Reporting and evaluation of HIV-related clinical endpoints in two multicenter international clinical trials.

PURPOSE: The processes for reporting and review of progression of HIV disease clinical endpoints are described for two large phase III international clinical trials. METHOD: SILCAAT and ESPRIT are multicenter randomized HIV trials evaluating the impact of interleukin-2 on disease progression and death in HIV-infected patients receiving antiretroviral therapy. We report definitions used for HIV progression of disease endpoints, procedures for site reporting of such events, processes for independent review of reported events by an Endpoint Review Committee (ERC), and the procedure for adjudication of differences of opinion between reviewers. RESULTS: Of 473 events reported through May 1, 2006, 28% were judged by an ERC to meet "confirmed" criteria and 38% to meet "probable" criteria; 34% were classified "does not meet criteria." For diseases with >5 case reports, the proportion accepted as either "confirmed" or "probable" events was highest for cervical cancer (100%), non-Hodgkin's lymphoma (88%), cryptococcosis (82%), and cryptosporidiosis (80%) and was lowest for HIV encephalopathy (25%), HIV wasting syndrome (33%), and multidermatomal herpes zoster (35%). 25% of cases required adjudication between reviewers before diagnostic certainty was assigned. CONCLUSION: Important requirements for HIV trials using clinical endpoints include objective definitions of "confirmed" and "probable," a formal reporting process with adequate information and supporting source documentation, evaluation by independent blinded reviewers, and procedures for adjudication.