CD-1 A-HeFT― Design and Study Population Anne Taylor, MD Professor of Medicine University of Minnesota Medical School
CD-2 A-HeFT Objective Demonstrate the safety and efficacy of BiDil ® compared with placebo in black patients with moderate to severe HF concurrently receiving standard HF treatment DV A-HeFT CSR P 15 6
CD-3 A-HeFT Study Design Double-blind, randomized, parallel-group, placebo-controlled study of black patients with stable symptomatic HF while on standard HF therapy DV A-HeFT CSR P. 15 CSR 6
CD-4 A-HeFT Dosing BiDil ® is an oral fixed-dose combination tablet –20 mg ISDN –37.5 mg HYD Patients were randomized to BiDil or placebo –1 tablet tid, with forced titration to a target of 2 tablets tid Target dose –120 mg/day ISDN –225 mg/day HYD 6 DV A-HeFT CSR P 37, 40
A-HeFT Study Design 6 DV A-HeFT CSR F 1 QoL Echo QoL Screening Randomize QoL Echo BiDil ® Placebo Visit no. Day/wk/mo–2 wk0 1 BaselineTitration 3-5 days 2 1 tab tid2 tabs tid 3 mo 3 QoL 6 mo 4 9 mo 5 12 mo 6 15 mo 7 18 mo 8 (final) Randomization stratified by β-blocker usage. CD-5
CD-6 Inclusion Criteria Self-identified African American (black) patients Symptomatically stable NYHA Class III-IV On standard HF treatment –If on β-blockers, treated for at least 3 mo prior to study entry Ejection fraction –LVEF ≤ 35% or –LVEF 2.9 cm/m 2 (or > 6.5 cm) DV A-HeFT CSR pp 37-38
CD-7 Exclusion Criteria (1) Unstable angina, myocardiaI infarction, acute coronary syndrome, cerebrovascular accident, cardiac surgery, percutaneous cardiac intervention within 3 mo Valvular disease, hypertrophic obstructive cardiomyopathy, or myocarditis Sustained VT unless implantable cardiac defibrillator Requirement for inotropes Women of childbearing age who were pregnant, nursing, or not using contraception Rapidly deteriorating or uncompensated HF such that cardiac transplantation would be likely over the ensuing 1 year DV A-HeFT CSR § 4.3.2
CD-8 Exclusion Criteria (2) Symptomatic hypotension Significant hepatic, renal, or other disease limiting survival over 1 year trial duration Any condition that would jeopardize the evaluation of efficacy or safety Any contraindications to the use of isosorbide dinitrate or hydralazine Receipt of another investigational drug or device within 3 mo Requirement for hydralazine, long-acting nitrates, or phosphodiesterase type 5 inhibitors DV A-HeFT CSR § 4.3.2
CD-9 Primary Endpoint Composite score –All-cause mortality –First HF hospitalization –Change in QoL at 6 mo relative to baseline DV A-HeFT CSR § 4.5.2
CD-10 Primary Endpoint Composite Score Score Death (at any time during the trial) –3 Alive at end of trial 0 First HF hospitalization (adjudicated) –1 No HF hospitalization 0 Change in QoL at 6 mo (or last measurement, if earlier than 6 mo) Improvement≥ 10 units +2 Improvement≥ 5 and < 10 units +1 Change< 5 units 0 Worsening≥ 5 and < 10 units –1 Worsening≥ 10 units –2 Possible score = –6 to +2 DV A-HeFT CSR 66, 67, T26, T24
CD-11 A-HeFT―Quality-of-Life Assessment The Minnesota Living With Heart Failure questionnaire (MLHFQ) is a self-administered, 21-question tool measuring physical and emotional effects of HF Scores range from 0 to 5 for each question (0 to 105 total possible score) Lower scores indicate better QoL QoL was measured at baseline and every 3 mo during the trial
CD-12 A-HeFT Design—Statistical Analysis for Primary Endpoint Composite Score Intention-to-treat analysis Worst-case score for missing data 3 components –Mortality (score 0 or –3) –Hospitalization for HF (score 0 or –1) –Change in QoL at 6 mo (score –2 to 2) Cui, Hung, and Wang (1999) group sequential method DV A-HeFT CSR pp 36, 76
CD-13 Secondary Endpoints Death from any cause –Time to death –Cause-specific mortality HF hospitalizations –Time to first hospitalization –Number of hospitalizations –Total days in hospital Change from baseline in overall QoL MLHFQ score at each timepoint DV A-HeFT CSR § 4.5.2
