Phase 1 Study of REGN1979, an Anti-CD20 x Anti-CD3 Bispecific Monoclonal Antibody, in Patients with CD20+ B-Cell Malignancies Previously Treated with CD20-Directed Antibody Therapy

Introduction: REGN1979 is a hinge-stabilized CD20xCD3bispecific full-length antibody (Ab) based on an IgG4isotype modified to reduce Fc binding. It is designed to bind T cells (via CD3) and CD20-expressing cells. Cross-linking results in specific, local T cell activation and engagement ofcytolytic functions, independent of T cell receptor mediated recognition. This mechanism of action (MOA) is distinct from approved anti-CD20 Abs, and may provide additional therapeutic benefit. This report describes safety and clinical outcomes in 25 patients (pts) treated with REGN1979.

Methods: The study uses a 3+3 design to assess safety, pharmacokinetics (PK), pharmacodynamics (PD), and activity of REGN1979 inpts with NHL and CLL, in separate dose escalation cohorts. REGN1979 is administered IV over 1-4 hours, weekly (QW) for 4 doses followed by every 4 weeks (Q4W) for 5 doses. Each dose level (DL) consists of an initial starting dose followed by a higher (step-up) dose.Pts with either progressive disease (PD) after initial response or suboptimal initial response could be retreated.

Results: As of clinical cut-off, 25pts had been enrolled and treated with REGN1979 at flat doses ranging from 0.03 - 3.0 mg: 20pts with NHLand 5pts with CLL. NHL subtypes included DLBCL (n=12), FL (n=6), and MCL (n=2). Allpts hada medianof3 (range 1-7) prior regimens, and received a median of 5 (range 2-9) doses of REGN1979. Sixpts remained on initial treatment, 3pts completed treatment, and 16pts discontinued treatment (14 for PD; 2 subject decision). Of 4pts who entered retreatment, 2pts remained ongoing.

The most common treatment-related treatment-emergent AEs (TR-TEAEs) were pyrexia (56%), infusionrelated reaction (IRR) (40%), chills (36%), cytokine release syndrome (CRS) (28%), fatigue (24%), tachycardia (24%), hypomagnesaemia (20%) and hypotension (20%). Grade >= 3 TR-TEAEs occurred in 8 pts(32%): IRR (12%), CRS (8%), hypotension (8%), and pyrexia, tachycardia, dyspnea, anemia, hypoxia, hypertension, tachypnea, transaminases increased, and tumor pain each in 1 pt(4%). All were grade 3 with the exception of 1 ptwith grade 4 hypotension. There were no protocol-defined DLTs, nor grade 5 TEAE. No ptsdiscontinued REGN1979 due to TR-TEAE. IRR and CRS events were most commonly reported with the first administration of both initial and step-up dose with incidence and severity decreasing with increasing exposure.

Tenptsreported at least one TR-SAE: CRS (24%), IRR (20%), and increased AST and tumor pain, each in 1pt(4%). Threeptswith aggressive lymphoma who discontinued REGN1979 due to PD died of their disease within 30 days of their last dose.

Notably, modifications to pre-medication(s) and drug administration instructions have improved tolerability (fewer and less severe IRR/CRS). No clinically significant CNS TEAEs have been observed.

CT-based (Cheson, 2007) anti-tumor activity was observed in 11 ptswith NHL: PR [n=4,median duration of 50d];SD [n=7, median duration of >196d]. The overall response rate (ORR) in NHL ptsacross DLs was 20%. In 2 highest dose levels, ORR was 27% (Fig 1). Of the 14pts evaluated by PET, 4pts had a partial metabolic response, withmedian duration of 244d perLugano Criteria. Twopts with CLL attained SD as best overall response.

PK: REGN1979 concentrations in serum increased with increasing DL and were variable across pts. Variability was less at higher DLs and also decreased over time during therapy. At DLs tested, REGN1979 concentrations did not notably increase over time during treatment.

PD/Cytokines: At the lowest dose tested, 0.03 mg, transient B-cell depletion was observed inpts with measurable circulating B cells.Samples from 12pts were available for cytokine analysis. Interim analysis indicates REGN1979 induces cytokine release as expected based on MOA. An increase in IL-6, IL-10 and to lesser extent IFN-γ was observed. The magnitude of cytokine response generally correlated with symptomatic IRR/CRS.

Conclusion: REGN1979 demonstrated an acceptable safety profile at flat doses of 0.03 - 3.0 mg (< 1% of rituximab dose) inpts with NHL and CLL. Most TEAEs were associated with IRR/CRS, and were managed with supportive therapy and modification to REGN1979 administration. No clinically significant neurological toxicity was observed. Preliminary antitumor activity demonstrates increased activity at higher doses. Dose escalation and treatment schedule optimization continue.

View largeDownload slide

View largeDownload slide

Close modal