Circ J 2006; 70: 530–535

Serum A as a Predictor of Cardiac Rupture in Acute Myocardial Infarction Patients Following Primary Coronary Angioplasty

Toshiro Katayama, MD; Hiroshi Nakashima, MD*; Chisa Takagi, MD*; Yukiharu Honda, MD*; Shin Suzuki, MD*; Yoshihiro Iwasaki, MD**; Tadashi Yamamoto, MD**; Masato Yoshioka, MD**; Katsusuke Yano, MD

Background The predictors of cardiac rupture (CR) in patients with acute myocardial infarction (AMI) treated with successful primary coronary angioplasty have not been identified. Methods and Results Of 433 consecutive AMI subjects who underwent reperfusion by primary coronary angioplasty within 24 h of onset, CR occurred in 11 (2.5%), free wall rupture in 9, and ventricular septal perfora- tion in 2. Rates of females, diabetes mellitus and anterior infarction were higher in the group of CR patients than in the others (p<0.05). There were no significant differences between the 2 groups in terms of left ventricular (LV) function soon after recanalization, such as LV ejection fraction, regional wall motion, or end-diastolic volume index. Plasma levels of both high-sensitivity C-reactive protein (hsCRP) and serum amyloid-A protein (SAA) were significantly higher in the CR patients than in the others (hsCRP: 6.7±6.7mg/dl vs 3.3±3.8 mg/dl, p=0.007; SAA: 699±812μg/dl vs 208±273μg/dl, p<0.0001). Multivariate analysis identified SAA as an indepen- dent predictor of CR (risk ratio: 8.8, 95%confidence interval: 1.7–25.6, p<0.05). Conclusions In patients with AMI treated with primary coronary angioplasty, inflammation may be closely related to CR, for which SAA is a useful predictor. (Circ J 2006; 70: 530–535) Key Words: Cardiac rupture; C-reactive protein; Myocardial infarction; Serum amyloid-A protein

n acute myocardial infarction (AMI), early recanaliza- of SAA are similar to those of CRP, but the clinical impli- tion of the culprit coronary artery is vital for the pres- cation of the relationship between SAA and CR in AMI I ervation of cardiac function and patient prognosis.1,2 patients is unknown. One method is direct percutaneous coronary intervention (PCI), now widely recognized as a useful and effective ther- apeutic strategy, and one associated with good outcomes Methods for the patient.3,4 Although rare, cardiac rupture (CR) is a Patient Population major adverse clinical consequence that is difficult to Between October 1999 and August 2005, 461 consecu- predict, and almost always fatal. tive patients with AMI who were admitted to hospital Current thinking suggests systemic inflammation is asso- within 24h of onset were enrolled in the study. The study ciated with ischemic heart diseases, including AMI.5,6 Pre- protocol, which included blood sampling for CRP and vious studies have demonstrated that C-reactive protein SAA, was approved by the hospital ethics committee, and (CRP), an acute phase representative protein, is associated informed consent was given by each patient and/or the with various complications in AMI, including cardiac family to 1 of the investigators before entry into the study. events.7,8 It has been reported that the peak CRP concentra- Diagnosis of AMI was based on: (1) complaint of chest tion in plasma is an important predictor of CR,9,10 but serum pain; (2) electrocardiographic ST segment elevation of amyloid A protein (SAA) is also an important acute phase >0.1mV in 2 or more limb leads, or >0.2mV in 2 or more inflammatory protein that has an expanded dynamic range precordial leads, or presentation of left-bundle branch with different kinetics compared with CRP and is reported block; and (3) elevated peak-serum creatine kinase (CK) to be a more sensitive indicator of inflammation in some more than twice the upper limit of the patient’s normal noncardiovascular conditions.11 When in vivo inflammation range. We excluded patients with renal failure on admission occurs, synthesis of SAA, as well as CRP, in the is (serum CK >2.5mg/dl), patients admitted >24h after onset, accelerated by inflammatory ; that is, the kinetics patients who died <24h after onset (ie, after blood samples had been taken) and patients who underwent drug-eluting (Received October 26, 2005; revised manuscript received January 5, stent implantation. Furthermore, to avoid other variables 2006; accepted February 15, 2006) that could influence the serum levels of high-sensitivity Department of Cardiovascular Medicine and Course of Medical and CRP (hsCRP) or SAA, we excluded patients with a history Dental Science, Graduate School of Biomedical Science, Nagasaki of recent surgery or trauma within the preceding 2 months, University, *Department of Cardiology, Nagasaki Citizens Hospital and **Department of Cardiology, Nagasaki Kouseikai Hospital, malignancy or liver cirrhosis, acute or chronic inflammatory Nagasaki, Japan disease, or collagen disease. We had no patients with Mailing address: Toshiro Katayama, MD, 2-5-5-1615 Akasako, primary or secondary . In the end, 433 is the Nagasaki 852-8156, Japan. E-mail: [email protected] number of patients after exclusions. There were 224 cases