CD-14 A-HeFT Design—Statistical Analysis for Secondary Endpoints Analysis of death: Kaplan-Meier, log-rank test Analysis of first hospitalization for HF: Kaplan-Meier, log-rank test Comparison of event rates of death and hospitalization for HF: 2-sample t test or Wilcoxon test Comparison of change in QoL: 2-sample t tests for the difference between groups in MLHFQ scores at each timepoint DV A-HeFT CSR, NitroMed BB
CD-15 Sample Size Calculation and Interim Analyses Protocol specified 2 interim analyses plus a final analysis Sample size re-estimation at the second interim analysis when 300 patients completed 6 mo –313 patients who reached 6 mo were included in this interim analysis (528 patients were randomized) Cui, Hung, and Wang method for analysis For an α = 0.05 –900 patients were required for 80% power Per advice of FDA used an α = 0.02 for sample size re-estimation but not for hypothesis testing –1100 patients were required for 80% power DV NitroMed BB pp 68-70
CD-16 Trial Termination (1) No boundaries to terminate trial for mortality had been formulated at start of study (May 2001) DSMB noted a disparity between treatment groups in deaths (March 2004) –O’Brien-Fleming type group sequential alpha spending function as described by Lan and DeMets to guide further decision making established at that time –Treatment difference in mortality in March 2004 fell just below the value specified by newly formulated boundaries DSMB recommended 1 additional safety review to take place 3 mo to 5 mo later DV NitroMed BB p 69-70
CD-17 Trial Termination (2) DSMB recommendation to stop trial at added safety review (July 2004) –Due to positive effect in mortality in BiDil group relative to placebo Results discussed with Steering Committee, who also recommended study be stopped NitroMed followed recommendations and stopped study on July 19, 2004 DV NitroMed BB pp 69-71
CD-18 Trial Overview 180 sites (169 sites randomized at least 1 patient) 1050 randomized patients (518 BiDil ®, 532 placebo) Up to 18 mo of follow-up No patient lost to follow-up for vital status First patient enrolled 5/29/01 Study terminated 7/19/04 for significant survival benefit in the BiDil group DV A-HeFT CSR< NitroMed BB
CD-19 Baseline Characteristics (1) * P < 0.05 BiDil ® n = 518 Placebo n = 532 Age (mean), yr57 Male sex * 56%64% NYHA class III95% IV 3% 5% Primary cause of heart failure Ischemic heart disease23% Hypertension40%37% Idiopathic25%28% Valvular heart disease3% Other10%9% DV A-HeFT CSR T 14, 15
CD-20 Baseline Characteristics (2) BiDil ® n = 518 Placebo n = 532 Systolic BP, mm Hg (mean ± SD)127 ± ± 18 Diastolic BP,* mm Hg (mean ± SD)78 ± 1076 ± 11 QoL score (mean ± SD)51 ± 2551 ± 26 LVIDD, cm (mean ± SD)6.5 ± ± 1.0 Ejection fraction, % (mean ± SD)24 ± 724 ± 8 Diabetes,* %4537 Renal insufficiency, %1618 Atrial fibrillation, %1518 Implantable cardiac defibrillator, %17 DV A-HeFT CSR T 14, 15, 16, PTT16.1 * P < 0.05
CD-21 Baseline Cardiovascular Medications Patients, n (%) BiDil ® (n = 518) Placebo (n = 532) ACE inhibitor 386 (74.5)400 (75.2) ARB 124 (23.9)112 (21.1) ACE inhibitor or ARB 478 (92.3)495 (93.0) β-blocker 434 (83.8)437 (82.1) Aldosterone antagonist 208 (40.2)201 (37.8) Digitalis glycoside 304 (58.7)324 (60.9) Diuretic 473 (91.3)494 (92.9) Calcium channel blocker 109 (21.0)104 (19.5) DV A-HeFT CSR T22
CD-22 Patient Disposition Patients, n (%) BiDil ® Placebo Total randomized patients518 (100.0)532 (100.0) Completed study469 (90.5)457 (85.9) Discontinued from study prematurely49 (9.5)75 (14.1) Death30 (5.8)54 (10.2) Investigator decision9 (1.7)13 (2.4) Patient withdrew consent5 (1.0)3 (0.6) Patients lost to follow-up (non-vital status follow-up) 2 (0.4)0 Cardiac transplantation3 (0.6) Not reported02 (0.4) 6 DV A-HeFT CSR P. 83
CD-23 Study Drug Prescribed as Assessed by Total Tablets/Day at Various Time Points A-HeFT Mean ± SD Timepoint BiDil ® n = 517 Placebo n = mo4.4 ± ± mo4.5 ± ± mo4.8 ± ± mo4.8 ± ± mo4.9 ± ± 1.7 Endpoint4.1 ± ± DV CSR T20
CD-24 Mean Daily Prescribed Dose of BiDil ® ISDNHYD 3 mo (n = 368) 88 (42)188 (71) 6 mo (n = 317) 90 (40)191 (68) 9 mo (n = 260) 96 (38)195 (64) 12 mo (n = 220) 96 (38)199 (60) 15 mo (n = 169) 98 (34)199 (64) Dose, mg/day (SD) DV A-HeFT CSR T 20