Obinutuzumab or Rituximab Plus CHOP in Patients with Previously Untreated Diffuse Large B-Cell Lymphoma: Final Results from an Open-Label, Randomized Phase 3 Study (GOYA)

Background: Rituximab (R) plus CHOP (R-CHOP) is standard-of-care treatment for previously untreated diffuse large B-cell lymphoma (DLBCL). Approximately 35-40% of patients (pts) will relapse following R-CHOP, and outcomes with salvage therapy remain poor. Obinutuzumab (GA101; GAZYVA/GAZYVARO; G) is a glycoengineered, type II anti-CD20 monoclonal antibody with greater direct cell death induction and antibody-dependent cellular cytotoxicity/phagocytosis activity than R. In the Phase 2 GATHER study (NCT01414855), G plus CHOP (G-CHOP) demonstrated manageable toxicity and promising efficacy in pts with advanced untreated DLBCL. GOYA (NCT01287741) is an open-label, multicenter, randomized Phase 3 study comparing the efficacy and safety of G-CHOP with R-CHOP in pts with previously untreated DLBCL. GOYA was sponsored by Roche with scientific support from the Fondazione Italiana Linfomi.

Methods: Eligible pts were aged ≥18 years and had adequate hematologic function, ≥1 bi-dimensionally measurable lesion, an ECOG performance status (PS) of ≤2 and an International Prognostic Index (IPI) score of ≥2 (high, high-intermediate or low-intermediate risk). Low-risk pts with an IPI score of 1 (but not due to age alone) or with an IPI score of 0 with bulky disease (one lesion ≥7.5cm) were also eligible. Pts were randomized 1:1 to receive 8 (21-day) cycles of G (1000mg i.v. on Days [D] 1, 8, and 15, Cycle [C] 1 and D1, C2-8) or R (375mg/m² i.v. on D1, C1-8) in combination with 6 or 8 cycles of CHOP (number of cycles preplanned in advance for all pts at each site). Preplanned radiotherapy was allowed for bulky or extranodal disease. The primary endpoint was investigator (INV)-assessed progression-free survival (PFS); for the target hazard ratio (HR) of 0.75, the 3-year PFS was expected to improve from 60% to 68%. Secondary endpoints included: PFS assessed by Independent Review Committee (IRC); overall survival (OS); complete response (CR) and overall response rate (ORR) with or without PET (assessed by INV or IRC according to modified Cheson 2007 criteria); and safety.

Results: 1418 pts were randomized to study treatment: 706 to G-CHOP and 712 to R-CHOP. Baseline characteristics were well balanced between the G-CHOP and R-CHOP arms: mean age, 62.0 years in both arms; ECOG PS ≥2, 12% vs. 14%; IPI score ≥3, 47% vs. 43%; Ann Arbor stage III-IV, 76% in both arms. Cell-of-origin distribution, as assessed by gene-expression profiling (NanoString), was similar in both treatment groups (GCB: 58% [271/471] G-CHOP, 58% [269/462] R-CHOP; ABC: 27% [125/471] G-CHOP, 26% [118/462] R-CHOP; Unclassified: 15.9% [75/471] G-CHOP, 16.2% [75/462] R-CHOP). For the primary endpoint of INV-assessed PFS, there was no significant difference between G-CHOP and R-CHOP (3-year PFS, 69% vs. 66%; stratified HR, 0.92; 95% confidence interval [CI], 0.76, 1.12; p=0.3868; Table). Secondary endpoints, including PFS by IRC, OS, and end-of-treatment ORR/CR rate (with and without PET), were consistent with the primary endpoint, with no clinically meaningful differences observed between the treatment arms (Table). In a prespecified subgroup analysis of INV-assessed PFS, a stratified HR of 0.72 (95% CI, 0.50, 1.01) in favor of G-CHOP over R-CHOP was determined for pts with GCB DLBCL (3-year PFS, 79% vs. 70%). No new safety signals were identified. Grade ≥3 adverse events (AEs; 74% vs. 65%) and serious AEs (43% vs. 38%) were more common in the G-CHOP than in the R-CHOP arm. Grade ≥3 AEs of particular interest that were numerically more common with G-CHOP than R-CHOP included neutropenia (57% vs. 48%), infusion-related reactions (45% vs. 32%), infections (54% vs. 44%), and thrombocytopenia (8% vs. 3%). AEs resulting in withdrawal from treatment (12% [84/704] G-CHOP; 9% [60/703] R-CHOP) and AEs with fatal outcome (6% [41/704] G-CHOP; 4% [30/703] R-CHOP) were slightly more common with G-CHOP. The most common AEs leading to death were pneumonia (5 G-CHOP; 6 R-CHOP) and sepsis/septic shock (7 G-CHOP; 3 R-CHOP).

Conclusions: The primary endpoint of this study was not met: G-CHOP did not significantly improve INV-assessed PFS compared with R-CHOP in previously untreated pts with DLBCL. No unexpected safety signals were identified. Further investigation of outcomes in subgroups is planned.

View largeDownload slide

View largeDownload slide

Close modal