Differentiating Nontuberculous Mycobacterial Lung Disease from Pulmonary Tuberculosis: A Retrospective Study on Clinical Features and a Diagnostic Model.

Background

Nontuberculous mycobacterial pulmonary disease (NTM-PD) and pulmonary tuberculosis (PTB) present with similar clinical and radiological features, frequently leading to misdiagnosis. This study aimed to compare NTM-PD with extrapulmonary NTM disease (ENTM) and PTB and to develop a predictive model for early differentiation.

Methods

We retrospectively analyzed 473 NTM patients (340 NTM-PD, 112 ENTM, 21 disseminated) and 218 PTB patients treated between January 2008 and December 2024. A diagnostic model was developed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm and multivariate logistic regression, with internal validation via 1000 bootstrap resamples and a pre-split validation cohort. Performance was evaluated by the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis.

Results

Compared to ENTM, NTM-PD patients were older and had lower BMI, with bronchiectasis (OR = 24.86) and prior PTB history (OR = 7.58) as the strongest risk factors. Relative to PTB, NTM-PD patients were more often female and older, with more bronchiectasis and mucus plugs on imaging, while PTB showed more pleural thickening and effusion. NTM isolates exhibited high resistance to first-line anti-TB drugs but maintained susceptibility to macrolides and amikacin. The nine-predictor nomogram achieved AUCs of 0.758 (training) and 0.731 (validation), with calibration and decision curve analysis confirming clinical utility.

Conclusions

NTM-PD and PTB exhibit distinct clinical, radiological, and microbiological profiles. The nomogram offers a useful preliminary screening tool for early differentiation, though external validation in multicenter studies is warranted.