Pulmonary Hypertension: Newer Insights

4 : 4 S Bhandari, PS Sandhu, New Delhi INTRODUCTION Most practicing cardiologists see patients with pulmonary hypertension (PH) on a regular basis. The explosion in knowledge of and treatment for PH over the past decade obligates cardiologists to be more cognizant of this disorder. PH has been defined as a resting mean pulmonary arterial pressure (mPAP) >25 mm Hg, or >30 mm Hg with exercise. The subgroup of PH known as pulmonary arterial hypertension (PAH) adds the criterion that the pulmonary arterial wedge pressure must be ≤15 mm Hg.1 CLASSIFICATION The clinical classification of PH has gone through a series of changes since the first version was proposed in 1973 at the first international conference on primary pulmonary hypertension endorsed by the WHO (Table 1).2 The initial classification designated only 2 categories, PPH or secondary PH. 25 years later, the 2nd World Symposium on PAH was held in Evian, France. The “Evian classification” attempted to create categories of PH that shared pathologic and clinical features as well as similar therapeutic options 3 The 3rd World Symposium on PAH was held in Venice, Italy, 5 years after the Evian conference. At this conference modest changes were made. The most notable change was to abandon the term PPH in favor of idiopathic pulmonary arterial hypertension (IPAH); familial PAH if there is a family history of PAH; or associated PAH if another cause, such as CTD or HIV, is present. The term “secondary PH” was abandoned since it did not help with diagnosis or in directing treatment.4 During the 4th World Symposium on PH held in 2008 in Dana Point, California, the general philosophy and organization of the Evian-Venice classifications was maintained, but modified it to clarify some areas that were unclear. Pathophysiology of PH Different pathological5,6,8–10 features characterize the diverse clinical PH groups. • Group 1, PAH: The increase in PVR is related to different mechanisms, including vasoconstric- tion, proliferative and obstructive remodeling of the pulmonary vessel wall, inflammation, and thrombosis. • Group 1′: includes mainly PVOD which involves septal veins and pre-septal venules with oc- clusive fibrotic lesions, venous muscularization, capillary proliferation (patchy), pulmonary oedema, occult alveolar haemorrhage, lymphatic dilatation and lymph node enlargement (vas- cular transformation of the sinus), and inflammatory infiltrates. • Group 2, PH due to left heart disease. • Group 3, PH due to lung diseases and/or hypoxia. The pathophysiological mechanisms involved in this setting are hypoxic vasoconstriction, mechanical stress of hyperinflated lungs, loss of capillaries, and inflammation.

