Impact of Bronchiectasis on Incident Nontuberculous Mycobacterial Pulmonary Disease: A 10-Year National Cohort Study

Abstract

Nontuberculous mycobacterial pulmonary disease (NTM-PD) is common in bronchiectasis patients and is associated with poor outcomes.1Maiz L. Giron R. Olveira C. Vendrell M. Nieto R. Martinez-Garcia M.A. Prevalence and factors associated with nontuberculous mycobacteria in non-cystic fibrosis bronchiectasis: a multicenter observational study.BMC Infect Dis. 2016; 16: 437Crossref PubMed Scopus (34) Google Scholar,2Kunst H. Wickremasinghe M. Wells A. Wilson R. Nontuberculous mycobacterial disease and Aspergillus-related lung disease in bronchiectasis.Eur Respir J. 2006; 28: 352-357Crossref PubMed Scopus (94) Google Scholar Regarding the relationship between bronchiectasis and NTM-PD, a recent study suggested that bronchiectasis patients are vulnerable to NTM infection.3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar However, studies evaluating this issue using a nationally representative database are lacking. Besides, only a few studies have evaluated the risk factors for NTM infection in bronchiectasis patients.1Maiz L. Giron R. Olveira C. Vendrell M. Nieto R. Martinez-Garcia M.A. Prevalence and factors associated with nontuberculous mycobacteria in non-cystic fibrosis bronchiectasis: a multicenter observational study.BMC Infect Dis. 2016; 16: 437Crossref PubMed Scopus (34) Google Scholar,3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar Hence, we aimed to evaluate the association of bronchiectasis and NTM-PD and the risk factors of NTM-PD in patients with bronchiectasis, using a nationwide longitudinal cohort. We used the National Health Insurance Service—National Sample Cohort (NHIS-NSC), a 2.2% representative sample data of Koreans. Bronchiectasis was defined as one or more claim under the International Classification of Diseases 10th revision code J47.4Choi H. Yang B. Nam H. et al.Population-based prevalence of bronchiectasis and associated comorbidities in South Korea.Eur Respir J. 2019; 54: 1900194Crossref PubMed Scopus (28) Google Scholar NTM-PD was defined by a diagnostic code of NTM disease (A31.0, A31.8, and A31.9) and claims data for acid-fast bacilli smears or mycobacterial cultures. Bronchiectasis-associated comorbidities were also defined using the International Classification of Diseases 10th revision codes.4Choi H. Yang B. Nam H. et al.Population-based prevalence of bronchiectasis and associated comorbidities in South Korea.Eur Respir J. 2019; 54: 1900194Crossref PubMed Scopus (28) Google Scholar To account for competing risks due to mortality, a proportional subdistribution hazards regression model for NTM-PD with death as a competing event was used. We used the Cox proportional hazards regression model to evaluate risk factors for NTM-PD in the bronchiectasis cohort (see footnotes in Table 1 for adjusted factors). The institutional review board approved the study and waived the requirement for informed consent because the NHIS-NSC was de-identified (IRB No. CBNUH 2020-07-020).Table 1Risk Factors for Nontuberculous Mycobacterial Pulmonary Disease in Patients With BronchiectasisUnivariable AnalysisMultivariable AnalysisaAdjusted for age, sex, type of insurance, comorbidities, and use of inhaled corticosteroids, oral corticosteroids, and long-term macrolide.HR95% CIPAdjusted HR95% CIPAge group ≤39 yRefRefRefRef 40-49 y5.351.86-15.42.0025.501.90-15.88.002 50-59 y5.692.01-16.13.0015.221.83-14.86.002 60-69 y5.802.06-16.34.0014.981.75-14.18.003 ≥70 y3.991.35-11.81.0123.581.20-10.71.023Sex MaleRefRefRefRef Female1.390.95-2.04.0911.641.11-2.41.014Type of insurance Self-employed health insuranceRefRefRefRef Employee health insurance1.210.82-1.79.3391.250.84-1.85.271 Medical aid1.400.34-5.82.6461.390.33-5.81.652Comorbidities COPD1.360.92-2.02.1250.900.58-1.38.620 Asthma1.310.90-1.92.1600.990.64-1.51.945 Previous pulmonary TB4.713.20-6.93<.0014.262.84-6.39<.001 Rheumatoid arthritis0.740.30-1.82.5160.580.23-1.44.239 Lung cancer0.990.37-2.69.9840.610.22-1.69.344Medication Use of inhaled corticosteroids0RefRefRefRefLess than 1 y0.930.54-1.62.8000.700.39-1.25.2291 or more years2.041.21-3.45.0081.180.64-2.17.592 Oral corticosteroid use<10 mg/dRefRefRefRef10 mg/d or more1.460.64-3.32.3711.170.50-2.77.720 Macrolide use9.265.97-14.37<.0016.824.26-10.90<.001HR = hazard ratio; Ref = reference.a Adjusted for age, sex, type of insurance, comorbidities, and use of inhaled corticosteroids, oral corticosteroids, and long-term macrolide. Open table in a new tab HR = hazard ratio; Ref = reference. Between January 1, 2003 and December 31, 2012, the database recorded data for 861,808 adult patients