A phase I/II study of regorafenib (R) plus avelumab (A) in digestive tumors.

Abstract

TPS3125 Background: Several preclinical studies have shown that simultaneous blockade of programmed death(PD) 1/PD-L1 and neoangiogenesis induces synergistic anti-tumour effect in vivo. R is a multikinase inhibitor approved for the treatment of metastatic colorectal (CRC) and gastro intestinal stromal tumor (GIST). A is a PD-L1 inhibitor approved for the treatment of Merkel carcinoma. We hypothesized that R in association with A could be synergistic and feasible in patients (pts) with tumors of the digestive tract. Methods: This is a multicenter, prospective phase I/II trial assessing R+A in 4 cohorts of pts with advanced pretreated: A) CRC not MSI-H or MMR-deficient, B) GIST, C) Oesophageal/gastric carcinoma, D) Biliary tract/hepatocellular carcinoma. Primary objectives are to determine the recommended phase II dose (RP2D) of R+A in phase I, and assess the best overall response defined as per RECIST v1.1 with A+R in phase II. In phase I, 2 doses of R will be investigated: 120mg, 160mg, daily, 3 weeks on/1 week off with fixed dose of A: 10mg/kg every 2 weeks. In phase II, all pts will received the RP2D of R with A. Main eligibility criteria are: -adult pts with metastatic, histologically confirmed tumor of 1 of the 4 cohorts, with measurable, progressive disease, after ≥ 1 previous line of systemic therapy. Primary endpoint is toxicity according to NCI-CTCAE v4.0 and incidence rate of DLT at each dose level during the first 28 days (I); antitumor activity in terms of best overall response (II). Secondary endpoints encompass: Objective Response Rate (ORR), Progression Free Survival (PFS), Growth modulation index, 6-months,1-year PFS and Overall Survival, PKs, Pharmacodynamics on (I) mandatory blood samples at baseline/on treatment, archived tumor tissue, (II) on optional biopsy at Baseline/after 4 weeks of R+A focusing on TAM, Lymphocytes infiltrates, PD-L1, VEGFR, PDGFR, HIF1alpha expression. Phase I will follow a classical 3+3 design with 2 dose levels and a maximum of 12 pts. A Bayesian approach will be used in phase II, with a maximum sample size of 50 pts/cohort. At each update of interim analysis, a stopping rule for inefficacy will recommend stopping the trial if there is a high predictive probability (≥80%) that ORR is < or = to the futility bound p0 = 20%. Clinical trial information: NCT03475953.