather than cardiologist was associated with faster Despite the increased emphasis on achieving appropriate door-to-balloon times.11 Prior studies actually implementing door-to-balloon times, only 32% of patients overall in the emergency department physician activation have shown im- United States receive PCI within 90 minutes.6,7 In addition, provements in door-to-balloon time. One report revealed a there has been limited temporal improvement in door-to- reduction in median door-to-balloon from 88 to 61 minutes.12

Received November 20, 2006; accepted April 20, 2007. From the Indiana Heart Physicians, Indianapolis (U.N.K., M.B.K., S.R.S., W.J.B.); St. Francis Hospital and Health Centers, Beech Grove (M.L.J.); Curtis Ramsey and Associates, Indianapolis (C.S.); and Emergency Physicians of Indianapolis, Beech Grove (R.T.), Ind. Correspondence to Umesh N. Khot, MD, Indiana Heart Physicians/St. Francis Heart Center, 5330 E Stop 11 Rd, Indianapolis, IN 46237. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.677401

67 68 Circulation July 3, 2007

However, the results of that study were confounded by ing, or balloon pump) and ongoing cardiopulmonary resuscitation; simultaneous conversion of reperfusion strategy from a com- these patients were prepared for immediate transfer but transferred bination of thrombolytics and PCI to solely PCI. In addition, by the nursing staff with the cardiologist. The emergency department physician contacted the hospital operator to activate the catheteriza- the study excluded data from a 6-month transition period tion laboratory. The operator subsequently paged the cardiology between strategies. In a recent retrospective study, emergency physician assistant, catheterization laboratory coordinator, and crit- department physician activation reduced door-to-balloon time ical care unit nurse during regular hours or the on-call cardiologist, from 118 to 89 minutes.13 There are limited prospective data interventional cardiologist, critical care unit nurse, chest pain unit on the effect of adopting emergency department physician nurse, and on-call catheterization team during off-hours. On catheterization laboratory activation, the critical care unit nurse proceeded activation of the catheterization laboratory on door-to-balloon to the emergency department and subsequently transferred the time in centers already dedicated to primary PCI. We there- patient to the catheterization laboratory with the emergency depart- fore prospectively determined the impact on door-to-balloon ment nurse. During transfer, in case of sustained hypotension or time of emergency department physician activation of the arrest, the critical care unit nurse could administer dopamine or catheterization laboratory, combined with a novel strategy of norepinephrine intravenous drips, perform defibrillation, and request intubation by respiratory therapy, all without prior physician ap- immediate physical transfer of the patient to an immediately proval. The critical care unit modified the work requirements for the available catheterization laboratory by in-house nursing staff. EHART nurse by assigning the nurse 1 patient instead of 2 patients. To make the catheterization laboratory immediately available Methods during regular hours, the catheterization laboratory coordinator identified a catheterization room and staff for the patient. The Study Design catheterization laboratory coordinator could remove an elective The present prospective study was conducted between October 1, patient from the catheterization laboratory if the case had not started 2004, and June 26, 2006, at St Francis Hospital and Health Center (defined as cardiologist fully scrubbed at bedside obtaining access). (Beech Grove and Indianapolis, Ind), a 591-bed tertiary care com- If all rooms were occupied with cases in progress, then the STEMI munity hospital consisting of 2 campuses 7 miles apart (13-minute patient went to the first available room. On patient placement on the drive). Both campuses have emergency departments staffed with catheterization laboratory table, the EHART team members trans- emergency medicine residency–trained physicians; cardiology ser- ferred the patient’s nursing care to the catheterization team. vices are located within the Indianapolis campus. During the study To make the catheterization laboratory immediately available period, the hospital performed primary PCI for all STEMI patients during off-hours, the chest pain unit nurse proceeded to the cathe- presenting at either campus. A single 20-physician group provides terization laboratory, activated the catheterization laboratory imag- cardiology services at both campuses (Indiana Heart Physicians, ing equipment, and confirmed that the temporary pacemaker, balloon Indianapolis). Cardiologists take in-hospital call. Call consists of a pump, defibrillator, and activated clotting time machine were in noninterventional cardiologist on call with interventional cardiolo- working order. This individual subsequently assisted the critical care gist on backup call at home Ϸ80% of the time; the other 20% of the unit nurse and emergency department nurse in the initial setup of the time, an interventional cardiologist takes primary call. Call pattern patient, including placement on the catheterization table, monitoring was unchanged during the study period. Four catheterization staff equipment setup, prepping of groin, and assistance with the sterile members take home call during off-hours and are expected to arrive catheterization laboratory table. The emergency department nurse to the hospital within 30 minutes of laboratory activation. and critical care nurse monitored the patient until the third and fourth We prospectively enrolled consecutive patients who presented to catheterization staff members arrived and subsequently transferred either the Beech Grove or Indianapolis emergency department with nursing care to the catheterization team. If the patient was unstable, STEMI who received PCI within 24 hours of presentation.4,5 We all staff attended to the patient until safe transfer of care was excluded STEMI patients who were hospital inpatients at the time of possible. Table 1 compares the processes in the 2 time periods. diagnosis. All activities in the emergency department, during the transfer to the catheterization laboratory, and during initial setup in the cathe- Protocol During Cardiology Activation/Routine terization laboratory did not require cardiologist presence or input Transfer Period (October 1, 2004, Through August (see the order set available at www.stfrancishospitals.org/heart). The cardiologist evaluated the patient and determined the appropriateness 31, 2005) for emergency catheterization in the emergency department, en route Emergency department physicians requested immediate cardiology to the catheterization laboratory, or in the catheterization laboratory. evaluation for all STEMI patients. After patient evaluation, the cardiologist activated the catheterization laboratory by contacting the Study End Points and Statistical Analysis catheterization laboratory coordinator during regular hours (7 AM to The primary end point was median door-to-balloon time.3 Door time 5 PM weekdays) or the hospital operator during off-hours (weekends represented the arrival time at the initial emergency department. and 5 PM to 7 AM weekdays). During regular hours, the catheterization laboratory coordinator notified the emergency department to Secondary end points included the individual components of door- transfer the patient when a catheterization room became available. to-balloon time (ie, door-to-ECG time), infarct size measured by 14 During off-hours, transfer to the catheterization laboratory occurred peak creatinine kinase within the first 24 hours, hospital costs on arrival of 2 catheterization staff members. (total, direct, and indirect), hospital length of stay, and all-cause in-hospital mortality. Hospital cost data reflect the actual costs Protocol During Emergency Department involved in the delivery of care to each patient and were determined by the cost-accounting software of the hospital (Alliance for Deci- Activation/Immediate Transfer Period (September sion Support, Avega, El Segundo, Calif). Cost data for all patients 1, 2005, Through June 26, 2006) (including outliers) were analyzed. All-cause in-hospital mortality On September 1, 2005, at 7 AM, we implemented a protocol mandating was presented in unadjusted fashion. We determined the prevalence (1) emergency department physician activation of the catheterization of “false-positive” activation, defined as a patient sent to the laboratory and (2) immediate transfer of the patient to an immedi- catheterization laboratory by the emergency department physician ately available catheterization laboratory by an in-house emergency but who subsequently was determined to be an inappropriate heart attack response team (EHART) consisting of an emergency activation by the cardiologist. All patients provided informed con- department nurse, a critical care unit nurse, and a chest pain unit sent. Our institutional review board approved the study. nurse. The only exceptions to immediate transfer by the nursing staff Time values are presented as medians with interquartile ranges were hemodynamic compromise (requiring pressors, temporary pac- and were analyzed using 2-sample Wilcoxon rank sum tests. Other

Khot et al Improving Door-to-Balloon Time in STEMI 69

TABLE 1. Summary of Processes During the 2 Study Time Periods

Cardiology Activation Routine Transfer, October 1, ED Activation Immediate Transfer, September 1, 2004, Through August 31, 2005 2005, Through June 26, 2006 PCI as primary reperfusion strategy for all STEMI Yes Yes Routine availability of prehospital ECG in No No ambulances Activation of cath lab based on prehospital ECG No No Standing order for obtaining ECG in triage area Yes Yes ED physician activation of cath lab No Yes 1 Call activates cath lab* Yes Yes Immediate transfer of patient to immediately No Yes available cath lab Cath lab staff arrives within 30 min of activation Yes Yes Cardiologist in-house call Yes Yes Door-to-balloon time feedback to staff and Yes Yes physicians ED indicates emergency department; cath lab, catheterization laboratory. *In the first period, the 1 call to activate the cath lab was to the cath lab coordinator during daytime and through the operator at night. In the second time period, all calls went through the operator. continuous data are presented as meanϮSD and were analyzed by change, although there was a modest but statistically signif- 2-sample t tests. Categorical data are presented as proportions and icant decrease in sheath-to-balloon time. Mean infarct size Ͻ were analyzed by Fisher exact test. Values of P 0.05 were consid- and hospital length of stay decreased. Total hospital costs, ered statistically significant. Stata software was used for statistical analyses (version 8.2, Stata Corp, College Station, Tex). direct hospital costs, and indirect hospital costs all decreased. The authors had full access to and take responsibility for the The proportion of patients treated within 90 minutes integrity of the data. All authors have read and agree to the increased from 28% to 71% (PϽ0.0001) (Table 4). There was manuscript as written. a Ͼ2-fold increase in the treatment within 60 minutes and a nearly 10-fold reduction in treatment requiring Ͼ120 minutes Results (PϽ0.0001). From October 1, 2004, to August 30, 2005, 68 consecutive During the 10-month emergency department activation/ patients presented with STEMI, and 60 patients met the immediate transfer period, the prevalence of “false-positive” inclusion criteria (Figure 1). From September 1, 2005, to June activation by an emergency department physician was 1% (1 26, 2006, 96 consecutive patients presented with STEMI, and of 97). A patient with flash-pulmonary edema was misrouted 86 patients met the inclusion criteria. The proportions of to the catheterization laboratory instead of the critical care patients with normal/mild coronary artery disease (4.4% unit. This case occurred early in the implementation of the ϭ versus 5.2%; P 1), significant coronary artery disease program, and focused review indicated that emergency de- Ͼ ( 50% stenosis) managed medically (1.5% versus 1%; partment physician-cardiologist misunderstanding was the ϭ P 1), or coronary artery bypass grafting (2.9% versus 1%, root cause of the problem (patient not included in Figure 1). Pϭ0.57) were similar. Unadjusted all-cause in-hospital mortality was similar (intention to treat, 7.4% versus 5.2%, Discussion Pϭ0.74; post-PCI, 5% versus 4.7%, Pϭ1). Emergency department physician activation of the catheter- Demographics, initial presentation details, and treatments ization laboratory and immediate transfer of the patient to an were similar (Table 2). Emergency department physician immediately available catheterization laboratory by an in- catheterization laboratory activation increased from 0% to house nursing team led to a substantial reduction in door-to- 87% (PϽ0.0001). balloon time. This reduction was seen regardless of time of Median door-to-balloon time decreased from 113.5 to 75.5 day, ambulance or nonambulance presentation, and clinical minutes (PϽ0.0001) (Table 3). Reductions in door-to-balloon characteristics of the patient (Table 3). The success of our time were seen during regular hours, during off-hours, and protocol came from transforming a rigid system of stepwise with transfer from 1 campus to another (Table 3). Improve- serial processes into a parallel process system with nearly ments were seen regardless of gender, ambulance or nonam- simultaneous performance of catheterization laboratory acti- bulance presentation, or need for additional procedures (de- vation, physical transfer to catheterization laboratory, initial fibrillation, pacemaker, balloon pump) before PCI. Patients catheterization laboratory setup, and cardiology evaluation presenting during regular hours had a median door-to-balloon (Figure 2). In addition, our data confirm a recent survey of time of 45 minutes. hospital practices highlighting the importance of emergency The most substantial decrease occurred in time spent in the department physician activation of the catheterization labo- emergency department and in transportation to the catheter- ratory and add an immediate transfer process to the list of ization laboratory (Table 3). Door-to-ECG and catheteriza- strategies that reduce door-to-balloon time.11 Finally, the tion laboratory–to–sheath placement times showed no present study is one of the first prospective studies to reveal 70 Circulation July 3, 2007

Figure 1. Summary of enrollment and outcomes in study during the 2 time periods. PCI denotes percutaneous intervention. Signiﬁcant coronary artery disease indicates a stenosis Ͼ50%. ED indicates emergency department; CCU, coronary care unit; CAD, coronary artery disease; and CABG, coronary artery bypass grafting. that decreasing door-to-balloon time leads to decreased in- mended measures have been multiple and often complex, farct size, length of stay, and hospital costs. typically requiring months to years for full implementa- Providing timely emergency PCI is a complex undertaking tion.18,19 In contrast, the present study showed that 2 simple, demanding rapid coordination of care by multiple physicians, focused modifications rapidly reduced door-to-balloon time nurses, and hospital staff. In prior reports, an audit process, and could be implemented rapidly within a day. in-depth continuous quality control improvement analysis, Our protocol was resisted initially because of concerns that and multidisciplinary quality initiatives have all improved emergency department physician activation of the catheter- door-to-balloon time.15–19 However, the specific steps recom- ization laboratory would not reduce door-to-balloon time mended have varied between these studies, making it chal- since our cardiologists took in-house call. However, this lenging for other institutions to adopt specific protocols to simple change allowed catheterization laboratory staff to improve their door-to-balloon time. In addition, the recom- arrive 20 to 40 minutes earlier (Table 3), indicating that there

Khot et al Improving Door-to-Balloon Time in STEMI 71

TABLE 2. Demographics, Initial Presentation Characteristics, and Treatment Outcomes

Cardiology Activation Routine Transfer, October 1, ED Activation Immediate Transfer, September 1, 2004, Through August 31, 2005 (nϭ60) 2005, Through June 26, 2006 (nϭ86) P Demographics Age, y 58Ϯ13 60Ϯ13 0.31 Female gender 17 (28.3) 25 (29.1) 1 Health insurance Private 28 (46.7) 48 (55.8) 0.51 Medicare 21 (35) 29 (33.7) ⅐⅐⅐ Medicaid 3 (5) 3 (3.5) ⅐⅐⅐ Self-pay 8 (13.3) 6 (7) ⅐⅐⅐ Medical history Current smoker 31 (51.7) 48 (55.8) 0.74 Diabetes 10 (16.7) 17 (19.8) 0.67 Hypertension 34 (56.7) 46 (53.5) 0.74 Hypercholesterolemia 19 (31.7) 31 (36.1) 0.60 Family history of CHD 22 (36.7) 28 (32.6) 0.72 Congestive heart failure 0 (0) 0 (0) ⅐⅐⅐ COPD 4 (6.7) 9 (10.5) 0.56 Prior PCI 10 (16.7) 23 (26.7) 0.17 Prior CABG 5 (8.3) 5 (5.8) 0.74 PVD 3 (5) 8 (9.3) 0.53 Stroke 0 (0) 5 (5.8) 0.08 Initial presentation Regular hours 26 (43.3) 30 (34.9) 0.39 Transferred for PCI 12 (20) 22 (25.6) 0.55 Symptom onset to arrival Յ1 h 20 (33.3) 39 (45.4) 0.40 Ͼ1–2 h 14 (23.3) 22 (25.6) ⅐⅐⅐ Ͼ2–6 h 12 (20) 8 (9.3) ⅐⅐⅐ Ͼ6–12 h 4 (6.7) 4 (4.7) ⅐⅐⅐ Ͼ12 h 6 (10) 6 (7) ⅐⅐⅐ Unknown 4 (6.7) 7 (8.1) ⅐⅐⅐ Chest pain at presentation 54 (90) 71 (82.6) 0.24 Prehospital ECG 2 (3.3) 7 (8.1) 0.31 Ambulance arrival 27 (45) 31 (36.1) 0.31 Heart rate, bpm 85Ϯ21 79Ϯ23 0.10 Systolic blood pressure, mm Hg 137Ϯ26 137Ϯ34 0.99 Diastolic blood pressure, mm Hg 83Ϯ19 83Ϯ22 0.89 Location of infarct Anterior 24 (40) 26 (30.2) 0.29 Inferior 34 (56.7) 56 (65.1) 0.39 Lateral (isolated) 2 (3.3) 4 (4.7) 1 LBBB 2 (3.3) 0 (0) 0.17 ECG leads with ST-elevation 2 16 (27.1) 19 (22.1) 0.80 3–4 29 (49.2) 45 (52.3) ⅐⅐⅐ Ն5 14 (23.7) 22 (25.6) ⅐⅐⅐ Cardiogenic shock 4 (6.7) 5 (5.8) 1 Cath lab activation ED physician 0 (0) 75 (87.2) Ͻ0.0001 Cardiologist 60 (100) 11 (12.8) ⅐⅐⅐ 72 Circulation July 3, 2007

TABLE 2. Continued

Cardiology Activation Routine Transfer, October 1, ED Activation Immediate Transfer, September 1, 2004, Through August 31, 2005 (nϭ60) 2005, Through June 26, 2006 (nϭ86) P Treatment Aspirin 57 (95) 85 (98.8) 0.31 ␤-Blocker 53 (88.3) 75 (87.2) 1 Heparin 59 (98.3) 86 (100) 0.41 Glycoprotein IIb/IIIa inhibitor 60 (100) 83 (96.5) 0.27 Defibrillation before PCI 5 (8.3) 8 (9.3) 1 Temporary pacemaker before 4 (6.7) 3 (3.5) 0.45 PCI IABP before PCI 1 (1.7) 1 (1.2) 1 Catheterization results Infarct-related artery Left main 2 (3.3) 0 (0) 0.32 Left anterior descending 23 (38.3) 33 (38.4) ⅐⅐⅐ Left circumflex 5 (8.3) 13 (15.1) ⅐⅐⅐ Right coronary 27 (45) 38 (44.2) ⅐⅐⅐ Bypass graft 3 (5) 2 (2.3) ⅐⅐⅐ Treatment Balloon angioplasty only 10 (16.6) 11 (12.8) 0.63 Balloon angioplasty/stent 50 (83.3) 75 (87.2) ⅐⅐⅐ Type of stent Bare metal stent 11 (22) 13 (17.3) 0.64 Drug-eluting stent 39 (78) 62 (82.7) ⅐⅐⅐ Values are expressed as meanϮSD or n (%). ED indicates emergency department; CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; CABG, coronary artery bypass grafting; PVD, peripheral vascular disease; cath lab, catheterization laboratory; and IABP, intra-aortic balloon pump. are considerable delays associated with even waiting for addressing the transfer process was underscored by a in-house cardiology evaluation. In addition, although emer- report showing that simply preparing patients for transfer gency department physician interpretation of ST-segment to the catheterization laboratory reduced emergency de- elevation typically is accurate,20 our cardiologists were con- partment time.15 Although a recent survey of hospital cerned about inappropriate activation of the catheterization practices did not identify transfer process as a factor in laboratory by the emergency department physicians. We door-to-balloon time, this is likely related to the fact that overcame this resistance by emphasizing that the decision to 93% of hospitals practiced in a manner similar to the activate the laboratory and the decision to perform catheter- routine transfer period of the present study and required ization and intervention were distinct. Thus, the final decision the catheterization laboratory to notify the emergency regarding appropriateness for emergency catheterization re- department when it was ready before transfer. In fact, none mained with cardiology, and our cardiologists were instructed of the hospitals in the survey had an immediate transfer to perform catheterization only if they agreed with the policy in place.11 The present study shows that the decision emergency department assessment. We further audited all to transfer the patient should originate in the emergency cases for appropriateness and provided this information to the department and that the catheterization laboratory should emergency department physicians and cardiologists. Ulti- be immediately prepared to receive the patient. mately, inappropriate activations occurred rarely11 and were Any transfer of a critically ill patient within the hospital is eventually accepted as necessary to improve the overall care of STEMI patients. associated with a potential risk of decompensation or adverse The largest component of door-to-balloon time is the event during transfer, and there was concern regarding the time spent within the emergency department and transfer- safety of transfer without a cardiologist. To maximize patient ring to the catheterization laboratory.15,18 Transfer out of safety during transfer, the transfer team included a critical the emergency department to the catheterization laboratory care nurse who had standing orders for events such as often is hampered by structural impediments such as strict unstable heart rhythms or hypotension. We also excluded requirements for cardiology consultation or catheterization from immediate transfer patients with ongoing cardiopulmo- laboratory readiness before transfer.11 Furthermore, during nary resuscitation or hemodynamic instability. In addition, day hours, the catheterization laboratory can be occupied because our cardiologists were in-house, the time without by elective cases, further impeding the STEMI patient’s cardiologist presence was minimized. Finally, even in criti- access to the catheterization laboratory. The importance of cally ill patients, proceeding to the catheterization laboratory Khot et al Improving Door-to-Balloon Time in STEMI 73

TABLE 3. Primary and Secondary End Points

Cardiology Activation Routine ED Activation Immediate Transfer, October 1, 2004, Transfer, September 1, Through August 31, 2005, Through June 26, No. 2005 (nϭ60) No. 2006 (nϭ86) P Primary end point door-to-balloon time, min All patients 60 113.5 (83, 143) 86 75.5 (64, 94) Ͻ0.0001 Regular hours 26 83.5 (63, 129) 30 64.5 (42, 85) 0.0048 Off-hours 34 123.5 (108, 157) 56 77.5 (69.5, 100) Ͻ0.0001 All patients, excluding outside transfers 47 109 (79, 130) 64 73.5 (54, 91) Ͻ0.0001 Regular hours 21 79 (61, 125) 23 45 (39, 75) 0.0019 Off-hours 26 120.5 (106, 139) 41 76 (68, 94) Ͻ0.0001 Outside transfer patients 13 147 (114, 157) 22 85 (75, 98) 0.0006 Regular hours 5 114 (83, 134) 7 87 (78, 98) 0.3709 Off-hours 8 152.5 (133.5, 239.5) 15 84 (74, 107) 0.0004 Female 17 139 (122, 149) 25 84 (74, 99) 0.0004 Male 43 108 (73, 126) 61 74 (61, 89) Ͻ0.0001 Ambulance arrival 27 106 (67, 125) 31 71 (44, 84) 0.0007 Nonambulance arrival 33 125 (97, 149) 55 78 (68, 99) Ͻ0.0001 Defibrillation, temporary pacemaker, or IABP before PCI 9 134 (114, 153) 11 75 (68, 77) 0.0002 Secondary end points Door-to-balloon components, min Door to ECG 60 5 (1, 9) 86 4 (1, 6) 0.2328 ECG to cath lab arrival, overall 60 71 (45, 88) 86 39 (25, 54) Ͻ0.0001 Regular hours 26 48.5 (36, 75) 30 27.5 (19, 52) 0.0006 Off hours 34 81.5 (66, 107) 56 41.5 (33, 54.5) Ͻ0.0001 ECG to cath lab arrival, excluding outside transfer patients 47 67 (43, 83) 64 33 (21.5, 47.5) Ͻ0.0001 Regular hours 21 44 (36, 69) 23 25 (16, 31) Ͻ0.0001 Off hours 26 77.5 (58, 97) 41 37 (28, 50) Ͻ0.0001 ECG to cath lab arrival, outside transfer patients 13 85 (78, 108) 22 55 (43, 62) 0.0012 Regular hours 5 78 (34, 79) 7 56 (42, 70) 0.5698 Off hours 8 100 (81.5, 199) 15 52 (43, 61) 0.0002 Cath lab arrival to sheath placement 60 16 (10.5, 22) 86 16 (10, 21) 0.3833 Sheath placement to balloon 60 17.5 (12.5, 25) 86 13 (9, 18) 0.0045 ECG to first cath lab staff arrival off-hours, overall 30 51 (42, 70) 51 30 (24, 39) Ͻ0.0001 Excluding Outside Transfers 23 48 (35, 58) 39 28 (24, 39) 0.0003 Outside Transfers Only 7 74 (59, 157) 12 32 (27.5, 38) 0.0007 ECG to second cath lab staff arrival off-hours, overall 30 56.5 (46, 75) 51 34 (28, 42) Ͻ0.0001 Excluding outside transfers 23 54 (39, 63) 39 32 (27, 42) 0.0011 Outside transfers only 7 75 (70, 170) 12 36 (30.5, 42) 0.0005 Mean infarct size, peak creatinine kinase, IU/L 60 2623Ϯ3329 83 1517Ϯ1556 0.0089 Mean total hospital costs, $ 60 26 826Ϯ29 497 86 18 280Ϯ8943 0.0125 Mean direct hospital costs, $ 60 19 585Ϯ21 946 86 13 060Ϯ6438 0.0102 Mean indirect hospital costs, $ 60 7240Ϯ7571 86 5220Ϯ2518 0.0228 Mean hospital length of stay, d 60 5Ϯ7863Ϯ2 0.0097 Mean time in coronary care unit, h 60 68Ϯ86 86 48Ϯ37 0.0574 All-cause in-hospital mortality, intention to treat 68 5 (7.4) 96 5 (5.2) 0.74 All-cause in-hospital mortality post-PCI 60 3 (5) 86 4 (4.7) 1 All time values are median (25th and 75th percentiles). IABP indicates intra-aortic balloon pump; cath lab, catheterization laboratory. as soon as possible was believed to allow more timely The present study included patients who were transferred delivery of lifesaving interventions, outweighing the potential for emergency PCI from another emergency department at risk of transfer. our other campus. Such transfer patients are typically ex- 74 Circulation July 3, 2007

TABLE 4. Proportion of Patients Treated Within Various Time Points

Cardiology Activation/ Routine Transfer, October ED Activation/Immediate Transfer, September 1, Door-to-Balloon (min) 1, 2004, Through August 31, 2005 (nϭ60) 2005, Through June 26, 2006 (nϭ86) P Ͻ60 5 (8.3) 17 (19.8) Ͻ0.0001 60–90 12 (20) 44 (51.2) 91–120 16 (26.7) 20 (23.3) Ͼ120 27 (45) 5 (5.8) Values are expressed as n (%). cluded from most analyses15,18 and are specifically excluded maintaining the delivery of care by a highly trained catheter- from public reporting of quality indicators despite having the ization staff. longest door-to-balloon times.4,5 In the National Registry of Myocardial Infarction, transfer patients had a median door- Study Limitations to-balloon time of 180 minutes, with only 4.2% achieving Although the 2 cohorts were similar, baseline differences reperfusion within 90 minutes.7 With our new protocol, the cannot be completely accounted for between the 2 time median door-to-balloon time decreased to 85 minutes, and periods because of the nonrandomized nature of the present 62% of these patients were treated within 90 minutes. Thus, study. Nevertheless, the present study design reflects the our protocol leads to improvement in the care of STEMI “real-life” manner in which physicians and hospitals imple- patients transferred directly for PCI. ment process improvements and is similar to other process Most patients undergoing emergency PCI present during improvement studies.21 The present study was performed in a off-hours, and only 26% undergo reperfusion in Յ90 min- community hospital setting with cardiologists taking in-house utes.6 It has been suggested that hospitals that perform PCI call. However, the protocol could be adapted for cardiologists have catheterization staff on site 24 hours to ensure timely taking home call only or for the inclusion of residents and revascularization or to cross-train in-house staff to perform fellows in academic settings. Our results could be explained catheterization staff duties.6 However, even in high-volume by increased attention to door-to-balloon time; however, our centers, 24-hour coverage would be prohibitively expensive, results during the baseline time period were widely presented and maintaining proficiency of cross-trained staff exposed to to physician and hospital staff with no door-to-balloon time nighttime myocardial infarction cases only would be chal- improvement. Our transfer patients traveled a 7-mile dis- lenging. Our use of an in-house transfer team allowed us to tance, and our results may not be applicable to longer transfer substantially improve the care of off-hours patients while distances. Emergency medicine residency–trained physicians

Figure 2. Serial vs parallel processing in achieving door-to-balloon time. Simultaneous performance of catheterization laboratory activation, physical transfer to catheterization laboratory, initial catheterization laboratory setup, and cardiology evaluation leads to a reduction in door-to-balloon time. Khot et al Improving Door-to-Balloon Time in STEMI 75 staffed our emergency departments, and staffing by a differ- Force on Practice Guidelines (Committee to Revise the 1999 Guidelines ent mix of physicians may not be able to duplicate our results. for the Management of Patients With Acute Myocardial Infarction). JAm Coll Cardiol. 2004;44:E1–E211. Our results occurred with limited availability of prehospital 3. Krumholz HM, Anderson JL, Brooks NH, Fesmire FM, Lambrew CT, ECG, but we believe the use of ECGs is complementary and Landrum MB, Weaver WD, Whyte J, Bonow RO, Bennett SJ, Burke G, could further reduce door-to-balloon time. However, 50% of Eagle KA, Linderbaum J, Masoudi FA, Normand SL, Pina IL, Radford STEMI patients nationwide do not present by ambulance, MJ, Rumsfeld JS, Ritchie JL, Spertus JA. ACC/AHA clinical performance measures for adults with ST-elevation and non-ST-elevation underscoring the importance of protocols that improve the myocardial infarction: a report of the American College of Cardiology/ care of all patients.22 American Heart Association Task Force on Performance Measures Emergency department activation of the catheterization (Writing Committee to Develop Performance Measures on ST-Elevation laboratory and immediate transfer of the patient to an imme- and Non-ST-Elevation Myocardial Infarction). J Am Coll Cardiol. 2006; 47:236–265. diately available catheterization laboratory by an in-house 4. Centers for Medicare & Medicaid Services and the Joint Commission on transfer team are 2 specific measures that allow the delivery Accreditation of Healthcare Organizations. Specifications Manual for of PCI in a timely fashion to a broad population of patients. National Hospital Quality Measures. Baltimore, Md: Centers for Medicare and Medicaid Services and the Joint Commission on Accredi- Additional benefits include reductions in myocardial infarct tation of Healthcare Organizations; 2006. Available at: http:// size, hospital length of stay, and total hospital costs. Wide- qnetexchange.org/public/hdc.do?hdcPageϭhosp_quality_manual. spread implementation of this simple strategy can substan- Accessed April 7, 2006. tially improve the quality of care of STEMI patients under- 5. Williams SC, Schmaltz SP, Morton DJ, Koss RG, Loeb JM. Quality of care in U.S. hospitals as reflected by standardized measures, 2002–2004. going emergency PCI. An electronic copy of the order set N Engl J Med. 2005;353:255–264. used in the EHART protocol is available at www. 6. Magid DJ, Wang Y, Herrin J, McNamara RL, Bradley EH, Curtis JP, stfrancishospitals.org/heart. Pollack CV Jr, French WJ, Blaney ME, Krumholz HM. Relationship between time of day, day of week, timeliness of reperfusion, and in-hospital mortality for patients with acute ST-segment elevation myo- Appendix cardial infarction. JAMA. 2005;294:803–812. The following individuals participated in the present study. 7. Nallamothu BK, Bates ER, Herrin J, Wang Y, Bradley EH, Krumholz Indiana Heart Physicians, Indianapolis: A. Akinwande, M. HM. Times to treatment in transfer patients undergoing primary percu- Barron, J. Christie, H. Genovely, J. Graham, D. Hadian, M. taneous coronary intervention in the United States: National Registry of Myocardial Infarction (NRMI)-3/4 analysis. Circulation. 2005;111: Jones, S. Karanam, D. Kovacich, I. Labin, S. Lall, J. Mossler, 761–767. G. Revytak, and R. Shea. Emergency Physicians of Indianap- 8. McNamara RL, Herrin J, Bradley EH, Portnay EL, Curtis JP, Wang Y, olis, Beech Grove, Ind: S. Antoine, D. Blank, S. Boha, M. Magid DJ, Blaney M, Krumholz HM. Hospital improvement in time to reperfusion in patients with acute myocardial infarction, 1999 to 2002. Brown, W. Corbett, D. Debikey, B. Dillman, K. Ernsting, M. J Am Coll Cardiol. 2006;47:45–51. Fitzpatrick, G. Godfrey, C. Hartman, B. Johnston, S. Kistler, 9. National Heart Attack Alert Program Coordinating Committee. National H. Levitin, R. Mara, W. McDaniel, E. Olson, M. Overfelt, M. Heart Attack Alert Program Coordinating Committee and Subcommittees Russell, A. Stern, M. Stone, and E. Weinstein. Meeting Summary Reports. In: Abstracts of the National Heart Attack Alert Program 10-Year Anniversary Meeting; June 25–26, 2001; Alex- andria, Va: 10. Acknowledgments 10. Bradley EH, Curry LA, Webster TR, Mattera JA, Roumanis SA, Radford We are indebted to Mechelle L. Peck, RN; Diana L. Brown, RN; MJ, McNamara RL, Barton BA, Berg DN, Krumholz HM. Achieving Mark Manning, CCT; Patricia L. Wray, RN; Stephen H. Kliman, rapid door-to-balloon times: how top hospitals improve complex clinical MD; Juan E. Weksler, MD; Carl L. Rouch, MD; and Horace O. systems. Circulation. 2006;113:1079–1085. Hickman, MD, who were intimately involved in the implementation 11. Bradley EH, Herrin J, Wang Y, Barton BA, Webster TR, Mattera JA, of the present study. Roumanis SA, Curtis JP, Nallamothu BK, Magid DJ, McNamara RL, Parkosewich J, Loeb JM, Krumholz HM. Strategies for reducing the door-to-balloon time in acute myocardial infarction. N Engl J Med. Sources of Funding 2006;355:2308–2320. Indiana Heart Physicians and St Francis Hospital and Health Centers 12. Thatcher JL, Gilseth TA, Adlis S. Improved efficiency in acute myo- provided funding for the present study. The funding organizations cardial infarction care through commitment to emergency department- had no role in the design and conduct of the study; collection, initiated primary PCI. J Invasive Cardiol. 2003;15:693–698. management, analysis, and interpretation of the data; and prepara- 13. Jacoby J, Axelband J, Patterson J, Belletti D, Heller M. Cardiac cathe- tion, review, or approval of the manuscript. terization lab activation by the emergency physician without prior consultation decreases door-to-balloon time. J Invasive Cardiol. 2005;17: Disclosures 154–155. 14. Haase J, Bayar R, Hackenbroch M, Storger H, Hofmann M, Schwarz CE, None. Reinemer H, Schwarz F, Ruef J, Sommer T. Relationship between size of myocardial infarctions assessed by delayed contrast-enhanced MRI after References primary PCI, biochemical markers, and time to intervention. J Interv 1. Berger PB, Ellis SG, Holmes DR, Jr., Granger CB, Criger DA, Betriu A, Cardiol. 2004;17:367–373. Topol EJ, Califf RM. Relationship between delay in performing direct 15. Ward MR, Lo ST, Herity NA, Lee DP, Yeung AC. Effect of audit on coronary angioplasty and early clinical outcome in patients with acute door-to-inflation times in primary angioplasty/stenting for acute myo- myocardial infarction: results from the Global Use of Strategies to Open cardial infarction. Am J Cardiol. 2001;87:336–338, A339. Occluded Arteries in Acute Coronary Syndromes (GUSTO-IIb) trial. 16. Shry EA, Eckart RE, Winslow JB, Rollefson WA, Simpson DE. Effect of Circulation. 1999;100:14–20. monitoring of physician performance on door-to-balloon time for primary 2. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, angioplasty in acute myocardial infarction. Am J Cardiol. 2003;91: Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, 867–869. Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL, 17. Caputo RP, Ho KK, Stoler RC, Sukin CA, Lopez JJ, Cohen DJ, Kuntz Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka RE, Berman A, Carrozza JP, Baim DS. Effect of continuous quality LF, Hunt SA, Jacobs AK, Ornato JP. ACC/AHA guidelines for the improvement analysis on the delivery of primary percutaneous trans- management of patients with ST-elevation myocardial infarction: a report luminal coronary angioplasty for acute myocardial infarction. of the American College of Cardiology/American Heart Association Task Am J Cardiol. 1997;79:1159–1164.

76 Circulation July 3, 2007

18. Caputo RP, Kosinski R, Walford G, Giambartolomei A, Grant W, Reger 21. Eagle KA, Montoye CK, Riba AL, DeFranco AC, Parrish R, Skorcz S, MJ, Simons A, Esente P. Effect of continuous quality improvement Baker PL, Faul J, Jani SM, Chen B, Roychoudhury C, Elma MA, Mitchell analysis on the delivery of primary percutaneous revascularization for KR, Mehta RH. Guideline-based standardized care is associated with acute myocardial infarction: a community hospital experience. Catheter substantially lower mortality in Medicare patients with acute myocardial Cardiovasc Interv. 64:428–433, 2005. infarction: the American College of Cardiology’s Guidelines Applied in 19. Zarich SW, Sachdeva R, Fishman R, Werdmann MJ, Parniawski M, Practice (GAP) projects in Michigan. J Am Coll Cardiol. 2005;46: Bernstein L, Dilella M. Effectiveness of a multidisciplinary quality 1242–1248. improvement initiative in reducing door-to-balloon times in primary 22. Canto JG, Zalenski RJ, Ornato JP, Rogers WJ, Kiefe CI, Magid D, angioplasty. J Interv Cardiol. 2004;17:191–195. Shlipak MG, Frederick PD, Lambrew CG, Littrell KA, Barron HV. Use 20. Brady WJ, Perron A, Ullman E. Errors in emergency physician interpre- of emergency medical services in acute myocardial infarction and sub- tation of ST-segment elevation in emergency department chest pain sequent quality of care: observations from the National Registry of patients. Acad Emerg Med. 2000;7:1256–1260. Myocardial Infarction 2. Circulation. 2002;106:3018–3023.

CLINICAL PERSPECTIVE The relationship between delays in door-to-balloon time and increased mortality is well established and has led to consensus guidelines and hospital quality-of-care programs recommending that ST-elevation myocardial infarction patients achieve a door-to-balloon time of Յ90 minutes. However, most patients do not receive treatment within the recommended 90 minutes, and there are limited prospective data on specific measures to significantly reduce door-to-balloon time. In the present study, we prospectively determined the impact on median door-to-balloon time of a protocol mandating (1) emergency department physician activation of the catheterization laboratory and (2) immediate transfer of the patient to an immediately available catheterization laboratory by an in-house transfer team consisting of an emergency department nurse, a critical care unit nurse, and a chest pain unit nurse. With this protocol, median door-to-balloon time fell substantially, and this reduction was seen regardless of time of day, ambulance or nonambulance presentation, and clinical characteristics of the patients. In addition, patients transferred from an outside affiliated emergency department to our main hospital also saw substantial reductions in door-to-balloon time. Treatment within 90 minutes increased from 28% to 71%. Other benefits included reductions in mean infarct size, hospital length of stay, and total hospital costs per admission. The present study identifies a simple and widely applicable strategy that leads to improvements in both door-to-balloon time and the quality of care of ST-elevation myocardial infarction patients undergoing emergency percutaneous intervention.

The Brain–Heart Connection

Martin A. Samuels, MD

eurocardiology has many dimensions, but it may be (1) the impact of the collapse or death of a close person; (2) Nconceptualized as divided into 3 major categories: the during acute grief; (3) on threat of loss of a close person; (4) heart’s effects on the brain (eg, cardiac source embolic during mourning or on an anniversary; (5) on loss of status or stroke), the brain’s effects on the heart (eg, neurogenic heart self-esteem; (6) personal danger or threat of injury; (7) after disease), and neurocardiac syndromes (eg, Friedreich dis- danger is over; (8) reunion, triumph, or happy ending. ease). The present review deals with the nervous system’s Common to all is that they involve events impossible for the capacity to injure the heart. This subject is inherently impor- victim to ignore and to which the response is overwhelming tant but also represents an example of a much more wide- excitation, giving up, or both. spread and conceptually fascinating area of neurovisceral In 1957, Curt Richter reported on a series of experiments damage in general. aimed to elucidate the mechanism of Cannon’s “voodoo” death.3 Richter, a former student of Cannon, pursued an History of Learning the Nature of the incidental discovery of an epidemic of sudden death in a Brain–Heart Connection colony of rodents, which was induced when a colleague, In 1942, at the culmination of his distinguished career as Gordon Kennedy, had clipped the whiskers of the animals to Professor of Physiology at Harvard Medical School, Walter prevent contamination of the urine collection. Richter studied B. Cannon published a remarkable paper entitled “‘Voodoo’ the length of time that domesticated rats could swim at Death,”1 in which he recounted anecdotal experiences, various water temperatures and found that at a water temper- largely from the anthropology literature, of death from fright. ature of 93°C these rats could swim for 60 to 80 minutes. These often remote events, drawn from widely disparate parts However, if the animal’s whiskers were trimmed, it would of the world, had several features in common. They were all invariably drown within a few minutes. When carrying out induced by an absolute belief that an external force, such as similar experiments with fierce wild rats, he noted that a a wizard or medicine man, could, at will, cause demise and number of factors contributed to the tendency for sudden that the victim himself had no power to alter this course. This death, the most important of which was restraint, which perceived lack of control over a powerful external force is the involved holding the animals and confinement in a glass sine qua non for all the cases recounted by Cannon, who swimming jar with no chance of escape. By trimming the postulated that death was caused “by a lasting and intense rats’ whiskers, a procedure that destroys possibly their most action of the sympathico-adrenal system.” Cannon believed important proprioceptive mechanism, the tendency for early that this phenomenon was limited to societies in which the demise was exacerbated. In the case of the calm domesticated people were “so superstitious, so ignorant, that they feel animals in which restraint and confinement were apparently themselves bewildered strangers in a hostile world. Instead of not significant stressors, removal of whiskers rendered these knowledge, they have fertile and unrestricted imaginations animals as fearful as wild rats with a corresponding tendency which fill their environment with all manner of evil spirits for sudden death. ECGs taken during the process showed capable of affecting their lives disastrously.” Over the years development of a bradycardia prior to death, and adrenalec- since Cannon’s observations, evidence has accumulated to tomy did not protect the animals. Furthermore, atropine support his concept that “voodoo” death is, in fact, a real protected some of the animals, and cholinergic drugs led to an phenomenon but, far from being limited to ancient peoples, even more rapid demise. All this was taken as evidence that may be a basic biological principle that provides an important overactivity of the sympathetic nervous system was not the clue to understanding the phenomenon of sudden death in cause of the death but rather it was caused by increased vagal modern society as well as providing a window into the world tone. of neurovisceral disease (also known as psychosomatic We now believe that the apparently opposite conclusions illness). of Cannon and Richter are not mutually exclusive, but rather George Engel collected 160 accounts from the lay press of that a generalized autonomic storm, which occurs as a result sudden death that were attributed to disruptive life events.2 of a life-threatening stressor, will have both sympathetic and He found that such events could be divided into 8 categories: parasympathetic effects. The apparent predominance of one

From the Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass. Correspondence to Dr Martin A. Samuels, Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis St, Boston, MA 02115. E-mail [email protected] (Circulation. 2007;116:77-84.) © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI:10.1161/CIRCULATIONAHA.106.678995 77

a 78 Circulation July 3, 2007 over the other depends on the parameter measured (eg, heart autonomic nervous system. The authors argued that Brod- rate, blood pressure) and the timing of the observations in mann area 13 on the orbital surface of the frontal lobe and relation to the stressor (eg, early events tend to be dominated area 24 on the anterior cingulate gyrus were the cortical by sympathetic effects, whereas late events tend to be centers for cardiovascular control. dominated by parasympathetic effects). Cerebral hemispheral There is clear evidence that cardiac lesions can be pro- dominance with regard to autonomic control (right predom- duced as the result of nervous system disease. The concept of inantly sympathetic and left predominantly parasympathetic) visceral organ dysfunction that occurs as a result of neuro- probably also contributes to the dominant mechanism of logical stimuli can be traced to Ivan Pavlov. Hans Selye, a sudden death (ie, sympathetic versus vagal) in a given student of Pavlov, described electrolyte–steroid–cardiopathy person.4 with necroses (ESCN).9 Selye’s view was that this cardiac In human beings, one of the easily accessible windows into lesion was common and often described by different names in autonomic activity is the ECG. Edwin Byer and colleagues the literature. He argued that this lesion was distinct from the reported 6 patients whose ECGs showed large upright T coagulation necrosis that occurred as a result of ischemic waves and long QT intervals.5 Two of these patients had disease, but that it could exist in the same heart. Selye hypertensive encephalopathy, 1 patient had a brain stem demonstrated that certain steroids and other hormones created stroke with neurogenic pulmonary edema, 1 patient had an a predisposition for the development of ESCN, but that other intracerebral hemorrhage, 1 patient had a postpartum ische- factors were required for ESCN to develop. The most mic stroke possibly related to toxemia of pregnancy, and 1 effective conditioning steroid was 2-␣-methyl-9-␣- patient had no medical history except a blood pressure of fluorocortisol. Among the factors that led to ESCN in 210/110 mm Hg. On the basis of experimental results of steroid-sensitized animals were certain electrolytes (eg, cooling or warming the endocardial surface of a dog’s left NaH2PO4), various hormones (eg, vasopressin, adrenaline, ventricle, Byer and colleagues concluded that these ECG insulin, thyroxine), certain vitamins (eg, dihydrotachysterol), changes were caused by subendocardial ischemia. Harold cardiac glycosides, surgical interventions (eg, cardiac reper- Levine reported on several disorders other than ischemic fusion after ischemia), and psychic or nervous stimuli (eg, heart disease that could produce ECG changes reminiscent of restraint, fright). The cardiac lesions could not be prevented coronary disease.6 Among these was a 69-year-old woman by adrenalectomy, which suggests that the process, if related who was admitted and remained in coma. Her admission to sympathetic hyperactivity, must exert its influence by ECG showed deeply inverted T waves in the anterior and direct neural connection to the heart rather than by a blood- lateral precordial leads. Two days later, it showed ST seg- borne route. ment elevation with less deeply inverted T waves, a pattern Cardiac lesions may be produced in rats by pretreatment suggestive of myocardial infarction. However, at autopsy a with either 2-␣-methyl-9-␣-fluorohydrocortisone (fluorocor- ruptured berry aneurysm was found and no evidence of tisol), dihydrotachysterol (calciferol), or thyroxine and then myocardial infarction or pericarditis was noted. Levine did restraint of the animals on a board for 15 hours or with cold not propose a specific mechanism but referred to experimen- stress.10 Agents that act by inhibition of the catecholamine- tal work on the production of cardiac arrhythmias by basal mobilizing reflex arc at the hypothalamic level (eg, chlor- ganglia stimulation and ST and T-wave changes induced by promazine) or by blockade of only the circulating but not the injection of caffeine into the cerebral ventricle. neurogenic intramyocardial catecholamines (eg, dibenamine) George Burch and colleagues7 reported on 17 patients who were the least effective for protection of cardiac muscle, were said to have “cerebrovascular accidents” (ie, strokes). In whereas those drugs that act by ganglionic blockade (eg, 14 of the 17, hemorrhage was demonstrated by lumbar mecamylamine) or by direct intramyocardial catecholamine- puncture. It is not possible to determine which of these depletion (eg, reserpine) were the most effective. Further- patients had hemorrhagic infarction, intracerebral hemor- more, it is clear that blood catecholamine levels are often rhage, and subarachnoid hemorrhage, and no data about the normal but that identical ECG findings are seen with high territory of the strokes are available. The essential features of systemic catecholamines. These clinical and pharmacological the ECG abnormalities were: (1) long QT intervals in all data support the concept that the cardiac necrosis is caused by patients; (2) large, usually inverted, T waves in all patients; catecholamine toxicity and that catecholamines released di- and (3) U waves in 11 of the 17 patients.7 rectly into the heart via neural connections are much more Cropp and Manning8 reported on the details of ECG toxic than those that reach the heart via the bloodstream, abnormalities in 29 patients with subarachnoid hemorrhage. though clearly the 2 routes could be additive in the intact, Twenty-two of these patients survived. Two of those who nonadrenalectomized animal. Intracoronary infusions of epi- died had no postmortem examination, which left 5 patients in nephrine reproduce the characteristic ECG pattern of neuro- whom autopsies confirmed the presence of a ruptured cere- cardiac disease, which is reminiscent of subendocardial bral aneurysm. In 3 of these 5 patients, the heart and coronary ischemia, though no ischemic lesion can be found in the arteries were said to be normal, but the details of the hearts of dogs euthanized after several months of infusions.11 pathological examination were not revealed. The point is In the years that followed, numerous reports emanated from made that ECG changes seen in the context of neurological around the world that documented the production of cardiac disease do not represent ischemic heart disease but are merely repolarization abnormalities in the context of various neuro- a manifestation of autonomic dysregulation, possibly caused logical catastrophes and that proposed that this was caused by by a lesion that affected the cortical representation of the an autonomic storm. It seemed likely that the connection Samuels The Brain–Heart Connection 79 between neuropsychiatric illness and the visceral organs contention that the ECG changes and cardiac lesions are due would be provided by the autonomic nervous system. to direct intracardiac release of catecholamines. Melville et al12 produced ECG changes and myocardial Hawkins and Clower20 injected blood intracranially into necrosis by stimulation of the hypothalamus of cats. With mice, which thereby produced the characteristic myocardial anterior hypothalamic stimulation, parasympathetic responses lesions. The number of lesions could be reduced but not occurred, predominantly in the form of bradycardia. Lateral obliterated by pretreatment with adrenalectomy and the use of hypothalamic stimulation produced tachycardia and ST seg- either atropine or reserpine, which suggested that the cause of ment depressions. With intense bilateral and repeated lateral the lesions was in part caused by sympathetic overactivity stimulation, persistent irreversible ECG changes occurred and (humoral arrival at the myocardium from the adrenal and by postmortem examination revealed a stereotyped cardiac le- direct release into the muscle by intracardiac nerves) and in sion characterized by intense cytoplasmic eosinophilia with part caused by parasympathetic overactivity. This supports loss of cross-striations and some hemorrhage. The coronary the concept that the cause is an autonomic storm with a arteries were normal without occlusion. Although Melville contribution from both divisions to the pathogenesis. referred to this lesion as “infarction,” it is probably best to Jacob et al21 produced subarachnoid hemorrhage experi- reserve that term for coagulation necrosis caused by ischemia. mentally in dogs and carefully studied the sequential hemo- This lesion is probably identical to Selye’s ESCN and would dynamic and ultrastructural changes that occurred. The he- now be called coagulative myocytolysis, myofibrillar degen- modynamic changes occurred in 4 stages and directly eration, or contraction band necrosis. More recently, Oppen- paralleled the effects seen with intravenous norepinephrine heimer and Cechetto have mapped the chronotropic organi- injections. These stages were: (1) dramatic rise in systemic zational structure in the rat insular cortex, which blood pressure; (2) extreme sinus tachycardia with various demonstrates that sympathetic innervation arises from a more arrhythmias (eg, nodal or ventricular tachycardia, bradycar- rostral part of the posterior insula then causes parasympa- dia, atrioventricular block, ventricular premature beats, ventricular tachycardia, ventricular fibrillation with sudden thetic innervation.13 The insula and thalamus had already been shown to have a sensory viscerotropic representation death), all of which could be suppressed by bilateral vagotomy or orbital frontal resection; (3) rise in left ventricular that included the termination of cardiopulmonary afferents.14 pressure parallel to rise in systemic pressure; and (4) up to The central role of the insula in the control of cardiovascular 2-fold increase in coronary blood flow. function has been supported by a robust experimental and Ultrastructurally, a series of 3 stereotyped events occurred clinical literature.15,16 that could be imitated exactly with norepinephrine injections. Despite the fact that myocardial damage could definitely These were: (1) migration of intramitochondrial granules that be produced in animals, until the mid-1960s there was little ϩ contained Ca2 to the periphery of the mitochondria; (2) recognition that this actually occurred in human beings with disappearance of these granules; and (3) myofilament disin- acute neurological or psychiatric illness until Koskelo and tegration at the I bands while the density of the I band was colleagues17 reported on 3 patients with ECG changes caused increased in the intact sarcomeres.21 by subarachnoid hemorrhage who were noted on postmortem Partially successful efforts to modify the developments of examination to have several small subendocardial petechial neurocardiac lesions were made with reserpine pretreatment 18 hemorrhages. Connor reported focal myocytolysis in 8% of in mice subjected to simulated intracranial hemorrhage22 and 231 autopsies, with the highest incidence seen in patients who by Hunt and Gore,23 who pretreated a group of rats with suffered fatal intracranial hemorrhages. The lesion reported propranolol and then attempted to produce cardiac lesions by Connor conforms to the descriptions of Selye’s ESCN or with intracranial blood injections. No lesions were found in what might now be called myofibrillar degeneration, coagu- the control animals, but they were found in 21 of the 46 lative myocytolysis, or contraction band necrosis. Connor untreated rats and in only 4 of the 22 treated rats. This pointed out that previous pathology reports probably over- suggested that neurological influences via catecholamines looked the lesion because of the fact that it was multifocal, may be at least partly responsible for cardiac cell death. More with each individual focus being quite small, which would modern studies have confirmed the fact that myocardial require extensive tissue sampling. It is clear now that even injury occurs in the context of subarachnoid hemorrhage and Connor underestimated the prevalence of the lesion and that that the likelihood of myocardial necrosis was correlated with serial sections are required to rigorously exclude its presence. the severity of the clinical neurological state as judged by the Greenshoot and Reichenbach19 reported on 3 new patients standard Hunt-Hess grading system for subarachnoid with subarachnoid hemorrhage and a review of 6 previous hemorrhage.24 patients from the same medical center. All 9 of these patients The phenomenology of the various types of myocardial had cardiac lesions of varying degrees of severity that ranged cell death was articulated most clearly by Baroldi,25 who from eosinophilia with preservation of cross-striations to pointed out that there were 3 main patterns of myocardial transformation of the myocardial cell cytoplasm into dense necrosis: (1) coagulation necrosis, the fundamental lesion of eosinophilic transverse bands with intervening granularity, infarction, in which the cell loses its capacity to contract and sometimes with endocardial hemorrhages. Both the ECG dies in an atonic state with no myofibrillar damage; (2) abnormalities and the cardiac pathology could be reproduced colliquative myocytolysis, in which edematous vacuolization in cats given mesencephalic reticular formation stimulation. with dissolution of myofibrils without hypercontraction oc- Adrenalectomy did not protect the hearts, which supports the curs in the low-output syndromes; and (3) coagulative myo-

