Kaposi's sarcoma (KS) is the most common type of cancer in HIV-1-infected individuals who are progressing to AIDS. In Uganda, KS accounts for four of every 10 cancer diagnoses made at the Uganda Cancer Institute and nine out of 10 of these patients are infected with HIV-1. Although highly active antiretroviral therapy (HAART) can promote KS-regression, nearly 50% of patients fail to achieve a total remission. Even with improved availability of antiretroviral therapy, KS remains a major clinical problem in sub-Saharan Africa. An additional problem for AIDS-KS patients initiating HAART is the development of immune reconstitution inflammatory syndrome (IRIS), an exaggerated and sometimes fatal inflammatory reaction. The pathogenesis of IRIS is poorly understood, and predicting which patients are likely to develop KS-IRIS is not possible. We hypothesize that flow cytometry can be utilized to identify proteins expressed by peripheral blood mononuclear cells (PBMCs) that predict whether or not the KS tumor will respond to HAART and whether or not the patient will develop IRIS. Such biomarkers would greatly facilitate prospective clinical management of these patients. To identify such biomarkers, we propose to analyze PBMC and plasma samples from 224 clinically well-characterized Ugandans with AIDS-related KS collected prior to HAART, at the initiation of treatment and during HAART. This 1-year study will be conducted at the Infectious Diseases Institute (IDI) at Makerere University in Kampala. The study has an added advantage of helping to build basic science capacity on site at the IDI. To analyze our data for statistical significance, we will work closely with Drs. Peter Bachetti and Jeff Martin, experienced biostatisticians. We believe this is the first study to undertake an in-depth characterization of AIDS-KS patients before and during HAART with the goal of identifying useful biomarkers predictive of incomplete tumor responses to anti-retroviral therapy or development of IRIS.