Circulation Journal Vol.70, May 2006 Predictors of Cardiac Rupture in AMI 531

Table 1 Patient and Lesion Characteristics in Cases of CR

Case Age Culprit location SAA Peak CK BNP hs-CRP at 24 h Day of occurrence Type Previous Diabetes Hyper Gender Hypertension Outcome no. (years) (segment no.) (μg/dl) (IU/L) (pg/ml) since onset (mg/dl) of CR of CR MI mellitus lipidemia 1 F 70 6 371 12,199 325 6.79 4 FWR(B) – – + – Dead 2 M 88 5 546 9,425 195 11.28 3 FWR(B) – – + – Dead 3 M 69 5 373 2,039 73 1.00 7 VSP – + + + Alive 4 M 71 6 493 2,748 434 0.15 3 FWR(B) – + – – Dead 5 F 89 7 940 8,414 221 12.32 2 FWR(O) – – – – Alive 6 M 56 6 328 2,311 183 4.11 5 VSP + + – + Alive 7 F 67 7 460 3,426 461 4.01 6 FWR(B) – + + + Dead 8 F 81 6 1,605 1,110 136 14.27 2 FWR(B) – + – – Alive 9 F 85 2 110 4,056 441 5.02 9 FWR(O) – – + – Dead 10 F 75 7 592 4,263 748 13.86 2 FWR(O) + – + + Alive 11 F 91 11 1,603 2,601 230 3.23 3 FWR(B) – + + + Dead CR, cardiac rupture; segment no., location of coronary artery by American Heart Association; SAA, serum amyloid A protein; Peak CK, peak value of creatine kinase; BNP, brain natriuretic peptide; CRP, C-reactive protein; Previous MI, previous myocardial infarction; FWR, free wall rupture; B, blow-out type free wall rupture; VSP, ventricular septal perforation; O, oozing type free wall rupture. of anterior myocardial infarction (MI). type), ventricular septal perforation (VSP) and papillary muscle rupture. The diagnosis of free wall rupture was Treatment Strategy based on echocardiography followed by pericardiocentesis Following oral administration of 200mg of aspirin and or surgical or postmortem examination. The blow-out type 200mg of ticlopidine, all patients underwent direct PCI of rupture was defined as a free wall rupture with electro- within 2h of admission. For 72h after reperfusion therapy, mechanical dissociation, and sudden onset of severe hypo- all patients were given intravenous nicorandil (96mg/day) tension without ECG change. The oozing type rupture was and heparin (10,000units/day). We did not use abciximab. defined as a gradual bleeding into the pericardial sac, causing progressive or recurrent cardiac tamponade. VSP Blood Sampling was diagnosed on the basis of abnormal shunt-flow at the On admission, blood samples were taken while patients interventricular septum on color Doppler echocardiogra- were supine, for analysis of hsCRP, CK, and troponin-T. phy. Our follow-up period was 30 days after onset of AMI. For 3 days after admission hsCRP was serially determined The primary endpoint was CR or cardiac death. every 24h, and CK every 4h. At 24h after onset, blood samples were taken, again while patients were supine, for Statistical Analysis analysis of brain natriuretic peptide (BNP), atrial natri- Values are expressed as mean±standard-deviation. All uretic peptide (ANP), renin, aldosterone, epinephrine, statistical tests were 2-tailed, and a p-value <0.05 was con- norepinephrine, and SAA. The SAA assay by the latex sidered statistically significant. The chosen cut-off points agglutination nephelometric immunoassay was contracted for the concentrations of the hsCRP at 24-h since onset, out (BioMedical Laboratories Inc). SAA, peak-CK, and BNP were based on the highest tertiles in the overall sample. Multivariate logistic regression Hemodynamics and Cardiac