Abiraterone acetate preferentially enriches for the gut commensal *Akkermansia muciniphila* in castrate-resistant prostate cancer patients

The human microbiome significantly influences cancer development, progression, and treatment response. *Akkermansia muciniphila* shows promise in improving responses to immunotherapy, and its inverse correlation with inflammation and metabolic disorders is well-documented. However, its role in prostate cancer (PC), a prevalent and costly malignancy, remains largely unexplored. PC progression relies on androgen receptor activation. Androgen deprivation therapy (ADT) is a cornerstone treatment, but castrate-resistant PC often develops, requiring further therapies like abiraterone acetate (AA). AA's poor absorption leads to high fecal excretion, exposing the gut microbiota to significant concentrations. This study investigates the influence of ADT and AA on the gut microbiome in PC patients, hypothesizing that androgen depletion impacts microbial homeostasis and that AA's metabolism by the gut microbiota may influence treatment outcomes.

Several studies have highlighted the role of the gut microbiome in cancer treatment efficacy. *Akkermansia muciniphila*, in particular, has been linked to positive responses to anti-PD-1 immunotherapy in various cancers. Its association with improved metabolic health and reduced inflammation has also been demonstrated. However, limited research exists on the interplay between the gut microbiota and PC treatments. The influence of serum sex steroid hormone levels on the composition of the gastrointestinal microbiota is increasingly recognized. Given the known impact of ADT on androgen levels, and the observation that a significant portion of AA is excreted in feces, suggesting metabolism within the gut, we set out to explore the direct influence of PC treatments on the gut microbiota and its possible impact on treatment response.

This study employed 16S rRNA gene sequencing to analyze fecal samples from 68 PC patients, categorized into three groups: control (no treatment), ADT alone, and ADT plus AA. Patient metadata, including age, BMI, antibiotic use, and other relevant factors, were also collected to account for potential confounders. Statistical analyses were performed to identify differentially abundant taxa and pathways between the treatment groups. To isolate the effect of AA, in vitro experiments were conducted, incubating fecal samples with AA or vehicle and analyzing microbiota changes. A simulated host-free gut model using a bioreactor was used to further elucidate the interaction between AA and the gut microbiota, independent of host factors. qPCR was used to validate 16S rRNA sequencing results. Pure culture experiments investigated bacterial metabolism of AA and its effect on bacterial growth. Functional inferencing using PICRUSt2 analyzed the predicted metabolic pathways and changes in vitamin K2 biosynthesis.

The study revealed significant differences in gut microbiota composition between the three patient groups. ADT and ADT + AA treatments both led to a decrease in androgen-utilizing *Corynebacterium* spp. Critically, oral AA treatment uniquely led to a significant enrichment of *A. muciniphila*. In vitro experiments using fecal samples showed that AA's impact on gut microbiota diversity was dependent on the baseline abundance of *A. muciniphila*. The host-free gut model confirmed AA's ability to selectively enrich *A. muciniphila* independent of host factors. Furthermore, the study demonstrated that while several bacterial strains metabolized AA, only *A. muciniphila* showed enhanced growth in the presence of AA, primarily due to acetate. Functional inferencing indicated that AA exposure, both in vitro and in vivo, correlated with an increase in bacterial biosynthesis of vitamin K2, a potential anticancer agent. The study also showed a homolog-specific increase in vitamin K2 biosynthesis pathways; in *A. muciniphila* for MK7, MK8, MK11, MK12, MK13, and in Enterobacteriaceae for MK6 and MK9. A potential mechanism involving cross-feeding between *A. muciniphila* and other bacteria was suggested, where *A. muciniphila*'s mucin degradation and acetate utilization support the growth of other bacteria, which in turn provide complementary quinones necessary for *A. muciniphila* growth.

These findings indicate that AA's impact on the gut microbiota is not solely attributed to host-mediated androgen depletion but also involves a direct interaction with specific bacterial species. The enrichment of *A. muciniphila* and the concomitant increase in vitamin K2 biosynthesis suggest a potential auxiliary mechanism contributing to AA's anticancer effects. The study's use of both in vitro and in vivo models helps to disentangle the complex interplay between the drug, the host, and the microbiota. The findings support the potential of targeting the gut microbiota to enhance cancer treatment efficacy. The homolog-specific increase in vitamin K2 biosynthesis pathways and a potential mutualistic relationship between *A. muciniphila* and other bacteria underscore the complexity of these interactions.

This study demonstrates that abiraterone acetate's effect on the gut microbiome is driven by a direct interaction with *A. muciniphila*, leading to its enrichment and increased vitamin K2 production. This suggests a potential novel mechanism for AA's anticancer activity. Further research should focus on the mechanistic details of the interaction between AA and *A. muciniphila*, the precise role of vitamin K2 in AA's efficacy, and the potential of targeted microbiota modulation to improve prostate cancer treatment.

The study's observational design does not establish causality. The simulated gut model, although valuable, is not a perfect representation of the human gut. Further research is necessary to confirm the clinical relevance of the findings and to explore the broader implications for other cancer types and treatments. The use of inferential methods for metabolic pathway prediction also presents limitations.