Table 2_Analytical validation of a metagenomic next-generation diagnostic platform for urinary tract infection in a Thai tertiary hospital setting: a BI-Biotia UTI cohort study.pdf

BackgroundThe BIOTIA-DX platform (BDX), a commercially available clinical-grade mNGS-based test in the United States, has not been analytically validated for urinary tract infections (UTIs) in a Southeast Asian cohort, where microbial epidemiology and antimicrobial resistance (AMR) patterns differ significantly.
ObjectiveOur primary objective was to evaluate the analytical performance and concordance with standard urine culture of the BIOTIA-DX platform in a Thai tertiary hospital setting, thereby assessing its transportability to a Southeast Asian population with distinct microbial epidemiology.
MethodsWe analyzed 398 retrospectively collected urine samples from patients with suspected UTI at a private hospital in Bangkok. Each sample was processed in parallel using standard-of-care urine culture and the BDX mNGS workflow. After excluding 30 samples with insufficient sequencing reads (<500 non-human reads), 368 samples (231 culture-positive, 137 culture-negative) were included. Diagnostic accuracy was assessed against culture, and genotypic AMR predictions were compared to phenotypic antimicrobial susceptibility testing (AST) N = 192.
ResultsThe BIOTIA-DX platform demonstrated high analytical sensitivity at the sample level (98.7% [95% CI: 0.95-0.99]; 228/231 culture-positive samples detected) and organism level (94.6%; 229/242 culture-identified organisms correctly detected). Among 137 culture-negative samples, BIOTIA-DX detected microbial DNA in 98 samples (71.5%), identifying 264 organisms not detected by standard culture. These additional detections predominantly comprised anaerobic organisms (150/264, 56.8%) and fastidious species (54/264, 20.5%); however, the clinical significance of these detections (infection vs. colonization vs. contamination) could not be determined without clinical correlation. For AMR prediction, genotype-phenotype concordance rates were 94.1% for fluoroquinolone resistance in E. coli (96/102 resistant isolates correctly predicted), 91.4% for beta-lactams (106/116), 91.3% for aminoglycosides (21/23), and 81.5% for sulfamethoxazole/trimethoprim (75/92). Specificity and positive predictive value could not be calculated because organisms detected by BIOTIA-DX but not by culture could not be definitively classified as true positives or false positives without independent confirmation.
ConclusionsThe BIOTIA-DX platform demonstrates robust analytical concordance with urine culture in a Thai patient population. Prospective clinical validation studies are needed to assess clinical utility and impact on patient outcomes, particularly in culture-negative and polymicrobial cases. This study represents the first analytical validation of this platform using Oxford Nanopore Technology and the first validation in Southeast Asia.