Function analyses were performed, with the independent variables On admission, we used the Killip classification to assess including age, gender, diabetes mellitus, culprit lesion loca- the severity of the patient’s condition. We also evaluated tion, the number of vessels involved, Killip classification major complications, defined as ventricular arrhythmia on admission, ANP, BNP, peak-CK value, hsCRP at 24-h (except reperfusion arrhythmia), pulmonary edema, car- since onset, and SAA. The odds ratios and 95%confidence diogenic shock, sub-acute thrombosis and cardiac death. intervals (CI) were also calculated. We defined cardiogenic shock as systolic blood pressure <90 mmHg (without inotropic or intra-aortic balloon (IABP) support), signs of hypoperfusion (cold extremities, Results impaired mental status, or urine output <30 ml/h), and A total of 433 patients with AMI were analyzed (291 evidence of pulmonary congestion on chest X-ray. male, 142 female; mean age 69±12 years). Before PCI ther- All subjects, except 29 with cardiogenic shock, under- apy, simple aspiration thrombectomy catheters were used went left ventriculography soon after recanalization from a in 184 patients with massive visible thrombus, but distal single plane with a right anterior deviation of 30° to evalu- protection devices were not used. Conventional plain old ate left ventricular (LV) ejection fraction (EF), to calculate balloon angioplasty was sufficient in 85 patients, and the the LV end-diastolic volume index (LVEDVI) using the remaining 348 underwent coronary stenting. Of the total, area-length method, and to evaluate regional wall motion 420 patients attained TIMI III grade flow, and the remaining (RWM), using the center-line method.12 We excluded 13 patients, TIMI II grade flow. Details of the 11 patients posterior infarctions from our evaluation because of the (2.5%) in whom CR occurred are summarized in Table1: associated issues of inaccuracy. We also assessed LV end- free wall rupture occurred in 9 (6 patients were blow-out diastolic pressure (LVEDP) at that time. We used the type, 3 patients were oozing type) and VSP in 2. None had Thrombolysis in Myocardial Infarction (TIMI) classifica- PMR. Six patients died and 2 underwent surgery. tion to evaluate epicardial coronary flow. The characteristics of the 2 groups (with or without CR) are shown in Table2. There were no significant differences Endpoints and Definitions between them in terms of time elapsed from onset, or the CR included free wall rupture (blow-out type, oozing existence of pre-infarction angina pectoris within 24h of

Circulation Journal Vol.70, May 2006 532 KATAYAMA T et al.

Table 2 Comparison of Patients With or Without CR

CR (+) CR (–) p value (n=11) (n=422) Age (years) 76±11 69±12 0.06 Gender, male (%) 4 (36) 287 (68) 0.03 Hypertension (%) 7 (64) 236 (56) NS Diabetes mellitus (%) 7 (64) 139 (33) 0.04 Hyperlipidemia (%) 5 (44) 181 (43) NS Time elapsed (h) 2.2±3.3 5.9±6.0 NS Angina within 24 h (%) 3 (27) 207 (49) NS Previous MI (%) 2 (18) 59 (14) NS

Continuous values expressed as mean±SD. CR, cardiac rupture; CR (+), patients with CR; CR (–), patients without CR; Time elapsed, time from onset to admission; Angina within 24 h, existence of pre-infarction angina pectoris within 24 h of onset; Previous MI, previous myocardial infarction.

