INSIGHTS FROM MONITORING THE CPCRA DIDANOSINE ZALCITABINE TRIAL

The design, conduct, and analysis of clinical trials that evaluate the safety and efficacy of treatment interventions in patients with HIV i nfection provide many scientific challenges. A recently completed rand omized trial of didanosine (ddI) and zalcitabine (ddC), sponsored by t he Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA ), is an especially valuable resource for illustrating these challengi ng issues and for providing insights into how they might be properly a ddressed. Establishing equivalence of treatment effects on clinical ef ficacy end points is illustrated through the use of the confidence int erval approach. The striking changes in treatment efficacy results tha t occurred during the course of the CPCRA trial provide important insi ghts into how a data and safety monitoring board can reduce the risk o f inappropriate early study termination. The trial also provides valua ble insights into how treatment effects should be assessed, revealing inconsistencies between effects on the CD4 surrogate end point and eff ects on primary clinical efficacy end points and showing the incomplet eness of the standardly employed definition of AIDS progression. Final ly, the results of this ddI/ddC trial are used to examine the role of covariate adjustment.