[New anxiolytic drugs: methodological issues]

Compared to benzodiazepinic anxiolytics, new anxiolytic drugs are supposed to have a comparable efficacy, but a better profile of safety (abuse liability, rebound effects, cognitive impairment, sedative effects for example). Methodological issues concerning the development of new anxiolytic drugs are numerous: The diagnostic definition and classification of "Anxiety Disorders" (DSM, ICD) remain discussed and regularly revised. Anxiety disorders, while being heterogeneous, often occur in combination with each other or with depression and anxiety appears to be an ubiquitous component of most psychiatric disorders. Studies of new anxiolytic drugs use to assess their efficacy and safety. In most cases, those studies are double-blind, randomized, parallel-group studies versus placebo and reference drug. Usually, the reference drugs remain benzodiazepinic anxiolytics. It would be interesting to develop trials comparing new anxiolytics with each other and with non-pharmacological therapeutics, to improve therapeutic strategies in anxiety disorders. But, in this topic, a lot of methodological difficulties remain unsolved. The choice of dosages may be difficult because dose-ranging studies are not always available for the drugs and the anxiety disorders studied: it actually the lowest effective dose is not well known. Moreover, it could be more useful to study the therapeutic index of the drug. The duration of short and medium term studies, which are the most performed with anxiolytics, is about 4 to 12 weeks. The duration of long term studies is about 4 to 6 months. Those durations are empirically determined and might not always be appropriate. For clinical efficacy assessment, the choice of evaluation criteria is difficult because no tool is perfect. Then, a main criterion and several different criteria have to be chosen among anxiety rating scales, self rating scales and global ratings of psychopathology. Safety assessment of a new anxiolytic specifically use to focus on sedative effects, cognitive impairment, rebound effects and abuse liability. Several methods are interesting to assess abuse liability: for example, looking for a withdrawal syndrome (the most common), determining the profile of subjective effects of the drug using questionnaires and auto-administration or discrimination studies. Sedative effects and cognitive impairment assessment use various batteries of tests. Concerning general assessment of side effects, both check-lists and open reporting have got advantages and drawbacks. Assessment intervals depend on the aim and the duration of the study: weekly in short-term studies and two-weekly or monthly in long-term studies. Those empirically determined intervals, which take into account the practical achievement of the studies, might not detect some drug effects. The number of subjects entering the study has to be determined before its beginning: about 30 to 40 subjects in each group in trials versus placebo and about 120 to 150 subjects in each group in studies versus reference drug. Ethical and legal issues are important in clinical trials of new anxiolytics. The french law (Loi Huriet 1988) has introduced the necessity for investigators to obtain from subjects their informed and written consent for their participation in the study. Also, the representatively of subjects participating in new anxiolytics trials is a matter of concern. However, those difficulties and the requirements of french health department have not yet obstructed clinical research of new anxiolytic drugs, as shown by the number of new beginning studies.