80 Circulation July 3, 2007

Figure 1. The neurocardiac lesion: Gross specimen of a patient Figure 3. Intense mineralization within minutes of the onset of who died during an acute psychological stress shows fresh contraction band necrosis. endocardial hemorrhages (1 of many is shown by the arrow). cytolysis, in which the cell dies in a hypercontracted state massive contracture followed by necrosis and enzyme re- with early myofibrillar damage and anomalous irregular lease. This phenomenon, known as the calcium paradox, can cross-band formations. be imitated almost exactly with reoxygenation after hypox- Coagulative myocytolysis is seen in reperfused areas emia. The latter, called the oxygen paradox, has been linked around regions of coagulation necrosis in transplanted hearts, to the calcium paradox by pathological calcium entry.27 This in “stone hearts,” in sudden unexpected and accidental death, major ionic shift is probably the cause of the dramatic ECG and in hearts exposed to toxic levels of catecholamines, such changes seen in the context of neurological catastrophe, a fact as in patients with pheochromocytoma. This is probably the that could explain the phenomenon of sudden unexpected major lesion described by Selye as ESCN and is likely to be the major lesion seen in animals and people who suffer acute death (SUD) in many contexts. neurological or psychiatric catastrophes. Although coagula- Although SUD is now recognized as a medical problem of tive myocytolysis is probably the preferred term, the terms major epidemiological importance, it has generally been myofibrillar degeneration and contraction band necrosis are assumed that neurological disease rarely results in SUD. In commonly used in the literature. This lesion tends to calcify fact, it has been traditionally held that neurological illnesses early and to have a multifocal subendocardial predisposition almost never cause sudden demise, with the occasional (Figure 1, 2, and 3). patient who dies during an epileptic convulsion or rapidly in Intense rapid calcification makes it likely that the subcel- the context of a subarachnoid hemorrhage as the exception. lular mechanisms that underlie the development of coagula- Further, it has been assumed that the various SUD syndromes tive myocytolysis involve calcium entry. Zimmerman and (eg, sudden death in middle-aged men, sudden infant death Hulsmann26 reported that the perfusion of rat hearts with syndrome, sudden unexpected nocturnal death syndrome, calcium-free media for short periods of time creates a frightened to death (“voodoo” death), sudden death during an situation such that on readmission of calcium there is a epileptic seizure, sudden death during natural catastrophe, sudden death associated with recreational drug abuse, sudden death in wild and domestic animals, sudden death during asthma attacks, sudden death during the alcohol withdrawal syndrome, sudden death during grief after a major loss, sudden death during panic attacks, sudden death from mental stress, and sudden death during war) are entirely separate and have no unifying mechanism. For example, it is generally accepted that sudden death in middle-aged men is usually caused by a cardiac arrhythmia (ie, ventricular fibrillation), which results in functional cardiac arrest, whereas most work on sudden infant death syndrome focuses on immaturity of the respiratory control systems in the brain stem. However, the connection between the nervous system and the cardiopulmonary system provides the unifying link that allows a coherent explanation for most, if not all, of the forms of neurocardiac damage. Powerful evidence from multiple Figure 2. Cardiac contraction band necrosis (also known as coagulative myocytolysis, myoﬁbrillar degeneration). The arrow disparate disciplines allows for a neurological explanation for shows 2 of the contraction bands. many forms of SUD.28

Samuels The Brain–Heart Connection 81

Neurogenic Heart Disease of the necrotic debris by mononuclear cells is often noted, sometimes with hemorrhage. Definition of Neurogenic Ultrastructurally, the changes in cardiac muscle are even Electrocardiographic Changes more widespread than they appear to be in light microscopy. A wide variety of changes in the ECG is seen in the context Nearly every muscle cell shows some pathological alteration, of neurological disease. Two major categories of change are which range from a granular appearance of the myofibrils to regularly noted: arrhythmias and repolarization changes. It is profound disruption of the cell architecture with relative likely that the increased tendency for life-threatening arrhyth- preservation of ribosomes and mitochondria. Intracardiac mias found in patients with acute neurological disease is a nerves can be seen and identified by their external lamina, result of the repolarization change, which increases the microtubules, neurofibrils, and the presence of intracytoplas- vulnerable period during which an extrasystole would be mic vesicles. These nerves can sometimes be seen immedi- likely to result in ventricular tachycardia and/or ventricular ately adjacent to an area of myocardial cell damage. The fibrillation. Thus, the essential and potentially most lethal pathological changes in the cardiac muscle are usually less at features of the ECG, which are known to change in the a distance from the nerve, often with a complete return to context of neurological disease, are the ST segment and T normalcy by a distance of 2 to 4 ␮m away from the nerve wave, which reflect abnormalities in repolarization. Most ending.21 often, the changes are seen best in the anterolateral or Myofibrillar degeneration (also known as coagulative inferolateral leads. If the ECG is read by pattern recognition myocytolysis and contraction band necrosis) is an easily by someone who is not aware of the clinical history, it will recognizable form of cardiac injury, distinct in several major often be said to represent subendocardial infarction or antero- respects from coagulation necrosis, which is the major lesion lateral ischemia. The electrocardiographic abnormalities usu- of myocardial infarction.25,31 In coagulation necrosis, the cells ally improve, often dramatically, with death by brain criteria. die in a relaxed state without prominent contraction bands. In fact, any circumstance that disconnects the brain from the This is not visible by any method for many hours or even heart (eg, cardiac transplantation, severe autonomic neurop- days. Calcification only occurs late, and the lesion elicits a athies caused by amyloidosis or diabetes, stellate ganglionec- polymorphonuclear cell response. In stark contrast, in myo- tomy for treatment of the long QT syndrome) blunts neuro- fibrillar degeneration the cells die in a hypercontracted state cardiac damage of any cause. with prominent contraction bands (Figures 2 and 3). The The phenomenon is not rare. In a series of 100 consecutive lesion is visible early, perhaps within minutes of its onset. It stroke patients, 90% showed abnormalities on the ECG elicits a mononuclear cell response and may calcify almost compared with 50% of a control population of 100 patients immediately.31,32 admitted for carcinoma of the colon.29 This of course does not mean that 90% of stroke patients have neurogenic ECG Stress Plus or Minus Steroids changes. Obviously, stroke and coronary artery disease have A similar, if not identical, cardiac lesion can be produced common risk factors, so that many ECG abnormalities in with various models of stress. This concept was applied to the stroke patients represent concomitant atherosclerotic coro- heart when Selye published his monograph The Chemical nary disease. Nonetheless, a significant number of stroke Prevention of Cardiac Necrosis in 1958.9 He found that patients have authentic neurogenic ECG changes. cardiac lesions probably identical to those described above could be produced regularly in animals that were pretreated Mechanism of the Production of Neurogenic with certain steroids, particularly 2-␣-methyl-9-␣- Heart Disease fluorohydrocortisone (fluorocortisol) and then subjected to various types of stress. Other hormones, such as dihydrotach- Catecholamine Infusion ysterol (calciferol) and thyroxine, could also sensitize animals Josué30 first demonstrated that epinephrine infusions could for stress-induced myocardial lesions, though less potently cause cardiac hypertrophy. This observation has been repro- than fluorocortisol. This so-called stress could be of multiple duced on many occasions, which documents the fact that types such as restraint, surgery, bacteremia, vagotomy, and systemically administered catecholamines are not only asso- toxins. He believed that the first mediator in the translation of ciated with ECG changes reminiscent of widespread ischemia these widely disparate stimuli into a stereotyped cardiac but with a characteristic pathological picture in the cardiac lesion was the hypothalamus and that it, by its control over muscle that is distinct from myocardial infarction. An iden- the autonomic nervous system, caused the release of certain tical picture may be found in human beings with chronically agents that were toxic to the myocardial cell. Since Selye’s elevated catecholamines, as seen with pheochromocytoma. original work, similar experiments have been repeated in Patients with stroke often have elevated systemic catechol- many different types of laboratory animals with comparable amine levels, a fact that may in part account for the high results. Although the administration of exogenous steroids incidence of cardiac arrhythmias and ECG changes seen in facilitates the production of cardiac lesions, it is clear that these patients. On light microscopy, these changes range from stress alone can result in the production of morphologically increased eosinophilic staining with preservation of cross- identical lesions. striations to total transformation of the myocardial cell Whether a similar pathophysiology could ever be repeated cytoplasm into dense eosinophilic transverse bands with in human beings is, of course, of great interest. Many intervening granularity. In severely injured areas, infiltration investigators have speculated on the role of stress in the

82 Circulation July 3, 2007 pathogenesis of human cardiovascular disease and, in partic- system can cause coronary vasospasm, but the possibility ular, on its relationship to the phenomenon of SUD. A few remains. Regardless of the precise mechanism, the fact autopsies on patients who experienced sudden death have remains that takotsubo-like cardiomyopathy occurs after an shown myofibrillar degeneration. Cebelin and Hirsch33 re- acute psychological stress and thereby represents an example ported on a careful retrospective analysis of the hearts of 15 of a neurocardiac lesion. It seems likely that this dramatic victims of physical assault who died as a direct result of the condition may be the tip of an iceberg under which lurks a assault, but without sustaining internal injuries. Eleven of the much larger, albeit less easily demonstrable, problem; namely 15 individuals showed myofibrillar degeneration. Age- and neurocardiac lesions that are not sufficiently severe and cardiac disease–matched controls showed little or no evi- widespread to produce gross heart failure but may predispose dence of this change. This appears to represent human stress to serious cardiac arrhythmias. cardiomyopathy. Whether such assaults can be considered murder has become an interesting legal correlate of the Nervous System Stimulation problem. Nervous system stimulation produces cardiac lesions that are Because the myofibrillar degeneration is predominantly histologically indistinguishable from those described for subendocardial, it may involve the cardiac conducting sys- stress and catecholamine-induced cardiac damage. It has been tem, which thus predisposes to cardiac arrhythmias. This known for a long time that stimulation of the hypothalamus lesion, combined with the propensity of catecholamines to can lead to autonomic cardiovascular disturbances,38 and produce arrhythmias even in a normal heart, may well raise many years ago lesions in the heart and gastrointestinal tract the risk of a serious arrhythmia. This may be the major have been produced with hypothalamic stimulation.39,40 It has immediate mechanism of sudden death in many neurological been clearly demonstrated that stimulation of the lateral circumstances, such as subarachnoid hemorrhage, stroke, hypothalamus produces hypertension and/or electrocardio- epilepsy, head trauma, psychological stress, and increased graphic changes reminiscent of those seen in patients with intracranial pressure. Even the arrhythmogenic nature of central nervous system damage of various types. Further- digitalis may be largely mediated by the central nervous more, this effect on the blood pressure and ECG can be system. Further evidence for this is the antiarrhythmic effect completely prevented by C2 spinal section and stellate of sympathetic denervation of the heart for cardiac arrhyth- ganglionectomy, but not by vagotomy, which suggests that mias of many types. the mechanism of the electrocardiographic changes is sym- Furthermore, it is known that stress-induced myocardial pathetic rather than parasympathetic or humoral. Stimulation lesions may be prevented by sympathetic blockade with many of the anterior hypothalamus produces bradycardia, an effect different classes of antiadrenergic agents, most notably, that can be blocked by vagotomy. Unilateral hypothalamic ganglionic blockers such as mecamylamine and catechol- stimulation does not result in histological evidence of myo- amine-depleting agents such as reserpine.10 This suggests that cardial damage by light microscopy, but bilateral prolonged catecholamines, which are either released directly into the stimulation regularly produces myofibrillar degeneration in- heart by sympathetic nerve terminals or reach the heart distinguishable from that produced by catecholamine injec- through the bloodstream after release from the adrenal me- tions and stress, as previously described.41 dulla, may be excitotoxic to myocardial cells. Other methods to produce cardiac lesions of this type Some people who are subjected to an extreme stress may include stimulation of the limbic cortex, the mesencephalic develop an acute cardiomyopathy that presents with chest reticular formation, the stellate ganglion, and regions known pain and/or symptoms of heart failure. This process is most to elicit cardiac reflexes such as the aortic arch. Experimental commonly seen in older women, whose echocardiograms and intracerebral and subarachnoid hemorrhages can also result in ventriculograms show a typical pattern of left ventricular cardiac contraction band lesions. These neurogenic cardiac apical ballooning, which was named takotsubo-like cardio- lesions will occur even in an adrenalectomized animal, myopathy34 because of the similarity in the appearance the although they will be somewhat less pronounced.20 This left ventricle to the Japanese octopus trapping pot, the evidence argues strongly against an exclusively humoral takotsubo. If a lethal arrhythmia does not intervene, the mechanism in the intact organism. High levels of circulating process is potentially completely reversible. Some debate catecholamines exaggerate the electrocardiographic findings exists regarding whether this syndrome (variously described and myocardial lesions, but high circulating catecholamine as myocardial stunning or myocardial hibernation) could be levels are not required for the production of pathological explained by ischemia, but it is striking that this pattern of changes. These electrocardiographic abnormalities and car- dysfunction is most consistent with a neural rather than a diac lesions are stereotyped and identical to those found in the vascular distribution.35,36 Wittstein and colleagues37 reported stress and catecholamine models already outlined. They are a series of such patients and referred to the problem as not affected by vagotomy and are blocked by maneuvers that myocardial stunning. In patients in whom endocardial biop- interfere with the action of the sympathetic limb of the sies were performed, contraction band lesions were found. autonomic nervous system, such as C2 spinal section, stellate The finding of contraction bands suggests either catechol- ganglion blockade, and administration of antiadrenergic amine effect and/or reperfusion. The 2 mechanisms are not drugs such as propranolol. mutually exclusive in that a neural stimulus could produce The histological changes in the myocardium range from both catecholamine release and coronary vasospasm followed normal muscle on light microscopy to severely necrotic (but by vasodilation. There is no direct evidence that the nervous not ischemic) lesions with secondary mononuclear cell infil- Samuels The Brain–Heart Connection 83 tration. The findings on ultrastructural examination are invariably more widespread, often involving nearly every muscle cell, even when the light microscopic appearance is unimpressive. The electrocardiographic findings undoubtedly reflect the total amount of muscle membrane affected by the pathophysiological process. Thus, the ECG may be normal when the lesion is early and demonstrable only by electron microscopy. Conversely, the ECG may be grossly abnormal when only minimal findings are present by light microscopy, since the cardiac membrane abnormality responsible for the electrocardiographic changes may be reversible. Cardiac arrhythmias of many types may also be elicited by nervous system stimulation along the outflow of the sympathetic nervous system.

Reperfusion Figure 4. Cascade of events that lead to neurocardiac damage. The fourth and last model for the production of myofibrillar degeneration is reperfusion, as is commonly seen in patients released as a result of reperfusion after ischemia or by the who die after a period of time on a left ventricular assist pump metabolism of catecholamines to the known toxic metabolite, or after they undergo extracorporeal circulation. Similar adrenochrome, may contribute to cell membrane destruction, lesions are seen in hearts that were reperfused with angio- which leads to leakage of cardiac enzymes into the blood.44,45 plasty or fibrinolytic therapy. The mechanism by which Thus, the cardiac toxicity of locally released norepinephrine reperfusion of ischemic cardiac muscle produces coagulative falls on a continuum that ranges from a brief reversible burst myocytolysis (also known as myofibrillar degeneration and of electrocardiographic abnormalities to a pattern that resem- contraction band necrosis) involves entry of calcium after a bles hyperkalemia and then finally to an irreversible failure of period of relative deprivation.41 the muscle cell with permanent repolarization abnormalities, Sudden calcium influx by one of several possible mecha- or even the occurrence of transmural cardiac necrosis with nisms (eg, a period of calcium deficiency with loss of enzyme (eg, troponin, creatine kinase) release and Q waves intracellular calcium, a period of anoxia followed by reoxy- seen on the ECG. genation of the electron transport system, a period of ische- Histological changes would also represent a continuum mia followed by reperfusion, or opening of the receptor- that ranges from complete reversibility in a normal heart operated calcium channels by excessive amounts of locally through mild changes seen only by electron microscopy to released norepinephrine) may be the final common pathway severe myocardial cell necrosis with mononuclear cell infil- by which the irreversible contractures occur, which leads to tration and even hemorrhages. The amount of cardiac en- myofibrillar degeneration. Thus reperfusion-induced myocar- zymes released and the electrocardiographic changes would dial cell death may be a form of apoptosis (programmed cell roughly correlate with the severity and extent of the patho- death) analogous to that seen in the central nervous system, in logical process. This explanation, summarized in Figure 4, which excitotoxicity with glutamate results in a similar, if not rationalizes all the observations in the catecholamine infu- identical, series of events.42 sion, stress plus or minus steroid, nervous system stimulation, The precise cellular mechanism for the electrocardio- and reperfusion models. graphic change and the histological lesion may well reflect the effects of large volumes of norepinephrine released into the myocardium from sympathetic nerve terminals.43 The fact Concluding Remarks and Potential Treatments that the cardiac necrosis is greatest near the nerve terminals in In conclusion, there is powerful evidence to suggest that the endocardium and is progressively less severe as one overactivity of the sympathetic limb of the autonomic ner- samples muscle cells near the epicardium provides further vous system is the common phenomenon that links the major evidence that catecholamine toxicity produces the lesion.19 cardiac pathologies seen in neurological catastrophes. These This locally released norepinephrine is known to stimulate profound effects on the heart may contribute in a major way synthesis of adenosine 3Ј,5Ј-cyclic phosphate, which in turn to the mortality rates of many primarily neurological condi- results in the opening of the calcium channel with influx of tions such as subarachnoid hemorrhage, cerebral infarction, calcium and efflux of potassium. This efflux of potassium status epilepticus, and head trauma. These phenomena may could explain the peaked T waves (a hyperkalemic pattern) also be important in the pathogenesis of SUD in adults, often seen early in the evolution of neurogenic electrocardio- sudden infant death, sudden death during asthma attacks, graphic changes.21 The actin and myosin filaments interact cocaine- and amphetamine-related deaths, and sudden death under the influence of calcium but do not relax unless the during the alcohol withdrawal syndrome, all of which may be calcium channel closes. Continuously high levels of norepi- linked by stress and catecholamine toxicity. nephrine in the region may result in failure of the calcium Investigations aimed at alteration of the natural history of channel to close, which leads to cell death, and finally to these events with catecholamine receptor blockade, calcium- leakage of enzymes out of the myocardial cell. Free radicals channel blockers, free-radical scavengers, and antioxidants

84 Circulation July 3, 2007

20. Hawkins WE, Clower BR. Myocardial damage after head trauma and simulated intracranial haemorrhage in mice: the role of the autonomic nervous system. Cardiovasc Res. 1971;5:524–529. 21. Jacob WA, Van Bogaert A, DeGroot-Lasseel MHA. Myocardial ultrastructural and haemodynamic reactions during experimental subarachnoid haemorrhage. J Moll Cell Cardiol. 1972;4:287–298. 22. McNair JL, Clower BR, Sanford RA. The effect of reserpine pretreatment on myocardial damage associated with simulated intracranial haemorrhage in mice. Eur J Pharmacol. 1970;9:1–6. 23. Hunt D, Gore I. Myocardial lesions following experimental intracranial hemorrhage: prevention with propranolol. Am Heart J. 1972;83:232–236. 24. Tung P, Kopelnik A, Banki N, Ong K, Ko N, Lawton MT, Gress D, Drew B, Foster E, Parmley W, Zaroff J. Predictors of neurocardiogenic injury after subarachnoid hemorrhage. Stroke. 2004;35:548–553. 25. Baroldi F. Different morphological types of myocardial cell death in man. In: Fleckstein A, Rona G, eds. Recent Advances in Studies in Cardiac Structure and Metabolism. Pathophysiology and Morphology of Myocardial Cell Alteration. Vol 6. Baltimore, Md: University Park Press, 1975. Figure 5. Possible therapeutic approaches aimed to prevent neu- 26. Zimmerman ANA, Hulsmann WC. Paradoxical influence of calcium ions rocardiac damage. GABA indicates gamma aminobutyric acid. on their permeability of the cell membranes of the isolated rat heart. Nature. 1966;211:616–647. 27. Hearse DJ, Humphrey SM, Bullock GR. The oxygen paradox and the are in progress in many centers around the world and are calcium paradox: two facets of the same problem? J Moll Cell Cardiol. summarized in Figure 5. 1978;10:641–668. 28. Samuels MA. Neurally induced cardiac damage. Neurol Clin. 1993;11: 273–292. Disclosures 29. Dimant J, Grob D. Electrocardiographic changes and myocardial damage None. in patients with acute cerebrovascular accidents. Stroke. 1977;8: 448–455. References 30. Josué O. Hypertrophie cardiaque cause par l’adrenaline and la toxine typhique. C R Soc Biol (Paris). 1907;63:285–287. 1. Cannon WB. “Voodoo” death. Am Anthropologist. 1942;44(new series): 31. Karch SB, Billingham ME. Myocardial contraction bands revisited. Hum 169–118. Pathol. 1986;17:9–13. 2. Engel G. Sudden and rapid death during psychological stress. Ann Intern 32. Rona G. Catecholamine cardiotoxicity. J Moll Cell Cardiol. 1985;17: Med. 1971;74:771–782. 291–306. 3. Richter CP. On the phenomenon of sudden death in animal and man. 33. Cebelin M, Hirsch CS. Human stress cardiomyopathy. Hum Pathol. Psychosomatic Med. 1957;19:191–198. 1980;11:123–132. 4. Oppenheimer SM, Gelb A, Girvin JP, Hachinski VC. Cardiovascular effects of human insular cortex stimulation. Neurology. 1992;42: 34. Sato H, Tateishi H, Uchida T. Takotsubo-type left ventricular dysfunction 1727–1732. due to multivessel coronary spasm. In: Kodama K, Haze K, Hori M, eds. 5. Byer E, Ashman R, Toth LA. Electrocardiogram with large upright T Clinical Aspects of Myocardial Injury: From Ischemia to Heart Failure. wave and long Q-T intervals. Am Heart J. 1947;33:796–801. Tokyo, Japan: Kagakuhyoronsha Publishing Co; 1990:56–64. 6. Levine HD. Non-specificity of the electrocardiogram associated with 35. Angelakos ET. Regional distribution of catecholamines in the dog heart. coronary heart disease. Am J Med. 1953;15:344–350. Circ Res. 1965;16:39–44. 7. Burch GE, Myers R, Abildskov JA. A new electrocardiographic pattern 36. Murphree SS, Saffitz JE. Quantitative autoradiographic delineation of the observed in cerebrovascular accidents. Circulation. 1954;9:719–726. distribution of beta-adrenergic receptors in canine and feline left ventric- 8. Cropp CF, Manning GW. Electrocardiographic change simulating myo- ular myocardium. Circ Res. 1987;60:568–579. cardial ischaemia and infarction associated with spontaneous intracranial 37. Wittstein IS, Thiemann DR, Lima JAC, Baughman KL, Schulman SP, haemorrhage. Circulation. 1960;22:25–38. Gerstenblith G, Wu KC, Rade JJ, Bivalaqua TJ, Champion HC. Neuro- 9. Selye H. The Chemical Prevention of Cardiac Necrosis. New York, NY: humoral features of myocardial stunning due to sudden emotional stress. Ronald Press; 1958. N Engl J Med. 2005;352:539–548. 10. Raab W, Stark E, MacMillan WH, Gigee WR. Sympathogenic origin and 38. Dikshit BB. The production of cardiac irregularities by excitation of the anti-adrenergic prevention of stress-induced myocardial lesions. hypothalamic centres. J Physiol. 1934;81:382–394. Am J Cardiol. 1961;8:203–211. 39. Karplus JP, Kreidl A. Gehirn und Sympathicus. Sympathicusleitung im 11. Barger AC, Herd JA, Liebowitz MR. Chronic catheterization of coronary Gehirn und Halsmark [German]. Pflugers Arch. 1912;143:109–127. artery induction of ECG pattern of myocardial ischemia by intracoronary 40. Karplus JP, Kreidl A. Gehirn und Sympathicus. Uber Beziehungen der epinephrine. Proc Soc Exp Biol Med. 1961;107:474–477. Hypothalamaszentren zu Blutdruck und innerer Sekretion [German]. 12. Melville KI, Blum B, Shister HE, Silver MD. Cardiac ischemic changes Pflugers Arch. 1927;215:667–674. and arrhythmias induced by hypothalamic stimulation. Am J Cardiol. 41. Braunwald E, Kloner RA. Myocardial reperfusion: a double-edged 1963;12:781–791. sword? J Clin Invest. 1985;76:13–19. 13. Oppenheimer SM, Cechetto DF. Cardiac chronotropic organization of the 42. Gottlieb R, Burleson KO, Kloner RA Babior BM, Engler RL. Reper- rat insular cortex. Brain Res. 1990;533:66–72. fusion injury induces apoptosis in rabbit cardiomyocytes. J Clin Invest. 14. Cechetto DF, Saper CB. Evidence for a viscerotopic sensory repre- 1994;94:1621–1628. sentation in the cortex and thalamus in the rat. J Comp Neurology. 43. Eliot RS, Todd GL, Pieper GM, Clayton FC. Pathophysiology of cate- 1987;262:27–45. cholamine-mediated myocardial damage. J S C Med Assoc. 1979;75: 15. Cheung RTF, Hachinski V. The insula and cerebrogenic sudden death. 513–518. Arch Neurol. 2000;57:1685–1688. 44. Singal PK, Kapur N, Dhillon KS, Beamish RE, Dhalla NS. Role of free 16. Cheshire WP, Saper CB. The insular cortex and cardiac response to radicals in catecholamine-induced cardiomyopathy. Can J Physiol stroke. Neurology. 2006;66:1296–1297. Pharmacol. 1982;60:1390–1397. 17. Koskelo P, Punsar SO, Sipila W. Subendocardial haemorrhage and ECG 45. Meerson FZ. Pathogenesis and prophylaxis of cardiac lesions in stress. changes in intracranial bleeding. BMJ. 1964;1:1479–1483. Adv Myocardiol. 1983;4:3–21. 18. Connor RCR. Myocardial damage secondary to brain lesions. Am Heart J. 1969;78:145–148. KEY WORDS: antioxidants Ⅲ apoptosis Ⅲ cardiomyopathy Ⅲ cerebral 19. Greenshoot JH, Reichenbach DD. Cardiac injury and subarachnoid haem- infarction Ⅲ death, sudden Ⅲ nervous system, autonomic Ⅲ nervous system, orrhage. J Neurosurg. 1969;30:521–531. sympathetic

Cardiovascular Involvement in General Medical Conditions

Chronic Kidney Disease Effects on the Cardiovascular System

Ernesto L. Schiffrin, MD, PhD, FRSC, FRCPC; Mark L. Lipman, MD, FRCPC; Johannes F.E. Mann, MD

Abstract—Accelerated cardiovascular disease is a frequent complication of renal disease. Chronic kidney disease promotes hypertension and dyslipidemia, which in turn can contribute to the progression of renal failure. Furthermore, diabetic nephropathy is the leading cause of renal failure in developed countries. Together, hypertension, dyslipidemia, and diabetes are major risk factors for the development of endothelial dysfunction and progression of atherosclerosis. Inflammatory mediators are often elevated and the renin-angiotensin system is frequently activated in chronic kidney disease, which likely contributes through enhanced production of reactive oxygen species to the accelerated atherosclerosis observed in chronic kidney disease. Promoters of calcification are increased and inhibitors of calcification are reduced, which favors metastatic vascular calcification, an important participant in vascular injury associated with end-stage renal disease. Accelerated atherosclerosis will then lead to increased prevalence of coronary artery disease, heart failure, stroke, and peripheral arterial disease. Consequently, subjects with chronic renal failure are exposed to increased morbidity and mortality as a result of cardiovascular events. Prevention and treatment of cardiovascular disease are major considerations in the management of individuals with chronic kidney disease. (Circulation. 2007;116:85-97.) Key Words: atherosclerosis Ⅲ hypertension Ⅲ kidney Ⅲ vasculature

t is increasingly apparent that individuals with chronic addressed because the emphasis will be on CKD before Ikidney disease (CKD) are more likely to die of cardiovas- ESRD is reached. In addition, the CV complications associ- cular (CV) disease (CVD) than to develop kidney failure.1,2 A ated with dialysis will not be discussed. The different stages large cohort study comprising Ͼ130 000 elderly subjects of CKD according to the level of glomerular filtration rate showed that increased incidence of CV events could be in (GFR) are shown in Table 1. ESRD corresponds to the stage part related to the fact that persons with renal insufficiency where patients need renal replacement therapy (ie, dialysis or are less likely to receive appropriate cardioprotective treat- renal transplantation), whereas stage 1 is mostly recognized ments.3 However, beyond the effects of lack of appropriate by either albuminuria or structural renal abnormality (eg, therapy, it is clear that accelerated CVD is prevalent in hyperechoic renal parenchyma on ultrasound). Table 2 pro- subjects with CKD. The first part of the present review will vides the approximate odds ratios (univariate) of CVD therefore focus on the epidemiological links between impair- according to stages of CKD on the basis of the literature cited ment of renal function and adverse CV events, between below. The increase in risk in comparison to people without albuminuria and CV events, and between serum cystatin C CKD depends on the age of the population studied: the and CVD. The second part of the present review will address younger the person, the higher the relative risk. Microalbu- the mechanisms that lead to the association of renal and CVD, minuria increases the CV risk 2- to 4-fold. which include hypertension, dyslipidemia, activation of the renin-angiotensin system, endothelial dysfunction and the Epidemiological Links Between Impaired role of asymmetric dimethyl arginine (ADMA), oxidative GFR and Adverse Cardiovascular Events stress, and inflammation. Finally, mechanisms that are in- Evidence for the relationship between renal dysfunction and volved in vascular calcification often found in CKD and adverse CV events was perhaps first recognized in the end-stage renal disease (ESRD) will be described. Addition- dialysis population in whom the incidence of CV death is ally, ESRD is associated with several specific complications strikingly high. Approximately 50% of individuals with caused by the uremic state per se, which can contribute to the ESRD die from a CV cause,2,4,5 a CV mortality that is 15 to development and progression of CVD through volume over- 30 times higher than the age-adjusted CV mortality in the load with consequent hypertension, anemia, uremic pericar- general population.4,6 This disparity is present across all ages, ditis, and cardiomyopathy. However, these issues will not be but it is most marked in the younger age group (25 to 34 years

From the Department of Medicine (E.L.S., M.L.L.), Sir Mortimer B. Davis Jewish General Hospital, McGill University, Montreal, Quebec, Canada; and the Department of Nephrology and Hypertension (J.F.E.M.), Schwabing General Hospital, Ludwig Maximilians University, Munich, Germany. Correspondence to Ernesto L. Schiffrin, MD, PhD, FRSC, FRCPC, Sir Mortimer B. Davis Jewish General Hospital, B-127, 3755 Côte Ste-Catherine Rd, Montreal, Quebec, Canada H3T 1E2. E-mail [email protected] © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.678342 85 86 Circulation July 3, 2007

