Telmisartan, Ramipril, or Both in Patients at High Risk for Vascular Events

Randomized, controlled trials involving about 150,000 patients have convincingly demonstrated that angiotensin-converting–enzyme (ACE) inhibitors reduce rates of death, myocardial infarction, stroke, and heart failure among patients with heart failure,1 left ventricular dysfunction,2,3,4 previous vascular disease alone,5,6,7 or high-risk diabetes.8 ACE inhibitors do not block the production of all angiotensin II, so direct receptor blockade might be more effective. ACE inhibitors reduce bradykinin degradation, which enhances vasodilatation, but increase the rates of angioedema and cough. In patients with heart failure, angiotensin II levels may increase and symptoms worsen, despite the use of ACE inhibitors.9 The use of an angiotensin-receptor blocker (ARB), as compared with placebo, reduced the rate of death or hospitalization for heart failure in patients with a low ejection fraction and heart failure who either could not tolerate an ACE inhibitor10 or were already receiving one.11,12 As compared with beta-blockers, ARBs also reduced vascular events in high-risk patients with hypertension and left ventricular hypertrophy.13 Nevertheless, in other high-risk populations, the role of ARBs as an alternative or in addition to ACE inhibitors to prevent cardiovascular outcomes is not known.

Discussion

ACE inhibitors have been convincingly shown to reduce rates of death, myocardial infarction, stroke, heart failure, and revascularization among patients with previous cardiovascular disease and high-risk diabetes. Therefore, to provide clinically relevant information, trials evaluating ARBs in these patients must include proven doses of an ACE inhibitor, either as background therapy or as a comparator.

In our study, we evaluated both approaches in a population similar to the one studied in the HOPE trial. Telmisartan was clearly not inferior to ramipril for both the prespecified primary outcome of death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for heart failure and for the primary outcome in the HOPE trial (death from cardiovascular causes, myocardial infarction, or stroke). Similarities in the telmisartan group and the ramipril group in the proportions of patients who had heart failure, underwent revascularization, or had renal dysfunction (outcomes that were reduced by ramipril in the HOPE trial) and the high rates of adherence to both drug regimens provided additional confidence in establishing the noninferiority of telmisartan. As compared with the ramipril group, the telmisartan group had significantly fewer episodes of cough or angioedema, but this benefit was partially offset by higher rates of hypotensive symptoms (but not syncope). Higher rates of hypotension-related symptoms are consistent with the slightly lower blood-pressure levels associated with telmisartan, although the lower levels did not lead to further benefit. The benefits of ARBs are being evaluated in three other placebo-controlled trials that are expected to be completed in 2008.14,18,19

Our results parallel the findings of the Valsartan in Acute Myocardial Infarction Trial (VALIANT),20 which established the noninferiority of valsartan, as compared with captopril, in patients with left ventricular dysfunction or heart failure after myocardial infarction. The upper boundaries of the confidence intervals and the noninferiority margins were almost identical in the two studies, even though they enrolled different patient populations. The side effects in our study were similar to those in the VALIANT study, which showed lower rates of cough and angioedema in the valsartan group than in the captopril group but higher rates of hypotension-related symptoms. There were more dose reductions (but not discontinuations) because of renal complications in the valsartan group than in the captopril group, an association that was not observed in our study.

In order to establish noninferiority, a rigorous trial design is required that includes a patient population similar to that in the reference trial, a similar drug regimen, high adherence rates, outcomes that the comparator is known to affect, and high statistical power to exclude clinically important differences. All these criteria were satisfied in our study. The entry criteria for our study and the event rates in the ramipril group were similar to those in the HOPE trial, with high follow-up rates in both trials. Moreover, the adherence rate was higher in the ramipril group (89.4% at 2 years and 84.7% at the end of the study among patients receiving either ramipril or an open-label ACE inhibitor) than that in the HOPE trial (85.0% and 78.8%, respectively). A sensitivity analysis that was restricted to patients who adhered to their allocated drug regimen for more than 50% of the study period showed the consistency of our results and confirmed the robustness of noninferiority.

In our study, we confirmed the statistical noninferiority of telmisartan, as compared with ramipril, since the upper boundary of the 97.5% confidence interval (1.09) was lower than the predefined margin of 1.13 for both the primary outcome (P=0.004) and the primary outcome used in the HOPE trial (P=0.001). Telmisartan preserved about 95% (95% CI, 83.2 to 106.3) of the benefits of ramipril over placebo with respect to the primary outcome and preserved 105% (95% CI, 91.6 to 119.0) of the benefits with respect to the outcome of death from cardiovascular causes, myocardial infarction, or stroke that were observed in the HOPE trial. Use of the method of Hasselblad and Kong21 to impute effects of telmisartan versus placebo (based on the benefits of ramipril over placebo in the HOPE trial) indicates a relative risk of 0.79 (95% CI, 0.70 to 0.89) for the primary outcome. The number of patients who discontinued therapy was significantly smaller in the telmisartan group than in the ramipril group, although the absolute difference in the discontinuation rate was modest, because we used an active run-in phase that selected patients for randomization only if they tolerated both medications. After randomization, we vigorously reinforced adherence and encouraged patients who stopped medications to restart them. In clinical practice, the rates of discontinuation might be higher.

