New European Union guidance on tackling problems associated with clinical trials when there are limited numbers of patients available has been published by the European Medicines Agency. Due to come into effect on February 1, 2007, the Guideline on Clinical Trials in Small Populations supplements the current EU and International Conference on Harmonisation directives and guidance. Most orphan drugs and paediatric indications submitted for regulatory approval are based on randomised controlled trials that follow generally accepted rules and guidance. However, the new guideline addresses exceptional cases where less conventional or less common methodological approaches might be acceptable.

Special trial designs and/or refined statistical approaches may improve randomised controlled trials where the interpretation will be less clear compared with typically sized Phase III trials. In situations where randomised controlled trials will be severely underpowered, controlled studies with low statistical power in case of an important treatment effect may be preferable to no controlled studies. When randomised controlled trials are not feasible and only case series - with external control groups - or only anecdotal case reports are available, alternative approaches are required. Such compromise positions, notes the guideline, will usually be at the cost of increased uncertainty concerning the reliability of the results and, hence, the reliability of the effectiveness, safety and risk-benefit of the product. Therefore, additional follow-up data, post-approval, will be necessary.

Although there are no special methods for designing, conducting or analysing clinical trials in small populations, approaches exist which increase the efficiency of clinical trials. The guideline stresses the need to weigh the need for statistical efficiency against the need for clinically relevant/interpretable results, with the latter being more important.

The guideline acknowledges that the choice of endpoint may pose considerable problems. Collation of data from several endpoints may be sufficiently compelling to justify marketing authorisation. Surrogate endpoints may be acceptable but need to be fully justified and their relation to clinical efficacy must be clear so that the balance of risks and benefits can be evaluated. Quality of life data are not considered an appropriate stand-alone endpoint.

The importance of controls and comparator groups is stressed, since their absence compromises the reliability of studies. Consideration of the use of a placebo as a comparator is advocated where there is no existing treatment, even in life-threatening diseases. Where a placebo may not be possible, the "best standard of care" may be an appropriate control. Superiority to available alternative treatments that do not have a good evidence base will usually be necessary.

Any deviation from existing guidance will require justification and the EMEA strongly encourages sponsors to seek scientific advice, or protocol assistance, during all phases of development to guide them as to the acceptability of their planned approaches for later marketing.

The Guideline on Clinical Trials in Small Populations was adopted by the Committee for Medicinal Products for Human Use (CHMP) on July 27, 2006. By Jeanette Marchant