181

Medicine Update 2012  Vol. 22

Table 1 Updated clinical classification of PH (Dana Point, • Group 5, PH with unclear and/or multifactorial mecha- 2008) nisms. 1 Pulmonary arterial hypertension (PAH) Natural History and Prognostic Factors 1.1 Idiopathic The natural history of IPAH was well described by the NIH 1.2 Heritable registry in 1985 before the availability of any disease-specific 1.2.1 BMPR2 therapy.11 The median survival was 2.8 years, with 1, 3, and 1.2.2 ALK1, endoglin (with or without hereditary haemorrhagic 5-year survival rates of 68%, 48%, and 34%, respectively.12 telangiectasia) Associated conditions influence outcomes: Patients with CTD 1.2.3 Unknown and HIV-associated PAH tend to have a worse prognosis, 1.3 Drugs and toxins induced whereas those with congenital heart disease–associated PAH 1.4 Associated with (APAH) tend to have a better prognosis. 1.4.1 Connective tissue diseases(CTD) Important prognostic indicators in PAH include symptoms, 1.4.2 HIV infection exercise endurance, and hemodynamics.13Most of these 1.4.3 Portal hypertension prognostic variables are related to RV function. In the NIH 1.4.4 Congenital heart disease registry, the median survival among patients presenting with 1.4.5 Schistosomiasis class I and II symptoms was ~ 6 years versus 2.5 years for patients 1.4.6 Chronic haemolytic anaemia with class III symptoms and just 6 months for patients who presented with class IV symptoms.12 Two large retrospective 1.5 Persistent pulmonary hypertension of the newborn series have confirmed the importance of functional class 1′ Pulmonary veno-occlusive disease and/or pulmonary capillary 14,15 haemangiomatosis as a prognostic variable, even during treatment. Among IPAH patients treated with epoprostenol, prognosis was worse 2 Pulmonary hypertension due to left heart disease for patients who commenced therapy with more advanced 2.1 Systolic dysfunction symptoms. Moreover, in both series, patients who improved 2.2 Diastolic dysfunction to class I or II status after 3 to 17 months of epoprostenol 2.3 Valvular disease therapy had a better prognosis than patients who remained 3 Pulmonary hypertension due to lung diseases and/or hypoxia in class III or IV. 3.1 Chronic obstructive pulmonary disease Exercise tolerance in PAH is commonly assessed by means 3.2 Interstitial lung disease of the 6-minute-walk distance (6MWD). In one of the first 3.3 Other pulmonary diseases with mixed restrictive and obstructive controlled trials, a 6MWD of <150 m was associated with a pattern very poor prognosis. 16 In a series of 178 IPAH patients treated 3.4 Sleep-disordered breathing with epoprostenol, those who walked further than the median 3.5 Alveolar hypoventilation disorders value of 380 m after 3 months of therapy had a better prognosis 16 3.6 Chronic exposure to high altitude than those who did not. 3.7 Developmental abnormalities Progress of medical treatment in pul- 4 Chronic thromboembolic pulmonary hypertension monary arterial hypertension 5 PH with unclear and/or multifactorial mechanisms In 1891, Ernst von Romberg, a German physician, described 5.1 Haematological disorders: myeloproliferative disorders, splenec- an autopsy subject as having “pulmonary vascular sclerosis”; tomy. however, it is only since 1995 with the introduction of 5.2 Systemic disorders: sarcoidosis, pulmonary Langerhans cell his- epoprostenol that disease-specific targeted medical therapies tiocytosis, lymphangioleiomyomatosis, neurofibromatosis, vascu- for PAH have become available. Furthermore, significant litis advances in the treatment of PAH have occurred during the 5.3 Metabolic disorders: glycogen storage disease, Gaucher disease, thyroid disorders past 15 years. Currently 9 medical therapies have received regulatory approval. These agents target the prostacyclin 5.4 Others: tumoural obstruction, fibrosing mediastinitis, chronic re- nal failure on dialysis pathway, the nitric oxide pathway, and the endothelin pathway. Combination trials have demonstrated additive or synergistic ALK-1 = activin receptor-like kinase 1; BMPR2 = bone morphogenetic pro- tein receptor 2, HIV = human immunodeficiency virus. benefit by targeting 2 or all 3 of these pathways (Figs. 1 and 2). Conventional Treatments • Group 4, CTEPH: non-resolution of acute embolic masses which later undergo fibrosis leading to mechani- Diuretics role has been limited to patients manifesting RV cal obstruction of pulmonary arteries. failure. However, patients with advanced PAH can have increased left ventricular filling pressures that contribute 182 Pulmonary Hypertension: Newer Insights

