Serial CT Findings of Nodular Bronchiectatic Mycobacterium avium Complex Pulmonary Disease With Antibiotic Treatment

Transcription

1 Cardiopulmonary Imaging Original Research Lee et al. Serial CT of MAC Pulmonary Disease Cardiopulmonary Imaging Original Research Geewon Lee 1 Hyun Su Kim 1 Kyung Soo Lee 1 Won-Jung Koh 2 Kyeongman Jeon 2 Byeong-Ho Jeong 2 Joonghyun Ahn 3 Lee G, Kim HS, Lee KS, et al. Keywords: CT, Mycobacterium avium, Mycobacterium avium complex, Mycobacterium intracellulare, nontuberculous mycobacteria DOI: /AJR Received September 3, 2012; accepted after revision January 28, Department of Radiology and Center for Imaging Science, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50, Ilwon-Dong, Gangnam-Gu, Seoul , Korea. Address correspondence to K. S. Lee (kyungs.lee@samsung.com). 2 Department of Medicine, Division of Pulmonary and Critical Care Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Gangnam-Gu, Seoul, Korea. 3 Samsung Biomedical Research Institute, Samsung Medical Center, Gangnam-Gu, Seoul, Korea. AJR 2013; 201: X/13/ American Roentgen Ray Society Serial CT Findings of Nodular Bronchiectatic Mycobacterium avium Complex Pulmonary Disease With Antibiotic OBJECTIVE. The objective of our study was to analyze the serial CT findings of patients with the nodular bronchiectatic form of Mycobacterium avium complex (MAC) pulmonary disease treated with antibiotic therapy. MATERIALS AND METHODS. Between January 2005 and December 2009, MAC lung disease was diagnosed in 475 patients at a single tertiary referral hospital. Of the 475 patients, 339 had a CT pattern of disease consistent with the nodular bronchiectatic form. Among these 339 patients, 110 patients treated with a combination of antibiotics for 1 year were selected for this study. Two independent chest radiologists reviewed retrospectively the chest CT scans of 101 patients (M. avium disease [n = 57] and M. intracellulare disease [n = 44]) in whom serial CT scans had been obtained at the beginning of and at 12 months after standardized therapy. Each CT study was assessed for the presence and extent of lung parenchymal abnormalities (maximum score, 30). RESULTS. After 12 months of antibiotic therapy, 84 patients (83%) had a decrease in the overall CT score, three (3%) had an increase, and 14 (14%) had no change in disease extent. The decrease in total CT score was statistically significant (overall score difference, 2.54; p < ). Cellular bronchiolitis showed the largest decrease in extent (difference in mean preand posttreatment scores, 1.02, 1.07, and 0.94 for MAC, M. avium, and M. intracellulare diseases, respectively). Before treatment, patients with M. intracellulare disease showed more extensive disease than patients with M. avium disease (total CT score, vs 11.10; p = 0.025). CONCLUSION. In the nodular bronchiectatic form of MAC pulmonary disease, lung parenchymal abnormalities show a significant decrease in extent on CT after antibiotic treatment and the decrease is mainly related to the improvement of cellular bronchiolitis. M ycobacterium avium complex (MAC) has been known as the most common cause of nontuberculous mycobacteria (NTM) pulmonary disease. MAC consists mainly of two mycobacterial species: M. avium and Mycobacterium intracellulare [1, 2]. Because M. avium and M. intracellulare are similar in morphology and cause clinical features that are indistinguishable from each other, the organisms have been called collectively MAC [3]. With the advancement of molecular biology such as DNA probe like 16S ribosomal RNA gene sequencing, those two species are now better and more confidently differentiated from each other than before [4]. While analyzing 127 cases of MAC pulmonary disease (65 cases of M. intracellulare disease and 62 cases of M. avium disease), Han et al. [4] reported statistically significant differences between the cases depending on Mycobacterium species in clinical significance, patient age at infection, and pathogenicity. They found that M. intracellulare is more pathogenic than M. avium and tends to infect older women (age 50 years). Moreover, the M. intracellulare prevalence rate of 1.86% in postmenopausal women suggests the need to further investigate the public health significance of M. intracellulare disease [4]. The reported treatment success rate of MAC lung disease with macrolide-containing regimens is approximately 60 80%, with success defined as the eradication of organisms without relapse over a several-year period after the cessation of treatment. The high failure rate is caused by the prolonged treatment period, the side effects of antibiotic treatment, and possible reinfection rather than relapse [3, 5]. MAC lung disease manifests as two types of clinicoradiologic findings: men with underlying lung disease who present with upper lobe fibrocavitary disease and postmenopausal women who present with nodular bronchiectatic disease involving typically the right mid- 764 AJR:201, October 2013

