Angiotensin receptor blockers and myocardial infarction

Editor—Verma and Strauss say that, compared with angiotensin converting enzyme (ACE) inhibitors, angiotensin receptor blockers do not reduce (and may increase) the risk of myocardial infarction.¹ Their claim represents an incomplete, inaccurate, and misleading “analysis” of the evidence.

They did not cite the two largest studies that randomised patients to an ACE inhibitor or angiotensin receptor blockers and had the statistical power to evaluate cardiovascular outcomes.²,³ These had twice as many myocardial infarctions as their trials combined (table). As none of their trials randomised these two treatments, their conclusions depend on indirect comparisons, small numbers of events and are unreliable. OPTIMAAL (379 patients with MI in the captopril group and 384 losartan) and VALIANT (798 total myocardial infarctions in captopril group, 796 valsartan) strongly refute the authors' hypothesis.

Other data were selectively and incorrectly cited—for example, mentioning the only CHARM trial with an excess of myocardial infarctions in the candesartan group (the other two trials had fewer) and inaccurate citation of the risk increase (table).⁴ That the losartan group in RENAAL had fewer myocardial infarctions was not mentioned. It is no surprise that angiotensin receptor blockers failed to reduce mortality in trials underpowered to test for this.⁵ The interpretation of trials using an active control is confounded when these can reduce myocardial infarction (for example, a β blocker) or lower blood pressure more (for example, amlodipine).^{w1 w2}

Table 1.

Treatments and cardiovascular outcomes in trials

Trial	Patients' condition	Treatments	No of patients	Follow-up in years	No of patients with myocardial infarction*
IDNT	Diabetic nephropathy	Placebo	569	2.6	51
		Irbesartan	579		48
		Amlodipine	567		29
RENAAL	Diabetic	Placebo	762	3.4	68
	nephropathy	Losartan	751		50
SCOPE	Elderly	Placebo	2460	3.7	63
	hypertension	Candesartan	2477		70
LIFE	Hypertension	Atenolol	4588	4.8	188
	Left ventricular	Losartan	4605		198
	hypertrophy
VALUE	Hypertension risk factors	Amlodipine	7596	4.2	313
		Valsartan	7649		369
CHARM	Heart failure	Placebo	3796	3.1	190
		Candesartan	3803		176
OPTIMAAL	Myocardial infarction	Captopril	2733	2.7	379
		Losartan	2744		384
VALIANT	Myocardial infarction	Captopril	4909	2.1	559
		Valsartan	4909		587
		Captopril and valsartan	4885		554

CHARM: candesartan cilexitil in heart failure: assessment of reduction mortality and morbidity. IDNT: irbesartan in diabetic nephropathy trial. LIFE: losartan intervention for endpoint reduction in hypertension. OPTIMAAL: optimal trial in myocardial infarction with the angiotensin II antagonist losartan. RENAAL: reduction in endpoints in patients with non—insulin-dependent diabetes mellitus with the angiotensin II antagonist losartan. SCOPE: study of cognition and prognosis in the elderly. VALIANT: valsartan in acute myocardial infarction trial. VALUE: valsartan antihypertensive long-term use evaluation.

^*Fatal or non-fatal.

A correct analysis would have considered all relevant data, appropriately weighted, and composite non-fatal and fatal outcomes, to take account of competing risks.

We endorse the need to obtain (and disclose) evidence from randomised trials to support the use of new drugs. Seeing such a misleading opinion in the BMJ does a disservice to the proper evaluation of drug efficacy and safety, as well causing unnecessary anxiety for patients.