Baseline Biomarkers of T-Cell Function Correlate with Clinical Responses to Odronextamab (REGN1979), and Loss of CD20 Target Antigen Expression Identified As a Mechanism of Treatment Resistance

BACKGROUND: Odronextamab (REGN1979) is a first-in-class, hinge-stabilized, fully human CD20 x CD3 IgG4-based bispecific antibody that binds to CD20-expressing cells and CD3 on T cells, targeting CD20+ cells via T-cell-mediated cytotoxicity independent of T-cell receptor recognition. Patients with relapsed/refractory B-cell non-Hodgkin lymphoma were treated with odronextamab in a first-in-human, Phase 1 study (NCT02290951). Patient biopsies were analyzed to investigate the association of clinical response and relapse with B- and T-cell markers.

METHODS: Tumor biopsies collected at baseline and at disease progression were analyzed by semi-quantitative CD20 chromogenic immunohistochemistry (IHC). B-cell antigen and immune cell multiplex immunofluorescence was also performed. Bulk tumor tissue nucleic acid isolates were analyzed by whole exome DNA sequencing. Peripheral blood mononuclear cells isolated from baseline samples were analyzed by highly multiplexed flow cytometry T-cell immunophenotyping assays. A single biopsy sample of fresh tumor tissue obtained at disease progression was analyzed by flow cytometry and single cell RNA sequencing.

RESULTS: At baseline, higher levels of tumor-infiltrating CD4 and CD8 T cells were present in patients with a complete or partial response to odronextamab (N=13) compared with patients with no response (N=14); median (interquartile range) in CD4 responders vs non-responders was 2496 cells/mm² (869–3940) vs 475 cells/mm² (208–2714), and in CD8 responders vs non-responders was 3289 cells/mm² (1981–7060) vs 489 cells/mm² (110–3811), respectively. However, clinical responses were also observed in patients with low levels of baseline T-cell infiltration. Clinical efficacy was also associated with systemic T-cell immunophenotypic subsets in peripheral blood at baseline, including T-cell subset distribution, co-stimulatory molecule expression, and checkpoint molecule expression. Patient response to odronextamab was independent of the baseline overall frequency of intratumoral CD20+ cells (N=51) or intensity of CD20 expression (N=28) in tumor cells. The presence of CD20(-)/Pax5(+) lymphoma cell subsets at baseline, a potential source of CD20(-) disease escape, did not preclude durable clinical responses (N=27). However, loss of CD20 expression was observed in 6/9 biopsy samples taken at relapse. Genomic analysis of relapse samples identified CD20 gene mutations in 3/8 cases (two truncating and one c-terminal); these three patients were treated at an active odronextamab dose level, and had experienced response before progressing. In one case with available repeat samples, CD20 mutation observed at disease progression had not been detected at Week 5. Loss of CD20 expression by IHC was observed in two cases in the absence of a CD20 gene mutation, suggesting an alternative epigenetic molecular mechanism of CD20 loss.

Single-cell analysis of a fresh tissue sample from a case of progressive disease, occurring after a prior complete response, showed a complete loss of CD20 expression on the cell surface while maintaining abundant tumor-infiltrating effector T cells.

CONCLUSIONS: Preliminary analyses suggest that high levels of baseline tumor-infiltrating T cells may be associated with clinical response to odronextamab. Systemic T-cell homeostasis at baseline may be a potential predictor of clinical benefit with odronextamab, and further investigation is warranted. Although baseline CD20 expression level did not correlate with efficacy, loss of CD20 expression was observed frequently in progressive disease. Several CD20 gene mutations were detected in patient samples at clinical progression, suggesting potential target antigen-dependent disease escape. A newly detected CD20 mutation at disease progression suggests resistance may not always be mediated by outgrowth of pre-existing CD20(-) disease subclones.

The use of baseline CD20 expression as a predictive biomarker is not supported at this time. Durable responses observed in patients with CD20(-) subclones suggest a bystander immune effect may be induced by odronextamab. CD20 loss appears to be an important mechanism of treatment resistance, which may help inform future clinical development strategies. Additional mechanisms of resistance are under investigation, including an evaluation of the immune inhibitory microenvironment at disease progression.