Purpose: Randomized, multicenter, open-label, phase 2/3 trial investigating lenalidomide versus investigator's choice (IC) in relapsed/refractory diffuse large B-cell lymphoma (DLBCL). Experimental Design: Patients with DLBCL who received ≥2 prior therapies were stratified by DLBCL subtype (germinal center B-cell [GCB] vs non-GCB; determined by immunohistochemistry [IHC]), then randomized 1:1 to lenalidomide (25 mg/day, 21 days of 28-day cycle) or IC (gemcitabine, rituximab, etoposide, or oxaliplatin). Crossover to lenalidomide was permitted for IC-treated patients with radiologically confirmed progressive disease. The primary endpoint was overall response rate (ORR). Progression-free survival (PFS), overall survival, and subtype analysis (GCB vs activated B-cell [ABC]) using gene expression profiling (GEP) were exploratory endpoints. Results: Stage 1: 102 DLBCL patients (by IHC: non-GCB, n=54; GCB, n=48) received ≥1 dose of lenalidomide or IC. Hematologic treatment-emergent adverse events with lenalidomide versus IC included neutropenia (42.6%; 36.4%), anemia (33.3%; 47.3%), thrombocytopenia (24.1%; 43.6%), and leukopenia (5.6%; 12.7%), respectively. Overall, lenalidomide-treated patients had an ORR of 27.5% versus 11.8% in IC (ORRs were similar regardless of IHC-defined DLBCL subtype). Median PFS was increased in patients receiving lenalidomide (13.6 weeks) versus IC (7.9 weeks; P=0.041), with greater improvements in non-GCB patients (15.1 weeks versus 7.1 weeks, respectively; P=0.021) compared with GCB (10.1 weeks versus 9.0 weeks, respectively; P=0.550). Conclusions: The clinical benefit of lenalidomide monotherapy in DLBCL patients was more evident in the non-GCB subtype. Exploratory analyses suggest this preferential benefit was more pronounced in the GEP-defined ABC population, demonstrating a need for additional studies of lenalidomide in DLBCL using GEP subtyping.
- Received November 9, 2016.
- Revision received February 28, 2017.
- Accepted March 31, 2017.
- Copyright ©2017, American Association for Cancer Research.