2 Serial CT of MAC Pulmonary Disease Nodular bronchiectatic form: n = 339 Patients who were given the standard antibiotic therapy for more than 1 year: n = 110 Sputum negative conversion: n = 96 Patients enrolled in study: n = 101 dle lobe and lingular division of the left upper lobe [1, 2]. Recent reports show that MAC disease incidence rates have changed, with higher incidence rates in the nodular bronchiectatic form than the upper lobe fibrocavitary form for the past several decades in North America, Japan, and Korea [6 8]. Compared with the considerable number of articles in the literature covering the traditional type of MAC pulmonary infection, there are substantially fewer articles focusing on the nodular bronchiectatic form of MAC. Furthermore, whereas there is widespread agreement on immediate treatment of the upper lobe fibrocavitary form, the guideline on by whom and when the nodular bronchiectatic form of MAC pulmonary disease should be treated is unclear. Also, quantitative evidence of radiologic improvement after the treatment of nodular bronchiectatic MAC pulmonary disease and its relationship with clinical outcome have yet to be documented. Therefore, the aim of our study was to analyze in a relatively large group of patients the serial CT findings of patients with the nodular bronchiectatic form of MAC pulmonary disease treated with antibiotic therapy. Materials and Methods Our institutional review board approved this retrospective study and patient informed consent was waived for reviewing medical records. New diagnosis of MAC lung disease during 5-year period: n = 475 patients Upper lobe fibrocavitary form: n = 78 Unclassifiable form: n = 58 Excluded patients: n = did not undergo treatment 56 did not complete standard treatment period of 1 year Failure of sputum negative conversion: n = 5 Excluded patients: n = 9 4 did not have a follow-up CT study 3 underwent lobectomy 2 had connective tissue disease associated interstitial lung disease Fig. 1 Flowchart shows inclusion and exclusion criteria applied for selection of study group. MAC = Mycobacterium avium complex. Patient Enrollment During the 5-year period between January 2005 and December 2009, MAC lung disease was diagnosed in 475 patients at a single tertiary referral hospital. The cases were sought from an NTM disease registry in which all NTM culture examination data obtained at our hospital are entered. The diagnosis of MAC lung disease was made by a pulmonologist, and none of the patients had been treated previously for the disease. All patients met the diagnostic criteria of NTM lung disease according to the American Thoracic Society (ATS) guidelines [1], and none had HIV infection. Before treatment, all patients underwent thoracic CT and were classified to have one of the following three forms of the disease: fibrocavitary, nodular bronchiectatic, or unclassifiable [9, 10]. Of the 475 patients, 339 (71%) patients had the nodular bronchiectatic form, and standard antibiotic therapy for MAC disease was started and given for more than 1 year in 110 patients who had substantial symptoms or signs or had advanced or progressive radiologic abnormalities [5]. Of the 110 patients, we chose 101 patients who fulfilled the following inclusion criteria: Patients had undergone both an initial CT study and a follow-up CT study at 12 months after the initiation of antibiotic therapy, and they did not have an underlying malignancy such as lung cancer, metastatic lung disease, or interstitial lung disease. Nine patients who did not have a follow-up CT study, underwent lobectomy, or had an underlying connective tissue disease associated interstitial lung disease were excluded from this study (Fig. 1). The median time interval between the initial CT before treatment and the initiation of treatment was 7 days (interquartile range, days). Etiologic organisms were verified