Background:  Programmed treatment interruption in adults has shown controversial results, but experience in children is still scarce. We aimed to evaluate the safety of programmed treatment interruption as a sparing regimen, in terms of clinical events and immunovirological evolution in HIV-vertically-infected patients.

Methods:  Prospective case-control study. Inclusion criteria for programmed treatment interruption were to be on a first-line HAART initiated during chronic infection; plasmatic HIV RNA <200 copies /mL, and a CD4 cell count or percentage within CDC immunological category 1, both for the last 24 months. Monthly controls during the first 6 months after programmed treatment interruption and 2 to 3 months thereafter were performed and HAART was resumed if any HIV-related clinical event in CDC clinical category C or a marked decrease in CD4 cells (<350 cells/mm³ for adolescents or <17% for children <12 years) was observed. A control group of children fulfilling the same inclusion criteria but not undergoing treatment interruption was used.

Results:  A total of 27 HIV-infected children (13 girls), 12 of whom underwent programmed treatment interruption (8 girls) were included. Baseline characteristics (age, gender, age at HIV infection diagnosis, CDC clinical category, nadir CD4 cell percentage, age at HAART initiation, and type of HAART regimens) were not different among cases and controls. At programmed treatment interruption (median age 12 years), patients had been a median time of 62 months on HAART and 57 months with undetectable HIV viral loads. Because of rapid decrease in CD4 cell percentage (down to 9% and 18% plus HIV-related thrombocytopenia, respectively), 2 girls had to resume HAART after 2 and 5 months. The rest of the children have remained a median time of 36 months (range 18 to 46 months) on programmed treatment interruption. No patients have developed an AIDS-defining clinical event during the follow-up. Among children on programmed treatment interruption, 2 non-complicated episodes of zoster and 1 bacterial pneumonia have been observed (none in the control group). A significant decrease in median CD4 cell counts and percentage have been observed after programmed treatment interruption when compared with controls (728 vs 923 /mm³, p = 0.05; 33% vs 51%, p = 0.001), with the following slopes –11 cells /mm³ and –0.6%/month. HIV viral load has remained stable overtime (median 4.3 log, range 2.4 ro 5.5 log) after the initial blip. An additional 3 girls have had to resume HAART because of immunologic criteria (n = 2) and HCV-related hepatopathy (n = 1). In all 5 cases when HAART was restarted, a rapid immunologic and virologic response was observed.

Conclusions:  Despite progressive decrease in CD4 cell counts, programmed treatment interruption may represent a safe therapeutic strategy in HIV-infected pediatric patients, provided close follow-up is warranted.