Immunological predictors of CD4+ T cell decline in antiretroviral treatment interruptions

27 HIV-infected patients participated in a prospective multicenter study in with a 24 month follow-up. Patients on stable highly active antiretroviral therapy (HAART), with CD4+ count > 600/μL, and HIV-RNA < 50 copies/ml for at least 6 months were offered the option to discontinue antiretroviral therapy. The main outcome was a decline in CD4+ cell count to < 350/μL.After 24 months of follow-up, 16 of 27 (59%) patients (who discontinued therapy) experienced declines in CD4+ cell count to < 350/μL. Patients with a CD4+ nadir of < 200 cells/μL had a greater risk of restarting therapy during the follow-up (RR (CI95%): 3.37 (1.07; 10.36)). Interestingly, lymphoproliferative responses to Mycobacterium tuberculosis purified protein derivative (PPD) below 10000 c.p.m. at baseline (4.77 (1.07; 21.12)), IL-4 production above 100 pg/mL at baseline (5.95 (1.76; 20.07)) in PBMC cultured with PPD, and increased IL-4 production of PBMC with p24 antigen at baseline (1.25 (1.01; 1.55)) were associated to declines in CD4+ cell count to < 350/μL.Both the number (CD4+ nadir) and the functional activity of CD4+ (lymphoproliferative response to PPD) predict the CD4+ decrease associated with discontinuation of ART in patients with controlled viremia.After more than ten years of use, combination antiretroviral therapy (ART) continues to have a significant impact on the morbidity and mortality of HIV infection [1,2]. Despite significant improvement in the toxicity profile and the convenience of administration of currently recommended therapies, the need for lifelong ART makes it desirable to find ways of decreasing the exposure to drugs [3]. Treatment interruption has been proposed as one such a possibility and it has been explored in several clinical studies in different contexts [4-6].Recently, large clinical trials have concluded that treatment interruptions can no longer be recommended due to the risk of clinical progression in some circumstances [4]. However, other studies, both retros