as M. avium in 57 patients (56%) 23 men (mean age, 61 ± 9.6 [SD] years; age range, years) and 34 women (55 ± 9.7 years; years) and as M. intracellulare in 44 patients (44%) 13 men (mean age, 65 ± 9.4 years; age range, years) and 31 women (59 ± 10.3 years; age range, years). A comprehensive review of the medical records of all patients was conducted to obtain clinical data including pulmonary symptoms or signs and medical history. Sputum Examination and Identification All patients had sputum smear and culture examinations performed with standard methods monthly for the first 6 months after the initiation of treatment and then at 3-month intervals until the end of treatment. Sputum smear was prepared from concentrated specimens and was stained using the carbolfuchsin method; a positive smear was defined as > 1 acid-fast bacillus (AFB) per 100 high-power fields. For sputum culture, specimens were decontaminated using N-acetyl-lcysteine and 2% NaOH, and processed specimens were placed in plates on 3% Ogawa medium (Shinyang). The inoculated tubes were incubated at 37 C and inspected weekly for 8 weeks. The confirmation of NTM species was conducted using a polymerase chain reaction restriction fragment length polymorphism method based on the rpob gene. Negative sputum conversion was defined as three consecutive AFB-negative cultures within 6 months [5]. Image Acquisition The following CT scanners of various types made by different vendors were used to perform a total of 202 CT studies of the 101 patients in the study group: a second-generation dual-source CT system (Somatom Definition Flash, Siemens Healthcare [seven studies]), a 64-MDCT scanner (Aquilion 64, Toshiba Medical Systems [46 studies]; or LightSpeed VCT, GE Healthcare [57 studies]), a 40-MDCT scanner (Brilliance 40, Philips Healthcare [32 studies]), a 16-MDCT scanner (LightSpeed 16, GE Healthcare [36 studies]; or LightSpeed QX/i, GE Healthcare [seven studies]), and an 8-MDCT scanner (LightSpeed Ultra, GE Healthcare [17 studies]). IV contrast medium was not given to any of the patients. Scanning was performed from the lung apices to bases. Images were reconstructed using a high-spatial-frequency algo- AJR:201, October

3 Lee et al. TABLE 1: CT Scoring System for Assessment of Mycobacterium avium Complex (MAC) Disease Extent Score CT Findings (Maximum Score Possible) 0 Points 1 Point 2 Points 3 Points Bronchiectasis (9 points) Severity Absent Mild a Moderate b Severe c Extent, no. of involved segments Absent > 9 Mucus plugging, no. of involved segments Absent > 9 Cellular bronchiolitis (6 points) Severity Absent Mild d Moderate e Severe f Extent, no. of involved segments Absent > 9 Cavity (9 points) Diameter, cm Absent < > 5 Wall thickness, mm Absent < > 5 Extent, no. of cavities Absent > 5 Nodule extent, no. of involved segments (3 points) Absent > 9 Consolidation extent, no. of involved segments (3 points) Absent < > 5 Note Maximum total score possible is 30 points. Bronchus diameter is greater than adjacent vessel diameter. Bronchus diameter is two to three times greater than vessel diameter. Bronchus diameter is three times greater than vessel diameter. Cellular bronchiolitis identified in, peripheral lung, 1 cm from pleura. Definite cellular bronchiolitis, with involvement more than 1 3 cm from pleura. Extensive cellular bronchiolitis, extending to central lung. TABLE 2: Intraclass Correlation Coefficients (ICCs) for Each Pattern of Parenchymal Lung Abnormality Before and After CT Patterns Before After ICC 95% CI ICC 95% CI Bronchiectasis Cellular bronchiolitis Cavity Nodule Consolidation Total CT score rithm. The CT images were obtained with the following parameters: detector collimation, 1.25 or mm; FOV, 36 cm; beam pitch, 1.35 or 1.375; gantry speed, 0.5 or 0.6 second per rotation; 120 kvp; ma; and reconstruction interval, mm. The mean estimated radiation dose for each CT examination was 4.8 mgy (range, mgy). The imaging data were reformatted with a 2.5- mm section thickness for transverse images and a 2.0-mm section thickness for coronal images. All images were directly displayed at our standard mediastinal (window width, 400 HU; window level, 20 HU) and lung (window width, 1500 HU; window level, 700 HU) window settings on monitors (four monitors, image matrices, 8-bit viewable gray-scale, and 60-ft-lambert luminescence) of a PACS (Centricity 2.0, GE Healthcare Integrated Imaging Solutions). CT Interpretation