Abstract
Patients with a DLBCL of GCB immunophenotype have a higher chance of cure than patients with non-GCB DLBCL when treated with conventional chemotherapy. In the relapsed/refractory setting, there are conflicting data regarding the impact of cell-of-origin classification on survival following ASCT. The addition of high-dose rituximab (HDR) (1000 mg /m2 days +1 and +8 after ASCT) has been shown to improve results for patients with relapsed DLBCL who undergo transplantation (Khouri et al, J Clin Oncol 2005;23:2240-7). The optimal rituximab dose to be used with ASCT [HDR vs. standard dose (SDR) 375 mg/m2 days +1 and +8)] has been under investigation at our center. More recently, we and others reported on the safety of incorporating 90YIT (0.4 mCi/Kg) in the conditioning. Herein, we compare outcomes of patients with GCB and non-GCB DLBCL after ASCT according to the conditioning regimen received.
121 de novo DLBCL patients were treated on 4 consecutive trials between 2000 and 2012. We determined the cell-of-origin, based upon the Visco/Young and Choi immunohistochemical algorithms, and classified 64 (53%) as GCB and 57 (47%) as non-GCB tumors. Median age, gender distribution, remission status (CR vs PR), # of prior therapies, IPI, beta-2 microglobulin and LDH distribution were similar in both groups. Patients were staged with CT, PET (whenever indicated) scans and bone marrow biopsies, every 3 months for the first year, every 6 months x 5 years, then yearly thereafter. Patients received BEAM conditioning with either SDR [GCB (n =14), non-GCB (n=7), HDR [GCB (n =31), non-GCB (n=40) or 90YIT[GCB (n =19), non-GCB (n=10)].
Determinants for progression (cell-of-origin, age, and gender, and disease status, # of prior therapies, IPI, beta-2 microglobulin and LDH distribution) were studied within each type of conditioning. Within the BEAM-SDR group, the cell-of-origin was the only significant predictor for relapse at 3-years (86% in non-GCB vs. 14% in GCB; P=0.01) (Figure), resulting in improved OS and PFS for the GCB group (both 71% vs 14%, respectively, P=0.06 for OS and 0.04 for PFS). Non-relapse mortality was 0% in this group. Within the BEAM-HDR and BEAM-90YIT groups, GCB and non-GCB patients had a similar rate of relapse (30% vs. 38%, respectively at 3-year, P=0.4) and similar 3-year PFS rates 66% vs 56%, P=0.3), and 3-year OS of 77% vs. 61% (P=0.09).
Our results suggest similar outcomes for patients with DLBCL of GCB and non-GCB immunophenotype when either HDR or 90YIT is added to the BEAM conditioning. A SDR has been found to be associated with a significantly higher risk of progression and inferior survival in non-GCB patients.
Khouri:Spectrum Pharmaceuticals: Consultancy, Research Funding. Off Label Use: Rituximab and ibritumomab tiuxetan for stem cell transplant.
