Exploring bladder cancer through urinary microbiota: innovative "urinetypes" classification and establishment of a diagnostic model.
Bladder cancer (BCa) is a prevalent and lethal malignancy of the urinary system. Recent evidence suggests a strong association between the urinary microbiota and the pathogenesis, progression, and prognosis of BCa. This study investigated the role of the urinary microbiota in BCa, aiming to develop a non-invasive diagnostic model based on microbial biomarkers. Additionally, we proposed a novel urine-based microbiota classification method to enhance diagnostic accuracy and guide treatment strategies.
The study included a discovery cohort (104 BCa patients, 56 with Other Malignant Urological Cancer, 98 with benign urinary diseases, and 42 healthy controls) and a validation cohort (66 BCa patients, 5 with Other Malignant Urological Cancer, 51 with benign urinary diseases, and 22 healthy controls). The urinary microbiota composition was analyzed using 16 S rRNA gene sequencing to assess diversity, identify biomarkers, and construct a diagnostic model for BCa. Finally, clustering analysis was used to establish "Urinetypes".
BCa patients exhibited greater richness and diversity in their urinary microbiota, with significant differences in beta diversity observed across the groups. Genera such as Sphingomonas, Anaerococcus, Acinetobacter, Stenotrophomonas, Aeromonas, and Novosphingobium were more abundant in BCa patients, while Lactobacillus and Gardnerella were less abundant, suggesting their potential as biomarkers. PICRUSt analysis revealed significant enrichment in carbohydrate and nucleotide metabolism in BCa patients, reflecting the increased metabolic demands of cancer cells. A biomarker prediction model employing random forest analysis based on 12 microbial genera achieved high accuracy in the discovery cohort (AUC = 89.08%) and demonstrated robust performance in the validation cohort (AUC = 70.8%). To facilitate potential clinical application, we developed a "Patient Differentiation Index" (PDI), which maintained predictive efficiency in both the discovery cohort (AUC = 86.17%) and the validation cohort (AUC = 78%). Additionally, we identified distinct "Urinetypes", including those dominated by Prevotella and Corynebacterium, which were more prevalent in BCa patients and might represent high-risk subtypes.
This study characterizes the urinary microbiota of BCa patients and, for the first time, provides a reliable non-invasive diagnostic method based on urinary microbiota. The introduction of the innovative concept of "Urinetypes" and the identification of high-risk subtypes associated with BCa offer the potential for improved diagnostic and therapeutic strategies.
This trial was registered on the Chinese Clinical Trial Registry (ChiCTR) with the registration number ChiCTR2300070969, registered on 27 April 2023, https://www.chictr.org.cn/ ChiCTR2300070969. The registration details are publicly accessible on ChiCTR for verification and reference.
The study included a discovery cohort (104 BCa patients, 56 with Other Malignant Urological Cancer, 98 with benign urinary diseases, and 42 healthy controls) and a validation cohort (66 BCa patients, 5 with Other Malignant Urological Cancer, 51 with benign urinary diseases, and 22 healthy controls). The urinary microbiota composition was analyzed using 16 S rRNA gene sequencing to assess diversity, identify biomarkers, and construct a diagnostic model for BCa. Finally, clustering analysis was used to establish "Urinetypes".
BCa patients exhibited greater richness and diversity in their urinary microbiota, with significant differences in beta diversity observed across the groups. Genera such as Sphingomonas, Anaerococcus, Acinetobacter, Stenotrophomonas, Aeromonas, and Novosphingobium were more abundant in BCa patients, while Lactobacillus and Gardnerella were less abundant, suggesting their potential as biomarkers. PICRUSt analysis revealed significant enrichment in carbohydrate and nucleotide metabolism in BCa patients, reflecting the increased metabolic demands of cancer cells. A biomarker prediction model employing random forest analysis based on 12 microbial genera achieved high accuracy in the discovery cohort (AUC = 89.08%) and demonstrated robust performance in the validation cohort (AUC = 70.8%). To facilitate potential clinical application, we developed a "Patient Differentiation Index" (PDI), which maintained predictive efficiency in both the discovery cohort (AUC = 86.17%) and the validation cohort (AUC = 78%). Additionally, we identified distinct "Urinetypes", including those dominated by Prevotella and Corynebacterium, which were more prevalent in BCa patients and might represent high-risk subtypes.
This study characterizes the urinary microbiota of BCa patients and, for the first time, provides a reliable non-invasive diagnostic method based on urinary microbiota. The introduction of the innovative concept of "Urinetypes" and the identification of high-risk subtypes associated with BCa offer the potential for improved diagnostic and therapeutic strategies.
This trial was registered on the Chinese Clinical Trial Registry (ChiCTR) with the registration number ChiCTR2300070969, registered on 27 April 2023, https://www.chictr.org.cn/ ChiCTR2300070969. The registration details are publicly accessible on ChiCTR for verification and reference.
Authors
Sheng Sheng, Liu Liu, Wang Wang, Chen Chen, Xu Xu, Zhang Zhang, Xu Xu, Zhang Zhang, Zhu Zhu, Qin Qin, Zeng Zeng, Zheng Zheng, Zhang Zhang
View on Pubmed