Compositional differences in gastrointestinal microbiota in prostate cancer patients treated with androgen axis-targeted therapies

Abstract

Background: It is well known that the gastrointestinal (GI) microbiota can influence the metabolism, pharmacokinetics, and toxicity of cancer therapies. Conversely, the effect of cancer treatments on the composition of the GI microbiota is poorly understood. We hypothesized that oral androgen receptor axis-targeted therapies (ATT), including bicalutamide, enzalutamide, and abiraterone acetate, may be associated with compositional differences in the GI microbiota.

Methods: We profiled the fecal microbiota in a cross-sectional study of 30 patients that included healthy male volunteers and men with different clinical states of prostate cancer (i.e., localized, biochemically recurrent, and metastatic disease) using 16S rDNA amplicon sequencing. Functional inference of identified taxa was performed using PICRUSt.

Results: We report a significant difference in alpha diversity in GI microbiota among men with versus without a prostate cancer diagnosis. Further analysis identified significant compositional differences in the GI microbiota of men taking ATT, including a greater abundance of species previously linked to response to anti-PD-1 immunotherapy such as Akkermansia muciniphila and Ruminococcaceae spp. In functional analyses, we found an enriched representation of bacterial gene pathways involved in steroid biosynthesis and steroid hormone biosynthesis in the fecal microbiota of men taking oral ATT.

Conclusions: There are measurable differences in the GI microbiota of men receiving oral ATT. We speculate that oral hormonal therapies for prostate cancer may alter the GI microbiota, influence clinical responses to ATT, and/or potentially modulate the antitumor effects of future therapies including immunotherapy. Given our findings, larger, longitudinal studies are warranted.

Fig. 1

Unsupervised clustering (log-transformed) of 16S rDNA Illumina-sequencing results from fecal samples by genus. The dendrogram was based on hierarchical clustering of the Euclidean distance between samples in the combined groups. L cancer localized prostate cancer, BCR biochemically recurrent prostate cancer, mHSPC metastatic hormone-sensitive prostate cancer, mCRPC metastatic castration-resistant prostate cancer. No cancer = no clinical and/or biopsy proven diagnosis of cancer (control and benign groups)

Fig. 2

Principal coordinates analysis (PCoA) and beta diversity (unweighted UniFrac) of each fecal sample bacterial profile, analyzed by the indicated groups. a–c Principal coordinate axis 2 showed the most distinction between medication groups, so statistical comparisons were limited to this dimension. d Statistical comparison of beta diversity between the indicated groups (Mann–Whitney test). Shown is the mean unweighted UniFrac distance (+SEM)

Fig. 3

Quantitative PCR (qPCR) for Akkermansia muciniphila. a Confirmation of enriched abundance of this species in the GI microbiota of men taking oral ATT versus the other men included in the study. b The qPCR results were in strong correlation (R² = 0.9938) with the results obtained by 16S rDNA Illumina amplicon sequencing