Table 3 Laboratory Findings in Patients With or Without CR

CR (+) CR (–) p value (n=11) (n=422) BNP (pg/ml) 313±194 361±462 NS ANP (pg/ml) 36±28 63±80 NS Peak-CK (IU/L) 4,778±3,575 2,461±2,661 0.005 hs-CRP on admission (mg/dl) 1.80±2.35 1.30±2.84 NS hs-CRP at 24 h since onset (mg/dl) 6.91±5.15 3.32±3.87 0.003 Peak hs-CRP (mg/dl) 11.94±5.03 7.38±6.42 0.04 SAA (μg/dl) 715±502 206±294 <0.0001 Renin (ng·ml–1·h–1) 10.1±13.8 4.6±6.7 0.01 Aldosterone (pg/ml) 99±66 72±73 NS Norepinephrine (pg/ml) 2,562±4,426 1,757±1,003 NS Epinephrine (pg/ml) 124±80 200±176 NS Troponin-T (ng/ml) 0.31±0.55 0.50±0.70 NS TC (mg/dl) 198±33 203±41 NS Lp (a) (mg/dl) 32±13 32±26 NS HDL-C (mg/dl) 42±10 51±16 NS

Continuous values expressed as mean±SD. ANP, atrial natriuretic peptide; hs-CRP, high-sensitivity C-reactive protein; TC, total ; Lp (a), (a); HDL-C, high-density lipoprotein cholesterol. Other abbreviations as shown in Tables 1,2.

Table 4 Hemodynamic Data on Admission for Patients with AMI

CR (+) CR (–) p value (n=11) (n=422) HR (beats/min) 91±30 81±20 NS Systolic BP (mmHg) 115±47 121±33 NS Killip classification (%) NS I 5 (45) 244 (58) II 3 (27) 86 (20) III 1 (9) 42 (10) IV 2 (18) 50 (12) Culprit vessel (%) 0.03 LMT 2 (18) 7 (2) LAD 7 (64) 208 (49) RCA 1 (9) 126 (30) LCX 1 (9) 81 (19) Multivessel (%) 3 (27) 181 (43) NS IABP use (%) 7 (64) 30 (7) <0.0001 Spontaneous recanalization (%) 3 (27) 148 (35) NS Collateral circulation (%) 1 (9) 76 (18) NS

Continuous values expressed as mean±SD. AMI, acute myocardial infarction; HR, heart rate; Systolic BP, systolic blood pressure; Culprit vessel, coronary artery responsible for AMI; LMT, left main trunk; LAD, left anterior descending artery; RCA, right coronary artery; LCX, left circumflex coronary artery; Multivessel, multivessel disease; IABP use, intra-aortic balloon pumping; Spontaneous recanalization, existence of sponta- neous recanalization (Thrombolysis in Myocardial Infarction II or III); Collateral circulation, good collateral circulation (Rentrop grade II or III). Other abbreviations as shown in Tables 1,2.

Circulation Journal Vol.70, May 2006 Predictors of Cardiac Rupture in AMI 533

Table 5 Major Complications

CR (+) CR (–) p value (n=11) (n=422) AC 0 1 (0.2) NS SAT 1 (9) 9 (2) NS Pulmonary edema 9 (82) 55 (13) <0.0001 Cardiogenic shock 9 (82) 45 (11) <0.0001 Cardiac death 6 (55) 21 (5) <0.0001

Continuous values expressed as mean±SD; values in parentheses as percentage. AC, abrupt vessel closure; SAT, sub-acute stent thrombosis. Other abbreviations as shown in Table 2.

Table 6 Left Ventricular Function in the Acute Phase

CR (+) CR (–) p value (n=9) (n=315) LVEF (%) 48±11 51±12 NS RWM (SD/chord) –2.5±1.2 –2.2±1.1 NS LVEDVI (ml/m2)61±17 67±18 NS LVEDP (mmHg) 17±8 15±7 NS

Continuous values expressed as mean±SD. LVEF, left ventricular ejection fraction; RWM, regional wall motion; LVEDVI, left ventricular end-diastolic volume index; LVEDP, left ventricular end-diastolic pressure. Other abbreviations as shown in Table 2.

