Dec 17, 2002

ALLHAT: Diuretic the best bet as a first step in hypertension

When ALLHAT was launched in 1994, the aim was to compare outcomes with newer antihypertensive agents with standard diuretic treatment with chlorthalidone. While they may have theoretical advantages over diuretics and have been shown to reduce cardiovascular events vs placebo, the value of the new agents relative to diuretics has not been shown, the researchers write.

The trial, supported by the National Heart, Lung, and Blood Institute, compared doxazosin, lisinopril, and amlodipine with chlorthalidone; agents were selected to represent their class, and the ALLHAT paper notes that it should be possible then to extrapolate the results seen in this study to all agents in these classes.

More than 44 000 patients were initially randomized, but that number dropped to 33 357 after the halt of the doxazosin arm. Eligible patients were 55 years of age or older, with hypertension and at least 1 other risk factor. They were randomized to receive chlorthalidone (12.5 mg to 25 mg/day, n=15 255), amlodipine (2.5 mg to 10 mg/day, n=9048), or lisinopril (10 mg to 40 mg/day, n=9054). Almost 50% of participants were women, and 35% were black.

Those previously treated for hypertension continued on their drugs until the morning after randomization, when they began the study drug, unless a tapering period was required for safety reasons. If blood pressure was not controlled below 140/90 mm Hg with the step-1 drug, open-label treatment with atenolol, clonidine, or reserpine was allowed at the physician's discretion, but none of the drugs in any of the classes of agents being tested was allowed according to the protocol.

The primary end point was the combined incidence of fatal CHD or nonfatal MI by intention to treat. Secondary outcomes were all-cause mortality, stroke, combined CHD (fatal CHD, nonfatal MI, coronary revascularization, or angina with hospitalization), or combined cardiovascular disease (combined CHD plus stroke, treated angina without hospitalization, heart failure, and peripheral arterial disease).

After a mean follow-up of 4.9 years, a primary outcome event occurred in 2956 of the patients, with virtually identical frequency in each group.

ALLHAT: Primary end point

All-cause mortality was also not different between groups. Five-year systolic pressures were significantly higher with both amlodipine (0.8 mm Hg, p<0.03) and lisinopril (2 mm Hg, p<0.001) compared with chlorthalidone, but 5-year diastolic pressures favored amlodipine (0.8 mm Hg lower than chlorthalidone, p<0.001).

While there was no difference in the primary outcome, differences were seen in several secondary outcomes. These outcomes were similar between amlodipine and chlorthalidone, with the exception of a higher 6-year rate of heart failure with amlodipine.

Secondary outcomes: Amlodipine vs chlorthalidone

When lisinopril and chlorthalidone were compared, secondary outcomes in lisinopril-treated patients showed higher 6-year rates of combined CVD, stroke, and heart failure.

Secondary outcomes: Lisinopril vs chlorthalidone

To download tables as slides, click on slide logo below

Safety outcomes showed similar rates of hospitalization for gastrointestinal bleeding between groups, but angioedema occurred 4 times more often in patients randomized to lisinopril than to chlorthalidone.

Cholesterol levels were higher and the prevalence of hypokalemia and new diabetes was more common with chlorthalidone compared with the other groups after 2 and 4 years of follow-up, the researchers note. "Overall, these metabolic differences did not translate into more cardiovascular events or into higher all-cause mortality in the chlorthalidone group compared with the other 2 groups," they add.

The treatment effects were consistent across all subgroups by sex and diabetic and baseline CHD status. For stroke and combined CVD, however, there was a significant differential effect by race, with higher risks for these outcomes seen among black participants on lisinopril.

ALLHAT steering committee member Dr Jackson T Wright Jr (Case Western Reserve University, General Clinical Research Center, Cleveland) told heartwire that the results would appear to run counter to findings in the African-American Study of Kidney Disease and Hypertension (AASK) trial, of which he was the principal investigator. In AASK, an ACE inhibitor was superior to both a CCB and a beta blocker in preventing progression to renal disease. However, he noted, in that study, cardiovascular outcomes were not the main interest.

"In patients with kidney disease or with major risk for renal events, ACE inhibitors clearly should be considered, but in those where cardiovascular events are the greater risk, the thiazide-type diuretic is the preferred agent," Wright said.

Several subgroup analyses of the ALLHAT data are under way.

