Investigating The Role Of The Gastrointestinal Microbiome In Response To Immune Checkpoint Inhibitors (Icis) Among Patients (Pts) With Metastatic Renal Cell Carcinoma (Mrcc)

730 Background: ICIs are effective in mRCC, but one pertinent clinical need is to identify predictive biomarkers for response. The PD-1 receptor has been implicated in regulating gastrointestinal commensal bacteria, with varied immune interactions, thereby impacting response to ICIs. We evaluated bacterial taxa and ICI outcomes in mRCC pts. Methods: Fecal samples from 22 mRCC pts were collected at baseline, week (wk)-4 on ICI, and upon disease progression. Pts were grouped as responders (R, complete or partial response) or non-responders (NR, stable or progressive disease). Microbial DNA was isolated by next generation DNA sequencing. The V4 region of bacterial 16S ribosomal RNA was amplified from extracted DNA and analyzed for bacterial abundance, as well as alpha diversity indices (number of amplicon sequence variants [ASVs], Shannon’s Index, Faith’s Phylogenetic Diversity, and Pielou’s evenness) and beta diversity indices on ASVs (Bray-Curtis, Jaccard, and unweighted/weighted UniFrac dissimilarity measures). Results: Beta diversity analysis at baseline showed no difference in microbial composition between Rs and NRs. However, beta diversity analysis did show a significant change in composition from baseline to wk 4 in R vs NR pts (Bray Curtis p-value=0.03). Among mRCC pts with CR to ICIs, counts of bacteria in the phylum Verrucomicrobia had an upward trend from baseline to wk 4. All mRCC pts with CR (n=3) had Akkermansia at wk 4. However, Akkermansia colonization was not sufficient for response, present in 7/9 Rs and 6/11 NRs. Conclusions: Baseline microbiome differences between ICI Rs and NRs are not enough to predict outcomes. Diversity changes between baseline and wk-4 on treatment could be an early predictor of response. Factors other than presence of Akkermansia (tumor or host-specific, Akkermansia strain variation, or other bacteria in the microenvironment) may contribute to response. Further species and strain-level profiling of the microbiota, tumor-specific genomic alterations, host immune response, and increasing sample size of ICI-treated patients may improve detection of significant differences between Rs and NRs.