Schizophrenia Research Forum - A Catalyst for Creative Thinking

Barbui C, Bighelli I. Regulatory science in Europe: the case of schizophrenia trials. Lancet. 2013 Jun 20 ; Pubmed Abstract

Comments on News and Primary Papers

Primary Papers: Regulatory science in Europe: the case of schizophrenia trials.

Comment by:  Stefan Leucht
Submitted 8 July 2013
Posted 8 July 2013

In their Lancet paper, Barbui and Bighelli criticize the current position of the European Medicines Agency (EMA), which does not ask pharmaceutical companies to submit all conducted trials and does not require trials to be registered in an international registry. The decision to install registers such as for the recording of clinical trials before their start was a major step forward in the combat of publication bias. So in my opinion, the EMA’s current position is a step backward.

I am more skeptical about the other major point made by the authors—that placebo-controlled trials should no longer be necessary for antipsychotic drugs. Despite all the problematic ethical issues with using placebo in schizophrenia, I think placebo control is still necessary. The most recent antipsychotics only yielded effect sizes of around 0.30-0.40 compared to placebo, similar to the effects of antidepressants compared to placebo in major depressive disorder. These differences are so small it could well be that both the new and the established antipsychotic are actually ineffective, but in a head-to-head trial without a placebo control, this cannot be found out.

Finally, the authors’ last point that a new drug should be significantly better than a standard comparator in a trial in at least one a priori defined outcome makes sense to me. However, whether this needs to be a broad, pragmatic outcome is more debatable. For people with schizophrenia it is important to have a variety of agents available that can be adapted to their needs, and the available second-generation antipsychotics are actually all quite different, in my opinion (more so than, e.g., the different SSRIs). So I would not make this a criterion for EMA approval but leave the question as to how much better a new compound is compared to available treatments to decisions about reimbursement prices.

View all comments by Stefan Leucht

Primary Papers: Regulatory science in Europe: the case of schizophrenia trials.

Comment by:  Wolfgang Fleischhacker
Submitted 3 December 2013
Posted 3 December 2013

The European Medicines Agency (EMA) and Schizophrenia Trials
In their critique of the EMA´s guidance for clinical trials in patients with schizophrenia, Drs. Barbui and Bighelli highlight some very pertinent points, such as that it should be mandatory for a sponsor to provide information on all clinical trials that have been performed and that all trials should be registered in an international repository of study protocols. However, they make a number of suggestions that appear problematic and curious.

First, they note that in a three-arm study, the experimental compound should not only be superior to placebo but also not inferior to an active comparator. Given that non-inferiority trials require a considerably larger number of patients (Fleischhacker et al., 2003), this would significantly enhance the cost of such studies and prolong the duration of innovative drug development. Second, the authors’ assumption that a non-inferiority trial tests “whether a new product is not unacceptably worse than a reference drug” is statistically incorrect. Such differences would only be borne out in superiority studies. Third, Barbui and Bighelli support the notion that experimental compounds must demonstrate superiority over active comparators, should effective agents be available. This, if considered at face value, would in essence create a monopoly for certain drugs until a better or safer one is developed, which not only flies in the face of free market enterprise, but will also minimize the choice clinicians have in selecting medications for their patients. Do the authors really envision healthcare with one antidepressant, one antihypertensive agent, one anti-parkinsonian drug, etc.?

Finally, it is also doubtful whether such superiority trials would “provide practical insight into the potential place of a new investigational product,” as the authors suggest. Practical insight is seldom achieved in preregistration clinical trials, which, given strict inclusion and exclusion criteria, hardly ever reflect everyday clinical practice. To counteract this problem, the EMA might consider stipulating more rigorous Phase 4 commitments.

In contrast to the authors, I fear that setting such a high threshold for the entry of new drugs would not force but rather hinder innovation, insofar as such standards will serve as a strong disincentive for the pharmaceutical industry to engage in central nervous system drug development research.


Fleischhacker WW, Czobor P, Hummer M, Kemmler G, Kohnen R, Volavka J. Placebo or active control trials of antipsychotic drugs? Arch Gen Psychiatry, 2003 May;60(5):458-64. Abstract

View all comments by Wolfgang Fleischhacker