Pooled Safety Analysis from Phase I and II Studies for Patients with Relapsed Indolent Non-Hodgkin's Lymphoma Treated with Intravenous Copanlisib

Introduction: Copanlisib is a pan-Class I PI3K inhibitor with predominant activity against PI3K-α and PI3K-δ isoforms. It has been studied in indolent and aggressive non-Hodgkin's lymphoma (NHL) and chronic lymphocytic leukemia (Patnaik et al., Anns Oncol 2016, 27:1928-40; Dreyling et al. Anns Oncol 2017, in press; Dreyling et al. ICML 2017); including a total of 168 patients with indolent NHL (iNHL). Late-onset adverse events (AE) have been reported with oral PI3K inhibitors (e.g. late-onset severe diarrhea leading to colitis). We therefore examined whether the intravenous intermittent weekly dosing of copanlisib (3 weeks on and 1 week off) influences the incidence, severity and timing of adverse events (AEs) for patients with iNHL being treated with copanlisib monotherapy.

Methods: Patients with histologically confirmed iNHL (4 subtypes: follicular, marginal zone, small lymphocytic and lymphoplasmacytoid/Waldenstrӧm macroglobulinemia) and relapsed after, or refractory to, ≥2 prior lines of treatment and who had been treated with copanlisib monotherapy in a phase I or II clinical trials were included in this pooled analysis. AEs were coded using MedDRA version 19.0 and the intensity/severity by CTCAE version 4.0/4.03 grading. (Worst-grade [G] AEs reported herein.) Patients were not excluded based on histories of gastrointestinal disorders and prophylaxis for opportunistic infection was not mandated.

Results: The median overall duration of treatment with copanlisib for the pool of 168 iNHL patients was 22 weeks (range 1-206); mean 30 weeks (±30). Treatment-emergent AEs included 94 (56%) G3 AEs and 42 (25%) G4 AEs as worst-grade AEs. A total of 115 (69%) patients had one or more dose modifications due to AEs and 41 (24%) discontinued due to AEs. Per cycle, discontinuations due to AEs were low and evenly distributed over the entire course of treatment (2-3% per patients at risk per cycle). Interruptions or delays due to AEs ranged from 10-20% per cycle over the entire course of treatment. Interruptions or delays (for any reason) were generally less than one week in duration (0.85 ± 0.45 weeks).

The most common AEs included hyperglycemia (51%), diarrhea (36%), hypertension (35%), fatigue (29%), nausea (26%), and neutropenia (25%). The highest G3/G4 events were principally hyperglycemia (32%/6%) and hypertension (27%/0%). These infusion-related events were asymptomatic, transient and largely self-limiting. When viewed as incidence rate per cycle for new or worsening grade events, G3 hyperglycemia was highest in cycle (C) 1 (~30%) and then generally 10% per cycle thereafter, with G4 events occurring in early cycles and generally very low in later cycles (1-2%). The percent of patients with new or worsening G3 hypertension per cycle was relatively constant over the course of treatment. The incidence per cycle of G3 and G4 neutropenia was ~5% and 3%, respectively, for early cycles and then diminished whereas there was gradual rise in G2 events (not exceeding 10%). Diarrhea was generally G1 and G2, with only 8 patients (5%) experiencing a G3 event overall; no reports of G4. The G3 events occurred in early cycles and isolated events in C12 and C13. One colitis event (G4) was reported (the event resolved with antibiotics and did not recur over 5 additional cycles at lower dose). Serious AEs (SAEs) were reported in 81 (48.2%) patients. Events reported in ≥4 patients included pneumonia (10%), pyrexia (5.4%), hyperglycemia (4.8%), pneumonitis (4.8%), neutropenia (2.4%), diarrhea (2.4%), and lung infection (2.4%). There were 6 treatment-emergent deaths overall, with 3 deemed to be drug-related: lung infection, respiratory failure, and embolism.

Conclusions: The incidence and pattern of severity for AEs associated with intermittent IV dosing of copanlisib in iNHL patients over the course of treatment indicates a very different safety profile than that reported with continuous daily oral dosing of PI3K inhibitors. The most prominent events, namely hypertension and hyperglycemia, were transient and mostly self-limiting, with G3 events observed at a constant rate over the course of treatment. There was a low incidence of individual SAEs and no evidence for a pattern of late-onset AEs.