Remission from Kaposi's sarcoma on HAART is associated with suppression of HIV replication and is independent of protease inhibitor therapy.

Highly active antiretroviral therapy (HAART) reduces the incidence and improves the prognosis of Kaposi's sarcoma (KS). This study was designed to identify factors associated with KS clinical responses in HIV-infected patients during HAART. We reviewed the files of 138 HIV-1-infected patients with KS. Epidemiologic and HIV-related clinical and biological parameters were recorded at KS diagnosis (baseline) and every 6 months thereafter. In a subset of 73 antiretroviral-naive patients, we compared the clinical outcome of KS according to the use or nonuse of protease inhibitors (PI). After 6 months of follow-up, KS remission was more frequent in patients who were naive of HAART and who were at ACTG stage S0 at baseline (P=0.03 and 0.02). Undetectable HIV viral load was strongly associated with KS remission (P⩽0.004 at all time points), while CD4 cell count was not. Among the 73 antiretroviral-naive patients at baseline, and who were studied for 24 months, KS outcome did not differ between patients who were prescribed PI-containing and PI-sparing regimens. Intercurrent multicentric Castleman's disease was associated with poor outcome after 60 months of follow-up (P⩽0.0001). Fourteen deaths occurred after a median follow-up of 37.5 months, eight of which were KS related. Suppression of HIV replication appears to be crucial to control KS. Non-PI-based regimens were equivalent to PI-based regimens as regards the clinical and virological outcome of antiretroviral-naive HIV-infected patients with KS.