Two independent chest radiologists with 3 and 5 years of experience in chest CT interpretation, respectively, evaluated retrospectively the chest CT scans. The observers were unaware of the clinical or microbiologic information; they performed their evaluations in random order irrespective of CT study sequence (i.e., pre- or posttreatment CT studies). Each lung lobe was evaluated for the presence and extent of parenchymal abnormalities including bronchiectasis, cellular (or inflammatory) bronchiolitis, nodules of mm in diameter, airspace consolidation, and cavities. Parenchymal abnormalities were defined according to the Glossary of Terms established by the Fleischner Society [11] or in consideration of a previous study [9]. Bronchiectasis was defined to be present when the bronchial lumen diameter was greater than the adjacent pulmonary artery without tapering of the bronchial lumen diameter. Because bronchial wall thickening and mucus plugging were frequently accompanied by bronchiectasis, they were considered under the category of bronchiectasis. Bronchial wall thickness was estimated by measuring the ratio of the airway wall thickness to the outer diameter of the corresponding bronchus. Mucus plugging was regarded to be present when a broad, linear or branching, attenuation lesion was observed within a proximal airway (lobar, segmental, or subsegmental bronchus) associated with airway dilatation. Cellular bronchiolitis was defined as the presence of centrilobular small nodules (< 10 mm in diameter) and branching nodular structures (i.e., tree-in-bud signs) on CT scans. Other abnormalities, including nodules (10 30 mm in diameter), cavities, and airspace consolidation, were also recorded. When new lesions, particularly areas of consolidation or nodules, appeared at follow-up CT (n = 3), confirmation that the new lesions were caused by MAC pulmonary disease was achieved. The confirmation was accomplished by ensuring that no other specific organisms and that no known tumorous or inflammatory lesions were associated with the new lesions (n = 2) or by culturing MAC organisms from the new lesion (n = 1). 766 AJR:201, October 2013

4 Serial CT of MAC Pulmonary Disease TABLE 3: Pattern and Distribution of Parenchymal Abnormalities at Initial CT in Patients With Mycobacterium avium Disease and Patients With Mycobacterium intracellulare Disease Laterality of M. intracellulare Disease (No. of Patients) Location of M. intracellulare Disease (No. of Lesions) Location of M. avium Disease (No. of Lesions) Laterality of M. avium Disease (No. of Patients) Total (n = 264) Right Lung Left Lung Right Lung Left Lung Total U M L U Li L (n = 342) U M L U Li L Uni Bi Uni Bi No. (%) of Patients With M. intracellulare Disease No. (%) of Patients With M. avium Disease CT Patterns Bronchiectasis 57 (100) 44 (100) Cellular 57 (100) 44 (100) bronchiolitis Cavity 9 (16) 16 (36) Nodules 45 (79) 36 (82) Consolidation 43 (75) 35 (80) Note Uni = unilateral, Bi = bilateral, U = upper lobe, M = middle lobe, L = lower lobe, Li = lingular division of the left upper lobe. Lesion Quantification The CT scores in terms of the severity of lung involvement in M. avium and M. intracellulare pulmonary diseases were calculated by modifying a previously published scoring system proposed by Song and coworkers [10] that appeared to correlate with measures of functional impairment in patients with MAC pulmonary disease (Table 1). This modified and detailed scoring system was used because simple visual quantification at several levels of CT images seemed to be inadequate for assessing the detailed changes in lung abnormalities seen at serial CT. Scores were given by considering the presence, severity, and extent of bronchiectasis (maximum score, 9), cellular bronchiolitis (maximum score, 6), and cavity (maximum score, 9) and by considering the presence and extent of nodules (maximum score, 3) and consolidation (maximum score, 3) in both lungs. The mean overall CT score for each pattern of parenchymal abnormality was defined as the mean score of the two observers divided by the total number of patients. The highest possible total CT score for the quantification of overall extent of lung disease was 30 points. Statistical Analyses For statistical analyses, statistics software (SAS, version 9.3, SAS Institute) for Microsoft Windows was used; p < 0.05 indicated statistical significance. For the frequency of lung