Custom scoring based on ecological topology of gut microbiota associated with cancer immunotherapy outcome.

Over the past decade, accumulating evidence pointed to the clinical impact of theintestinal microbiota on immunotherapeutic outcomes across various cancers. Althoughspecific gut microbial species have been associated with beneficial responses in meta-analyses, no consensus exists on a gut microbiome signature for personalized predictionof immunoresistance in clinical practice. Based on fecal shotgun metagenomics (MG)sequencing data from a retrospective discovery cohort of 245 patients with advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICI), weconstructed species-level co-abundance networks that were clustered into speciesinteracting groups (SIG) correlating with overall survival. Thirty-seven and forty-fivemetagenomic species (MGS) belonged to clusters associated with resistance (SIG1) andresponse (SIG2) to ICI, respectively. For each patient, we calculated the SIG1-SIG2 ratioand integrated the trichotomic relative abundance of Akkermansia muciniphila (Akk)when this ratio was poorly discriminant, to obtain a final TOPOSCORE, associatedwith clinical outcome in a validation cohort of 254 NSCLC patients and in a third cohortof 216 renal and urothelial carcinoma amenable to ICI. The TOPOSCORE outperformedPD-L1 tumor scoring. Finally, the 83 MGS-based TOPOSCORE was translated into a 21bacterial probe set qPCR-based scoring that was validated in a prospective cohort ofNSCLC. Custom scoring of intestinal bacteria by the TOPOSCORE could represent adynamic diagnosis tool of gut dysbiosis, that stratifies patients for microbiota-relatedinterventions.