mutation and the degree of differentiation or node invasion (p¼0.46 and p¼0.24 respectively). Conclusion:The prognostic impact of RAS mutation was demonstrated in our patients. Treatment for metastatic cancers remains challenging in Tunisia despite recent advances. Thanks to the identification of RAS status, unjustified expenses of anti-EGFR targeted therapy could be avoided.Legal entity responsible for the study: The authors.

Background: Systemic immunity can be regulated via transient interactions of non-colonizing bacteria and immune cells in the small intestinal mucosa. The potential to harness these interactions for IO treatment was first described by Sivan in 2015. EDP1503 is a non-colonizing preparation of a single strain of Bifidobacterium animalis lactis that, when administered orally, in preclinical models, invokes an anti-tumor immune response by activation of innate and adaptive immune mechanisms. The pleiotropic effects of EDP1503 are initiated in the small intestine and mediated by interactions with multiple pattern-recognition receptors inducing a proinflammatory signature in human PBMCs, activating CD8 and NK cell IFNγ production and cytolytic activity. Preclinically, EDP1503 inhibits tumor growth as a monotherapy and in combination with anti-PD-1/L1. Single agent PD-1/L1 therapies have showed limited clinical benefit in previously treated triple-negative breast cancer (TNBC) patients. Here, we report the safety, tolerability, and efficacy of EDP1503 in combination with pembrolizumab in patients with TNBC (NCT03775850). Methods: EDP1503-101 is an open label Phase 1/2 study of EDP1503 in combination with pembrolizumab. TNBC subjects who had received ≥ 1 treatment regimen for metastatic disease were enrolled. Prior anti-PD-1/L1 treatment was permitted. Subjects initially receive 14 days of EDP1503 monotherapy orally twice a day (b.i.d) and then EDP1503 b.i.d. in combination with pembrolizumab 200 mg iv every 3 weeks. The initial 3 subjects received 2 capsules of EDP1503 b.i.d. All other subjects received 4 capsules b.i.d. Safety and tolerability were assessed using CTCAE v5.0. Evidence of anti-tumor activity was based on investigator-assessed objective response (OR), by RECIST v1.1 and iRECIST, and disease control rate defined as OR and/or stable disease (SD) after 4 cycles of therapy. Paired biopsies at baseline and day 14 were used to establish PD-L1 status and investigate pharmacodynamic biomarker changes. Results: As of 1 October 2020, 15 TNBC subjects had been treated (median age of 52, median of 2 prior lines of therapy and ECOG of 0-1). Additional subjects are being recruited, up to a maximum of 30. EDP1503 was well-tolerated as monotherapy and in combination with pembrolizumab. 53% of subjects experienced treatment-related adverse events (TRAEs). Most common TRAEs were GI-related: abdominal distension (20%), decreased appetite (20%), diarrhea (13%), flatulence (13%). No Grade 4-5 TRAEs were observed. 1 Grade 3 TRAE (diarrhea) leading to treatment discontinuation was reported. In 12 subjects receiving the 4 capsules b.i.d dose, 2 partial responses (PR) and 1 SD were observed, giving an objective response rate (ORR) of 18% and a disease control rate (DCR) of 27% in evaluable patients (n=11). Both responders had tumor burden reductions of >65% by RECIST. One responder had 4 prior lines of therapy for metastatic disease, including a PD-L1 inhibitor combination, has a PD-L1 combined positive score (CPS) of 30 and has been on study treatment for 192 days with a 66% reduction in target lesions. The only residual lesion is 6 mm which is PET-negative. The other responder had 2 prior lines of therapy for metastatic disease, was checkpoint inhibitor naive, has a PD-L1 CPS of 2 and remained on study treatment for 275 days with a 73% reduction in target lesions before discontinuing due to treatment-related AEs. The patient with SD had also previously relapsed on a PD-L1 inhibitor and was on study treatment for 226 days before progressing. Conclusions: EDP1503 administered with pembrolizumab is safe and well-tolerated with no Grade 4-5 TRAEs or SAEs. Clinical benefit was observed in a subset of TNBC patients treated with the combination of EDP1503 and pembrolizumab. This study is continuing to recruit TNBC patients at the high dose of EDP1503. Citation Format: Loise Francisco-Anderson, Shamira Shariffudin, Humphrey Gardner, Peter Sandy, Michael Goldberg, Shubhra Kashyap, Mary Abdou, Maria Sizova, Valeria Kravitz, Holly Ponichtera, Mark Carlson, Shannon Argueta, Chris Davitt, Pooja Parameswaran, Michael Chisamore, Mark Bodmer, Duncan McHale. A phase I/II clinical trial of EDP1503 with pembrolizumab for triple-negative breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS11-27.

BackgroundThe small intestinal axis (SINTAX) is a network of anatomic and functional connections between the small intestine and the rest of the body. It acts as an immunosurveillance system, integrating signals from the environment that affect physiological processes throughout the body. The impact of events in the gut in the control of tumor immunity is beginning to be appreciated. We have previously shown that an orally delivered single strain of commensal bacteria induces anti-tumor immunity preclinically via pattern recognition receptor-mediated activation of innate and adaptive immunity. Some bacteria produce extracellular vesicles (EVs) that share molecular content with the parent bacterium in a particle that is roughly 1/1000th the volume in a non-replicating form. We report here an orally-delivered and gut-restricted bacterial EV which potently attenuates tumor growth to a greater extent than whole bacteria or checkpoint inhibition.MethodsEDP1908 is a preparation of extracellular vesicles produced by a gram-stain negative strain of bacterium of the Oscillospiraceae family isolated from a human donor. EDP1908 was selected for its immunostimulatory profile in a screen of EVs from a range of distinct microbial strains. Its mechanism of action was determined by ex vivo analysis of the tumor microenvironment (TME) and by in vitro functional studies with murine and human cells.ResultsOral treatment of tumor-bearing mice with EDP1908 shows superior control of tumor growth compared to checkpoint inhibition (anti-PD-1) or an intact microbe. EDP1908 significantly increased the percentage of IFNγ and TNF producing CD8+ CTLs, NK cells, NKT cells and CD4+ cells in the tumor microenvironment (TME). EDP1908 also increased tumor-infiltrating dendritic cells (DC1 and DC2). Analysis of cytokines in the TME showed significant increases in IP-10 and IFNg production in mice treated with EDP1908, creating an environment conducive to the recruitment and activation of anti-tumor lymphocytes.ConclusionsThis is the first report of striking anti-tumor effects of an orally delivered microbial extracellular vesicle. These data point to oral EVs as a new class of immunotherapeutic drugs. They are particularly effective at harnessing the biology of the small intestinal axis, acting locally on host cells in the gut to control distal immune responses within the TME. EDP1908 is in preclinical development for the treatment of cancer.Ethics ApprovalPreclinical murine studies were conducted under the approval of the Avastus Preclinical Services’ Ethics Board. Human in vitro samples were attained by approval of the IntegReview Ethics Board; informed consent was obtained from all subjects.