TABLE 1. Stages of CKD

Stage Description GFR, mL/min per 1.73 m2 1* Kidney damage with normal or increased GFR Ն90 2 Kidney damage with mildly decreased GFR 60 to 89 3 Moderately decreased GFR 30 to 59 4 Severely decreased GFR 15 to 29 5 Kidney failure Ͻ15 or dialysis ESRD is defined as the need for renal replacement therapy (ie, need for dialysis or renal transplantation). *Stage 1 CKD is mostly recognized by either albuminuria or structural renal abnormality (eg, hyperechoic renal parenchyma on ultrasound). old), where the CV mortality is 500-fold greater in ESRD A growing number of studies have demonstrated that the patients compared with age-matched controls with normal relationship between renal dysfunction and increased CV renal function.1 It is therefore unsurprising that established morbidity and mortality extends across the spectrum of renal CVD is easily demonstrable in CKD. For example, 40% of dysfunction to encompass the mildest degrees of renal im- patients who have started dialysis treatments have evidence pairment. Moreover, this relationship appears to hold across of coronary artery disease, and 85% of these patients have populations with widely varying degrees of baseline CV abnormal left ventricular structure and function.7 health. The relationship between renal disease and CV mortality has also been shown to extend to subjects with more CVD Associated With Renal Disease in the moderate degrees of renal functional impairment. In fact, the General Population majority of patients with stage 3 to 4 CKD (ie, a GFRϽ60 The Framingham Heart Study was among the first to assess mL/min per 1.73 m2) die of CV causes rather than progress to mild renal insufficiency and its association with death and ESRD. Here too, objective evidence of structural and func- adverse CV events in the general population.10 Of the 6233 tional cardiac abnormalities has been demonstrated by echo- participants in the study, mild renal insufficiency was present cardiography. Levin et al determined left ventricular mass in 246 men and 270 women (serum creatinine, 1.4 to 3.0 index in a population of 115 men and 60 women with an mg/dL). Of these individuals, 81% had no prevalent CVD at average creatinine clearance (CrCl) of 25.5Ϯ17 mL/min with entry. Over the 15-year follow-up period, there was no 2-dimensional targeted M-mode echocardiography. The pop- significant association between mild renal insufficiency and ulation was stratified into 3 groups according to renal either death or adverse CV events in women. However, in function. The prevalence of left ventricular hypertrophy men there was a trend toward more CV events with mild renal (LVH) was 26.7% in subjects with CrClϾ50 mL/min, 30.8% insufficiency, and a significant association was demonstrated in those with CrCl between 25 and 49 mL/min, and 45.2% in with age-adjusted all-cause mortality (hazard ratio, 1.42). individuals with CrClϽ25 mL/min.8 Given the relatively small number of subjects followed in this Tucker et al9 reported a similar finding in a population of study and the low number of outcome events, these findings 85 persons with renal insufficiency. With the same echocar- were suggestive but not definitive of a correlation between diographic techniques, as well as comparable criteria for the mild renal dysfunction and increased CV morbidity and diagnosis of LVH, these investigators found an LVH preva- mortality. lence of 16% in subjects with a CrClϾ30 mL/min and 38% in More recently, Go et al11 examined the relationship of those with a CrClϽ30 mL/min. These studies demonstrate GFR and adverse CV events in a low-risk population. They that LVH is common in patients with renal insufficiency even analyzed the database of a large healthcare provider in before they progress to dialysis, and that the prevalence of northern California and used the Modification of Diet in LVH correlates with the degree of renal functional Renal Disease (MDRD) formula12 to estimate the baseline impairment. GFR from measurements of serum creatinine in Ͼ1.1 million adults, with only those who were on dialysis or TABLE 2. CV Risk According to Stages of CKD who had undergone a kidney transplant excluded. The Stage CV Risk (Odds Ratio, Univariate) primary outcomes examined included death from any cause, CV events, and hospitalizations. The end-point 1 Depending on degree of proteinuria information was obtained from the health-plan database 2 1.5 and the California death registry with a mean follow-up 3 2 to 4 period of 2.84 years. After adjustment for age, sex, race, 4 4 to 10 coexisting illnesses, and socioeconomic status, a stepwise 5 10to50 increase in the rate of each of the 3 primary outcomes was ESRD 20 to 1000 seen for every sequential decrease in GFR. With the best Ͼ 2 The increase in risk in comparison with people free of CKD depends on the cohort (GFR 60 mL/min per 1.73 m ) as the point of age of the population studied: The younger the person, the higher the relative reference, the adjusted hazard ratio for death from any risk. Microalbuminuria increases the CV risk 2- to 4-fold. cause and any CV event increased to 1.2 and 1.4, respec- Schiffrin et al Kidney Disease and the Cardiovascular System 87 tively, for a GFR between 45 to 59 mL/min per 1.73 m2; LVEFϾ40%. The primary end point was a composite of 1.8 and 2.0 for a GFR between 30 to 44 mL/min per 1.73 death from CV causes, MI, and coronary revascularization. m2; 3.2 and 2.8 for GFR between 15 to 29 mL/min per 1.73 Patients with a serum creatinine Ͼ2.0 mg/dL were excluded m2; and 5.9 and 3.4 for a GFRϽ15 mL/min per 1.73 m2. and the median duration of follow-up was 4.8 years. A post The adjusted risk of hospitalization with a reduced GFR hoc analysis of 8280 subjects, in whom baseline renal followed a similar pattern. This large study, which incor- function was separated into quartiles with the MDRD for- porated a diverse population of adults, clearly demon- mula, demonstrated significant stepwise increases in event strated an independent and graded (inverse) correlation rates as the baseline GFR declined. Interestingly, unlike in between decreasing levels of renal function and increasing HOPE, there was a significant interaction between GFR and event rates of CV morbidity and death. treatment group with respect to CV and all-cause mortality in that the angiotensin-converting enzyme inhibitor benefited CVD Associated With Renal Disease in only those individuals with a GFRϽ60 mL/min per 1.73 m2. Hypertensive Subjects The association between renal function and mortality in the Effect of Renal Disease in Patients With Established hypertensive population was evaluated by the Hypertension Heart Failure or Postmyocardial Infarction Detection and Follow-up Program Cooperative Group, which Hillege et al examined whether renal dysfunction was a followed and treated 10 940 hypertensive subjects to compare predictor of mortality in stable patients with advanced heart stepped care to referred care.13 The primary end point of the failure.18 They studied 1906 subjects with New York Heart study was all-cause mortality. Persons with a baseline serum Association class III and IV heart failure and evidence of left creatinine Ն1.7 mg/dL experienced an 8-year mortality rate ventricular dysfunction (LVEFϽ35%) who were enrolled in that was Ͼ3 times higher than that of all other participants. the Second Prospective Randomized study of Ibopamine on Data from the Hypertension Optimal Treatment (HOT) Mortality and Efficacy (PRIME II).19 Hillege et al correlated study support this finding. In the HOT study, 18 790 hyper- baseline GFR, as calculated with the Cockroft-Gault equa- tensive subjects, only 10% of whom had evidence of athero- tion, with overall mortality after a median follow-up of 277 sclerotic disease, were assigned to 3 diastolic blood pressure days. The authors found that patients in the lowest quartile of target groups and followed for a mean of 3.8 years. Persons GFR (Ͻ44 mL/min) had relative risk of mortality of 2.85 with a serum creatinine Ͼ3 mg/dL were excluded and the compared with subjects in the highest quartile (Ͼ76 mL/min). Cockroft-Gault14 equation was used to calculate baseline Somewhat surprisingly, baseline GFR was independent of GFR. The adjusted relative risks for total mortality and for impaired LVEF and was a stronger predictor of mortality than major CV events (nonfatal myocardial infarction [MI], non- either LVEF or New York Heart Association class. In fact, fatal stroke, CV death) were 1.65 and 1.58, respectively, in GFR was the strongest predictor of mortality of all factors subjects with GFRϽ60 mL/min compared with those with a analyzed, which included parameters of neurohormonal GFRϾ60 mL/ min.15 activation. Hillege et al also explored the prognostic ability of baseline Effect of Renal Disease on Individuals With renal function to predict the development of heart failure after Preexisting Stable CVD or Risk Factors for CVD an anterior-wall MI.20 Patients with a serum creatinine Ͼ180 A post hoc analysis of the Heart Outcomes and Prevention ␮mol/L (2.0 mg/dL) were excluded. Baseline GFR was Evaluation (HOPE) study examined the impact of baseline calculated with the Cockroft-Gault formula, and the 298 serum creatinine on the incidence of the composite primary patients were divided into tertiles of renal function. At 1 year outcome (CV death, MI, or stroke).16 The HOPE population of follow-up the incidence of congestive heart failure by included individuals with objective evidence of vascular tertile of decreasing GFR was 24.0%, 28.9%, and 41.2%. disease or diabetes combined with another CV risk factor and Risk of de novo congestive heart failure was 1.86-fold higher was designed to test the benefit of add-on ramipril versus in the lowest tertile (Ͻ81 mL/min) than in the highest tertile placebo in this population. Patients with heart failure or a (Ͼ103 mL/min). As the mean GFR in the lowest tertile was serum creatinine concentration Ͼ2.3 mg/dL were excluded. 67.0 mL/min, the study by Hillege et al highlights the impact The follow-up period was Ϸ5 years. There were 980 subjects of even mild GFR reductions on cardiac outcomes. with mild renal insufficiency (serum creatinineՆ1.4 mg/dL) In a post hoc analysis of the Valsartan in Acute Myocardial and 8307 subjects with normal renal function (serum creati- Infarction Trial (VALIANT), Anavekar et al examined the nineϽ1.4 mg/dL). The cumulative incidence of the primary relationship between baseline renal function and adverse outcome was 22.2% in individuals with mild renal insuffi- outcomes in 14 527 subjects with acute MI complicated by ciency versus 15% in those with normal renal function clinical or radiologic signs of heart failure and/or left ven- (PϽ0.001). The impact of renal insufficiency was indepen- tricular dysfunction.21 Subjects were randomly assigned to dent of both the baseline CV risk factors as well as the receive captopril, valsartan, or both, and they were followed treatment group. for a mean of 24.7 months. Individuals with a serum A similar relationship between renal function and CV creatinine Ͼ2.5 mg/dL were excluded from the study. The events was demonstrated in the Prevention of Events with primary end point was death from any cause, and secondary Angiotensin-Converting Enzyme Inhibition (PEACE) trial.17 end points included death from CV causes, heart failure, In PEACE, add-on trandolapril was compared with placebo in recurrent MI, resuscitation after cardiac arrest, stroke, and a a population with chronic stable coronary artery disease and composite of these.22 Anavekar et al21 stratified these subjects 88 Circulation July 3, 2007 into 4 groups; the investigators used the MDRD formula to events with a 1.3-fold increased relative risk for each natural estimate baseline GFR (mL/min per 1.73 m2) and found that, log increase of1Uinurine ACR. irrespective of treatment group, there was a progressive A similar population was studied in the Reduction of End increase in both the primary end point as well as each of the points in NIDDM with the Angiotensin II Antagonist Losar- secondary end points as GFR declined across the 4 groups. tan (RENAAL). Here, 1513 persons with type 2 diabetes, These findings remained significant even when an extensive, hypertension, and macroalbuminuria (mean baseline ACR, 70-candidate, variable model was used to adjust for higher 1810 mg/g) were randomized to either losartan or placebo comorbidities in patients with the poorest renal function. If and followed for a mean of 3.4 years. The primary end point the group with a GFR Ͼ75 mL/min per 1.73 m2 is considered was the same as in IDNT, namely a composite of mainly the reference point, the adjusted hazard ratio for adverse CV adverse nephrological events (serum creatinine doubling, events was 1.10 in the GFR group between 60.0 to 74.9 ESRD, or death), and, consistent with IDNT, the angiotensin mL/min per 1.73 m2 and 1.49 in the GFR group Ͻ45.0 antagonist provided superior nephroprotection but conferred mL/min per 1.73 m2. When GFR was analyzed as a contin- no statistically significant benefit on the secondary CV uous variable, each decrease in GFR of 10 mL/min per 1.73 outcomes,28 although de novo heart failure was less fre- m2 below 81.0 was associated with a 1.1-fold increase in risk quently noted in the losartan group. Nevertheless, in a post of death and nonfatal CV complications.21 hoc analysis of RENAAL, baseline albuminuria was again shown to be a predictor of both the prespecified composite Epidemiological Links Between Albuminuria CV end point (composite of MI, stroke, first hospitalization and Adverse Cardiovascular Events for heart failure or unstable angina, coronary or peripheral Renal disease may not only be identified by low GFR but also revascularization, or CV death) as well as of heart failure by the presence of abnormal quantities of albumin in the alone. With subjects stratified into 3 groups on the basis of urine. In fact, the appearance of pathological albuminuria baseline ACR (Ͻ1500, 1500 to 3000, Ͼ3000 mg/g), compar- often precedes the functional deterioration that is evidenced ison of the highest tertile with the lowest revealed an adjusted by a decline in GFR. Importantly, albuminuria has also been hazard ratio of 1.92 for the composite CV end point and 2.70 shown to be a potent independent marker of CV risk in both for heart failure. In multivariate analysis, baseline albumin- diabetic and nondiabetic persons. Similar to GFR, the link uria was the strongest independent predictor of both these between albuminuria and adverse CV events was first recog- outcomes. Perhaps more significant was the finding that the nized in the more overt situations of macroalbuminuria (urine change in urine albumin excretion from baseline to 6 months albumin:creatinine ratio [ACR] Ͼ300 mg/g),23,24 and then was the only dynamic correlate of adverse CV outcomes. A this link was extended to more modest elevations such as 50% reduction in baseline albuminuria translated into an 18% microalbuminuria (ACR, 30 to 300 mg/g).25 More recently, it reduction in the composite CV end point and a 27% reduction has become increasingly recognized that CV risk begins to in the risk of heart failure. Thus, albuminuria is not only a risk rise within currently defined normal levels of albuminuria factor for adverse CV outcomes but may also be a therapeutic (ACRϽ30 mg/g). Thus, urinary albumin is a continuous CV target or an indicator of therapeutic response.29 risk factor, whereas microalbuminuria is a designated threshold for renal functional deterioration in individuals with and CVD in Patients With Microalbuminuria without diabetes. Microalbuminuria also correlates with adverse CV events. In a multivariate analysis of CHD mortality in a type-2 diabetic CVD in Patients With Macroalbuminuria population, Mattock et al reported that microalbuminuria was The Irbesartan Diabetic Nephropathy Trial (IDNT) enrolled the strongest predictor of adverse CV outcomes with an odds subjects with type 2 diabetes, hypertension, and macroalbu- ratio of 10.02, which outranked smoking (odds ratio, 6.52), minuria.26 A total of 1715 subjects with mean urine ACR of diastolic blood pressure (odds ratio, 3.20), and serum choles- 1416.2 mg/g were randomized into 3 treatment groups that terol (odds ratio, 2.32).30 received irbesartan, amlodipine, or placebo and were fol- The HOPE study investigators reported on the risk of CV lowed for a mean period of 2.6 years. The primary outcome events associated with baseline ACR Ͼ2.0 mg/mmol (equiv- of the main trial was a renal-centric composite of serum alent to 17.7 mg/g). This amount of albuminuria was present creatinine doubling, ESRD, or death. Although irbesartan at baseline in 1140 (32.6%) subjects of the diabetic cohort proved to be the superior treatment with respect to the and in 823 (14.8%) subjects of the nondiabetic cohort. In the primary outcome, no difference was detected between treat- overall population a baseline ACR Ͼ2.0 mg/mmol increased ment groups on the secondary outcome of CV events. With the adjusted relative risk of CV events (1.83), all-cause death this data, a post hoc analysis was performed by Anavekar et (2.09), and hospitalization for congestive heart failure (3.23). al27 to assess the relationship between baseline albumin The impact of microalbuminuria on the primary composite excretion and the CV composite (CV death, nonfatal MI, outcome (CV death, MI, or stroke) was significant in both hospitalization for heart failure, stroke, amputation, and diabetics (relative risk, 1.97) and nondiabetics (relative risk, coronary and peripheral revascularization). A univariate anal- 1.61).31 ysis revealed that the proportion of patients who experienced The ability of microalbuminuria to predict adverse CV the CV end point progressively increased with increasing events is not restricted to a high-risk population like that of quartiles of baseline urine ACR. A multivariate analysis the HOPE trial. In fact, Hillege et al demonstrated the ability confirmed albuminuria as an independent risk factor for CV of microalbuminuria to predict CV and non-CV mortality in Schiffrin et al Kidney Disease and the Cardiovascular System 89 the general population.32 The investigators mailed medical a statistically significant stepwise increase in the primary questionnaires and a vial to collect early morning urine composite end point that started with the group with the low samples to all inhabitants of the city of Groningen between baseline/low year-1 group ratios (5.5%). Intermediate risk 1997 and 1998. More than 40 000 subjects responded and was found in the groups with low baseline/high year-1 (8.6%) were followed for a mean period of 961 days. Vital statistics ratios and high baseline/low year-1 ratios (9.4%). The highest and the causes of death were available from government risk group had high baseline/high year-1 values (13.5%). registries. The percentage of subjects who manifested base- These results were independent of in-treatment blood pres- line microalbuminuria was 22.5% in those who succumbed to sure and indicated that reductions in urine ACR over time CV death, 16.0% in patients who died as a result of non-CV translated into diminished CV risk. death, and 7.0% in patients who remained alive at the end of the study period. After adjustment for other known CV risk Epidemiological Links Between Serum Cystatin C factors, a doubling of the urine albumin excretion rate was and Adverse Cardiovascular Events associated with a relative risk of 1.29 for CV mortality and Recently, serum cystatin C has gained recognition as an 1.12 for non-CV mortality. Here again, microalbuminuria excellent endogenous marker of kidney function. Cystatin C outranked the predictive power of other classic CV risk is a cysteine proteinase with a molecular weight of 13 kDa factors. that is produced by almost all human cells and released into the blood. Cystatin C is freely filtered by the glomerulus and CVD in Patients With Albuminuria in the metabolized by proximal tubular cells, but it is not secreted Normal Range into the tubules. Cystatin C does not appear to be affected by The relationship between CV events and albuminuria has age, gender, or muscle mass, and there is evidence to suggest been extended further by several studies that suggest CV risk that it may be a more sensitive detector of incipient renal associated with increased levels of urinary albumin excretion dysfunction than creatinine-based estimates of GFR such as begins to emerge at levels previously defined as normal the Cockroft-Gault or MDRD formulas.37 Several recent (ACRϽ30 mg/g). Here too, the association appears to apply reports have indicated that cystatin C may be a better to a wide spectrum of patient populations. Analysis of the predictor of adverse CV events and all-cause mortality than HOPE study population supports albuminuria as a continuous either serum creatinine or creatinine-based estimating equa- risk factor for adverse CV events from an ACR as low as 0.5 tions.38–41 Ix et al categorized a population of 990 ambulatory mg/mmol (equivalent to 4.4 mg/g).33 For every subsequent persons with stable coronary heart disease into quartiles on 0.4 mg/mmol increase in the ratio, the adjusted hazard of the basis of baseline serum cystatin C levels and followed major CV events increased by 5.9%. these subjects for a median of 37 months.42 Subjects in the Similarly, Klausen et al34 reported that the risk of CV highest cystatin C quartile (Ն1.30 mg/dL), when compared events in the general population began to increase at urinary with the lowest quartile (Յ0.91 mg/dL), had a hazard ratio of albumin excretion levels below the defined threshold for 3.6 for all-cause mortality, 2.0 for CV events (composite of microalbuminuria. Klausen et al followed subjects in the CHD death, MI, and stoke), and 2.6 for incident heart failure. Third Copenhagen City Heart Study, which included These statistically significant results were adjusted for tradi- Ϸ10 200 randomly selected participants who underwent a tional CV risk factors. Potentially the most important finding detailed CV investigation program and provided a timed in this study was that higher cystatin C levels were predictive overnight urine sample. Subjects were classified into quartiles of these adverse outcomes even among people without on the basis of urinary albumin with a follow-up period that microalbuminuria or a diminished GFR as estimated by the ranged from 5 to 7 years. A urinary albumin excretion above MDRD formula (Յ60 mL/min per 1.73 m2). Currently the upper quartile of 4.8 ␮g/min (equivalent to ACR Ϸ9 cystatin C is not routinely measured in clinical practice. mg/g) was associated with an increased adjusted relative risk In summary, the presence of renal dysfunction, whether of 2.0 for CHD and 1.9 for death. detected by GFR, urine albumin excretion, or serum cystatin A post hoc analysis of the Losartan Intervention for C, predicts adverse CV outcomes. These relationships appear Endpoint Reduction in Hypertension (LIFE) study related not to extend to individuals with and without diabetes, those with just baseline albuminuria to CV risk but also the impact of and without preexisting CVD, and subjects with minimal to reduction of urinary albumin excretion on CV events.35 The marked perturbations in their renal parameters. LIFE study followed 8206 hypertensive individuals with LVH for a mean period of 4.8 years. The principal finding of Mechanisms of Cardiovascular Complications LIFE was that losartan proved superior to atenolol in the in Renal Disease reduction of the composite primary end point (CV death, As described in the preceding paragraphs, there is growing nonfatal stroke, nonfatal MI) for the same degree of blood evidence that relatively minor renal abnormalities such as a pressure reduction.36 In the post hoc study by Ibsen et al,35 the slightly reduced GFR or microalbuminuria even within the LIFE population was stratified into 4 groups according to normal range may be associated with increased risk of CV mean ACR at baseline (1.21 mg/mmol, equivalent to 10.6 events. One of the principal pathophysiological mechanisms mg/g) and at year 1 (0.67 mg/mmol, equivalent to 5.9 mg/g). involved in this association has been proposed to be endo- The percentage of subjects who experienced an adverse CV thelial dysfunction. Whether micro- or macroalbuminuria is event was reported on the basis of whether their ACR was an expression of generalized endothelial cell dysfunction above or below the mean values. This analysis demonstrated remains to be demonstrated. However, many studies have 90 Circulation July 3, 2007 demonstrated the correlation of albuminuria with endothelial exacerbate endothelial dysfunction and contribute to acceler- dysfunction as measured in peripheral blood vessels. Many of ation of atherogenesis. It has been postulated that glomerular the traditional and nontraditional CV risk factors that could endothelial dysfunction is an early feature of essential hyper- affect endothelial function can be found in association with tension that may precede blood pressure elevation. Microal- CKD. Related conditions such as diabetes, obesity, and buminuria may itself contribute to renal dysfunction, which hypertension, as well as the presence of renal dysfunction per progresses with uncontrolled blood pressure elevation. Endo- se lead to activation of the renin-angiotensin system, oxida- thelial dysfunction in turn may contribute to CV mortality tive stress, elevated ADMA, low-grade inflammation with already in mild renal insufficiency as suggested by the Hoorn increased circulating cytokines, and dyslipidemia, which are Study.59 Reduced bioavailability of nitric oxide (NO) appears all common pathophysiological mechanisms that play a role to be one of the main factors involved in chronic renal in the association of renal failure and CVD.43 failure–associated endothelial dysfunction,48,52,60 in large measure because of increased oxidative stress in the vascular Hypertension wall (see Dyslipidemia, Inflammation, and Oxidative Stress Hypertension in and of itself represents a powerful risk factor in Renal Disease).48,49 Prevalence of impaired endothelial for CVD in CKD and is almost invariably present in patients function, low-grade inflammation, and dyslipidemia associ- with renal failure. Sodium retention and activation of the ated with incipient and progressive renal disease may explain renin-angiotensin system have been considered the most the acceleration of atherosclerosis and, together with hyper- important mechanisms involved in the elevation of blood tension, may explain the high prevalence of coronary ische- pressure in subjects with kidney disease.44 Sympathetic ner- mia and CV events in CKD. The presence of hypertension, vous system activation also plays a role. Plasma catechol- sometimes difficult to control, in subjects with the previously amine concentrations are elevated, and increased nerve sym- mentioned risk factors may underlie the prevalence of cere- pathetic traffic has been demonstrated in renal failure.45,46 brovascular disease and stroke in patients with renal disease. The participation of the sympathetic system has become more Paradoxically, a recent report showed that lowest systolic complex with the recent discovery of renalase, a new regu- blood pressure was associated with stroke in stage 3 to 4 lator of cardiac function and blood pressure produced by the CKD.61 kidney. Xu et al47 screened libraries of the Mammalian Gene ADMA is a competitive inhibitor of NO synthase.62 Collection Project and identified a 37.8-kDa oxidase, which ADMA is synthesized potentially in many tissues, but in the contained flavin-adenine-dinucleotide, expressed mainly in CV system it is produced in the heart, endothelium, and glomeruli and proximal tubules of the kidney but also in smooth muscle cells. It is derived from the catabolism of cardiomyocytes and other tissues; the investigators called this proteins that contain methylated arginine residues, and it is oxidase renalase. Renalase metabolizes catecholamines in the released as the proteins are hydrolyzed. The synthesis of following order: dopamine 3 epinephrine 3 norepinephrine. ADMA requires the enzyme protein arginine methyltrans- In contrast to other oxidases, renalase is secreted into plasma ferase type I, which methylates arginine residues, and the and urine of healthy persons. However, it is not detectable in protein arginine methyltransferase type II forms symmetric uremic individuals. Recombinant renalase exerts a powerful dimethylarginine, which is a stereoisomer of ADMA and is and rapid hypotensive effect on rats. To what extent the not an inhibitor of NO synthase. ADMA and symmetric impairment of renalase production contributes to sympathetic dimethylarginine enter endothelial cells through the cationic hyperactivity and blood pressure elevation in CKD remains to amino acid yϩ transporter. The activity of this transporter be established. Also, endothelial dysfunction48–52 (see below) colocalizes with caveolin-bound NO synthase, which sug- ϩ and remodeling of blood vessels53 may participate not only in gests that y transporter activity may be a determinant of the vascular complications in patients with kidney disease but local concentrations of ADMA. The ADMA and symmetric also in the maintenance of elevated blood pressure. dimethylarginine compete with each other and L-arginine for Hypertension also plays a major role in cardiac damage in transport into the cell. Thus, ADMA may block entry of CKD via LVH induction.54,55 In addition, a reduction in L-arginine, with the resulting decrease in synthesis of NO. coronary reserve and capillary density that occurs in CKD ADMA is metabolized mainly by dimethylarginine dimeth- patients exposes them to coronary ischemia,56 which in turn ylaminohydrolase and cleared by the kidney. Exogenous leads to worsening of ventricular dysfunction. ADMA inhibits NO generation in vitro, and in humans it reduces forearm blood flow and cardiac output and increases Endothelial Dysfunction, Nitric Oxide systemic vascular resistance and blood pressure.63 Subpressor Bioavailability, and ADMA in Renal Disease ADMA infusion increases renovascular resistance, induces Impairment of endothelial function is recognized as one of intimal hyperplasia, and affects small and large vessels.64–66 the initial mechanisms that lead to atherosclerosis. Endothe- Plasma concentrations of ADMA are increased in association lial dysfunction, which occurs in both large and small with endothelial dysfunction and/or reduced NO production, arteries, is present in renal disease.51 Microalbuminuria, a particularly in renal failure.67,68 Increased ADMA in renal marker of glomerular hyperfiltration, has been correlated with failure may result from both increased activity of protein and may be a manifestation of impaired endothelial func- arginine methyltransferase and decreased metabolism by tion.57 Experimental evidence suggests that microvascular dimethylarginine dimethylaminohydrolase.69 It is unclear endothelial dysfunction participates in the mechanisms that whether endogenous ADMA concentrations increase suffi- lead to progression of renal disease,58 which in turn may ciently to inhibit NO production in vivo. Interestingly, plasma Schiffrin et al Kidney Disease and the Cardiovascular System 91 norepinephrine and ADMA concentrations are closely corre- TABLE 3. Effects of Renal Failure and Inflammation on lated in patients with ESRD and are likely to act through Lipoprotein and Endothelial Structure and Function 70 common mechanisms that contribute to CV events. ADMA HDL is now considered one of the strongest markers of atheroscle- Effects of renal failure: decreased synthesis of apoA-I; decreased LCAT 71 rosis. Elevated plasma concentrations of ADMA are asso- activity; increased apoC-III; increased triglycerides; decreased levels of ciated not only with endothelial dysfunction and atheroscle- mature HDL 72 rosis but predict mortality and CV complications in CKD Effects of inflammation: replacement of apoA-I with serum amyloid A; and ESRD.68 In subjects with mild to advanced CKD, plasma decreased levels of mature HDL; decreased paroxynase and AHH activity; ADMA was inversely related to GFR73 and was an indepen- decreased ability to protect against cytokine action of endothelium; dent risk marker for progression to ESRD and mortality.74 In decreased ability to reduce oxidized LDL the Mild to Moderate Kidney Disease Study, ADMA was Remnants significantly associated with progression of nondiabetic kid- Effects of renal failure: increased levels linked to increase apoC-III; ney disease.75 Elevated plasma ADMA has been shown to be decreased clearance; interaction with blood vessels to induce a marker of CV morbidity in early nephropathy associated vasoconstriction with type 1 diabetes.76 In the Ludwigshafen Risk and Car- LDL diovascular Health Study, ADMA independently predicted Effects of renal failure: accumulation of small dense atherogenic LDL; total and CV mortality in individuals with angiographic results in increased AngII and also upregulation of the AT1 receptor coronary artery disease.77 Although reduced bioavailability of Activation of the renin-angiotensin system NO and accumulation of ADMA cause endothelial dysfunc- Stimulates NAD(P)H oxidase, xanthine oxidase, etc., which leads to tion, there is little evidence for coronary artery endothelial production of superoxide; induces IL-6 and other cytokines as well as dysfunction in renal failure. Recently, Tatematsu et al78 PAI-1 gene expression; increased superoxide leads to decreased induced renal failure in dogs and evaluated coronary vasodi- bioavailability of nitric oxide, endothelial dysfunction, vascular remodeling, lator response to acetylcholine, which demonstrated blunted and hypertension responses in the CKD dogs. mRNA expression of dimethyl- HDL indicates high-density lipoprotein; apo, apolipoprotein; LACT, decreased arginine dimethylaminohydrolase-II and endothelial NO syn- lecithin cholesterol ester transferase; AHH, aryl hydrocarbon hydrolase; LDL, thase in coronary arteries were downregulated, which dem- low-density lipoprotein; AT, angiotensinogen; and PAI-1, plasminogen activator inhibitor 1. Adapted from Kaysen and Eiserich80 with permission from the onstrated a possible mechanism for coronary endothelial American Society of Nephrology. Copyright 2004. dysfunction in early stages of CKD.

Dyslipidemia, Inflammation, and Oxidative Stress to reduce proteinuria modestly, and results in a small reduc- in Renal Disease tion in the rate of loss of kidney function, especially in Individuals with CKD become progressively malnourished, populations with CVD.82 as evidenced by low levels of albumin, prealbumin, and The changes in lipoprotein composition and structure as transferrin, which has been suggested to be a mechanism for well as angiotensin II–mediated alterations in endothelial activation of inflammation.79 Diseases in which low-grade function stimulate and amplify the effect of inflammatory inflammation is found, such as diabetes and hypertension, are mechanisms.83 Between 30 and 50% of CKD patients have often associated with CKD. Thus it is difficult to conclude elevated serum levels of inflammatory markers such as whether there is a direct effect of renal failure on inflamma- C-reactive protein, fibrinogen, interleukin-6, tumor necrosis tion in early CKD. Renal failure causes changes in plasma factor-␣, factor VIIc, factor VIIIc, plasmin-antiplasmin com- components and endothelial structure and function that favor plex, D-dimer, and the adhesion molecules E-selectin, vascular injury, which may play a role as a trigger for VCAM-1 and ICAM-1.84,85 Mechanisms are unclear but inflammatory response. Dyslipidemia associated with increased inflammatory mediators have been attributed to CKD80,81 contributes to the inflammatory response in renal increased oxidative stress, accumulation of postsynthetically failure. The changes in blood-lipid composition and their modified proteins, advanced glycation end products, and relation to renal dysfunction and inflammation are summa- other agents normally cleared by the kidney. Thus, causes of rized in Table 3. Hepatic apolipoprotein A-I synthesis de- inflammation may include comorbidities, oxidative stress, creases and high-density lipoprotein levels fall. High-density infections, and hemodialysis-related factors that depend on lipoprotein is an important antioxidant and also protects the membrane biocompatibility and the dialysate.86 Progressive endothelium from the effects of proinflammatory cytokines. deterioration of renal function in CKD may lead to dyslipid- Apolipoprotein C-III, a competitive inhibitor of lipoprotein emia or accumulation of uremic toxins, which can stimulate lipase, is increased in CKD. Serum triglyceride levels in- oxidative stress and inflammation, which in turn may con- crease as a result of accumulation of intermediate-density tribute to endothelial dysfunction and progression of lipoprotein, which comprise very low-density lipoprotein and atherosclerosis. chylomicron remnants. These impair endothelial function and A major contributor to the increase in circulating inflam- are associated with CVD. matory biomarkers in CKD may be enhanced oxidative Because dyslipidemia associated with CKD appears to play stress.85–87 Mechanisms of oxidative stress in uremia may a role in the enhanced CV risk of these patients, treatment of involve activation of reduced nicotinamide adenine dinucle- dyslipidemia conversely should reduce proteinuria and ame- otide (NAD(P)H) oxidase, xanthine oxidase, uncoupled en- liorate the progression of CKD. Indeed, statin therapy appears dothelial NO synthase, myeloperoxidase (MPO), and mito-

92 Circulation July 3, 2007 chondrial oxidases. NAD(P)H oxidase is probably the most TABLE 4. Promoters and Inhibitors of Vascular Calcification important source in the vasculature, and it is stimulated by Promoters of calcification angiotensin II and other agents (see Renin-Angiotensin Sys- Traditional factors tem).88 Increased production of reactive oxygen species (ROS) by uncoupled endothelial NO synthase49 as well as Older age, male gender, hypertension, diabetes, smoking high LDL cholesterol, low HDL cholesterol, genetics reduced inactivation of ROS by antioxidant systems such as Uremia-related factors superoxide dismutase87 also play an important role. MPO is present in neutrophils and monocytes/macrophages, and has Uremic serum, hyperphosphatemia, increased CaϫP product, been shown to be expressed to a significant degree in human exogenous vitamin D therapy, elevated parathyroid hormone levels, dialysis vintage, calcium load and hypercalcemia, chronic inflammation, 89 atheroma. It may thus play a role in the accelerated warfarin, elevated leptin levels atherosclerosis of renal failure. It has recently been reported Inhibitors of calcification that a single nucleotide polymorphism in the promoter region Circulating inhibitors of the MPO gene associated with reduced expression of MPO is accompanied by a lower prevalence of CVD in ESRD Fetuin-A, bone morphogenetic protein-7, parathyroid hormone–related peptide, HDL, magnesium patients.90 Active MPO is released from white blood cells during hemodialysis, and this could be a mechanism whereby Locally acting inhibitors MPO plays a role in vascular injury in subjects with ESRD. Matrix Gla protein, osteopontin, pyrophosphate, osteoprotegerin, genetics Renin-Angiotensin System Adapted from Qunibi93 with permission from the American Society of Activation of the renin-angiotensin system occurs in many Nephrology. Copyright 2005. forms of renal disease. Angiotensin II stimulates NAD(P)H oxidase, which leads to generation of superoxide anion and phosphate (P) product without presence of an active osteo- contributes to endothelial dysfunction and vascular remodel- genic process, and it must be differentiated from other forms ing and growth.91 Mechanisms whereby the renin-angiotensin of calcification of the skin that do not affect blood vessels and system may be activated by kidney disease are multiple and from medial calcific sclerosis, which affects larger vessels. It beyond the scope of the present review, but such mechanisms is a rare complication of renal failure present in up to 4% of may in part depend on the adaptation to loss of renal mass that hemodialysis patients, typically in obese diabetic females, results in changes in renal hemodynamics. When angiotensin associated often with secondary hyperparathyroidism, hyper-

II acts through the AT1 receptor, it stimulates generation of calcemia, hyperphosphatemia, malnutrition, and sometimes ROS by NAD(P)H oxidase and other enzymes systems, with warfarin therapy or hypercoagulability. However, al- which leads to upregulation of inflammatory mediators, though warfarin and hypercoagulability have both been im- which include cytokines, chemokines, adhesion molecules, plicated, the latter on the basis of an association of protein C and plasminogen activator inhibitor 1, and superoxide scav- deficiency and calciphylaxis, some studies suggest that nei- enging of NO. These events, together with the mechanisms ther hypercoagulability nor warfarin play a role in this rare already mentioned, promote endothelial dysfunction, vascular condition.99 Similarly, parathyroidectomy has been reported remodeling, and the progression of atherosclerosis.92 to lead to the resolution of the skin ulcers of calciphylaxis in some series100 but not all.101 Vascular Calcification, Inducers and Inhibitors Mechanisms involved in vascular calcification in CKD of Calcification, and the Role of Phosphate in include passive precipitation of Ca and P in the presence of Renal Failure excessively high extracellular concentrations, effects of in- Accelerated calcifying atherosclerosis and valvular heart ducers of osteogenic transformation and hydroxyapatite for- disease occur with high frequency in CKD.93–95 A recent mation, and deficiency of calcification inhibitors.94,102 Table study showed that 40% of patients with CKD and a mean 4 summarizes some of the inducers and inhibitors of vascular GFR 33 mL/min exhibited coronary artery calcification calcification that induce an osteoblast phenotype in vascular compared with 13% in matched control subjects with no renal smooth muscle cells in CKD. Patients with ESRD often have impairment.96 Calcification can be found in atherosclerotic severe changes in their CaϫP product, which induces a trend plaques and in the vascular media, smooth muscle cells, and toward ectopic calcification. Aortic stiffening associated with elastic laminae of large elastic and medium muscular arteries calcification103 will cause LVH, which results in increased as well as in cardiac valves.93–95 Subjects with renal failure CV risk. Increased phosphate levels are also a source of who exhibit medial calcification are typically middle-aged increased CV risk, probably as a result of worsening vascular and have been dialyzed for some time, although some calcification.104 Precipitation associated with a raised CaϫP individuals may already have calcified vessels before dialy- product may contribute to soft-tissue calcification, but calci- sis.97 There is a specific dialysis-related type of vascular fication of the media of blood vessels appears to involve calcification called calciphylaxis, or calcific uremic arteriopa- active transport through the Na-P cotransporter PiT-1 which thy, that is characterized by diffuse calcification of the media occurs in part as a result of a phenotypic switch of vascular of small to medium arteries and arterioles with intimal smooth muscle cells into osteoblast-like cells as a conse- proliferation and thrombosis that results in skin ulcers98 and quence of high intracellular Ca and P, which induce osteo- can lead to life-threatening skin necrosis or acral gangrene. genic differentiation of smooth muscle cells.94,95 In an in vitro Calciphylaxis is the result of an elevated calcium (Ca) ϫ model, elevated Ca or P induced human vascular smooth Schiffrin et al Kidney Disease and the Cardiovascular System 93 muscle cell calcification, which was initiated by release of may occur in ESRD as a consequence of removal of PPi membrane-bound matrix vesicles and apoptotic bodies.105 during hemodialysis,120 which may be one of the mechanisms Vesicles released by cells exposed to Ca and P calcified to an that contribute to accelerated vascular calcification in hemo- important degree, but those released in the presence of serum dialysis patients. were minimally calcified and were found to contain the BMPs are important regulators of bone formation. They are calcification inhibitors fetuin-A and matrix Gla protein members of the largest subclass of the transforming growth (MGP) (see next paragraph). Thus, vascular calcification is a factor-␤ superfamily and have been localized in areas of cell-mediated process regulated by calcification inhibitors, vascular calcification.121 BMP-2 is generated from a 60-kDa functional impairment of which leads to accelerated vascular precursor, which is processed to an 18-kDa monomer that calcification. associates with another monomer to form the active ho- Among the inhibitors of calcification, fetuin-A (␣2-Schmid modimer, which then binds to its receptor. The BMPR is a Heremans glycoprotein; molecular weight, 60 kDa), which is heterodimer that consists of types 1 and 2 serine/threonine produced by the liver and circulates in blood, appears to be of kinases. BMPR2 phosphorylates BMPR1, which in turn prime importance. Fetuin-A has a transforming growth phosphorylates the Smad 1/5/8 complex, which, with Smad 4, factor-␤ receptor II–like domain and may function as a then modulates target gene expression.122 Of the different soluble transforming growth factor-␤ antagonist that inter- BMPs, BMP-2 or BMP-4 may induce osteogenic differenti- feres with insulin receptor autophosphorylation and tyrosine ation of vascular smooth muscle cells through induction of kinase activity.106 It forms stable colloidal spheres with Ca transcription factors Cbfa1, osterix, and the msh homeobox and P (calciprotein particles) and is the main component of a homolog MSX-2. Other effects of BMP-2/BMP-4 that con- high molecular mass complex that contains Ca, P, and tribute to calcification of the vasculature are the triggering of MGP.107 Low serum fetuin-A levels in subjects with CKD apoptosis and inhibitory effects on MGP. In addition, BMP-4 have been associated with enhanced vascular calcification102 has been shown to exert vascular effects that lead to increased and increased CV mortality.108,109 MGP belongs to a family oxidative stress and impaired endothelial function, and to of N-terminal ␥-carboxylated (Gla) proteins that require a what extent these effects are related to media calcification vitamin K–dependent ␥-carboxylation for their biological remains to be established. BMP-7 on the other hand inhibits activation and prevent bone morphogenetic protein (BMP)- vascular calcification by upregulation of ␣-smooth muscle 2/BMP receptor-2 (BMPR2) interactions.110 The actin expression via induction of p21 and upregulation of ␥-carboxylated MGP, but not the non–␥-carboxylated MGP, Smad 6 and 7. BMP-7 is expressed mainly in the kidney, and is carried in plasma by fetuin-A. Mice that lack MGP develop its expression decreases with progression of renal failure, spontaneous calcification of arteries and cartilage.111 Ele- which results in reduction of its ability to inhibit calcification. vated concentrations of MGP may be found in the vicinity of Lowering of BMP-7 affects bone metabolism with conse- atherosclerotic plaques112 and have been shown to be associ- quent increase in serum phosphate levels, which adversely ated with calcification of vascular smooth muscle cells in affects the CaϫP product and induces phenotypic changes in vitro.113 MGP levels in blood have been reported to correlate vascular smooth muscle cells, which leads to metastatic negatively with coronary artery calcification.114 calcification. Osteoprotegerin regulates osteoclast activation. It acts as a Increased leptin levels may participate in the process of soluble decoy receptor that prevents the binding of the vascular calcification in CKD because serum leptin concen- osteoclast stimulator receptor activator of nuclear factor-␬B trations are increased in renal failure as a result of reduced ligand to its receptor. Osteoprotegerin deficiency in mice leptin excretion. Leptin induces heterotopic calcification via leads to vascular calcification, but its mechanism of action its receptors in the hypothalamus that induce an increased has not been elucidated.115 Osteoprotegerin levels are ele- sympathetic activity, which stimulates osteoblast vated in ESRD,116 correlate with vascular calcification, and ␤-adrenergic receptors.123 Leptin increases bone marrow predict mortality in hemodialysis patients, in particular in stem cell differentiation into an osteoprogenitor phenotype individuals with high C-reactive protein levels.117 Interest- and may act on vascular smooth muscle cells to induce ingly, lower soluble receptor activator of nuclear factor-␬B calcification,124 in part by an increase in ROS generation and ligand concentrations were associated with better induction of BMP-2.95 outcomes.117 In summary, BMP-2/BMP-4 binds the BMPR1/BMPR2 Elevated pyrophosphate (PPi) concentrations prevent hy- receptor complex and phosphorylates the regulatory Smads, droxyapatite crystal formation and calcification. PPi is syn- which then signal downstream to upregulate the expression of thesized by the rate-limiting enzyme nucleotide pyrophos- transcription factors Cbfa1, osterix, and MSX-2. BMP-4 also phatase phospho-diesterase-1. Mice that lack nucleotide stimulates generation of ROS. Cbfa1 expression is also pyrophosphatase phospho-diesterase-1 develop PPi defi- enhanced by ROS, leptin, vitamin D, high phosphate levels, ciency, which results in an altered vascular smooth muscle and PiT-1.87 The result is a phenotypic change in vascular cell phenotype and vascular calcification.118 The cellular PPi smooth muscle cells to an osteogenic phenotype. These cells exporter ankyrin, which is encoded by the transmembrane express alkaline phosphatase and produce hydroxyapatite transporter progressive ankylosis locus, mediates PPi exit crystals. Calcification inhibitors such as fetuin-A, MGP, from cells.119 Vascular calcification may also result from osteoprotegerin, osteopontin, BMP-7, and Smad 6 antagonize enhanced activity of the membrane-bound tissue-nonspecific BMP-2/BMP-4 signaling and metastatic calcification alkaline phosphatase, which degrades PPi to P. PPi deficiency (Figure). 94 Circulation July 3, 2007

Angiotensin II Thrombosis Inflammation Fetuin-A Endothelium LOX-1

Hydroxyapatite PO Fetuin-A 4 Ca x P BMP-2/4 OPG Pit-1 Calcification Leptin MGP Calcification Osteogenic promoters Vitamin D OPN inhibitors VSMC ALP ROS BMP-7 PPi

NADH/ VSMC NADPH oxidase AT1R

Mechanisms depicted here are some of those involved in vascular calcification in chronic kidney disease. Activation of the renin- angiotensin system results in stimulation of AT1R, which stimulates reduced NAD(P)H oxidase, the main source of vascular ROS. BMP- 2/4 binds the BMP receptor BMPR1/BMPR2 receptor complex and phosphorylates the Smad 1/5/8 complex, which, with Smad 4, signals downstream to upregulate expression of transcription factors Cbfa1, osterix, and MSX-2. Cbfa1 expression is also enhanced by ROS, leptin, vitamin D, increased CaϫP product, or high PO4 levels induced by Pit-1, the sodium-phosphate cotransporter, activated in part as a result of the phenotypic switch of VSMCs into osteoblast-like cells. VSMCs that have acquired an osteogenic phenotype express ALP and produce hydroxyapatite crystals. Calcification inhibitors such as PPi inhibit hydroxyapatite precipitation, whereas fetuin-A, MGP, OPG, OPN, BMP-7, and Smad 6 antagonize BMP2/4 signaling and calcification. AT1R indicates angiotensin AT1 receptor; NAD(P)H, nicotinamide adenine dinucleotide; ROS, reactive oxygen species; BMP, bone morphogenic protein; PO4, phosphate; VSMC, vascular smooth muscle cells; ALP, alkaline phosphatase; PPi, pyrophosphate; MGP, matrix Gla protein; OPG, osteoprotegerin; and OPN, osteopontin.

Conclusion 3. Shlipak MG, Heidenreich PA, Noguchi H, Chertow GM, Browner WS, The present review underlines the CV risk to which patients McClellan MB. Association of renal insufficiency with treatment and outcomes after myocardial infarction in elderly patients. Ann Intern with CKD are exposed and summarizes some of the mecha- Med. 2002;137:555–562. nisms that lead to the increased risk of adverse CV events. It 4. Foley RN, Parfrey PS, Sarnak M. Clincial epidemiology of cardiovas- is also clear that some of this risk is modifiable and can be cular disease in chronic renal disease. Am J Kidney Dis. 1998;32: improved with currently available therapy by reduction of 112–119. 5. Herzog CA, Ma JZ, Collins AJ. Poor long-term survival after acute blood pressure according to guidelines, aggressive treatment myocardial infarction among patients on long-term dialysis. N Engl of dyslipidemia, control of protein intake, minimization of J Med. 1998;339:799–805. bone resorption, optimization of Ca and P metabolism, and 6. Parfrey PS, Foley RN. The clinical epidemiology of cardiac disease in combat of hypercoagulability, with the caveat that warfarin chronic uremia. J Am Soc Nephrol. 1999;10:1606–1615. 7. Foley RN, Parfrey PS, Harnett JD. Clinical and echocardiographic may be implicated in calciphylaxis of the latter. Therapeutic disease in patients starting end-stage renal disease therapy. Kidney Int. aspects that may require new approaches include manage- 1995;47:186–193. ment of the increased oxidative stress and low-grade inflam- 8. Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalent LVH in mation, as well as development of novel strategies to increase the predialysis population: identifying opportunities for intervention. the concentrations of inhibitors of calcification and to mod- Am J Kidney Dis. 1996;27:347–354. 9. Tucker B, Fabbian F, Giles M, Thuraisingham RC, Raine AE, Baker LR. erate the agents that promote calcification. Left ventricular hypertrophy and ambulatory blood pressure monitoring in chronic renal failure. Nephrol Dial Transplant. 1997;12:724–728. Sources of Funding 10. Culleton BF, Larson MG, Wilson PWF, Evans JC, Parfrey PS, Levy D. The work by Dr Schiffrin was supported by a Canada Research Chair Cardiovascular disease and mortality in a community-based cohort with on Hypertension and Vascular Research, and by grants 37917 and mild renal insufficiency. Kidney Int. 1999;56:2214–2219. 82790 from the Canadian Institutes of Health Research. 11. Go AS, Chertow GM, Fan D, McCullock CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004;351:1296–1305. Disclosures 12. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more None. accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal References Disease Study Group. Ann Intern Med. 1999;130:461–470. 1. Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm 13. Shulman NB, Ford CE, Hall WD, Blaufox MD, Simon D, Langford HG, LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, Parfrey P, Schneider KA. Prognostic value of serum creatinine and the effect of Pfeffer M, Raij L, Spinosa DJ, Wilson PW. Kidney disease as a risk treatment of hypertension on renal function. Results from the hyper- factor for development of cardiovascular disease: a statement from the tension detection and follow-up program. The Hypertension Detection American Heart Association Councils on Kidney in Cardiovascular and Follow-up Program Cooperative Group. Hypertension. 1989; Disease, High Blood Pressure Research, Clinical Cardiology, and Epi- 13(suppl):I80–I93. demiology and Prevention. Hypertension. 2003;42:1050–1065. 14. Cockroft DW, Gault MH. Prediction of creatinine clearance from serum 2. Tonelli M, Wiebe N, Culleton B, House A, Rabbat C, Fok M, McAlister creatinine. Nephron. 1976;16:31–41. F, Garg AX. Chronic kidney disease and mortality risk: a systematic 15. Ruilope LM, Salvetti A, Jamerson K, Hanson L, Warnold I, Wedel H, review. J Am Soc Nephrol. 2006;17:2034–2047. Zanchetti A. Renal function and intensive lowering of blood pressure in

Schiffrin et al Kidney Disease and the Cardiovascular System 95

hypertensive participants of the Hypertension Optimal Treatment (HOT) Albuminuria and cardiovascular events, death and heart failure in Study. J Am Soc Nephrol. 2001;12:218–225. diabetic and non-diabetic individuals. JAMA. 2001;286:421–46. 16. Mann JFE, Gerstein HC, Pogue J, Bosch J, Yusuf S. Renal insufficiency 34. Klausen K, Borch-Johnsen K, Feldt-Rasmussen B, Jensen G, Clausen P, as a predictor of cardiovascular outcomes and the impact of ramipril: Scharling H, Appleyard M, Jensen JS. Very low levels of microalbu- The HOPE randomized trial. Ann Intern Med. 2001;134:629–636. minuria are associated with increased risk of coronary heart disease and 17. Solomon SD, Rice MM, Jablonski KA, Jose P, Domanski M, Sabatine death independently of renal function, hypertension, and diabetes. Cir- M, Gersh BJ, Rouleau J, Pfeffer MA. Renal function and effectiveness culation. 2004;110:32–35. of angiotensin-converting enzyme inhibitor therapy in patients with 35. Ibsen H, Olsen MH, Wachtell K, Borch-Johnsen K, Lindholm LH, chronic stable coronary disease in the prevention of events with ACE Mogensen CE, Dahlöf B, Devereux RB, De Faire U, Fyhrquist F, Julius inhibition (PEACE) trial. Circulation. 2006;114:26–31. S, Kjeldsen SE, Lederballe-Pedersen O, Nieminen MS, Omvik P, Oparil 18. Hillege HL, Girbes AR, De Kam PJ, Boomsma F, De Zeeuw D, Char- S, Wan Y. Reduction in albuminuria translates to reduction in cardio- lesworth A. Renal function, neurohormonal activation and survival in vascular events in hypertensive patients: losartan intervention for end patients with chronic heart failure. Circulation. 2000;102:203–210. point reduction in hypertension study. Hypertension. 2005;45:198–202. 19. Hampton JR, Van Veldhuisen DJ, Kleber FX, Cowley AJ, Ardia A, 36. Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, De Faire U, Block P, Cortina A, Cserhalmi L, Follath F, Jensen GK, Lie KL, Mancia Fyhrquist F, Ibsen H, Kristiansson K, Lederballe-Pedersen O, Lindholm G, Skene AM. Randomised study of effect of ibopamine on survival in LH, Nieminen MS, Omvik P, Oparil S, Wedel H, LIFE Study Group. patients with advanced severe heart failure: second prospective ran- Cardiovascular morbidity and mortality in the Losartan Intervention For domised study of ibopamine on mortality and efficacy (PRIME II) End point reduction in hypertension study (LIFE): a randomised trial investigators. The Lancet. 1997;349:971–977. against atenolol. The Lancet. 2002;359:995–1003. 20. Hillege HL, Van Gilst WH, Van Veldhuisen DJ, Navis G, Grobbee DE, 37. Perkins BA, Nelson RG, Ostrander BE, Blouch KL, Krolewski AS, De Graeff PA. Accelerated decline and prognostic impact of renal Myers BD, Warram JH. Detection of renal function decline in patients function after myocardial infarction and the benefits of ACE inhibition: with diabetes and normal or elevated GFR by serial measurements of the CATS randomized trial. Eur Heart J. 2003;24:412–420. serum cystatin C concentration: results of a 4-year follow-up study. JAm 21. Anavekar NS, McMurray JJV, Velazquez EJ, Solomon SD, Kover L, Soc Nephrol. 2005;16:1404–1412. Rouleau JL, White HD, Nordlander R, Maggioni A, Dickstein K, Zelen- 38. Shlipak MG, Sarnak MJ, Katz R, Seliger SL, Newman AB, Siscovick kofske S, Leimberger JD, Califf RM, Pfeffer MA. Relation between DS, Stehman-Breen C. Cystatin C and the risk of death and cardiovas- renal dysfunction and cardiovascular outcomes after myocardial cular events among elderly persons. N Engl J Med. 2005;352: infarction. N Engl J Med. 2004;351:1285–1295. 2049–2060. 22. Pfeffer MA, McMurray JJ, Velazquez EJ, Rouleau JL, Kober L, 39. Sarnak MJ, Katz R, Stehman-Breen CO, Fried LF, Jenny NS, Psaty BM, Maggioni AP, Solomon SD, Swedberg K, Van de Werf F, White H, Newman AB, Siscovick D, Shlipak MG. Cystatin C concentration as a Leimberger JD, Henis M, Edwards S, Zelenkofske S, Sellers MA, Califf risk factor for heart failure in older adults. Ann Intern Med. 2005;142: RM. Valsartan, captopril, or both in myocardial infarction complicated 497–505. by heart failure, left ventricular dysfunction, or both. N Engl J Med. 40. Fried LF, Katz R, Sarnak MJ, Shlipak MG, Chaves PH, Jenny NS, 2003;349:1893–1906. Stehman-Breen C, Gillen D, Bleyer AJ, Hirsch C, Siscovick D, Newman 23. Kannel WB, Stampfer MJ, Castelli WP, Verter J. The prognostic sig- AB. Kidney function as a predictor of noncardiovascular mortality. JAm nificance of proteinuria: the Framingham study. Am Heart J. 1985;108: Soc Nephrol. 2005;16:3728–3735. 1347–1352. 41. Shlipak MG, Wassel Fyr CL, Chertow GM, Harris TB, Kritchevsky SB, 24. Grimm RH, Svendsen KH, Kasiske B, Keane WF, Wahi MM. Pro- Tylavsky FA, Satterfield S, Cummings SR, Newman AB, Fried LF. teinuria is a risk factor for mortality over 10 years of follow-up. MRFIT Cystatin C and mortality risk in the elderly: the health, aging, and body Research Group. Kidney Int. 1997;63:S10–S14. composition study. J Am Soc Nephrol. 2006;17:254–261. 25. Keane WF, Eknoyan G, NKF PC. Proteinuria, albuminuria, risk, 42. Ix JH, Shlipak MG, Chertow GM, Whooley MA. Association of cystatin assessment, detection, elimination (PARADE): a position paper of the C with mortality, cardiovascular events, and incident heart failure National Kidney Foundation. Am J Kidney Dis. 1999;33:1004–1010. among persons with coronary heart disease. Circulation. 2006;115: 26. Lewis EJ, Hunsicker LG, Clarke WR, Berl T, Pohl MA, Lewis JB, Ritz 173–179. E, Atkins RC, Rohde R, Raz I, Collaborative Study Group. Renopro- 43. Amann K, Wanner C, Ritz E. Cross-talk between the kidney and the tective effect of the angiotensin-receptor antagonist irbesartan in patients cardiovascular system. J Am Soc Nephrol. 2006;17:2112–2119. with nephropathy due to type 2 diabetes. N Engl J Med. 2001;345: 44. Guyton AC, Coleman TG, Wilcox CS. Quantitative analysis of the 851–860. pathophysiology of hypertension. J Am Soc Nephrol. 1999;10: 27. Anavekar NS, Gans DJ, Berl T, Rohde RD, Cooper W, Bhaumik A, 2248–2249. Hunsicker LG, Rouleau JL, Lewis JB, Rosendorff C, Porush JG, Drury 45. Converse RL Jr, Jacobsen TN, Toto RD, Jost CM, Cosentino F, Fouad- PL, Esmatjes E, Raz I, Vanhille P, Locatelli F, Goldhaber S, Lewis EJ, Tarazi F, Victor RG. Sympathetic overactivity in patients with chronic Pfeffer MA. Predictors of cardiovascular events in patients with type 2 renal failure. N Engl J Med. 1992;327:1912–1918. diabetic nephropathy and hypertension: A case for albuminuria. Kidney 46. Neumann J, Ligtenberg G, Klein II, Koomans HA, Blankestijn PJ. Int. 2004;66(suppl 92):S50–S55. Sympathetic hyperactivity in chronic kidney disease: pathogenesis, 28. Brenner BM, Cooper ME, De Zeeuw D, Mitch WE, Parving HH, clinical relevance, and treatment. Kidney Int. 2004;65:1568–1576. Remuzzi G, Snapinn SM, Zhang ZX, Shahinfar S, RENAAL Study. 47. Xu J, Li G, Wang P, Velazquez H, Yao X, Li Y, Wu Y, Peixoto A, Effects of losartan on renal and cardiovascular outcomes in patients with Crowley S, Desir GV. Renalase is a novel, soluble monoamine oxidase type 2 diabetes and nephropathy. N Engl J Med. 2001;345:861–869. that regulates cardiac function and blood pressure. J Clin Invest. 2005; 29. De Zeeuw D, Remuzzi G, Parving HH, Keane WF, Zhang ZX, Shahinfar 115:1275–1280. S, Snapinn S, Cooper ME, Mitch WE, Brenner BM. Albuminuria, a 48. Wever R, Boer P, Hijmering M, Stroes E, Verhaar M, Kastelein J, therapeutic target for cardiovascular protection in type 2 diabetic Versluis K, Lagerwerf F, Van Rijn H, Koomans H, Rabelink T. Nitric patients with nephropathy. Circulation. 2004;110:921–927. oxide production is reduced in patients with chronic renal failure. Arte- 30. Mattock MB, Barnes DJ, Viberti G, Keen H; Burt D, Hughes JM; rioscler Thromb Vasc Biol. 1999;19:1168–1172. Fitzgerald AP, Sandhu B, Jackson PG. Microalbuminuria and coronary 49. Vaziri ND, Ni Z, Oveisi F, Liang K, Pandian R. Enhanced nitric oxide heart disease in NIDDM: an incidence study. Diabetes. 1998;47: inactivation and protein nitration by reactive oxygen species in renal 1786–1792. insufficiency. Hypertension. 2002;39:135–141. 31. Mann JFE, Yi QL, Gerstein HC. Albuminuria as a predictor of cardio- 50. Stehouwer CDA, Henry RMA, Dekker JM, Nijpels G, Heine RJ, Bouter vascular and renal outcomes in people with known atherosclerotic car- LM. Microalbuminuria is associated with impaired brachial artery, flow- diovascular disease. Kidney Int. 2004;66:S59–S62. mediated vasodilation in elderly individuals without and with diabetes: 32. Hillege HL, Fidler V, Diercks GFH, Van Gilst WH, De Zeeuw D, Van Further evidence for a link between microalbuminuria and endothelial Veldhuisen DJ, Gans ROB, Janssen WMT, Grobbee DE, De Jong PE. dysfunction: the Hoorn study. Kidney Int. 2004;66:S42–S44. Urinary albumin excretion predicts cardiovascular and noncardiovascu- 51. Endemann DH, Schiffrin EL. Endothelial dysfunction. JAmSoc lar mortality in general population. Circulation. 2002;106:1777–1782. Nephrol. 2004;15:1983–1992. 33. Gerstein HC, Mann JFE, Qilong Y, Zinman B, Dinneen SF, Hoogwerf 52. Passauer J, Pistrosch F, Bussemaker E, Lassig G, Herbrig K, Gross P. B, Halle JP, Young J, Rashkow A, Yoyce C, Nawaz S, Yusuf S. Reduced agonist-induced endothelium-dependent vasodilation in uremia 96 Circulation July 3, 2007

