An Open-Label, Phase Ib Study of Duvelisib (IPI-145) in Combination with Bendamustine, Rituximab or Bendamustine/Rituximab in Select Subjects with Lymphoma or Chronic Lymphocytic Leukemia

Background: Duvelisib is a potent inhibitor of the δ and γ isoforms of phosphoinositide-3-kinase (PI3K) being developed as a potential therapeutic in hematologic malignancies including B and T cell lymphoma and chronic lymphocytic leukemia (CLL). In a phase I study of single agent duvelisib (D), ORR of 52% was seen in pts with indolent non-Hodgkin’s lymphoma (iNHL) and 47% in CLL. Bendamustine (B), rituximab (R), and their combination have demonstrated proven activity in iNHL and CLL. Combining duvelisib with either bendamustine or rituximab alone or in combination with each other may improve response rates and the durability of remission. The goal of this Phase 1b, open-label, three-arm, non-randomized, dose escalating, safety and tolerability trial is to characterize the safety, maximum tolerated dose (MTD) and preliminary efficacy profile of duvelisib given in combination with rituximab (Arm 1-DR), bendamustine plus rituximab (Arm 2-DBR) or bendamustine (Arm 3-DB) in subjects with select relapsed/refractory lymphoma or CLL.

Methods: Pts had relapsed CLL or NHL, ECOG performance status (PS) ≤2, and adequate organ function. The subject population during dose escalation was limited to relapsed NHL. During the dose expansion phase, each treatment arm enrolled to population specific cohorts to continue to assess efficacy. Arm 1 (DR) received rituximab 375 mg/m² IV weekly for 2, 28 day cycles plus duvelisib PO BID up to 12 cycles. Arm 2 (DBR)received rituximab 375 mg/m² IV weekly for 2, 28 day cycles, bendamustine 90 mg/m² IV on Days 1 and 2 of the first six cycles plus duvelisib PO BID up to 12 cycles. Arm 3 (DB) received bendamustine 120 mg/m² IV on Days 1 and 2 of the first six cycles plus duvelisib PO BID up to 12 cycles. Three different dose levels of duvelisib were explored, 25, 50, and 75 mg PO BID. DLTs were defined as: febrile neutropenia, G4 neutropenia ≥7 days, G4 thrombocytopenia ≥ 7 days or G3 thrombocytopenia with bleeding, Grade 4 AST/ALT, Grade 2 hyperbilirubinemia ≥7 days, ≥ Grade 3 non-hematologic toxicity ≥7 days (excluding alopecia), Treatment delay of ≥7 days due to unresolved toxicity that prevents re-dosing, hepatocellular injury (defined as ALT>2 x ULN and (ALT/ULN)/(ALP/ULN) >5) and bilirubin >2 x ULN or jaundice ± alkaline phosphatase <2 x ULN. Patients were evaluated for response every 3 cycles according to specific criteria for their disease.

Results: Between August 2013 and May 2014, 32 pts, median age 66 years (44-78) were enrolled to the study, 12 NHL pts on the dose escalation portion and 20 pts on dose expansion (13 CLL, 7 NHL). Patients had a median of 4 prior therapies (1-11). In arms 1 (DR) and 2 (DBR), no dose limiting toxicities were seen at the highest dose level of duvelisib (75 mg bid). In arm 3 (DB) in which a higher dose of bendamustine is used 1 pt developed a DLT at the 50 mg BID dose level of duvelisib (febrile neutropenia, neutropenia ≥ 7 days, thrombocytopenia ≥ 7 days, and liver toxicities which resulted in a treatment delay of ≥ 7 days). Dose escalation continues in this arm as the MTD has not reached. Patients on the dose expansion portion of the study are receiving duvelisib at 25 mg BID due to emerging data of duvelisib monotherapy showing no advantage in doses greater than 25 mg BID in these histologies. The AE profile is consistent with the toxicities of the single agents. The most common AEs > grade 3 were neutropenia (28% overall; [Arm 1 (DR), 27%]; [Arm 2 (DBR), 38%]), and rash (16% overall; [Arm 1, 14%]; [Arm 2, 25%]). Grade 3 or higher AST/ALT increases were seen in 2 out of 12 patients on Arm 1, 2 out of 8 patients on Arm 2 and no patients on Arm 3. There have been 2 deaths (cardiac arrest and pneumonia), both on Arm 1. Twenty one pts were evaluable for response with an ORR of 81% (10% CR, 71% PR, 14% SD and 5% PD). With a median follow up of 4.0 months, time to event analyses are immature. However, Kaplan-Meier estimate of PFS at 3 months is 87%. PK analysis is consistent with the monotherapy Phase I trial of duvelisib.

Conclusions: Initial early analysis of duvelisib administered in combination with bendamustine and rituximab suggests these combinations to be generally well-tolerated with encouraging. Further follow-up is required to better characterize response rates and durability of remissions.