A phase Ib study of guadecitabine and durvalumab in patients with advanced hepatocellular carcinoma, pancreatic adenocarcinoma, and biliary cancers.

Abstract

574 Background: Pancreatic (PC) and biliary cancers (BC) are cold tumors with limited activity of single agent immune checkpoint inhibitors. DNA methyltransferase inhibitors (DNMTi) have immunomodulatory effects manifested by upregulation of interferon pathways and expression of endogenous retroviral signatures. We performed a phase Ib study of the DNMTi guadecitabine (G) and durvalumab (D) in patients (pts) with hepatocellular carcinoma, PC and BC. We report initial results from the PC and BC cohorts. Methods: This is a phase Ib study to establish the maximum tolerated dose (MTD) of the combination (dose escalation; 3+3 design) and evaluate the objective response rate (ORR) in expansion cohorts of PC and BC. G was given at escalating doses of 30 mg/m2 and 45 mg/m2 subQ for 5 days q 28 days. D was given at 1500 mg IV on day 8 of each cycle. Expansion was started at the MTD. Eligibility criteria included ECOG 0-1, ANC ≥ 1,500, platelets > 100,000, albumin ≥ 2.5 g/dL, total bilirubin ≤ 2.5 x upper limit of normal, failure of ≥ 1 prior line of therapy for advanced disease. Prior anti PD-1/PDL-1 was not allowed. Tumor biopsies were performed during screening and on cycle 3 day 1. Results: A total of 11 pts were treated in dose escalation; 3 at dose level 1, and 8 (6 evaluable for DLT) at dose level 2. Given lack of dose-limiting toxicities, MTD was the highest planned dose of G at 45 mg/m2. 24 pts with PC and 23 pts with BC were treated in dose escalation and expansion. For the PC cohort: median age was 66 (43, 93), 29% female, 67% ECOG 1, and median number of prior therapies 2 (1,3). For the BC cohort: median age was 61 (41, 85), 52% female, 78% ECOG 1, and median number of prior therapies 1 (1,3). All grade treatment related AEs in ≥10% of pts were neutropenia (55%), leukopenia (50%), anemia (33%), fatigue (33%), thrombocytopenia (17%), nausea (15%), and anorexia (10%). Grade 3/4 AEs in ≥10% of pts were neutropenia (40%), leukopenia (35%), and anemia (13%). There was 1(5%) PR in PC cohort lasting > 24 mo and ongoing and 1(5%) in BC cohort lasting 12 mo; both were in MSS pts. SD was noted in 7/24 (29%) PC and 5/23 (22%) BC pts, 8 of which lasted ≥4 mo. Median PFS for PC and BC was 2.1 mo [1.9, 3.8] and 1.9 mo [1.4, 2] respectively. Median OS for PC and BC was 4.4 mo [3.4, NR] and 8.6 mo [6.4, NR]. Six and 12 mo OS rates are 38% [21, 66] and 27% [13, 56] for PC; 69% [52, 91] and 35% [19, 63] for BC. 4% of PC pts and 42% of BC pts received another therapy after progression. Conclusions: The combination of G and D has a manageable safety profile in pts with advanced PC and BC; grade 3/4 AEs were limited to myelosuppression. The combination had limited clinical activity based on ORR and PFS in this unselected, pretreated population; however, a subset of pts appeared to derive prolonged clinical benefit, and OS rates were comparable to standard second line chemotherapy, despite a minority of pts receiving subsequent treatment. Biomarker analyses are ongoing. Clinical trial information: NCT03257761.