Final Update of Safety, Efficacy and T-Cell Predictive Biomarkers from a Phase I Trial of Copanlisib+Nivolumab in Patients with Richter's Transformation (RT) or Transformed Non-Hodgkin Lymphoma (tNHL)

Background. Despite novel therapies, outcomes among patients with tNHL and in particular RT remain poor. Copanlisib (COPA) is a selective, small molecule PI3K-α/δinhibitor which has shown clinical efficacy in NHL. Pre-clinical studies in lymphoma have demonstrated synergy between PI3K and PD-1 inhibitors. Here we report the final clinical and translational results of a phase 1 study of COPA in combination with PD-1 inhibitor nivolumab (NIVO) in patients (pts) with R/R RT or tNHL (NCT03884998). Methods. This multicenter, open-label, phase I, investigator-sponsored study enrolled 27 pts with RT or tNHL, age ≥18 years, whose disease has relapsed or was refractory to ≥1 prior line of therapy. The phase I portion of the study followed a standard 3+3 design, followed by a dose expansion of 16 evaluable pts, with two planned dose levels of COPA administered IV (dose level [DL]1 - 45 mg on days 1, 8 and 15 and DL2 - 60 mg on days 1, 8 and 15 of a 28-day cycle). NIVO 240 mg was given IV on days 1 and 15. Pts received up to 24 cycles of therapy. The primary study objective was the maximum tolerated dose (MTD) of the combination; secondary objectives included measures of efficacy (Lugano criteria). Dose limiting toxicities (DLT) were defined as grade (g)≥4 hematologic lasting > 7 days or grade ≥3 non-hematologic toxicities. Exploratory objectives included T-cell functionality assessed prior to treatment, after 1 and 5 cycles of therapy and at the end of treatment using flow cytometry and scRNA-Seq analyses. Results. Of 27 pts, 11 were enrolled in dose finding (8 at DL1 and 3 at DL2) and 16 in dose expansion cohorts. Fourteen pts had RT and 13 had tNHL (12 tFL and 1 lymphoplasmacytic lymphoma). Median (med) age was 65 years (range, 32-77), 89% had ECOG performance status 0-1. Pts had received a med of 4 prior lines of therapy (range, 1-10); 8 pts (2 RT, 6 tNHL) had prior CAR T cells. During dose finding, 8 pts were treated at DL1 (45 mg COPA), of which 2 pts did not complete the DLT period due to rapidly progressive disease and were replaced. No DLTs were observed in 6 evaluable pts at DL1. At DL2 (60 mg COPA), three DLTs occurred in 2 RT pts (g4 febrile neutropenia and g4 neutropenia in one patient, g4 thrombocytopenia in a second patient) leading to the MTD of COPA as 45 mg. The most common treatment-related adverse events (AEs, any grade) were diarrhea (44%), anemia, fatigue, alkaline phosphatase increased, neutropenia, thrombocytopenia and hyperglycemia (37% each). Most common g3-4 AEs were neutropenia (22%) and lymphopenia (15%). All 27 pts are off protocol after a med of 3 cycles (range 1-24). Eighteen pts discontinued therapy due to progressive disease, 7 due to AEs (2 colitis, 1 prolonged lung infection, 1 esophagitis, 1 diarrhea, 1 neck pain), 1 due to patient refusal because of an infusion reaction, and 1 due to preexisting CNS disease discovered after enrollment. Overall response rate (ORR) was 46% among the 26 efficacy-evaluable pts who completed ≥1 cycle of therapy. ORR was 67% in pts with tFL (2 CR and 6 PR) with med progression free survival (mPFS) 4.4 months (95%CI: 1.4-12.2) and med duration of response (mDOR) 3 months (95%CI: 1.7-13.6). ORR was 31% in pts with RT (2 CR and 2 PR) with mPFS 2.0 months (95%CI: 0.7-4.9) and mDOR 15.2 months (95%CI: 3.0-NA). ORR was 63% after prior CAR T (2 CR and 3 PR). Flow cytometry analysis of 5 longitudinal RT samples documented an increase in circulating CD8+ T cell population and a decrease in exhaustion markers (PD1, CTLA-4) with treatment. scRNA-Seq demonstrated downregulation of MYC in tumor cells in 2 responding RT pts. Concurrently, we observed upregulation of IFN-α and IFN-γ signaling pathway in CD4+ and CD8+ T cells from responders. Such pattern was not observed in 3 RT non-responders. By contrast, T-cells from non-responders exhibited downmodulation of the IFN response after cycle 1 of treatment and concurrent activation of p53-regulated genes and apoptotic genes. Thus, we observed key differences in gene expression patterns with T cell subpopulations between the responders and NRs after COPA/NIVO exposure. Discussion. The combination of COPA and NIVO was well-tolerated and highly effective, in pts with tFL and RT. COPA 45 mg IV on days 1, 8 and 15 was the MTD/RP2D. Downregulation of MYC in the tumor and enhanced T-cell IFN signaling following treatment were associated with response in RT.