Table 7 Predictors of CR by Multivariate Analysis

Odds ratio 95%CI p value H-SAA 3.8 1.7–25.6 0.01 H-hs-CRP at 24 h since onset 8.8 0.9–12.6 0.06 Gender female 3.0 0.6–11.6 0.09 Anterior infarction 2.6 0.6–10.7 0.09 H-Peak-CK 2.0 0.5–8.8 NS DM 1.8 0.5–9.0 NS Age 1.7 0.4–7.9 NS Killip class. on admission 1.0 0.3–8.9 NS H-BNP 1.1 0.2–6.4 NS CI, confidence interval; H-SAA, serum amyloid A protein levels in the upper tertile; H-hs-CRP, high sensitivity C-reactive protein levels in the upper tertile; H-Peak-CK, peak-creatine kinase levels in the upper tertile; DM, diabetes mellitus; Killip class. on ad- mission, Killip calssification on admission; H-BNP, brain natriuretic peptide levels in the upper tertile. onset, or previous MI. Rates of female (64% vs 32%, p<0.0001), cardiogenic shock (82% vs 11%, p<0.0001), p<0.05) and diabetes mellitus (64% vs 33%, p<0.05) were and cardiac death (55% vs 5%, p<0.0001) in the group with higher in the CR group and there was a tendency for CR. patients with CR to be older, but it was not a significant Table6 shows LV function in the acute phase, soon after trend. recanalization. There were no significant differences in Table3 is a comparison of the laboratory findings for the LVEF, RWM, LVEDVI, or LVEDP. 2 groups. Plasma levels of peak-CK, hsCRP, and SAA The follow-up period was 30 days, during which CR were significantly higher in the CR group (peak-CK: occurred in 11 patients (2.5%). The predictors of CR by 4,778±3,575 IU/L vs 2,461±2,661 IU/L, p<0.01; hsCRP: multivariate analysis in these AMI patients are shown in 6.91±5.15mg/dl vs 3.32±3.87mg/dl, p<0.01; SAA: 715± Table7. When the CR subjects were divided into 2 groups 502μg/dl vs 206±294μg/dl, p<0.0001). However, there according to the threshold of the upper tertile (290μg/dl) of were no significant differences in the plasma levels of SAA, only 1 patient with CR was in the low SAA group. BNP. Multivariate analysis identified SAA concentration as an Hemodynamic data at the time of admission are shown independent predictor of CR during 30 days (risk ratio: 8.8, in Table4. There were no significant differences between 95%CI: 1.7–25.6, p=0.01). the 2 groups for heart rate, systolic blood pressure, Killip classification and culprit vessel location. There were sig- nificantly more patients with anterior infarction in the CR Discussion group (p<0.05) and significantly more frequent use of There are some patients with AMI who have a poor IABP (p<0.0001). clinical course despite early recanalization therapy by PCI. Major complication rates in the first 30 days are shown in There were some reports that CR occurs in 1–4% of cases Table5. There were significantly more patients with major of AMI,9,13,14 but although rare, CR is a major adverse clini- complications, such as pulmonary edema (82% vs 13%, cal consequence with a high mortality and is the second

