Comparison of Computed Tomography Findings of Tuberculosis and Non-Tuberculous Mycobacterial Lung Infections

The treatment of non-tuberculous mycobacterial infections, which are becoming increasingly common, is different and longer than tuberculosis. Correct diagnosis and subtyping are vital for patient management.

To determine the radiological findings of mycobacterial infections, detect differences, and contribute to the most accurate and rapid diagnosis.

This retrospective and single-center study included a total of 264 patients diagnosed with mycobacterial lung infections, comprising 152 cases of tuberculosis and 112 cases of non-tuberculous mycobacterial infections. Patient demographics (age and gender), the specific subtypes of non-tuberculous mycobacteria, and detailed computed tomography findings were systematically evaluated.

Mean age of the patients was 45.84±16.84 (min 14, max 87). More than half of the mycobacteria infections (54.5%) were in the patient group aged 35-65. As the patient's age decreased, the incidence of tuberculosis increased, and as the patient's age increased, the incidence of non-tuberculous mycobacteria infections increased. 85.9% of patients in the <35 age group were infected with tuberculosis, and 64.3% of patients in the >65 age group were infected with non-tuberculous mycobacteria. Male patients were the majority in all mycobacteria infections, but male dominance was especially evident in non-tuberculous mycobacteria infections (p<0.001 for both). Tree-in-bud and pleural effusion were more common in tuberculosis; while cavity, bronchiectasis, and sequela fibrotic changes were more common in non-tuberculous mycobacteria infections (p<0.05). The cavitation wall was thicker in tuberculosis (p=0.015). In addition, the anterior segment of the left upper lobe was more frequently affected in tuberculosis patients (p=0.018).

It is necessary to know the radiological findings well and to guide clinicians in terms of possible non-tuberculous mycobacterial infection, especially in cases resistant to treatment. Accurate differentiation between tuberculosis and non-tuberculous infections is essential, as it has a direct impact on the treatment algorithm. Multi-center studies with larger patient populations may provide additional contributions to our current knowledge.