is attributable to an impairment of vascular nitric oxide. JAmSoc with incipient primary chronic renal disease. J Am Soc Nephrol. 2002; Nephrol. 2005;16:959–965. 13:170–176. 53. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. 74. Ravani P, Tripepi G, Malberti F, Testa S, Mallamaci F, Zoccali C. Impact of hypertension on cardiomyopathy, morbidity and mortality in Asymmetrical dimethylarginine predicts progression to dialysis and end-stage renal disease. Kidney Int. 1996;49:1379–1385. death in patients with chronic kidney disease: a competing risks 54. Locatelli F, Bommer J, London GM, Martin-Malo A, Wanner C, modeling approach. J Am Soc Nephrol. 2005;16:2449–2455. Yaqoob M, Zoccali C. Cardiovascular disease determinants in chronic 75. Fliser D, Kronenberg F, Kielstein JT, Morath C, Bode-Boger SM, Haller renal failure: clinical approach and treatment. Nephrol Dial Transplant. H, Ritz E. Asymmetric dimethylarginine and progression of chronic 2001;16:459–468. kidney disease: the Mild to Moderate Kidney Disease study. JAmSoc 55. Amann K, Breitbach M, Ritz E, Mall G. Myocyte/capillary mismatch in Nephrol. 2005;16:2456–2461. the heart of uremic patients. J Am Soc Nephrol. 1998;9:1018–1022. 76. Tarnow L, Hovind P, Teerlink T, Stehouwer CDA, Parving HH. 56. Amann K, Miltenberger-Miltenyi G, Simonoviciene A, Koch A, Orth S, Elevated plasma asymmetric dimethylarginine as a marker of cardiovas- Ritz E. Remodeling of resistance arteries in renal failure: effect of cular morbidity in early diabetic nephropathy in type 1 diabetes. endothelin receptor blockade. J Am Soc Nephrol. 2001;12:2040–2050. Diabetes Care. 2004;27:765–769. 57. Stehouwer CDA, Smulders YM. Microalbuminuria and risk for cardio- 77. Meinitzer A, Seelhorst U, Wellnitz B, Halwachs-Baumann G, Boehm vascular disease: analysis of potential mechanisms. J Am Soc Nephrol. BO, Winkelmann BR, Marz W. Asymmetrical dimethylarginine inde- 2006;17:2106–2111. pendently predicts total and cardiovascular mortality in individuals with 58. Fujihara CK, De Nucci G, Zatz R. Chronic nitric oxide synthase inhi- angiographic coronary artery disease (the Ludwigshafen Risk and Car- bition aggravates glomerular injury in rats with subtotal nephrectomy. diovascular Health study). Clin Chem. 2007;53:273–283. J Am Soc Nephrol. 1995;5:1498–1507. 78. Tatematsu S, Wakino S, Kanda T, Homma K, Yoshioka K, Hasegawa K, 59. Stam F, van Guldener C, Becker A, Dekker JM, Heine RJ, Bouter LM, Sugano N, Kimoto M, Saruta T, Hayashi K. Role of nitric oxide- Stehouwer CDA. Endothelial dysfunction contributes to renal function- producing and -degrading pathways in coronary endothelial dysfunction associated cardiovascular mortality in a population with mild renal in chronic kidney disease. J Am Soc Nephrol. 2007;18:741–749. insufficiency: the Hoorn study. J Am Soc Nephrol. 2006;17:537–545. 79. Stenvinkel P, Heimburger O, Lindholm B, Kaysen GA, Bergstrom J. 60. Hasdan G, Benchetrit S, Rashid G, Green J, Bernheim J, Rathaus M. Are there two types of malnutrition in chronic renal failure? Evidence Endothelial dysfunction and hypertension in 5/6 nephrectomized rats are for relationships between malnutrition, inflammation and atherosclerosis mediated by vascular superoxide. Kidney Int. 2002;61:586–590. (MIA syndrome). Nephrol Dial Transplant. 2000;15:953–960. 61. Weiner DE, Tighiouart H, Levey AS, Elsayed E, Griffith JL, Salem DN, 80. Kaysen GA, Eiserich JP. The role of oxidative stress-altered lipoprotein Sarnak MJ. Lowest systolic blood pressure was associated with stroke in structure and function and microinflammation on cardiovascular risk in stage 3–4 chronic kidney disease. J Am Soc Nephrol. 2007;18:960-966. patients with minor renal dysfunction. J Am Soc Nephrol. 2004;15: 538–548. 62. Vallance P, Leiper J. Cardiovascular biology of the asymmetric dim- 81. Trevisan R, Dodesini AR, Lepore G. Lipids and renal disease. JAmSoc ethylarginine:dimethylarginine dimethylaminohydrolase pathway. Arte- Nephrol. 2006;17(suppl 2):S145–S147. rioscler Thromb Vasc Biol. 2004;24:1023–1030. 82. Sandhu S, Wiebe N, Fried LF, Tonelli M. Statins for improving renal 63. Kielstein JT, Impraim B, Simmel S, Bode-Boger SM, Tsikas D, Frolich outcomes: a meta-analysis. J Am Soc Nephrol. 2006;17:2006–2016. JC, Hoeper MM, Haller H, Fliser D. Cardiovascular effects of systemic 83. Pecoits-Filho R, Heimburger O, Barany P, Suliman M, Fehrman- nitric oxide synthase inhibition with asymmetrical dimethylarginine in Ekholm I, Lindholm B, Stenvinkel P. Associations between circulating humans. Circulation. 2004;109:172–177. inflammatory markers and residual renal function in CRF patients. Am J 64. Zoccali C, Benedetto FA, Maas R, Mallamaci F, Tripepi G, Malatino Kidney Dis. 2003;41:1212–1218. LS, Boger R, CREED Investigators. Asymmetric dimethylarginine, 84. Jofre R, Rodriguez-Benitez P, Lopez-Gomez JM, Perez-Garcia R. C-reactive protein, and carotid intima-media thickness in end-stage renal Inflammatory syndrome in patients on hemodialysis. J Am Soc Nephrol. disease. J Am Soc Nephrol 2002;13:490–496. 2006;17(suppl 3):S274–S280. 65. Kielstein JT, Zoccali C. Asymmetric dimethylarginine: a cardiovascular 85. Vaziri ND, Oveisi F, Ding YX. Role of increased oxygen free radical risk factor and a uremic toxin coming of age? Am J Kidney Dis. activity in the pathogenesis of uremic hypertension. Kidney Int. 1998; 2005;46:186–202. 53:1748–1754. 66. Yilmaz MI, Saglam M, Caglar K, Cakir E, Sonmez A, Ozgurtas T, 86. Himmelfarb J, Stenvinkel P, Ikizler TA, Hakim RM. The elephant in Aydin A, Eyileten T, Ozcan O, Acikel C, Tasar M, Genctoy G, Erbil K, uremia: oxidant stress as a unifying concept of cardiovascular disease in Vural A, Zoccali C. Pathogenesis and treatment of kidney disease and uremia. Kidney Int. 2002;62:1524–1538. hypertension: the determinants of endothelial dysfunction in CKD: oxi- 87. Vaziri ND, Dicus M, Ho ND, Boroujerdi-Rad L, Sindhu RK. Oxidative dative stress and asymmetric dimethylarginine. Am J Kidney Dis. 2006; stress and dysregulation of superoxide dismutase and NADPH oxidase 47:42–50. in renal insufficiency. Kidney Int. 2003;63:179–185. 67. Vallance P, Leone A, Calver A, Collier J, Moncada S. Accumulation of 88. Touyz RM, Yao G, Quinn MT, Pagano PJ, Schiffrin EL. p47phox an endogenous inhibitor of nitric oxide synthesis in chronic renal failure. associates with the cytoskeleton through cortactin in human vascular The Lancet. 1992;339:572–575. smooth muscle cells: role in NAD(P)H oxidase regulation by angioten- 68. Zoccali C, Bode-Boger SM, Mallamaci F, Benedetto F, Tripepi G, sin II. Arterioscler Thromb Vasc Biol. 2005;25:512–518. Malatino L, Cataliotti A, Bellanuova I, Fermo I, Frolich J, Boger R. 89. Daugherty A, Dunn JL, Rateri DL, Heinecke JW. Myeloperoxidase, a Plasma concentration of asymmetrical dimethylarginine and mortality in catalyst for lipoprotein oxidation, is expressed in human atherosclerotic patients with end-stage renal disease: a prospective study. The Lancet. lesions. J Clin Invest. 1994;94:437–444. 2001;358:2113–2117. 90. Pecoits-Filho R, Stenvinkel P, Marchlewska A, Heimburger O, Barany 69. Matsuguma K, Ueda S, Yamagishi SI, Matsumoto Y, Kaneyuki U, P, Hoff CM, Holmes CJ, Suliman M, Lindholm B, Schalling M, Shibata R, Fujimura T, Matsuoka H, Kimoto M, Kato S, Imaizumi T, Nordfors L. A functional variant of the myeloperoxidase gene is asso- Okuda S. Molecular mechanism for elevation of asymmetric dimethyl- ciated with cardiovascular disease in end-stage renal disease patients. arginine and its role for hypertension in chronic kidney disease. JAm Kidney Int. 2003;63:S172–S176. Soc Nephrol. 2006;17:2176–2183. 91. Touyz RM, Schiffrin EL. Reactive oxygen species in vascular biology: 70. Mallamaci F, Tripepi G, Maas R, Malatino L, Boger R, Zoccali C. implications in hypertension. Histochem Cell Biol. 2004;122:339–352. Analysis of the relationship between norepinephrine and asymmetric 92. Schiffrin EL, Touyz RM. From bedside to bench to bedside: role of dimethyl arginine levels among patients with end-stage renal disease. renin-angiotensin-aldosterone system in remodeling of resistance J Am Soc Nephrol. 2004;15:435–441. arteries in hypertension. Am J Physiol Heart Circ Physiol. 2004;287: 71. Cooke JP. Asymmetrical dimethylarginine: the uber marker? Circu- H435–H446. lation. 2004;109:1813–1818. 93. Qunibi WY. Reducing the burden of cardiovascular calcification in 72. Kielstein JT, Frolich JC, Haller H, Fliser D. ADMA (asymmetric dim- patients with chronic kidney disease. J Am Soc Nephrol. 2005;16(suppl ethylarginine): an atherosclerotic disease mediating agent in patients 2):S95–S102. with renal disease? Nephrol Dial Transplant. 2001;16:1742–1745. 94. Ketteler M, Schlieper G, Floege J. Calcification and cardiovascular 73. Kielstein JT, Boger RH, Bode-Boger SM, Frolich JC, Haller H, Ritz E, health: new insights into an old phenomenon. Hypertension. 2006;47: Fliser D. Marked increase of asymmetric dimethylarginine in patients 1027–1034.

Schiffrin et al Kidney Disease and the Cardiovascular System 97

95. Johnson RC, Leopold JA, Loscalzo J. Vascular calcification: pathobio- 109. Stenvinkel P, Wang K, Qureshi AR, Axelsson J, Pecoits-Filho R, Gao P, logical mechanisms and clinical implications. Circ Res. 2006;99: Barany P, Lindholm B, Jogestrand T, Heimburger O, Holmes C, 1044–1059. Schalling M, Nordfors L. Low fetuin-A levels are associated with 96. Russo D, Palmiero G, De Blasio AP, Balletta MM, Andreucci VE. cardiovascular death: impact of variations in the gene encoding fetuin. Coronary artery calcification in patients with CRF not undergoing Kidney Int. 2005;67:2383–2392. dialysis. Am J Kidney Dis. 2004;44:1024–1030. 110. Zebboudj AF, Imura M, Bostrom K. Matrix GLA protein, a regulatory 97. Blacher J, Guerin AP, Pannier B, Marchais SJ, London GM. Arterial protein for bone morphogenetic protein-2. J Biol Chem. 2002;277: calcifications, arterial stiffness, and cardiovascular risk in end-stage 4388–4394. renal disease. Hypertension. 2001;38:938–942. 111. Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, 98. Budisavljevic MN, Cheek D, Ploth DW. Calciphylaxis in chronic renal Karsenty G. Spontaneous calcification of arteries and cartilage in mice failure. J Am Soc Nephrol. 1996;7:978–982. lacking matrix GLA protein. Nature. 1997;386:78–81. 112. Shanahan CM, Cary NR, Metcalfe JC, Weissberg PL. High expression 99. Rudwaleit M, Schwarz A, Trautmann C, Offermann G, Distler A. Severe of genes for calcification-regulating proteins in human atherosclerotic calciphylaxis in a renal patient on long-term oral anticoagulant therapy. plaques. J Clin Invest. 1994;93:2393–2402. Am J Nephrol. 1996;16:344–348. 113. Proudfoot D, Skepper JN, Shanahan CM, Weissberg PL. Calcification of 100. Hafner J, Keusch G, Wahl C, Sauter B, Hurlimann A, von Weizsacker human vascular cells in vitro is correlated with high levels of matrix Gla F, Krayenbuhl M, Biedermann K, Brunner U, Helfenstein U. Uremic protein and low levels of osteopontin expression. Arterioscler Thromb small-artery disease with medial calcification and intimal hyperplasia Vasc Biol. 1998;18:379–388. (so-called calciphylaxis): a complication of chronic renal failure and 114. Jono S, Ikari Y, Vermeer C, Dissel P, Hasegawa K, Shioi A, Taniwaki benefit from parathyroidectomy. J Am Acad Dermatol. 1995;33: H, Kizu A, Nishizawa Y, Saito S. Matrix Gla protein is associated with 954–962. coronary artery calcification as assessed by electron-beam computed 101. Budisavljevic MN, Cheek D, Ploth DW. Calciphylaxis in chronic renal tomography. Thromb Haemost. 2004;91:790–794. failure. J Am Soc Nephrol. 1996;7:978–982. 115. Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, Capparelli C, 102. Moe SM, Reslerova M, Ketteler M, O’Neill K, Duan D, Koczman J, Scully S, Tan HL, Xu W, Lacey DL, Boyle WJ, Simonet WS. Westenfeld R, Jahnen-Dechent W, Chen NX. Role of calcification Osteoprotegerin-deficient mice develop early onset osteoporosis and inhibitors in the pathogenesis of vascular calcification in chronic kidney arterial calcification. Genes Dev. 1998;12:1260–1268. disease (CKD). Kidney Int. 2005;67:2295–2304. 116. Nitta K, Akiba T, Uchida K, Kawashima A, Yumura W, Kabaya T, 103. Pannier B, Guèrin AP, Marchais SJ, Safar ME, London GM. Stiffness of Nihei H. The progression of vascular calcification and serum osteopro- capacitive and conduit arteries prognostic significance for end-stage tegerin levels in patients on long-term hemodialysis. Am J Kidney Dis. renal disease patients. Hypertension. 2005;45:592–596. 2003;42:303–309. 104. Tonelli M, Sacks F, Pfeffer M, Gao Z, Curhan G, for the Cholesterol 117. Morena M, Terrier N, Jaussent I, Leray-Moragues H, Chalabi L, Rivory And Recurrent Events (CARE) Trial Investigators. Relation between JP, Maurice F, Delcourt C, Cristol JP, Canaud B, Dupuy AM. Plasma serum phosphate level and cardiovascular event rate in people with osteoprotegerin is associated with mortality in hemodialysis patients. coronary disease. Circulation. 2005;112:2627–2633. J Am Soc Nephrol. 2006;17:262–270. 105. Reynolds JL, Joannides AJ, Skepper JN, McNair R, Schurgers LJ, 118. Johnson K, Polewski M, van Etten D, Terkeltaub R. Chondrogenesis Proudfoot D, Jahnen-Dechent W, Weissberg PL, Shanahan CM. Human mediated by PPi depletion promotes spontaneous aortic calcification in vascular smooth muscle cells undergo vesicle-mediated calcification in NPP1-/- mice. Arterioscler Thromb Vasc Biol. 2005;25:686–691. 119. Harmey D, Hessle L, Narisawa S, Johnson KA, Terkeltaub R, Millan JL. response to changes in extracellular calcium and phosphate concen- Concerted regulation of inorganic pyrophosphate and osteopontin by trations: a potential mechanism for accelerated vascular calcification in akp2, enpp1, and ank: an integrated model of the pathogenesis of ESRD. J Am Soc Nephrol. 2004;15:2857–2867. mineralization disorders. Am J Pathol. 2004;164:1199–1209. 106. Demetriou M, Binkert C, Sukhu B, Tenenbaum HC, Dennis JW. 120. Lomashvili KA, Khawandi W, O’Neill WC. Reduced plasma pyro- Fetuin/alpha 2-HS glycoprotein is a transforming growth factor-beta phosphate levels in hemodialysis patients. J Am Soc Nephrol. 2005;16: type ii receptor mimic and cytokine antagonist. J Biol Chem. 1996;271: 2495–2500. 12755–12761. 121. Hruska KA, Mathew S, Saab G. Bone morphogenetic proteins in 107. Price PA, Thomas GR, Pardini AW, Figueira WF, Caputo JM, Wil- vascular calcification. Circ Res. 2005;97:105–114. liamson MK. Discovery of a high molecular weight complex of calcium, 122. Chen D, Zhao M, Mundy GR. Bone morphogenetic proteins. Growth phosphate, fetuin, and matrix gamma-carboxyglutamic acid protein in Factors. 2004;22:233–241. the serum of etidronate-treated rats. J Biol Chem. 2002;277:3926–3934. 123. Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, 108. Ketteler M, Bongartz P, Westenfeld R, Wildberger JE, Mahnken AH, Armstrong D, Ducy P, Karsenty G. Leptin regulates bone formation via Bohm R, Metzger T, Wanner C, Jahnen-Dechent W, Floege J. Asso- the sympathetic nervous system. Cell. 2002;111:305–317. ciation of low fetuin-A (AHSG) concentrations in serum with cardio- 124. Parhami F, Tintut Y, Ballard A, Fogelman AM, Demer LL. Leptin vascular mortality in patients on dialysis: a cross-sectional study. The enhances the calcification of vascular cells: artery wall as a target of Lancet. 2003;361:827–833. leptin. Circ Res. 2001;88:954–960.

ACCF/AHA/SCAI Clinical Competence Statement

ACCF/AHA/SCAI 2007 Update of the Clinical Competence Statement on Cardiac Interventional Procedures A Report of the American College of Cardiology Foundation/American Heart Association/American College of Physicians Task Force on Clinical Competence and Training (Writing Committee to Update the 1998 Clinical Competence Statement on Recommendations for the Assessment and Maintenance of Proficiency in Coronary Interventional Procedures) WRITING COMMITTEE MEMBERS

Spencer B. King III, MD, MACC, FAHA, FSCAI, Chair; Thomas Aversano, MD, FACC; William L. Ballard, MD, FACC, FSCAI; Robert H. Beekman III, MD, FACC, FAHA; Michael J. Cowley, MD, FACC, FSCAI*; Stephen G. Ellis, MD, FACC; David P. Faxon, MD, FACC, FAHA, FSCAI*; Edward L. Hannan, PhD, FACC; John W. Hirshfeld, Jr, MD, FACC, FAHA; Alice K. Jacobs, MD, FACC, FAHA, FSCAI; Mirle A. Kellett, Jr, MD, FACC, FSCAI; Stephen E. Kimmel, MD, FACC, FAHA; Joel S. Landzberg, MD, FACC; Louis S. McKeever, MD, FACC, FSCAI; Mauro Moscucci, MD, FACC; Richard M. Pomerantz, MD, FACC, FSCAI; Karen M. Smith, MD, FACC, FSCAI; George W. Vetrovec, MD, FACC, FSCAI*

TASK FORCE MEMBERS

Mark A. Creager, MD, FACC, FAHA, Chair; John W. Hirshfeld, Jr, MD, FACC, FAHA†; David R. Holmes, Jr, MD, FACC; L. Kristin Newby, MD, FACC, FAHA; Howard H. Weitz, MD, FACC, FACP; Geno Merli, MD, FACP; Ileana Piña, MD, FACC, FAHA; George P. Rodgers, MD, FACC, FAHA; Cynthia M. Tracy, MD, FACC†

*Society for Cardiovascular Angiography and Interventions Representative. †Former Task Force member during the writing effort. This document was approved by the American College of Cardiology Board of Trustees in May 2007, the American Heart Association Science Advisory and Coordinating Committee in May 2007, and the Society for Cardiovascular Angiography and Interventions in May 2007. When this document is cited, the American College of Cardiology, American Heart Association, and Society for Cardiovascular Angiography and Interventions would appreciate the following citation format: King SB III, Aversano T, Ballard WL, Beekman RH III, Cowley MJ, Ellis SG, Faxon DP, Hannan EL, Hirshfeld JW Jr., Jacobs AK, Kellett MA Jr., Kimmel SE, Landzberg JS, McKeever LS, Moscucci M, Pomerantz RM, Smith KM, Vetrovec GW. ACCF/AHA/SCAI 2007 update of the clinical competence statement on cardiac interventional procedures: a report of the American College of Cardiology Foundation/American Heart Association/American College of Physicians Task Force on Clinical Competence and Training (Writing Committee to Update the 1998 Clinical Competence Statement on Recommendations for the Assessment and Maintenance of Proficiency in Coronary Interventional Procedures). Circulation. 2007;116:98–124. This article has been copublished in the July 3, 2007, issue of the Journal of the American College of Cardiology. Copies: This document is available on the World Wide Web sites of the American College of Cardiology (www.acc.org), the American Heart Association (www.americanheart.org), and the Society for Cardiovascular Angiography and Interventions (www.scai.org). For copies of this document, please contact Elsevier Inc. Reprint Department, fax (212) 633-3820, e-mail [email protected]. To purchase Circulation reprints, call 843-216-2533 or e-mail [email protected]. Permissions: Multiple copies, modification, alteration, enhancement, and/or distribution of this document are not permitted without the express permission of the American College of Cardiology or the American Heart Association. Instructions for obtaining permission are located at http:// www.americanheart.org/presenter.jhtml?identifierϭ4431. A link to the “Permission Request Form” appears on the right side of the page. (Circulation. 2007;116:98-124.) © 2007 by the American College of Cardiology Foundation and the American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.107.185159 98 King et al ACCF/AHA/SCAI Clinical Competence Statement 99

TABLE OF CONTENTS

Preamble...... 99 The Quality Assessment Process...... 113 Introduction ...... 100 Conclusions and Recommendations Purpose ...... 101 for PCIs ...... 113 Writing Group Composition ...... 101 Success and Complication Rates...... 113 Literature Review...... 101 Risk Adjustment...... 114 Percutaneous Coronary Intervention ...... 101 Volume–Activity Relationships...... 114 Evolution of Competence and Training Recommendations for Institutional Standards ...... 101 Maintenance of Quality ...... 114 Evolution of Coronary Interventional Recommendations for Individual Capabilities...... 102 Maintenance of Quality ...... 114 Procedural Success and Complications of Percutaneous Noncoronary Interventions ...... 114 Coronary Interventional Procedures ...... 102 Introduction ...... 114 Patient, Lesion, and Institutional Variables Disorders of the Atrial Septum ...... 115 Influencing Success and Complication Criteria for Competency ...... 115 Rates ...... 103 Cardiologists in Training Programs ...... 115 Measures/Definitions of Success...... 103 Cardiologists in Practice ...... 115 Patient and Lesion Characteristics Related to Procedural Success and Complication Maintenance of Competency for Rates ...... 104 Percutaneous ASD/PFO Closure ...... 115 Strategies for Risk Stratification and Quality Assurance ...... 115 Operator Evaluation ...... 104 Hypertrophic Cardiomyopathy and Alcohol Septal Ablation ...... 116 Impact of the Facility on Procedural Success ...... 104 Criteria for Competency ...... 116 Components of Operator Competence ...... 105 Valvular Heart Disease ...... 116 Cognitive Knowledge Base ...... 105 Cognitive Knowledge Base ...... 116 Technical Skills ...... 106 Criteria for Competency ...... 116 Nonballoon Devices...... 106 Percutaneous Ventricular Assist Devices ...... 116 Relationships of Operator and Institutional Experience and Activity to Outcomes in Laboratory and Staff Competence...... 116 Coronary Interventional Procedures ...... 106 Conclusions and Recommendations ...... 117 Evidence Reviewed ...... 106 Percutaneous Noncoronary Relationship of Institutional Volume to Interventions...... 117 Procedural Outcome...... 107 References ...... 117 Volume and Outcomes Relationship for Appendix 1...... 121 Primary PCI in Acute MI ...... 108 Appendix 2...... 122 Relationship of Individual Operator Volume to Procedural Outcome ...... 109 Preamble Combination of Individual Operator The granting of clinical staff privileges to physicians is a Volume and Institutional Volume on primary mechanism used by institutions to uphold the quality Procedural Outcome...... 110 of care. The Joint Commission on Accreditation of Health Ongoing Quality Improvement and Care Organizations requires that the granting of continuing Maintenance of Competence ...... 111 medical staff privileges be based on the criteria specified in Institutional Maintenance of Quality...... 111 the medical staff bylaws. Physicians themselves are thus Individual Maintenance of Quality ...... 112 charged with identifying the criteria that constitute profes- Quality Assurance ...... 112 sional competence and with evaluating their peers accord- Definition of Quality in PCI ...... 112 ingly. Yet, the process of evaluating physicians’ knowledge Institutional Quality Assurance and competence is often constrained by the evaluator’s own Requirement ...... 112 knowledge and ability to elicit the appropriate information, Role of Risk Adjustment in Assessing problems compounded by the growing number of highly Quality ...... 112 specialized procedures for which privileges are requested. Challenges in Determining Quality...... 112 The American College of Cardiology Foundation/Amer- Requirement for Institutional Resources ican Heart Association/American College of Physicians and Support ...... 113 (ACCF/AHA/ACP) Task Force on Clinical Compe- 100 Circulation July 3, 2007

tence and Training was formed in 1998 to develop The AHA estimated that more than 1,000,000 PCIs recommendations for attaining and maintaining the cogni- were performed in the United States in 2003 (4). Physicians tive and technical skills necessary for the competent perfor- performing these procedures represent approximately 25% mance of a specific cardiovascular service, procedure, or of board-certified cardiologists in the United States (5). technology. These documents are evidence based and, As a result of the maturation of PCI as a discipline and where evidence is not available, expert opinion is utilized to the ongoing clarification of its role in the management of formulate recommendations. Indications and contraindica- coronary heart disease, the public can and should appropri- tions for specific services or procedures are not included in ately expect consistent access to high-quality PCI capability. the scope of these documents. Recommendations are in- However, there is potential for substantial variation in the tended to assist those who must judge the competence of quality of PCI services. PCI is often a complex, demanding cardiovascular health care providers entering practice for the procedure. To perform PCI optimally, an operator must first time and/or those who are in practice and undergo possess a substantial cognitive knowledge base as well as periodic review of their practice expertise or who apply for considerable technical skill. In addition, the technical diffi- privileges at a new institution. The assessment of compe- culty of a particular procedure can vary greatly from one tence is complex and multidimensional, therefore, isolated patient to another. Furthermore, serious complications of recommendations contained herein may not necessarily be coronary interventional procedures may occur unpredictably sufficient or appropriate for judging overall competence. in procedures that initially appear to be straightforward. The current document addresses competence in cardiac Recognition and management of complications are critical interventional procedures and is authored by representatives components of PCI procedures that require skill, knowl- of the ACCF, the AHA, and the Society for Cardiovascular edge, experience, and judgment. Since there can be variation Angiography and Interventions (SCAI). This document among operators in cognitive knowledge and skill and applies to specialists trained in internal medicine and/or among procedures in technical difficulty, there is a potential adult cardiology and is not meant to be a clinical compe- for substantial variation in procedure safety and efficacy. tence statement on procedures for congenital heart disease Credentialing physicians to perform procedures is the in the child or young adult. responsibility of the governance of the local health care The ACCF/AHA/ACP Task Force makes every effort to facility. The Joint Commission on the Accreditation of avoid any actual or potential conflicts of interest that might Health Care Organizations requires that medical staff priv- arise as a result of an outside relationship or personal interest ileges be granted to applicants only after assessment based of a member of the ACCF/AHA/ACP Writing Commit- on professional criteria. Physicians are charged with the tee. Specifically, all members of the Writing Committee responsibility to establish the criteria that constitute profes- were asked to provide disclosure statements of all such sional competence and to evaluate their peers on the basis of relationships that might be perceived as real or potential such criteria. The U.S. health care system relies, in part, on conflicts of interest relevant to the document topic. These this process of granting and renewing clinical privileges to statements were reviewed by the Writing Committee and maintain quality. updated as changes occurred. The relationships with indus- The issue of determining quality standards and creden- try for authors and peer reviewers are published in the tialing criteria has presented a major challenge to the appendices of the document. medical profession. Developing standards has been difficult Mark A. Creager, MD, FACC, FAHA because, until recently, there were few data available on Chair, ACCF/AHA/ACP Task Force on which to base them and because PCI techniques, indica- Clinical Competence and Training tions, and capability have evolved rapidly. During the past several years, documents have been published that have offered guidelines and standards for the training and main- Introduction tenance of competence (6–15). Because of the paucity of clinical data, the earlier standards were developed principally Coronary intervention has evolved from an investigational through observation, experience, and intuition. These stan- procedure to a widely practiced, mature mainstream clinical dards relied heavily on operator activity level as a surrogate therapy (1). Conventional balloon angioplasty, while still a for skill and quality. core procedure in interventional cardiology, has been aug- The most recent document published by the ACC was mented by adjunctive stenting, which greatly improves based on the information available in 1998 (16). The procedure efficacy and modestly reduces the risk of resteno- recommendations of this and other similar documents sis (2). Bare-metal stents have been replaced by drug-eluting require updating as technology and training evolve (17). stents in the majority of cases, which further reduce the risk Percutaneous noncoronary cardiac interventions, such as of restenosis (3). Because stents or other interventional aortic and mitral valvuloplasty, atrial septal defect (ASD) devices are commonly used, the coronary angioplasty pro- and patent foramen ovale (PFO) closure, and alcohol septal cedure is more aptly termed “percutaneous coronary inter- ablation therapy, were not addressed in the previous docu- vention” (PCI). ment (16). These procedures, although constituting a King et al ACCF/AHA/SCAI Clinical Competence Statement 101

small minority of interventional activity, are performed Literature Review by interventional cardiologists and are included in the A literature search was conducted with 5 goals: Accreditation Council on Graduate Medical Education (ACGME) curriculum and the American Board of In- 1. To identify published coronary and other cardiac inter- ternal Medicine (ABIM) certifying exam. There have ventional outcomes data that could be used as bench- been no statements addressing clinical competence in marks for quality assessment. In addition, the process noncoronary interventions. sought to identify those risk adjustment variables that The ACC, the ACP, the SCAI, the Society for Vascular affect the likelihood of success and complications. This Medicine and Biology (SVMB), and the Society for Vas- review focused on outcomes of coronary interventions, cular Surgery (SVS) have jointly developed a document on including the latest interventional devices as of the date acquisition and maintenance of competence in vascular of this revision. medicine and catheter-based vascular interventions (18); 2. To identify data that examines the relationships between however, PCI and other percutaneous cardiac procedures operator and institutional experience, and activity levels, are not addressed by the current document. This document and their impact on procedural success and complication is divided into 2 sections: PCI and percutaneous noncoro- rates. nary cardiac interventions. 3. To assess the issues and problems associated with judging operator and institutional proficiency based on out- Purpose come statistics—in particular, the challenge of accurately This document was developed to review the currently assessing the performance of low-volume operators and available scientific data with the following purposes: institutions. 4. To expand the recommendations beyond coronary inter- 1. To characterize the expected success and complication ventions to other cardiac interventional procedures. rates for coronary interventional procedures when per- 5. To identify methods for monitoring appropriateness of formed by highly skilled operators. performance of PCI. 2. To identify comorbidities and other risk factors that may be used for risk adjustment when assessing procedure- Percutaneous Coronary Intervention specific expected success and complication rates. 3. To assess the relationship between operator activity level and success rates in PCI procedures as assessed by Evolution of Competence and Training Standards risk-adjusted outcome statistics. Initially, because experience was limited, the coronary an- 4. To assess the relationship between institutional activity gioplasty technique was disseminated informally among level and success rates in PCI procedures as assessed by physicians who were highly experienced at diagnostic car- risk-adjusted outcome statistics. diac catheterization. During this period, physicians acquired 5. To develop recommendations for standards to assess angioplasty skills through “on-the-job” experience, and no operator proficiency and institutional program quality. standards existed for either training requirements or for These include standards for data collection to permit demonstration of competence. monitoring of appropriateness and effectiveness of PCI As the coronary angioplasty knowledge base grew and procedures both at the level of the operator and the techniques evolved, standards were developed for training institution. (19). Formal angioplasty training programs were first orga- 6. To expand the scope of this competency document, nized in the early 1980s. The most recent recommendations previously limited to coronary procedures, to also include were published by the ACC in 1999 (20). The ABIM noncoronary cardiac interventions. developed an Examination in Interventional Cardiology that was first administered in 1999. As of 2005, 5,020 Writing Group Composition physicians had successfully passed the examination and The Writing Group was selected to represent a broad range become board certified in interventional cardiology. Cur- of experience and expertise to bear on this issue. The rently, eligibility to sit for the ABIM interventional cardi- members of the Writing Group were identified on the basis ology examination requires completion of a fourth-year of 1 or more of the following attributes: PCI operators with fellowship in interventional cardiology in an ACGME- a broad range of experience (in practice and in academic accredited program. During academic year 2004 to 2005, settings); individuals who have performed clinical research there were 122 accredited interventional cardiology pro- studying the outcome of PCI procedures; individuals who grams in the United States that had 240 filled training direct catheterization laboratories with a broad cross section positions. of interventional operators; and individuals with broad Professional organizations have addressed the issue of clinical experience who have had considerable previous standards and criteria for proficiency in PCI procedures involvement with PCI. since 1986, with an increasing focus on the issue of 102 Circulation July 3, 2007

maintenance of proficiency and skills (6–15). These docu- Procedural Success and Complications ments have universally endorsed an annual caseload goal for of Coronary Interventional Procedures maintenance of proficiency. The most commonly endorsed Recent clinical studies have demonstrated that despite a activity level has been 75 procedures per year per operator. continuing increase in clinical and angiographic complexity, This standard was initially based on general consensus of procedural and clinical success rates have remained high and experts. In recent years, considerable research has examined complication rates have remained low (37–45)(Table 1). the volume–outcome relationship and, in general, has af- Angiographic success occurs in over 95% of patients. firmed it (21,22). Among patients without ST-segment elevation myocardial Since the previous guidelines were published, there has infarction (STEMI), PCI is associated with an average been debate over the relationship between volume and mortality rate of less than 1%, a Q-wave MI rate of less than quality. While a relationship between volume and outcomes 1%, and an emergency coronary artery bypass surgery exists, volume alone does not determine quality. Also, the (CABG) rate of less than 1%. Table 1 contains data from 5 ABIM interventional cardiology board exam has been large contemporary registries of PCI procedures and the established to certify a level of knowledge and experience in first 2 National Heart, Lung, and Blood Institute (NHLBI) the field. This competency document addresses these factors registries for historical comparison. These data constitute a as they relate to determinations of overall operator and point of departure for developing benchmarking standards. institutional quality. Adverse events related to PCI procedures are categorized either by the mechanism of the complication or by the Evolution of Coronary Interventional Capabilities adverse event caused by the procedure. A given adverse event, such as death, may be caused by a variety of The cognitive and technical knowledge base required for complications. proficiency in PCI has expanded. The fundamental con- Complications can be divided into 3 mechanistic categories: cepts of coronary angioplasty technique, namely the coaxial guide catheter and the dilation catheter with a minimally 1. Coronary vascular injury. Coronary arterial injury can compliant cylindrical balloon, were formulated by Andreas occur when devices are introduced into coronary vessels Gruntzig (23). Because of the initial comparatively primitive or result from embolization of thrombotic or atheroscle- equipment design and capability, coronary angioplasty was rotic material from devices or vessel walls. Examples only applicable to readily accessible discrete proximal coro- include coronary dissection, thrombosis, perforation, and nary stenoses. Subsequent refinement in instrumentation embolization. has greatly enhanced procedural success and extended the 2. Other vascular events. Other vascular events are caused indications for the performance of PCI. Complex anatomic either by injury to a peripheral vessel by catheter inser- situations now considered technically suitable for PCI tion, manipulation, or removal, or by embolization of procedures include multivessel disease (24–30), distal and thrombotic or atherosclerotic material. Examples include bifurcation stenoses, total occlusions (31), saphenous vein pseudoaneurysm, retroperitoneal hemorrhage, arterio- graft stenoses (32), and complex stenoses. Challenging venous fistula, and stroke. clinical situations now considered appropriate for coronary 3. Systemic nonvascular events. Systemic nonvascular ad- intervention include patients with unstable angina (33,34) verse events are caused by the procedure but are not due and myocardial infarction (MI) (35,36) and those who are to vascular injury. They include all the systemic hazards not considered candidates for coronary bypass surgery. of cardiovascular radiographic angiography procedures. Nonballoon devices, including coronary stents, and direc- Examples include contrast agent-induced nephropathy tional, rotational, and laser atherectomy devices, have been and acute pulmonary vascular congestion. introduced. These devices augment conventional balloon For the purpose of assessing clinical competence, com- angioplasty and extend its capability; however, they all plications may be divided into 8 basic outcome categories: require specific training and mentoring by a previously experienced operator. To become competent in the use of 1. Death: related to the procedure, regardless of mechanism any of these newer interventional devices, an operator must 2. Stroke acquire the additional knowledge and technical skills spe- 3. MI: related to the procedure, regardless of mechanism cific to each device. 4. Ischemia requiring emergency CABG: either as a result A number of adjunctive antithrombotic and antiplatelet of procedure failure or a procedure complication medications have been introduced for the purpose of reduc- 5. Vascular access site complications ing acute thrombus-related treatment site complications. 6. Contrast agent nephropathy 7. Excessive bleeding, requiring treatment Understanding the appropriate indications for and compli- 8. Other (such as coronary perforation and tamponade) cations associated with the use of these medications, which are powerful anticoagulants, requires knowledge of hemo- The first 4 of these categories are generally considered static mechanisms. major adverse cardiac and cerebral events (MACCE). Be- King et al ACCF/AHA/SCAI Clinical Competence Statement 103

cause adverse events are deﬁnite end points, they are easily recognized and captured for statistical summary purposes. The ACC-National Cardiovascular Data Registry

) (NCDR)® has developed a comprehensive data dictionary 45 ( with rigorous deﬁnitions of recognized adverse events (46). 2001–2003 It may be impossible to determine conclusively whether Heart Association Task Force death or a complication was caused by a procedure. None- New York State Registry

Nonemergent Emergent theless, for the purposes of monitoring performance, rate of complications or deaths substantially above that expected, after adjustment for patient risk factors, is a cause for concern. ) 44 ( Cross Patient, Lesion, and Institutional Variables Consortium Michigan Blue Inﬂuencing Success and Complication Rates A number of factors have improved the overall success and complication rates of PCI procedures. These include increased operator experience, modiﬁcations in conventional ) National Heart, Lung, and Blood Institute.

43 instrumentation (balloon catheters, guide catheters, guide ϭ ( 12 18.9 0 53 94 NA 97.5 97.5 England ). wires), newer interventional devices (stents and emboliza- Consortium 15 Northern New tion protection devices), and advances in adjunctive phar- macologic therapy. Concurrently, these improvements have led to the extension of interventional treatment to higher- not available; NHLBI 06;47:216–35 ( vention—summary article: a report of the American College of Cardiology/American

ϭ risk patients with more complex coronary anatomy and comorbid disease. These factors have inﬂuenced overall

) acute and long-term outcome associated with PCI procedures. 42 (

Registry Measures/Deﬁnitions of Success 1,082,690 36,831 5,901 124,096 14,946 ACC National

Cardiovascular Data Anatomic success. The deﬁnition of anatomic success myocardial infarction; NA nonstented lesions: 86

ϭ focuses exclusively on the enlargement of the lumen at the

Clinical Characteristics target site and blood ﬂow through the epicardial coronary artery. Although there has been disagreement, the current Success and Complication Indicators deﬁnition of success of PCI with stenting is the achieve- )

41 ment of a minimal diameter stenosis of less than 20% as ( 6,183 NHLBI Registry Dynamic visually assessed by angiography and maintenance of multicenter database; MI

ϭ Thrombolysis In Myocardial Infarction (TIMI) ﬂow grade 3(15). Anatomic success of PCI without stenting is deﬁned as stenosis diameter reduction greater than 20% with resid- ) ual stenosis less than 50%. Notably, there is frequently a 1,802 38,39 disparity between the visual estimate of lumen diameter and ( NHLBI-2 quantitative measurements (47,48).

