Association of gut microbiome signatures with iron metabolism response to roxadustat in peritoneal dialysis patients with anemia: a cross-sectional study.

Background: Anemia in chronic kidney disease (CKD) is associated with increased cardiovascular risk, impaired quality of life, and reduced survival. Roxadustat, a hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI), has demonstrated non-inferior efficacy to erythropoiesis-stimulating agents (ESAs) for anemia correction in CKD. Gut microbiota modulate the intestinal HIF-iron metabolism axis, thereby regulating intestinal iron absorption. This cross-sectional study investigated the association between gut microbiome composition and iron-metabolism response to roxadustat, and developed a logistic regression model to identify factors associated with iron metabolism non-response in anemic patients undergoing peritoneal dialysis (PD). Methods: Demographic and clinical data were collected at study enrollment, and fecal samples underwent 16S rRNA gene sequencing. Microbial taxa associated with the iron-metabolism response to roxadustat were identified using linear discriminant analysis effect size (LEfSe), differential abundance analysis with DESeq2, and Spearman's rank correlation analysis. Key microbial features were further selected using random forest analysis. Multivariable logistic regression models were constructed using R software (version 4.2.3). Variable selection was performed through stepwise selection based on the Akaike information criterion. Model performance was evaluated with the area under the receiver operating characteristic curve, calibration curves, and decision curve analysis. Internal validation was performed using 10-fold cross-validation and bootstrapping with 1,000 iterations. Results: The overall iron-metabolism response rate to roxadustat was 39% among the enrolled participants. Random forest analysis identified microbial features associated with the iron-metabolism response to roxadustat. The final multivariable model included Clostridium perfringens, Propionibacterium acnes, Bacilli, Paraeggerthella hongkongensis, compound α-ketoacid tablets, and antihypertensive medication. The final model achieved an AUC exceeding 0.8, with favorable calibration and clinical utility, and showed good discriminative performance for iron metabolism non-response to roxadustat in PD patients. Conclusion: Distinct gut microbiome signatures are associated with the iron metabolism response to roxadustat in anemic PD patients.