Background
Phase III, open-label RCT (GOYA). 1,418 patients with previously untreated CD20-positive DLBCL. Obinutuzumab (GA101) is a type II glycoengineered anti-CD20 antibody with enhanced ADCC and direct cell death activity compared to rituximab (type I). The hypothesis was that obinutuzumab's superior CD20 binding and immune effector function would improve DLBCL outcomes beyond rituximab.
Interventions and follow up
Arm A: G-CHOP — obinutuzumab 1000 mg days 1, 8, 15 (cycle 1) then day 1 (cycles 2–6 or 8) + CHOP × 6–8 cycles (n=706)
Arm B: R-CHOP — rituximab 375 mg/m² day 1 × 6–8 cycles (n=712)
Primary endpoint: PFS (investigator-assessed, ITT)
Median follow up: 29 months
Arm B: R-CHOP — rituximab 375 mg/m² day 1 × 6–8 cycles (n=712)
Primary endpoint: PFS (investigator-assessed, ITT)
Median follow up: 29 months
Results
3-yr PFS (G-CHOP vs. R-CHOP): 70.1% vs. 67.3%, HR 0.92 (95% CI 0.76–1.12), P=.40 — negative
CR rate: 57.5% vs. 54.8% — no significant difference
3-yr OS: 77.2% vs. 76.1% — no difference
Subgroups (GCB, non-GCB): no benefit in either subtype
CR rate: 57.5% vs. 54.8% — no significant difference
3-yr OS: 77.2% vs. 76.1% — no difference
Subgroups (GCB, non-GCB): no benefit in either subtype
Adverse events
Infusion-related reactions: 37% (G-CHOP) vs. 22% (R-CHOP), primarily cycle 1 day 1
Hematologic/infectious (grade ≥3): neutropenia 48% vs. 41%; infections 22% vs. 20%
Other: overall grade ≥3 AEs comparable; no excess cardiac toxicity
Hematologic/infectious (grade ≥3): neutropenia 48% vs. 41%; infections 22% vs. 20%
Other: overall grade ≥3 AEs comparable; no excess cardiac toxicity
Conclusions
Obinutuzumab-CHOP did not improve PFS or OS compared to R-CHOP in untreated DLBCL (HR 0.92, P=.40), with higher infusion-related reaction rates. Obinutuzumab is not approved or used for frontline DLBCL.
Key Limitations
The design allowed 6 or 8 cycles per investigator discretion, introducing heterogeneity. DLBCL has lower surface CD20 expression than follicular lymphoma (where obinutuzumab succeeded in GALLIUM), potentially limiting benefit from enhanced CD20 engagement. COO subtype analysis was not prospectively planned as primary stratification; no subgroup benefited. The rationale for type II vs. type I anti-CD20 advantage in fast-dividing, high-apoptosis DLBCL is weaker than in indolent lymphomas. GOYA and GALLIUM together suggest obinutuzumab's benefit is follicular-specific.
Clinical Context
GOYA is the definitive negative trial establishing that obinutuzumab provides no advantage over rituximab in frontline DLBCL, in contrast to follicular lymphoma (GALLIUM). Rituximab remains the anti-CD20 backbone for DLBCL; obinutuzumab is approved only for CLL and follicular lymphoma. The failure of GOYA, ROBUST, and PHOENIX to improve upon R-CHOP emphasizes that the standard evolved only with a fundamentally different approach (ADC: POLARIX in 2022).
References