Circulation Journal Vol.70, May 2006 534 KATAYAMA T et al. most common cause of in-hospital death, after pump dominant on plasma high density lipopro- failure.15,16 To date, however, it remains difficult to predict. tein (HDL)-cholesterol particles, where it was thought to be In our study, there were no significant differences in apolipoprotein A-1, and after HDL-mediated cholesterol terms of acute phase Killip classification or BNP concen- delivery to cells.26 These observations may explain its tration between subjects with or without CR. It is difficult increased concentration in patients with acute coronary to predict CR from the severity of either hemodynamics or syndromes. Accordingly, SAA may be a more sensitive heart failure on admission marker of disease severity and may represent a different There are some reports that thrombolytic therapy in- type of acute-phase response than CRP. creases the risk of CR,14,17,18 which is why we did not use it. There is 1 report that CRP is an excellent marker for And yet, CR occurred in 2.5% of the study population, sig- plaque instability or post-stent inflammatory status, and nificantly more frequently in women than men. This con- that its source might be the inflamed site of the plaque or firmed previous data showing that the female gender is the coronary arterial wall injured by stenting.27 Furthermore, independently associated with adverse outcomes including coronary levels of SAA at the site of plaque rupture were CR after AMI.19 The predisposition of female patients to increased relative to the systemic circulation, indicating CR is attributed to a susceptible collagen framework and local production of biologically active inflammatory medi- differences in the collagen matrix within the infarcted myo- ators. In AMI, the acute-phase reactant SAA is produced cardium.20 at the site of coronary occlusion by macrophages and Rupture occurs when transmural pressure exceeds the passively vascular cells.24 tensile strength of the heart’s connective tissue substrate. To date, the exact mechanism of CR remains poorly Increased intramural pressure from bleeding within the understood. It can be precipitated by infarct expansion, reperfused AMI is a factor in rupture. Two types of col- which, in turn, is influenced by infarct healing and wall lagen damage have been identified.21 The first is loss of stress.28 This complication usually occurs during the first birefringence, observed in silver-impregnated sections. week after a transmural infarction when the necrotic myo- Degeneration of ground substance attached to collagen has cardium is vulnerable to wall stress.28 During this period, been reported after myocardial ischemia.22 The second type pathological specimens show a central core of necrotic is loss of the collagen struts that form lateral connections myocardium surrounded by a zone of hemorrhage and between neighboring myocytes.21 acute .29 The degree of polymorphonuclear It has been reported that the peak-CRP concentration in infiltration in the area of infarction could correlate with the plasma is an important predictor of CR in patients with incidence of CR. Severe inflammation resulting in tissue AMI.9,10 Intramyocardial hemorrhage and pericarditis, vulnerability is a possible cause of the extreme elevation in resulting from slow leakage of blood into the pericardial serum SAA level in patients with CR. In recent reports, space early after MI, can precipitate local cardiac inflam- predictors of CR were older age, female gender, anterior mation, causing marked and persistent elevation of serum location, and thrombolysis. In the present study, however, CRP. As for SAA, we reported that acute-phase plasma these were not independent predictors by multivariate SAA concentrations after successful direct PCI are closely analysis. related to major adverse complications, including mortali- ty.23 However, the clinical implication of the relationship Study Limitations between inflammation and CR in AMI patients is unknown. First, the small size of the sample population means that The present study demonstrated a significant association our results require confirmation in a large-scale trial before between the SAA level in the acute phase of AMI and CR. any concrete significance can be extrapolated. Second, we Furthermore, the results also clarified that SAA is a predic- excluded from the study patients with renal failure, on the tor of CR. basis that they would not be able to tolerate coronary It has been reported that the plasma SAA concentration angioplasty. This may have introduced bias into our results. is elevated from several hours to a few days after the onset Third, the follow-up was short, at 30 days after onset, of AMI,24 and reaches its peak value approximately 24h which does not allow for long-term prognosis. Fourth, both after onset.24 We considered that the peak value of SAA anterior and inferior MI were included in the evaluation of was the most accurate indicator of the severity of AMI, cardiac function. Fifth, the value of SAA was assayed only which is why we took blood samples 24h after onset. at a single time point. Therefore, the present study was not The exact mechanism for the association between in- always satisfactory in either the number of patients or the flammation and future cardiovascular risk is not known. observation period, and therefore, more patients will need Laboratory studies have demonstrated that inflammation to be accumulated in a future study that includes long-term plays a pathophysiological role in atherogenesis and may prognosis. promote the development of atherosclerotic plaques in coronary arteries. Acute-phase reactants may play a patho- physiological role in atherosclerotic plaque instability as Conclusions manifested by elevated hsCRP levels in the presence of We have shown that in patients with AMI treated with ruptured coronary artery plaques. Similar to hsCRP, SAA primary coronary angioplasty, elevation of inflammation is synthesized in the liver in response to infection and in- markers such as hsCRP and SAA may be closely related to flammation and thus is also a representative inflammatory the occurrence of CR. Furthermore, the acute-phase plasma reactant. In AMI the inflammatory -6 SAA concentration may be a useful predictor of CR. and the acute-phase reactant SAA are produced at the site of coronary occlusion by macrophages and possibly vascu- References lar cells, whereas CRP is produced mainly in the liver and 1. Brodie BR, Stuckey TD, Muncy DB, Hansen CJ, Wall TC, Pulsipher taken up locally by phagocytosing white blood cells.25 Dur- M, et al. Importance of time-to-reperfusion in patients with acute ing the acute-phase reaction, SAA is secreted as the pre- myocardial infarction with and without cardiogenic shock treated