We also evaluated whether the combination of telmisartan and ramipril (with both drugs targeted to the full dose) would be superior to ramipril alone. Surprisingly, despite a reduction in systolic blood pressure of 2 to 3 mm Hg in the combination-therapy group, as compared with the ramipril group — a decrease that should have translated into a risk reduction of 4 to 5% — no significant benefit was seen in the primary outcome among patients receiving the two-drug therapy. However, combination therapy significantly increased the risk of hypotension, syncope, renal dysfunction, and hyperkalemia, with a trend toward an increased risk of renal dysfunction requiring dialysis. These results are similar to the analysis of the combined effects of an ARB and an ACE inhibitor, as compared with an ACE inhibitor alone, in four previous trials.22 Therefore, even though combination therapy had a larger biologic effect (as suggested by greater blood-pressure lowering and changes in potassium), dual blockade did not produce any additional clinical benefit in this population.

In this regard, our results are also similar to those of the VALIANT study, in which the combination of a full dose of captopril plus valsartan (the latter at a dose of 80 mg per day, which was lower than the 160 mg per day used in the valsartan-only group) did not significantly reduce the occurrence of the primary outcome but did increase hypotension.20 Taken together, these two studies showed no additive effect for an ARB in conjunction with a full dose of a proven ACE inhibitor.

However, our findings contrast with those of two other studies. In the Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) study,12 which involved patients who had symptomatic heart failure and had been hospitalized in the previous 6 months, candesartan, when added to existing therapy with any ACE inhibitor used at variable doses (with less than half the patients receiving full doses), was superior to placebo in reducing death or hospitalization for heart failure. Similarly, a reduction in hospitalization for heart failure was observed in the Valsartan Heart Failure Trial,11 which compared valsartan with placebo in patients with heart failure, with about 90% of patients receiving background therapy with ACE inhibitors at variable doses. Both trials differed from both our study and the VALIANT study in that decisions regarding the dose and choice of an ACE inhibitor were left to individual physicians, and there was no attempt to titrate the ACE inhibitor to the maximum dose. Furthermore, patients had symptomatic heart failure despite receiving an ACE inhibitor.

The lack of an additional benefit of a substantial lowering of blood pressure is puzzling, both in our study and in the VALIANT study. These results may have been due to previous successful treatment of patients with other drugs so that little further clinical benefit was possible with the addition of full doses of multiple drugs that block the renin–angiotensin system. Alternatively, some harm from combined therapy with ACE inhibitors and ARBs used at full doses may offset any potential gains. Further information is expected regarding the role of ARBs as compared with placebo in patients after stroke,18 in high-risk patients with vascular disease who are unable to tolerate an ACE inhibitor,14 and in patients with atrial fibrillation.19

In conclusion, our data show that in patients who have vascular disease or high-risk diabetes but do not have heart failure, telmisartan is an equally effective alternative to ramipril and is less likely to cause angioedema. The choice between the two agents will depend on the preferences of patients and physicians and the individual patient's susceptibility to specific adverse events. There is no additional advantage (and there is some harm) from the combination of telmisartan and ramipril used in full doses in this population, as compared with ramipril alone.

Supported by a grant from Boehringer Ingelheim, the Heart and Stroke Foundation of Ontario, and a Senior Scientist Award from the Canadian Institutes of Health Research (to Dr. Yusuf).

Dr. Yusuf reports receiving consulting and lecture fees and research grants from Boehringer Ingelheim, AstraZeneca, Sanofi-Aventis, Servier, Bristol-Myers Squibb, and GlaxoSmithKline; Dr. Teo, receiving consulting and lecture fees and grant support from Boehringer Ingelheim; Dr. Schumacher, being an employee of Boehringer Ingelheim; Dr. Dagenais, receiving consulting and lecture fees from Boehringer Ingelheim and Sanofi-Aventis and grant support from Sanofi-Aventis; Dr. Sleight, receiving consulting and lecture fees from Boehringer Ingelheim and lecture fees from AstraZeneca and Sanofi-Aventis; and Dr. Anderson, receiving consulting fees from Boehringer Ingelheim, Servier, Novo Nordisk, and AstraZeneca, lecture fees from Boehringer Ingelheim, Servier, AstraZeneca, and Sanofi-Aventis, and grant support from Boehringer Ingelheim. No other potential conflict of interest relevant to this article was reported.
NEJM