High-Resolution CT Findings as Predictive Factors for Recurrent Nontuberculous Mycobacterial Pulmonary Disease after Successful Treatment

• Main Authors: Hyewon Choi, Min Jae Cha, Yang Soo Kim, Jae Chol Choi
• Format: Article
• Language: English
• Published: MDPI AG 2021-01-01
• Series: Journal of Clinical Medicine
• Subjects: nontuberculous mycobacteria; recurrence; computed tomography
• Online Access: https://www.mdpi.com/2077-0383/10/2/172

Despite long-term treatment for nontuberculous mycobacterial pulmonary disease (NTM-PD), recurrence is common. We aim to identify computed tomography (CT) findings that predict recurrence after successful treatment of NTM-PD. This retrospective study included 44 patients (12 men, 60 ± 11.2 years) successfully treated for NTM-PD between March 2009 and September 2016. Recurrence developed in 18 patients (40.9%) during follow-up (median, 852 days). CT scores for bronchiectasis, bronchiolitis, consolidation, cavities, and nodules at the initiation and termination of treatment were evaluated, then determined association with recurrence. We also assessed the diagnostic performance and reproducibility of CT scores. Patients with recurrent NTM-PD showed higher CT scores for bronchiectasis (<i>p</i> = 0.008), nodules (<i>p</i> = 0.006), consolidation (<i>p</i> = 0.033), and total CT scores (<i>p</i> = 0.017) at the time of treatment termination. On the contrary, only nodule score differed among the initial CT scores (<i>p</i> = 0.014). Regression analysis showed that the scores for bronchiectasis (odds ratio (OR) = 1.638, 95% confidence interval (CI) = 1.049–2.558, <i>p</i> = 0.030) and nodules (OR = 5.246, 95% CI = 1.370–20.087, <i>p</i> = 0.016) at treatment termination were significant predictors. The AUC of the regression model was 0.814 (95% CI = 0.689–0.939, <i>p</i> = 0.005). The interreader agreement for the total CT score was excellent (intraclass correlation coefficient = 0.841, <i>p</i> < 0.001). CT scores at the time of treatment termination can predict disease recurrence with good reproducibility.