Clinical evaluation of metagenomic next-generation sequencing in unbiased pathogen diagnosis of urinary tract infection

Background Early availability of pathogen identification in urinary tract infections (UTIs) has critical importance in disease management. Metagenomic next-generation sequencing (mNGS) has the potential to transform how acute and serious infections are diagnosed by offering unbiased and culture-free pathogen detection. However, clinical experience with application of the mNGS test is relatively limited. Methods We therefore established a MinION-based mNGS pathogens diagnostic platform and evaluated its potential for clinical implementation in UTIs with clinical samples. 213 urine samples from patients with suspected UTIs were included and subjected to mNGS testing using the MinION platform. mNGS results were compared to the gold standard of clinical culture and composite standard of combining clinical testing, confirmatory qPCR testing, and clinical adjudication by doctors. Results The mNGS exhibited a sensitivity of 81.4% and a specificity of 92.3%, along with a positive predictive value of 96.6%, a negative predictive value of 64.9%, and an overall accuracy of 84.4%, all of which were determined based on the gold standard of routine culture results. When assessed against the composite standard, the sensitivity and specificity both increased to 89.9% and 100%, respectively, while the accuracy rose to 92.4%. Notably, the positive predictive value and negative predictive value also saw improvements, reaching 100% and 76.8%, respectively. Moreover, this diagnostic platform successfully identified dsDNA viruses. Among the 65 culture-negative samples, the viral detection rate reached 33.8% (22/65) and was subsequently validated through qPCR. Furthermore, the automatic bioinformatics pipeline we developed enabled one-click analysis from data to results, leading to a significant reduction in diagnosis time. Conclusion These results demonstrate that the pathogen detection performance of mNGS is sufficient for diagnostic testing in clinical settings. As the method is generally unbiased, it can improve diagnostic testing of UTIs and other microbial infections.