Terbogrel STEP its vasodilator and antithrombotic effects. Patients may have Treprostinil Sildenafil a reduction in PVR of ≥50%, even if no acute hemodynamic AIR Sildenafil STRIDE2 ARIES Epoprostenol SUPER PHIRST BREATHE1 BREATHE2 COMBI EARLY effects are noted. Epoprostenol is administered through Epoprostenol SERAPH TRIUMPH ALPHABET STRIDE1 BREATHE5 PACES a central venous catheter that is surgically implanted and Epoprostenol Bosentan Beraprost delivered by an ambulatory infusion system. 1990 1996 2000 ‘01 ’02 ‘03 ’04 ‘05 ’06 ‘08 ’09 ‘10 Most complications are due to the delivery system and include Monotherapy catheter-related infections and temporary interruption of the Monotherapy and/or sequential combination infusion because of pump malfunction, which causes rebound Upfront combination pulmonary hypertension. Side effects related to epoprostenol Colour code identifies the design of the study: monotherapy (black): in- are flushing, headache, nausea, diarrhea, and a jaw discomfort vestigational drug vs. placebo in patients naive for PAH approved drugs. that occurs with eating. In most patients, these symptoms are Monotherapy and/or sequential combination (red): investigational drug vs. placebo in patients either naive for, or treated with PAH approved drugs. minimal and well tolerated. Upfront combination (light blue): single drug vs. combination of two drugs in patients naive for pulmonary arterial hypertension approved drugs. The experience with epoprostenol in patients with IPAH for more than 10 years has been reported by two large Fig. 1: Time-course of 25 published RCT in pulmonary arterial centers.14,15 Survival rates markedly improved; predictors of hypertension survival included NYHA class, exercise tolerance, and acute vasodilator responsiveness. Both studies provided important to the symptoms of dyspnea and orthopnea, which can be data for identifying patients who would do well over the relieved with diuretics. long term, versus those in whom transplantation should be Because hypoxemia is a potent pulmonary vasoconstrictor, considered. oxygen supplementation is given to maintain SPO2 > 90%. Treprostinil is a stable prostacyclin analogue that has Anticoagulants The evidence for favorable effects of oral pharmacological actions similar to those of epoprostenol, anticoagulant treatment is based on retrospective analyses from but differs in that it is chemically stable at room temperature 7 studies, of which 5 were positive and 2 were negative.18,20,22 and neutral pH and has a longer half-life. It is administered The survival of anticoagulated patients selected on the basis through continuous S/C infusion. In a large RCT in patients of clinical judgment was improved from 21% to 49%. Most with PH, it was effective in increasing 6MWD, decreasing experts recommend warfarin anticoagulation targeted to an dyspnea, and hemodynamics.25 INR of 2.0 to 3.0. Iloprost, an inhalational analogue of prostacyclin. In RCT, Short-term IV digoxin in IPAH produces a modest increase inhaled iloprost was shown to have an acute effect on in cardiac output and a significant reduction in circulating hemodynamics similar to those of inhaled nitric oxide and norepinephrine21; no data is available on the effects of long- when given chronically, causes improvement in 6MWD, and term treatment. Therefore, the use is based primarily on the in hemodynamics.26 Because of the short half-life of iloprost, judgment of the physician. however, it requires frequent (up to 12/day) inhalations. CCBs Favorable clinical and prognostic effects of high doses Beraprost is an orally active prostacyclin analogue that has of oral CCB drugs in acutely vasoreactive patients with been evaluated in RCT trials in patients with PAH. In one IPAH have been shown in single-center, nonrandomized, large European trial (ALPHABET study), beraprost improved uncontrolled studies.17, 22 exercise capacity and symptoms over a 12-week period but Acute vasodilator testing is recommended for all PAH had no significant effect on hemodynamics or functional patients, even though patients with IPAH and anorexigen class. A similar trial conducted in the United States, however, induced PAH are more likely to respond. Empirical treatment showed similar efficacy at 12 weeks, only to document the 27 with CCBs without a positive response with acute vasodilator loss of effectiveness over 1 year. At present, beraprost is only testing using either inhaled nitric oxide or IV epoprostenol is approved for use in Japan. contraindicated.23 Endothelin Receptor Blockers (ERA) Synthetic prostacyclin and prostacyc- ET-1 exerts vasoconstrictor and mitogenic effects and is lin analogues activated in PAH. Three endothelin receptor blockers have Prostacyclins have been found to be effective in the therapy been approved for PAH. Although there have never been direct 28 of PAH.15 Continuous IV infusion of epoprostenol has been comparative trials, all three appear to have similar efficacy. shown in RCT to improve quality of life and symptoms related Bosentan is a non-selective ET receptor blocker. 9 RCTs to PH. The long-term effects of epoprostenol in PH include using 1 of 3 ERAs as monotherapy have been performed

183 Medicine Update 2012  Vol. 22

Vasodilator and Vasodilator Therapies Antiplatelet Therapies

Oxygen Prostacyclin analogues Calcium-channel blockers Nitric oxide donors Endothelin-receptor antagonists L-arginine Brain natriuretic peptide Phosphodiesterase inhibitors Anticoagulant Calcitonin gene-related peptide Prostacyclin synthase Therapies

Warfarin

Antiinflammatory Therapies

Prostacyclin analogues Smooth Nitric oxide donors muscle cells Platelets Fibrin Endothelin-receptors antagonists Statins 5-lipoxygenase inhibitors Monocyte-macrophage chemoattractant protein-1

