Response to immune checkpoint inhibition is associated with the gut microbiome in advanced KRAS-mutated non-small cell lung cancer

Background KRAS-mutated non-small cell lung cancer (NSCLC) is associated with a poor prognosis to standard therapies. Despite advances of immune checkpoint inhibitors (ICIs), not all patients show durable responses. In this study, we aim to identify associations between ICI-response and the gut microbiome in patients with KRAS-mutated NSCLC. Methods We performed shotgun metagenomic sequencing of stool samples collected before ICI initiation from 33 patients with KRAS-mutated NSCLC. Microbiome composition within (α-diversity) and between samples (β-diversity) was calculated using Shannon diversity index and principal component analysis on Aitchison distances, respectively. A Bayesian logistic-normal regression model ( Pibble ) was implemented to identify associations between gut microbial features and disease control rate (DCR), progression free survival at 12 months (PFS12) and immune related adverse events (irAEs), adjusting for ICI-regimen, metastatic disease stage, age, sex and BMI. Results Responders were enriched with several saccharolytic species, including Agathobaculum butyriciproducens, Fusicatenibacter saccharivorans, Bifidobacterium longum and Eubacterium ramulus . Non-responders harbored higher abundances of several Bacteroides and Blautia species. Patients unaffected by irAEs demonstrated higher abundances of biotin and butyrate synthesis pathways. Development of irAEs was associated with higher Alistipes finegoldii, Bifidobacterium longum and Bacteroides uniformis abundance. No differences were observed between responders and non-responders in Shannon diversity index ( P =0.69) and overall microbial composition ( P =0.82). Conclusions We show gut microbial species and pathways that are differentially abundant between responders and non-responders to ICI in the setting of KRAS-mutated NSCLC. We find overlap with microbial signatures of response to ICI in other tumor types, potentially reflecting tumor-independent microbial mechanisms.