Intestinal microbiota predict response and toxicities during anti-PD-1/anti-CTLA-4 immunotherapy

Abstract

Immunotherapies targeting PD-1/PD-L1 and CTLA-4 have revolutionised the treatment of malignant melanoma. Whilst combining strategies is associated with improved response rates, this is accompanied by increased incidence of severe immune related adverse events (irAEs). Given the ability of the gut microbiota to modulate both local and systemic immunity, this study aimed to examine the association of the gut microbiome with the subsequent efficacy and development of irAEs during combination immunotherapy in the neoadjuvant setting. Pre-treatment faecal microbiomes of stage III melanoma patients (n=38) were analysed using 16S sequencing. Low microbial diversity and a reduction in the abundance of butyrate-producing Ruminococcaceae and methanogenic-archaea were associated with lack of response and the development of severe irAEs. Machine learning applied to the data was able to predict patients who would develop severe irAEs in the absence of tumour efficacy with 87% accuracy. Mass cytometry of matched pre-treatment PBMCs indicated that differences in peripheral immune cells were associated with changes in microbial diversity. Together, the data suggests that pre-treatment microbiomes influence systemic immunity and can be used to predict immunotherapeutic outcomes, and the maintenance of a robust microbial ecosystem that supports barrier function is key to developing successful microbial interventions to improve patient outcomes.