lesion laterality and distribution, the readings from the two observers were averaged and the average was rounded. The proportions of increase, decrease, and no change of each CT pattern were assessed by calculating the difference values between the initial CT and followup CT studies in MAC, M. avium, and M. intracellulare diseases, respectively. To test the significance of differences in CT scores rated for overall extent and for each parenchymal abnormality before and after treatment in MAC pulmonary disease and in both M. avium and M. intracellulare diseases, we compared the mean data of disease extent of a total disease and each pattern of parenchymal abnormalities using a two-way analysis of variance mixed model. Additionally, we defined an objective value for quantitative comparison as the difference value, which was calculated simply by subtracting the pretreatment score from the posttreatment score for each CT pattern in each patient. A negative difference value meant improvement, whereas a positive difference value meant progression; a difference of 0 was defined as stable disease. First, the difference values of the five CT patterns were tested with the Friedman test, followed by a pairwise comparison between all pairs of the difference values of the five CT patterns using the Wilcoxon signed rank test in MAC, M. avium, and M. intracellulare diseases, respectively. From these results, a disease pattern that was most related to the improvement or progression of the disease was sought. The Bonferroni correction was applied when multiple comparisons were performed. Interobserver agreement for the extent and severity of lung parenchymal abnormalities was tested using a repeated measures of data analysis for the intraclass correlation coefficient (ICC). For the extent of lung lesions before and after treatment, readings from the two observers were averaged and rounded off. The 95% CIs for the ICC were estimated. The ICC was regarded to show poor agreement when it was < 0.4, moderate agreement when 0.4 ICC < 0.75, and high agreement when it was Results Clinical Features and Ninety-six of 101 (95%) patients (53 and 43 patients with M. avium and M. intracellulare disease, respectively) had symptoms such as cough, sputum, hemoptysis, and weight loss. Five asymptomatic patients (5%; four of 57 patients with M. avium disease and one of 44 with M. intracellulare disease) were also treated on the basis of deteriorating radiologic findings and pulmonologists discretion. In all patients, a standardized combination antibiotic therapy consisting of oral macrolides (clarithromycin or azithromycin), rifampin, and ethambutol was given usually for 12 months after sputum negative reconversion was achieved [1, 2]. Therefore, in most patients, treatment continued for at least months after the initiation of antibiotic therapy; the mean treatment period was 732 ± (SD) days (range, days; M. avium disease, 728 ± days; M. intracellulare disease, 738 ± days). Of the 96 symptomatic patients, 74 patients (73%) had improvement of their symptoms with treatment: 40 of 53 patients (75%) with M. avium disease and 34 of 43 patients (79%) with M. intracellulare disease. The remaining 22 patients (22%; 13 of 53 patients [25%] with M. avium disease and nine of 43 patients [21%] with M. intracellulare disease) showed no change in their symptoms. Sputum conversion was achieved in 96 of 101 (95%) patients (54 of 57 patients [95%] with AJR:201, October

5 Lee et al. Fig. 2 Improvement of Mycobacterium avium pulmonary disease with antibiotic treatment in 41-year-old woman. A, Lung window CT scan (2.5-mm-section thickness) obtained at level of basal trunk before treatment shows bronchiectasis (arrowheads) and cellular bronchiolitis (arrows) of tree-in-bud sign in right middle lobe, lingular division of left upper lobe, and right lower lobe. CT score was 13. B, Annual follow-up CT scan obtained at level similar to A after treatment shows decreased extent of lung involvement with improvement of cellular bronchiolitis. Bronchiectasis remains. CT score at this time was 11. M. avium disease and 42 of 44 patients [95%] with M. intracellulare disease). The mean total treatment period was 719 days in patients with negative sputum conversion, and 985 days in those whose sputum culture positivity persisted even with standard antibiotic treatment. Thirty-four patients had a history of tuberculosis treatment (18 and 