Development and validation of a predictive model for calcium oxalate kidney stone recurrence integrating gut microbiome and clinical features
摘要
To characterize the gut microbiome and clinical profiles of patients with recurrent calcium oxalate kidney stones, identify risk factors for recurrence, and develop an integrated predictive model.
MethodsThe development and validation of the prediction model followed the reporting standards outlined in the TRIPOD checklist. In this prospective study, patients with a first calcium oxalate stone episode were enrolled and followed for two years. Gut microbiota were profiled using 16 S rDNA sequencing. Independent risk factors were identified by logistic regression, and a nomogram was constructed and validated with receiver operating characteristic curves, calibration plots, and decision curve analysis.
ResultsAmong 268 patients, 75 (27.99%) developed recurrence. The recurrent group showed significantly lower gut microbial alpha diversity. LEfSe analysis revealed enrichment of Proteobacteria, Enterococcaceae, Limosilactobacillus, and Escherichia-Shigella, with reduced Firmicutes. Family history of stones (OR = 10.684), elevated serum creatinine (OR = 1.025), and increased Escherichia-Shigella relative abundance (OR = 1.063) were independent risk factors. The nomogram achieved an AUC of 0.978 in the training cohort and 0.943 in the validation cohort, with excellent calibration and net clinical benefit.
ConclusionRecurrent calcium oxalate stone patients exhibit gut dysbiosis characterized by reduced diversity and Escherichia-Shigella enrichment. Family history, serum creatinine, and Escherichia-Shigella abundance independently predict recurrence. The nomogram integrating these factors provides a reliable tool for recurrence risk assessment.