Procedural success. Procedural success has been defined as the achievement of anatomic success of all treated lesions ) coronary artery bypass graft surgery; MCD 40 ϭ without the major complications of death, MI, or emer- ( NHLBI-1 gency CABG (14,40). Although emergency CABG during hospitalization and death are easily identified end points, the definition of periprocedural MI has been more problematic. Some definitions require the development of Q waves in addition to a threshold value for creatine kinase (CK) elevation. However, more recent reports have included non-STEMIs with CK elevations greater than 3 or 5 times Variable

American College of Cardiology; CABG the upper limit of normal as clinically signiﬁcant, since they ϭ have been shown to correlate with long-term mortality (49). ACC Years of entryNo. of patientsStent use (%)Mean patient age (yrs)Unstable angina (%)ST-segment elevation MI (%) 1977–1981 1,155 54 0 1985–1986 37 0 1997–2002 58 0 49 0 1998–2005 63 25 44 78 2000–2004 61 2002 13 35 91.6 62 86 43 63 84.0 32.0 65 87.5 27.8 60 92.7 83.6 Angiographic success (%)Emergency CABG (%)Mortality (%) 68 5.80 91 1.2 3.40 93 1.0 1.00 2005: stented lesions: 99; 1.33 0.4 1.2 unadjusted rate 1.17 0.4 1.27 0.51 0.36 0.20 3.25 0.54 on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention). J Am Coll Cardiol 20 Table 1. Changes in Coronary Interventional Practice and Outcome From Registry Data Reprinted with permission from Smith SC Jr., Feldman TE, Hirshfeld JW Jr., et al. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary inter Although major adverse cardiac events (MACE) have been 104 Circulation July 3, 2007

used to judge success, some recent studies also include have higher success rates in more complex lesions (53–56). MACCE. As a result, lesion morphology may be less predictive of complications currently than it has been in the past (57). Short-term clinical success. Short-term clinical success requires, in addition to procedural success, the relief of signs Strategies for Risk Stratification and symptoms of myocardial ischemia. and Operator Evaluation Several large retrospective studies of patients undergoing Longer-term clinical success. Longer-term clinical success PCI have identified clinical and angiographic characteristics requires that the initial clinical success remains durable and that correlate with procedural success, in-hospital morbid- that the patient has persistent relief of signs and symptoms ity, and mortality (21,22,44)(Table 2). These observations of myocardial ischemia for 6 to 9 months after the proce- have been used to develop multivariate logistic regression dure. Restenosis remains the principal cause of a lack of models that can stratify patients before the procedure. clinical success over the first year following a successful Model reliability is best assessed by relative predictive procedure. This directly leads to target lesion revasculariza- accuracy (C-statistic: moderate is greater than 0.80, excel- tion (TLR), target vessel revascularization, and target vessel lent is greater than 0.90) and scaling accuracy (the Hosmer- failure. Thereafter, clinical events are usually caused by Lemeshow statistic). Several models predict periprocedural progression of disease at other sites. Clinically important mortality with C-statistic greater than 0.80 (Table 2). restenosis may be judged by the frequency with which Prediction of other events is typically less accurate (58–60). subsequent TLR procedures are performed after the index Model utility also must consider the frequency and clinical procedure. Incomplete revascularization, new lesion forma- importance of the event measured. Very infrequently occur- tion, and stent thrombosis may also limit long-term clinical ring events, even if severe, may not allow adequate evalua- success, especially in subsequent years (50). tion of operators with low volume. Results of several years of experience should be considered in order to have sufficient Patient and Lesion Characteristics Related to numbers of events to be adequately assessed from a statis- Procedural Success and Complication Rates tical standpoint. Operators and catheterization laboratories Angioplasty procedural success and complication rates are should be encouraged to submit information to large data- influenced by a variety of patient and target lesion charac- bases that allow for evaluation of risk-adjusted outcomes. teristics. These characteristics must be taken into consider- ation through risk adjustment when assessing adverse event Impact of the Facility on Procedural Success rates. In addition, they must also be weighed in determining Physical facility requirements. The physical facility in procedure appropriateness. which interventional procedures are performed has an im- Patient clinical characteristics. The clinical factors associ- portant impact on procedural success. The facility must ated with an increased risk of an adverse outcome after provide radiologic equipment, monitoring, and patient sup- intervention include advanced age, female gender, acute port equipment to enable operators to perform at the best of coronary syndrome (especially STEMI), chronic renal in- their ability. The video and “cine” image quality of radio- sufficiency, heart failure, and multivessel coronary disease logic imaging equipment must be optimal to facilitate (7,12,14,15). Patients with impaired renal function, partic- accurate catheter and device placement and enable proper ularly patients with diabetes, are at increased risk for assessment of procedure results. Physiologic monitoring contrast-induced nephropathy (51). equipment must provide continuous, accurate information about the patient’s condition. Requisite support equipment Target lesion anatomic factors. Particular lesion morpho- must be available and in good operating order to respond to logic characteristics are predictive of immediate outcome emergency situations. with coronary intervention (7,12,14,52). Lesion length, presence of thrombus, and degenerated saphenous vein Overall institutional system requirements. The interven- grafts are independently associated with abrupt vessel clo- tional laboratory must have an extensive support system of sure and major ischemic complications. Chronic total oc- specifically trained laboratory personnel. Cardiothoracic clusions (greater than or equal to 3 months) are associated surgical, respiratory, and anesthesia services should be avail- with a lower procedural success rate. On the basis of these able to respond to emergency situations in order to mini- observations, a previous ACC/AHA Clinical Task Force on mize detrimental outcomes. The institution should have Clinical Privileges in Cardiology (13) proposed a classifica- systems for credentialing, governance, data gathering, and tion scheme based on lesion morphology to estimate the quality assessment. Prospective, unbiased collection of key likelihood of procedural success and complications. This data elements on consecutive patients and consistent feed- scheme was subsequently modified by others (52) and has back of results to providers brings important quality control served as a useful guide for assessing the risk of an adverse to the entire interventional program. The ACC/AHA/ outcome associated with a particular lesion. More recent SCAI 2005 Guideline Update for PCI (15) recommends experience indicates that improved devices and techniques that each interventional program performing elective PCI

King et al ACCF/AHA/SCAI Clinical Competence Statement 105

Table 2. Odds Ratios* for Signiﬁcant Independent Risk Factors† for Short-Term Mortality Related to PCI

New York Northern Michigan Source State New England BMC2 ACC-NCDR ACC-NCDR Update COAP No. of patients 50,046 15,331 10,796 100,253 No acute MI Acute MI 19,358 (142,817) (30,926) Incidence (%) 0.58 1.1 1.6 1.4 N/A N/A 1.6 Years 2003 1994–1996 1997–1999 1998–2000 1998–2001 1998–2001 1999–2000 Clinical Acute MI less than 12–24 h 8.6 5.5 2.8 1.3 ϩ Age ϩϩ ϩϩ ϩ ϩϩ Cardiac arrest 3.7 CHF 3.2 8.6 1.6 COPD 1.3 1.7 1.5 1.8 Diabetes 1.4 1.25 Female 1.5 1.8 1.4 IABP pre 26.2 1.7 1.9 Peripheral vascular disease 2.6 3.3 1.6 1.6 Prior CABG 1.4 Priority (salvage, emergent urgent, elective) ϩϩϩϩϩ Renal insufﬁciency 3.1 6.4 5.5 3.0 3.5 2.0 3.5 Shock 22.1 32.2 11.5 8.5 9.8 8.8 9.8 Anatomic ACC lesion score, C 2.9 Ejection fraction ϩϩ ϩ ϩ ϩϩ LMT lesion 2.0 1.5 2.1 Number of diseased vessels ϩϩ Prox LAD lesion 2.0 1.3 1.3 ϩ SCAI lesion score ϩϩϩ Thrombus ϩ Procedural Lytic use 1.4 1.25 Nonstent use 1.6 1.6 1.4 C-statistic 0.905 0.88 0.90 0.89 0.85 0.87 0.87

*Values are odds ratios for binary variables unless otherwise noted; †specific definitions of risk factors may vary from series to series; ϩrelationship exists for continuous or ordinal variables (61–66). ACC-NCDR ϭ American College of Cardiology National Cardiovascular Data Registry; BMC2 ϭ Blue Cross Blue Shield of Michigan Cardiovascular Consortium; CABG ϭ coronary artery bypass graft; CHF ϭ congestive heart failure; COAP ϭ clinical outcome assessment program; COPD ϭ chronic obstructive pulmonary disease; IABP ϭ intra-aortic balloon pump; LAD ϭ left anterior descending; LMT ϭ left main trunk; MI ϭ myocardial infarction; PCI ϭ percutaneous coronary intervention; SCAI ϭ Society for Cardiovascular Angiography and Interventions. should have in-house surgical support. Institutions that do program are eligible to take the certifying examination. not have in-house surgical support and are performing Individuals who train in interventional cardiology should primary PCI only for STEMI, should have an established, become ABIM certified in interventional cardiology. well-organized system for emergency transfer to surgery at Training programs and the qualifying examination another institution. (20,69) require that interventional cardiologists be knowl- Components of Operator Competence edgeable in anatomy, physiology, and pathophysiology of the cardiovascular system. In particular, one should under- Cognitive Knowledge Base stand the biology of coronary artery disease, be knowledge- The knowledge needed to perform PCI, including that able about the pathophysiology of myocardial ischemia and expected to be acquired in ACGME-approved interven- MI, and understand the dynamics of cardiac dysfunction. tional training programs, has been addressed by expert Interventionalists should possess a fundamental knowledge panels (7,8,20,67,68). The core knowledge is now tested by of stents and be familiar with the polymers and drugs that the ABIM Interventional Cardiology certifying examina- are incorporated into stents, coagulation cascade, thrombo- tion which has been administered since 1999. Through sis, and the pharmacology, therapeutic application, and risks 2003, physicians trained by a nontraditional pathway were of antiplatelet, antithrombin, and fibrinolytic drugs that are eligible to take the examination based on either practice- used in association with PCI. Competent operators must based procedure activity and experience or by completion of have knowledge of the indications for PCI and adjunctive an interventional training program. Since 2003, only indi- and alternative use of medical therapy and surgery for viduals who have completed an ABIM-qualified training patients with coronary artery disease based on an in-depth 106 Circulation July 3, 2007

understanding of published clinical trials. Coronary inter- during training and by performing actual procedures under ventionalists must understand the role of primary angio- the direction of an experienced interventionalist. These plasty compared with fibrinolytic therapy for STEMI and include the manipulation and operation of guide catheters, the alternative therapeutic approaches for treating STEMI coronary angioplasty guide wires, coronary angioplasty bal- that depend upon the time of presentation, anticipated loon catheters, specialized atherectomy devices, stents, and door-to-balloon time, and the presence or absence of intracoronary ultrasound catheters. Such training appro- ongoing symptoms and/or electrocardiographic abnormalities. priately occurs in standardized training programs that are Cognitive knowledge must be bolstered by clinical skills ACGME-approved and lead to eligibility for board and experience that support the rational selection of optimal certification. treatment strategies for each patient. Such decisions are Nonballoon Devices based on symptoms, anatomy, and associated risk factors. Thus, equally important to knowing the indications for PCI A special area of competence involves use of lesion assess- is an understanding of its limitations and contraindications, ment tools. Intracoronary devices commonly used by inter- particularly as these relate to comorbid systemic diseases and ventional cardiologists for assessment of intraluminal coro- special anatomical subsets. Physicians performing these nary anatomy and/or physiology include intravascular procedures should be conversant with the applicable guide- ultrasound (IVUS) or intracoronary ultrasound (ICUS), lines (e.g., PCI, CABG, STEMI, unstable angina/ Doppler flow wires, and pressure wires. Competency in the NSTEMI [15,70–72]). use of angioscopy, optical coherence tomography, spectros- Coronary interventionalists must also have a thorough copy, intravascular thermography, and intravascular mag- knowledge of specialized equipment, techniques, and de- netic resonance imaging is beyond the scope of this docu- vices used to perform PCI competently, including: ment. Expertise in device manipulation and image interpretation is required to use these intravascular assess- 1. The theoretical and practical aspects of X-ray imaging, ment devices safely and effectively. The risks of these devices radiation physics and safety, and other equipment to is the same as those with PCI and include vessel spasm; generate digital images; quality control of images; image myocardial ischemia; coronary artery dissection; plaque archiving; consequences of exposure of patients and disruption; thrombosis; air, plaque, or thrombotic emboli- personnel to ionizing radiation; and methods of reducing zation; acute occlusion; coronary artery perforation; and patient and staff radiation exposure (73). contrast nephropathy, stroke, and access site complications. 2. Specialized catheterization recording and safety equip- Therefore, only an interventional cardiologist skilled in ment (physiological data recorders, pressure transducers, transluminal coronary techniques such as balloon angio- blood gas analyzers, defibrillators) (74). plasty and stenting who is able to diagnose and treat 3. Catheters, guide wires, balloon catheters, stents, complications of interventional procedures should employ atherectomy devices, ultrasound catheters, intra-aortic these devices. Recommendations regarding the use of balloon pumps, puncture site sealing devices, contrast IVUS, Doppler flow wires, and pressure wires are published agents, distal protection devices, and thrombus extrac- in Appendix C of the ACC/AHA Guidelines for Coronary tion devices. Angiography (75). It is also important to ensure quality image acquisition, Operators must be knowledgeable about the prevention, measurement, and reporting for each of the intravascular prompt recognition, and treatment of procedural complica- assessment devices. For ICUS, the reader is referred to the tions. It is extremely important to have the knowledge and ACC Clinical Expert Consensus Document on Standards skills to diagnose and manage vessel perforation, no reflow, for Acquisition, Measurement and Reporting of Intravas- coronary dissection, expanding hematoma, pseudoaneu- cular Ultrasound Studies (76). No such documents are rysm, arterial venous fistulas, and retroperitoneal hemor- available for Doppler analysis of coronary flow reserve and rhage. Interventionalists must also be cognizant of systemic pressure wire analysis of fractional flow reserve, but many of complications, including cerebrovascular events and the general principles in the IVUS document may be of contrast-related nephropathy. some benefit in guiding appropriate use of these other modalities. Technical Skills Many of the skills required to perform coronary interven- Relationships of Operator and Institutional tional procedures are closely related to those needed to Experience and Activity to Outcomes perform diagnostic cardiac catheterization and coronary in Coronary Interventional Procedures angiography. These include manual dexterity and the ability to obtain percutaneous arterial and venous access and Evidence Reviewed maintain sterile surgical technique. Computerized literature searches of English language pub- Most of the other required technical skills are unique to lications, review of recent abstract publications, and solici- coronary interventional procedures and can only be acquired tation of manuscripts under review for publication from King et al ACCF/AHA/SCAI Clinical Competence Statement 107

Table 3. Published Data Relating Hospital Coronary Angioplasty Volume to Complication Rates

No. of Patients/ Study Data Source Hospitals Studied Conclusions Comments Hartz et al. (78) 1989–1991 Wisconsin Medicare 2,091/16 No relation between volume and outcome Very low number of cases and hospitals examined Ritchie et al. (86) 1989 California State (Adm) 24,883/110 Increased CABG (not death) less than 20 cases per yr; ﬁnding is valid for both acute MI and nonacute MI patients Jollis et al. (85) 1987–1990 MEDPAR (Adm) 217,836/1,194 Death and CABG increased with low volume (risk increases with Medicare patient volume* (less than 100–200 total per yr for death, 200–300 per yr for CABG) Kimmel et al. (84) 1992–1993 SCAI 19,594/48 Fewer major complications for labs with Able to risk adjust more greater than 400 cases per yr completely than most other analyses GUSTO (llb) Angioplasty GUSTO llb trial 565/59 No difference, 200–625 vs. greater than All operators greater than or Substudy Group (36) 625 cases per yr for acute MI patients equal to 50 cases per yr Kato et al. (79) 1991 HCFA (RAND Corp.) 113,576/862 Except for Medicare volume* less than 50, higher volume hospitals had higher mortality rates Stone et al. (80) PAMI II trial 1,100/34 No difference, less than 500, 501–1,000, greater than 1,000 cases per yr for acute MI patients Jollis et al. (77) 1992 Medicare (Adm) 97,498/984 Incremental decrease in death and bypass surgery as hospital Medicare volume* less than 100, 100–200, greater than 200 per yr Tiefenbrunn et al. (83) Second National Registry of MI 4,939/? Increased acute MI mortality for hospital (U.S.) less than 25 acute MI cases per yr Hannan et al. (82) 1991–1994 NY State 62,670/31 Death alone and same-stay CABG Risk-adjusted increased with annual caseloads less than 600 Zahn et al. (81) 1992–1995 German Hospital 4,625/? For patients with acute MI; increased No risk-adjusted Consortium mortality in hospitals with less than or equal to 40 acute MI PTCA per yr Moscucci et al. (22) 1998–1999 NY State and MI 11,374/8 In-hospital death increased for hospital Risk-adjusted volume less than 400 Hannan et al. (21) 1998–2000 NY State 107,713/34 Death, same-day CABG, same-stay CABG Risk-adjusted increased for hospital volume less than 400

*Medicare patients usually constitute 35% to 50% of total interventional caseload. Adm ϭ administrative data set; CABG ϭ coronary artery bypass graft; GUSTO ϭ Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes; HCFA ϭ Health Care Financing Administration; MEDPAR ϭ Medicare provider analysis and review; MI ϭ myocardial infarction; PAMI ϭ Primary Angioplasty in Myocardial Infarction; PTCA ϭ percutaneous transluminal coronary angioplasty; SCAI ϭ Society for Cardiovascular Angiography and Interventions. many physicians and epidemiologists expert in the field were have a greater incidence of procedure-related complications, used to compile the relevant available scientific evidence notably death and need for bypass surgery for failed inter- relating institutional and operator activity level to outcomes vention, than hospitals performing more procedures. Mul- (Table 3). In general, greater weight was given to recent, tiple data sources support the existence of a curvilinear, fully peer-reviewed publications of high quality. No single perhaps logarithmic, statistical relation between caseload work was considered definitive. It was recognized that many and outcome (Fig. 1). However, for CABG, the continued analyses were limited to some extent by lack of capacity to importance of the relationship between volume and out- fully adjust expected outcomes for differences in patient comes has been recently confirmed using contemporary characteristics, changes and advances in the field of inter- clinical data (87). For PCI, the majority of the studies ventional cardiology, and inability to generalize the results available either predate the widespread introduction in to a broader population. interventional practice of coronary stenting and adjunctive Relationship of Institutional Volume use of glycoprotein receptor blockers, or were obtained to Procedural Outcome through analysis of Medicare claims data or other admin- The preponderance of data suggest that, on average, hospi- istrative data. Recognized limitations of Medicare data tals in which fewer coronary interventions are performed include the need to extrapolate the total number of proce-

108 Circulation July 3, 2007

2.0

1.8

1.6

1.4

1.2

1.0 State-Wide Rate 0.8

0.6

Risk-Adjusted MortalityRisk-Adjusted (%) Rate 0.4

0.2

0.0

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000

Mean Annual Volume

Figure 1. Mean Annual Hospital PCI Volume and Risk-Adjusted In-Hospital Mortality Rate in New York State, 1998–2000

Reprinted with permission from Hannan EL, Wu C, Walford G, et al. Volume-outcome relationships for percutaneous coronary interventions in the stent era. Circulation 2005; 112:1171–9 (21). PCI ϭ percutaneous coronary intervention. dures from the number of Medicare procedures, the incom- was found that over time, the disparity in outcomes between plete reporting in Medicare claims of comorbidities that low- and high-volume hospitals had narrowed, and that might be important predictors of adverse outcomes (16,17), outcomes had improved significantly for all hospitals (88). and the possibility of miscoding complications as comor- The author of this study concluded that given these im- bidities (18). provements, lower minimum volume standards might be The direct relationship between institutional volume and justifiable in less populated areas, where the alternative is no outcomes has been recently confirmed by 2 more contem- access to angioplasty at all. Importantly, procedural volume porary analyses of large clinical registries. The first study is only one of many factors contributing to the variability of compared data collected between 1998 and 1999 in a measured outcomes (58,82,89). Furthermore, there is no multicenter PCI registry in Michigan with data from the clear “cut-off ” above or below which hospitals, or groups of New York State data registry (22). An institutional annual hospitals in aggregate, perform well or poorly. There are volume less than 400 cases per year was found to be institutions with low volumes that appear to achieve very independently associated with an increased risk of in- acceptable results. For an individual institution, however, hospital death compared with hospitals with annual volumes such an impression must be tempered by the statistical of at least 400 (adjusted odds ratio [OR] 1.77, 95% imprecision of the estimate of risk. confidence interval [CI] 1.16 to 2.70). The second study Volume and Outcomes Relationship (21), based on the New York State data registry, evaluated for Primary PCI in Acute MI 107,713 procedures performed in 34 hospitals in New York State during 1998 to 2000. The same hospital volume The relationship between operator and institutional volume threshold of less than 400 procedures per year was found to and outcome of primary PCI for acute MI has been be associated with an increased risk of in-hospital mortality examined nearly exclusively at hospitals with onsite cardiac (adjusted OR 1.98, 95% CI 1.17 to 3.35), “same day” surgery. In an analysis including 62,299 patients with acute CABG surgery (adjusted OR 2.07, 95% CI 1.36 to 3.15) or MI and enrolled in the National Registry of Myocardial “same stay” CABG surgery (adjusted OR 1.51, 95% CI 1.03 Infarction, Magid et al. (90) analyzed data from 446 to 2.21). Figure 1 from the New York study presents the acute-care hospitals providing primary angioplasty services. continuous relationship between hospital volume and risk- Hospitals were classified as low volume (less than 16 adjusted in-hospital mortality. procedures per year), intermediate volume (17 to 48 proce- It is important to underscore that advancements in dures per year), and high volume (more than 49 procedures technology have resulted in a progressive improvement in per year). In high-volume hospitals, mortality for acute MI outcomes of PCI, and that this improvement has at least in patients was significantly lower with primary angioplasty part offset the adverse institution volume–outcome relation- when compared with fibrinolysis, while in low-volume ship. In a recent study evaluating temporal trends in the hospitals, there were no differences in mortality rates be- volume–outcome relationship in the state of California, it tween primary angioplasty and fibrinolysis. Two other King et al ACCF/AHA/SCAI Clinical Competence Statement 109

analyses from the same registry and 2 studies using the New surgery onsite may be similar to the relationship in hospitals York State data registry have shown a direct relationship with surgery onsite. For facilities without onsite surgery, it between hospital volume of primary angioplasty and mor- is mandatory that there be an established, well-organized tality. In the analysis by Canto et al. (91), hospital volume plan for transfer for surgery if needed. was divided in quartiles. In-hospital mortality was 28% Relationship of Individual Operator Volume to lower in patients treated in the highest volume quartile (greater than 33 primary PCIs per year) when compared Procedural Outcome with patients treated in the lowest volume quartile (less than Several large studies have assessed the potential relation 12 primary PCIs per year). Similar results were obtained by between individual operator caseload and procedural com- Cannon et al. (92). In this analysis, a procedure volume plications (93). Recently, McGrath et al. (94) analyzed greater than 3 PCIs per month was found to be associated relatively contemporary data (calendar year 1997) from the with a lower in-hospital mortality rate when compared with Medicare database. Based on a slightly different assumption a procedure volume of less than 1 PCI per month, or with than Wennberg et al. (93) that Medicare patients represent a procedure volume between 1 and 3 PCIs per month. 35% to 45% of total PCI procedure volume, they estimated Recently, Hannan et al. (21) reported data from the New that 30 PCIs per operator per year on Medicare patients York State Coronary Angioplasty Reporting System Regis- could be extrapolated to a total procedure volume of 70 try collected in the years 1998 to 2000, a period when PCIs per operator per year (94). A significant relationship stenting was used in a large majority of the STEMI between operator volume and outcomes was also reported in patients. A trend toward an increased odds ratio of in- their study, with better outcomes observed in patients hospital mortality was observed for low-volume operators treated by high-volume operators when compared with when compared with high-volume operators both for a patients treated by low-volume operators. Similar results volume cut of 8 procedures per year (OR 1.40, 95% CI 0.89 were obtained in the study by Hannan et al. (21)inthe to 2.20) and with a volume cut of 10 procedures per year analysis of data collected from the 107,713 procedures (OR 1.27, 95% CI 0.87 to 1.87). Importantly, a significant performed in the 34 hospitals performing PCI in New York increase in the odds of in-hospital mortality was observed State during 1998 to 2000. Operator volume thresholds with lower institutional volume of primary PCI, regardless were set at 75 procedures per year based on ACC/AHA of whether the institutional volume cut point was set at 36 recommendations, and at slightly higher levels of 100 and procedures per year (OR 2.01, 95% CI 1.27 to 3.17), 40 125 procedures per year. There were no differences in procedures per year (OR 1.73, 95% CI 1.1 to 2.71), or 60 risk-adjusted mortality between patients undergoing PCI procedures per year (OR 1.45, 95% CI 1.01 to 2.09). performed by lower volume operators and patients under- Volume and outcomes relationship for PCI in hospitals going PCI performed by higher volume operators for any of without onsite cardiac surgery. There is only 1 report the 3 volume thresholds that were examined. However, for indicating a relationship between institutional PCI volume all 3 volume thresholds, significant differences for “same and outcome in hospitals without onsite cardiac surgery. day” CABG surgery and for “same stay” CABG surgery Wennberg et al. (93) reported that among Medicare recip- were observed. For example, patients undergoing PCI with ients, there was no difference in mortality after primary/ operators performing less than 75 procedures per year had a rescue PCI (emergency procedure on the same day for 65% increased odds of undergoing same-day CABG sur- STEMI) performed at hospitals with or without cardiac gery, and a 55% increased odds of undergoing “same-stay” surgery onsite. However, they did report a higher mortality CABG surgery. for PCI patients, excluding primary/rescue PCI, at hospitals Further confirmation of the adverse operator volume– without cardiac surgery onsite (adjusted OR 1.38, 95% CI outcome relationship with contemporary PCI comes from 1.14 to 1.67; p ϭ 0.001). The relationship between insti- an analysis by Moscucci et al. (95) of another regional, tutional volume and PCI affecting this outcome was con- audited, clinical PCI registry. In that analysis including fined mainly to hospitals without cardiac surgery onsite 18,504 procedures performed in 14 Michigan hospitals in performing 50 or fewer nonprimary/rescue PCIs in Medi- calendar year 2002, operator volume was subdivided in care recipients per year. Among hospitals performing more quintiles (1 to 33 PCIs per year, 34 to 89 PCIs per year, 90 than 100 PCIs in Medicare recipients, mortality was not to 139 PCIs per year, 140 to 206 PCIs per year, and 207 to higher in hospitals without surgery onsite (adjusted OR 582 PCIs per year). The primary end point was a composite 0.76, 95% CI 0.52 to 1.11; p ϭ 0.16). These hospitals likely of MACE, including death, CABG, stroke, transient isch- perform more than 200 PCIs per year based on the emic attack, MI, and repeat PCI at the same lesion site. assumption that 100 Medicare PCIs represent approxi- Stent utilization was greater than 80%, and greater than mately 200 total PCIs per year. 70% of patients received a glycoprotein (GPIIb/IIIa) receptor Taken together, these data suggest that the relationship inhibitor. After adjustment for comorbidities, patients between institutional volume of PCI patients (excluding treated by operators in the 2 lower volume quintiles (Quin- primary/rescue PCI) and mortality in hospitals without tiles 1 and 2) had a 63% increase in the odds of MACE 110 Circulation July 3, 2007

Adjusted Odds ratios for MACE in Quintiles of Operator Volume vs. Quintile 5 ( Clustering model by Hospital)

2.5

2 ** ** 1.63 1.63

1.5

1.24 1.10 1 1 Adjusted ORAdjusted and 95% CI

0.5

0 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 1-33 34-89 90-139 140-206 207-582 Quintiles of Operator Volume with Ranges. ** P<0.0001

Figure 2. Adjusted Odds Ratios for MACE by Quintile of Operator Volume

Reprinted with permission from Moscucci M, Share D, Smith D, et al. Relationship between operator volume and adverse outcome in contemporary percutaneous coronary intervention practice: an analysis of a quality-controlled multicenter percutaneous coronary intervention clinical database. J Am Coll Cardiol 2005; 46:625–32 (95). MACE ϭ major adverse cardiac events.

(Fig. 2). No signiﬁcant relationship was observed between Combination of Individual Operator Volume and operator volume and risk of in-hospital death. The adverse Institutional Volume on Procedural Outcome relationship between operator volume and outcomes appeared to be relatively independent of patient risk. A The combined impact of hospital volume and operator detailed analysis of individual operator risk-adjusted out- volume on adverse outcomes was assessed by Hannan et al. comes revealed the presence of several low-volume operators (21). Patients undergoing PCI performed by operators with with better than expected outcomes, and of a few high- volumes below 75 per year in hospitals with volumes below volume operators with worse than expected outcomes, thus 400 per year were found to have signiﬁcantly higher odds of suggesting that there are exceptions to the rule, and that dying in the hospital than patients undergoing PCI per- low-volume operators should be tracked over a longer formed by operators with volumes of 75 or more in hospitals period of time to ascertain their true performance (Fig. 3). with volumes of 400 or more (OR 5.92, 95% CI 3.25 to

Figure 3. Linear Plot of Standardized MACE Ratios (Observed/Predicted Rates) Versus Annual Operator Volume

King et al ACCF/AHA/SCAI Clinical Competence Statement 111

10.97). Also, patients undergoing PCI performed by oper- using volume as a surrogate, and as more registry data ators with annual volumes of below 75 in hospitals with become available, procedure volume will likely no longer be annual volumes below 400 experienced significantly higher used as a replacement or a surrogate for quality assessment. same-day CABG rates than patients with high-volume Yet, limitations related to the effect of random variation operators (greater than 75 annually) in high-volume hospi- and to the evaluation of rare events continue to exist. These tals (greater than 400 annually), with an OR of 4.02. For limitations make it difficult to assess the true performance of same-stay CABG surgery the respective OR was 3.19. It very-low-volume operators. In such situations, close scru- should be noted that the magnitude of these ORs demon- tiny of case selection and close monitoring of outcomes on strates that the increase in adverse outcomes compound a case-by-case basis might serve as a substitute/complement when patients undergo PCIs performed by low-volume to risk adjustment. operators (less than 75 annually) in low-volume hospitals In summary, while there are inherent limitations in using (less than 400 annually). procedure volume as a surrogate of quality and outcomes, In summary, analysis of more contemporary data supports recent data suggest that there is still a relationship between the hypothesis that technological advancements have not experience and outcomes. In the analysis of the New York completely offset the influence of “practice” in determining State data, the relationship appeared to be at a level of 75 proficiency of contemporary PCIs. However, procedure procedures per year, with further improvement in outcomes volume is only a poor substitute for quality and outcomes; observed at a volume threshold of greater than 100 proce- therefore, it should not be used as a replacement for dures per year. In the analysis of the Michigan data, the appropriately risk-adjusted outcomes. Nevertheless, it is relationship was at a level of 100 procedures per year. On easy to measure, and its potential implications are easily the basis of these data, it is recommended that the operator understood by patients undergoing PCI. As such, it seems volume threshold continue to be 75 procedures per year. appropriate to continue to include procedure volume among Independent of procedure volume, all operators should the several indirect quality indicators of contemporary PCI participate in a regional or national program for outcome practice. assessment and quality improvement. In addition, it is However, it is also important to underscore that there are recognized that there are limitations in the application of significant limitations to the simplistic interpretation of the risk-adjustment methodology in the evaluation of rare procedure volume statistics as a measure of competence and events and of low-volume operators, and that there might be quality. First, it is uncertain whether this relationship is a substantial variations in the volume–outcome relationship. result of the “practice makes perfect” principle, or the fact For operators that do not meet a threshold of 75 cases per that patients are more frequently referred to high-quality year measured in 2-year intervals, it is recommended that a operators. Second, it remains unclear where the “cut-off” case-by-case review, case selection, and prior experience number should be set. Third, studies have shown significant including the total number of cases in a lifetime career be variability in the volume–outcome relationship within the included in their evaluation. They could also partner with same registry, with some low-volume operators having higher volume operators to perform cases together to gain better than expected outcomes, and a few high-volume further experience. operators having worse-than-expected outcomes. Further- more, at present, few or no data exist linking operator Ongoing Quality Improvement and volume to case selection, appropriateness of procedures, Maintenance of Competence periprocedural MI, long-term clinical outcome, or cost- Maintenance of competence in interventional procedures effectiveness, each of which measures a component of should be accomplished for both the individual physician quality of care, or linking clinical outcomes to operator operator and for the institutions in which cardiac interven- experience as measured by the number of years in practice, tional procedures are performed. The goals in setting total procedure volume over a lifetime career, or board criteria for maintaining competence include: certification. The development of national, regional, and state regis- 1. Ensuring quality of patient care and outcomes; tries for outcome assessment is also promoting a shift of the 2. Enabling quality interventionalists and institutions to paradigm surrounding quality of PCI from a mere collection continue to perform PCI; of procedure volume to objective assessment of clinical 3. Providing standards that all institutions and operators outcomes. In addition, the past decade has been character- should strive to achieve. ized by substantial advancement in methodology, scientific Institutional Maintenance of Quality rigor, and acceptance of risk adjustment. Factors related to in-hospital mortality following PCI are now well defined, It is recommended that all institutions have a regular (at and progress is being made toward the development of least monthly) catheterization laboratory conference. The statistical models for other outcomes. Clearly, the calcula- opportunity for ongoing dialogue and collaboration among tion of risk-adjusted outcomes using data from clinical angiographers, interventional operators, and cardiothoracic registries is a more accurate way to assess outcomes than surgical colleagues is highly desirable. New developments in 112 Circulation July 3, 2007

the angioplasty literature should be reviewed, and proce- public to ensure that patient care conducted under its dural complications should be discussed. jurisdiction is of acceptable quality. Quality assessment Maintenance of competence also requires that patient review should be conducted both at the level of the entire outcomes be determined longitudinally for each procedure program and at the level of the individual practitioner. by the institution’s quality assessment program. Participa- Each institution that performs PCI must establish an tion in a state, regional, or national database is highly ongoing mechanism for valid peer review of its quality and encouraged. This allows institutions to measure risk- outcomes. The program should provide an opportunity for adjusted outcomes and compare them to regional and interventionalists as well as physicians who do not perform national benchmarks for improving quality of care. angioplasty, but are knowledgeable about it, to review its It is recommended that lower volume institutions (less overall results on a regular basis. The review process should than 400 interventions per year) consider holding confer- tabulate the results achieved both by individual physician ences with a partnering, more highly experienced institu- operators and by the overall program and compare them to tion. It is also recommended that any institution that falls national benchmark standards with appropriate risk adjust- outside the risk-adjusted national benchmarks in mortality ment. Valid quality assessment requires that the institu- or emergency same-stay CABG during 2 of 3 contiguous tion maintain meticulous and confidential records that 6-month periods have an external audit looking for oppor- include the patient demographic and clinical characteris- tunities to improve quality of care. tics necessary to assess appropriateness and to conduct Individual Maintenance of Quality risk adjustment. To maintain a cognitive knowledge base, it is recommended Role of Risk Adjustment in Assessing Quality that individual operators attend at least 30 h of interven- A raw adverse event rate that is not appropriately risk tional cardiology continuing medical education (CME) every 2 years. This could include catheterization conferences adjusted has little meaning. Data compiled from large and PCI meetings in addition to expanding the use of registries of procedures performed in recent years have simulation cases for procedure use and competence. generated multivariate risk adjustment models for adverse To ensure appropriate patient selection and quality of event rates for PCI in the current era. Six multivariate technical skills, it is recommended that all operators have 5 models of the risk of mortality following PCI have been randomly selected cases and all major complications re- published (62,64,96–99). viewed each year by the catheterization laboratory director Although these models differ somewhat, they are consis- or a Quality Assessment Committee at the institution. Any tent in identifying acute MI, shock, and age as important operator performing less than 75 cases per year should have risk stratification variables for mortality. The ACC- 10 cases reviewed per year. These performance evaluations NCDR® reported an univariate in-hospital mortality of should include feedback to the operator. If it is determined 0.5% for patients undergoing elective PCI, mortality of that the quality of PCI care being provided does not meet 5.1% for patients undergoing primary PCI within6hofthe national benchmarks, the catheterization laboratory director onset of STEMI and mortality of 28% for patients under- should have the discretion of making recommendations for going PCI for cardiogenic shock (64). Thus, it is clear that, improving quality and reassessing over the next 6 months. If in order to assess PCI mortality rates, patients should be disagreements concerning corrective action occur, external stratified by whether they are undergoing elective PCI, review is often helpful. primary PCI for acute STEMI without shock, or primary PCI for STEMI with shock. Quality Assurance Challenges in Determining Quality Definition of Quality in PCI Given the complexity of case selection and procedure Satisfactory quality in PCI may be defined as selecting conduct, quality is difficult to measure in PCI and is not patients appropriately for the procedure and achieving determined solely by adverse event rates even when properly risk-adjusted outcomes that are comparable to national risk adjusted. Accurate assessment of quality becomes more benchmark standards in terms of procedure success and problematic for low-volume operators and institutions be- adverse event rates. To achieve optimal quality and out- cause absolute event rates are expected to be small. Thus, comes in PCI it is necessary that both the physician operator particularly in low-volume circumstances, quality may be and the supporting institution be appropriately skilled and better assessed by an intensive case-review process con- experienced. ducted by recognized experts who can properly judge all of the facets of the conduct of a case. Case review also has Institutional Quality Assurance Requirement merit in high-volume situations as it can identify subtleties In the United States, responsibility for quality assurance is of case selection and procedure conduct that may not be vested in the health care institution that is responsible to the reflected in pooled statistical data. King et al ACCF/AHA/SCAI Clinical Competence Statement 113

Table 4. Key Components of a Quality Assurance Program

Clinical proﬁciency

● General indications/contraindications

● Institutional and individual operator complication rates, mortality, and emergency coronary artery bypass grafting

● Institutional and operator procedure volumes

● Training and qualiﬁcations of support staff Equipment maintenance and management

● Quality of laboratory facility (see ACC/SCAI Expert Consensus Document on Catheterization Laboratory Standards [100]) Quality improvement process

● Establishment of an active concurrent database to track clinical and procedural information and patient outcomes for individual operators and the institution. Participation in multicenter database is highly encouraged. Radiation safety

● Educational program in the diagnostic use of X-ray

● Patient and operator exposure

Requirement for Institutional Resources and Support rare, a valid estimate of a properly risk-adjusted adverse A high-quality PCI program requires appropriately trained, event rate generally requires tabulating the results of a large experienced, and skilled physician operators. However, the number of procedures. This adds an additional challenge to operator does not work in a vacuum. An operator needs a the valid assessment of low-volume operators and institu- well-maintained high-quality cardiac catheterization facility tions. The responsible supervising authority should monitor to practice effectively. In addition, the operator depends on the issues outlined in Table 4. a multidisciplinary institutional infrastructure for support In addition, mere tabulation of adverse event rates, even and response to emergencies. Thus, to provide quality PCI with appropriate risk adjustment, is inadequate to judge services, the institution must ensure that its catheterization operator or program quality. Such tabulations do not ad- facility is properly equipped and managed, and that all of its dress numerous other quality issues—in particular, appro- necessary support services, including data collection, are of priateness. Thus, the quality assessment process should also high quality and are readily available. conduct detailed reviews of both cases that have adverse outcomes, to determine the cause(s) of the adverse event, The Quality Assessment Process and of uncomplicated cases, in order to judge case selection Quality assessment is a complex process that includes more appropriateness and procedure execution quality. These than a mere tabulation of success and complication rates. reviews should be conducted by recognized experienced Components of quality in coronary interventional proce- interventionalists, drawn either from within the institution dures include appropriateness of case selection; quality of or externally, if a requisite number of appropriately qualified procedure execution; proper response to intraprocedural unconflicted individuals are not available. problems; accurate assessment of procedure outcome both Conclusions and Recommendations for PCIs short- and long-term; and appropriateness of postprocedure management. It is important to consider each of these In formulating conclusions and recommendations it is parameters when conducting a quality assessment review. A important to emphasize that the ultimate goal of setting quality program performs appropriately selected procedures standards is to facilitate the attainment of optimal patient while achieving risk-adjusted outcomes, in terms of proce- outcomes. Optimal outcome is most likely when operators dure success and complication rates, that are comparable to select clinically appropriate patients for interventional pro- national benchmark standards. It is accepted that quality cedures and perform these procedures at a requisite level of assurance monitoring is best conducted through the peer- proficiency. Institutional and programmatic quality is ulti- review process despite the political challenges associated mately determined by its success in achieving that goal. with colleagues evaluating each other. There has been Success and Complication Rates considerable controversy surrounding efforts to define standards, criteria, and methodologies for conducting quality Coronary interventional procedures may be complex and assessment. There are many challenges to conducting this technically demanding to perform. Complications of these process in a fair and valid manner. procedures may be life-threatening and can occur unpre- The cornerstone of quality assurance monitoring is the dictably. Nonetheless, recent clinical studies have demon- assessment of procedure outcomes in terms of success and strated that despite increased clinical and angiographic adverse event rates. Other components of quality assurance complexity, procedural and clinical success has remained monitoring include establishing criteria for assessing proce- high and complications have remained low. Angiographic dure appropriateness and applying proper risk adjustment to success (at least 1 lesion successfully dilated by greater than interpret adverse event rates. As adverse events should be 20%, with a residual stenosis of less than 50%), excluding 114 Circulation July 3, 2007

STEMI patients, occurs in over 95% with an average For both institutional and individual volume assessments, mortality rate of less than 1%, a Q-wave MI rate of less than ongoing 2-year volumes should be measured, then averaged 1%, and an emergency CABG rate of less than 1%. to arrive at annual statistics. It is recommended that lower volume institutions (less than 400 per year) consider holding Risk Adjustment conferences with a more experienced partnering institution, Several large retrospective studies have identified both with all staff expected to attend on a regular basis. clinical and angiographic characteristics of PCI that corre- It is also recommended that any institution that falls more late with procedural success, hospital morbidity, and mor- than 2 standard deviations outside the risk-adjusted national tality. These studies have been used to develop multivariate benchmarks in mortality or emergency same-stay CABG logistic regression models that can stratify patients into risk during 2 of 3 contiguous 6-month periods have an external groups before the procedure which have moderate predictive audit looking for opportunities to improve quality of care. value for mortality (C-statistic 0.85 to 0.90), and slightly An institution offering coronary interventional proce- less predictive value for morbidity (C-statistic 0.67 to 0.78). dures should have a physician-director who is responsible for the program’s overall quality. The director should be Volume–Activity Relationships certified in interventional cardiology by the ABIM, with a Analysis of more contemporary data supports the hypothesis career experience of more than 500 procedures. The director that technological advancements have not offset the influ- should perform procedures at the facility that he or she ence of “practice” in determining proficiency of contempo- directs. rary PCIs. There are statistical associations between activity Recommendations for Individual Maintenance levels and short-term complication rates (emergency CABG of Quality and mortality) (17,58,85,89,97,101) for both institutions and for individual operators. In particular, low-volume To maintain an appropriate cognitive knowledge base for operators operating at low-volume hospitals had an in- PCIs, it is recommended that individual operators attend at creased mortality rate. However, procedural volume is only least 30 h of PCI CME every 2 years. The overall perfor- one of many factors contributing to the variability of mance of physicians whose complication rates exceed na- measured outcomes. Furthermore, there is no clear “cut-off” tional benchmark standards for 2 of 3 contiguous 6-month above or below which hospitals or individual operators periods should be reviewed by the program director, with perform well or poorly. Procedural volume continues to be careful attention to statistical power and risk-adjustment correlated with outcomes, but should not serve as a substi- issues. It is recommended that the operator volume thresh- tute for a well-controlled analysis of results and does not old continue to be 75 procedures per year. Monitoring of ensure quality. The development of national, regional and physicians with an annual procedural volume of less than 75 state registries for outcome assessment is promoting objec- should be particularly detailed because of the difficulty of tive assessment of clinical outcomes. estimating their true complication rate. These performance The expected low complication rate for coronary inter- evaluations should include feedback to the operator. ventional procedures presents a major statistical power If it is determined that the quality of PCI care being problem when attempting to estimate the true complication provided does not meet national benchmarks, the catheter- rate of the low-volume operator with meaningful precision. ization laboratory director should have the discretion of In such situations, close scrutiny of case selection and close making recommendations for improving quality and reas- monitoring of outcomes on a case-by-case basis would serve sessing over the next 6 months. These recommendations as a complement to risk adjustment. could include establishing a defined mentoring relationship Highly complex procedures require much more skill and with an experienced operator. If the operator in question experience, and should be undertaken by operators possess- disputes this assessment, then external review may be ing these attributes. Complex cases appropriate for inter- helpful in determining the most appropriate methods of ventions should be referred, not denied. assuring quality performance. Recommendations for Institutional Maintenance of Quality Percutaneous Noncoronary Interventions It is recommended that all institutions have a regular (at Introduction least monthly) catheterization laboratory conference. Pa- tient outcomes should be determined longitudinally for each Noncoronary interventions are a growing and important procedure by the institution’s quality assessment program. contribution to the field of interventional cardiology. The Participation in a state, regional, or national registry is majority of procedures have had their origin in the pediatric highly encouraged to allow institutions to measure risk- population, and several have expanded to the adult patient. adjusted outcomes and compare them to national bench- The purpose of this section is to discuss the training and marks for improving quality of care. experience necessary for the safe and successful performance King et al ACCF/AHA/SCAI Clinical Competence Statement 115

of valvuloplasty, alcohol septal ablation, and percutaneous mum number of cases required for maintenance of compe- repair of ASD/PFO. tency and proficiency. A survey of Pediatric Cardiology The knowledge, skills, and training necessary for compe- Interventional Catheterization training programs concluded tency in noncoronary interventional procedures are different that a minimum of 10 percutaneous ASD closures is from that required for coronary interventions. Therefore, necessary for a trainee to gain clinical competence with the special study of the anatomy, physiology, and pathology of procedure (102). these conditions is a prerequisite for safe and effective With this in mind, it is recommended that interventional treatment. Furthermore, an in-depth understanding of the cardiologists who intend to perform these procedures inde- clinical indications for treatment and the unique complica- pendently, should be involved in these procedures during tions of these treatments are essential. training with at least 10 of these cases being secundum ASD Although the scope of this document is focused on closures. Furthermore, as part of the procedure, the fellow competency, this section will expand the discussion some- should be fully conversant in the use of transesophageal what to describe some anatomical and procedural details. echocardiography and/or intracardiac echocardiography. He Such details are well known for PCI, and their performance or she should understand how to obtain the appropriate is widespread, whereas these noncoronary procedures, in the views to image necessary structures in order to perform the estimation of this Writing Committee, warrant some dis- procedure safely and to exclude other anatomical problems cussion of background information and procedural alternatives. such as a primum or sinus venosus ASD, anomalous pulmonary venous drainage, fenestrated or multiple ASD, Disorders of the Atrial Septum or lipomatous hypertrophy of the septum. Obviously, not all fellows in training will be able to gain this experience and, Criteria for Competency therefore, concentrating the experience in training should be The knowledge base required for performing PCI is differ- limited to a few trainees. ent than that required for percutaneous closure of ASD and Cardiologists in Practice PFO. Extensive knowledge of structural cardiac anatomy, especially that of the atrial septum and the adjacent struc- Interventional cardiologists in practice who were not specifi- tures, is required, as is the understanding of the impact of cally trained in ASD/PFO closure but would like to perform abnormal anatomy and function, and the relative value of these procedures should be fully credentialed in interventional therapeutic options (85,101–105). Therefore, specific train- techniques in their institution. The first several cases should be ing and experience is necessary to safely and successfully done with a proctor. To ensure safety and success, it seems treat this subgroup of patients. The Food and Drug Ad- prudent that the first 10 cases be proctored by someone fully ministration guidelines on the use of device closure of PFOs credentialed in these techniques such as a pediatric cardiologist in these patients state that only patients who have failed or adult cardiologist trained in congenital heart disease. Proc- tors should also be present for the first 3 to 5 cases if a different anticoagulation or have a compelling medical reason to not device is to be used after the initial credentialing proctorship. be anticoagulated are appropriate for device closure. These guidelines should be fully discussed with patients during the Maintenance of Competency for Percutaneous informed consent process. In addition, complications such ASD/PFO Closure as cardiac perforation, device embolization, thrombus for- To maintain physician proficiency and competency in percu- mation on the device, infective endocarditis, arrhythmias, taneous ASD/PFO closure, a minimum of 10 cases per year is and early as well as late erosion of the device through the recommended. Similarly, to maintain catheterization labora- atrial wall or aorta should be disclosed. Currently, 2 studies tory proficiency, a minimum of 10 cases per year should be are underway comparing percutaneous closure of PFO to performed in each institution each year. To achieve this standard oral anticoagulation, which should clarify the experience, it may be necessary to concentrate the procedures indications for interventional treatment. in the hands of only a few operators. A multidisciplinary Since these procedures are relatively new to intervention- program, including neurology consultation for PFO closure, alists trained in adult cardiology, no pre-existing guidelines prospective evaluation of case selection, and evaluation of are available on which to base current opinion. In the clinical outcomes is critical to ensure appropriateness and absence of such guidelines, we arrived at these recommen- maintain safety and efficacy. Laboratories and individual oper- dations from discussions with colleagues actively performing ators that are not active enough to maintain quality outcomes these percutaneous closures. should reconsider treating these patients. Cardiologists in Training Programs Quality Assurance Acquisition of the knowledge and skills necessary to per- The quality improvement process used for oversight of ASD/ form percutaneous procedures to treat ASD and PFO PFO closure should include concurrent case review, and will should be incorporated into the formal training of interven- also benefit from regular case conferences to discuss indica- tional cardiologists. There are no data regarding the mini- tions, procedural techniques, and case outcomes. It is particu- 116 Circulation July 3, 2007

larly useful in any developing procedural area to share results acquisition of clinical skills for the evaluation of indications for with other institutions through informal and formal confer- the procedure and the assessment of suitable valve morphology. ences. Because there are, as of yet, no large databases of It requires the development of proficiency in the performance outcomes for these procedures, participation in local, regional, of transseptal cardiac catheterization, device manipulation, and and national registries is encouraged. Focusing the perfor- online evaluation of hemodynamic parameters. The interven- mance of these procedures in the hands of a few experienced tionalist must be able to recognize and manage complications operators is also recommended. specific to mitral valvuloplasty, including acute mitral regurgi- Hypertrophic Cardiomyopathy and tation, cardiac perforation, pericardial tamponade, and stroke. Alcohol Septal Ablation Although a learning curve has been well described, there are currently no specific data regarding the minimum numbers Hypertrophic cardiomyopathy is the most common genetic needed for competency. Nonetheless, 5 to 10 cases should be cardiovascular disease, with a prevalence in the general popu- done with an experienced colleague before attempting to lation estimated to be 0.2% (103). Physicians performing these perform balloon valvuloplasty independently. Any program procedures should have extensive knowledge of the outcomes, offering mitral valvuloplasty as an alternative to mitral valve limitations and complications of medical therapy (104), dual replacement or surgical commissurotomy for the treatment of chamber pacing and surgical myectomy (105–107), and alcohol mitral stenosis should include a thorough quality assurance septal ablation (105–114). No comparative trial against surgical program and close monitoring of case selection and clinical myectomy has been performed. outcomes. As with other infrequently performed procedures, Criteria for Competency concentration of experience among a small subset of interven- Acquisition of competence. It is strongly recommended that tional cardiologists within an institution is appropriate. alcohol septal ablation be offered within a multidisciplinary Maintenance of competence. With the low prevalence of program that includes the contribution of experienced cardiac mitral stenosis in the United States, maintaining experience surgeons, echocardiographers, general cardiologists, and elec- is difficult. Given this limitation, concentration of this trophysiologists. Although there are currently no data regard- experience among institutional and perhaps regional centers ing the minimum number of procedures required for training may be appropriate. and for credentialing, a minimum number of 10 procedures seems to be appropriate. Quality assurance. Quality assurance in such low-volume procedures requires an approach similar to that outlined for Maintenance of competence. It is recommended that indi- ASD and PFO closures, as previously described. vidual operators perform a minimum of 6 cases per year to maintain competence in performance of septal ablation for Percutaneous Ventricular Assist Devices hypertrophic cardiomyopathy. Each institution should employ Percutaneous ventricular assist devices are becoming avail- a multidisciplinary program with prospective evaluation of case able. They require training and proctored supervision to selection and clinical outcomes. Such an approach is critical for attain competence, as well as periodic use or refresher drills any institution offering alcohol septal ablation as a treatment to maintain competence. As with other seldom-used tech- option for symptomatic patients with hypertrophic obstructive niques, experience should be concentrated among a limited cardiomyopathy. number of operators and laboratory staff who have received Quality assurance. Quality assurance in such low-volume appropriate training. procedures requires an approach similar to that outlined for ASD and PFO closures, as previously described. Laboratory and Staff Competence Valvular Heart Disease In order for laboratories to become competent in the performance of noncoronary cardiac procedures, the super- Cognitive Knowledge Base vising or performing operator should be fully credentialed in Physicians performing invasive procedures on stenotic cardiac the procedure. Initially, this may require off-site training, valves must have extensive knowledge of the pathoanatomy, simulation training, a visiting proctor, or a combination of the hemodynamic alterations, the clinical course, and the these approaches. The operator responsible for the perfor- outcomes of various therapeutic options. Complications of mance of the procedure in the catheterization laboratory aortic (115,116) and mitral (117–119) valvuloplasty should be should supervise the staff in acquiring the necessary skills well understood. and equipment for the procedure. As is the case for the operators of lower volume procedures, there should be a Criteria for Competency small number of dedicated staff members trained to perform Acquisition of competence. Mitral valvuloplasty is one of specific noncoronary interventions, concentrating the expe- the most challenging cardiac procedures. The presence of a rience. If and when a specific procedure becomes more “learning curve” has been well described (120,121). Thus, common, then the training may be expanded to the remain- training in the performance of mitral valvuloplasty requires the der of the staff and operators. King et al ACCF/AHA/SCAI Clinical Competence Statement 117

