Cell-Intrinsic and Immune-Mediated Effects in Lymphoma Patients Treated with Golcadomide and Associations with Treatment Outcome

Introduction: Compounds inducing cereblon-mediated degradation of Ikaros and Aiolos (I/A), like lenalidomide (LEN) and avadomide, have demonstrated antitumor activity in non-Hodgkin lymphoma (NHL), with significant but limited cell-autonomous activity in diffuse large B-cell lymphoma (DLBCL) cell lines. Their pharmacodynamic activities also include T-cell activation, increased macrophage tracking, and increased regulatory T lymphocytes (T-regs) in peripheral blood (PB) [Carpio C et al Blood 2020]. High immune infiltration in tumors at baseline (BL) was correlated with improved outcome (Risueño A et al Blood 2020). Golcadomide (GOLCA) is a potential first-in-class cereblon E3 ligase modulatory drug (CELMoD™) that, compared with immunomodulatory (IMiD) agents like LEN, has shown more profound cell autonomous activity and immune stimulation in preclinical DLBCL models (Lopez-Girona A et al Hematological Oncology, 2021). Its potent I/A degradation, prolonged exposure, and high tissue penetration predicted superior GOLCA activity vs IMiDs, including in tumors with low immune infiltration. In the CC-99282-NHL-001 study, GOLCA ± rituximab showed a manageable, predictable safety profile characterized by an on-target toxicity of neutropenia and promising efficacy in relapsed/refractory NHL in heavily pretreated pts (Chavez JC et al Blood 2023). Here, we describe pt immune profiles, their modulation by GOLCA and association with objective clinical response in the CC-99282-NHL-001 DLBCL biomarker cohort.

Methods: Cell intrinsic and immune-mediated activity of GOLCA in evaluable pts with DLBCL enrolled in the study were assessed by RNA-Seq, whole genome sequencing, immune-fluorescence and flow cytometry (FC) in screening and on-treatment biopsies, and by FC in PB samples.

Results: I/A degradation was necessary but not always sufficient for GOLCA activity. We observed deep I/A degradation in biopsies and PB samples from all evaluable pts independent of response, suggesting additional tumor-related features influence response to GOLCA. In contrast to IMiDs, efficacy of GOLCA (best overall response and progression-free survival) was independent of BL tumor microenvironment (TME) content. Despite this, BL CD4 naive T cells, T follicular helper and CD8 T cell levels were higher in responders while mono/macrophages were lower. GOLCA sensitivity analysis was run separately in immune infiltration (TME+) vs high B-cell content (TME−) tumors to explore immune mediated and tumor intrinsic mechanisms. Low leukocyte migration and low chemokine signaling were enriched in TME− responders. TME+ tumors showed distinct expression levels of genes involved in immune activation and trafficking (↑AIRE, ↑CLDN12, ↑IQGAP3, ↓MADD, ↓UBE7). GOLCA responders showed a significant increase in pathways of chemokine and cytokine receptor binding and chemokine signaling. Interferon (IFN)-stimulating genes and other proteins involved in immunoregulatory/inflammatory processes (IFI6, OAS3, CCL3L etc) were upregulated in responders, with a modest increase in pathways like tumor necrosis factor alpha and IFNα/β. GOLCA-induced I/A degradation led to increased natural killer- (NK) and T-cell activation with dose/schedule-dependent T-reg increase and decrease in naive T cells in PB. Response was associated with increased PD-1, potentially due to increased T-cell activation. GOLCA responders showed a significant decrease in B-cell content and increase in NK and dendritic cells, macrophage and T-cell subsets. In these responders, there was positive correlation of T-regs between PB at Cycle 1 Day 15 and on-treatment biopsies. Other T-cell populations like naive and memory T cells showed anti-correlation, suggesting peripheral changes may be due to trafficking. In non-responders, these changes were less pronounced.

Conclusion: These results show GOLCA fulfills its design intent, causing deep and prolonged I/A degradation in NHL lesions resulting in BL TME-independent activity, compared with IMiDs like LEN which are TME-dependent. BL immune profiles and their modulation on treatment link to treatment effect and imply an immune contribution to GOLCA efficacy. These data support broader activity of GOLCA than IMiDs in NHL, with high potential for combination with anti-lymphoma agents acting on tumor B cells and/or TME.