Objectives: To evaluate the prognostic value of surrogate markers (HIV RNA copy number, CD4 counts and CDC clinical and immunologic categories) in HIV-infected children through a 2-year period.
Methods: Eighty-six HIV-infected children followed by the Duke Pediatric HIV Clinic in the fall of 1994 were evaluated for plasma HIV RNA concentration (viral load), CD4 lymphocyte percentage, age, antiretroviral treatment status and CDC clinical and immunologic categories. Follow-up evaluations were performed for approximately 2 years, and the time to progression to a new CDC category C diagnosis or death was noted.
Results: Of 86 children 22 had progression to new Category C diagnosis or death. Seven children died, 17 had a new Category C diagnosis and 2 had both. Among children who progressed, the median CD4 percentage at entry was 3% (absolute count, 75 cells/mm3), whereas children who had no disease progression entered with a median of 29% (868 cells/mm3). The overall median viral load at study entry was 4.58 log10 copies/ml (38,019 copies/ml, with a range of 1.7 to 6.78 logs). Children who had no disease progression had a median log copy number of 4.43, whereas 5.18 was the median for children whose disease progressed. Log copy number declined over time in children < 3 years of age, whereas it remained fairly consistent for children 3 years or older. Progression rates were determined by entry plasma HIV RNA concentration quartiles [quartile boundaries < 4.18, 4.58, > 5.08 log RNA copy/ml (< 15,136, 38,019 and > 120,226 copies/ml, respectively)]. Progression rates by quartile were 0 of 21, 4 of 22, 5 of 21 and 13 of 22. Kaplan-Meier survival curves defined by CD4% less than or greater than 15 and log RNA less than or greater than 5.0 (100,000) revealed that patients with CD4% less than 15 and plasma HIV RNA concentration > 5 log10 copies/ml did least well: 11 of 12 (92%) had a progression event at a median of 179 days. Patients with a high CD4 percentage and high viral load, or a low CD4 percentage and low viral load did similarly; 5 of 14 (36%) and 4 of 12 (33%) had progression events, respectively. Patients with high CD4 percentage and low viral load did best: only 2 of 48 (4%) had a progression event.
Conclusions: The two most significant prognostic indicators of disease progression were the initial CD4 percentage and the plasma HIV RNA concentration, and a combination of CD4 percentage and virus load best predicted which children had progression events. Progression was less common in children who had < 100,000 HIV RNA copies/ml initially (6 of 60 vs. 16 of 26; P < 0.001; relative risk 0.16). Therefore it seems reasonable that in a child for whom complete suppression is not possible, a threshold of 100,000 (5 log10 copies/ml) can be used to mandate a change in therapy.