Abstract: 【Abstract】Objective To assess the performance of metagenomic second-generation sequencing(mNGS) in early detection of bacterial infection in patients with peritoneal dialysis-related peritonitis and to investigate whether the integration of cytokines and mNGS assay could improve diagnostic accuracy. To construct and evaluate the etiological value of a risk prediction model to differentiate peritoneal dialysis-associated peritonitis (PDAP) caused by different types of bacteria. Methods PDAP patients admitted to the First Affiliated Hospital of Anhui Medical University from 2020 to 2023 were included in this study, and randomly divided into groups based on the results of mNGS and conventional microbiology testing (CMT). The clinical data were simultaneously collected. PD fluids were sent for conventional microbiological test (CMT), mNGS, and cytokine measurement in parallel. Independent predictors were selected by univariate and multivariate Logistics regression analysis, and prognostic nomogram was constructed for G- and G+ related . The accurate and discriminative abilities of the nomogram were evaluated by Objective  To assess metagenomic second-generation sequencing(mNGS) in the early detection of bacterial infection of peritoneal dialysis-related peritonitis(PDAP), to investigate whether mNGS combined with cytokines could improve the diagnostic accuracy, and to construct a risk prediction model to predict PDAP caused by different types of bacteria and its unfavorable outcome.  Methods   PDAP patients admitted to the First Affiliated Hospital of Anhui Medical University from 2020 to 2023 were included in this study. They were randomly divided into groups based on the results of mNGS and conventional microbiology testing (CMT). Their clinical data were collected, and peritoneal dialysis fluids were tested for CMT, mNGS, and cytokine measurement. Univariate and multivariate logistics regression analyses were used to find out the independent predictors, and a nomogram was constructed for the prediction of G- and G+ bacterial related prognosis. Efficacy of the nomogram was evaluated by receiver operating characteristic curve (ROC) and clinical impact curve (CIC).  Results  A total of 50 PDAP patients were included in this study. Peritoneal fluids were tested for pathogens by mNGS in 41 patients, showing higher positive detection rate (82% vs. 76%, P=0.440) and less detection time (23.0h vs. 29.4h, P＜0.001) than CMT. A model to predict the PDAP by different types of bacteria and a nomogram were constructed,  from which we found that IL-10 and IL-6 in peritoneal dialysis fluid, serum procalcitonin and serum C-reactive protein were the most important biomarkers to predict the prognosis of PDAP by different types of bacteria.  Conclusions   mNGS demonstrates its powerful potentials in the early and rapid detection of pathogens of PDAP. Comprehensive use of IL6 and IL10 in peritoneal dialysis fluid, serum procalcitonin, and serum C-reactive protein will increase the accuracy in the interpretation of mNGS results. A risk prediction model constructed based on cytokines can effectively distinguish G- and G+ bacterial PDAP. 

Key words: Peritoneal dialysis, Peritoneal dialysis related peritonitis, Metagenomic second-generation sequencing, Diagnostic biomarker, Clinical prediction model