Gut microbial signature in lung cancer patients highlights specific taxa as predictors for durable clinical benefit

Abstract We aimed to determine microbial signature linked with lung cancer (LC) diagnosis and to define taxa linked with durable clinical benefit (DCB) of advanced LC patients. Stool samples for microbial 16S amplicon sequencing and clinical data were collected from 75 LC patients (50 of which were treated with checkpoint inhibitors) and 31 matched healthy volunteers. We compared LC to healthy controls and patients with DCB to those without. LC patients had lower α-diversity and higher between-subject diversity. Random Forests model to differentiate LC cases from controls ROC-AUC was 0.74. Clostridiales, Lachnospiraceae, and Faecalibacterium prausnitzii taxa abundance was decreased in LC compared to controls. High Akkermansia muciniphila correlated with DCB (HR 4.26, 95% CI 1.98–9.16), not only for the immunotherapy-treated patients. In addition, high Alistipes onderdonkii (HR 3.08, 95% CI 1.34–7.06) and high Ruminococcus (HR 7.76, 95% CI 3.23–18.65) correlated with DCB.Our results support the importance of gut microbiome in LC. We have validated the apparent predictive value of Akkermansia muciniphila, and highlighted Alistipes onderdonkii and Ruminococcus taxa correlation with DCB. Upon additional validations those can be used as biomarkers or as targets for future therapeutic interventions.