Editors' ChoiceBiostatistics

Phase 2B Clinical Trials—A Useful Interim Step to a Phase 3 Clinical Trial

See allHide authors and affiliations

Science Translational Medicine  12 May 2010:
Vol. 2, Issue 31, pp. 31ec76
DOI: 10.1126/scitranslmed.3001254

The field of translational research arose from the desire to expeditiously move scientific ideas from the bench to successful testing in humans. To do this, new ways of designing and conducting the three phases of clinical trials are needed. Phase 1 and 2 trials establish proof-of-concept, and phase 3 trials establish the efficacy and safety of the treatment in humans. Despite positive results in phase 1 and 2 trials, many treatments leave phase 2 with a low pre-test plausibility of efficacy. Given the large jump in time and resources to conduct a phase 3 trial, many potential treatments go unstudied.

Recently, the HIV vaccine field has begun to use phase 2B trials to initially investigate the efficacy of increasing the number of potential vaccinations that can be investigated with limited resources. Now, Peter B. Gilbert shows that one use of the phase 2B trial is to essentially conduct a small-scale efficacy trial. When designing phase 2B trials, Gilbert advocates expanding the traditional study design criteria of power (the probability of declaring a treatment efficacious when truly it is efficacious) and type I error (the probability of declaring a treatment efficacious when truly it is not efficacious) to include other important considerations, such as the notion of treatment efficacy, cost, and time-to-results as envisaged by researchers at the onset of the trial. Using decision theory, Gilbert outlines how to effectively surmise one of five decisions at the end of a phase 2B trial: (i) treatment is harmful, (ii) treatment is useless, (iii) treatment is plausible and a phase 3 trial should be conducted, (iv) treatment meets phase 3 criteria for efficacy, or (v) treatment meets criteria for efficacy from two phase 3 trials. The decision theory framework is the mathematical formula that allows the study team to place weights as to the importance of each of the above five concepts in determining a study outcome while still arriving at a single decision on the basis of the size of the effect of the treatment in the study. This makes decision theory no more difficult to interpret than interpreting a P value above or below 0.05. Using this study design framework, Gilbert theorizes that moving to a phase 3 study from a phase 2 study is only warranted when the pre-test plausibility of treatment success is extremely high. This work establishes the role that phase 2B studies can play in moving more treatments from the bench to final clinical testing in humans and shows how decision theory can be used to expand the considerations included in a studies design.

P. B. Gilbert, Some design issues in phase 2B vs phase 3 prevention trials for testing efficacy of products or concepts. Stat. Med. 29, 1061–1071 (2010). [Abstract]

Navigate This Article