Circulation Journal Vol.70, May 2006 Predictors of Cardiac Rupture in AMI 535

with primary percutaneous coronary intervention. Am Heart J 2003; free wall or ventricular septum among necropsy cases of fatal acute 145: 708–715. myocardial infarction since introduction of coronary care units. Am J 2. Sakurai K, Watanabe J, Iwabuchi K, Koseki Y, Kon-no Y, Fukuchi Cardiol 1989; 63: 906–911. M, et al. Comparison of the efficacy of reperfusion therapies for early 16. Pollak H, Nobis H, Mlczoch J. Frequency of left ventricular free wall mortality from acute myocardial infarction in Japan. Circ J 2003; 67: rupture complicating acute myocardial infarction since the advent of 209–214. thrombolysis. Am J Cardiol 1994; 74: 184–186. 3. Degeare VS, Dangas G, Stone GW, Grines CL. Interventional proce- 17. Bueno H, Martinez-Selles M, Perez-David E, Lopez-Palop R. Effect dures in acute myocardial infarction. Am Heart J 2001; 141: 15–24. of thrombolytic therapy on the risk of cardiac rupture and mortality 4. Aversano T, Aversano LT, Passamani E, Knatterud GL, Terrin ML, in older patients with first acute myocardial infarction. Eur Heart J Williams DO, et al. Thrombolytic therapy vs primary percutaneous 2005 Apr 26; (Epub ahead of print). coronary intervention for myocardial infarction in patients presenting 18. Okino S, Nishiyama K, Ando K, Nobuyoshi M. Thrombolysis in- to hospitals without on-site cardiac surgery: A randomized controlled creases the risk of free wall rupture in patients with acute myocardial trial. JAMA 2002; 287: 1943–1951. infarction undergoing percutaneous coronary intervention. J Interv 5. Stefanadis C, Diamantopoulos L, Dernellis J, Economou E, Tsiamis Cardiol 2005; 18: 167–172. E, Toutouzas K, et al. Heat production of athelosclerotic plaques and 19. Solodky A, Behar S, Herz I, Assali A, Porter A, Hod H, et al. Com- imflammation assesed by the acute phase in acute coronary parison of incidence of cardiac rupture among patients with acute syndromes. J Mol Cell Cardiol 2000; 32: 43–52. myocardial infarction treated by thrombolysis versus percutaneous 6. Kosuge M, Kimura K, Ishikaewa T, Shimizu T, Takamura T, transluminal coronary angioplasty. Am J Cardiol 2001; 87: 1105– Tsukahara K, et al. Relation between white blood cell counts and 1108. myocardial reperfusion in patients with recanalized anterior acute 20. Becker RC, Hoshman JS, Cannon CP, Spencer FA, Ball SP, Rizzo myocardial infarction. Circ J 2004; 68: 526–531. MJ, et al. Fatal cardiac rupture among patients treated with throm- 7. Berton G, Cordiano R, Palmieri R, Pianca S, Pagliara V, Palatini P. bolytic agents and adjunctive thrombin antagonists: Observations C-reactive protein in acute myocardial infarction: Association with from the Thrombolysis and Thrombin Inhibition in Myocardial heart failure. Am Heart J 2003; 145: 1094–1101. Infarction 9 Study. J Am Coll Cardiol 1999; 33: 479–487. 8. Sano T, Tanaka A, Namba M, Nishibori Y, Nishida Y, Kawarabayashi 21. Whittaker P, Boughner DR, Klonar RA. Role of collagen in acute T, et al. C-reactive protein and lesion morphology in patients with myocardial infarct expansion. Circulation 1991; 84: 2123–2134. acute myocardial infarction. Circulation 2003; 108: 282–285. 