Analysis of Immune Pharmacodynamic and Baseline Biomarkers in Patients with Relapsed or Refractory Follicular Lymphoma Treated with Mosunetuzumab in Combination with Lenalidomide

Background: Mosunetuzumab (M) is a bispecific monoclonal antibody targeting CD20 and CD3 that redirects T cells to eliminate B cells and has shown efficacy in patients (pts) with relapsed/refractory (R/R) follicular lymphoma (FL; Budde et al. Lancet Oncol 2022). Lenalidomide (Len) is clinically active in FL, and its immunomodulatory activity offers potential synergistic efficacy when combined with M. Here, we present preliminary data from a Phase Ib/II trial (NCT04246086), which aimed to characterize pharmacodynamic biomarkers and investigate baseline (BL) features of early progressive disease (PD), in pts with R/R FL treated with M+Len.

Methods: Eligible pts had Grade 1–3a R/R FL and received ≥1 prior systemic anti-cancer therapy. Pts received ≤12 cycles of M+Len (Cycle [C] 1: 21 days; C2–12: 28 days). In C1, step-up doses of intravenous M were given on C1 Day (D) 1 (1mg) and C1D8 (2mg), with the target dose (30mg) given on C1D15 and then on D1 of C2‒12. Len (20mg) was administered orally on D2‒22 (n=10) and on D1–21 of C2–12 for the remaining 19 pts. Activation of immune cells during M+Len treatment was assessed in longitudinal collections of peripheral blood by flow cytometry, and plasma IL-6 levels were evaluated by enzyme-linked immunosorbent assay. CD20 expression was assessed by immunohistochemistry in BL and PD biopsies. CD20 loss was defined as ≤5% CD20+PAX5+ cells. Mutation profiling was performed by next-generation sequencing (NGS). A modified version of the AVENIO circulating tumor (ct)DNA analysis workflow and pipeline (Roche; for research use only) was used for NGS, based on the previously described CAPP-Seq technology (Kurtz et al. J Clin Oncol 2018).

Results: A transient reduction in circulating T-cell counts (likely due to margination) and increase in percentage (%) of T cells expressing the activation marker CD69 were observed after M monotherapy in C1, and these kinetics were maintained in subsequent cycles, particularly in the CD4 T-cell population. Maturation of T cells, measured as % of T cells expressing human leukocyte antigen DR, increased after C2, following the addition of Len, and was sustained throughout M+Len treatment. The % of circulating programmed cell death protein 1 (PD-1)+ CD4 T cells increased after the first cycle of M, implying a rapid activation of T cells by M, and progressively decreased after C2. A decrease in % of PD1+ CD4 T cells suggested that sustained PD-1 expression, often a sign of T cell exhaustion, does not occur during M+Len treatment. Natural killer (NK) cells showed transient activation following M infusion (assessed as margination and % of CD69+ NK cells), and a sustained increase in functional markers, like granzyme B, during treatment with M+Len. IL-6 secretion was often enhanced after the first M infusion in C1. However, increased IL-6 levels were not observed after M+Len co-administration in subsequent cycles.

As of April 25, 2022, 29 pts had been enrolled and were evaluable for efficacy. Best objective response rate was 89.7%, with complete metabolic response (CMR) observed in 21 pts (72.4%). Five cases of early PD (≤6 cycles of treatment) were observed. BL CD20 expression was heterogeneous (n=25 BL biopsies), and one CD20-negative pt experienced early PD. Two pts with early PD showed CD20 loss at progression. ctDNA analysis was performed in a subgroup of 5 pts with PD vs 5 with CMR for signal seeking. A trend towards higher median ctDNA levels at BL, measured as mutant molecules per mL, was observed in pts with early PD (400.1) vs CMR (58.6). Moreover, a higher median number of somatic variants was found in plasma (145.0) vs tissue (120.0) at BL, with a lower median concordance of plasma variants in tissue in pts with early PD (n=4; 43.8%) compared with those achieving CMR (n=5; 94.1%), suggesting higher tumor heterogeneity in early PD.

Conclusions: In pts with R/R FL, our preliminary data suggest that M+Len activates the immune system. Increased % of T and NK cells expressing maturation and functional markers were observed during M+Len treatment. In contrast, addition of Len to M did not enhance IL-6 secretion, a cytokine known to be related with cytokine release syndrome, which is consistent with the manageable safety profile previously reported for this combination. Future investigations with a larger sample size are needed to further assess the prognostic value of CD20 expression in response to the M+Len combination and the prognostic value of ctDNA measurement at BL.