SOME STATISTICAL ISSUES IN HIV VACCINE TRIALS

Efficacy trials of prophylactic HIV vaccines will be among the most di fficult clinical trials ever attempted. Not only will there be challen ges with the recruitment and retention of high-risk uninfected individ uals, there will be many statistical challenges to the design, conduct , analysis, and interpretation of these trials. General features of an efficacy trial are described, including choice for the primary endpoi nt and testing for and estimating vaccine efficacy. Secondary objectiv es of trials are also discussed. These include determining the correla tes of protective immunity, assessing the impact of HIV genetic variat ion on vaccine efficacy, and using biological markers such as viral lo ad and CD4(+) lymphocyte cell count to gain insight on a vaccine's abi lity to prevent or delay disease. The use of biological markers as sur rogates for disease outcome is discussed. Last, trial designs for stud ying several candidate vaccines or other HIV prevention strategies in a single trial are examined.