Gut Bacteria Composition Drives Primary Resistance to Cancer Immunotherapy in Renal Cell Carcinoma Patients

Background: The development of immune checkpoint blockade (ICB) has revolutionized the clinical outcome of renal cell carcinoma (RCC). Nevertheless, improvement of durability and prediction of responses remain unmet medical needs. While it has been recognized that antibiotics (ATBs) decrease the clinical activity of ICB across various malignancies, little is known about the direct impact of distinct intestinal nonpathogenic bacteria (commensals) on therapeutic outcomes of ICB in RCC. Objective: To evaluate the predictive value of stool bacteria composition for ICB efficacy in a cohort of advanced RCC patients. Design, setting, and participants: We prospectively collected fecal samples from 69 advanced RCC patients treated with nivolumab and enrolled in the GETUG-AFU 26 NIVOREN microbiota translational substudy phase 2 trial (NCT03013335) at Gustave Roussy. We recorded patient characteristics including ATB use, prior systemic therapies, and response criteria. We analyzed 2994 samples of feces from healthy volunteers (HVs). In parallel, preclinical studies performed in RCC-bearing mice that received fecal transplant (FMT) from RCC patients resistant to ICB (NR-FMT) allowed us to draw a cause-effect relationship between gut bacteria composition and clinical outcomes for ICB. The influence of tyrosine kinase inhibitors (TKIs) taken before starting nivolumab on the microbiota composition has also been assessed. Outcome measurements and statistical analysis: Metagenomic data (MG) from whole genome sequencing (WGS) were analyzed by multivariate and pairwise comparisons/fold ratio to identify bacterial fingerprints related to ATB or prior TKI exposure and patients’ therapeutic response (overall response and progression-free survival), and compared with the data from cancer-free donors. Results and limitations: Recent ATB use (n = 11; 16%) reduced objective response rates (from 28% to 9%, p < 0.03) and markedly affected the composition of the microbiota, facilitating the dominance of distinct species such as Clostridium hathewayi, which were also preferentially over-represented in stools from RCC patients compared with HVs. Importantly, TKIs taken prior to nivolumab had implications in shifting the microbiota composition. To establish a cause-effect relationship between gut bacteria composition and ICB efficacy, NR-FMT mice were successfully compensated with either FMT from responding RCC patients or beneficial commensals identified by WGS-MG (Akkermansia muciniphila and Bacteroides salyersiae). Conclusions: The composition of the microbiota is influenced by TKIs and ATBs, and impacts the success of immunotherapy. Future studies will help sharpen the role of these specific bacteria and their potential as new biomarkers. Patient summary: We used quantitative shotgun DNA sequencing of fecal microbes as well as preclinical models of fecal or bacterial transfer to study the association between stool composition and (pre)clinical outcome to immune checkpoint blockade. Novel insights into the pathophysiological relevance of intestinal dysbiosis in the prognosis of kidney cancer may lead to innovative therapeutic solutions, such as supplementation with probiotics to prevent primary resistance to therapy. Antibiotics prior to immune checkpoint inhibitors have a deleterious clinical impact, reduce the microbiome diversity, and increase Clostridium hathewayi bacteria associated with resistance. Higher baseline microbiome diversity and Akkermansia muciniphila are associated with longer progression-free survival. In murine fecal microbiome transplantation experiments, A. muciniphila can restore the anticancer activity of the combination of anti–PD-1 and CTLA-4.