Where the European regulator draws the line on placebo run-in in depression trials
- The Committee for Medicinal Products for Human Use (CHMP) of the European Medicines Agency agreed Revision 3 of its guideline on clinical investigation of medicines for depression on 20 January 2025; it came into effect on 30 September 2025 and replaces Revision 2. The document is addressed to drug developers.
- For short-term efficacy the guideline requires two randomised, double-blind, controlled trials, at least one of them placebo-controlled. Comparison with placebo is also described as useful for distinguishing disease manifestations from adverse reactions of the medicine.
- Short-term trials usually run around 6 weeks; a 50% improvement on the usual rating scale is accepted as a clinically relevant response; success of blinding is to be assessed in all trials at the last visit.
- Enrichment strategies with a placebo run-in are acceptable only in phase II, not in phase III studies, since the external validity of the studies may be affected.
Most prescriptions for depression rest on a few registration trials, and a regulator's guideline sets out what it expects from them. In March 2025 the European Medicines Agency published the third revision of its depression guideline, in force since 30 September 2025. It is written for drug developers, not for clinicians.
What the guideline asks of a pivotal trial
For short-term efficacy it requires two randomised, double-blind, controlled trials, at least one of them placebo-controlled. Placebo also helps to separate disease manifestations from adverse reactions of the medicine. Short-term trials usually run around 6 weeks, and a 50% improvement on the usual rating scale counts as a clinically relevant response. Success of blinding is assessed in all trials at the last visit.
The run-in and the client in your office
A high placebo response has been observed in depression trials submitted for approval, and the guideline gives it a section of its own. One device is singled out: enrichment with a placebo run-in is acceptable in phase II but not in phase III, "since the external validity of the studies may be affected". Another passage gives the reason. A run-in with subsequent patient selection is "considered problematic with regard to the generalizability of the results to the population treated in clinical practice", because the included patients "may not correspond to the target population", for example non-responders to a placebo run-in.
That is the guideline's statement about one selection step. Diagnosis by DSM-5 or ICD-10 and rating-scale cut-offs stay in the design, and the text does not claim they make the sample resemble practice.
Author's comment: when you read the methods of a phase III depression trial, find what happened between screening and randomisation. If a placebo run-in sits there, the participants were selected by their reaction to placebo, which is the case the Agency names as limiting generalizability. If it does not, they entered randomisation without that filter.
Before applying a phase III result to your client, read what happened between screening and randomisation: a placebo run-in is the selection step the EMA names as limiting generalizability to clinical practice.
This is a regulatory guideline for drug developers, not a study, and it states expectations for trial design rather than describing any particular trial. The placebo run-in is the one selection device it names; other entry criteria remain part of each trial's design.