PURPOSE We examined a consecutive series of 78 patients with non-Hodgkin's lymphoma treated on prospective protocols with high-dose cyclophosphamide, carmustine (BCNU), and etoposide (CBV) plus autotransplantation to determine prognostic factors for time to treatment failure. PATIENTS AND METHODS Patients with relapsed, refractory, or poor-risk intermediate- and high-grade non-Hodgkin's lymphoma were treated with CBV with autologous marrow or peripheral-blood progenitor cell support. Patient characteristics before transplantation were examined in univariate analyses by the log-rank test and simultaneously in a Cox proportional hazards regression analysis. A best-predictive model was determined from those variables significant (P < .10) in the univariate test. RESULTS In univariate analysis, intermediate-grade and immunoblastic lymphoma, responsiveness to pretransplant salvage chemotherapy, and transplantation after primary therapy (first complete response [CR] or partial response [PR]) were associated with prolonged time to treatment failure. In proportional hazards multiple regression analysis, intermediate-grade and immunoblastic histology, responsive disease, and autotransplantation in first CR or PR were positive prognostic factors, and these characteristics are the basis of the best-predictive model for prolonged time to failure. Actuarial 3-year failure-free survival of patients with stable or responding disease at autotransplant was 54%. CONCLUSION CBV is an effective conditioning regimen in intermediate-grade and immunoblastic non-Hodgkin's lymphoma. Patients with these histologies transplanted while responding to primary therapy, or those with stable disease or disease responding to salvage therapy at the time of autotransplant, are most likely to benefit. Patients with lymphoblastic lymphoma or diffuse undifferentiated lymphoma did poorly with CBV and should be offered alternative therapy.
Original article: Chemotherapy with MACOP-B and VACOP-B for intermediate- and high-grade non-Hodgkin's lymphoma: Clinical results and analysis of prognostic factors