CNS drugs approved by the centralised European procedure: true innovation or dangerous stagnation?

The European Medicines Agency (EMEA) is the regulatory body that provides the institutions of the European community with the best possible scientific advice on the quality, safety, and efficacy of medicinal products (Garattini and Bertele 2001; Garattini et al. 2003). Since the beginning of the EMEA’s activities, a total of 14 drugs acting on the Central Nervous System (CNS) received marketing authorization. To ascertain the degree of innovation of these new drugs, we examined the scientific documentation supporting the approval of drugs reported in the European Public Assessment Reports (EPARs; Table 1). The majority of trials supporting the approval of CNS drugs were short-term studies (75/83). Only 29 studies adopted a three-arm design, where the active substance was compared with placebo and with an active comparator. However, no attempts were made to establish the value of the new drugs in relation to the reference drug, as can be inferred from the absence of statistical power to detect a difference between active treatments in all comparisons submitted to the European regulatory authority. Rating scale scores were used as primary outcome measures in the evaluation of drugs for Alzheimer disease and psychiatric disorders, although three long-term trials of olanzapine (time to symptomatic relapse) and all zaleplon trials (total sleep time) adopted functional outcomes. Functional outcomes were more often employed in the evaluation of drugs for Parkinson disease (daily on time) and for the treatment of partial seizures (seizure frequency). All drugs were approved on the basis of a demonstration of superiority over placebo in terms of primary outcomes (Table 1). With the exception of olanzapine for schizophrenia and for mania, active substances were never shown to be more effective than active comparators. The two olanzapine studies showing superiority over active control drugs, however, used high doses of a control drug in one case (haloperidol, administered at a mean modal maintenance dose 11.8 mg/die; Barbui 1998) and did not compare olanzapine to the best comparator in the second case (valproate was selected as active control; Table 1). A clear reporting of the discontinuation rates, a pragmatic proxy indicator of overall tolerability and effectiveness was never presented, with the exception of one olanzapine trial. Safety warnings leading to changes in the summary of product characteristics were issued for nine drugs after approval (Table 1). Clearly, this analysis of the scientific documentation supporting the approval of the CNS drugs showed very few elements of innovation. Two crucial elements seem to prevent innovation. First, the demonstration of a difference against placebo, and not necessarily against an active comparator, makes a new drug eligible for registration in Europe. Consequently, when active comparators were employed, the applicant relied on demonstrating therapeutic “non-inferiority” because this is in agreement with the current EMEA requirements (Barbui et al. 2001). This resulted in a high degree of uncertainty about the therapeutic role of new drugs acting on the CNS. Moreover, safety data Psychopharmacology (2007) 190:265–268 DOI 10.1007/s00213-006-0629-3