Conclusions and Recommendations 7. Cowley M, Faxon DP, Holmes DR Jr. Guidelines for training, credentialing, and maintenance of competence for the performance of Percutaneous Noncoronary Interventions coronary angioplasty: a report from the Interventional Cardiology Committee and the Training Program Standards Committee of the Noncoronary cardiac interventions require special training Society for Cardiac Angiography and Interventions. Cathet Cardio- that is not possible for all operators to obtain because of the vasc Diagn 1993;30:1–4. 8. Cowley MJ, King SB. ACC/AHA guidelines for credentialing and small number of these procedures. Therefore, it is necessary facilities for performance of coronary angioplasty. Circulation 1988; to concentrate the activity both in training and practice so 15:136–8. that adequate experience can be obtained to allow for quality 9. Douglas JSJ, Levin DC, Pepine CJ, et al. Recommendations for development and maintenance of competence in coronary interven- performance. Hospitals should develop clear credentialing tional procedures. American College of Cardiology Cardiac Cathe- criteria, despite the small number of cases and empiric data terization Committee. J Am Coll Cardiol 1993;22:629–31. from which to judge appropriateness, as well as success and 10. Parker DJ, Birkhead JS, Balcon R, et al. Planning for coronary angioplasty: Guidelines for training and continuing competence. complication rates of these procedures. British Cardiac Society (BCS) and British Cardiovascular Interven- The quality improvement process used for oversight of tion Society (BCIS) Working Group on Interventional Cardiology. percutaneous noncoronary interventions should include Heart 1996;75:419–25. 11. Pepine CJ, Babb JD, Brinker JA, et al. Guidelines for training in concurrent case review, and will also benefit from regular adult cardiovascular medicine. Core Cardiology Training Symposium case conferences to discuss indications, procedural tech- (COCATS). Task Force 3: training in cardiac catheterization and niques, and case outcomes. It is particularly useful in any interventional cardiology. J Am Coll Cardiol 1995;25:14–6. 12. Ryan TJ, Faxon DP, Gunnar RM, et al. Guidelines for percutaneous developing procedural area to share results with other transluminal coronary angioplasty: a report of the American College institutions through informal and formal conferences. Since of Cardiology/American Heart Association Task Force on Assess- there are, as of yet, no large databases of outcomes for these ment of Diagnostic and Therapeutic Cardiovascular Procedures procedures, participation in local, regional, and national (Subcommittee on Percutaneous Transluminal Coronary Angio- plasty). J Am Coll Cardiol 1988;12:529–45. registries is encouraged. Focusing the performance of these 13. Ryan TJ, Klocke FJ, Reynolds WA. Clinical competence in percu- procedures in the hands of a few experienced operators is taneous transluminal coronary angioplasty: a statement for physicians also recommended. from the ACP/ACC/AHA Task Force on Clinical Privileges in Cardiology. Circulation 1990;81:2041–6. 14. Ryan TJ, Bauman WB, Kennedy JW, et al. Guidelines for percutaneous transluminal coronary angioplasty: a report of the American Staff College of Cardiology/American Heart Association Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Proce- American College of Cardiology Foundation dures (Committee on Percutaneous Transluminal Coronary Angio- John C. Lewin, MD, Chief Executive Officer plasty). J Am Coll Cardiol 1993;22:2033–54. 15. Smith SC Jr., Feldman TE, Hirshfeld JW Jr., et al. ACC/AHA/ Thomas E. Arend, Jr, Esq., Chief Operating Officer SCAI 2005 guideline update for percutaneous coronary interven- Lisa Bradfield, Associate Director, Practice Guidelines tion—summary article: a report of the American College of Cardi- Erin A. Barrett, Senior Specialist, Clinical Policy and ology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the Documents 2001 Guidelines for Percutaneous Coronary Intervention). J Am Coll American Heart Association Cardiol 2006;47:216–35. 16. Hirshfeld JW Jr., Ellis SG, Faxon DP. Recommendations for the M. Cass Wheeler, Chief Executive Officer assessment and maintenance of proficiency in coronary interventional Rose Marie Robertson, MD, FACC, FAHA, Chief procedures: atatement of the American College of Cardiology. J Am Science Officer Coll Cardiol 1998;31:722–43. 17. Jollis JG, Peterson ED, Nelson CL, et al. Relationship between Kathryn A. Taubert, PhD, FAHA, Senior Scientist physician and hospital coronary angioplasty volume and outcome in elderly patients. Circulation 1997;95:2485–91. REFERENCES 18. Creager MA, Goldstone J, Hirshfeld JW Jr., et al. ACC/ACP/ SCAI/SVMB/SVS clinical competence statement on vascular med- 1. Gruntzig A. Transluminal dilatation of coronary-artery stenosis. icine and catheter-based peripheral vascular interventions: a report of Lancet 1978;1:263. the American College of Cardiology/American Heart Association/ 2. Fischman DL, Leon MB, Baim DS, et al. A randomized comparison American College of Physician Task Force on Clinical Competence of coronary-stent placement and balloon angioplasty in the treatment (ACC/ACP/SCAI/SVMB/SVS Writing Committee to Develop a of coronary artery disease. Stent Restenosis Study Investigators. Clinical Competence Statement on Peripheral Vascular Disease). N Engl J Med 1994;331:496–501. J Am Coll Cardiol 2004;44:941–57. 3. Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus 19. Hodgson JM, Tommaso CL, Watson RM, Weiner BH. Core standard stents in patients with stenosis in a native coronary artery. curriculum for the training of adult invasive cardiologists: report of N Engl J Med 2003;349:1315–23. the Society for Cardiac Angiography and Interventions Committee 4. Thom T, Haase N, Rosamond W, et al. Heart disease and stroke on Training Standards. Cathet Cardiovasc Diagn 1996;37:392–408. statistics—2006 update: a report from the American Heart Associa- 20. Hirshfeld JW Jr., Banas JS Jr., Brundage BH, et al. American College tion Statistics Committee and Stroke Statistics Subcommittee. Cir- of Cardiology training statement on recommendations for the struc- culation 2006;113:e85–151. ture of an optimal adult interventional cardiology training program: a 5. American Board of Internal Medicine Certification Examination. report of the American College of Cardiology task force on clinical Available at: http://www.abim.org/resources/statcert.shtm. Last up- expert consensus documents. J Am Coll Cardiol 1999;34:2141–7. date 2005. 21. Hannan EL, Wu C, Walford G, et al. Volume-outcome relationships 6. Conti CR, Faxon DP, Gruentzig A, Gunnar RM, Lesch M, Reeves for percutaneous coronary interventions in the stent era. Circulation TJ. Training in cardiac catheterization. Bethesda Conference 17: 2005;112:1171–9. Adult Cardiology Training Task Force III. J Am Coll Cardiol 22. Moscucci M, Eagle KA, Share D, et al. Public reporting and case 1986;7:1205–6. selection for percutaneous coronary interventions: an analysis from 118 Circulation July 3, 2007

two large multicenter percutaneous coronary intervention databases. of patients with multivessel coronary disease after CAGB or PCI: J Am Coll Cardiol 2005;45:1759–65. analysis of BARI-like patients in northern New England. Circulation 23. Gruntzig A. Percutaneous dilatation of experimental coronary artery 2005;112:I371–6. stenosis—description of a new catheter system. Klin Wochenschr 44. Moscucci M, Rogers EK, Montoye C, et al. Association of a 1976;54:543–45. continuous quality improvement initiative with practice and outcome 24. CABRI Trial Participants. First-year results of CABRI (Coronary variations of contemporary percutaneous coronary interventions. Angioplasty versus Bypass Revascularisation Investigation). Lancet Circulation 2006;113:814–22. 1995;346:1179–84. 45. Percutaneous Coronary Interventions (PCI) in New York State, 25. The Bypass Angioplasty Revascularization Investigation (BARI) 2001–2003. New York State Department of Health. Available at: Investigators. Comparison of coronary bypass surgery with angio- http://www.health.state.ny.us/nysdoh/heart/pdf/pci_2001–2003.pdf. plasty in patients with multivessel disease. N Engl J Med 1996;335: Last update 2005. 217–25. 46. Weintraub WS, McKay CR, Riner RN, et al. The American College 26. Writing Group for the Bypass Angioplasty Revascularization Inves- of Cardiology National Database: progress and challenges. American tigation (BARI) Investigators. Five-year clinical and functional out- College of Cardiology Database Committee. J Am Coll Cardiol come comparing bypass surgery and angioplasty in patients with 1997;29:459–65. multivessel coronary disease: a multicenter randomized trial. JAMA 47. Brown BG, Bolson EL, Dodge HT. Quantitative computer tech- 1997;277:715–21. niques for analyzing coronary arteriograms. Prog Cardiovasc Dis 27. King SB III, Lembo NJ, Weintraub WS, et al. A randomized trial 1986;28:403–18. comparing coronary angioplasty with coronary bypass surgery. Emory 48. Detre KM, Wright E, Murphy ML, Takaro T. Observer agreement Angioplasty versus Surgery Trial (EAST). N Engl J Med 1994;331: in evaluating coronary angiograms. Circulation 1975;52:979–86. 1044–50. 49. Califf RM, Abdelmeguid AE, Kuntz RE, et al. Myonecrosis after 28. Serruys PW, Unger F, Sousa JE, et al. Comparison of coronary-artery revascularization procedures. J Am Coll Cardiol 1998;31:241–51. bypass surgery and stenting for the treatment of multivessel disease. 50. Cutlip DE, Chhabra AG, Baim DS, et al. Beyond restenosis: N Engl J Med 2001;344:1117–24. five-year clinical outcomes from second-generation coronary stent 29. Sharma S, Forrester J, Makkar RM. The SoS trial. Lancet 2003;361: trials. Circulation 2004;110:1226–30. 614–5. 51. Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for 30. Hamm CW, Reimers J, Ischinger T, Rupprecht HJ, Berger J, prediction of contrast-induced nephropathy after percutaneous cor- Bleifeld W. A randomized study of coronary angioplasty compared onary intervention: development and initial validation. J Am Coll with bypass surgery in patients with symptomatic multivessel coro- Cardiol 2004;44:1393–9. nary disease. German Angioplasty Bypass Surgery Investigation 52. Ellis SG, Vandormael MG, Cowley MJ, et al. Coronary morphologic (GABI). N Engl J Med 1994;331:1037–43. and clinical determinants of procedural outcome with angioplasty for 31. Myler RK, Shaw RE, Stertzer SH, et al. Lesion morphology and multivessel coronary disease. Implications for patient selection. Mul- coronary angioplasty: current experience and analysis. J Am Coll tivessel Angioplasty Prognosis Study Group. Circulation 1990;82: Cardiol 1992;19:1641–52. 1193–2. 32. de Feyter PJ, van Suylen RJ, de Jaegere PP, Topol EJ, Serruys PW. 53. Altmann DB, Racz M, Battleman DS, et al. Reduction in angioplasty Balloon angioplasty for the treatment of lesions in saphenous vein complications after the introduction of coronary stents: results from a bypass grafts. J Am Coll Cardiol 1993;21:1539–49. consecutive series of 2242 patients. Am Heart J 1996;132:503–7. 33. Ambrose JA, Torre SR, Sharma SK, et al. Adjunctive thrombolytic 54. Williams DO, Holubkov R, Yeh W, et al. Percutaneous coronary therapy for angioplasty in ischemic rest angina: results of a double- intervention in the current era compared with 1985–1986: the blind randomized pilot study. J Am Coll Cardiol 1992;20:1197–204. National Heart, Lung, and Blood Institute Registries. Circulation 34. Perry RA, Seth A, Hunt A, Shiu MF. Coronary angioplasty in 2000;102:2945–51. unstable angina and stable angina: a comparison of success and 55. Malenka DJ, Wennberg DE, Quinton HA, et al. Gender-related complications. Br Heart J 1988;60:367–72. changes in the practice and outcomes of percutaneous coronary 35. Grines CL, Browne KF, Marco J, et al. A comparison of immediate interventions in Northern New England from 1994 to 1999. J Am angioplasty with thrombolytic therapy for acute myocardial infarc- Coll Cardiol 2002;40:2092–101. tion. The Primary Angioplasty in Myocardial Infarction Study 56. Krone RJ, Kimmel SE, Laskey WK, et al. Evaluation of the Society Group. N Engl J Med 1993;328:673–9. for Coronary Angiography and Interventions lesion classification 36. GUSTO (IIb) Angioplasty Substudy Group. The Global Use of system in 14,133 patients with percutaneous coronary interventions Strategies to Open Occluded Coronary Arteries in Acute Coronary in the current stent era. Catheter Cardiovasc Interv 2002;55:1–7. Syndromes: a clinical trial comparing primary coronary angioplasty 57. Krone RJ, Shaw RE, Klein LW, et al. Evaluation of the American with tissue plasminogen activator for acute myocardial infarction. College of Cardiology/American Heart Association and the Society N Engl J Med 1997;336:1621–8. for Coronary Angiography and Interventions lesion classification 37. Bittl JA. Advances in coronary angioplasty. N Engl J Med 1996;24; system in the current “stent era” of coronary interventions (from the 335:1290–302. ACC-National Cardiovascular Data Registry). Am J Cardiol 2003; 38. Detre KM, Holubkov R, Kelsey S, et al. Percutaneous transluminal 92:389–94. coronary angioplasty in 1985–1986 and 1977–1981: the National 58. Ellis SG, Omoigui N, Bittl JA, et al. Analysis and comparison of Heart, Lung, and Blood Institute Registry. N Engl J Med 1988;318: operator-specific outcomes in interventional cardiology: from a mul- 265–70. ticenter database of 4860 quality-controlled procedures. Circulation 39. Holmes DRJ, Holubkov R, Vlietstra RE, et al. Comparison of 1996;93:431–9. complications during percutaneous transluminal coronary angioplasty 59. Krone RJ, Laskey WK, Johnson C, et al. A simplified lesion from 1977 to 1981 and from 1985 to 1986: the National Heart, classification for predicting success and complications of coronary Lung, and Blood Institute Percutaneous Transluminal Coronary angioplasty. Registry Committee of the Society for Cardiac Angiog- Angioplasty Registry. J Am Coll Cardiol 1988;12:1149–55. raphy and Intervention. Am J Cardiol 2000;85:1179–84. 40. Kent KM, Bentivoglio LG, Block PC, et al. Percutaneous translu- 60. Singh M, Lennon RJ, Holmes DR Jr., Bell MR, Rihal CS. minal coronary angioplasty: report from the Registry of the National Correlates of procedural complications and a simple integer risk score Heart, Lung, and Blood Institute. Am J Cardiol 1982;49:2011–20. for percutaneous coronary intervention. J Am Coll Cardiol 2002;40: 41. Abbott JD, Choi EJ, Selzer F, Srinivas VS, Williams DO. Impact of 387–93. coronary collaterals on outcome following percutaneous coronary 61. Maynard C, Goss JR, Malenka DJ, Reisman M. Adjusting for intervention (from the National Heart, Lung, and Blood Institute patient differences in predicting hospital mortality for percutaneous Dynamic Registry). Am J Cardiol 2005;96:676–80. coronary interventions in the Clinical Outcomes Assessment Pro- 42. Hirshfeld JW Jr. ACC-National Cardiovascular Data Registry®, gram. Am Heart J 2003;145:658–64. CathPCI data repository, 1998–2005. November 28, 2005. 62. Moscucci M, Kline-Rogers E, Share D, et al. Simple bedside additive 43. Malenka DJ, Leavitt BJ, Mearne MS, et al. Northern New England tool for prediction of in-hospital mortality after percutaneous coro- Cardiovascular Disease Study Group. Comparing long-term survival nary interventions. Circulation 2001;104:263–8.

King et al ACCF/AHA/SCAI Clinical Competence Statement 119

63. O’Connor GT, Malenka DJ, Quinton H, et al. Multivariate predic- 77. Jollis JG, Peterson ED, Nelson CL, et al. Relationship between tion of in-hospital mortality after percutaneous coronary interven- physician and hospital coronary angioplasty volume and outcome in tions in 1994–1996. Northern New England Cardiovascular Disease elderly patients. Circulation 1997;95:2485–91. Study Group. J Am Coll Cardiol 1999;34:681–91. 78. Hartz AJ, Kuhn EM, Kayser KL, et al. Assessing providers of 64. Shaw RE, Anderson HV, Brindis RG, et al. Development of a risk coronary revascularization: a method for peer review organizations. adjustment mortality model using the American College of Am J Public Health 1992;82:1631–40. Cardiology-National Cardiovascular Data Registry (ACC-NCDR) 79. Kato NS, Ergun ME, Carter GM. Health policy implications of experience: 1998–2000. J Am Coll Cardiol 2002;39:1104–12. volume recommendations on percutaneous transluminal coronary 65. Shaw RE, Anderson HV, Brindis RG, et al. Updated risk adjustment angioplasty in the United States (abstr). Circulation 1996;94:I532. mortality model using the complete 1.1 dataset from the American 80. Stone GW, Marsalese D, Brodie BR, et al., on behalf of the Second College of Cardiology National Cardiovascular Data Registry (ACC- Primary Angioplasty In Myocardial Infarction (PAMI-II) Trial NCDR). J Invasive Cardiol 2003;15:578–80. Investigators. A prospective, randomized evaluation of prophylactic 66. Wu C, Hannan EL, Walford G, et al. A risk score to predict intraaortic balloon counterpulsation in high risk patients with acute in-hospital mortality for percutaneous coronary interventions. J Am myocardial infarction treated with primary angioplasty. J Am Coll Coll Cardiol 2006;47:654–60. Cardiol 1997;29:1459–67. 67. Pepine CJ, Babb JD, Brinker JA, et al. Guidelines for training in 81. Zahn R, Vogt A, Seidl K, et al, Balloon dilatation in acute myocardial adult cardiovascular medicine. Core Cardiology Training Symposium infarct in routine clinical practice: results of the register of the (COCATS). Task Force 3: training in cardiac catheterization and Working Society of Leading Cardiologic Hospital Physicians in interventional cardiology. J Am Coll Cardiol 1995;25:14–6. 4,625 patients. Z Kardiol 1997;86:712–21. 68. Beller GA, Bonow RO, Fuster V. ACC revised recommendations for 82. Hannan E, Racz M, Ryan TJ, et al. Coronary angioplasty volume- training in adult cardiovascular medicine. Core Cardiology Training outcome relationships for hospitals and operators in New York State: II (COCATS 2). (Revision of the 1995 COCATS training state- 1991–1994. JAMA 1997;277:892–8. ment). J Am Coll Cardiol 2002;39:1242–6. 83. Tiefenbrunn AJ, Chandra NC, French WJ, Gore JM, Rogers 69. Smith SC Jr., Dove JT, Jacobs AK, et al. ACC/AHA guidelines of WJ. Clinical experience with primary PTCA compared with alteplase percutaneous coronary interventions (revision of the 1993 PTCA (rt-PA) in patients with acute myocardial infarction: a report from guidelines)—executive summary. A report of the American College the Second National Registry of Myocardial Infarction (NRMI 2). of Cardiology/American Heart Association Task Force on Practice J Am Coll Cardiol 1998;31:1240–5. Guidelines (Committee to Revise the 1993 Guidelines for Percuta- 84. Kimmel SE, Berlin JA, Strom BL, Laskey WK. Development and neous Transluminal Coronary Angioplasty). J Am Coll Cardiol validation of simplified predictive index for major complications in 2001;37:2215–39. contemporary percutaneous transluminal coronary angioplasty prac- 70. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guide- tice. The Registry Committee of the Society for Cardiac Angiogra- lines for the management of patients with ST-elevation myocardial phy and Interventions. J Am Coll Cardiol 1995;26:931–8. 85. Jollis JG, Peterson ED, DeLong ER, et al. The relation between the infarction: a report of the American College of Cardiology/American volume of coronary angioplasty procedures at hospitals treating Heart Association Task Force on Practice Guidelines (Committee to Medicare beneficiaries and short-term mortality. N Engl J Med Revise the 1999 Guidelines for the Management of Patients With 1994;331:1625–29. Acute Myocardial Infarction). J Am Coll Cardiol 2004;44:E1–211. 86. Ritchie JL, Phillips KA, Luft HS. Coronary angioplasty: statewide 71. Braunwald E, Antman EM, Beasley JW, et al. ACC/AHA 2002 experience in California. Circulation 1993;88:2735–43. guideline update for the management of patients with unstable 87. Peterson ED, Coombs LP, DeLong ER, Haan CK, Ferguson TB. angina and non-ST-segment elevation myocardial infarction— Procedural volume as a marker of quality for CABG surgery. JAMA summary article: a report of the American College of Cardiology/ 2004;291:195–201. American Heart Association task force on practice guidelines (Com- 88. Ho V. Evolution of the volume-outcome relation for hospitals mittee on the Management of Patients With Unstable Angina). J Am performing coronary angioplasty. Circulation 2000;101:1806–11. Coll Cardiol 2002;40:1366–74. 89. Kimmel SE, Berlin JA, Laskey WK. The relationship between 72. Eagle KA, Guyton RA, Davidoff R, et al. ACC/AHA 2004 guideline coronary angioplasty procedure volume and major complications. update for coronary artery bypass graft surgery: summary article. A JAMA 1995;274:1137–42. report of the American College of Cardiology/American Heart 90. Magid DJ, Calonge BN, Rumsfeld JS, et al. Relation between Association Task Force on Practice Guidelines (Committee to hospital primary angioplasty volume and mortality for patients with Update the 1999 Guidelines for Coronary Artery Bypass Graft acute MI treated with primary angioplasty vs thrombolytic therapy. Surgery). J Am Coll Cardiol 2004;44:e213–e310. JAMA 2000;284:3131–8. 73. Hirshfeld JW Jr., Balter S, Brinker JA, et al. ACCF/AHA/HRS/ 91. Canto JG, Every NR, Magid DJ, et al. The volume of primary SCAI clinical competence statement on physician knowledge to angioplasty procedures and survival after acute myocardial infarction. optimize patient safety and image quality in fluoroscopically guided National Registry of Myocardial Infarction 2 Investigators. N Engl invasive cardiovascular procedures: a report of the American College J Med 2000;342:1573–80. of Cardiology Foundation/American Heart Association/American 92. Cannon CP, Gibson CM, Lambrew CT, et al. Relationship of College of Physicians Task Force on Clinical Competence and symptom-onset-to-balloon time and door-to-balloon time with mor- Training. J Am Coll Cardiol 2004;44:2259–82. tality in patients undergoing angioplasty for acute myocardial infarc- 74. Kern MJ, Lim M. Evaluation of myocardial blood flow and metab- tion. JAMA 2000;283:2941–7. olism. In: Baim DS, editor. Grossman’s Cardiac Catheterization, 93. Wennberg DE, Lucas FL, Siewers AE, Kellett MA, Malenka DJ. Angiography, and Intervention. Philadelphia, PA: Lippincott Wil- Outcomes of percutaneous coronary interventions performed at liams & Wilkins, 2006:335–70. centers without and with onsite coronary artery bypass graft surgery. 75. Scanlon PJ, Faxon DP, Audet AM, et al. ACC/AHA guidelines for JAMA 2004;292:1961–8. coronary angiography. A report of the American College of Cardi- 94. McGrath PD, Wennberg DE, Dickens JD Jr., et al. Relation ology/American Heart Association Task Force on Practice Guide- between operator and hospital volume and outcomes following lines (Committee on Coronary Angiography). Developed in collab- percutaneous coronary interventions in the era of the coronary stent. oration with the Society for Cardiac Angiography and Interventions. JAMA 2000;284:3139–44. J Am Coll Cardiol 1999;33:1756–824. 95. Moscucci M, Share D, Smith D, et al. Relationship between operator 76. Mintz GS, Nissen SE, Anderson WD, et al. American College of volume and adverse outcome in contemporary percutaneous coronary Cardiology clinical expert consensus document on standards for intervention practice: an analysis of a quality-controlled multicenter acquisition, measurement and reporting of intravascular ultrasound percutaneous coronary intervention clinical database. J Am Coll studies (IVUS). A report of the American College of Cardiology Cardiol 2005;46:625–32. Task Force on Clinical Expert Consensus Documents. J Am Coll 96. Ellis SG, Weintraub W, Holmes D, Shaw R, Block PC, King SB III. Cardiol 2001;37:1478–92. Relation of operator volume and experience to procedural outcome of 120 Circulation July 3, 2007

percutaneous coronary revascularization at hospitals with high 108. Bhargava B, Narang R, Aggarwal R, Bahl VK, Manchanda SC. interventional volumes. Circulation 1997;95:2479–84. Conduction blocks following transcatheter septal ablation for hyper- 97. Hannan EL, Racz M, Ryan TJ, et al. Coronary angioplasty volume- trophic cardiomyopathy. Eur Heart J 1997;18:2011–2. outcome relationships for hospitals and cardiologists. JAMA 1997; 109. Boltwood CM Jr., Chien W, Ports T. Ventricular tachycardia 277:892–8. complicating alcohol septal ablation. N Engl J Med 2004;351: 98. Holmes DR, Selzer F, Johnston JM, et al. Modeling and risk 1914–5. prediction in the current era of interventional cardiology: a report 110. Faber L, Meissner A, Ziemssen P, Seggewiss H. Percutaneous from the National Heart, Lung, and Blood Institute Dynamic transluminal septal myocardial ablation for hypertrophic obstructive Registry. Circulation 2003;107:1871–6. cardiomyopathy: long term follow up of the ﬁrst series of 25 patients. 99. McGrath PD, Malenka DJ, Wennberg DE, et al. Changing out- Heart 2000;83:326–31. comes in percutaneous coronary interventions: a study of 34,752 111. Fernandes VL, Nagueh SF, Wang W, Roberts R, Spencer WH III. procedures in northern New England, 1990 to 1997. Northern New A prospective follow-up of alcohol septal ablation for symptomatic England Cardiovascular Disease Study Group. J Am Coll Cardiol hypertrophic obstructive cardiomyopathy—the Baylor experience 1999;34:674–80. (1996–2002). Clin Cardiol 2005;28:124–30. 100. Bashore TM, Bates ER, Berger PB, et al. American College of Cardiology/Society for Cardiac Angiography and Interventions clin- 112. Kern MJ, Holmes DG, Simpson C, Bitar SR, Rajjoub H. Delayed ical expert consensus document on cardiac catheterization laboratory occurrence of complete heart block without warning after alcohol standards: a report of the American College of Cardiology Task septal ablation for hypertrophic obstructive cardiomyopathy. Cathe- Force on Clinical Expert Consensus Documents. J Am Coll Cardiol ter Cardiovasc Interv 2002;56:503–7. 2001;37:2170–214. 113. Lakkis NM, Nagueh SF, Kleiman NS, et al. Echocardiography- 101. Ellis SG, Weintraub W, Holmes DR Jr., et al. Relation of operator guided ethanol septal reduction for hypertrophic obstructive cardio- volume and experience to procedural outcome of percutaneous myopathy. Circulation 1998;98:1750–5. coronary revascularization at hospitals with high interventional vol- 114. Nagueh SF, Ommen SR, Lakkis NM, et al. Comparison of ethanol umes. Circulation 1997;95:2479–84. septal reduction therapy with surgical myectomy for the treatment of 102. Beekman RH III, Hellenbrand WE, Lloyd TR, et al. ACCF/ hypertrophic obstructive cardiomyopathy. J Am Coll Cardiol 2001; AHA/AAP recommendations for training in pediatric cardiology. 38:1701–6. Task force 3: training guidelines for pediatric cardiac catheteriza- 115. Kuntz RE, Tosteson AN, Berman AD, et al. Predictors of event-free tion and interventional cardiology endorsed by the Society for survival after balloon aortic valvuloplasty. N Engl J Med 1991;325: Cardiovascular Angiography and Interventions. J Am Coll Cardiol 17–23. 2005;46:1388–90. 116. Saﬁan RD, Berman AD, Diver DJ, et al. Balloon aortic valvuloplasty 103. Maron BJ, Gardin JM, Flack JM, Gidding SS, Kurosaki TT, Bild in 170 consecutive patients. N Engl J Med 1988;319:125–30. DE. Prevalence of hypertrophic cardiomyopathy in a general popu- 117. The National Heart, Lung, and Blood Institute Balloon Valvulo- lation of young adults. Echocardiographic analysis of 4111 subjects in plasty Registry Participants. Multicenter experience with balloon the CARDIA study. Coronary Artery Risk Development in (Young) mitral commissurotomy. NHLBI Balloon Valvuloplasty Registry Adults. Circulation 1995;92:785–9. Report on immediate and 30-day follow-up results. Circulation 104. Maron BJ, McKenna WJ, Danielson GK, et al. American College of 1992;85:448–61. Cardiology/European Society of Cardiology clinical expert consensus 118. Complications and mortality of percutaneous balloon mitral commis- document on hypertrophic cardiomyopathy: a report of the American surotomy. A report from the National Heart, Lung, and Blood College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Institute Balloon Valvuloplasty Registry. Circulation 1992;85: Committee for Practice Guidelines. J Am Coll Cardiol 2003;42: 2014–24. 1687–713. 119. Feldman T. Hemodynamic results, clinical outcome, and complica- 105. Maron BJ. Surgery for hypertrophic obstructive cardiomyopathy: tions of Inoue balloon mitral valvotomy. Cathet Cardiovasc Diagn alive and quite well. Circulation 2005;111:2016–8. 1994;Suppl 2:2–7. 106. Williams WG, Wigle ED, Rakowski H, Smallhorn J, LeBlanc J, 120. Rihal CS, Nishimura RA, Holmes DR Jr. Percutaneous balloon Trusler GA. Results of surgery for hypertrophic obstructive cardio- mitral valvuloplasty: the learning curve. Am Heart J 1991;122: myopathy. Circulation 1987;76:V104–8. 1750–6. 107. Woo A, Williams WG, Choi R, et al. Clinical and echocardiographic 121. Sanchez PL, Harrell LC, Salas RE, Palacios IF. Learning curve of determinants of long-term survival after surgical myectomy in obstruc- the Inoue technique of percutaneous mitral balloon valvuloplasty. tive hypertrophic cardiomyopathy. Circulation 2005;111:2033–41. Am J Cardiol 2001;88:662–7. King et al ACCF/AHA/SCAI Clinical Competence Statement 121

APPENDIX 1. AUTHOR RELATIONSHIPS WITH INDUSTRY—ACCF/AHA/SCAI WRITING COMMITTEE TO UPDATE THE CLINICAL COMPETENCE STATEMENT ON CARDIAC INTERVENTIONAL PROCEDURES

Research Scientiﬁc Speakers’ Steering Name Consultant Grant Advisory Board Bureau Committee Stock Holder Other Dr. Thomas None None None None None None None Aversano

Dr. William L. None None None None None None None Ballard

Dr. Robert H. ● AGA Medical None None None None None None Beekman, III

Dr. Michael J. None None None None None None None Cowley

Dr. Stephen G. ● Boston ● Celera ● Boston Scientiﬁc None None None None Ellis Scientiﬁc ● Centacor/Lilly ● Cordis ● Celera ● Cordis ● Viacon ● Cordis ● Guidant ● Viacon

Dr. David P. ● Bristol-Myers None ● Boston Scientiﬁc None None ● Medical Technology None Faxon Squibb/Sanoﬁ Informational

Dr. Edward L. None None None None None None None Hannon

Dr. John W. None None ● Bracco Inc. None None None None Hirshfeld, Jr. ● Bristol-Myers Squibb/Sanoﬁ

Dr. Alice K. None None None None None None ● Wyeth–Spouse’s Jacobs Employer

Dr. Mirle A. None None None None None None None Kellett, Jr.

Dr. Stephen E. None None None None None None None Kimmel

Dr. Spencer B. ● Bristol-Myers None ● Medtronic ● Bristol-Myers None None ● Novoste–Royalties King, III Squibb Squibb ● CV Therapeutics ● Sanoﬁ ● Sanoﬁ/Aventis

Dr. Joel S. None None None None None None None Landzberg

Dr. Louis S. None None None None None None None McKeever

Dr. Mauro None ● BlueCross/ None ● Aventis None None ● Cordis–Fellowship Moscucci BlueShield ● Pﬁzer Training Grant

Dr. Richard M. ● Medacorp. None None ● Sanofi/Aventis None ● Pfizer None Pomerantz ● Sanofi/Aventis

Dr. Karen M. None None None None None None None Smith

Dr. George W. Merck ● Cordis/Johnson None ● Lilly None ● Johnson & Johnson None Vetrovec & Johnson ● Pﬁzer ● NHLBI

This table represents the relationships of committee members with industry that were reported orally at the initial writing committee meeting and updated in conjunction with all meetings and conference calls of the writing committee during the document development process. It does not necessarily reﬂect relationships with industry at the time of publication. 122 Circulation July 3, 2007

APPENDIX 2. PEER REVIEWER RELATIONSHIPS WITH INDUSTRY—ACCF/AHA/SCAI 2007 UPDATE OF THE CLINICAL COMPETENCE STATEMENT ON CARDIAC INTERVENTIONAL PROCEDURES

Scientiﬁc Advisory Speakers’ Steering Stock Name Representation Consultant Research Grant Board Bureau Committee Holder Other

Dr. John C. Giacomini ● Ofﬁcial–AHA None None None None None None None

Dr. Lawrence Laslett ● Ofﬁcial–ACC None None None None None ● General None Board of Electric Governors

Dr. Carl J. Pepine ● Official–ACC ● Abbott ● Abbott None None None None ● Educational grant– Board of ● CV Therapeutics ● AstraZeneca AstraZeneca, Trustees ● Berlex CV Therapeutics, Laboratories GlaxoSmithKline, ● Pfizer King Pharmaceuticals, Monarch Pharmaceuticals, Pfizer, Sanofi-Aventis, Schering-Plough, Wyeth-Ayerst Laboratories

Dr. Albert P. Rocchini ● Ofﬁcial–AHA None None None None None None None

Dr. Samuel J. Shubrooks ● Ofﬁcial–ACC None None None None None None None Board of Governors

Dr. Alan Yeung ● Official–AHA ● Medtronic ● Abbott ● Abbott None None ● Boston None ● Boston ● Boston Scientific Scientific Scientific ● Cordis

Dr. Gregory Dehmer ● Organizational– None None None None None None None Society for Cardiovascular Angiography and Interventions

Dr. John Hodgson ● Organizational– ● Volcano Corp ● Boston ● Volcano Corp ● GE Medical None ● Technology ● Management– Society for Scientiﬁc ● Pﬁzer Solutions Mytogen Cardiovascular ● GE Medical Group Technology Angiography ● Lilly ● BioInfo Solutions Group and ● RADI Medical Accelerator ● BioInfo Accelerator Interventions ● Volcano Corp Fund Fund ● Volcano Corp

Dr. Morton J. Kern ● Organizational– ● Bracco Inc. None None ● RADI None None None Society for ● Meritt Medical Medical Cardiovascular ● Therox Inc. Angiography and Interventions

Dr. Ronald Krone ● Organizational– None None None None None None None Society for Cardiovascular Angiography and Interventions

Dr. Douglass A. Morrison ● Organizational– None None None None None None None Society for Cardiovascular Angiography and Interventions

Dr. Mark Reisman ● Organizational– None None ● Abbott ● Boston None None None Society for ● Boston Scientiﬁc Cardiovascular Scientiﬁc ● Cordis Angiography ● Cordis and ● Medtronic Interventions

Dr. Barry Uretsky ● Organizational– None None None None None None None Society for Cardiovascular Angiography and Interventions

Dr. Mazen Abu-Fadel ● Content–ACCF None None None None None None None Cardiac Catheterization and Intervention Committee

Dr. Peter Berger ● Content– ● Boston ● Cardiokinetix ● Arginox None None ● Lumen, Inc None Individual Scientific ● Conor ● Bristol-Myers Reviewer ● Cordis/Johnson ● Cordis/Johnson Squibb & Johnson & Johnson ● Sanofi- ● Genentech ● Datascope Aventis ● Guilford ● Guilford ● Schering- ● Lilly Plough ● The Medicine Company ● Medtronic ● Sankyo ● Sanofi-Aventis Continued on next page King et al ACCF/AHA/SCAI Clinical Competence Statement 123

APPENDIX 2. Continued Scientiﬁc Advisory Speakers’ Steering Stock Name Representation Consultant Research Grant Board Bureau Committee Holder Other

Dr. Robert O. Bonow ● Content–PCI None None None None None None None Guideline Writing Committee

Dr. Jose G. Diez ● Content–ACCF None None None None None None None Cardiac Catheterization and Intervention Committee

Dr. Ted E. Feldman ● Content–ACCF ● Boston ● Abbott None None None None None Cardiac Scientiﬁc ● Atritech Catheterization ● Cardiac ● Boston and Dimensions Scientiﬁc Intervention ● Cordis ● Cardiac Committee ● Myocor Dimensions ● Cordis ● Evalve

Dr. James Ferguson ● Content– None None None None None None None Individual Reviewer

Dr. Tommaso Gori ● Content–AHA None None None None None None None Diagnostic & Interventional Cardiac Catherization Committee

Dr. Hani Jneid ● Content–AHA None Pﬁzer None None None None None Diagnostic & Interventional Cardiac Catheterization Committee

Dr. Fred M. Krainin ● Content–ACCF None None None None None ● Boston None Cardiac Scientiﬁc Catheterization ● Johnson & and Johnson Intervention ● Medtronic Committee

Dr. Glenn Levine ● Content–AHA None None None None None None None Diagnostic & Interventional Cardiac Catheterization Committee

Dr. Charanjit S. Rihal ● Content–ACCF None None None None None None None Cardiac Catheterization and Intervention Committee

Dr. Dan M. Roden ● Content– ● Abbott None None None None None None Individual ● Alza Reviewer ● Arpida ● AstraZeneca ● Bristol-Myers Squibb ● CV Therapeutics ● EBR Systems ● First Genetic Trust ● GlaxoSmithKline ● Genzyme ● Johnson & Johnson ● Lexicon ● Lundbeck ● Medtronic ● Merck ● NPS Pharmaceuticals ● Novartis ● Pﬁzer ● Sanoﬁ- Synthelabo Groupe ● Solvay ● Thornton Medical ● Wyeth ● Yamanouchi

Dr. Carlos Ruiz ● Content–ACCF None None None None None None None Cardiac Catheterization and Intervention Committee

Dr. Michael J. Silka ● Content– None None None None None None ● General Electric Individual Reviewer Continued on next page 124 Circulation July 3, 2007

APPENDIX 2. Continued Scientiﬁc Advisory Speakers’ Steering Stock Name Representation Consultant Research Grant Board Bureau Committee Holder Other

Dr. Thoralf M. Sundt ● Content–ACCF None None None None None ● Medtronic None Cardiac (son has Catheterization stock) and Intervention Committee

Dr. Cynthia M. Tracy ● Content– None ● Guidant Corp None None None None None Individual ● Medtronic Reviewer

Dr. E. Murat Tuzcu ● Content–AHA None None None None None None None Diagnostic & Interventional Cardiac Catheterization Committee

Dr. Matthew Wolff ● Content–ACCF None None None None None None None Cardiac Catheterization and Intervention Committee

Dr. Yerem Yeghazarans ● Content– None None None ● Pﬁzer None None None Individual ● Sanoﬁ- Reviewer Aventis

This table represents the relationships of peer reviewers with industry that they reported as relevant to this topic. It does not necessarily reﬂect relationships with industry at the time of publication. Participation in the peer review process does not imply endorsement of the document. Names are listed in alphabetical order within each category of review. An Unusual Site for a Common Disease

Maysaa Alzetani, MRCP, MSc; Joseph J. Boyle, MRCP, PhD; David Lefroy, MA, MB, BChir, FRCP; Petros Nihoyannopoulos, MD, FRCP

75-year-old Asian woman presented with a 5-month Isolated cardiac tuberculosis is extremely rare. However it Ahistory of night sweats, lethargy, and malaise. On should be included in the differential diagnosis of intracardiac admission she was found to have low-grade pyrexia and masses. elevated inflammatory markers. Septic screen, which included repeated blood cultures and chest x-ray (Figure 1), were negative. A transthoracic and transesophageal echocardiography revealed a doughnut-shaped mass (online Data Sources of Funding Supplement Movie I) that surrounded the mitral valve annu- Dr Boyle has received research support from the British Heart lus and extended up to the left atrial walls and atrial septum. Foundation, KRUK, HHRTC, and the Broad Foundation. A surgical biopsy was taken (Figure 2), which showed epithelioid and langhans giant cell granulomas with central caseating necrosis consistent with tuberculosis. Special staining with high-sensitivity immunoperoxidase confirmed the Disclosures diagnosis of tuberculosis. The patient was treated for tuber- Dr Boyle has received honoraria from the International Journal of Experimental Pathology, has served as a speaker for the Histochemical culosis with complete resolution of her symptoms. Repeated Society and the British Atherosclerosis Society as a member of the echocardiography 6 months later showed a dramatic reduc- editorial board for the International Journal of Experimental Pathology, tion of the mass size (Data Supplement Movies II and III). and as an expert witness to UK coroners and mesothelioma panels.

Figure 1. Posterior-anterior chest x-ray, taken on admission, Figure 2. Biopsy taken from the mass that surrounded the shows clear lung fields and no signs of infection. mitral valve annulus and extended up to the left atrial walls and atrial septum. Hematoxylin and eosin staining. Magnifica- tion, ϫ20. Ep indicates epithelioid macrophages; L, Langhans giant cells; and N, necrosis. Inset, immunohistochemistry with monoclonal antibody to mycobacterium (Dako, Glostrup, Den- mark) and immunoperoxidase (Menarini Diagnostic, Woking- ham, UK). B indicates bacillus; M, macrophage. Magnification, ϫ100 (cropped for space).

From the Hammersmith Hospital NHS Trust (M.A.), and the Histopathology Department (J.J.B.) and Hammersmith Hospital (D.L., P.N.), Imperial College, London, UK. The online-only Data Supplement, consisting of movies, is available with this article at http://circ.ahajournals.org/cgi/content/full/116/1/e1/DC1. Correspondence to Dr Maysaa Alzetani, 42 Tryfan Close, Ilford, London IG4 5JY, United Kingdom. E-mail [email protected] (Circulation. 2007;116:e1.) © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.677120

Sine-Wave Pattern Arrhythmia and Sudden Paralysis That Result From Severe Hyperkalemia

Maurice J.H.M. Pluijmen, MD; Ferry M.R.J. Hersbach, MD

54-year-old man with a history of end-stage renal pattern.1 Progression of hyperkalemia causes further widening of Adisease treated with hemodialysis presented to the emer- the QRS complex, often with the configuration of a left or right gency department because of a sudden inability to move his bundle-branch block. Eventually, merger of QRS complex and T limbs. wave will lead to the appearance of a typical sine-wave pattern. Physical examination revealed a complete quadriplegia. Contrary to our patient, a sine-wave pattern often precedes Blood pressure was 145/90 mm Hg, with a regular pulse of 45 ventricular fibrillation or asystole.2 Furthermore, rapidly pro- beats per minute. Initial 12-lead electrocardiography showed gressing flaccid motor weakness may result in a quadriplegia, a sinus bradycardia with atrial, atrioventricular, and intraven- which was the presenting symptom in this case and is an tricular conduction delay (Figure 1). Subsequently, the pa- uncommon manifestation of severe hyperkalemia that may tient developed several episodes of a nonsustained wide- ultimately result in respiratory failure.2,3 complex tachycardia with a right bundle-branch block Recognition of the combination of sudden paralysis and configuration that gradually evolved into and from a “sine- electrocardiographic abnormalities as demonstrated in this wave” pattern (Figures 2 and 3). Remarkably, the patient case can lead to early diagnosis and treatment of severe remained hemodynamically stable. Because hyperkalemia hyperkalemia. was suspected, serum potassium was determined in venous as well as arterial blood samples to be 9.9 mmol/L. Calcium gluconate was administered, and emergency hemodialysis was performed. After normalization of the serum potassium Disclosures concentration, sinus rhythm was maintained, the cardiac None. conduction times returned to normal (Figure 4), and the quadriplegia resolved completely. This case illustrates some typical features of severe hyper- References kalemia. Initial characteristic electrocardiographic abnormal- 1. Sims DB, Sperling LS. Images in cardiovascular medicine. ST-segment elevation resulting from hyperkalemia. Circulation. 2005;111: ities in hyperkalemia are tall and peaked T waves, followed e295–e296. by an increasing cardiac conduction delay. As demonstrated 2. 2005 American Heart Association Guidelines for Cardiopulmonary in this case, this results in flattened and broadened P waves, Resuscitation and Emergency Cardiovascular Care, Part 10.1. Life- an atrioventricular block of first or higher degrees, and threatening electrolyte abnormalities. Circulation. 2005;112: IV121–IV125. widening of the QRS complex. In rare instances, ST-segment 3. Freeman SJ, Fale AD. Muscular paralysis and ventilatory failure caused elevation may occur, which leads to a “pseudoinfarction” by hyperkalaemia. Br J Anaesth. 1993;70:226–227.

From the Department of Cardiology, Medisch Centrum Rijnmond-Zuid, Rotterdam, the Netherlands. Correspondence to Dr Maurice J.H.M. Pluijmen, Groene Hilledijk 315, 3075EA Rotterdam, The Netherlands. E-mail [email protected] (Circulation. 2007;116:e2-e4.) © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.687202

Pluijmen and Hersbach Severe Hyperkalemia e3

Figure 1. The initial 12-lead ECG (25 mm/sec, 5 mm/mV) obtained on presentation to the emergency department demonstrates a sinus bradycardia with prolonged atrial conduction (ﬂattened and broadened P waves, see arrows), a ﬁrst-degree atrioventricular block (PQ, 300 ms), and an intraventricular conduction delay (QRS, 160 ms).

Figure 2. Rhythm strip recording of lead II, with a wide-complex tachycardia with right bundle-branch block conﬁguration that gradually evolves into a sine-wave pattern.

e4 Circulation July 3, 2007

Figure 3. This 12-lead ECG (25 mm/sec, 10 mm/mV) of the wide-complex tachycardia (QRS, 240 ms) demonstrates the right bundle- branch block conﬁguration that evolves into and from a sine-wave pattern.

Figure 4. The 12-lead ECG (25 mm/sec, 5 mm/mV) after normalization of the serum potassium concentration reveals the normalization of atrial, atrioventricular (PQ, 140 ms), and intraventricular (QRS, 80 ms) conduction times, as well as a preexistent left ventricular hypertrophy pattern. Images in Cardiovascular Medicine

Lipomatous Metaplasia in Ischemic Cardiomyopathy A Common but Unappreciated Entity

Matthias Schmitt, MD, PhD, MRCP; Nilesh Samani, BSc, MB, ChB, MD, FRCP, FmedSci; Gerry McCann, BSc, MB, ChB, MD, MRCP

68-year-old man with a 14 1year prior history of anterior idiopathic dilated cardiomyopathy, or chronic valvulopathy, Amyocardial infarction was referred for viability assess- respectively.1 Myocardial perfusion imaging (lipohilic myo- ment. He had established 3-vessel coronary artery disease cardial perfusion agents such as tetrofosmin and sestamibi with a proximally occluded left anterior descending coronary may well be taken up into fat cells) and echocardiography fail artery. He complained of worsening shortness of breath and to diagnose LM (which accounts for the lack of recognition of diminishing exercise tolerance. Cardiac magnetic resonance this not uncommon entity) with the consequent implications imaging demonstrated a nondilated left ventricle with minor for overestimation of viable myocardium or underestimation aneurysmal transformation that affected the mid-anterior wall of scar size. and extended into mid- and antero-septum as well as the apex. Importantly, LM on magnetic resonance imaging must not Gradient (Figure, A) and T1-weighted spin-echo images be mistaken for late enhancement after gadolinium contrast. demonstrated bright signal intensity in the mid-myocardium of the anterior wall (see arrows in Figure, B), which disap- peared with fat saturation (see arrows in Figure, C and D). These findings are indicative of lipomatous metaplasia (LM). Disclosures The term LM describes fat that is present in and seemingly None. replaces scar tissue in the myocardium. The exact etiology of LM is unknown, but it is not seen in the absence of substitutive myocardial fibrosis. Histological evidence of LM has been found in up to 68%, 24%, and 37% of areas of left References 1. Baroldi G, Silver MD, De Maria R, Parodi O, Pellegrini A. Lipo- ventricular myocardial scars in explanted hearts of patients matous metaplasia in left ventricular scar. Can J Cardiol. 1997;13: who underwent transplantation for ischemic heart disease, 65–71.

