Safety and Preliminary Clinical Activity of REGN1979, an Anti-CD20 x Anti-CD3 Bispecific Antibody, in Patients with B-NHL Previously Treated with CD20-Directed Antibody Therapy

Introduction: REGN1979 is a CD20xCD3 bispecific antibody (Ab) based on an IgG4 isotype modified to reduce Fc binding. It is designed to bind T cells (via CD3) and CD20-expressing cells and is being evaluated for the treatment of B-lymphoid malignancies. Binding of the Ab to both targets results in specific, local T-cell activation and cytotoxicity, and this mechanism of action is distinct from that of current anti-CD20 Abs. Updated findings are presented for this open-label, multi-center, dose escalation Phase 1 trial of REGN1979 in patients (pts) with B-NHL.

Methods:This study uses a 3+3 design to assess the safety, pharmacokinetics (PK), pharmacodynamics (PD), and clinical activity of REGN1979 in dose escalation cohorts. Pts must have progressed on or after at least 1 prior line of therapy and must have received at least 1 CD20-directed therapy. Based on the initial PK data, the regimen was revised to 12 weekly doses followed by maintenance q2w dosing for 12 doses.Tolerability is determined using pre-specified dose limiting toxicity criteria; responses are assessed by Cheson 2007 and Lugano 2014 criteria for metabolic response.

Results: 38 pts with B-NHL comprising DLBCL (22), FL (11), MCL (3), and MZL (2) subtypes were treated with REGN1979. Pts had a median of 3 (range 1-6) prior regimens. Pts were treated at doses ranging from 0.03 to 7.0 mg REGN1979. At the time of data cut-off, 10 B-NHL pts were on treatment, 6 pts completed treatment, and 22 pts discontinued therapy (the majority [17] due to progressive disease). The median number of doses received was 5 (range 1-21).

There were no DLTs. In patients with B-NHL, the most common treatment-related adverse events (TR-AEs) were pyrexia (66%), chills (50%), cytokine release syndrome (CRS) (42%), fatigue (26%), hypotension (23%), and infusion-related reaction (IRR) (21%). The most common grade ≥3 TR-AEs were CRS, IRR, hypotension, anemia, AST increased, hypophosphatemia, and dyspnea (each 5.3%). Five pts died due to their disease while on treatment or within 30 days of discontinuation; one had grade 5 multi-organ failure deemed unrelated to study treatment. Three pts discontinued REGN1979 due to Gr3 hemolysis, Gr3 fatigue, or Gr3 bilateral lower extremity leg pain. IRR and CRS events were most commonly reported with the initial infusions of REGN1979. Notably, modifications to pre-medications and drug administration instructions decreased the incidence and severity of IRR/CRS.

At REGN1979 doses of 5-7 mg (11 pts), 5 patients achieved a PR (Fig 1). Overall, 10 pts had a best response of PR by Cheson criteria (median duration of response 80 days), and these were distributed among those with FL (4 pts), DLBCL (4 pts), MZL (1 pt), and MCL (1 pt). 12 pts had a best response of SD that was durable (median duration of 287 days). In all cohorts using Lugano criteria, 2 pts had a complete metabolic response, and 7 pts had a partial metabolic response, with a median duration of response of 209 days. After observing rapid clearance, the dosing regimen was changed, and trough REGN1979 concentrations in serum became measurable during qw and q2w dosing and in higher dose groups. Ctrough levels of 350 mcg/L were noted in the 7 mg cohort, close to the level observed after a 0.1 mg/kg single dose in a murine tumor model. In this animal model, tumor control was observed after a 0.4 mg/kg single dose at exposures that were 2-fold (Cmax, Cav and AUC) to 4-fold (Ctrough) higher than those observed in the 7 mg cohort in patients; therefore further dose escalation continues.

PD/Cytokines: An interim analysis of cytokine levels (n= 35 pts) indicates REGN1979 induces cytokine release with a trend towards dose response, as expected based on mechanism of action. An increase in IL-6, IL-10, TNFα, and to lesser extent IFNγ and IL-2 was observed. The occurrence of cytokine response was associated with symptomatic IRR/CRS.

Conclusion: At the doses evaluated, REGN1979 demonstrated an acceptable safety profile in pts with B-NHL; most TR-AEs comprised IRR/CRS and were well managed with supportive care. No clinically significant neurological toxicity was observed. At dose levels of 5-7 mg REGN1979, and with a revised dosing schedule that allows for measurable Ctrough levels, the preliminary ORR in B-NHL is 45% in a diverse range of B-NHL subtypes. Dose escalation continues towards drug exposure levels commensurate with optimal tumor control in a murine lymphoma model.

View largeDownload slide

View largeDownload slide

Close modal