Preirradiation chemotherapy with cyclophosphamide, doxorubicin, vincristine, and dexamethasone for primary CNS lymphomas: initial report of radiation therapy oncology group protocol 88-06.

Abstract

PURPOSE: This study was a prospective phase I/II trial performed by the Radiation Therapy Oncology Group (RTOG) to test the tolerance and efficacy of preirradiation cyclophosphamide, doxorubicin, vincristine, and dexamethasone (CHOD) chemotherapy followed by large-volume, high-dose brain radiation therapy (RT) for patients with primary CNS lymphoma (PCNSL). PATIENTS AND METHODS: Fifty-four (52 assessable) human immunodeficiency virus (HIV)-negative patients with PCNSL were entered on study and received two (n = 20) or three (n = 32) cycles of CHOD (six patients with positive CSF cytology received intrathecal methotrexate in addition to CHOD). Whole-brain RT to 41.4 Gy and tumor boost to 18 Gy (total dose, 59.4 Gy) followed chemotherapy. RESULTS: As of July 1994, with a minimum potential follow-up time of 20 months, 12 of 52 assessable patients remain alive without evidence of progression. The median survival time for the entire group is 16.1 months, with a 2-year survival rate of 42%. By univariate analysis, patient age was found to be a significant prognostic factor with respect to survival (P = .005) in favor of age less than 60 years. Karnofsky performance status (KPS) was of borderline significance (P = .057). Survival for patients treated on RTOG 88-06 was compared with that of patients treated on RTOG 83-15, which tested RT alone. No difference in overall survival was found (P = .53). Grade 4 neutropenia developed in 29 of 51 patients during chemotherapy. There were two deaths during chemotherapy: one as a result of sepsis and one of a pulmonary embolus. The worst toxicity during RT was < or = grade 2 in 50 of 52 patients. CONCLUSION: Preirradiation CHOD chemotherapy does not significantly improve survival over RT alone for patients with PCNSL. Age remains a powerful prognostic factor independent of therapy and must be considered in testing alternative combined approaches.