Quantitative PCR Bacterial Levels Cannot Distinguish Recurrent UTI from Other Non-infectious Urinary Symptoms

Quantitative polymerase chain reaction (qPCR) has high sensitivity in detecting uropathogens. While qPCR is increasingly used in urinary tract infection (UTI) diagnosis, its accuracy distinguishing UTI from non-infectious lower urinary tract symptoms (LUTS) is unclear. This study sought to assess qPCR utility in distinguishing UTI from non-infectious LUTS and controls.

qPCR examined urinary DNA from 502 women with UTI (UTI, n=104), non-infectious LUTS (n=354), or controls (n=44) for uropathogens (Escherichia coli, Klebsiella pneumoniae, Enterococcus), Lactobacilli, BV-associated bacteria (Gardnerella, Prevotella, Burkholderia), and Candida albicans. Relative bacterial DNA quantities were compared across groups and menopausal strata by logistic regression. Receiver operating characteristic analysis of continuous qPCR values assessed the ability of individual and combined microbial species to distinguish UTI from LUTS and controls, with AUC ≥0.7 as the discrimination threshold.

Almost all women had detectable urinary E. coli (97.3%), Enterococcus (99.2%), K. pneumoniae (73.4%), and C. albicans (94.2%) DNA. No qPCR detection threshold could distinguish symptomatic UTI subjects from controls or non-infectious LUTS. Decreased L. crispatus, L. iners, and Burkholderia were seen in UTI subjects. Postmenopausal women had higher E. coli (p=0.023) and Enterococcus (p=0.002) levels and lower Lactobacillus (p<0.001) than premenopausal women. In premenopausal women, vaginal species alone and in combination with uropathogens modestly discriminated UTI from LUTS (AUC > 0.7), although no model achieved discrimination in postmenopausal women.

Women with UTI, non-infectious LUTS, and controls exhibit similar uropathogen levels by qPCR. Providers should exercise caution using qPCR-based tests in women without a high suspicion for UTI.