From the Department of Cardiology, Glenfield Regional Cardiac Centre, Leicester University Hospital Trust, Leicester, UK. The online-only Data Supplement, consisting of movies, is available with this article at http://circ.ahajournals.org/cgi/content/full/116/1/e5/DC1. Correspondence to Dr Matthias Schmitt, Department of Cardiology, Glenfield Regional Cardiac Centre, Leicester University Hospital Trust, Groby Road, Leicester LE3 9QP, UK. E-mail [email protected] (Circulation. 2007;116:e5-e6.) © 2007 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.107.690800 e5 e6 Circulation July 3, 2007

Short-axis images at mid-cavity level (A and C) and long-axis images (B and D). White arrows mark high signal intensity (no gadolinium given) in the anterior wall (A and B) that turns black in corresponding fat suppression sequence (C and D).

Correspondence

Letter by Brewster and van Montfrans Disclosures Regarding Article, “Risks Associated With None. Statin Therapy: A Systematic Overview of Lizzy M. Brewster, MD Randomized Clinical Trials” Gert A. van Montfrans, MD To the Editor: Departments of Internal and Vascular Medicine Kashani et al1 report that in randomized trials, statin Academic Medical Center F4-222 therapy (with the exclusion of cerivastatin) did not result in University of Amsterdam significant absolute increases in myalgias (risk difference per Amsterdam, The Netherlands 1000 patients, 2.7 (95% CI, Ϫ3.2 to 8.7), or mild creatine kinase (CK) elevations 0.2 (95% CI, Ϫ0.6 to 0.9). The authors 1. Kashani A, Phillips CO, Foody JM, Wang Y, Mangalmurti S, Ko DT, used data from published trials to reach this conclusion. Krumholz HM. Risks associated with statin therapy: a systematic overview However, Kashani et al did not take into account that many of randomized clinical trials. Circulation. 2006;114:2788–2797. 2. LaRosa JC, Grundy SM, Waters DD, Shear C, Barter P, Fruchart JC, Gotto statin trials disclose that eligible patients with muscle complaints, AM, Greten H, Kastelein JJ, Shepherd J, Wenger NK; Treating to New previous adverse responses to cholesterol-lowering therapy, or Targets (TNT) Investigators. Intensive lipid lowering with atorvastatin in Ͼ even a mild asymptomatic elevation of CK ( 1.5 to 6.0 times the patients with stable coronary disease. N Engl J Med. 2005;352:1425–1435. upper limit of normal) are not randomized.2–5 3. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection On the basis of the complete exclusion criteria, up to 76% of Study of cholesterol lowering with simvastatin in 20,536 high-risk indi- the screened participants in statin trials are not randomized and viduals: a randomised placebo-controlled trial. Lancet. 2002;360:7–22. excluded.2–5 Thus, the incidence of hyperCKemia with the use of 4. Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe SM, statins that emerges from these trials may mainly concern Ford I, Gaw A, Hyland M, Jukema JW, Kamper AM, Macfarlane PW, subjects in the lower part of the CK distribution, with a low Meinders AE, Norrie J, Packard CJ, Perry IJ, Stott DJ, Sweeney BJ, a priori risk to develop highly elevated CK levels. Twomey C, Westendorp RG; PROSPER Study Group. Pravastatin in elderly individuals at risk of vascular disease (PROSPER): a randomised The exclusion of these patients before randomization may lead controlled trial. Lancet. 2002;360:1623–1630. to biased reports on the frequency of occurrence of side effects 5. Shepherd J, Cobbe SM, Ford I, Isles CG, Lorimer AR, MacFarlane PW, with statin use. This should be acknowledged to be a limitation McKillop JH, Packard CJ. Prevention of coronary heart disease with when adverse effects associated with statins are assessed in pravastatin in men with hypercholesterolemia. West of Scotland Coronary published trials. Prevention Study Group. N Engl J Med. 1995;333:1301–1307.

Letter by Rosenberg and Uretsky Regarding Article, drug could still account for myalgias in Ͼ250 000 patients— “Risks Associated With Statin Therapy: A not quite an insignificant number! Systematic Overview of Randomized Clinical Trials” To the Editor: In regard to the article by Kashani et al,1 the authors Disclosures conclude that “...statin therapy is associated with small excess None. risk of transminase elevations, but not of myalgias, creatine Lauren Rosenberg, MD kinase elevations, rhabdomyolysis...”. We think this conclu- Seth Uretsky, MD sion is misleading. On closer inspection of the data, atorva- St Luke’s-Roosevelt Medical Center statin has 3 times the occurrence of myalgias compared with Columbia University College of Physicians and Surgeons Ͻ placebo (P 0.04). According to the authors’ data, these New York, NY adverse events seem to be unique to atorvastin and were not observed with other statins. Admittedly the absolute risk is 1. Kashani A, Phillips CO, Foody JM, Wang Y, Mangalmurti S, Ko DT, Krumholz small; however, when one considers that in the US alone Ϸ5 HM. Risks associated with statin therapy: a systematic overview of randomized million patients are presently treated with atorvastatin, this clinical trials. Circulation. 2006;114:2788–2797.

Response to Letters Regarding Article, “Risks Amir Kashani, MS, MD Associated With Statin Therapy: A Systematic JoAnne M. Foody, MD Overview of Randomized Clinical Trials” Yongfei Wang, MS Harlan M. Krumholz, MD, SM The letters that address our analysis of risks associated with statin Section of Cardiovascular Medicine 1 therapy raise 2 important points: the generalizability of the data and Department of Medicine the excess incidence of myalgias observed with atorvastatin therapy. Yale University School of Medicine Although inclusion/exclusion criteria required for clinical trials limit New Haven, Conn generalizability to patients in clinical practice, randomized controlled trials remain the best unbiased source of data to assess Christopher O. Phillips, MD, MPH adverse effects.2,3 However, there remains a need for additional, Department of General Internal Medicine large, safety studies in populations previously not studied. The Cleveland Clinic Foundation In their letter, Drs Brewster and van Montfrans indicate that Cleveland, Ohio some statin trials exclude patients with elevated creatine kinase Sandeep Mangalmurti, MD levels (Ͼ1.5 to 6 times the upper limit of normal). We believe Ambulatory Health Clinic that these patients are appropriately excluded, as patients with United States Navy extreme creatine kinase elevations have an underlying pathology Groton, Conn and may represent a population inappropriate for statin therapy. The issue of excess incidence of myalgia observed with Dennis T. Ko, MD atorvastatin, raised in the letter from Drs Rosenberg and Uretsky, Schlich Heart Centre merits further investigation. This observation reached marginal ϭ Sunnybrook Health Sciences Centre statistical significance (P 0.04) and was based on only 567 Ontario, Canada patients (19 of 375 patients versus 3 of 192 patients in the treatment and placebo groups, respectively). Accordingly, this finding is worth further pursuit, but should not be considered 1. Kashani A, Phillips CO, Foody JM, Wang Y, Mangalmurti S, Ko DT, definitive at this time. Krumholz HM. Risks associated with statin therapy: a systematic overview of randomized clinical trials. Circulation. 2006;114: 2788–2797. 2. Curtin F, Altman DG, Elbourne D. Meta-analysis combining parallel Disclosures and cross-over clinical trials. I: Continuous outcomes. Stat Med. 2002;21:2131–2144. Dr Foody received honoraria from and served as a consultant/ 3. Curtin F, Elbourne D, Altman DG. Meta-analysis combining parallel and advisory board member for Merck, BMS/Sanofi, and Pfizer. The cross-over clinical trials. III: The issue of carry-over. Stat Med. other authors have nothing to disclose. 2002;21:2161–2173.

The editors express appreciation to the following referees who served from July 1, 2006, though December 31, 2006.

Einari Aavik Michael H. Alderman Toshihisa Anzai Peter Backx Inmaculada B. Aban Salvatore Alesci Caroline M. Apovian Larry Baddour Nicola Abate Anthony Aletras Christina L. Aquilante Michael Bader Amr E. Abbas Barbara Alexander Andrew E. Arai Nitish Badhwar Antonio Abbate M. Yvonne Alexander Jack L. Arbiser Juan Jose Badimon Jinnette Dawn Abbott Mark E. Alexander Abbas Ardehali Emilio Badoer Mohamed Abdel-Wahab R. Wayne Alexander Ross Arena Hala M. Badran Hazem Abdul Hussien Andrei V. Alexandrov Miguel A. Arias Stephen F. Badylak E. Dale Abel Ayyaz A. Ali Robert A. Ariens Man Jong Baek Jamil A. AboulHosn Ziad A. Ali Chiara Armani Emilia Bagiella M. Roselle Abraham Kari Alitalo Gary C. Armitage Steven Bailey Nader G. Abraham Lindsey D. Allan Ehrin Johnson Armstrong Jean-Patrice Baillargeon Theodore P. Abraham Larry Alexander Allen Paul W. Armstrong Donald S. Baim Hugues Abriel Paul D. Allen Donna K. Arnett Alison E. Baird Elias Abrutyn Maurits A. Allessie Doron Aronson Andrew H. Baker Stephan Achenbach Matthew A. Allison Rishi Arora George L. Bakris Paul E. Achouh Thomas G. Allison Rohit Arora Stephan Baldus Michael Acker Carlos Alonso-Villaverde Umesh Arora H. Scott Baldwin Michael J. Ackerman Nicholas Alp Hiroshi Asanuma Christie M. Ballantyne Volker Adams Bahaaldin Alsoufi Raimondo Ascione Jean-Luc Balligand Philip A. Ades Peter A. Altman Muhammad Ashraf Scott W. Ballinger Srilakshmi M. Adhyapak Salomon Amar Samuel J. Asirvatham Andriy Bandos Gail K. Adler Pierre Amarenco Peter Aspelin Eddy Barasch Vahid Afshar-Kharghan John A. Ambrose Gerd Assmann John C. Barbato Salvatore Agati Giuseppe Ambrosio Birgit Assmus Silvia Stella Barbieri Piergiuseppe Agostoni Peter Ammann Nimer N. Assy John C. Barefoot Eustachio Agricola Enrico Ammirati Carmela Asteria Amber E. Barnato Pietro Maria G. Agricola Ezra Amsterdam Anne Sofie Astrup Adrian G. Barnett David Aguilar Dhakshinamurthy Vijay Anand Bela F. Asztalos Francesco Barone-Adesi Seyedhossein Aharinejad Inder S. Anand Dianne L. Atkins Jose A. Barrabes Ferhaan Ahmad Sonia S. Anand John P. Atkinson Robyn J. Barst Ali Ahmed Nagesh Sadanand Anavekar Pavan Atluri Philip Barter Alsir A.M. Ahmed Ulla Overgaard Andersen Douwe E. Atsma John R. Bartholomew Ismayil Ahmet Garnet L. Anderson Ha˚vard Attramadal Matthias Barton Youngkeun Ahn H. Vernon Anderson Andrew M. Atz Peter Ba¨rtsch Seema S. Ahuja Jeffrey L. Anderson Angelo Auricchio Riyaz Bashir Diana Aicher Kelley P. Anderson Gerard P. Aurigemma Thomas M. Bashore Elena Aikawa Mark E. Anderson Michael V. Autieri Theodore A. Bass Masanori Aikawa Peter G. Anderson Pablo Avanzas Eric R. Bates Ryuichi Aikawa Robert H. Anderson Maurizio R. Averna Philip M.W. Bath Judith A. Airey Todd J. Anderson Antonio Aversa D. Battle Anthony Aizer Stefan Andreas Nili Avidan Paulo Fernando Dotto Bau Adesuyi A. Ajayi Arne Kristian Andreassen Abraham Aviv Robert Bauernschmitt Nadine Ajzenberg Maria Grazia Andreassi Angelo Avogaro Johann Bauersachs Tankut Hakki Akay Felicita Andreotti Alberto Avolio Kenneth Lee Baughman Shahab A. Akhter Douglas Andres Ola A. Awad Jan H. Baumert Olakunle O. Akinboboye Dominick J. Angiolillo Vitor M.P. Azevedo Daniela Baumgartner Masahiro Akishita Stefan D. Anker Helmut Baumgartner Amin Al-Ahmad Brian H. Annex Gerard Babatasi Iris Baumgartner Dagmar G. Alber Benjamin Ansell Fritz H. Bach Ralf W. Baumgartner Alexander Albert Jack Ansell Richard G. Bach William Baumgartner Christine M. Albert Amedeo Anselmi Emile Bacha Jeroen J. Bax Jeffrey Albert David Antoniucci Stephen L. Bacharach Jeroen Bax Michelle Albert Jovan P. Antovic Robert J. Bache B. Timothy Baxter George Alberti Charles Antzelevitch Jean E. Bachet Gary F. Baxter Gabriel Aldea Piero Anversa Markus Michael Bachschmid Antoni Bayés de Luna Yvette B.J. Aldenhoff Ani Anyanwu Carl Backer W. Scott Beattie e10 Acknowledgment of Reviewers e11

James Beck Joyce Bischoff Mark R. Boyett Denis Buxton Diane M. Becker Nanette H. Bishopric Andrew J. Boyle Graham B. Byrnes Edmund R. Becker Gianluigi Bisleri E. H. Bradley Richard C. Becker John A. Bittl T. Douglas Bradley Candido Cabo Joshua A. Beckman Vera Bittner Susan Brain Howard Cabral Thomas R. Behrenbeck Martin Björck Randy W. Braith Evren Caglayan Vidar Beisvag Henry W. Blackburn Stefan-Martin Brand-Herrmann Michael E. Cain Bernard Belhassen Eugene H. Blackstone Ruediger C. Braun-Dullaeus Francesc Calafell Jonathan N. Bella Stefan Blankenberg Eugene Braunwald Antonio Maria Calafiore Carsten J. Beller James C. Blankenship Alan C. Braverman A. Louise Calder George A. Beller Andrew D. Blann George A. Bray James H. Caldwell Marek Belohlavek Sabine Bleiziffer Bernhard R. Brehm Mary Caldwell Lotfi Ben Mime Rudiger Blindt Ole A. Breithardt David A. Calhoun David G. Benditt Wilhelm Bloch Sorin J. Brener Robert M. Califf Umberto Benedetto Peter C. Block Robert T. Brennan Hugh Calkins Federico Jose Benetti Hanna Bloomfield Hermann Brenner Ronald Callahan David J. Callans Yanai Ben-Gal David A. Bluemke Kate M. Brett Francois A. Cambien Frank M. Bengel Elizabeth D. Blume Sally E. Brett Vicky A. Cameron Ivor J. Benjamin Roger S. Blumenthal Martin Breuer Paolo G. Camici Ralf Benndorf Yuri V. Bobryshev Gregorio Brevetti A. John Camm Joel S. Bennett Jorge B. Boczkowski Charles R. Bridges Duncan J. Campbell D. Woodrow Benson Christoph Bode David Brieger Julie H. Campbell Robert A. Berg William E. Boden Francoise Briet Umberto Campia Alan K. Berger Johannes Boehm Ju¨rgen Brinckmann Ann E. Canfield Martin W. Bergmann Matthijs Boekholdt Michael R. Bristow Christopher P. Cannon Zekarias Berhane Cornelis Boersma Susanne Bro Richard O. Cannon Bradford C. Berk Eric Boersma Craig Broberg John G. Canto Daniel S. Berman Daniel J. Boffa Otto-Erich Brodde Noel M. Caplice Jose M. Bernal Rainer H. Boger Sergey V. Brodsky Maurizo C. Capogrossi Sheilah A. Bernard Frank Bogun Ulrich Broeckel Thomas P. Cappola Luciano Bernardi Barry Anthony Boilson Robert D. Brook Massimo Caputo Michael C. Berndt William Boisvert Maria Mori Brooks Blase A. Carabello Ann Bolger James M. Brophy Susannah Bernheim Christopher M. Carlin Daniel Bernstein Roberto Bolli Meg E. Brousseau Mark David Carlson Stuart S. Berr Marvin O. Boluyt Ingeborg A. Brouwer Robert M. Carney Donald M. Bers Massimo Bonacchi Gregory L. Brower John Alfred Carr Giuseppe S. Berton Nikolaos Bonaros B. Greg Brown Thierry P. Carrel Alain G. Bertoni Mark Bond Clive M. Brown Oscar A. Carretero Gabriele B. Bertoni Enzo Bonora David A. Brown Alain Carrie Joline W.J. Beulens Robert O. Bonow Jeremiah R. Brown Michel Carrier Steve Bevan Munir Boodhwani Nancy J. Brown John D. Carroll Friedhelm Beyersdorf George W. Booz Martina Brueckmann Andrew J. Carter Connie R. Bezzina Jeffrey S. Borer Martina Brueckner Angela M. Carter Dwaipayan Bharadwaj Michael A. Borger Ramon Brugada Raymond Cartier Aruni Bhatnagar Susanna Etje Borggreve Eric J. Brunner Barbara Casadei Deepak Bhatnagar Claudio Borghi Chris L. Bryson Juan Pablo Casas Deepak L. Bhatt Christina A. Boros Michael R. Buchanan Ivan Pearse Casserly Jinsong Bian Dirk Böse V.M. Buckalew Daniel F. Catanzaro Giuseppe Bianchi Kritina Bostrom Matthew J. Budoff Martha Cathcart Giorgio M. Biasi René M. Botnar Brian Buijsse Christophe Caussin Luigi Marzio Biasucci Michiel L. Bots L. Maximilian Buja Erdal Cavusoglu Kirsten Bibbins-Domingo Morten Bottcher Harry R. Buller Serghei Cebotari J. Thomas Bigger Elias H. Botvinick T. Jared Bunch Frank Cecchin Diane E. Bild Tarek Bouhali Allen P. Burke David S. Celermajer George E. Billman Anne Bouloumie Gregory L. Burke Antonio Ceriello Ofer Binah Martial G. Bourassa John C. Burnett Matteo Cesari Christian Binggeli Pascal Bousquet Michel Burnier Juan C. Chachques Andreia Biolo Edward L. Bove Ivo Buschmann Claudia U. Chae Rainer Birck Thierry Bove Sebastian J. Buss Bernard R. Chaitman Yochai Birnbaum Frieke Box Rudi Busse Subrata Chakrabarti Jonathan Birns Penelope A. Boyden Peter M. Buttrick Aravinda Chakravarti e12 Acknowledgment of Reviewers

Lorraine Chalifour Leslie L. Clark Michael H. Ciqui James A. de Lemos John Chalmers Robert Clarke Julia Alison Critchley Moniek P.M. de Maat Hunter Clay Champion Matthias A. Clauss Bernard Lewis Croal Ramon de Nooijer Wing Bun Chan John G.F. Cleland CarrollE. Cross Anne-Cornelie J.M. de Pont Krishnaswamy Chandrasekaran Denis L. Clement Luigi X. Cubeddu Albert de Roos Y. Chandrashekhar Richard T. Clements Bruce F. Culleton Giovanni de Simone Chih-Jen Chang Ton J. Cleophas William C. Culp Bianca L. De Stavola Joyce Chung-Chou Ho Chang Jack P.M. Cleutjens Anne B. Curtis Pieter P. de Tombe Ruey-Kang R. Chang Roger L. Click Ricardo C. Cury Roberto De Vogli Keith M. Channon William T. Clusin Mary Cushman Leon J. De Windt Jean-Pierre Chanoine Michael A. Coady Daniele M. Cusi Robbert J. de Winter Miguel Chaput Christopher S. Coffey Jeffrey A. Cutler Dick de Zeeuw Marietta Charakida David J. Cohen Donald E. Cutlip Barbara J. Deal Fadi Joseph Charchar Hillel W. Cohen John Eric Deanfield Israel F. Charo Meryl S. Cohen Michael W. Dae Arjun Deb John C. Chatham Michael V. Cohen Mat J.A.P. Daemen G. William Dec Michael Daffertshofer Kanu Chatterjee Richard A. Cohen Jeanne M. DeCara Marinos C. Dalakas Sarwat Chaudhry Smadar Cohen Carole J. Decker Darshan Dalal Irshad H. Chaudry Stanley N. Cohen Ulrich K.M. Decking James E. Dalen Vijay S. Chauhan Jay N. Cohn Prakash C. Deedwania Ronald L. Dalman Melvin D. Cheitlin Lawrence H. Cohn Carolin Deiner Caroline A. Daly Alex F. Chen William E. Cohn Etienne Delacre´taz Jan Kristian Damas Chunguang Chen Duncan Robert Coles Brian P. Delisle Dorte Damgaard Frederick Y.Y. Chen Andrea Colli Mario Delmar Michael Dandel Hua Yun Chen Alan R. Collins Anthony N. DeMaria George Dangas Ian Y. Chen Robert W. Colman Linda L. Demer Jean-Marie Daniel Lamaziere Jiu-Chiuan Chen Antonio Colombo Nikolaos Demiris Stephen R. Daniels Ju Chen Catherine Communal Martin den Heijer A.H. Jan Danser Kai Chen Gianluigi Condorelli Mario C. Deng Lara Danziger-Isakov Peng-Sheng Chen Paul R. Conlin David DeNofrio Jeanine M. D’Armiento Shih-Ann Chen John E. Connett Christophe Depre Dipak K. Das Yiu-Fai Chen Agostino Consoli Akshay S. Desai Sandeep Das Milind Desai Bin Cheng Nancy R. Cook Kaberi Dasgupta Debbie Cheng Stephen R. Cook James Daubert Anne M. Deschamps Tsung O. Cheng Thomas Cook Jean Claude Daubert Christopher A. DeSouza Yu Cheng John P. Cooke Siamak Davani Jean-Pierre Despres Glenn M. Chertow Joshua M. Cooper Anthony P. Davenport Alexander Deten Adrian H. Chester Leslie T. Cooper Tirone E. David Christian Detter Bernard M.Y. Cheung Mark E. Cooper Charles J. Davidson Marcelo F. Di Carli Mordechai Chevion Rhonda M. Cooper-DeHoff Karina W. Davidson Michele Di Mauro Shu Chien James Coromilas Michael H. Davidson Nuno V. Dias John S. Child Domenico Corrado Peter F. Davies David A. Dichek William M. Chilian Egle Corrado Victor G. Davila-Roman Hans Dieplinger Michael T. Chin James P. Corsetti Bryce H. Davis D.B. Diercks Randolph W. Chitwood-Jr Roberto Corti Roger B. Davis Rodney J. Dilley Ray C.J. Chiu Francesco Cosentino Robin L. Davisson Wolfgang H. Dillmann Aram V. Chobanian Marco A. Costa Kevin P. Davy Vasken Dilsizian Yeon Hyeon Choe Salvatore Costa Buddhadeb Dawn J. Michael DiMaio Michel B. Chonchol Nathalie Costedoat-Chalumeau Jeffrey D. Dawson John P. DiMarco Magdalena Chottova Dvorakova William G. Cotts Jonathan R.S. Day Pawel Petkow Dimitrow Benjamin J.W. Chow Patrick Anthony Coughlin Mariza de Andrade Stefanie Dimmeler Phil Chowienczyk David Couper Dirk De Bacquer Robert A.E. Dion Torsten Christ Francis Couturaud Jacques de Bakker Donald J. DiPette Geir Christensen Linda D. Cowan Bernard De Bruyne Dobromir Dobrev Timothy F. Christian Allen W. Cowley Marco De Carlo Douglas W. Dockery Karkos D. Christos Dermot Cox Raffaele De Caterina Frances Dockery Sumeet S. Chugh Ciprian M. Crainiceanu Carlos R. De Diego Jean-Michel P.N. Dogné Moo K. Chung Sybil Crawford Sarah D. de Ferranti Anuja Dokras Juan Cinca Filippo Crea Pim J. de Feyter Raul J. Domenech Francesco Cipollone Mark A. Creager Dominique P. de Kleijn J. Kevin Donahue Marco Cirillo Harry Crijns Peter W. de Leeuw Rosario Donato

Acknowledgment of Reviewers e13

Chunming Dong Suzette Elias-Smale Zahi A. Fayad Thomas Joseph Forbes G.A. Donnan Uri Elkayam Sergio Fazio Thomas Force Marjo Donners Ronald C. Elkins William F. Fearon Ian Ford Vincent Dor Kenneth A. Ellenbogen Paul W.M. Fedak Daniel Forman Andrea Doria C. Gregory Elliott Martin Feelisch Myriam Fornage Paul Dorian Stephen G. Ellis Jeffrey A. Feinstein James S. Forrester Pedro D’orléans-Juste R Curtis Ellison Frederick Feit Tom Forsen Gerald W. Dorn Costanza Emanueli Arthur M. Feldman Elyse Foster Mirko Doss Michael Emerson David S. Feldman Jean-Claude Fouron John S. Douglas Mary Emond Ted Feldman Caroline S. Fox Pamela S. Douglas Jean-Philippe Empana Peter Ferdinandy Keith A.A. Fox James M. Downey Masao Endoh Maros Ferencik Harry A. Fozzard Luciano Ferreira Drager Matthias Endres James J. Ferguson Gabriele Fragasso Christopher Drake S. Engeli T. Bruce Ferguson Renerio Fraguas Mark H. Drazner Richard M. Engelman Francisco Fernandez-Aviles Alain Fraisse Helmut Drexler Marguerite M. Engler Olivier Feron Mark W. Frampton Daniel L. Dries Robert L. Engler Christiane Ferran Gary S. Francis Dayue Duan Mark L. Entman Carlos M. Ferrario Veronica Franco Naihua Duan Andrew Epstein Victor A. Ferraris John V. Frangioni Raghvendra K. Dubey Stephen E. Epstein Paolo Ferrazzi Nikolaos G. Frangogiannis Anique Ducharme Raimund R. Erbel Robert E. Ferrell John A. Frangos Stephen J. Duffy Sandra Erbs Julio J. Ferrer-Hita Barry A. Franklin Jean G. Dumesnil John M. Erikson Andreas Festa Charles Fraser Brian W. Duncan Urs K. Eriksson Loren J. Field Sohrab Fratz J. Michael Duncan Aycan F. Erkan David S. Fieno Gunilla Nordin Fredrikson Daniel Duprez Korhan Erkanli Hans R. Figulla David S. Freedman Jocelyn Dupuis David Erlinge Marcin Fijalkowski S. Ben Freedman Brent A. French William Durante Sabine Ernst Janos G. Filep Greg Dusting Georg Ertl John K. French Kristian B. Filion Lance D. Dworkin Nilda Gladys Espinola Michael P. Frenneaux Jeffrey R. Fineman Jason R. Dyck Christine Espinola-Klein Ulrich H. Frey Toren Finkel Peter Dyck Katherine Esposito Gary S. Friedman Joel Finkelstein Barry C. Esrig Matthias G. Friedrich Louis D. Fiore Kim A. Eagle Faadiel Essop Soren Friis Paolo Fiorina Mark J. Earley Mohammed Rafique Essop M. Kent Froberg Christian Firschke Cara A. East N.A. Mark Estes Victor Froelicher Peter Fischbach Robert T. Eberhardt Anthony L. Estrera Edward D. Frohlich Roman Fischbach Franz R. Eberli Susan P. Etheridge Jiri J. Frohlich Marcus Fischer Shah Ebrahim Charles J. Everett Peter C. Frommelt Uwe M. Fischer Dwain L. Eckberg Gordon A. Ewy Alexandra Theresia Fuchs Rodolphe Fischmeister Robert H. Eckel Vernat Exil Flavio Danni Fuchs Ce´line Fiset Elazer R. Edelman Michael D. Ezekowitz Robert C. Fuhlbrigge Michael C. Fishbein Sarah Eder Bianca Fuhrman Thomas S. Edgington Gianluca Faggioli Edward A. Fisher Yoshihiro Fukumoto Richard Edwards Colomba Falcone Marc Fisher John W. Funder Igor R. Efimov Pierre-Emmanuel Falcoz Patrick W. Fisher Colin D. Funk Kensuke Egashira Erling Falk Steven A. Fisher Curt D. Furberg Hannelore Ehrenreich Rodney H. Falk Garret A. FitzGerald George Fust Oliver Eickelberg James A. Fallavollita Peter J. Fitzgerald Valentin Fuster Andreas Eicken John T. Fallon Greg C. Flaker Benjamin W. Eidem James C. Fang Scott D. Flamm William H. Gaasch John F. Eidt James I. Fann Marcus D. Flather Alain-Pierre Gadeau John W. Eikelboom Søren Fanø Jerome L. Fleg James V. Gainer Michael Eikmans Frank Faraci Lee A. Fleisher Peter A. Gaines Graeme Eisenhofer Andrew Farb Ingrid Fleming Maurizio Galderisi David A. Eisner Jeronimo Farre Danilo Fliser Catharine R. Gale Tony Eissa Marc Fatar John S. Floras Zorina S. Galis Daniel T. Eitzman Farzin F. Fath-Ordoubadi Alan M. Fogelman William Gallo Ulf Ekelund Diane Fatkin Robert N. Foley Apoor S. Gami Garabed Eknoyan Rossella Fattori Warren Foltz Ronald Gangnon Amir Elami Desiderio Favarato Gregg C. Fonarow Peter Ganz John A. Elefteriades William P. Fay Guo-Hua Fong Feng Gao

e14 Acknowledgment of Reviewers

Susan M. Gapstur Stanton A. Glantz William J. Groh Robert A. Haworth Mario J. Garcia Ruchira Glaser Christian Grohe Kenshi Hayashi Julius M. Gardin Christopher Glass Garrett J. Gross Toshio Hayashi Timothy J. Gardner Stephen J. Glatt Gil J. Gross Michael R. Hayden Alistair N. Garratt David K. Glover Paul D. Grossfeld David L. Hayes Peter Garred Peter D. Gluckman Eugene A. Grossi Daniel Hayoz Daniel J. Garry Robert J. Glynn Blair P. Grubb Stanley L. Hazen Jean-Michel Théodule Gaspoz Alan S. Go Eberhard Grube Mary Fran Hazinski Michael A. Gatzoulis David C. Goff Scott M. Grundy Guo-Wei He Glenn R. Gaudette Noyan Gokce Gary L. Grunkemeier Jiang He Kimberlee Gauvreau Diane R. Gold William H. Guilford Geoffrey Head Haralambos P. Gavras Jeffrey P. Gold Christian Guilleminault John P. Headrick Irene Gavras Alexander Goldberg Gerard Marcel Guiraudon Susan R. Heckbert J. William Gaynor Andrew P. Goldberg Giosue Gulli Markus Hecker David C. Gaze Jeffrey J. Goldberger Paul A. Gurbel Timothy Heeren Thomas Gaziano Ilan Goldenberg Geoffrey C. Gurtner Christopher Heeschen Raul J. Gazmuri Joshua I. Goldhaber Björn I. Gustafsson Robert A. Hegele Carmine Gazzaruso Samuel Z. Goldhaber David D. Gutterman Paul A. Heidenreich Carolyn L. Geczy Lee Goldman Robert A. Guyton Jay W. Heinecke Bruce D. Gelb Martin E. Goldman Tomasz J. Guzik Christian Heiss Caroline Genco Steve Goldman Claes Held Constance Kay Haan Yong-Jian Geng James A. Goldstein Johanna Helmersson Judith Haendeler Thomas L. Gentles Larry B.Goldstein Harry Hemingway Steven M. Haffner Alfred L. George Sidney Goldstein Jeroen Hendrikse Dominik Georg Haider James F. George Paolo Golino Marc Hendrikx David E. Haines Sarah Jane George Jonathan Golledge Timothy D. Henry Michel Haissaguerre Bernhard L. Gerber Michael H. Gollob Moonseong Heo Katherine Hajjar Bernard J. Gersh Gershon Golomb Dirk Hermann Roger J. Hajjar Welton M. Gersony Ana M. Gomez Ramon C. Hermida Julian P.J. Halcox Gary Gerstenblith Celso E. Gomez-Sanchez Adrian F. Hernandez Charles A. Hales Edward P. Gerstenfeld Philimon Gona Victoria L.M. Herrera Andrew Halestrap Robert E. Gerszten Isabel Goncalves Amy Herring Michael E. Halkos David M. Herrington Leonard S. Gettes Mario D. Gonzalez Hermann Haller Godfrey S. Getz Mark O. Goodarzi Michael Hallman Howard C. Herrmann Tal Geva John Gorcsan Alfred P. Hallstrom Ray E. Hershberger Henry Gewirtz Tommaso Gori Naomi M. Hamburg Charles A. Herzog Michael Gewitz Agnes Gorlach Mohamed H. Hamdan Otto M. Hess Mihai Gheorghiade Joseph H. Gorman Christian W. Hamm Gerd Heusch Giorgio Ghilardi Shinya Goto Wayne W. Hancock Stephane Heymans Raymond J. Gibbons Roberta A. Gottlieb Aase Handberg William R. Hiatt C. Michael Gibson K. Lance Gould Anthony J. Hanley Charles B. Higgins Frank C. Gibson Andrew A. Grace Edward L. Hannan Robert Higgins Samuel S. Gidding Juan F. Granada Tim Hanson Thomas Hilberg Stephan Gielen Scott Grandy Goran K. Hansson Karl F. Hilgers Wayne R. Giles Christopher B. Granger Tomonori Haraguchi John S. Hill Linda D. Gillam Augustus O. Grant Violet I. Haraszthy Joseph A. Hill Brenda W. Gillespie Richard A. Gray Joshua M. Hare Hans L. Hillege A. Marc Gillinov William A. Gray Robert A. Harrington L. David Hillis Matthew W. Gillman Paul A. Grayburn William S. Harris Gerhard Hindricks Richard F. Gillum J. Thomas Grayston David G. Harrison Aroon D. Hingorani Robert F. Gilmour Sally C. Greaves Paul Harrison Rabea Hinkel Jeffrey M. Gimble Barry Greenberg David Hasdai Loren F. Hiratzka Frank J. Giordano Roy K. Greenberg Gerd Hasenfuss Yoshitaka Hirooka Domenico Girelli Stephen E. Greenwald Vic Hasselblad Karen K. Hirschi Adriana C. Gittenberger-de Darren C. Greenwood Richard N.W. Hauer Valeria Hirschler Groot Kathy K. Griendling Paul J. Hauptman John W. Hirshfeld Robert P. Giugliano Brian P. Griffin Derek J. Hausenloy Mark A. Hlatky Carla Giustetto John H. Griffin Richard J. Havel Michael Ho Michael M. Givertz Cindy L. Grines Axel Haverich Judith S. Hochman David Gjertson Steven K. Grinspoon Edward P. Havranek Morrison Hodges Mark T. Gladwin Johannes C. Grisar Nat Hawkins Barbara H. Hoffmann Acknowledgment of Reviewers e15

Martin H. Hoffmann John P. Ioannidis Gregory T. Jones Rosemary J. Keogh Udo Hoffmann Carlos Iribarren Robert H. Jones Richard E. Kerber Peter Höglund Thomas A. Ischinger Steven P. Jones Dean J. Kereiakes Thomas Hohlfeld Shun Ishibashi Habo J. Jongsma Karl B. Kern Stefan H. Hohnloser Masaharu Ishihara Jens Jordan Morton J. Kern Brian D. Hoit Tetsuya Ishikawa Marit Eika Jørgensen Steven J. Keteyian Fernando Holguin Ami E. Iskandrian Jacob Joseph Paul Khairy David R. Holmes Hiroyasu Iso Mark E. Josephson Amit Khera Paul Holvoet Mitsuaki Isobe Pekka Jousilahti Ali Reza Khoshdel Michael Holzer Hideki Itoh Michael J. Joyner Stefan Kiechl Shunichi Homma Bernard Iung J. Wouter Jukema Jan T. Kielstein Yuling Hong D. Dunbar Ivy Philip Jung Shinji Kihara L.N. Hopkins Kohichiro Iwasaki Suh-Hang Hank Juo Lois A. Killewich Paul N. Hopkins Philip J. Kilner Richard Hopkins Wael A. Jaber Stefan Kaab David Kilpatrick Susan Hopkins Edwin K. Jackson Ramanathan Kadirvel Hyo-Soo Kim Maria T.E. Hopman Alice K. Jacobs Alan H. Kadish Shokei Kim-Mitsuyama David R. Jacobs Hiroyuki Kageyama Masatsugu Hori Spencer B. King Marshall L. Jacobs Richard Kahn Masatsugu Horiuchi Bronwyn A. Kingwell Donald W. Jacobsen Hisashi Kai Benjamin D. Horne Scott Kinlay Paul Jacques Jan Kajstura John D. Horowitz Margaret L. Kirby Sae Young Jae Afksendiyos Kalangos Keith A. Horvath Paulus F. Kirchhof Michael R. Jaff Jonathan M. Kalman Lawrence D. Horwitz James Kirklin Allan S. Jaffe Timothy J. Kamp Akiko Hosler Lorrie A. Kirshenbaum Farouc A. Jaffer David E. Kandzari Barbara V. Howard Toru Kita Mukesh Kumar Jain Sachiko Kanki-Horimoto George Howard Masafumi Kitakaze Jose Jalife Prince J. Kannankeril Vicky Y. Hoymans Michelle Kittleson Richard W. James Ronald J. Kanter Patrick C.H. Hsieh Arthur L. Klatsky Daphne T. Hsu Konrad Jamrozik Jørgen K. Kanters Neal S. Kleiman Fang-Chi Hsu Ik-Kyung Jang Emmanouil Ioannis Kapetanakis Allan L. Klein Priscilla Hsue Joseph S. Janicki Tara Karamlou George J. Klein Chengcheng Hu Warren R. Janowitz Richard H. Karas Lloyd W. Klein Gang Hu Maurits A. Jansen Elissavet Kardami Michael D. Klein Howard Hu Stefan P. Janssens Tom R. Karl Charles S. Kleinman Rae-Chi Huang Craig T. January Johan Karlberg Paul D. Kligfield Johnny Huard James Louis Januzzi Joel S. Karliner Uwe Klima Sally Ann Huber Patrick Y. Jay Aly Karsan Francis J. Klocke Susanna Y. Huh John Jefferies S. Ananth Ananth Karumanchi Robert A. Kloner James C. Huhta J. Richard Jennings David Alan Kass John L. Knight Heikki V. Huikuri Jong Hyeon Jeong Robert S. Kass Anne A. Knowlton P.P. Hujoel Jamie Yancey Jeremy Jerome P. Kassirer Juhani Knuuti Roger Hullin Tomas Jernberg Adnan Kastrati Jon A. Kobashigawa Per M. Humpert Michael Jerosch-Herold Sekar Kathiresan Lars Kober William Gregory Hundley Paula Jerrard-Dunne Masahiko Kato Colleen Gorman Koch Judy Hung Ashish K. Jha Tomohiro Katsuya Sharon Ann Hunt Ishwarlal Jialal Hugo A. Katus Walter J. Koch Stephen Hunyor Hongyu Jiang Zvonimir S. Katusic Werner Koch Ahsan Husain Zhihua Jiang Stuart D. Katz Patrick M. Kochanek Zhezhen Jin Marc P. Kaufman Paul V. Kochupura Gianluca Iacobellis Hanjoong Jo Philipp A. Kaufmann Todd M. Koelling Ioannis Iakovou Magnus Carl Johansson Padma Kaul Wolfgang Koenig Fumito Ichinose Roger A. Johns Sanjiv Kaul Theo Kofidis Raymond E. Ideker Arnold Johnson Rae-Ellen Kavey Kwang Kon Koh Masaaki Ii B. Delia Johnson Chuichi Kawai Martin Köhrmann John S. Ikonomidis Jason L. Johnson Ziya Kaya Pipin Kojodjojo Erkki Ilveskoski Lynne L. Johnson David M. Kaye Pappachan E. Kolattukudy Armin Imhof Paula A. Johnson Mark T. Kearney Theodore J. Kolias Akihiro Inazu Richard J. Johnson Bernard Keavney Frank D. Kolodgie Robin Ingalls Robert L. Johnson Daniel P. Kelly MichelM. K. Komajda Julie R. Ingelfinger Robert Graham Johnson Darren J. Kelly Tatsuya Komaru Erik Ingelsson Timothy D. Johnson Ralph A. Kelly Masashi Komeda Joanne S. Ingwall James G. Jollis Kenneth M. Kent Issei Komuro

e16 Acknowledgment of Reviewers

Takahisa Kondo Wyman W. Lai Amir Lerman Eng H. Lo Marvin A. Konstam Edward G. Lakatta Bruce B. Lerman Amanda Lochner Stavros V. Konstantinides Evanthia Lalla Lilach O. Lerman James E. Lock Michael C. Kontos Karen S.L. Lam Michelle Letarte Andrew J. Lodge Marianne Eline Kooi Benoit Lamarche Hanno H. Leuchte Frank W. LoGerfo Willem J. Kop John J. Lamberti Adeera Levin Barry London Bruce A. Koplan Michael J. LaMonte Max Levin Ge´rard M. London Gideon Koren Rachel Lampert Benjamin D. Levine Eva M. Lonn Andreas Koster Kathryn G. Lamping R.J. Levine Gary D. Lopaschuk Thomas Erling Kottke Steve Lancel Robert A. Levine Matthias W. Lorenz Darrell N. Kotton R. Clive Landis Sidney Levitsky Douglas W. Losordo Alexei Kouroedov Ulf Landmesser Bodo Levkau Stavros P. Loukogeorgakis Petri T. Kovanen Michael J. Landzberg Bruce D. Levy Gordon D. Lowe Peter R. Kowey Florian Lang Daniel Levy Tse Min Lu Jun Koyama Irene Marthe Lang Finn Olav Levy Russell V. Luepker Andrew D. Krahn Roberto M. Lang Jerrold H. Levy Friedrich C. Luft Dara L. Kraitchman Jonathan Langberg Robert J. Levy Ketil Lunde Aldi Kraja Elisabetta Lapenna Wayne C. Levy Kathryn Lunetta Christopher M. Kramer Gina LaRocca Marilyne Le´vy Keith G. Lurie Evangelia G. Kranias John C. LaRosa Elad I. Levy Thomas F. Luscher William E. Kraus Martin Larson E. Douglas Lewandowski Aldons J. Lusis Ronald M. Krauss Warren K. Laskey Martin M. LeWinter Deborah Lyn Eswar Krishnan Robert D. Lasley Eldrin F. Lewis John W. Lynch Steen Dalby Kristensen Johan P.E. Lassus Gary F. Lewis Xin Liang Ma Michael H. Kroll Roberto Latini Chuanfu Li Christoph Maack Irving L. Kron Wei C. Lau Fan Li David M. Maahs Mitchell W. Krucoff Michael S. Lauer Gui-Rong Li Charles A. Mack Warren Kruger Ulrich Laufs Liang Li Henry Krum Michael J. Mack Jari A. Laukkanen Na Li Isao Kubota Wendy J. Mack Lenore J. Launer Nan Li Karen Kuehl Todd A. MacKenzie Geoffrey J. Laurent Ren-Ke Li Nino Kuenzli Rachel H. Mackey Kenneth R. Laurita Ronald A. Li Donald M. Kuhn Nigel Mackman Peter C. Laussen Chang-seng Liang Helena Kuivaniemi Michael I. Mackness Michael LaValley James K. Liao Thomas J. Kulik Kenneth N. MacLean Carl Lavie Ronglih Liao Lewis H. Kuller William Robb MacLellan Debbie A. Lawlor Peter Libby Iftikhar J. Kullo Paolo Madeddu Harold L. Lazar Andrew H. Lichtman Koichiro Kumagai Aldo P. Maggioni Mitchell A. Lazar David Richard Light Cheng-Deng Kuo William T. Mahle Thierry H. Le Jemtel Chee Chew Lim Christian Kupatt William T. Mahle Bruce J. Leavitt Valter C. Lima Sabina Kupershmidt Jonathan D. Mahnken Alexander Wolfgang Leber Marian C. Limacher Hiromi Kurosawa Lynn Mahony Nathan K. LeBrasseur Jing Ping Lin Tobias Kurth Lars S. Maier Sandrine Lecour Julie Lin Irving Kushner Willibald Maier JoAnn Lindenfeld Johanna Kuusisto Amanda J. Lee Francesco Maisano Jeffrey T. Kuvin Douglas S. Lee Marshall D. Lindheimer William H. Maisel Masafumi Kuzuya Duk-Hee Lee Jes S. Lindholt Amy S. Major Martijn Kwaijtaal Hon-Chi Lee Jonathan R. Lindner Koon-Hou Mak Raymond Y. Kwong I-Min Lee Karl H. Lindner Jonathan C. Makielski Yoshiki Kyo Kerry L. Lee Ken A. Lindstedt Marek Malik Alan Patrick Kypson Richard T. Lee Mark S. Link Robert T. Mallet Frans H.H. Leenen Axel Linke Giuseppe Mancia David E. Laaksonen David J. Lefer MacRae F. Linton Donna M. Mancini Carlos Alberto Labarrere Michael H. Lehmann Gregory Y.H. Lip G.B. John Mancini Vinod Labhasetwar Stephan E. Lehnart Steven E. Lipshultz Kaushik Mandal Arthur Labovitz Stephanie Lehoux Lewis A. Lipsitz Ravi Mandapati Daniel Lackland James M. Leiper Richard Lipton Jayawant N. Mandrekar Karl J. Lackner Paul LeLorier William C. Little Dennis T. Mangano Peter S. Lacy Karl B. Lemstrom Laszlo Littmann Arduino A. Mangoni Stephanie Laeer Steven R. Lentz Kiang Liu Douglas L. Mann Huichuan Lai David A. Leon Donald M. Lloyd-Jones Giovanni E. Mann Shenghan Lai Philipp M. Lepper Cecilia Wen-Ya Lo Stewart Mann

