Calcium channel blockers or beta receptor antagonists for patients with ischaemic heart disease. What is the best choice?

The safety of calcium channel blockers has recently been challenged. A case control study by Psaty et a/.' and a new meta-analysis by Furberg et a/.'' of 16 randomized secondary-prevention trials with nifedipine initiated this debate. In the case-control study, antihypertensive therapy based on calcium channel blockers increased the risk for a subsequent myocardial infarction compared with the outcome for therapy based on diuretics or beta-blockers. In the meta-analysis, nifedipine used for secondary prevention in patients with ischaemic heart disease was associated with an increase in total mortality. Both studies claimed that an unfavourable outcome was more prevalent with increasing dosages. Furberg and co-workers are of the opinion that 'mortality data from randomized clinical trials of short acting nifedipine are alarming' and their review 'suggests that the problem may go beyond short-acting nifedipine' while 'extrapolation to slow release dihydropyridines and non-dihydropyridines represents a greater leap'. The authors also discuss mechanisms which may contribute to the increased risk such as pro-ischaemia, negative inotropism, neuroendocrine activation, nocturnal hypotension and increased risk for bleedings. Not surprisingly, these two reports initiated a vigorous debate, followed by several editorials in leading medical journals. Yusuf correctly emphasizes that no study of calcium antagonists in patients with coronary artery disease has yet demonstrated a convincing reduction of morbidity or mortality. His opinion, obviously influenced by the case-control study and the meta-analysis, is therefore that it 'at present is prudent to use drugs from alternative classes of agent as initial treatment for angina pectoris and hypertension'. He also claims that widespread use of calcium antagonists should await results of ongoing trials. Horton' shares the opinion that concerns raised by the new data are serious. Opie and Messerli, Buring et al. and Kloner would also like to see prospective mortality studies on the subject. These authors are hesitant, however, about some of the conclusions made by Psaty et a/. and Furberg et alP\ who they consider overemphasize and generalize the risks. It is, nevertheless, admitted that there may be risks associated with the use of nifedipine, in particular with high dosages of short acting formulations. Case-control studies have their well known limits. One of the most apparent is the problem of adjusting for confounding factors. In contrast to Psaty et a/., Aurnes et al.\ in another recent case-control study, could not verify that calcium antagonists increased the risk for myocardial infarction. One reason for these conflicting results may be the inherent problems with this type of analysis. Even meta-analyses have their flaws. In particular, Opie and Messerli are critical of the way Furberg et al. handled the data upon which their conclusions were based. The choice of studies that were cited, the interpretation of end-points and the time when endpoints were calculated are some aspects on which they expressed a difference in opinion. Diverging opinions regarding details always exist. The point is, however, that neither case-control studies nor meta-analyses are capable of giving us a final solution to a problem. Their strength is to create hypotheses to be tested in subsequent prospective, clinical trials. Accordingly, the debate regarding benefits and drawbacks with calcium antagonists will continue until the release of data from such investigations. In this issue there are two well conducted trials of considerable importance in this context. Fox et al. and Dargie et a/.' present the results of the Total Ischaemic Burden European Trial (TIBET). Atenolol, nifedipine SR and their combination were given to patients with chronic, stable angina pectoris. In all, 682 patients with an established diagnosis of mild to moderate angina pectoris were recruited to a randomized, double-blind, parallel group study. The beta-blocker atenolol was given in a dosage of 100 mg daily and nifedipine SR in a dose of 40 mg daily. The same dosages were used in the combined atenolol+nifedipine group. The two drugs and their