Background: It is hypothesized that chemotherapy alters the tumor microenvironment (TME) by increasing T cell infiltration and enhancing antigen presentation to T cells, and thereby may enhance immune-mediated cytotoxicity and tumor response to immunotherapy. The goal of this trial (DURVA+) is to evaluate and compare the immune-modulating effects of six different chemotherapy agents with Durvalumab (D), a PD-L1 checkpoint inhibitor. We report here the first analysis of one of the arms of DURVA+: D plus capecitabine (C).

Methods: This is a non-randomized phase 2 trial for adult patients (pts) with refractory advanced solid tumors. Pts received C at a dose of 1250 mg/m2 bid on days 1-14 of 21-day cycle. For the first 3 cycles, D is administered every 2 weeks at a dose of 750 mg IV starting on C1D8. From cycle 4 onwards, D was given at a dose of 1125 mg IV every 3 week (day 1 of each 21-day cycle). Tumor biopsies were required at baseline, C1D8 (prior to durvalumab), C3D8, and optional at progression. Radiologic tumor assessment occurs every 4 cycles by RECIST. iRECIST is assessed as a secondary endpoint. Baseline and at least one on-treatment biopsy were obtained for 11 patients, at least 8 of which are evaluable for pharmacodynamic (PD) response (treatment-induced changes in immune cell tumor content and function).

Results: As of Nov 10, 2021, 14 pts were enrolled to this arm. Median age 57 (range 31-73 years), 7 males and 7 females, 4 black, 8 white, 2 unknown race. Majority of pts had colorectal carcinoma (ca) (n=8), other include (n=1 each): appendiceal, anal squamous cell ca, leiomyosarcoma, pancreatic ca, breast ca, and nasopharyngeal squamous cell ca. Median lines of prior treatment was 3.5 (1-13), 4 pts previously received ICI. Median number of days on study treatment were 83 days. One pt experience partial response (breast) and five pts experienced SD (3 colon, 1 pancreatic, 1 non-uterine leiomyosarcoma) for a median duration of 136 days as of Jan 3, 2022 (data cutoff); one pt with pancreatic ca experienced a 27% reduction in sum of target lesions. Treatment related adverse events (TrAE) of all grades occurred in 13 pts, serious TrAEs (more than grade 2) occurred in 8 pts and included anemia, pulmonary embolism, hyperbilirubinemia, palmar plantar syndrome, neutropenia, lymphopenia and diarrhea. Immune-mediated AEs included transaminitis/hepatitis (n=4, all grade 1/2), hypothyroidism (n=2, grade 1/2), rash (n=1, grade 1) and colitis (grade 3). PD studies are still ongoing.

Conclusion: The combination of durvalumab with capecitabine is feasible with no unexpected adverse events. PD assessment is ongoing to examine the hypothesis that chemotherapy alters tumor micronenvironment (TME) and T cell activation. The study was sponsored by NCI, and was funded in part by NCI Contract HHSN261200800001E. DURVA+ is registered with, NCT03907475.

Citation Format: Mohamad Adham Salkeni, Geraldine O’Sullivan Coyne, Naoko Takebe, Elad Sharon, Howard Streicher, Jessica Mukherjee, Ashley Bruns, Arjun Mitra, Abdul R. Naqash, Larry Rubinstein, Kristin K. Fino, King L. Fung, Katherine Ferry-Galow, Ralph Parchment, Helen Chen, James Doroshow, Alice Chen. Safety and efficacy of Anti-PD-L1 antibody MEDI4736 (durvalumab) in combination with capecitabine in patients with advanced solid tumors (DURVA+) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT115.