The human gut microbiome is a diverse, dynamic, and complex ecosystem that modulates host processes including metabolism, inflammation, and cellular and humoral immune responses. Recent studies have suggested that the microbiome may also influence the development of certain cancers such as colorectal cancer, and equally importantly, tumor response to systemic therapy, especially immunotherapy. Multiple groups are exploring the therapeutic utility of the microbiome to enhance clinical response through the use of defined oral therapeutics comprising living commensal bacteria, which would represent a new therapeutic modality. Exploiting the microbiome for therapeutic benefit is not without its challenges due to the heterogeneity of the gut microbiota across healthy donors and patients. In addition, many aspects of conventional small molecule and biologics drug discovery and development do not apply to this novel class of living drugs.

We present an approach that leverages the concept of “reverse translation,” using genomic and immunologic characterization of patient samples from interventional studies to define and better understand the organisms and mechanisms that contribute to response or non-response to immunotherapy.

We are investigating the relationship between the composition of the gut microbiome prior to therapy and the antitumor response in patients receiving checkpoint inhibitors (CPI), as well as how CPI treatment modulates the microbiome in both responders and nonresponders. Fecal and blood samples are collected before and during therapy from cancer patients who receive approved CPI; tumor types include renal, bladder, and NSCLC. Whole metagenomic shotgun sequencing of patient microbiomes is used to identify higher order (e.g., order- and family-level) “microbial signatures” that associate with response to CPI treatment. We then utilize proprietary algorithms that enable species- and strain-level resolution of microbial signatures. In addition, global and targeted metabolomics are used to identify functional pathways associated with outcome, and these pathways can be linked to species and strains identified by genomic analysis. Our discovery strategy iterates computational analyses and machine learning approaches with empirical in vitro and ex vivo screening of strains and consortia to inform selection and drive drug design. Data from such a comprehensive approach is invaluable for designing compositions of bacteria that form “functional ecological networks” that can impact response to CPI therapy. Finally, our microbial library of >14,000 isolates from healthy human subjects captures the phylogenetic diversity and functional breadth of the gastrointestinal microbiome, and provides a robust platform to build unique compositions. Such compositions, when tested in syngeneic tumor models in germ-free mice, can provide a preliminary readout of the contributions of members of the consortia and enable candidate identification.

We present examples of reverse translation in patients with recurrent Clostridium difficile infection and ulcerative colitis, a form of inflammatory bowel disease, that have led to the translation of three drugs that are currently in clinical trials. This roadmap provides insight into how similar drugs can be discovered and developed in the setting of immunotherapy to augment the efficacy of CPIs by altering the cancer-immune set point.

This abstract is also being presented as Poster A06.

Citation Format: David N. Cook, Jonathan Peled, Marcel van den Brink, Lata Jayaraman. Drugging the human microbiome for combination with tumor immunotherapy [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2017 Oct 1-4; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2018;6(9 Suppl):Abstract nr PR06.