U.S. Department of Health and Human
Services
National Institutes of Health
National Heart, Lung, and Blood Institute
Review of Heart Failure Events in the Antihypertensive and Lipid Lowering
Treatment to Prevent Heart Attack Trial: ALLHAT Heart Failure Validation Study
Paula Einhorn, Barry Davis, Henry Black, William Cushman, John Kostis, Daniel Levy, Barry Massie, Barbara deLeon, Linda
Piller, Lara Simpson, Chuke Nwachuku
for
The ALLHAT Collaborative Research Group
Sponsored by the National Heart, Lung, and Blood Institute (NHLBI)
ALLHAT
www.allhat.org
Heart Failure Validation Study
Aimed to centrally review all hospitalized HF events in 42,418 ALLHAT participants.
ALLHAT
Randomized Design of ALLHAT BP Randomized Design of ALLHAT BP TrialTrial
42,41842,418
High-risk High-risk hypertensive hypertensive patientspatients
Consent / Consent / RandomizeRandomize
AmlodipineAmlodipine
ChlorthalidoneChlorthalidone
DoxazosinDoxazosin
LisinoprilLisinopril
Follow until death or end of study (4-8 years, mean 4.9 years)Follow until death or end of study (4-8 years, mean 4.9 years)
ALLHATALLHAT
JAMA. 2002;288:2981-2997
Participants
Hypertensive men and women Age 55 and older (mean 67) At least one additional CVD risk factor [old (>6 months) MI or stroke, LVH, h/o
type 2 diabetes, current smoker, HDL<35, or documented other ASCVD]
Excluded: h/o symptomatic HF and/or known LVEF <35%
ALLHAT
ALLHAT provides a double-blind comparison of 4 drug classes in preventing transition from hypertension to overt heart failure
ALLHAT
ALLHATHypertension and
Heart Failure
In Framingham Heart Study, hypertension was the most common risk factor for HF, antedating it in 91% of cases.
(JAMA. 1996;275:1557-1562)
In BP Lowering Treatment Trialists’ Collaboration, there were differences among drug classes in occurrence of HF. (Lancet 2003, 362:1527-35)
Hypertension Trial ALLHAT
Randomized, double-blind
Large, simple, practice-based trial
No protocol-mandated central review of HF events
Diagnoses assigned by clinic investigators guided by protocol-defined diagnostic criteria
ALLHAT HF Criteria
Must have one from each category:
Category “A” Category “B”Paroxysmal nocturnal dyspnea Rales
Orthopnea 2+ ankle edema
NYHA Classification III Tachycardia >= 120/min
Dyspnea at rest Cardiomegaly by CXR
CXR characteristic of CHF
S3 gallop
Jugular venous distention
ALLHAT Manual of Operations, 5.3.4 JAMA.1997;278:212-216
ALLHAT
Origin of the HF Validation Study
HF endpoint defined as treated in hospital or as outpatient, or fatal
A component of combined CVD (CHD, stroke, HF, PAD)
Systematic review of hospitalized HF events was initiated in 2001, on advice of the DSMB
ALLHAT
HF Validation Study ObjectivesALLHAT
Evaluate ALLHAT investigator-assigned diagnoses
Compare treatment effects to those previously reported (JAMA. 2002;288:2981-2997)
HF Validation Study
Relevant hospital records requested for 2952 hospitalizations in 2047 patients with a diagnosis of HF on admission or during hospital course Summary: face sheet, discharge /death summary,
admitting H&P, ER notes
Imaging reports: chest X-ray, cardiac catheterization, echocardiography, radionuclide imaging
Other reports: pulmonary function tests, cardiology or pulmonary consultation, autopsy
ALLHAT
HF Validation Study Definitions
ALLHAT
ALLHAT/SHEP
Traditional Framingham Criteria
Updated Framingham Criteria
Reviewers’ global clinical judgment
2962 hospital records for 2045 patients received.
2811 records of 1951 patients suitable for review.
Abstracted by cardiology fellows blinded to treatment assignment.
Each record independently reviewed by two reviewers.
For algorithmic criteria (ALLHAT and Framingham), diagnoses were assigned by computer.
Reviewers’ clinical judgment entered as yes, no, don’t know.
ALLHAT HF Validation Study
Framingham HF CriteriaTwo major or one major plus two minor criteria:
MajorParoxysmal nocturnal dyspnea
Jugular venous distention
Rales
Cardiomegaly on CXR
Acute pulmonary edema
S3 gallop
Central venous pressure
Circ. time >25 seconds
Hepatojugular reflux
Autopsy findings: pulmonary edema, congestion, or cardiomegaly
MinorBilateral ankle edema
Nocturnal cough
Dyspnea on exertion
Hepatomegaly
Pleural effusion
Vital capacity 1/3 from max.
Tachycardia > 120/min.
