Characterization of the microbial resistome in a prospective trial of CBM588 in metastatic renal cell carcinoma (mRCC) offers mechanism for interplay between antibiotic (abx) use and immune checkpoint inhibitor (ICI) activity.

Abstract

4510 Background: The negative association between ICI response and abx therapy is well defined (Derosa et al Cancer Discov 2021). Paradoxically, a retrospective assessment of the live bacterial product (LBP) CBM588 in patients (pts) with advanced lung cancer showed improved outcome with ICIs when the combination of CBM588 and abx (as compared to CBM588 alone) was employed (Tomita et al Cancer Immunol Res 2020). We postulated that the microbial resistome (genes encoding antimicrobial resistance) could shift in a manner with CBM588 therapy that facilitated ICI response. Methods: Pts with newly diagnosed mRCC with clear cell and/or sarcomatoid histology and intermediate/high risk disease per IMDC criteria were randomized to nivolumab/ipilimumab (nivo/ipi) or nivo/ipi/CBM588 in a 1:2 ratio. Stool samples were collected at baseline and week 12. Whole-metagenome sequencing was performed to analyze stool microbiome composition. Abx resistance genes (RGs) were inferred using publicly available database (McArthur et al. Antimicrob Agents Chemother 2013), and groups of abx RGs for various classes of abx were characterized. Wilcoxon signed-rank test was used for comparison of abx RG abundance between baseline and week 12 in each treatment arm and in responders (R) and non-responders (NR). Results: The study enrolled 30 pts, with the final analysis including 29 eligible pts (median age: 66 years, M:F 21:8, nivo/ipi: 19 pts, nivo/ipi/CBM588:10 pts). Objective response was 20% and 58% in nivo/ipi and nivo/ipi/CBM588 arms, respectively. The overall abundance of abx RGs remained unchanged between baseline and week 12 in pts receiving nivo/ipi alone. In contrast, a decrease in abx RGs was observed in pts receiving nivo/ipi with CBM588 arm from baseline to week 12 (p = 0.042 in Rs; p = 0.078 in NRs). More specifically, nivo/ipi/CBM588 treatment led to a significant reduction in fosfomycin RGs and nitroimidazole (e.g., metronidazole) RGs in both pts with R (p = 0.019 and 0.042, respectively) and NR (p = 0.031 and p = 0.031, respectively). A multitude of other clinically relevant abx RGs were downregulated in pts receiving CBM588, including those mediating resistance to glycopeptide (e.g., vancomycin) and lincosamide (e.g., clindamycin) abx. Conclusions: In the first interrogation of the resistome in mRCC, we demonstrate that CBM588 decreases abx RGs associated with multiple commonly used classes of abx. Abx clear commensals and increase pathogenic (abx resistant) bacteria in the gut. Based on our data, we formulate the hypothesis that combining abx with CBM588 may decrease potentially pathobionts and favor butyrogenic species, thereby improving CPI response. Clinical studies using CBM588 with abx priming may be warranted. Clinical trial information: NCT03829111.