A Phase I Study of Copanlisib and Venetoclax in Patients with Relapsed or Refractory Diffuse Large B-Cell Lymphoma

Introduction:

Diffuse large B-cell lymphoma (DLBCL) is a genomically heterogenous disease, which can be classified into distinct subgroups or clusters (Chapuy et al., Nat. Medicine, 2018, Wright et al., Cancer Cell, 2020). Leveraging knowledge of genetic alterations to design novel personalized treatment strategies may improve outcomes. Preclinical data suggest that the pan-phosphoinositide 3-kinase (PI3K) inhibitor, copanlisib, plus the BCL-2 inhibitor, venetoclax, are synergistic in genetically defined, high-risk, subtypes of DLBCL (Bojarczuk et al., Blood, 2019). We therefore hypothesized that copanlisib plus venetoclax would be an active treatment for patients with relapsed/refractory DLBCL, particularly in those with perturbed B-cell receptor (BCR)/PI3K signaling, which were predicted to be enriched in the relapsed setting. Here, we report the results of a phase I study of copanlisib plus venetoclax for patients (pts) with R/R DLBCL.

Methods:

This is an investigator-initiated, phase I, multicenter trial (NCT04572763) with the primary objective of determining the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D) of copanlisib and venetoclax in pts with R/R DLBCL. A standard 3+3 dose escalation design was employed, in which pts were treated with 60 mg IV copanlisib on days 1, 8, and 15 in addition to escalating doses of daily venetoclax for each 28-day cycle. During cycle 1, a venetoclax dose ramp-up was performed in the outpatient setting to a target dose of 200 mg (dose level (DL) +1), 400 mg (DL +2) and 800 mg (DL +3). The DLT observation period was the first cycle. Patients with progressive disease prior to the end of the DLT window were replaced. Key eligibility criteria included a confirmed diagnosis of DLBCL or high-grade B-cell lymphoma (HGBCL), relapse after or not a candidate for autologous stem cell transplant or chimeric antigen receptor (CAR) T-cell therapy, and adequate hematologic and organ function. Key exclusion criteria were poorly controlled hypertension or diabetes. CTCAE v5 and Lugano criteria were used to evaluate toxicity and efficacy, respectively.

Results:

As of July 21, 2023, 12 pts were evaluable. Patients were treated in DL +1 (n=5), DL +2 (n=3), and DL +3 (n=4). The median age was 62 years (range 24-87). Eighty three percent were male. Median number of prior treatments was 6 (range 1-11) and 10 pts had received prior CAR T-cell therapy. No DLTs were observed, though 2 pts in DL +3 were not able to escalate to 800 mg due to recurrent neutropenia and noncompliance, respectively. The most common toxicities were neutropenia (50%, all gr3/4), thrombocytopenia (42%, 8% ≥ gr3), nausea (33%, 8% ≥ gr3), fatigue (25%, 8% ≥ gr3), and headache (25%, 8% ≥ gr3) (Figure). Hyperglycemia occurred in 2 pts (both gr1). Serious adverse events occurred in 50% of pts and included fracture (n=1), hypoxia (n=2), and death (n=3), which were not related to study treatment. The deaths were related to progressive disease. Fifty percent of pts required drug holds (n=5 for venetoclax, n=5 for copanlisib). Sixty percent required growth factor support. There were no cases of laboratory or clinical tumor lysis syndrome. Sponsor support was withdrawn prior to completion of enrollment to DL+3, but DL+2 (copanlisib plus 400 mg venetoclax) was safe and well tolerated.

The overall response rate (ORR) and complete metabolic response (CMR) rate was 8% (90% CI: 0 - 34%), as 1 pt in DL +1, who had received >5 lines of prior therapy, achieved a CMR, with response lasting 3 months. The median follow-up was 9.1 months. The median overall survival (OS) and progression-free survival (PFS) were 3.5 and 1.8 months, respectively.

Conclusions:

Copanlisib plus venetoclax is feasible and safe in pts with R/R DLBCL, with no DLTs observed. While it is possible that copanlisib plus venetoclax is active in select genomic subgroups of DLBCL, there was limited activity in this unselected, heavily pre-treated patient population.