A proof-of-principle study for the point-of-care detection of ESBL (CTX-M) by NG-Test® CTX-M MULTI lateral flow assay in urine samples using a simplified method for use in a resource-limited setting

Abstract

Background The rise of extended-spectrum β-lactamase-producing Enterobacterales (ESBL-E) in low- and middle-income countries limits treatment options, leading to the frequent use of broad-spectrum antibiotics. Reducing time-to-result for a urinary infection can facilitate correct antibiotic treatment and support antimicrobial and diagnostic stewardship measures. This study compared two simplified enrichment methods for detecting CTX-M directly from urine specimens. Methods Two enrichment methods, namely centrifugation of 2 mL urine and filtration of 1 mL urine using the DirecTool adaptor, were compared using 20 culture-positive urine samples (20 suspected ESBL-E and 20 non-ESBL-E). CTX-M production was detected using a lateral flow assay (LFA), NG-Test® CTX-MMULTI. The presence of blaCTX-M genes was confirmed by whole-genome sequencing (WGS). Results The results of both enrichment methods were identical, with a sensitivity of 87.5% and a specificity of 100%. In 19/20 (95%) of the urine samples, the results of the CTX-M LFA were identical with the phenotypic confirmation and WGS. Both methods could detect ESBL-E bacteriuria with ≥104 cfu/mL. All ESBL-E-negative samples were identified accurately. Both enrichment methods yielded negative results in one ESBL-E-positive (CTX-M-15) sample despite phenotypic and genotypic confirmation of ESBL production. High leukocyte count (>500 cells/µL), the presence of boric acid or polymicrobial samples did not appear to impact the performance of both enrichment methods. Conclusions Our study underscores the feasibility of directly detecting CTX-M in urine. Simplified enrichment methods, particularly with a filtration kit, enhance the assay’s practicality, rendering it suitable for use in primary care, emergency departments or remote laboratories without sophisticated equipment.