Association between serum lysyl oxidase-like 2 and chitinase-3-like protein 1 levels and Mycobacterium tuberculosis drug resistance in children with pulmonary tuberculosis

Abstract Objective To investigate whether serum lysyl oxidase-like 2 (LOXL2) and chitinase-3-like protein 1 (YKL-40) levels are associated with drug-resistant Mycobacterium tuberculosis in children with pulmonary tuberculosis. Methods In this prospective observational study, 235 pediatric pulmonary tuberculosis patients treated at Hebei Provincial Chest Hospital from February 2021 to February 2025 were enrolled. Based on drug-susceptibility testing, participants were classified into a drug-resistant group (n = 63) and a drug-susceptible group (n = 172). Serum LOXL2 and YKL-40 concentrations were measured and correlated with resistance status. Multivariable logistic regression was used to identify independent risk factors for resistance. Receiver operating characteristic (ROC) curves assessed predictive performance, and a nomogram was constructed for further validation. Results Serum LOXL2 and YKL-40 levels were significantly higher in the drug-resistant group than in the drug-susceptible group (P < 0.05). Spearman correlation revealed a positive association between the two biomarkers (r = 0.433, P < 0.001). Multivariable analysis identified prior tuberculosis contact, elevated C-reactive protein, and increased LOXL2 and YKL-40 levels as independent predictors of drug resistance (P < 0.05). ROC analysis showed that combined LOXL2 and YKL-40 testing had good discriminatory ability (AUC = 0.912). The nomogram achieved AUCs of 0.945 in the training cohort and 0.802 in the validation cohort, with satisfactory calibration and Hosmer-Lemeshow goodness-of-fit. Conclusions In children diagnosed with tuberculosis, elevated serum levels of LOXL2 and YKL − 40 are significantly associated with drug resistance. These biomarkers may serve as potential predictors of drug resistance in children with confirmed tuberculosis, and their combined measurement can enhance the accuracy of risk assessment.