Objective: To estimate the probability of remaining free of AIDS for up to 25 years after infection with HIV by extrapolation of changes in CD4 lymphocyte count.
Design: Cohort study of subjects followed from time of HIV seroconversion until 1 January 1993. Creation of model by using extrapolated linear regression slopes of CD4 count to predict development of AIDS after 1993.
Setting: Regional haemophilia centre in teaching hospital.
Subjects: 111 men with haemophilia infected with HIV during 1979-85. Median length of follow up 10.1 years, median number of CD4 counts 17. The model was not fitted for three men because only one CD4 measurement was available.
Main outcome measures: Development of AIDS.
Interventions: From 1989 prophylaxis against candida and Pneumocystis carinii pneumonia and antiretroviral drugs when CD4 count fell below 200 x 10(6)/l.
Results: 44 men developed AIDS up to 1 January 1993. When AIDS was defined as a CD4 count of 50 x 10(6)/l the model predicted that 25% (95% confidence interval 16% to 34%) would survive for 20 years after seroconversion and 18% (11% to 25%) for 25 years. Changing the CD4 count at which AIDS was assumed to occur did not alter the results. Younger patients had a higher chance of 20 year survival than older patients (32% (12% to 52%) for those aged < 15, 26% (14% to 38%) for those aged 15-29, and 15% (0% to 31%) for those aged > or = 30).
Conclusions: These results suggest that even with currently available treatment up to a quarter of patients with HIV infection will survive for 20 years after seroconversion without developing AIDS.