Clinical evaluation of bacterial DNA using an improved droplet digital PCR for spontaneous bacterial peritonitis diagnosis

OBJECTIVE: Bacterial DNA (bactDNA) detection can be used to quickly identify pathogenic bacteria and has been studied on ascitic fluid. We aimed to retrospectively analyze the diagnostic value and applicational prospect of the bactDNA load in spontaneous bacterial peritonitis (SBP). METHOD: We extracted viable bactDNA from ascitic samples of 250 patients with decompensated cirrhosis collected from October 2019 to April 2021 and detected the bactDNA by droplet digital polymerase chain reaction (ddPCR). We used ascitic samples of a baseline cohort of 191 patients to establish diagnostic thresholds for SBP and analyze the patients’ diagnostic performance based on ascites polymorphonuclear (PMN) and clinical manifestation. We performed bactDNA quantification analysis on 13 patients with a PMN less than 250 cells/mm(3) but with clinical symptoms. The dynamic changes of the bactDNA load from eight patients (before, during, and after SBP) were analyzed. RESULTS: After the removal of free DNA, the bactDNA detected by ddPCR was generally decreased (1.75 vs. 1.5 log copies/µl, P < 0.001). Compared with the traditional culture and PMN count in the SBP diagnosis, the bactDNA showed that the ddPCR sensitivity was 80.5%, specificity was 95.3%, positive predictive value was 82.5%, and negative predictive value was 94.7%, based on clinical composite criteria. In patients with a PMN less than 250 cells/mm(3), the bactDNA load of 13 patients with symptoms was significantly higher than those without symptoms (2.7 vs. 1.7 log copies/µl, P < 0.001). The bactDNA in eight patients had SBP that decreased by 1.6 log copies/µl after 48 h of antibiotic treatment and by 1.0 log copies/µl after 3 days of continued treatment. CONCLUSION: BactDNA detection can be used to further enhance the diagnostic efficiency of SBP. Therefore, the application of ddPCR assay not only can be used to discriminate and quantify bacteria but also can be used in the clinical assessment for antibiotics treatment.