753    

Background:  CD4 cell count is partially correlated with disease progression in children. We investigated whether prediction of disease progression after age 1 year can be improved by considering both slopes of immunologic markers (absolute lymphocytes, CD4⁺ cell count and CD4⁺ percentage) over the first year of life and absolute values of these markers at age 1 year.

Methods:  Data on 98 HIV-infected children enrolled in a European birth cohort study were assessed. Predictive values of immunological markers at age one year were assessed using Cox proportional hazards models, with disease progression after age one year. Non-parametric tree structured survival analysis models were used to determine which measures of the same lymphocyte marker best predicted disease progression risk (CDC stage C or death).

Results:  Of the 98 children, 18 (18.4%) were treated with combination ART during the first year of life before progression and 41 (41.8%) initiated ART after 1 one year but prior to progression; the remaining 39 (39.8%) progressed before initiation of ART; 22 children progressed to serious disease or death after age 1 year, and 76 did not. Mean values and 95% confidence intervals for the absolute values and slopes relating to the first year of life differed significantly between the 2 groups for the 3 markers. For instance, the mean log₁₀ CD4⁺ cell count around 1 year of age for the 22 progressors was 2.89 log (cells/mL) compared with 3.25 log (cells/mL) for non-progressors (p < 0.0001). Both absolute values at 1 year of age and slopes during the first year of life for all 3 markers assessed were associated with risk of subsequent progression to serious disease. The largest effects in terms of a reduction in relative risks of disease progression were seen in the absolute value and in slope of absolute lymphocyte cell counts. In the best multivariable model, allowing for gender and race, a 0.5 log₁₀ increase in absolute lymphocytes at age one was associated with a 83.9% reduced disease progression, and a 10-fold increase in slope of CD4⁺ percentage with a  29.9% reduction. In the tree models for each marker, the absolute value was the most predictive but slope refined the prediction, especially for median absolute values.

Conclusions:  Although absolute values of the 3 immunologic markers were strongly indicative of serious disease progression, their slopes over the first year of life provide additional information for therapy initiation decisions.

Keywords: disease progression; immunological markers; Europe