Related Experiment Videos
Designing phase II studies of chemotherapy for HIV infection using CD4 as an end-point
D A Schoenfeld1, D M Finkelstein, D D Richman
1Biostatistics Center, Massachusetts General Hospital, Boston 02114.
Objective:
To provide information that will be helpful in designing AIDS clinical trials that use CD4 as an end-point.
Design:
Meta-analysis of randomized AIDS clinical trials comparing zidovudine and placebo.
Setting:
Tertiary care.
Patients:
Eight hundred and twenty-seven patients with HIV infection.
Interventions:
Treatment with zidovudine at various dosages compared with placebo.
Main Outcome Measure:
Differences in the log-ratio CD4 count.
Results:
The mean difference in the log-ratio CD4 count divided by its standard deviation varied from 0.31 to 0.76 depending on the study. Of the variation in CD4 count 63% is short-term variation.
Conclusion:
Trials of 12 weeks duration can be used to test for the effect of a new drug on CD4 counts. Testing combination therapies may require somewhat longer trial periods. The sample size for clinical trials can be reduced by replicating baseline and follow-up measurements. Trials of new agents should test for an increase in CD4 over baseline. Such trials will require between 25 and 190 patients per arm depending on the patient population. Trials that compare combinations to standard therapies will require between 32 and 235 patients per arm depending on disease stage and prior therapy of study participants.