New tool for assessing risks of disease progression and death based on age, CD4 percentage and viral load in HIV-infected children

Gareth Tudor-Williams, for HIV i-Base

A revision of the 1994 CDC classification system for HIV-infected children has been long-overdue [1].

This system included immunological categories that recognised age-related changes in CD4 counts in children, but was only crudely related to clinical disease progression, especially for younger children.

Trinh Duong and her colleagues from the HIV Paediatric Prognostic Markers Collaborative Study group have analysed data from a number of cohorts of children followed prospectively in the pre-combination-ART era [2].

They have published new graphs that show the relationship between age and CD4 percentages and clinical events (disease progression and death in the next 6 or 12 months). It is striking that CD4 percentages in children provide a much more sensitive prognostic marker than viral loads. For example, across the range of CD4 percentages for a 12 month old child, the risk of disease progression or death within the next 12 months from about 10-75%. In contrast the risk based on viral loads ranging from 3.5 to 6.5 logs is only about 8-30%.

A tool has been developed based on these data that allows the risks to be calculated by entering either CD4 percentage or viral load and age for individual children. This tool can be accessed and downloaded via the Pediatric European Network for the Treatment of AIDS (PENTA) website at

It is likely that this approach will revolutionise the criteria for starting treatment in children. The PENTA Steering Committee are currently working on the next update of the European treatment guidelines, which will reflect how these data might be used.


  1. Centers for Disease Control. 1994 revised classification system for human immuno-deficiency virus infection in children less than 13 years of age. MMWR 1994; 43 / No. RR-12: 1-10.
  2. HPPMCS group. Short-term risk of disease progression in HIV-1-infected children receiving no antiretroviral therapy or zidovudine monotherapy: a meta-analysis. Lancet 2003; 362: 1605–11.

Links to other websites are current at date of posting but not maintained.