Oral combination chemotherapy in conjunction with filgrastim (G-CSF) in the treatment of AIDS-related non-Hodgkin's lymphoma: evaluation of the role of G-CSF; quality-of-life analysis and long-term follow-up.

In 1993 we reported the efficacy and toxicity profile of an oral combination regimen administered to 18 patients with AIDS-related lymphoma (NHL-1 study). We observed a 61% response rate; 39% one-year survival rate; nearly two-thirds of patients developed > or = grade 3 leukopenia; and 28% of cycles were associated with febrile neutropenia. These results prompted us to shorten the duration of therapy and to add G-CSF to ameliorate the myelosuppression. Twenty patients with biopsy-proven AIDS-related lymphoma were treated with three 6-week cycles of oral chemotherapy consisting of lomustine (CCNU) 100 mg/m2 on day 1, cycles no. 1 and 3; etoposide 200 mg/m2 days 1-3; cyclophosphamide and procarbazine both 100 mg/m2 days 22-31; and G-CSF 5 microg/kg subcutaneously days 5-21 and days 33-42 (NHL-2 study). The following analyses were undertaken: (1) evaluation of toxicity and efficacy parameters for patients in the current (NHL-2) study; (2) analysis of the clinical role of G-CSF by (historical) comparison with the NHL-1 study of the same regimen without G-CSF; (3) quality-of-life assessments using the Functional Living Index-Cancer (FLIC) and Brief Symptom Inventory (BSI) instruments for all 38 patients (NHL-1+2); and (4) long-term follow-up for all 38 patients. In the current study the overall objective response using ECOG criteria was 70% (95% CI, 50-90%) with 6 CRs (30%) and 8 PRs (40%). The median survival duration was 7.3 months (range: 0.5-51+ months). One patient developed CNS relapse. There were no significant differences with respect to demographics or prognostic factors between the patient populations of the NHL-1 study and the current study (P > 0.2 for each factor). Myelosuppression was the major toxicity in both studies. In the current study versus the NHL-1 study, although the lower incidences of grade 3/4 myelosuppression (51% vs. 64%) and febrile neutropenia (17% vs. 28%) on a per cycle basis were not statistically significant, fewer patients (40% vs. 60%) were affected. However, the severity of myelotoxicity was lessened with the addition of G-CSF, measured in terms of the discontinuation of therapy, myelotoxic deaths, and freedom from grade 3/4 myelotoxicity ( P < 0.02). The number of hospitalizations for febrile neutropenia (7 in the NHL-2 study vs. 13 in the NHL-1 study) was also significantly different (P < 0.05). Quality-of-life analysis confirmed no significant functional or psychological deterioration during therapy except for patients experiencing febrile neutropenia, whose functional capacity deteriorated (P < 0.04). The 1-year, 18-month, and 2-year survival rates for the combined studies (38 patients) were 32%, 21%, and 13%, respectively. At time of death 49% of patients were free from progression of their lymphoma. Administration of the oral regimen has resulted in 13% of patients surviving two years, and half of patients surviving free from progression of their lymphoma. This regimen is efficacious and considerate of patient quality-of-life issues. The addition of G-CSF to the regimen decreases the frequency of hospitalization for febrile neutropenia.