Rituximab vs Observation after High-Dose Consolidative First-Line Chemotherapy (HDC) with Autologous Stem Cell Transplantation in Poor Risk Diffuse Large B-Cell Lymphoma. Final Analysis of the LNH98-B3 GELA Study.

We have shown (Haioun, JCO 2000) that consolidative HDC provides a better survival for patients (pts) with aggressive non-Hodgkin’s lymphoma presenting with 2 or 3 adverse age-adjusted International-Prognostic-Index (aa-IPI) factors in complete remission after induction treatment. In the same time, rituximab had been evaluated as a single therapeutic agent in aggressive lymphomas with promising efficacy (Coiffier, Blood 1998). More recently, rituximab maintenance therapy has been successfully used after chemotherapy or after ASCT to keep responding patients in remission.

The primary objective of the present study was to evaluate the potential benefit, as randomly compared to observation, of rituximab - 375 mg/m2/week for 4 weeks - 2 months after HDC, in decreasing the relapse rate (second randomization: R2). A secondary objective was to improve the response rate before HDC treatment using the intensified ACE chemotherapy regimen (doxorubicin 75mg/m2 d1, cyclophosphamide 1g/m2 d1-d2, etoposide 150mg/m2 d1–d3) as compared to the standard GELA ACVBP induction regimen (first randomization: R1). Induction regimens were delivered every 15 days for 4 cycles. In responding pts, peripheral blood stem cell (PBSC) collection was performed after the third or fourth cycle. HDC (mitoxantrone 45 mg/m2, cyclophosphamide 1500 mg/m2 x 4d, etoposide 250 mg/m2 x 4d and carmustine 300 mg/m2) was started between d80 and d90 after one or two courses of high-dose methotrexate.

From 10/99 to 05/03 (closing date), 476 pts younger than 60 years with diffuse large B-cell lymphoma and aa-IPI 2 or 3 (aa-IPI 3: 29%). were enrolled. We here present the results of the final analysis, performed in July 2005 with 235 pts assigned to receive ACE and 241 to ACVBP. Complete response (CR+CRu) rates to induction treatment did not significantly differ between the 2 regimens (65% and 63%, respectively, p=0.71) and toxic death rates were similar (2% and 3%, respectively) despite higher NCIC grade 3–4 infectious toxicity with the ACE regimen. Among the 331 pts who received HDC, 269 were randomized (R2) to receive either rituximab (n=139) or observation (n=130). 545 infusions of rituximab were administered with no clinically relevant infectious toxicity except two resolutive VZV infections. With a median follow-up of 3 years after R2, a trend toward a better 3y-EFS was observed in the rituximab group (80% vs 72%, p=0.10).

Early and brief rituximab maintenance is feasible after HDC. We hypothesize that prolongation of treatment may improve the results as well as selection of patients more likely to respond to rituximab therapy.