Inhalation Therapies

Endothelium Oxygen Prostacyclin analogues Nitric oxide donors Ethyl nitrite

Leukocytes Remodeling Therapies

Nitric oxide donors Endothelin-receptor antagonists

Fig. 2 : Therapeutic Approaches to Pulmonary Hypertension 24 in PAH patients, bosentan was evaluated in 4 RCTs in PAH receptor blocker that can be given once daily at a 100 mg patients,29-31 including 1 RCT performed in a cohort of patients dose. It has been assessed in PAH patients in 2 RCTs, both of with the Eisenmenger syndrome31 and 1 RCT performed in which demonstrated improvement in exercise capacity and a cohort of patients with only mildly symptomatic PAH.32 hemodynamic status.33,34 Overall, bosentan improved exercise capacity, functional Ambrisentan is an orally active ET -selective endothelin class, hemodynamic status, echocardiographic and Doppler A receptor blocker that can be given once daily at a 5-mg dose, variables, and time to clinical worsening. The approved which can be increased to 10 mg if the drug is well tolerated. It dosage of bosentan is 125 mg twice daily. has been evaluated in RCTs.35 Results showed improvements

Sitaxsentan is an orally active ETA-selective endothelin 184 Pulmonary Hypertension: Newer Insights

Algorithm for the medical treatment of PAH.42

Oral anticoagulants (E/B)-IPAH/HPAH Supportive therapy and general measures Avoid excessive physical exertion (E/A) Diuretics (E/A) Birth control (E/A) Oxygen* (E/A) Psychological and social support (E/C) Digoxin (E/C) Expert referral (E/A) Infection prevention (E/A) Supervised rehabilitation (E/B) Acute vasoreactivity test (A for IPAH) (E/C for APAH) ACUTE RESPONDER NON-RESPONDER

WHO Class I-IV Strength of Amlodipine, diltiazem, Recommendation WHO Class II WHO Class III WHO Class IV nifedipine (B) Ambrisentan, Bosentan, Ambrisentan, Bosentan, Epoprostenol IV A Sildenafil Epoprostenol IV, Iloprost inh, Sildenafil Sitaxsentan, Tadalafil Sitaxsentan, Tadalafil, Iloprost inh B Sustained response Treprostinil SC (WHO I-II) C Beraprost Treprostinil SC Iloprost IV, Treprostinil IV Iloprost IV, Treprostinil IV E/B Initial combination therapy (see below) Ambrisentan, Bosentan, YES NO E/C Sildenafil, Sitaxsentan, Tadalafil Not approved Treprostinil inh+ Treprostinil inh+ INADEQUATE CLINICAL RESPONSE Amlodipine, diltiazem, nifedipine (B) Sequential combination therapy

Prostanoids INADEQUATE CLINICAL RESPONSE + (B) + (B) Atrial septostomy (E/B) and/or PDE-5 I + (B) ERA lung transplant (E/A) in exercise capacity and clinical events that seem similar to In the pivotal tadalafil RCT, ~50% of the patients had oral the results observed with the other 2 ERAs. tadalafil added to background oral bosentan; it improved exercise capacity, hemodynamic status, and clinical On the basis of the results of RCTs using ERAs, the incidence events. Inhaled treprostinil has also been studied as add-on of elevated hepatic transaminases >3 times the upper limit of therapy to either background bosentan or sildenafil; in both normal seems to be ~10, 4, 2% with bosentan, sitaxsentan, combinations, the addition of inhaled treprostinil improved and ambrisentan respectively. They have interactions with exercise capacity.39 These studies support the efficacy of warfarin that require careful monitoring of the INR and dose combination treatment in patients who remain symptomatic adjustments when used together. on monotherapy. The optimal combination on the basis of COMBINATION THERAPY overall risk-benefit considerations remains unknown. Combination treatment has been evaluated to address the EARLY INTERVENTION multiple pathobiologic mechanisms of PH. The combination For functional class II or III patients, the role of early of oral bosentan and IV epoprostenol was investigated in aggressive intervention, either as monotherapy or in 1 small study, with inconclusive results.35,36The addition of conjunction with either a PDE-5 inhibitor and/or an ERA, inhaled iloprost to background oral bosentan demonstrated remains unknown. Although the first RCTs in PAH focused improved hemodynamic status and clinical events in 1 RCT40; primarily on functional class III and IV patients, results from however, these results were not confirmed in an open trial.37 a more recent RCT evaluating the efficacy of bosentan in only In another study, the addition of oral sildenafil to background mildly symptomatic PAH patients support early intervention.35 IV epoprostenol demonstrated improved exercise capacity, In addition, prespecified subgroup analyses of the sildenafil, hemodynamic status, and clinical events; furthermore, the tadalafil, and ambrisentan RCTs did not show any significant addition of oral sildenafil to background IV epoprostenol differences in the therapeutic efficacy of these drugs between increased survival versus IV epoprostenol alone.38 patients in WHO functional classes II and III.35 185 Medicine Update 2012  Vol. 22

