Certain NHLs are included in the surveillance case definition of AIDS, and account for greater than 3% of all AIDS-defining illnesses in the United States. Studies have previously identified several mechanisms for B cell activation in the setting of infection with HIV which may lead to development of NHL. We have followed 2 cohorts of patients receiving long-term antiretroviral therapy for the development of NHL: 55 patients receiving AZT-based therapy, of whom 8 have so far developed NHL, and 61 patients receiving ddI, of whom 4 have so far developed NHL. Combining the AZT-based treatment and the ddI patients into 1 cohort containing 116 patients yields a total of 12 high grade B cell lymphomas. For this combined cohort, the estimated probability of developing a NHL after 24 months of antiretroviral therapy was 8.4% (95% Cl of 4.1 to 16.6%), increasing to 19.4% (95% Cl of 10.9% to 32%) after 36 months. There was a statistically significant difference in the development of NHL in patients with less than 50 CD4 cellS/MM3 compared to those with greater than 50 CD4 cells/MM3 irrespective of time on antiretroviral therapy (p2=0.015). Thus, patients with severe HIV infection who survive for prolonged periods, particularly with less than 50 CD4 cells/mm3 are at significantly increased risk of developing NHL. We also examined the patients in the AZT-based cohort for a variety of entry parameters that may have been predictive for the subsequent development of NHL. Patients developing NHL had slightly higher entry IL-6 levels compared with those not developing NHL (p2=0.048). We currently are collaborating with Dwight Kaufman, M.D. on a trial administering infusional chemotherapy with G-CSF and ddI to patients with AIDS-NHL. A previous trial, administering modified PrOMACE-CytaBOM chemotherapy with AZT and GM-CSF demonstrated that AZT could be administered simultaneously with combination chemotherapy using GM-CSF, and yielded a CR rate of approximately 40%.