22. Sato S, Ashraf M, Millard RW, Fujiwara H, Schwartz A. Connective 9. Yip HK, Wu CJ, Chang HW, Wang CP, Cheng CI, Chua S, et al. Car- tissue changes in early ischemia of porcine myocardium: An ultra- diac rupture complicating acute myocardial infarction in the direct structural study. J Mol Cell Cardiol 1983; 15: 261–275. percutaneous coronary intervention reperfusion era. Chest 2003; 23. Katayama T, Nakashima H, Takagi C, Honda Y, Suzuki S, Iwasaki Y, 124: 565–571. et al. Prognostic value of serum amyloid A protein in patients with 10. Anzai T, Yoshikawa T, Shiraki H, Asakura Y, Akaishi M, Mitamura acute myocardial infarction. Circ J 2005; 69: 1186–1191. H, et al. C-reactive protein as a predictor of infarct expansion and 24. Casl MT, Surina B, Glojnaric I, Pape E, Jagarinec N, Kranjcevic S. cardiac rupture after a first Q-wave acute myocardial infarction. Serum amyloid A protein with acute myocardial infarction. Ann Clin Circulation 1997; 96: 778–784. Biochem 1995; 32: 196–200. 11. Malle E, De Beer FC. Human serum amyloid A protein: A prominent 25. Maier W, Altwegg LA, Corti R, Gay S, Hersberger M, Mary FE, et al. acute-phase reactant for clinical practice. Eur J Invest 1996; 26: Inflammatory markers at the site of ruptured plaque in acute myocar- 427–435. dial infarction: Locally increased interleukin-6 and serum amyloid A 12. Florence HS, Joachim S, Detlef GM, Mirle AK, Hugh S, Edward but decreased C-reactive protein. Circulation 2005; 111: 1355 – LB, et al. Measurement of regional wall motion from biplane con- 1361. trast ventriculograms: A comparison of the 30 degree right anterior 26. Pussinen PJ, Malle E, Mesto J, Sattler W, Raynes JG, Jauhiainen M. oblique projections in patients with acute myocardial infarction. Acute-phase HDL in phospholipids transfer protein (PLTP)- medi- Circulation 1986; 74: 796–804. ated HDL conversion. 2001; 155: 297–305. 13. Ikeda N, Yasu T, Kubo N, Hirahara T, Sugawara Y, Kobayashi N, et 27. Inoue T, Kato T, Uchida T, Sakuma M, Nakajima A, Shibazaki M, al. Effect of reperfusion therapy on cardiac rupture after myocardial et al. Local release of C-reactive protein from vulnerable plaque or infarction in Japanese. Circ J 2004; 68: 422–426. coronary arterial wall injured by stenting. J Am Coll Cardiol 2005; 14. Moreno R, Lopez-Sendon J, Garcia E, Perez de Isla L, Lopez de Sa 46: 239–245. E, Ortega A, et al. Primary angioplasty reduces the risk of left ven- 28. Shuster EH, Bulkley BH. Expansion of transmural myocardial in- tricular free wall rupture compared with thrombolysis in patients farction: A pathophysiologic factor in cardiac rupture. Circulation with acute myocardial infarction. J Am Coll Cardiol 2002; 39: 598– 1979; 60: 1532–1538. 603. 29. Richard V, Murry CE, Reimer KA. Healing of myocardial infarcts in 15. Reddy SG, Roberts WC. Frequency of rupture of the left ventricular dogs: Effect of late reperfusion. Circulation 1995; 92: 1891–1901.

Circulation Journal Vol.70, May 2006