A Randomized Phase II Study of Mosunetuzumab SC Plus Polatuzumab Vedotin Demonstrates Improved Outcomes Versus Rituximab Plus Polatuzumab Vedotin in Patients (Pts) with Relapsed or Refractory (R/R) Large B-Cell Lymphoma (LBCL)

Background: The ongoing Phase Ib/II study (NCT03671018) evaluates mosunetuzumab (Mosun) plus polatuzumab vedotin (Pola; M-Pola) in pts with R/R LBCL. The Phase II single-arm expansion cohort showed a manageable safety profile and highly durable responses with M (IV administration)-Pola at primary analysis (Budde et al. Nat Med 2024). To delineate the contribution of Mosun to the M-Pola combination, we conducted an analysis of a Phase II randomized cohort evaluating M (SC administration)-Pola vs rituximab (R)-Pola in R/R LBCL. Methods: Eligible pts had confirmed R/R LBCL (diffuse LBCL [DLBCL] not otherwise specified, follicular lymphoma [FL] Grade [Gr] 3b, high-grade B-cell lymphoma [HGBCL], or transformed [tr]FL) and had received ≥1 prior therapy (including an anti-CD20 antibody). Pts were randomized 1:1 to M (SC)-Pola or R-Pola (stratified by number of prior therapies [1 vs ≥2]). Treatment cycles were 21 days. Mosun SC was given with Cycle (C)1 step-up dosing (5mg on C1 Day [D]1; 45mg on C1D8, C1D15, and D1 of C2-8) to mitigate cytokine release syndrome (CRS). Hospitalization was not mandatory. Pola IV (1.8mg/kg) was given on D1 of C1-6. Rituximab IV (375mg/m2) was given on D1 of C1-8. Pts on R-Pola with disease progression during or at end of treatment (EOT) or stable disease at EOT could cross over to receive M-Pola (up to 8 cycles of cumulative Pola). Primary endpoint was best ORR using Lugano 2014 criteria. CRS events were defined per ASTCT criteria. Results: As of January 30, 2024, 80 pts were enrolled to receive M (SC)-Pola (n=40) or R-Pola (n=40). 20 pts on R-Pola received crossover treatment with M-Pola, and 1 pt in the R-Pola arm did not receive treatment (study withdrawal). For M-Pola vs R-Pola, median age was 71.5 vs 67.0 years, 48% vs 54% had IPI 3-5, and 78% vs 85% had Ann Arbor Stage III/IV disease, 68% vs 82% of pts had DLBCL, 8% vs 3% of pts had Gr 3b FL, and 25% vs 15% of pts had HGBCL; among pts with DLBCL and HGBCL, 13% vs 23% had trFL. Median number of prior therapies in the M-Pola vs R-Pola arms was 2 (range 1-5) vs 3 (range 1-9), 35% vs 39% of pts received prior CAR-T, and 15% vs 23% had prior ASCT. A total of 50% vs 62% of pts on M-Pola vs R-Pola were primary refractory (relapse <6 months after first-line [1L] therapy), 71% (n=10/14) vs 80% (n=12/15) were refractory to CAR-T, and 13% vs 8% had early relapse (6-12 months after 1L therapy). For M-Pola vs R-Pola, ORR was 78% (95% CI: 61.6-89.2) vs 50% (95% CI: 33.8-66.2), CR rate was 58% (95% CI: 40.9-73.0) vs 35% (95% CI: 20.6-51.7), median duration of response was NR vs 10.1 months (95% CI: 3.6-NR; HR 0.40 [95% CI: 0.1-1.2]), median duration of CR was NR in both arms (HR 0.38 [95% CI: 0.1-1.3]), and median PFS was NR vs 6.4 months (95% CI: 4.7-NR; HR 0.48 [95% CI: 0.2-1.0]). OS was NR in the M-Pola and R-Pola arms (median follow-up 18 months). Adverse events (AEs) occurring in ≥30% of pts with M-Pola were injection-site reaction (55%), diarrhea (48%), fatigue (35%), and constipation (30%). AEs in ≥30% of pts with R-Pola were nausea (36%), diarrhea (33%), and fatigue (33%). Serious AE rates were similar between arms (M-Pola, 33%; R-Pola, 26%). The most common Gr 3/4 AE in both arms was neutropenia (M-Pola, 30%; R-Pola, 21%). One Gr 3 febrile neutropenia event occurred with M-Pola. Gr 5 AEs occurred in 2 (5%) pts on M-Pola (both COVID-19 related) vs 1 (3%) pt on R-Pola (hepatic failure). CRS events occurred in 4 (10%) pts on M-Pola (all Gr 1/2, all during C1, all resolved). Median CRS duration was 3 days (range 2-5), and 1 pt was treated with tocilizumab. In pts experiencing CRS, mean IL-6 and IFN-γ levels were elevated ~2-fold 24 hours after Mosun SC treatment. Peripheral neuropathy (PN) occurred in 4 (10%) pts on M-Pola (Gr 1, n=3; Gr 2, n=1) and 2 (5%) pts on R-Pola (all Gr 1). One neurologic event potentially consistent with ICANS was observed with M-Pola (Gr 2 seizure). Serious infections occurred in 9 (23%) pts on M-Pola and 7 (18%) pts on R-Pola (mainly COVID-19/COVID-19 pneumonia). Treatment was discontinued due to AEs in 3 pts on M-Pola (PN, n=2 [1 Pola related]; COVID-19 pneumonia, n=1]); and 2 pts on R-Pola (PN, n=1; pain in extremity, n=1 [neither treatment related]). Conclusions: Fixed-duration outpatient M (SC)-Pola improved efficacy, with higher ORR and CR rate, promising durability, low rates of CRS events, and no excess toxicities vs R-Pola in pts with R/R LBCL. These results strongly support the continued evaluation of the M (SC)-Pola regimen being investigated in the Phase III SUNMO study (NCT05171647).