Major or minorWeight loss > 4.5 kg in 5 days in response to treatment of HF
KKL Ho et al, Circulation 88:107, 1993
ALLHAT
Updated Framingham CriteriaRequire 2 major or 1 major plus 2 minor criteria;
one has to be diagnostic and one clinical Part 1: Clinical Findings
Major
PND or orthopnea
Jugular Venous Distention
Pulmonary rales
S3 gallop
Hepatojugular reflux
Diuresis of 10 lbs or 5 kg in response to diuretic treatment with clinical improvement in congestive symptoms
Minor
Bilateral ankle edema
Nocturnal cough
Dyspnea on exertion
Hepatomegaly
ALLHAT
Updated Framingham CriteriaRequire 2 major or 1 major plus 2 minor criteria;
one has to be diagnostic and one clinical Part 2: Diagnostic Findings
MajorAcute pulmonary edema on CXR
PCWP>= 20 mm Hg
LVEF =< 35%
Cardiac Index < 2.0
Evidence of severe valvular disease
Pulmonary edema, congestion, or cardiomegaly on autopsy
MinorPleural effusion or vascular engorgement or redistribution on CXR
PCWP 16-19 mm Hg
LVEF 36 – 44%
Cardiac Index 2.0 – 2.4
Evidence of moderate valvular disease
ALLHAT
Percent agreement with investigator-assigned
diagnosis
Data refer to percentage of participants with at least one event of hospitalized or fatal heart failure that meet the criteria by at least one of the reviewers.
ALLHAT
0102030405060708090
100
ALLHAT
Fram
ingham
1
Fram
ingham
2
Rev
iewer
s
% a
gre
eme
nt
Percent Agreement with investigator-assigned diagnosisALLHAT
Chlor Amlod Lisin TOTAL
N (patients with event reviewed)
617 508 415 1093
ALLHAT 1 69% 73% 72% 71%
ALLHAT 2 69% 72% 71% 70%
ALLHAT 3 66% 69% 70% 68%
ALLHAT 4 56% 58% 56% 57%
Framingham 1 79% 81% 79% 80%
Framingham2 72% 73% 71% 72%
Reviewers agree HF 82% 86% 84% 84%
Percent Agreement with investigator-assigned diagnosisALLHAT
Chlor Dox TOTAL
N (patients with event reviewed)
418 411 829
ALLHAT 1 67% 69% 68%
ALLHAT 2 66% 68% 67%
ALLHAT 3 64% 67% 65%
ALLHAT 4 53% 53% 53%
Framingham 1 77% 76% 77%
Framingham2 69% 69% 69%
Reviewers agree HF 80% 83% 82%
HF Validation StudyVerification of Treatment Effects
For the various Validation Study definitions of HF, relative risks and 95% CIs were calculated using first events confirmed by a given definition.
ALLHAT
0.50 1 2
HF* 1.19 (1.07 - 1.31)
Hosp/fatal HF 1.10 (0.98 - 1.23)
ALLHAT 1 1.18 (1.02 - 1.36)
ALLHAT 2 1.18 (1.02 - 1.36)
ALLHAT 3 1.21 (1.04 - 1.40)
ALLHAT 4 1.13 (0.96 - 1.33)
Framingham 1 1.13 (0.99 - 1.30)
Framingham 2 1.12 (0.97 - 1.30)
Reviewer agree 1.15 (1.01 - 1.32)
Favors Lisinopril Favors Chlorthalidone
Definition, Relative Risk and 95% Confidence Intervals
Validation Study ACEI versus diuretic
ALLHAT
•RR estimates calculated from 2-by-2 table•Pre-specified endpoint of treated in hospital or as outpatient or fatal
0.50 1 2
HF* 1.37 (1.24 - 1.50)
Hosp/fatal HF 1.35 (1.21 - 1.50)
ALLHAT 1 1.46 (1.27 - 1.68)
ALLHAT 2 1.46 (1.27 - 1.67)
ALLHAT 3 1.45 (1.26 - 1.67)
ALLHAT 4 1.45 (1.24 - 1.69)
Framingham 1 1.42 (1.25 - 1.62)
Framingham 2 1.41 (1.23 - 1.61)
Reviewer agree 1.45 (1.28 - 1.64)
Definition, Relative Risk and 95% Confidence Intervals
Validation Study CCB versus diuretic
Favors Amlodipine Favors Chlorthalidone
ALLHAT
•Pre-specified endpoint of treated in hospital or as outpatient or fatal•RR estimates calculated from 2-by-2 table
HF* 1.77 (1.58 - 1.97)
Hosp/fatal HF 1.60 (1.41 - 1.81)
ALLHAT 1 1.71 (1.45 - 2.01)
ALLHAT 2 1.70 (1.45 - 2.01)
ALLHAT 3 1.72 (1.46 - 2.04)
ALLHAT 4 1.67 (1.39 - 2.01)
Framingham 1 1.63 (1.40 - 1.91)
Framingham 2 1.64 (1.40 - 1.93)
Reviewer agree 1.72 (1.48 - 1.99)
0.50 1 2 3
Validation Study α-blocker versus diuretic
Definition, Relative Risk and 95% Confidence Intervals
Favors Doxazosin Favors Chlorthalidone•Pre-specified endpoint of treated in hospital or as outpatient or fatal•RR estimates calculated from 2-by-2 table
ALLHAT
Summary An independent review of hospital records
showed a high degree of agreement with the diagnoses assigned by ALLHAT investigators
(71 – 84% agreement)
Relative risks calculated for several stringent definitions of HF confirm superiority of a thiazide-type diuretic over a CCB, an ACE-I and an alpha-blocker in preventing the onset of symptomatic HF in hypertensive patients with at least one additional risk factor.
ALLHAT
Implications
HF is a costly and deadly complication of hypertension
Thiazide-type diuretics have greater efficacy for HF prevention and should be considered first-step therapy for prevention of HF in high risk patients with hypertension
ALLHAT