Abstract
Early prediction and diagnosis of systemic inflammatory response syndrome (SIRS) following percutaneous nephrolithotomy (PCNL) are critical. This study aimed to investigate differences in clinical characteristics and the renal pelvis urobiome between patients with and without post-PCNL SIRS to identify potential predictive biomarkers. Patients undergoing unilateral PCNL were categorized into SIRS(+) and SIRS(-) groups based on postoperative SIRS status. Renal pelvis urine samples were collected for 2bRAD-M sequencing to profile the urobiome. Clinical data and urobiome composition were compared between the groups. Logistic regression identified preoperative serum albumin-globulin ratio (AGR) as an independent protective factor and operative time as an independent risk factor for post-PCNL SIRS, with an area under the receiver operating characteristic curve (AUC) of 0.76. Diversity analysis revealed distinctive microbial differences between the two groups. Through differential analysis and random forest, we screened six species, including Sphingomonas paucimobilis, Ralstonia sp000620465, Ralstonia pickettii, Pelomonas puraquae, Comamonas tsuruhatensis, and Lawsonella clevelandensis_A, to build the microbial prediction model, which achieved an AUC of 0.81. The combination of microbial data and clinical factors further improved predictive accuracy, achieving an AUC of 0.94. Functional profiling of the urobiome also demonstrated significant intergroup differences. This is the first study to explore renal pelvis urobiome dysbiosis in post-PCNL SIRS. Both clinical and microbial factors showed strong predictive value, with their combination offering the greatest discriminatory power. This research could pave the way for the early prediction of post-PCNL SIRS.IMPORTANCEGiven the significant morbidity associated with postoperative percutaneous nephrolithotomy (PCNL) systemic inflammatory response syndrome (SIRS), early prediction and diagnosis are crucial for preventing severe complications like sepsis, which can lead to multiple organ dysfunction or death. Our study uniquely explores how renal pelvis urobiome dysbiosis contributes to post-PCNL SIRS. By utilizing the novel 2bRAD-M sequencing, the research identifies key microbial species in the renal pelvis and integrates them with clinical factors like albumin-globulin ratio and operative time. The resulting prediction model, with an impressive area under the curve, significantly outperforms traditional clinical models. This offers a more precise approach to stratify patients at high risk of developing SIRS. This work suggests that microbial imbalances may actively drive SIRS, pointing to the potential to revolutionize the predictive strategies for post-PCNL SIRS.