Acknowledgment of Reviewers e17

Warren J. Manning William J. McKenna Ali H. Mokdad Sherif F. Nagueh Teri A. Manolio Vallerie V. McLaughlin Ernesto Molina Matthias Nahrendorf Nicolas Mansencal C. Alex McMahan Giuseppe Molinari Samer Najjar Alberto Mantovani John J.V. McMurray David J. Moliterno Hiroshi Nakagawa Michael Stephen Marber Elizabeth M. McNally Jeffery D. Molkentin Brahmajee K. Nallamothu Francis E. Marchlinski Robert L. McNamara Nico R. Mollet Bin Nan Frank I. Marcus Helene McNulty Tom E. Mollnes Claudio Napoli Andrew O. Maree Charles F. McTiernan Donald A. Molony Carlo Napolitano Raffaele Marfella Mandeep R. Mehra Laurent Monassier Sanjiv M. Narayan James R. Margolis Roxana Mehran Nicola Montano Jagat Narula Kenneth B. Margulies Jawahar L. Mehta Farouk Mookadam David Nash Daniel B. Mark Shamir R. Mehta Jeffrey Moore Andrea Natale Roger Markwald Bernhard Meier Samia Mora Peter Nathanielsz Martin Morad Jonathan D. Marmur James B. Meigs Stanley Nattel Fred Morady Barry J. Maron Silke Meiners Matthew T. Naughton Pierre Moreau David J. Maron Cynthia J. Meininger Mohamad Navab L.A. Moreno Martin Maron Christa Meisinger Frank Naya Marie-Claude Morice Luc Maroteaux Daniel Meldrum Mona Nemer Carlos A. Morillo Philip A. Marsden Russell H. Mellor Dario Neri Ryuichi Morishita Steven P. Marso Philippe Menasche Richard W. Nesto Randolph P. Martin Ulrike Mende Toshisuke Morita Gregory E. Morley Stefan Neubauer Ulrich Martin Jean-Jacques Mercadier Donna S. Neuberg Thomas H. Marwick Patrick Mercie Nicholas W. Morrell Brian J. Morris Ellis J. Neufeld Gerald R. Marx Yahye Merhi Franz-Josef Neumann Nikolaus Marx Alan F. Merry Laurie J. Morrison Joachim Neumann Daniele Maselli C. Noel Merz David A. Morrow Richard Mortensen Peter Newburger Attilio Maseri Emmanuel Messas Ralph S. Mosca David E. Newby Frederick A. Masoudi Franz H. Messerli Mauro Moscucci L. Kristin Newby Joseph M. Massaro Luisa Mestroni Jeffrey W. Moses John H. Newman Barry M. Massie Peter Meyer Ivan P. Moskowitz Mark F. Newman Serge Masson Theo E. Meyer Arthur J. Moss Christopher H. Newton-Cheh Kazuko Masuo J.A. Michaels Karen S. Moulton Ludwig Neyses Fiona Mathews Evangelos D. Michelakis J. Paul Mounsey Dusko G. Nezic Knut Matre Erin Donnelly Michos Matthew Movsesian Stephen James Nicholls Hiroaki Matsubara Shigetoshi Mieno Dariush Mozaffarian Wilmer W. Nichols Hikaru Matsuda Richard V. Milani Thomas Muenzel Georg Nickenig Akira Matsumori D. Douglas Miller Andreas Mugge Martin John Nicklin Francesco U.S. Mattace-Raso D. Craig Miller Andrew Mugglin Alfred C. Nicolosi Christian M. Matter Jordan D. Miller Debabrata Mukherjee Marco L.S. Matteucci Leslie W. Miller James T. Niemann Rupak Mukherjee Nilanjana Maulik Todd D. Miller Christoph A. Nienaber Douwe J. Mulder Mathew S. Maurer Virginia M. Miller Petros Nihoyannopoulos James E. Muller Sigrid Nikol Laura Mauri Susumu Minamisawa Jochen Muller-Ehmsen Simon Maxwell L. LuAnn Minich Rick A. Nishimura Michael J. Mulvany Steven E. Nissen Charles Maynard Kenji Minoguchi Tomoatsu Mune Dorothea Nitsch Bongani Mawethu Mayosi Gary S. Mintz Paul Muntner Timothy David Noakes Todor N. Mazgalev Israel Mirsky Elizabeth Murphy Yoshihiro Noji PatrickM. McCarthy Manisha Mishra Timothy P. Murphy Georg Nollert Michael V. McConnell Seema Mital Charles E. Murry Fumikazu Nomura Michael Leon McCormick Brett M. Mitchell Anthony J. Muslin Lars Norgren Brian W. McCrindle Gary F. Mitchell Aviva Must Sharon-Lise T. Normand Peter A. McCullough R. Scott Mitchell Robert J. Myerburg Kari E. North David H. McDermott Richard N. Mitchell Daniel D. Myers Mary McGrae McDermott Arnold Mitnitski Jonathan Myers Gavin R. Norton Doff B. McElhinney Suneet Mittal Leann Myers Michael G.A. Norwood CarmelM. McEniery Friedrich Mittermayer Ann-Trude With Notø Edward Mcfalls M.D. Murray A. Mittleman Elizabeth G. Nabel Gian M. Novaro Daniel McGee Hiroto Miura Koonlawee Nademanee William C. Nugent John C. McGiff Mihaela M. Mocanu Ryozo Nagai Satoshi Numata John L. McGregor Peter Mohler Shoichiro Nagasaka Ju¨rg Nussberger Darren K. McGuire Emile R. Mohler III Eike Nagel Pirjo Nuutila

e18 Acknowledgment of Reviewers

Timothy Daniel O’Connell Michele Pasotti Don Poldermans Azaria J.J.T. Rein Christopher M. O’Connor Gerard Pasterkamp Jaimie W. Polson Olaf Reinhartz Gerald T. O’Connor Amit N. Patel Giulio Pompilio Steven E. Reis Christopher J. O’Donnell Anushka Alankar Patel Philip A. Poole-Wilson Willem J. Remme Peter Oettgen Ayan Patel Barry M. Popkin Serge C. Renaud Patrick T. O’Gara Jeetesh V. Patel Thomas R. Porter Frederic S. Resnic Jae K. Oh Carlo Patrono Wendy S. Post Kathyrn M. Rexrode Seil Oh Richard D. Patten Luciano Potena Matthew R. Reynolds Takahiro Ohara Cam Patterson Lincoln Potter Shereif H. Rezkalla Ann M. O’Hare Walter J. Paulus Jeffrey T. Potts Edward K. Rhee Patrick Ohlmann Aime´e D.C. Paulussen Neil Poulter Jonathan Rhodes E. Magnus Ohman Jeffrey M. Pearl Andrew J. Powell Jorge P. Ribeiro John Ohrvik Jeremy D. Pearson Janet T. Powell Fernando F. Ribeiro-Filho Peter M. Okin Daniel Pella Ashwin Prakash Paul M. Ribisl Johannes Oldenburg Patricia A. Pellikka Abhiram Prasad Ken Rice Donal S. O’Leary Michael Pencina Sanjay K. Prasad Jean-Paul Richalet Jeffrey W. Olin Marc S. Penn Susan J. Pressler Vincent Richard Michael Hecht Olsen Dudley J. Pennell Jack F. Price Paul M Ridker Eric N. Olson Carl J. Pepine Ronald Prineas Johannes Rieger Lyle J. Olson Paul E. Peppard Frits W. Prinzen Nader Rifai Anders G. Olsson Emerson C. Perin Silvia G. Priori Charanjit S. Rihal Ray A. Olsson John Pernow Kirkwood A. Pritchard, Jr. Eric B. Rimm Patrick G. O’Malley James C. Perry Vincent Probst P.A. Ringleb Eileen O’Meara Stephen D. Persell Karin Przyklenk Rasmus Sejersten Ripa Steve R. Ommen Sharina D. Person William T. Pu James M. Ritter James O. O’Neill G. Rutger Persson John D. Puskas Eberhard Ritz Henry Ooi Inga Peter Reed Pyeritz Alain Rivard Kalevi Pyorala Suzanne Oparil Karlheinz Peter Jeffrey Robbins Lionel H. Opie Nicholas S. Peters Robert Roberts Hakan Oral Zhaohui Steve Qin Eric D. Peterson William C. Roberts E. John Orav Miguel A. Quiñones Linda R. Peterson Sander J. Robins Jose M. Ordovas Pamela Peterson Jennifer G. Robinson Joseph P. Ornato Ton J. Rabelink Eva Petkova Simon C. Robson Michael F. O’Rourke Alejandro A. Rabinstein Michael E. Phelps Frederic Roche Leiv Ose Vittorio Racca Gerald B. Phillips Dan M. Roden Clive Osmond Frank E. Rademakers Richard P. Phipps Brian Rodrigues Yutaka Otsuji Daniel J. Rader Colin K. Phoon Alfredo E. Rodriguez David Ott R. Radermecker Robert N. Piana Alicia Rodriguez-Pla Harald C. Ott Martha J. Radford Philippe Pibarot Matthew T. Roe Catherine M. Otto Paolo Raggi Eugenio Picano Mark Roest Noriyuki Ouchi Shahbudin H. Rahimtoola Michael H. Picard Marco Roffi Michel Ovize Leopoldo Raij J. Geoffrey Pickering Veronique L. Roger Al Ozonoff Elaine W. Raines Thomas G. Pickering Olli T. Raitakari Campbell Rogers Pal Pacher Luc A. Pierard Sanjay Rajagopalan Joseph G. Rogers Sandosh Padmanabhan Burkert Pieske Nalini M. Rajamannan Mary J. Roman Francis D. Pagani Bruce Pihlstrom Venkatesh Rajapurohitam Mats Rönnback Massimo Pagani Nico H.J. Pijls Harry Rakowski Dieter Ropers Richard L. Page Louise Pilote Vivek Rao Wayne D. Rosamond Jennifer Pai David R. Pimentel Ursula Rauch Eric A. Rose Paolo Palatini Ileana Pina Ursula Ravens Noel R. Rose Wulf Palinski H. Michael Piper Reza S. Razavi Michael R. Rosen Julio C. Palmaz Tobias Pischon Peter Razeghi David S. Rosenbaum Colin N.A. Palmer Bertram Pitt Richard Re Gary A. Rosenberg Demosthenes Panagiotakos Jeffrey L. Platt Patricia Reant Michael E. Rosenfeld Natesa G. Pandian Jonathan F. Plehn Rita F. Redberg Robert S. Rosenson James S. Pankow Mark J. Pletcher Margaret M. Redfield David N. Rosenthal Nazareno Paolocci Jorge Plutzky Josep Redon Anthony Rosenzweig Domenico Paparella Bruno K. Podesser Thomas C. Register Bernard Rosner Carlo Pappone Gerald M. Pohost Jalees Rehman Allan M. Ross Thomas G. Parker Paul Poirier Nathaniel Reichek David L. Ross Juan C. Parodi Joseph F. Polak Muredach Reilly Gian Paolo D. Rossi Acknowledgment of Reviewers e19

Marco L. Rossi John Lewis Sapp Christian Seiler Daniel I. Simon J.E. Rossouw Dennis Sarabi Akira Sekikawa Joel A. Simon Stephen J. Roth Wim Saris Donald F. Sellitti Michael Simons Dietrich Rothenbacher Ferdinando Carlo Sasso Frank W. Sellke Maarten L. Simoons Richard B. Rothman Masataka Sata Elizabeth Selvin Paul C. Simpson Peter M. Rothwell Naveed Sattar Andrew P. Selwyn Ross J. Simpson Joris Rotmans Kurt W. Saupe Craig H. Selzman Alan R. Sinaiko Melvyn Rubenfire Giorgio Savazzi Gregg L. Semenza Mervyn Singer Frederick L. Ruberg Stephen G. Sawada Marc J. Semigran Krishna Singh Lewis J. Rubin Tatsuya Sawamura Chris Sempos Michael N. Singh Israel Rubinstein Douglas B. Sawyer Chandan K. Sen Steven N. Singh Neil B. Ruderman Leslie A. Saxon Roxy Senior Tajinder P. Singh Goran Rudez Tiziano Scarabelli Victor L. Serebruany Albert J. Sinusas Marc Ruel Pierre-Yves Scarabin Charles N. Serhan Deborah A. Siwik Luis M. Ruilope Volker Schachinger Patrick W. Serruys Allan C. Skanes Carlos E. Ruiz Thomas Schachner Marc J. Servant Susan A. Slaugenhaupt Pilar Ruiz-Lozano Alvin Schadenberg Sudha Seshadri Karen Sliwa John S. Rumsfeld Hartzell V. Schaff Howard D. Sesso Richard W. Smalling Marschall S. Runge Martin J. Schalij Magnus Settergren Otto A. Smiseth Frank Ruschitzka Bernhard Schaller Ralph Shabetai Craig Smith Raymond R. Russell Wolfgang Schaper Robert E. Shaddy George Davey Smith Vincenzo Russo Doug E. Schaubel Ajay M. Shah Grace L. Smith Wolfgang Rutsch Patrick Schauerte Maully J. Shah Jonathan Smith Carolyn Rutter Debra Ann Schaumberg Pravin M. Shah Nicholas L. Smith Elfriede Ruttmann Dierk Scheinert Prediman K. Shah Stephen Mark Smith Thomas Ryan Melvin M. Scheinman David M. Shahian Timothy William Smith Thomas J. Ryan Sebastian M. Schellong Catherine M. Shanahan Warren Morrison Smith Jack Rychik Benjamin J. Scherlag Richard P. Shannon Ryszard T. Smolenski Lars Ryden Ralph Theo Schermuly Behrooz G. Sharifi David B. Snead Bernhard Schieffer Arya M. Sharma Allan D. Sniderman Tobias Saam Giuseppe Schillaci Samin K. Sharma Burton E. Sobel Samir Saba Martin Schillinger Ken Sharpe Birgitta Söder Hani N. Sabbah Alexandru Schiopu Palma Shaw Stefan Soderberg Joseph F. Sabik Ralph L. Sacco Thomas Schlosser Amanda M. Shearman Kyoko Soejima Michael N. Sack Alvin Schmaier Michael Shechter Manoocher Soleimani Frank M. Sacks Axel Schmermund Soren P. Sheikh Scott D. Solomon Michael S. Sacks Claudia Schmidtke Prem S. Shekar Virend K. Somers J. Evan Sadler Neil Schneiderman Rhidian John Shelton Robert J. Sommer Junichi Sadoshima Albert Schoemig Stanton K. Shernan Ali Sonel Michel E. Safar Frederick J. Schoen Mark V. Sherrid Dan Sorescu Jeffrey E. Saffitz Mark Howard Schoenfeld Guo-Ping Shi Vincent L. Sorrell Robert D. Safian Jurgen Schrader Weibin Shi P.C. Souverein Alexander Sagie Richard B. Schuessler Rei Shibata Arthur A. Spector David J. Sahn Gerhard C. Schuler Mei-Chiung Shih J. David Spence Genichi Sakaguchi Kevin A. Schulman Koichi Shimizu Markus Sperandio Sanjeev Saksena Heinz-Peter Schultheiss Tatsuya Shimizu John A. Spertus Tomas A. Salerno Richard Schulz Wataru Shimizu Martin Spiecker Veikko Salomaa Eric Schulze-Bahr Hiroaki Shimokawa Bruce Spiess Flora Sam Heribert Schunkert Ichiro Shiojima Francis G. Spinale Frederick F. Samaha Arnold Schwartz Girish S. Shirali Paolo Spirito Nilesh J. Samani Gary L. Schwartz Kalyanam Shivkumar David H. Spodick Jeffrey Samet Kenneth A. Schwartz Michael G. Shlipak Martinus Spoor Jonathan M. Samet Peter J. Schwartz Yehuda Shoenfeld Matthew L. Springer Jane-Lise Samuel Robert Stockton Schwartz Stephen R. Shorofsky Henri M.H. Spronk Timothy Allen Sanborn David S. Schwartzman Matthias Siepe Francesco Squadrito Dirk Sander Ernst R. Schwarz Hans-H. Sievers Iain B. Squire Mikael Sander P.E. Schwarz Ulrich Sigwart Ray W. Squires Stephen P. Sanders Karie Scrogin Donald S. Silverberg V.S. Srinivas John E. Sanderson Paola Sebastiani David I. Silverman Martin G. St. John Sutton Anthony J. Sanfilippo Christine E. Seidman Jean-Sebastien Silvestre Earl R. Stadtam Michael C. Sanguinetti Jonathan G. Seidman Robert D. Simari Jan A. Staessen e20 Acknowledgment of Reviewers

Meir J. Stampfer David P. Taggart Gordon F. Tomaselli Thomas E. Vanhecke Kenneth Stanley Peter Taggart Naruya Tomita Mani A. Vannan William C. Stanley Masato (Mike) Takahashi Marcello Tonelli Cristina Varas-Lorenzo William Stanley Shinji Takai Andrew M. Tonkin Nerea Varo Alice V. Stanton Satoshi Takeshita Eric J. Topol Mariuca Vasa-Nicotera Norbert Stefan Johanna J.M. Takkenberg Per Tornvall Thomas A. Vassiliades Michael W. Steffes William T. Talman Robert Daniel Toto Dorothy E. Vatner Philippe Gabriel Steg Masashi Tanaka Rhian M. Touyz Stephen F. Vatner Coen D. Stehouwer Toshihiro Tanaka Jeffrey A. Towbin Mary S. Vaughan Sarrazin Helmut O. Steinberg Lilong Tang Dwight A. Towler Hector O. Ventura Julia Steinberger Rajendra K. Tangirala Jonathan N. Townend Paolo Verdecchia Robin H. Steinhorn Lloyd Y. Tani Kazunori Toyoda Freek W.A. Verheugt Steve R. Steinhubl Laszlo B. Tanko Russell P. Tracy Subodh Verma Kurt R. Stenmark Felix C. Tanner John K. Triedman Cees Vermeer Andrew Steptoe Masaya Tanno Han-Mou Tsai Richard L. Verrier Michael P. Stern Yoshihisa Tanoue Thomas T. Tsai Sara Vesely Lynne Warner Stevenson Kahraman Tanriverdi Teresa S.M. Tsang George W. Vetrovec William G. Stevenson Victor F. Tapson Sotirios Tsimikas Victoria Vetter Duncan J. Stewart Jean-Claude Tardif Tim K. Tso Aristidis Veves Julian M. Stewart Giovanni Targher Yukiomi Tsuji Flordeliza S. Villanueva Ralph A.H. Stewart Mark B. Taubman Igor Tudorache Francisco Villarreal Simon Stewart Ahmed Tawakol Paul A. Tunick Karen A. Vincent Roland Stocker Allen J. Taylor José Tuñón Jakob Vinten-Johansen Karen Stokes Andrew M. Taylor Tanya N. Turan Renu Virmani Monika Stoll Anne L. Taylor Fiona Turnbull Sami Viskin Gregg W. Stone Alain Tedgui Alexander G.G. Turpie Eric Vittinghoff Gus J. Vlahakes Peter H. Stone Usha Tedrow E. Murat Tuzcu Volkan Tuzcu Robert A. Vogel Roslyn A. Stone John R. Teerlink Julie St-Pierre James S. Tweddell Manfred Vogt Paul S. Teirstein Ruth H. Strasser Christophe Tzourio Pierre Voisine David F. Teitel Bodo E. Strauer Arnold von Eckardstein George Tellides H. William Strauss Thomas Unger Finn Edler von Eyben Martin Tepel Bruno H. Stricker Zoltan Ungvari Marc A. Vos Masahiro Terashima Erik S.G. Stroes Philip Urban Robert Voswinckel Ronald L. Terjung Matthias Stuber Zsolt Urban Sari Voutilainen Norma Terrin Christian Stumpf Barry F. Uretsky Naren Vyavahare Dellara F. Terry Rodney Sturdivant Masuko Ushio-Fukai Pierre Theroux Yan Ru Su Bernard Waeber Aravinda Thiagalingam Krishnankutty Sudhir Viola Vaccarino Anja Wagner Perumal Thiagarajan Cathie L.M. Sudlow Miguel Valderrabano Ron Waksman Ravi R. Thiagarajan Koichi Sughimoto Marco Valgimigli Albert L. Waldo Chris Thiemermann M.-Saadeh Suleiman Patrick J.T. Vallance Lars Wallentin Gaetano Thiene Jerome L. Sullivan Jesus G. Vallejo Susanna M. Wallerstedt Anita C. Thomas Lisa M. Sullivan Joannis E. Vamvakopoulos Reidar Wallin James D. Thomas Jack C.J. Sun Eric Van Belle Edward P. Walsh Thoralf M. Sundt III Randal J. Thomas Peter Van Buren Thomas Walther H. Robert Superko Shane R. Thomas Ruud M.A. Van de Wal Paul J. Wang Mark A. Sussman William Thomas Frans J. Van de Werf Ping H. Wang Thomas M. Suter Douglas Thompson Greta Van den Berghe Qing Wang Allison J. Sutherland MaryLou Thompson Johanna Gerarda van der bom Thomas J. Wang Kim Sutton-Tyrrell Paul D. Thompson Yvonne T. van der Schouw Wenyu Wang Erik J. Suuronen Richard Thompson Ernst E. van der Wall Xuejun Wang Alan F. Sved Jens Jakob Thune Berry M. van Gelder Yun Wang Lars G. Svensson Johan Thyberg Bethany Van Guelpen James W. Warnica Elisabet Svenungsson Lu Tian George F. Van Hare Manabu Watanabe Michael O. Sweeney Rong Tian Linda Van Horn Mari A. Watanabe E.R. Swenson Tomasz Timek Johannes J. van Lieshout Nozomi Watanabe Bernard Swynghedauw Brian Timmons Joost P. van Melle David D. Waters Laurence Tiret A.M. van Rij Sergio Waxman Koichi Tabayashi Asa Tivesten J. Peter van Tintelen W. Douglas Weaver Stefano Taddei Geoffrey H. Tofler Dirk J. van Veldhuisen Catherine Webb Heinrich Taegtmeyer Cheng-Hock Toh David R. van Wagoner David J. Webb

Acknowledgment of Reviewers e21

Gary D. Webb Andrew S. Weyrich Paul Wolkowicz Mitsuhiro Yokoyama John G. Webb Gillian A. Whalley Kai C. Wollert Shi-Joon Yoo Steven A. Webber John Wharton Lennie Wong Young-sup Yoon Christian Weber Christopher J. White Nathan D. Wong Lawrence H. Young Karl T. Weber Guy StJ. Whitley John C. Wood Cheuk-Man Yu Michael A. Weber J. Lindsay Whitton Mark A. Wood Chun Yuan Mark W.I. Webster Mark H. Wholey Elizabeth A. Woodcock Kevin Wei Samuel A. Wickline Angela Woodiwiss Ian C. Zachary L. Wei Julian Widder R. Scott Wright Ralf Zahn Wei Wei Susan E. Wiegers Alan H.B. Wu Peter Zahradka Franz Weidinger William Wijns Chuntao Wu Andrew Zalewski Dorothee Weihrauch David J. Wilber Joseph C. Wu Faiez Zannad Max Harry Weil Arthur A.M. Wilde Kenneth K. Wu Wojciech Zareba Myron Weinberger Ian B. Wilkinson D. George Wyse Barry L. Zaret Neal L. Weintraub Bruce L. Wilkoff Xiao Xiao William S. Weintraub Andrew R. Willan Alan M. Zaslavsky Susan Xu Richard D. Weisel Walter C. Willett Cuihua Zhang Jianyi Zhang Mary C. Weiser-Evans Bryan Williams Magdi H. Yacoub Rong Zhang Eric S. Weiss David O. Williams Hitoshi Yaku Shetuan Zhang Harvey Richard Weiss David M. Williams Hafize Yaliniz Zefeng Zhang Robert G. Weiss Mark A. Williams Norikazu Yamada Li-Ru Zhao Robert M. Weiss Scott R. Willoughby Yoshiji Yamada Shankuan Zhu Neil J. Weissman Andrew M. Wilson Takashi Yamaki Jeffrey I. Weitz Peter W. Wilson Clyde W. Yancy Xinsheng Zhu Carrie Welch Stephan Windecker Homer Yang Felix Zijlstra David J. Welsh Karl Winkler Qinglin Yang Michael R. Zile Frederick G. Welt Jeffrey A. Winkles Zhihong Yang Thomas Zimmer Peter Wenaweser Jonathan Allan Winston Katsusuke Yano M. Bridget Zimmerman Nanette Kass Wenger Jacqueline C.M. Witteman Masafumi Yano Jean-Marc Zingg Jolanda J. Wentzel Janet Wittes Jack Yanovski Douglas P. Zipes Nikos S. Werner Joseph L. Witztum Derek M. Yellon Edgar Zitron Malcolm West Stephen D. Wiviott Midori Anne Yenari Carmine Zoccali Cynthia M. Westerhout Philip A. Wolf Seppo Yla-Herttuala Irving H. Zucker Dirk Westermann Eugene E. Wolfel Paul G. Yock Jay L. Zweier Cornelia M. Weyand Michael S. Wolin Mervin C. Yoder AHA Issues New Products lipid abnormalities with early atherosclerosis, discuss chal- The following new products for the public and the healthcare lenges with previous guidelines, and highlight results of professional are available through your local American Heart clinical trials with statin therapy in children and adoles- Association or by calling 1-800-AHAUSA1. cents with familial hypercholesterolemia or severe hypercholesterolemia. Product code 71-0406. ● AHA Conference Proceedings: Understanding the Com- ● AHA Scientific Statement: Essential Features of a Surveil- plexity of Trans Fatty Acid Reduction in the American Diet: lance System to Support the Prevention and Management American Heart Association Trans Fat Conference 2006. of Heart Disease and Stroke. This statement provides a Read about the current status and future implications of brief overview of the Healthy People 2010 goals, preven- reducing trans fatty acids without increasing saturated fats tion and management strategies, and the role of surveil- in the food supply, while functionality and consumer lance in monitoring the impact of prevention and treatment acceptance of packaged, processed, and prepared foods are efforts. It also provides a review of the existing surveil- maintained. Product code 71-0326. lance system for monitoring progress toward preventing ● AHA Guideline: Evidence-Based Guidelines for Cardio- heart disease and stroke in the United States and recom- vascular Disease Prevention in Women: 2007 Update. This mendations for filling important gaps in that system. update provides the most current clinical recommendations Product code 71-0386. for the prevention of cardiovascular disease in women Ն20 ● AHA Scientific Statement: Exercise and Acute Cardiovas- years of age and is based on a systematic search of the cular Events: Placing the Risks Into Perspective. This highest-quality science, interpreted by experts in the fields scientific statement discusses the potential cardiovascular of cardiology, epidemiology, family medicine, gynecology, complications of exercise, their pathological substrate, and internal medicine, neurology, nursing, public health, statis- their incidence and suggests strategies to reduce these tics, and surgery. These guidelines also cover the primary complications. Product code 71-0400. and secondary prevention of chronic atherosclerotic vascu- ● AHA Scientific Statement: Genetic Basis for Congenital lar diseases. Product code 71-0401. Heart Defects: Current Knowledge. This statement pro- ● AHA Guideline: Prevention of Infective Endocarditis. This vides the clinician with a summary of what is currently guideline updates the recommendations for the prevention known about the contribution of genetics to the origin of of infective endocarditis that were last published in 1997. congenital heart disease. Product code 71-0376. Product code 71-0407. ● AHA Scientific Statement: Indications for Heart Trans- ● AHA Policy Statement: Nonfinancial Incentives for Qual- plantation in Pediatric Heart Disease. Learn about the ity. Four principles were crafted to guide the structure and evaluation that led to the development and refinement of metrics used in pay-for-quality programs, and they identi- indications for heart transplantation for patients with con- fied at least 6 areas that required additional research to genital heart disease and pediatric cardiomyopathies in serve as criteria that should be considered when designing addition to indications for pediatric heart retransplantation. and evaluating pay-for-quality programs. Product code Product code 71-0393. 71-0387. ● AHA Scientific Statement: Noninherited Risk Factors and ● AHA Scientific Statement: Acute Coronary Care in the Congenital Cardiovascular Defects: Current Knowledge. Elderly, Part I: Non–ST-Segment-Elevation Acute Coro- This statement summarizes the currently available litera- nary Syndromes. The first part of this 2-part statement ture on potential fetal exposures that might alter risk for summarizes evidence on patient heterogeneity, clinical cardiovascular defects. Product code 71-0377. presentation, and treatment of non–ST-segment elevation ● AHA Scientific Statement: Physical Activity Intervention acute coronary syndromes in relation to age (Ͻ65, 65 to 74, Studies: What We Know and What We Need to Know. An 75 to 84, and Ն85 years). Product code 71-0404. overview is provided of existing physical activity interven- ● AHA Scientific Statement: Acute Coronary Care in the tion research, focusing on subpopulations and intervention Elderly, Part II: ST-Segment-Elevation Myocardial Infarc- modalities. New ideas and recommendations are also tion. This second part summarizes evidence on presentation offered to improve the state of the science within each area and treatment of ST-segment–elevation myocardial infarc- and, where possible, to propose ideas to help bridge the tion in relation to age (Ͻ65, 65 to 74, 75 to 84, and Ն85 gaps between these existing categories of research. Product years). Product code 71-0405. code 71-0369. ● AHA Scientific Statement: Cardiovascular Risk Reduction ● AHA Scientific Statement: Recommendations and Consid- in High-Risk Pediatric Patients. A panel of experts re- erations Related to Preparticipation Screening for Cardio- viewed what is known about very premature cardiovascular vascular Abnormalities in Competitive Athletes: 2007 Up- disease in 8 high-risk pediatric diagnoses and, from the date. Preparticipation cardiovascular screening is the science base, developed practical recommendations for systematic practice of medically evaluating large, general management of cardiovascular risk. Product code 71-0378. populations of athletes before participation in sports for the ● AHA Scientific Statement: Drug Therapy of High-Risk purpose of identifying (or raising suspicion of) abnormal- Lipid Abnormalities in Children and Adolescents. This ities that could provoke disease progression or sudden statement examines new evidence on the association of death. Identifying the relevant diseases may prevent some B1

m B2 News

instances of sudden death after temporary or permanent ation/American College of Cardiology Secondary Preven- withdrawal from sports or targeted treatment interventions. tion Guidelines, including baseline patient assessment, Product code 71-0399. nutritional counseling, risk factor management (lipids, ● AHA Scientific Statement: Relevance of Genetics and blood pressure, weight, diabetes mellitus, and smoking), Genomics for Prevention and Treatment of Cardiovascular psychosocial interventions, and physical activity counsel- Disease. Approaches that researchers are using to advance ing and exercise training. Product code 71-0394. ● understanding of the genetic basis of cardiovascular dis- AHA/ACC/HRS Scientific Statement: Recommendations for ease are discussed, as well as details of the current state of the Standardization and Interpretation of the Electrocar- knowledge regarding the genetics of myocardial infarction, diogram: Part I: The Electrocardiogram and Its Technol- atherosclerotic cardiovascular disease, hypercholesterol- ogy. This statement examines the relation of the resting ECG to its technology and to foster an understanding of emia, and hypertension. Product code 71-0410. how the modern ECG is derived and displayed so that ● AHA Scientific Statement: Treatment of Hypertension in standards are established that will improve the accuracy the Prevention and Management of Ischemic Heart Dis- and usefulness of the ECG in practice. Product code ease. This statement summarizes the published data relat- 71-0389. ing to the treatment of hypertension in the context of coronary ● AHA/ACC/HRS Scientific Statement: Recommendations for artery disease prevention and management and attempts, on the Standardization and Interpretation of the Electrocar- the basis of the best available evidence, to develop recom- diogram: Part II: Electrocardiography Diagnostic State- mendations that will be appropriate for blood pressure reduc- ment List. This statement provides a concise list of diag- tion and the management of coronary artery disease in its nostic terms for ECG interpretation that can be shared by various manifestations. Product code 71-0412. students, teachers, and readers of electrocardiography. An ● AHA Scientific Statement: Use of Nonsteroidal Antiinflam- intended outcome of this statement list is greater unifor- matory Drugs: An Update for Clinicians. Read about the mity of ECG diagnosis and a resultant improvement in current evidence that indicates that selective COX-2 inhib- patient care. Product code 71-0390. itors have important adverse cardiovascular effects, which ● AHA/ACC/SCAI/ACS/ADA Science Advisory: Prevention include increased risk for myocardial infarction, stroke, of Premature Discontinuation of Dual Antiplatelet Therapy heart failure, and hypertension. Product code 71-0396. in Patients With Coronary Artery Stents. This advisory ● AHA/ASA Guideline. Guidelines for the Early Management stresses the importance of dual antiplatelet therapy after of Adults With Ischemic Stroke. This guideline is an placement of a drug-eluting stent and educating the patient overview of the evaluation and treatment of adults with and healthcare providers about hazards of premature dis- acute ischemic stroke for physicians and other emergency continuation. Product code 71-0395. ● healthcare providers who treat patients within the first 48 AHA/ADA Scientific Statement: Primary Prevention of hours after stroke. Information for healthcare policy mak- Cardiovascular Diseases in People With Diabetes Mellitus. ers is included. Recommendations for management from The ADA and AHA have issued separate recommendations for each of the cardiovascular risk factors in patients with the first contact by emergency medical services personnel diabetes. This statement attempts to harmonize the recom- through initial admission to the hospital are also provided. mendations of both organizations where possible and Product code 71-0398. recognizes areas in which AHA and ADA recommenda- ● AHA/ASA Guideline: Guidelines for the Management of tions differ. Product code 71-0379. Spontaneous Intracerebral Hemorrhage in Adults: 2007 ● AHA/HRS Scientific Statement: Addendum to “Personal Update. Current and comprehensive recommendations are and Public Safety Issues Related to Arrhythmias That May presented for the diagnosis and treatment of acute sponta- Affect Consciousness: Implications for Regulation and neous intracerebral hemorrhage. Product code 71-0411. Physician Recommendations: A Medical/Scientific State- ● AHA/AACVPR Scientific Statement: Core Components of ment From the American Heart Association and the North Cardiac Rehabilitation/Secondary Prevention Programs: American Society of Pacing and Electrophysiology”: Public 2007 Update. This updated statement presents current Safety Issues in Patients With Implantable Defibrillators. This information on the evaluation, interventions, and expected statement extends the original 1996 recommendations and outcomes in each of the core components of cardiac provides specific recommendations on driving for individuals rehabilitation/secondary prevention programs, in agree- with implantable cardioverter-defibrillators (implanted for ment with the 2006 update of the American Heart Associ- primary prevention). Product code 71-0392.

Meetings Calendar

AHA Meetings forum in which to present recent scientific work Many of these meetings are sponsored by the American related to stroke and cerebrovascular disease. More Heart Association (AHA) and its Scientific Councils. For than 600 abstract presentations and lectures will be information, contact AHA, Scientific Meetings, 7272 featured. This year, special symposia will focus on Greenville Avenue, Dallas, TX 75231-4596; Fax 214- numerous topics, including controversies in vascular 373-3406; E-mail [email protected]; or cognitive impairment, genetics of stroke, getting visit the Web site http://my.americanheart.org/portal/ therapies across the blood–brain barrier, metabolic professional/conferencesevents down regulation in cerebral ischemia, stroke in neonates, stroke in women, platelet resistance in 2007 stroke prevention, diagnosis and management of July 30–Aug 2: 4th Annual Symposium of the Ameri- AVM, aneurysm and intracranial hemorrhage, the can Heart Association Council on Basic Cardio- role of exercise in stroke rehabilitation, the latest in vascular Sciences: Cardiovascular Repair and stroke prevention, advancing stroke systems of care, Regeneration: Structural and Molecular Ap- and other informative symposia. Sessions in clinical proaches in the Cellular Era. Keystone, Colo. This categories will center on diagnosis, acute manage- 3.5-day conference is a multifaceted symposium ment, in-hospital treatment, rehabilitation and recov- highlighting research under development in the car- ery, pediatric stroke, prevention and community/risk diovascular community targeted at slowing and/or factors, outcomes, vascular cognitive impairment, reversing the pathogenesis of disease. The confer- and systems of stroke care. Experimental categories ence will focus on how cellular-based approaches will address neurons/glia/inflammation and vascular are being manipulated to enhance the repair and pathophysiology/thrombosis. See Web site: regeneration capabilities of the cardiovascular sys- http://www.heart.org/presenter.jhtml?identifierϭ3045505 tem with the goal of therapeutic based interventions. Mar 11–15: Joint Conference – 48th Cardiovascular See Web site: http://www.heart.org/presenter. Disease Epidemiology and Prevention and the jhtml?identifierϭ3044056 Nutrition, Physical Activity, and Metabolism Sept 26–29: 61st Annual High Blood Pressure Research Conference 2008. Colorado Springs, Colo. The Conference 2007. Tucson, Ariz. The 2.5-day scien- Annual Conference offers participants the opportu- tific program gives physicians and research investi- nity to learn about: population trends in cardiovas- gators an opportunity to enhance their knowledge, cular diseases and their risk factors; causes and advance their skills, and learn about the latest mechanisms of atherosclerosis and other vascular developments in research pertaining to hyperten- diseases; results of cardiovascular disease treatment sion, stroke, kidney function, obesity, and genetics. and prevention trials; methods of population surveil- The program will include state-of-the-art lectures lance for cardiovascular disease and risk factors; and more than 350 oral and poster abstract presenta- techniques in preventive cardiology nutrition and tions and discussions led by authorities. See Web site: cardiovascular disease; and outcomes research in http://www.heart.org/presenter.jhtml?identifierϭ3043476 cardiovascular disease. See Web site: http://www. Nov 4–7: Scientific Sessions 2007. Orlando, Fla. Scien- heart.org/presenter.jhtml?identifierϭ3046690 tific Sessions encompasses 4 days of invited lectures and investigative reports. Simultaneous presenta- Other Meetings of Interest—Domestic tions represent all fields of cardiovascular and related disciplines. The program will include more 2007 than 4,000 basic, clinical, and population science Sept 5–8: 8th Annual New Cardiovascular Horizons abstract presentations; plenary, special, and how-to and Management of the Diabetic Foot & Wound sessions, morning programs and cardiovascular Healing. New Orleans, La. For more information, seminars; clinical practice sessions focusing on cur- contact [email protected], phone rent standards of care for practicing clinicians; 337-261-0944, or fax 337-572-9778. See Web site: translational science sessions that bring together http://www.newcvhorizons.com basic scientists and clinicians; and Ask the Experts luncheons and in-depth subspecialty updates. There will Other Meetings of Interest—International also be 7 pre-Sessions symposia. See Web site: http:// scientificsessions.americanheart.org 2007 Sept 17–20: 4th European Meeting on Vascular Biology 2008 and Medicine. Bristol, United Kingdom. Keynote Feb 20–22: International Stroke Conference 2008. New and plenary speakers’ topics include atherosclerosis Orleans, La. This 2.5-day conference provides a treatment, heart disease treatment, clinical modula- B3 B4 Meetings Calendar

tion of angiogenesis, vulnerable plaque, and pheno- many. A series of thematic lectures will run through- type of vascular cells. Tracks include atherosclero- out the congress that describe progresses and inno- sis, endothelium and angiogenesis, and cellular vations from 14 separate topics. Pre-conference dysfunction. For more information, contact meetings are also offered. Educational sessions will [email protected], phone ϩ44-1922- take the form of lectures, round tables, pro-con 457-984, or fax ϩ44-1922-455-238. See Web site: debates, clinical presentations, core competencies, http://2007.emvmb.org tutorials, and interactive education. For more infor- Oct 7–10: 7th International Congress on Coronary mation, contact [email protected], phone ϩ32-2- Artery Disease–From Prevention to Intervention 559-03-55, or fax ϩ32-2-527-00-62. See Web site: (ICCAD 7). Venice, Italy. The meeting will follow http://www.esicm.org the format of the very successful previous ICCAD Oct 14–16: 5th International Meeting on Intensive Coronary Artery Disease meetings and will provide Cardiac Care. Tel Aviv, Israel. Three parallel a comprehensive update on coronary disease in all sessions, with more 100 presentations, will be fea- its aspects. Keynote lectures will be delivered by a tured. A parallel nursing stream will also be pro- distinguished international faculty, while a large vided. For details, contact [email protected]. See number of selected free communications will report Web site: http://isas.co.il/cardiaccare2007 new data from basic research laboratories and clin- Nov 8–11: Fifth International Congress on Vascular ical centers around the globe. The program will Dementia. Budapest, Hungary. Attendees shall include sessions on molecular mechanisms, gene have an opportunity to deliberate on large and therapy and cell therapy, epidemiology and preven- small vessel brain diseases and how they contrib- tion, and clinical aspects. There will be a major ute to cognitive decline. There will also be an focus on new frontiers in interventional cardiology opportunity to identify the specific psychological and to the surgical management of coronary disease. markers, if any, of vascular dementia, and also the A new feature will be a fast track for recent and genetic factors involved. The overlap with Alz- “about to break” clinical trials. For more informa- heimer’s disease will be a central issue, as will tion, contact [email protected] or phone ϩ41- be the white matter changes frequently seen in 22-908-0488. See Web site: http://www.kenes.com/cad7 vascular dementia. For details, contact Oct 7–10: 20th Annual Congress of the European [email protected]. See Web site: http://www. Society of Intensive Care Medicine. Berlin, Ger- kenes.com/vascular

The American Heart Association welcomes announcements of interest to physicians, scientists, researchers, and others concerned with cardiovascular and cerebrovascular medicine. All copy is reviewed by the Scientific Publishing Department. Content may be edited for style, clarity, and length. Copy should be sent to Publications– AHA News & Meetings Calendar, American Heart Association, 7272 Greenville Ave, Dallas, TX 75231-4596; Fax 214-691-6342; E-mail [email protected] American Heart Association Newly Elected Fellows, Spring 2007

Sixty-six Premium Professional Members were elected Fellows and International Fellows of the American Heart Association (AHA) in the spring 2007. Fellows are elected on the basis of their outstanding credentials, achievements, and community contributions to the study of cardiovascular disease and stroke. Persons elected to fellowship are entitled to use FAHA, Fellow of the American Heart Association, as a professional designation. Fellows who reside outside the United States and Canada are designated International Fellows. For more information on the AHA Fellowship program, please visit our Web site at http://www.americanheart.org/presenter.jhtml?identifierϭ3033104

COUNCIL ON CLINICAL Eileen Michelle Hsich, MD, Ira S. Nash, MD, FAHA Martin St. John Sutton, CARDIOLOGY FAHA Associate Professor MBBS, FAHA Assistant Professor Cardiovascular Institute Director, Cardiovascular Imaging MARTIN R. BERK, MD, FAHA Case Western Reserve Mount Sinai Medical Center Department of Partner/Physician University New York, NY Medicine/Cardiology Cardiology & Interventional Hospital of the University of Cleveland, Ohio Jeffrey W. Olin, DO, FAHA Vascular Associates Pennsylvania Professor and Director, Dallas, Tex Atul R. Hulyalkar, MD, Philadelphia, Pa FAHA Vascular Medicine David J. D’Agate, DO, Cardiologist Mount Sinai School of Medicine Jon Walter Wahrenberger, FACC, FCCP, FAHA MD, FAHA North Ohio Heart Center, Inc. New York, NY Suffolk Heart Group, LLP Assistant Professor of Medicine Westlake, Ohio Armen Ovsepian, MD, FAHA Smithtown, NY Dartmouth Hitchcock Medical Staff Physician Birgit Kantor, MD, PhD, Center Harold L. Dauerman, MD, Suffolk Heart Group, LLP FAHA Lebanon, NH FAHA Smithtown, NY Senior Associate Consultant III Director, CV Catheterization Pavlos Toutouzas, MD, FAHA Robert A. Pelberg, MD, Labs Mayo Clinic, Rochester Athens, Greece FAHA University of Vermont Rochester, Minn Ohio Heart and Vascular Dimitrios N. Tziakas, MD, Burlington, Vt Kelley D. Kennedy, MD, Center FAHA Mark H. Drazner, MD, MSc, FAHA Cincinnati, Ohio Assistant Professor in Cardiology FAHA Cardiologist Alexandropolis, Greece Bloomington Heart Institute Jan J. Piek, MD, PhD, FAHA Associate Professor/Medical Professor, Cardiology Yee Guan Yap, BMedSci, Normal, Ill Director Academic Medical Center MBBS, MD, FAHA University of Texas Southwest Daniel M. Kolansky, MD, Amsterdam, the Netherlands Assoc. Professor/Head of Medical Center FAHA Cardiology Manuel A. Quiles-Lugo, MD, Dallas, Tex Associate Professor of University of Putra Malaysia FACC, FAHA Medicine Petaling Jaya, Selangor, Mark B. Effron, MD, FAHA Cardiologist University of Pennsylvania Malaysia Medical Fellow San Juan, Puerto Rico Lilly Corporate Center Philadelphia, Pa Srinivas Ramaka, MBBS, COUNCIL ON Indianapolis, Ind Robert J. Lederman, MD, MD, DM, FAHA CARDIOVASCULAR Victor A. Ferrari, MD, FAHA Consultant Cardiologist DISEASE FAHA Investigator, CV Branch Srivinvas Heart Centre IN THE YOUNG Associate Director, Non National Heart, Lung, and India Luis E. Alday, MD, FAHA Invasive Imaging Blood Institute Dimitrius Richter, MD, Head, Division of Cardiology University of Pennsylvania National Institutes of Health FAHA Hospital Acronautico Medical Center Division of Intramural Head of Cardiac Department Argentina Philadelphia, Pa Research Athens Gurolliwic Bethesda, Md Stuart Berger, MD, FAHA Athens, Greece Robert A. Harrington, MD, Medical Director, Herma Heart Pedro Lozano, MD, FAHA FAHA Spanos Vassilios, MD, FAHA Center Professor of Medicine Assistant Professor of Medicine Consultant, Interventional Professor of Pediatrics Director, Duke Clinical University of Oklahoma Health Cardiology Medical College of Wisconsin Research Institute Sciences Center Euroclinic Hospital Milwaukee, Wis Duke University Medical Oklahoma City, Okla Athens, Greece Center David Jonathan Sahn, MD, Durham, NC Wayne L. Miller, MD, PhD, Laurence S. Sperling, MD, FAHA FAHA FAHA Professor, Pediatrics John A. Hildreth, MD, Associate Professor of Director, Preventive Cardiology (Cardiology) FAHA Medicine Emory University Hospital/The Oregon Health and Science Medical Director, FPL Group Mayo Clinic and Foundation Emory Clinic University Palm Beach Gardens, Fla Rochester, Minn Atlanta, Ga Portland, Ore B5 B6 AHA Newly Elected Fellows, Spring 2007

Craig Andrew Sable, MD, Madhav Swaminathan, MD, Hirotsugu Ueshima, MD, Sherene Schlegel, Associate FAHA FASE, FAHA PhD, FAHA RN, FAHA Director, Echocardiography, Assistant Professor Professor, Health Sciences Stroke Program Clinical Fellowship Training, and Duke University Medical Shiga University of Medical Effectiveness Coordinator Telemedicine Center Science Swedish Medical Center Children’s National Medical Durham, NC Tsukinowa-cho, Otsu Seattle, Wash Center Yasuyuki Shimada, MD, Japan Washington, DC Tammy L. Cress, RN, BSN, PhD, FAHA MSN, FAHA L. LuAnn Minich, MD, FAHA Director, Cardiovascular Surgery COUNCIL FOR HIGH Stroke Program Manager University of Utah School of Yuri General Hospital BLOOD PRESSURE Swedish Medical Center Medicine Japan RESEARCH Seattle, Wash Children’s Medical Center Keiichiro Atarashi, MD, PhD, Pediatrics Afshin Andre Divani, MSc, COUNCIL ON FAHA Salt Lake City, Utah PhD, FAHA EPIDEMIOLOGY AND Physician of the Crown Director, Zeenat Qureshi PREVENTION COUNCIL ON Prince’s Household Stroke Research Center Imperial Household Agency CARDIOVASCULAR Javed Butler, MD, MPH, UMDNJ – New Jersey Medical Tokyo, Japan RADIOLOGY AND FAHA School INTERVENTION Associate Professor of Olakunle O. Akinboboye, Newark, NJ Medicine David A. Bluemke, MD, PhD, MD, MPH, MBA, FAHA Wende N. Fedder, RN, BSN, Department of MSB, FAHA Associate Director MBA, FAHA Medicine/Cardiology Clinical Director, MRI New York Hospital/Queens Director of Nursing, Stroke Emory University Associate Professor, Radiology Roslyn, NY Center and Medicine Atlanta, Ga Anil K. Bidani, MD, FAHA Alexian Brothers Hospital Johns Hopkins University J. Jeffrey Carr, MD, MSCE, Professor of Medicine and Network School of Medicine FAHA Division Director Chicago, Ill Baltimore, Md Professor/Vice Chair Loyola University Medical Division of Radiologic Elias A. Giraldo, MD, FAHA Seung Woon Rha, MD, PhD, Center and Sciences/Clinical Research Director, UTHSC Stroke FAHA Hines VA Hospital Wake Forest University Program Professor Maywood, Ill Department of Neurology Korea University Guro Hospital Durham, NC University of Tennessee Seoul, South Korea Ellen Demerath, PhD, FAHA STROKE COUNCIL Memphis, Tenn Associate Professor COUNCIL ON CV Community Health Erfan A. Albakri, MD, Keith Lowell Hull, Jr, MD, SURGERY Boonshoft School of Medicine FAHA FAHA AND NESTHESIA A Wright State University Medical Director Partner, Raleigh Neurology Jerrold Henry Levy, MD, Kettering, Ohio Florida Neuro Vascular Associates, PA FAHA James M. Galloway, MD, Institute Raleigh, NC Professor and Director, CT Tampa, Fla FAHA Robert H. Rossenwasser, Anesthesia Director, Native American Andrew Butler, PhD, PT, MD, FAHA Emory Healthcare Cardiology FAHA Professor and Chairman Atlanta, Ga Program & Senior Cardiologist Emory University School of Thomas Jefferson University Indian Health Service Yoshitaka Hayashi, MD, Medicine Philadelphia, Pa PhD, FAHA Flagstaff, Ariz Atlanta, Ga Sandeep Sachdeva, MBBS, Overseas Surgeon Jiang He, MD, PhD, FAHA Department of Cardiothoracic Professor and Chair Joseph Y. Chu, MD, FRCPC, MD, FAHA Surgery Department of Epidemiology FACP, FAHA Lead Hospitalist Stroke Monash Medical Centre Tulane University School of Neurologist Program Monash University Public Health University of Toronto Swedish Medical Center Clayton, Victoria, Australia and Tropical Medicine Toronto, Ontario, Canada Seattle, Wash New Orleans, La Hitoshi Hirose, MD, PhD, Niloufar Hadidi, RN, MS, Sean Isaac Savitz, MD, FAHA FAHA Angela Dorthea Liese, PhD, CNS, FAHA Neurologist, Stroke Service Faculty Staff MPH, FAHA Neuroscience Clinical Nurse Assistant Professor, Neurology Cardiothoracic Surgery Associate Director Specialist Beth Israel Deaconess Medical Drexel University College of Epidemiology and Biostatistics University of Minnesota Center Medicine University of South Carolina Medical Center Harvard Medical School Philadelphia, Pa Columbia, SC Shoreview, Minn Boston, Mass