The International Prognostic Index can be used as a guide to treatment decisions regarding patients with human immunodeficiency virus?related systemic non-Hodgkin lymphoma

Abstract

The International Prognostic Index (IPI) effectively separates aggressive lymphomas into four groups with significantly different responses to therapy and survival. The authors have applied the IPI to evaluate a series of patients with human immunodeficiency virus (HIV) related lymphoma, a disease for which treatment strategies are still controversial and prognostic indicators are therefore particularly important. Sixty-nine consecutive evaluable patients with HIV-related systemic non-Hodgkin lymphoma (NHL) diagnosed at a single Institution during a 10-year period were analyzed. Primary cerebral lymphoma was not considered. Forty-nine patients (71%) received aggressive combination chemotherapy (CT), 45 of whom were treated with the same program: cyclophosphamide, doxorubicin, etoposide, cytarabine, bleomycin, vincristine, and methotrexate with lecuovorin and prednisone (ProMACE-CytaBOM). Univariate and multivariate methods were used for statistical analysis. End points were response to treatment in patients receiving aggressive CT and survival of treated patients and all patients. According to age-adjusted IPI, 5 patients (7%) belonged to the low risk group, 12 (17%) to the low-intermediate risk group, 16 (23%) to the high-intermediate risk group, and 36 (52%) to the high risk group. Among the four groups with increasing IPI scores, the mean CD4 cell count at NHL diagnosis was 313, 230, 151, and 72/microL, respectively (P = 0.0085). The complete response (CR) rates were 100%, 88%, 50%, and 32% (P = 0. 0001) and the median survival of all patients was >60, 17, 10.9, and 6.8 months (P = 0.0002) for patients with low, low-intermediate, high-intermediate, and high risk IPI scores, respectively. In multivariate analysis, among patients receiving aggressive CT, high risk IPI (P = 0.013) and systemic symptoms (P = 0.014) were the only parameters related to CR, and high risk IPI (P = 0.016) and achievement of CR (P < 0.001) were the only parameters related to survival. When all patients were considered, high risk IPI had significant prognostic value for overall survival (P = 0.01), as did age (P = 0.019) and achievement of CR (P < 0.001). IPI was a reliable prognostic indicator in an unselected series of patients with HIV-related systemic NHL. The outcomes of patients without high risk IPI treated with aggressive CT were similar to those expected for HIV negative patients with lymphoma. However more than half of patients with HIV-related NHL had IPI high risk disease, and their outcomes were poor even after aggressive CT. The degree of immunodeficiency was related to increasing IPI score, suggesting that immunodeficiency may be an important factor contributing to the aggressive clinical presentation of lymphoma.