Virologic and immunologic parameters that predict clinical response of AIDS-associated Kaposi's sarcoma to highly active antiretroviral therapy

Abstract

The purpose of the work was to assess the predictive value of biologic factors on the efficacy of highly active antiretroviral therapy alone or combined with chemotherapy on AIDS-associated Kaposi's sarcoma. Twenty-six AIDS-Kaposi's sarcoma patients who started therapy with protease inhibitors were investigated. No baseline chemotherapy was associated with less severe initial clinical status. Median follow-up was 652 d. The main outcome measures were as follows: best Kaposi's sarcoma clinical response; Kaposi's-sarcoma-associated herpesviral load in peripheral blood mononuclear cells using real-time quantitative polymerase chain reaction (non-detectable if less than 100 copies per μg); human immuno-deficiency viral charge in plasma (non-detectable if less than 200 copies per ml); and CD4 lymphocyte count. Time to undetectable Kaposi's-sarcoma-associated herpesviral load, time to undetectable human immunodeficiency viral charge, and time to CD4 ≥ 150 per μl were also recorded over time, from 2 mo measurements. Patients were staged according to the AIDS Clinical Trials Group-based tumor, immune, systemic staging system criteria. At baseline, Kaposi's sarcoma was progressive for 25 (96%) of the 26 enrolled patients. Complete or partial response to highly active antiretroviral therapy alone or combined with chemotherapy was achieved in 22 patients (85%). Median time to clinical response was estimated at 251 d. Clinical response was faster in patients without chemotherapy at baseline (p = 0.003) as well as in patients not previously treated with reverse transcriptase inhibitors (p = 0.0012). Using univariable analyses, predictive factors of clinical response were undetectable Kaposi's-sarcoma-associated herpesviremia (p = 0.013), undetectable human immunodeficiency viremia (p = 0.03), and relative variation of CD4 lymphocytes (p = 0.004). Using multivariable analysis, undetectable Kaposi's-sarcoma-associated herpes-viremia (p = 0.009) and relative variation of CD4 (p = 0.005) were independently selected as having a predictive value for clinical response. Occurrence of nondetection of either Kaposi's-sarcoma-associated herpesvirus or human immunodeficiency virus was not associated with baseline CD4 value. Kaposi's-sarcoma-associated herpesvirus quantitative viral charge is an independent predictive factor of the efficacy of highly active antiretroviral therapy on AIDS-Kaposi's sarcoma. Our results support immune reconstitution as a mechanism of response of Kaposi's sarcoma to highly active antiretroviral therapy.