INVASIVE TECHNIQUES 6. Tuder RM et.al. Pulmonary circulation: development and patholo- gy. JACC 2009; 54:S3–S9. Atrial septostomy is a palliative procedure, the rationale of which is based on experimental and clinical observations 7. Fedullo PF, et al. CTEPH. NEJM 2001; 345:1465–1472. suggesting that intraatrial defect allowing right-to-left 8. Humbert M et.al. Cellular and molecular pathobiology of PAH. shunting in the setting of severe PH might be of benefit.19 JACC 2004; 43:S13–S24. Indications for the procedure include recurrent syncope and/ 9. Hassoun PM et.al. Inflammation, growth factors, and pulmonary or RV failure, despite maximum medical therapy, as a bridge vascular remodeling. JACC 2009; 54:S10–S19. to transplantation, or when no other option exists. 10. Morrell N et.al. Cellular and molecular basis of PAH. JACC 2009; 54:S20–S31. Heart-Lung And Lung Transplantation has been performed 11. Rich S. Primary pulmonary hypertension. A national prospective successfully in patients with PH since 1981. Currently, bilateral study. Ann Intern Med 1987; 107:216–223. lung transplantation has become the procedure of choice.41 12. D’Alonzo GE et al. Survival in patients with primary pulmonary Hemodynamic studies have shown a moderate reduction in hypertension: results from a national prospective registry. Ann In- PAP and PVR associated with improvement in RV function. tern Med 1991; 115:343–349. The 1-year survival rate is 70-75%, the 2-year survival rate 13. McLaughlin VV. et al. American College of Chest Physicians. is 55-60%, and the 5-year survival rate is between 40-45%. Prognosis of pulmonary arterial hypertension: ACCP evidence- Transplantation should be reserved for patients with PH when based clinical practice guidelines. Chest 2004; 126:78S–92S. they are in NYHA functional Class III/IV despite therapy with 14. McLaughlin VV. et al. Survival in primary pulmonary hyperten- a prostacyclin. sion: the impact of epoprostenol therapy. Circulation 2002; 106: 1477–1482. Meta-analysis to assess effect on mor- 15. Sitbon O et al. Long-term intravenous epoprostenol infusion in tality primary pulmonary hypertension: prognostic factors and survival. Meta-analysis of 25 RCTs in 3363 PH patients showed a JACC 2002; 40:780 –788. reduction in mortality ranging 38-43% after an average 16. Barst RJ et al. A comparison of continuous intravenous epoproste- treatment period of 14.3 weeks.43 A subgroup analysis showed nol with conventional therapy for primary pulmonary hypertensi- that all three classes of PH-approved drugs achieved a similar on. NEJM 1996; 334:296 –301. favorable reduction in mortality, although no statistical 17. Sitbon O.et al. Long-term response to calcium channel blockers significance was achieved individually. in idiopathic pulmonary arterial hypertension. Circulation 2005; 111:3105–11. Conclusions 18. Fuster V. PPH: natural history and the importance of thrombosis. The epidemiology of PAH is changing. Remarkable advances Circulation 1984; 70:580 –7. in understanding the pathobiology and clinical care in PAH 19. Klepetko W et al. Interventional and surgical modalities of treat- have resulted in improved exercise capacity and survival. ment for PAH. JACC 2004; 43:73S– 80. Despite such important progress, however, neither exercise 20. Johnson SR. Anticoagulation in PAH: a qualitative systematic re- capacity nor survival is normal. Controlled studies have only view. Eur Respir J 2006; 28:999 –1004. just begun to evaluate the role of combinations of therapies, as 21. Rich S et al. The short-term effects of digoxin in patients with well as the utility of genetic and other biomarkers. It is hoped right ventricular dysfunction from pulmonary hypertension. Chest that such studies will allow for better targeting of therapy to 1998;114:787–92 individual patients who have PH. 22. Rich S. The effect of high doses of calcium channel blockers on survival in primary pulmonary hypertension. NEJ M 1992; REFERENCES 327:76–81. 1. Kovacs G et.al. . Pulmonary arterial pressure during rest and exer- 23. McLaughlin VM et al. ACCF/AHA clinical expert consensus do- cise in healthy control subjects: a systematic review. Eur Respir J cument on pulmonary hypertension: a report of the ACC Foun- 2009. dation Task Force on Expert Consensus Documents. JACC 2009; 2. Hatano S et.al. Primary Pulmonary Hypertension. Report on a 53:1573– 619. WHO Meeting. October 15–17, 1973, Geneva: World Health Or- 24. Harrison WF et al. Pulmonary Arterial Hypertension NEJM 2004; ganization, 1975. 351:1655-65. 3. Fishman AP. Clinical classification of PH. Clin Chest Med 2001; 25. Simonneau G et al. Continuous subcutaneous infusion of trepro- 22:385–91. stinil, a prostacyclin analogue, in patients with PAH: A double- 4. Simonneau G et.al. Clinical classification of PH. JACC 2004; blind, RCT. Am J Respir Crit Care Med 2002; 165:800. 43:5S–12S. 26. Olschewski H et al. Inhaled iloprost for severe pulmonary hyper- 5. Pietra GG et.al. Pathologic assessment of vasculopathies in PH. tension. NEJM 2002; 347:322. JACC 2004; 43:S25–S32. 27. Barst RJ et al. Beraprost therapy for PAH. JACC 2003; 41:2119.

