A-HeFT? Design and Study Population

1
A-HeFT?Design and Study Population

• Anne Taylor, MD
• Professor of MedicineUniversity of Minnesota
  Medical School

2
A-HeFT Objective
6
DV A-HeFT CSR P 15

• Demonstrate the safety and efficacy of BiDil
  compared with placebo in black patients with
  moderate to severe HF concurrently receiving
  standard HF treatment

3
A-HeFT Study Design
6
DV A-HeFT CSR P. 15 CSR

• Double-blind, randomized, parallel-group,
  placebo-controlled study of black patients with
  stable symptomatic HF while on standard HF
  therapy

4
A-HeFT Dosing
6
DV A-HeFT CSR P 37, 40

• 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

5
A-HeFT Study Design
6
CD-5
DV A-HeFT CSR F 1
RandomizeQoLEcho
QoL Echo
QoL
QoL
QoL
QoL
QoL
BiDil
Placebo
Screening
Baseline
Titration
1 tab tid
2 tabs tid
4
Visit no.
1
2
3
5
6
7
8 (final)
6mo
Day/wk/mo
2 wk
0
3-5days
3mo
9mo
12mo
15mo
18mo
Randomization stratified by ß-blocker usage.
6
Inclusion Criteria
DV A-HeFT CSR pp 37-38

• Self-identified African American (black) patients
• Symptomatically stable NYHA Class III-IV
• On standard HF treatment
• If on ß-blockers, treated for at least 3 mo prior
  to study entry
• Ejection fraction
• LVEF 35 or
• LVEF lt 45 with resting LVIDD gt 2.9 cm/m2 (or gt
  6.5 cm)

7
Exclusion Criteria (1)
DV A-HeFT CSR 4.3.2

• 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

8
Exclusion Criteria (2)
DV A-HeFT CSR 4.3.2

• 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

9
Primary Endpoint
DV A-HeFT CSR 4.5.2

• Composite score
• All-cause mortality
• First HF hospitalization
• Change in QoL at 6 mo relative to baseline

10
Primary Endpoint Composite Score
DV A-HeFT CSR 66, 67, T26, T24

• 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 lt 10 units 1
• Change lt 5 units 0
• Worsening 5 and lt 10 units 1
• Worsening 10 units 2
• Possible score 6 to 2

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

12
A-HeFT DesignStatistical Analysis for Primary
Endpoint Composite Score
DV A-HeFT CSR pp 36, 76

• 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

13
Secondary Endpoints
DV A-HeFT CSR 4.5.2

• 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

14
A-HeFT DesignStatistical Analysis for Secondary
Endpoints
DV A-HeFT CSR, NitroMed BB

• 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

15
Sample Size Calculation and Interim Analyses
DV NitroMed BB pp 68-70

• 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 a 0.05
• 900 patients were required for 80 power
• Per advice of FDA used an a 0.02 for sample
  size re-estimation but not for hypothesis
  testing
• 1100 patients were required for 80 power

16
Trial Termination (1)
DV NitroMed BB p 69-70

• 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)
• OBrien-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

17
Trial Termination (2)
DV NitroMed BB pp 69-71

• 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

18
Trial Overview
DV A-HeFT CSRlt NitroMed BB

• 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

19
Baseline Characteristics (1)
DV A-HeFT CSR T 14, 15
BiDiln 518 Placebon 532
Age (mean), yr Age (mean), yr 57 57
Male sex Male sex 56 64
NYHA class NYHA class
III III 95 95
IV IV 3 5
Primary cause of heart failure Primary cause of heart failure
Ischemic heart disease Ischemic heart disease 23 23
Hypertension Hypertension 40 37
Idiopathic Idiopathic 25 28
Valvular heart disease Valvular heart disease 3 3
Other Other 10 9
P lt 0.05
20
Baseline Characteristics (2)
DV A-HeFT CSR T 14, 15, 16, PTT16.1
BiDiln 518 Placebon 532
Systolic BP, mm Hg (mean SD) Systolic BP, mm Hg (mean SD) 127 18 125 18
Diastolic BP, mm Hg (mean SD) Diastolic BP, mm Hg (mean SD) 78 10 76 11
QoL score (mean SD) QoL score (mean SD) 51 25 51 26
LVIDD, cm (mean SD) LVIDD, cm (mean SD) 6.5 0.9 6.5 1.0
Ejection fraction, (mean SD) Ejection fraction, (mean SD) 24 7 24 8
Diabetes, Diabetes, 45 37
Renal insufficiency, Renal insufficiency, 16 18
Atrial fibrillation, Atrial fibrillation, 15 18
Implantable cardiac defibrillator, Implantable cardiac defibrillator, 17 17
P lt 0.05
21
Baseline Cardiovascular Medications
DV A-HeFT CSR T22
Patients, n () Patients, n ()
BiDil(n 518) Placebo(n 532)
ACE inhibitor ACE inhibitor 386 (74.5) 400 (75.2)
ARB ARB 124 (23.9) 112 (21.1)
ACE inhibitor or ARB ACE inhibitor or ARB 478 (92.3) 495 (93.0)
ß-blocker ß-blocker 434 (83.8) 437 (82.1)
Aldosterone antagonist Aldosterone antagonist 208 (40.2) 201 (37.8)
Digitalis glycoside Digitalis glycoside 304 (58.7) 324 (60.9)
Diuretic Diuretic 473 (91.3) 494 (92.9)
Calcium channel blocker Calcium channel blocker 109 (21.0) 104 (19.5)
22
Patient Disposition
6
DV A-HeFT CSR P. 83
Patients, n () Patients, n ()
BiDil Placebo
Total randomized patients 518 (100.0) 532 (100.0)
Completed study 469 (90.5) 457 (85.9)
Discontinued from study prematurely 49 (9.5) 75 (14.1)
Death 30 (5.8) 54 (10.2)
Investigator decision 9 (1.7) 13 (2.4)
Patient withdrew consent 5 (1.0) 3 (0.6)
Patients lost to follow-up (non-vital status follow-up) 2 (0.4) 0
Cardiac transplantation 3 (0.6) 3 (0.6)
Not reported 0 2 (0.4)
23
Study Drug Prescribed as Assessed by Total
Tablets/Day at Various Time PointsA-HeFT
17-DV
CSR T20
Mean SD Mean SD
Timepoint BiDiln 517 Placebon 527
3 mo 4.4 2.1 5.0 1.9
6 mo 4.5 2.0 5.1 1.8
9 mo 4.8 1.9 5.2 1.7
12 mo 4.8 1.9 5.3 1.6
15 mo 4.9 1.7 5.3 1.7
Endpoint 4.1 2.0 5.2 1.5
24
Mean Daily Prescribed Dose of BiDil
DV A-HeFT CSR T 20
Dose, mg/day (SD)
ISDN HYD
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)