of at least 20 years of age. Among the 17,679 patients with bronchiectasis, we excluded those diagnosed with NTM-PD before bronchiectasis diagnosis (n = 22) and those who died within 1 year after enrollment in the cohort (n = 690). Among the 843,414 patients without bronchiectasis, we excluded those diagnosed with NTM-PD before matching (n = 2) and those who died within 1 year after enrollment (n = 4,665). After 1:4 matching for age and sex using a greedy match algorithm, we established the NTM-naive bronchiectasis cohort (n = 16,967) and the NTM-naive matched cohort (n = 67,868), and they were followed (after a 1-year washout period) until the development of NTM-PD, death, or December 31, 2013, whichever occurred first. The bronchiectasis and matched cohorts were well balanced in terms of age and sex (standardized differences were 0% for both). The proportion of patients who received medical aid was higher in the bronchiectasis cohort than in the matched cohort (2.9% vs 2.4%, P < .001). Regarding pulmonary comorbidities, the rates of asthma (34.4% vs 9.1%), COPD (27.8% vs 5.3%), previous pulmonary TB (12.1% vs 2.8%), pulmonary hypertension (0.2% vs 0%), and lung cancer (3.8% vs 0.4%) were greater in the bronchiectasis cohort than in the matched cohort (P < .001 for all). Concerning extrapulmonary comorbidities, diabetes mellitus (20.4% vs 15.6%), cardiovascular disease (10.6% vs 6.1%), gastroesophageal reflux (24.5% vs 13.7%), chronic liver disease (6.0% vs 3.3%), inflammatory bowel disease (0.6% vs 0.3%), and rheumatoid arthritis (5.8% vs 3.4%) were significantly more frequent in the bronchiectasis cohort than in the matched cohort (P < .001 for all). During the median follow-up duration of 6.1 years (interquartile range, 3.3-8.7 years), the age- and sex-adjusted incidence of NTM-PD was 109.1/100,000 person-years in the bronchiectasis cohort and 5.6/100,000 person-years in the matched cohort (subdistribution hazard ratio [HR] = 19.27; 95% CI = 12.82-28.95). The cumulative incidence of NTM-PD was significantly higher in the bronchiectasis cohort than in the matched cohort (Gray’s test, P < .001), which was consistent in the subgroups (Fig 1). Regarding risk factors of NTM-PD among the bronchiectasis cohort, as shown in Table 1, age ≥ 40 years (greatest HR among patients aged 40-49 years [adjusted HR (aHR) = 5.50; 95% CI = 1.90-15.88]), females (aHR = 1.64; 95% CI = 1.11-2.41), previous pulmonary TB (aHR = 4.26; 95% CI = 2.84-6.39), and long-term macrolide (aHR = 6.82; 95% CI = 4.26-10.90) were associated with an increased incidence of NTM-PD in a multivariable analysis. A previous study showed that bronchiectasis was associated with increased odds of NTM-PD.5Andréjak C. Nielsen R. Thomsen V. Duhaut P. Sørensen H.T. Thomsen R.W. Chronic respiratory disease, inhaled corticosteroids and risk of non-tuberculous mycobacteriosis.Thorax. 2013; 68: 256-262Crossref PubMed Scopus (205) Google Scholar However, they used a cross-sectional design. Overcoming this limitation, a previous study showed that bronchiectasis patients are at a higher risk of NTM culture than the general population.3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar Supporting this view, our study showed that bronchiectasis patients have a higher risk of NTM-PD compared with age- and sex-matched populations, using a nationally representative database. As shown in a previous study,3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar the age group of 40 to 59 years had the highest risk of NTM-PD, and females had a higher risk of NTM-PD than males in our study. Whereas smoking and socioeconomic status were risk factors for NTM infection in previous studies,3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar,6Huang H.L. Cheng M.H. Lu P.L. et al.Epidemiology and predictors of NTM pulmonary infection in Taiwan: a retrospective, five-year multicenter study.Sci Rep. 2017; 7: 16300Crossref PubMed Scopus (34) Google Scholar these factors were not associated with NTM-PD in this study. Previous pulmonary TB was associated with a more than fourfold increase in NTM-PD in patients with bronchiectasis, in agreement with previous findings.6Huang H.L. Cheng M.H. Lu P.L. et al.Epidemiology and predictors of NTM pulmonary infection in Taiwan: a retrospective, five-year multicenter study.Sci Rep. 2017; 7: 16300Crossref PubMed Scopus (34) Google Scholar Inhaled or oral corticosteroids were previously reported to increase the risk of NTM-PD in patients with chronic airway diseases.3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar,5Andréjak C. Nielsen R. Thomsen V. Duhaut P. Sørensen H.T. Thomsen R.W. Chronic respiratory disease, inhaled corticosteroids and risk of non-tuberculous mycobacteriosis.Thorax. 