The Incidence of Nontuberculous Mycobacterial Pulmonary Infection in Bilateral Bronchiectasis and Bronchiolitis

Overview

Nontuberculous mycobacteria (NTM) are ubiquitous organisms in the environment and are now increasingly being recognized as significant causes of chronic pulmonary infection in immunocompetent individuals (1). The most frequently encountered NTM lung disease worldwide is caused by Mycobacterium avium-intracellular complex (MAC) (2-4). In several studies with chest computed tomography (CT), researchers have demonstrated that the presence of bilateral multifocal bronchiolitis (well-defined small nodules and branching centrilobular nodules, or tree-in-bud pattern) and bronchiectasis distributed mainly in the right middle lobe and lingular segment are indicative of NTM pulmonary infection (7-11). Accordingly, it is believed that radiologic findings of bilateral bronchiolitis and bronchiectasis on chest CT scans specifically suggest NTM pulmonary infection (1). These CT findings, however, may not be specific for NTM pulmonary infection. CT patterns of bronchiectasis and bronchiolitis in the pulmonary infections caused by various NTM organisms have been reported, and these organisms include Mycobacterium kansasii, Mycobacterium xenopi, and rapidly growing mycobacteria such as Mycobacterium abscessus, Mycobacterium fortuitum, and Mycobacterium chelonae (12-14). In addition, not all patients with bronchiectasis and bronchiolitis have NTM pulmonary infection. Two recent studies showed that only about 50% of patients with such CT features have MAC pulmonary infection (9,15). To the best of our knowledge, however, there is no report about the incidence of NTM in patients with bronchiectasis or bronchiolitis in countries with low incidence of TB. Thus, the purpose of our study was to determine the frequency of NTM pulmonary infection in patients with bilateral bronchiectasis and bronchiolitis at chest CT and to investigate whether these CT findings are specifically indicative of MAC infection or other specific pathogen.

Study Population During the study period helical thin-section thoracic CT scans will obtained in patients in whom a diagnosis of bronchiectasis was known or who were suspected of having bronchiectasis on the basis of clinical symptoms and signs of a chronic cough and sputum production. One chest radiologist will review consecutive patients with bilateral multifocal bronchiectasis and bronchiolitis (well-defined small nodules and branching centrilobular nodules), regardless of the presence of cavitary lesions, who were suspected of having NTM pulmonary infection on the basis of these chest CT findings. Those patients with unilateral bronchiectasis, although the bronchiectasis was combined with bronchiolitis, will be excluded from the study. Those with isolated findings of bilateral bronchiectasis or bilateral bronchiolitis will also also be excluded. The patients will undergo a diagnostic work-up for NTM pulmonary infection. For a diagnosis of NTM infection, sputum AFB (Ziehl-Neelsen method) staining and culture examinations for mycobacteria will be performed at least three times. Bronchoscopy also will be performed for bronchial washing or transbronchial lung biopsy in patients with no sputum. Isolation and Identification of Mycobacteria Expectorated sputum and samples that will be obtained with bronchoscopy will be examined after AFB staining and will be cultured for mycobacteria. Colony numbers will be counted after incubation for as long as 8 weeks. Mycobacterium tuberculosis isolates will be identified (Gen-Probe Amplified Mycobacterium Tuberculosis Direct Test; Gen-Probe, San Diego, Calif), and NTM species identification will be confirmed by using a polymerase chain reaction-restriction fragment length polymorphism method based on the rpoB gene (16). In all cases, sputum examinations and bronchoscopic samples will be obtained within 6 months after CT. Diagnostic Criteria The diagnosis of definite NTM pulmonary infection will be determined when patients fulfilled the 1997 American Thoracic Society diagnostic criteria, which include clinical, radiologic, and microbiologic criteria (1). Probable NTM pulmonary infection will be diagnosed if the patients will be satisfied the clinical and radiologic criteria of the 1997 American Thoracic Society diagnostic criteria but will not fully satisfy the microbiologic criteria. Namely, the patients either will have only one or two cultures positive for organisms specifically identified as NTM with smear results negative for AFB or had more than three cultures positive for NTM without precise species identification; in addition, the patients will have symptoms and radiologic evidence of disease, which consisted of bilateral bronchiectasis and bronchiolitis. A diagnosis of pulmonary tuberculosis will be made when patients will have a culture positive for M tuberculosis. The clinical diagnostic criteria for diffuse panbronchiolitis will be as follows (17): (a) symptoms-chronic cough, sputum, and dyspnea on exertion; (b) physical signs-crackles and rhonchi; (c) chest radiographs-diffusely disseminated fine nodular opacities, mainly in the lower zones of the lung, with hyperinflation of the lungs; (d) lung function studies, with at least three of the following four abnormalities-forced expiratory volume in 1 second of less than 70%, vital capacity of less than 80% of the predicted value, residual volume of more than 150% of the predicted value, and partial pressure of oxygen (arterial) of less than 80 mm Hg. When no definite cause will be identified after evaluation of the causative factors, patients will received a diagnosis of bronchiectasis or nonspecific bronchiolitis. Chart Review The medical records of all patients will be reviewed by one of the authors for the following information, if available: height, body weight, history of tobacco use, history of previous antituberculous treatment, results of AFB staining, isolation and identification of mycobacteria, and the number of NTM-positive isolates. Imaging Evaluation Chest radiologists will retrospectively evaluate the chest CT scans. The observer will be unaware of the microbiologic test results and final diagnoses; A total of six lung lobes in each patient (the lingular segment will be considered a separate lobe) will be assessed for the presence of lung lesions and other abnormal findings. Each lobe in the lungs will be evaluated with regard to the presence or absence of bronchiectasis, well-defined small nodules (\<10 mm in diameter), and branching centrilobular nodules (ie, tree-in-bud pattern). Bronchiolitis will be defined as the presence of well-defined small nodules and branching centrilobular nodules on chest CT scans. The extents of involvement of bronchiectasis and bronchiolitis will be estimated by counting the number of involved lobes. The presence of other abnormalities, including nodules (10-30 mm in diameter), lobular consolidation (consolidation of 10-20 mm in diameter with a polygonal shape), segmental consolidation, cavities, and volume reduction, will be also recorded. Comparison between Patients with NTM Disease and Patients with Non-NTM Disease After classification of patients according to the previously mentioned diagnostic criteria, the patients will be categorized into two groups: the NTM pulmonary infection group, including the patients with definite or probable NTM pulmonary infection, and the non-NTM pulmonary infection group. Comparisons of the clinical and chest CT findings in patients with NTM diseases with those in patients with other airway diseases will be determined. Statistical Analysis Values are expressed as the mean ± standard deviation. By stratifying the age distribution of included patients according to birth sex and decades of patients' ages, the investigators will be tested whether there will be statistically significant differences between the distributions by using the χ2 test. All continuous variables will be tested for normality with the Kolmogorov-Smirnov test, and all had a Gaussian distribution. To statistically evaluate differences in continuous variables between the two groups, the investigators will use the unpaired t test. Frequencies will be analyzed by using the χ2 test or the Fisher exact test, as appropriate. A difference with a P value of less than .05 will be considered statistically significant. Statistical software (SPSS 11.0; SPSS, Chicago, Ill) will be used throughout.