186 Pulmonary Hypertension: Newer Insights

28. Dupuis J et al. Endothelin receptor antagonist in PAH. Eur Respir 36. Humbert M, Barst RJ, Robbins IM, et al. Combination of bosen- J 2008; 31:407. tan with epoprostenol in pulmonary arterial hypertension: BREA- 29. Channick RN, Simonneau G, Sitbon O, et al. Effects of the dual THE-2. Eur Respir J 2004; 24:353–9. endothelin-receptor antagonist bosentan in patients with pulmona- 37. Hoeper MM et al. Combining inhaled iloprost with bosentan in ry hypertension: a randomised placebo-controlled study. Lancet patients with IPAH. Eur Respir J 2006; 28:691– 4. 2001; 358:1119 –23. 38. Simonneau G et al., for the PACES Study Group. Addition of sil- 30. Rubin LJ et al., for the Bosentan Randomized Trial of Endothe- denafil to long-term intravenous epoprostenol therapy in patients lin Antagonist Therapy Study Group. Bosentan therapy for PAH. with PAH: a randomized trial. Ann Intern Med 2008; 149:521–30. NEJM 2002; 346: 896–903. 39. McLaughlin V et al. TRIUMPH I: efficacy and safety of inhaled 31. Galiè N et al. Bosentan therapy in patients with Eisenmenger syn- treprostinil sodium in patients with PAH. Am J Respir Crit Care drome (BREATHE-5): a multicenter, double-blind, randomized, Med 2008; 177:A965. placebo controlled study. Circulation 2006; 114:48–54. 40. McLaughlin VV et al. Randomized study of adding inhaled ilo- 32. Galiè N et al. Treatment of patients with mildly symptomatic PAH prost to existing bosentan in PAH. Am J Respir Crit Care Med with bosentan (EARLY study): a double-blind, RCT. Lancet 2008; 2006; 174:1257– 63. 371: 2093–100. 41. Orens JB: General overview of lung transplantation and review of 33. Barst RJ et al. Sitaxsentan therapy for PAH. Am J Respir Crit Care organ allocation. Proc Am Thorac Soc 2009; 6:13. Med 2004; 169:441–7. 42. Robyn JB et al. Updated Evidence-Based Treatment Algorithm in 34. Barst RJ. Treatment of PAH with the selective endothelin-A recep- PAH. JACC 2009; 54:S78–84. tor antagonist sitaxsentan. JACC 2006; 47:2049 –56. 43. Macchia A, et al. Systematic review of trials using vasodilators in 35. Galiè N et al. Ambrisentan therapy for PAH. JACC 2005; 46:529 pulmonary arterial hypertension: why a new approach is needed. –35. Am Heart J 2010; 159:245–257.

187