BACKGROUND: The objective was to analyse which baseline factors could predict a favourable outcome after structured therapy interruption (STI). METHODS: Data of three Spanish pilot studies of STI in early stage chronic HIV-1-infected patients were analysed. A set of 37 variables at baseline was used. Plasma and tonsillar tissue viral load (VL), lymphocyte immunophenotyping and proliferative responses (LPR) to mitogens and specific antigens, and HIV-1 specific cytotoxic T lymphocyte responses were assessed at baseline. Response was defined as a VL set-point after 6 months off antiretroviral therapy after the last interruption of < 5000 copies/ml and 0.5 log(10) below baseline PVL before any antiretroviral therapy. RESULTS: After STI, the 44 patients were classified as follows: 18 (41%) as responders, 26 (59%) as non-responders. In the univariate analysis patients who responded had a significantly lower baseline level of CD4CD38 (P = 0.0068) and naive CD4 T cells (P = 0.03), and a higher level of memory CD4 T cells (P = 0.03) and proliferative response to tetanus toxoid (TT) (P = 0.01) and HIV-1 p24 (P = 0.03) than non-responders. A model incorporating five qualitative variables transformed according to the median value (CD4CD38, CD4 naive and memory T cells and stimulation index to TT and HIV-1 p24) at baseline could classify 97% of patients correctly (P = 0.0001). CONCLUSIONS: A level of memory CD4 T cells and proliferative response to recall antigens above the median may predict a good response to STI, suggesting that preserved memory response in CD4 T cells is important factor. Copyright 2002 Lippincott Williams & Wilkins
BACKGROUND: The objective was to analyse which baseline factors could predict a favourable outcome after structured therapy interruption (STI). METHODS: Data of three Spanish pilot studies of STI in early stage chronic HIV-1-infectedpatients were analysed. A set of 37 variables at baseline was used. Plasma and tonsillar tissue viral load (VL), lymphocyte immunophenotyping and proliferative responses (LPR) to mitogens and specific antigens, and HIV-1 specific cytotoxic T lymphocyte responses were assessed at baseline. Response was defined as a VL set-point after 6 months off antiretroviral therapy after the last interruption of < 5000 copies/ml and 0.5 log(10) below baseline PVL before any antiretroviral therapy. RESULTS: After STI, the 44 patients were classified as follows: 18 (41%) as responders, 26 (59%) as non-responders. In the univariate analysis patients who responded had a significantly lower baseline level of CD4CD38 (P = 0.0068) and naive CD4 T cells (P = 0.03), and a higher level of memory CD4 T cells (P = 0.03) and proliferative response to tetanus toxoid (TT) (P = 0.01) and HIV-1 p24 (P = 0.03) than non-responders. A model incorporating five qualitative variables transformed according to the median value (CD4CD38, CD4 naive and memory T cells and stimulation index to TT and HIV-1 p24) at baseline could classify 97% of patients correctly (P = 0.0001). CONCLUSIONS: A level of memory CD4 T cells and proliferative response to recall antigens above the median may predict a good response to STI, suggesting that preserved memory response in CD4 T cells is important factor. Copyright 2002 Lippincott Williams & Wilkins