Abstract:

Objective: To evaluate the individual and combined diagnostic/application value of Mycobacterium tuberculosis/rifampicin resistance real-time fluorescent quantitative nucleic acid amplification detection technology (GeneXpert MTB/RIF, referred to as “Xpert”) and BACTEC MGIT 960 liquid culture method(referred to as “culture”)in bronchoalveolar lavage fluid (BALF), as well as tuberculosis-specific interferon-gamma release assay (referred to as “TB-IGRA”) in peripheral blood, for the etiologically negative pulmonary tuberculosis (PTB). Methods: A retrospective analysis was conducted. According to the setting inclusion criteria, we enrolled outpatients and inpatients with suspected PTB and negative sputum-smear and sputum-Xpert results who admitted to Qingdao Chest Hospital between October 2021 and November 2023. All included patients simultaneously received BALF Xpert and culture, and peripheral blood TB-IGRA tests. The examination results were recorded and analyzed. A clinically confirmed diagnosis was as the benchmark standard, 134 clinically confirmed etiologically negative PTB patients were enrolled in the tuberculosis group (TB group) and 92 non-PTB patients were included in the no-TB group. Multivariate logistic regression analysis and the receiver operating characteristic (ROC) curve were used to compare the positive detection rates of the above-mentioned three testing methods, and to evaluate their diagnostic/application efficacy for etiologically negative PTB, by using them individually and in combination. Results: Among 226 patients with suspected PTB, the positive detection rates of Xpert, mycobacterial culture and IGRA were 41.15% (93/226), 34.07% (77/226) and 57.96% (131/226), respectively. The positive detection rate of IGRA was significantly higher than that of Xpert (χ²=24.446, P<0.001) and mycobacterial culture (χ²=45.306, P<0.001). The positive detection rates of combined tests of Xpert+culture, Xpert+IGRA, culture+IGRA, and the three methods were 42.48% (96/226), 61.95% (140/226), 59.73% (135/226) and 61.95% (140/226) respectively. The positive detection rates of the three methods combined and Xpert+IGRA combined were markedly higher than those of single Xpert (χ²=45.021, P<0.001) and single mycobacterial culture (χ²=61.016, P<0.001). Xpert showed superior specificity (100.00% (92/92)) versus IGRA (78.26% (72/92)) and higher accuracy (81.86% (185/226)) versus culture (72.12% (163/226)). IGRA had higher sensitivity (82.84% (111/134)) than both Xpert (69.40% (93/134)) and culture (55.22% (74/134)). The combination of culture+IGRA had the highest sensitivity (84.33% (113/134)). The combined detection of Xpert+IGRA and the triple combination achieved the highest specificity (100.00% (92/92)), accuracy (89.82% (203/226)) and positive detection rate (61.95% (140/226)). The areas under the ROC curves (AUC) were as follows: the triple combination (0.933), Xpert+IGRA (0.928), culture+IGRA (0.890), Xpert+culture (0.867), IGRA (0.866), Xpert (0.863) and culture (0.769). Conclusion: This study validated the applicable scenarios of different diagnostic methods. For instance, IGRA can be used for screening and auxiliary diagnosis of etiologically negative PTB due to its high sensitivity. Xpert can be used for early and rapid diagnosis of etiologically negative PTB due to its high specificity and accuracy. The combination of culture and IGRA yielded the highest sensitivity, thereby can effectively reduce the risk of missed diagnosis in etiologically negative PTB. The combination of Xpert and IGRA achieved higher sensitivity, specificity, accuracy, positive detection rate and AUC values, demonstrating optimal diagnostic efficacy, which contributes to improving clinical diagnostic consistency and is clinically feasible.

Key words: Bronchoalveolar lavage, Mycobacterium tuberculosis, Molecular diagnostic techniques, Diagnosis, differential