2013; 68: 256-262Crossref PubMed Scopus (205) Google Scholar,7Sexton P. Harrison A.C. Susceptibility to nontuberculous mycobacterial lung disease.Eur Respir J. 2008; 31: 1322-1333Crossref PubMed Scopus (117) Google Scholar However, except for one study,3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar bronchiectasis was evaluated as one of the chronic lung diseases in the previous studies.5Andréjak C. Nielsen R. Thomsen V. Duhaut P. Sørensen H.T. Thomsen R.W. Chronic respiratory disease, inhaled corticosteroids and risk of non-tuberculous mycobacteriosis.Thorax. 2013; 68: 256-262Crossref PubMed Scopus (205) Google Scholar,7Sexton P. Harrison A.C. Susceptibility to nontuberculous mycobacterial lung disease.Eur Respir J. 2008; 31: 1322-1333Crossref PubMed Scopus (117) Google Scholar In contrast to the previous findings, inhaled or oral corticosteroid use did not increase NTM-PD in bronchiectasis patients in our study. Because a possibility exists that the results might have been caused by a relatively small number of NTM-PD, future studies are needed to address this issue. This study showed that long-term macrolide therapy was associated with incident NTM-PD in patients with bronchiectasis. Although macrolide plays a central role in the treatment of NTM-PD, long-term use in subjects with chronic lung disease may increase NTM infection.8Renna M. Schaffner C. Brown K. et al.Azithromycin blocks autophagy and may predispose cystic fibrosis patients to mycobacterial infection.J Clin Invest. 2011; 121: 3554-3563Crossref PubMed Scopus (227) Google Scholar The blockade of autophagy caused by macrolide was suggested as a mechanism for this phenomenon.8Renna M. Schaffner C. Brown K. et al.Azithromycin blocks autophagy and may predispose cystic fibrosis patients to mycobacterial infection.J Clin Invest. 2011; 121: 3554-3563Crossref PubMed Scopus (227) Google Scholar Because previous studies showed conflicting results regarding this issue,3Shteinberg M. Stein N. Adir Y. et al.Prevalence, risk factors and prognosis of non tuberculous mycobacteria infection among people with bronchiectasis: a population survey.Eur Respir J. 2018; 51: 1702469Crossref PubMed Scopus (15) Google Scholar,8Renna M. Schaffner C. Brown K. et al.Azithromycin blocks autophagy and may predispose cystic fibrosis patients to mycobacterial infection.J Clin Invest. 2011; 121: 3554-3563Crossref PubMed Scopus (227) Google Scholar,9Metersky M. Choate R. Addrizzo-Harris D. et al.The association of chronic macrolide treatment and incident isolation of non-tuberculous mycobacteria in patients with bronchiectasis.in: D12. Epidemiology and Therapy in CF and Non-CF Bronchiectasis. American Thoracic Society, 2020Crossref Google Scholar future research is warranted to elucidate. Plus, considering long-term macrolide monotherapy increases the risk for macrolide-resistant NTM-PD—difficult-to-treat disease10Choi H. Kim S.Y. Lee H. et al.Clinical characteristics and treatment outcomes of patients with macrolide-resistant Mycobacterium massiliense lung disease.Antimicrob Agents Chemother. 2017; 61Crossref Scopus (21) Google Scholar—it would be very important to collect sputum samples for NTM culture in bronchiectasis patients undergoing long-term macrolide treatment. There were several limitations to this study. First, the NHIS-NSC database did not provide data regarding BMI, pulmonary function, and smoking. Second, our study lacked NTM culture results. The NHIS-NSC provides claims of NTM cultures; however, it does not indicate whether the results are positive or negative. Thus, the definition of NTM-PD without NTM culture results might have exaggerated the risk of NTM-PD in bronchiectasis patients. Third, because of the lack of microbiological data, we could not identify the species of NTM-PD, which is important from a global perspective and extrapolation. In conclusion, the incidence of NTM-PD in bronchiectasis patients was approximately 19-fold higher than that in patients without bronchiectasis. Age, female sex, previous pulmonary TB, and long-term macrolide use were associated with increased NTM-PD in patients with bronchiectasis. Author contributions: H. Choi and H. Lee are guarantors of manuscript. J. Ryu did the data analysis. B, Yang, J. Ryu, H. Choi, and H. Lee wrote the initial draft of the manuscript and all authors were involved at all stages of critical revision of manuscript. All of the authors read and approved the final manuscript. Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript. Other contributions: We thank Sun-Young Kong, MD, PhD (National Cancer Center, Goyang, Korea), for providing us the NHIS-NSC dataset.