Introduction: Clinicians often extend antibiotic therapy for refractory nontuberculous mycobacterial pulmonary disease (NTM-PD), despite potential side effects and the uncertain benefits of extended therapy. In this study, we present our real-world experience with prolonged antibiotic therapy in patients with treatment-refractory NTM-PD. Methods: We reviewed adult patients with treatment-refractory NTM-PD treated at a tertiary referral center in South Korea between April 2003 and July 2024. Eligible patients did not achieve sustained culture conversion within 6 months and received macrolide-based therapy for at least 18 months. Patients were grouped into prolonged treatment (≥24 months) and shorter treatment (18–24 months) groups. The microbiological, radiographic, and clinical outcomes were analyzed. Results: Among the 123 patients, 50 (40.7%) received prolonged therapy. Patients who underwent prolonged treatment had a more frequent history of tuberculosis (44.0% vs. 24.7%, P=0.040) and autoimmune diseases (14.0% vs. 1.4%, P=0.016), although the other baseline characteristics were similar. The use of clofazimine, intravenous aminoglycosides, and inhaled amikacin was significantly higher in the prolonged group, along with a higher incidence of anemia as a side effect (14.0% vs. 2.7%, P=0.030). Acquired macrolide resistance appeared to be more common in the prolonged group (14.0% vs. 6.8%), but the difference was not statistically significant (P=0.315). Over a median follow-up of 4.4 years, prolonged treatment was not associated with a lower risk of retreatment or all-cause mortality. The rates of microbiological cure, sputum smear conversion, radiographic progression, and weight change did not differ between groups. Conclusion: Extending antibiotic therapy beyond 24 months in patients with treatment-refractory NTM-PD does not necessarily confer additional microbiological, radiographic, or survival benefits, and may increase the risk of adverse events. Therefore, prolonged therapy should be reserved for selected patients, with close monitoring for side effects and macrolide resistance.
Kim et al. (Tue,) studied this question.