RELATIONSHIP BETWEEN CD4 LYMPHOCYTE COUNT AND AIDS MORTALITY, 1986-1991

Objective: To determine the relationship between CD4 lymphocyte count and short-term AIDS mortality, and to determine whether this relationship changed during 1986-1991. Design: A retrospective analysis of CD4 lymphocyte counts in patients dying with AIDS and estimation of median survival in patients with a CD4 count < 50 x 10(6)/l. Methods: Absolute CD4 lymphocyte count in the 6 months before death was available for 178 patients. The terminal CD4 count was compared in five cohorts of patients dying in 12-month periods from July 1986. Kaplan-Meier survival curves were constructed using survival from date of first absolute CD4 count < 50 x 10(6)/l in 271 patients and compared for each yearly cohort. Results: The median terminal CD4 lymphocyte count for all patients was 10 x 106/l. The median terminal CD4 counts for each yearly cohort were: 1986/1987, 100 (n = 13); 1987-1988, 10 (n = 27); 1988-1989, 10 (n = 30); 1989/1990, 20 (n = 59); 1990-1991, 10 (n = 58). There was a significant difference in the terminal CD4 count in the 1986-1987 period compared with all other years combined, but no further changes after 1987-1988. The median survival of all patients from date of first CD4 count < 50 x 10(6)/l was 11.9 months. There was no significant difference in median survival in each yearly cohort. Conclusions: There is a close correlation between CD4 lymphocyte count and short-term mortality in AIDS. Therapeutic advances over the past 5 years are not reflected by changes in CD4 count before death or improved survival in patients with very low CD4 counts.