Quarrel with conclusions

The surprising ALLHAT results, in particular the poor showing made by the ACE inhibitors, are already being met with some resistance in the clinical community. Perhaps most vocal of the physicians heartwire spoke to is Dr Michael Weber (SUNY Downstate College of Medicine, Brooklyn, NY), who says he "strongly" disagrees with the conclusions made by the ALLHAT investigators that diuretics should be the preferred first step for hypertension therapy.

"The results of the study were driven very much by the experience in the black patients, in whom I would acknowledge that the data do seem to support the idea that a diuretic may be the most appropriate starting point for their therapy," Weber said.

The findings with lisinopril, though, may reflect the fact that ACE inhibitors are known not to work as well in African Americans, as evidenced by the overall poorer blood pressure control in the lisinopril group, he said. Compounding this was that by design, the add-on therapy was a beta blocker, which, he said, "is not logical because they're not sufficiently complementary in their actions. The treatment of choice to add to an ACE inhibitor like lisinopril would be a diuretic or a calcium channel blocker and of course they were prohibited by the study design. That is a major issue, and as a result, blood pressure control was significantly poorer in the lisinopril group, and that can explain a lot of observations in this paper." He also found the high incidence of heart failure in the amlodipine and to some extent, lisinopril groups "surprising," given that ACE inhibitors in particular are actually an effective treatment for heart failure. He speculated that the diuretic effects of chlorthalidone, which would relieve the clinical signs of heart failure such as edema and rales, would perhaps mask the existence of HF in some of those patients.

Finally, he pointed out that all-cause mortality was not different among the 3 groups, as would be expected if chlorthalidone were really preventing serious CV events. Mortality even trended toward being higher with chlorthalidone in nonblack subjects than in either of the other drug groups, he said, and "to me, that's not the basis of superiority."

"Here is a huge clinical trial that's taken a lot of time, a lot of money, a lot of energy, and the participation of a lot of investigators, and in the end, it comes out with a superficial economic and political conclusion that seems to run counter to the scientific findings," Weber concluded.

As for the "passionate debate" described by Appel, Weber dismissed this as an old quarrel with no current clinical relevance. Whichever agent is used to start therapy in new hypertensives, the fact is most patients end up on a combination of therapies. "It really seems to be arguing over trivia when you say, well, I would start with A rather than B, when in the end you're going to finish up with A plus B, or at least you ought to finish up with A plus B."

Dr Franz Messerli (Ochsner Clinic Foundation, New Orleans) said the "most disappointing aspect" of the ALLHAT findings was the results seen with lisinopril. "It deflates HOPE (the Heart Outcomes Prevention Evaluation) completely, because it would indicate that there is no specific cardioprotective or vascular protective effect of the ACE inhibitor," Messerli said. "This is particularly true with the stroke protection that was touted in HOPE to be independent of blood pressure."

He also made the point that the ACE inhibitor was probably disadvantaged by the fact that add-on therapy was often with a beta blocker. "In most physicians' minds, this is not a very efficacious combination," Messerli said.

Another trial similar to ALLHAT, called the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT), is comparing an ACE inhibitor/calcium channel blocker combination, amlodipine and perindopril, with a diuretic/beta blocker combination, atenolol and bendrofluazide. "That's a logical way of doing it, and obviously we look forward to the results of this trial," he said.

Messerli and Weber are currently preparing an editorial outlining some of these views that they expect will be published in an upcoming issue of the Lancet.

Still a place for ACE inhibitors

Furberg acknowledged that the order of add-on therapies probably disadvantaged the ACE inhibitor in this trial, noting that in "real life" patients on ACE inhibitors would receive diuretics, not beta blockers.

"The complicating issue with the interpretation of ALLHAT is that we didn't get the same blood pressure reduction with the ACE inhibitor, and so if the ACE inhibitors come out a little bit behind, we don't know if that is because of an ACE inhibitor effect or that we didn't get the same blood pressure reduction, and that applies primarily to African Americans," he said. Because blood pressure reduction was similar with the diuretic and the CCB, the superiority of the diuretic in that comparison is clear, he said, but more information from studies other than ALLHAT will be required to get a clearer answer on the ACE inhibitors, particularly in African American populations.

But, he adds, "ACE inhibitors have proven effects in a lot of other patient groups heart failure, postinfarction, and so on and so particularly in patients with comorbidity, I think they should be considered."