Mosunetuzumab with response-driven lenalidomide augmentation achieves high response rates and immune reprogramming in untreated follicular and marginal zone lymphoma: A multicenter Phase 2 trial

Abstract Background:Mosunetuzumab (Mosun) is a CD20xCD3 bispecific antibody with high efficacy in treatment of relapsed/refractory follicular lymphoma (FL). It is studied as first-line treatment alone or with lenalidomide (Len), an immunomodulator which may enhance T-cell effector function. However, full-dose Len is associated with frequent adverse events (AEs), while lower doses are effective in FL (Zucca et al, Blood 2019). In this multi-center, investigator-initiated trial, we studied the efficacy of Mosun with selectively applied, response-adapted, immune augmentation using low-dose Len, in previously untreated patients (pts) with FL or marginal zone lymphoma (MZL). Methods:Pts with high-burden FL or MZL requiring therapy received first-line subcutaneous Mosun for 8 planned cycles (C) with step-up dosing in C1 (day 1: 5mg, day 8 & 15: 45 mg; then 45mg on day 1 of C2-C8). Pts who did not attain a complete response (CR) at interim response assessment (IRA) after C4 received Len (10mg daily, dosed continuously) during C5-C8. Primary response assessment occurred after C8, with an option to extend Mosun+Len to C12 in case of ongoing PR. The primary endpoint was CR rate (CRR) at end of treatment (EOT) using PET-CT Lugano or international splenic MZL criteria. Secondary objectives included safety, overall response rate (ORR), progression-free (PFS) and overall survival (OS). Len-induced immunomodulation was analyzed by peripheral blood flow cytometry and RNAseq on CD8+ T cells isolated by MACS microbeads (Miltenyi Biotec). Results:The trial completed enrollment with 52 pts (median age, 65 [range, 30-89]; 51% women), including 67% with FL and 33% with MZL (14% extranodal, 13% nodal/unspecified, 6% splenic). As of July 2025, 43 pts were evaluable for EOT response, which will be updated at the meeting. The most common AE's included injection site reaction (53%; grade [G] 1, 51%, G2, 2%), fatigue (45%; G≥3, 1%), and rash (43%; G≥3, 8%). Serious AE's (in 33% of pts) included cytokine release syndrome (CRS, 12%), abdominal pain (4%), and pneumonia (4%). Three pts discontinued treatment for toxicity (1 pneumonia; 1 CRS; 1 G4 hepatitis on Mosun+Len). CRS occurred in 14 pts (27%), all G1 and during C1. Two pts received tocilizumab and 2 received dexamethasone. CRS was associated with higher baseline lymphocyte count (P=0.042) and splenic MZL (2 of 3 pts). Len augmentation was administered to 17 pts (33%) who did not attain a CR at IRA after C4; 4 pts extended therapy to 12 cycles. Common AEs related to Len (in n=17 pts) included rash (24%), fatigue (18%), and neutropenia (18%; G3/4, 9%). The ORR was 86% (95% CI=72-95). The CRR was 84% (95%CI=69-93), 81% in FL and 88% in MZL. Minimal residual disease (MRD) by ctDNA (Roche Avenio NHL assay) was examined in 15 initial FL pts in CR, and 14 (93%) had undetectable ctDNA (median limit of detection 4.5x10⁻5). With median follow up of 14 months (range, 2-35), PFS at 12 months is 87% (95%CI=73-94). Two pts died (one from transformed large B-cell lymphoma [LBCL], one from a neurodegenerative disorder). Of 6 pts with progression/early relapse events, 4 showed LBCL transformation, and CD20 expression was retained in 4 cases. At the time of IRA, pts who required Len augmentation had lower circulating NK cells (P=0.0021), HLADR+ T cells (P=0.04) and Tregs (P=0.0072) than those with early CR; these differences resolved at the EOT after Mosun + Len. RNAseq showed early (C2) upregulation of CD8+ T cell activation (TNFRSF9/4-1BB, IRF4, CD38, GZMB) and regulation (HAVCR2/TIM-3, CTLA4, LAG3, PDCD1) genes. However, at IRA we found no significant differentially expressed relevant genes in CD8+ T cells between pts with or without early CR. In pts who started Len augmentation, T cells at C6 showed upregulation of IFNG, RPTOR, ITK, KLRG1, and downregulation of CCR7, LEF1, TCF7, consistent with modulation towards effector memory states. Among 14 pts on Mosun+Len who reached EOT response evaluation, 11 (79%) attained CR while 3 had progressive disease (PD) with LBCL transformation. Conclusions:Fixed-duration Mosun with response-adapted low-dose Len augmentation attained a high CR rate, including MRD negativity, sparing most patients the toxicities of Len. Low-dose Len induced CD8⁺ T-cell effector reactivation but did not prevent progression in cases of occult LBCL transformation. Progression was not usually associated with CD20 antigen loss, suggesting immune evasion as an alternative resistance mechanism.