[How to develop an anti-arrhythmia agent in 1990. The cardiologist's viewpoint]

The development of an antiarrhythmic drug in 1990 follows well established pathways. The first step is animal experimentation to establish the antiarrhythmic effect and to estimate the risk of toxicity, but clinical trial is the fundamental stage which confers on a product the title of antiarrhythmic. Several models may be used, one of which is ventricular extrasystoles. However, the data so obtained must not be extrapolated to a wider context. All methods for evaluating an antitachycardia effect have imperfections. Obviously, arrhythmias treated in hospital are privileged although they may only represent a special subgroup of more severe arrhythmias. Potential antiarrhythmics are more often assessed by their effects on tachycardias reproduced by programmed pacing, leaving other arrhythmias, including atrial fibrillation, to one side. In addition, long-term efficacy is more often suspected than demonstrated. The action on malignant ventricular tachyarrhythmias escapes all valid assessment mainly because of the difficulty of using a control group of patients with life-threatening arrhythmias. The use of automatic defibrillator devices could, in the near future, bring a more rigorous approach to this problem. One of the most important objectives of antiarrhythmic drugs is the prevention of sudden death. Results have been relatively disappointing until now. This may be due to deleterious, above all proarrhythmic, side-effects counterbalancing the benefits of the drug. The unwanted side-effects must be evaluated systematically in the future. The methodology needed is not yet established and requires further research.