Angiotensin receptor antagonists

Editor, – It was with a good deal of scepticism that I read the recent article by Professor C.I. Johnston, 'Angiotensin receptor antagonists for the treatment of hypertension' (Aust Prescr 1998;21:95-7). His statement that 'unlike ACE inhibitors, angiotensin receptor antagonists have not been associated with cough or angioedema' appears to be based solely on clinical trials and ignores the recent literature. As I am sure your readers would be aware, trials conducted prior to marketing are small, with highly selected groups of patients, and with highly structured protocols. They are limited in their power to detect rare adverse drug reactions, such as angioedema, and to identify 'at risk' individuals or sub-groups not represented in the trials. It is only when the general population is exposed to drugs that many effects become apparent. The effective monitoring of drug safety is also reliant on spontaneous reports of adverse drug reactions submitted to the literature and national pharmacovigilance bodies such as the Adverse Drug Reactions Advisory Committee (ADRAC). The true incidence of adverse effects is probably higher than ADRAC reports suggest.

In the literature, there are at least 4 case reports of angioedema in association with angiotensin receptor antagonists^1,2,3,4,as well as a report from the Netherlands Pharmacovigilance Foundation which describes 13 cases of angioedema involving losartan.⁵ In the most recent issue of the Australian Adverse Drug Reactions Bulletin, ADRAC reported on adverse reactions in association with the two angiotensin II receptor antagonists marketed in Australia, losartan and irbesartan.⁶ There have been 37 reports of cough associated with these two drugs and 24 reports of angioedema. Three reports of cough were particularly convincing as the adverse reaction recurred on rechallenge. Studies in the U.K. using prescription-event monitoring have demonstrated that, although rates of cough are significantly less with losartan than ACE inhibitors, coughing is associated with losartan use at a rate of 3.1 per 1000 patient months.⁷

Ian W. Boyd
Executive Secretary
Adverse Drug Reactions Advisory Committee
Canberra, A.C.T.

Professor C.I. Johnston, the author of the article, comments:

Dr Ian Boyd is right to point out that reports of angioedema and cough associated with angiotensin receptor antagonists have now appeared in the literature and by two pharmacovigilance bodies. Perhaps I would have been wiser if I had used the phrase 'have not yet been' when I first wrote the article. I concur totally with him that clinical trials conducted prior to marketing are a very poor way of determining the scope and incidence of adverse effects of new drugs. I have long advocated that drug authorities should require less premarketing information and demand more postmarketing surveillance.

Angioedema can be a life-threatening adverse effect and, although it occurs with a very low frequency in the general population, it is prudent that patients who have a history of angioedema or have had angioedema or rash with ACE inhibitors not be given AT1receptor antagonists.

Cough is a more contentious issue. Cough is a common symptom in the community and, indeed, in the placebo-controlled trials of AT1 receptor antagonists, the incidence of cough in the placebo arm was up to 5% in some of the studies. Dr Boyd has selectively quoted the literature as there have now been reported 4 trials of AT₁ receptor antagonists including losartan,⁸ candesartan,⁹ eprosartan¹⁰ and telmisartan¹¹ in which the AT₁ receptor antagonists were given to patients with a known history of ACE inhibitor cough. In all of these studies, the incidence of cough was considerably less than in those patients rechallenged with the ACE inhibitor and the same frequency as in the placebo arm. Furthermore, a recent review¹² of the published results of all the placebo-controlled and comparative trials of AT1 blockers demonstrated that the incidence of cough was similar to that of placebo. These challenge studies contribute stronger direct evidence, than reports to pharmacovigilance bodies, that cough is not a characteristic of AT1receptor antagonists.

Furthermore, the recent Australian Adverse Drug Reactions Bulletin¹³ recorded only 37 reports of cough out of a total population of >150 000 patients who have been prescribed angiotensin receptor antagonists. The Adverse Drug Reactions Advisory Committee has a responsibility to put the report into some perspective, which leads to the suggestion that, with a common symptom, the Committee should at least give some idea of the denominator, i.e. the number of patients who have been treated with a particular drug.