16 patients with M. avium and M. intracellulare disease, respectively). However, none of the patients had active tuberculosis disease at the time of NTM disease diagnosis, and a considerable length of time had elapsed between previous tuberculosis treatment and initial CT for NTM disease diagnosis (range, 1 45 years). The mean body mass index was 20.3 ± 2.08 (range, ) for patients with M. avium disease and 19.9 ± 2.23 (range, ) for patients with M. intracellulare disease. Interobserver Agreement Interobserver agreement values (ICCs) for the extent of each pattern of abnormality on images obtained before and after treatment belonged to the category of high agreement (Table 2). The agreement was high in terms of the total CT score and scores of each pattern of parenchymal lesions. Initial CT Findings The frequency, laterality, and location of the lung lesions and the averaged values from the two observers are summarized in Table 3. All patients had bronchiectasis and cellular bronchiolitis owing to the patient selection criteria. Additional initial CT findings were nodules in 45 patients (79%) with M. avium disease and in 36 patients (82%) with M. intracellulare disease and consolidation in 43 (75%) and 35 (80%) patients, respectively. The CT scores for each pattern and for overall disease of lung parenchyma before and after antibiotic treatment are summarized in Table 4. Before treatment, there were no significant differences between the two groups based on individual scores for bronchiectasis (mean score, 4.30 vs 4.96 for M. avium disease and M. intracellulare disease, respectively), cellular bronchiolitis (3.89 vs 4.26), cavity (0.66 vs 1.73), nodule (0.97 vs 1.07), and consolidation (1.29 vs 1.30). In terms of overall extent, patients with M. intracellulare disease showed more extensive disease than those with M. avium disease before treatment (total CT score for M. avium vs M. intracellulare, vs 13.31, respectively; p = 0.025). A Changes of CT Findings After Of a total of 101 patients, 84 (83%) patients had a decrease in the overall score (Figs. 2 and 3), three (3%) had an increase (Fig. 4), and 14 (14%) had no change in the extent of the disease. In the five patients in whom sputum negative conversion had not been accomplished, four (80%) patients showed a decrease in the overall score and one (20%) showed stable disease. Specifically, of the patients with M. avium disease (n = 57), 47 (82%) had a decrease in the overall CT score, two (4%) patients had an increase, and eight (14%) had no change in the extent of the disease. Of the patients with M. intracellulare disease (n = 44), 37 (84%) had a decrease in the overall CT score, one (2%) patient had an increase, and six (14%) had no change in the extent of the disease. The mean total CT score of MAC disease showed a significant decrease in extent (overall score difference, 2.54; p < ) (Table 4). Although statistically not significant (p = 0.178), the difference was greater in M. intracellulare disease (overall change in extent of disease, 2.68) than in M. avium disease (2.42), and the disease extent was still larger in M. intracellulare disease (10.63) than M. avium disease (8.68) even after treatment (Table 4). Changes were statistically different in all five CT patterns and overall CT scores before and after the treatment in MAC disease, M. avium disease, and M. intracellulare disease with the exception of cavities in M. avium disease (p = 0.259). A decrease in the extent of the cavities, which was present even in the nodular bronchiectatic form of MAC disease, was significantly greater in patients with M. intracellulare disease (0.64) than in those with M. avium disease (0.18) after treatment (p = ). However, even after treatment, CT of patients with M. intracellulare disease showed a higher extent of cavitary lesions (mean score, 0.48 vs 1.09 for M. avium disease and M. intracellulare disease). In other words, except cavitary lesions in M. avium disease, the lung abnormalities showed obvious improvement after treatment and the amount of improvement was not different between M. avium and M. intracellulare diseases. B 768 AJR:201, October 2013

6 Serial CT of MAC Pulmonary Disease A B Fig. 3 Improvement of Mycobacterium intracellulare pulmonary disease with antibiotic treatment in 65-year-old woman. A, Lung window CT scan (2.5-mm-section thickness) obtained at level of inferior pulmonary veins shows bronchiectasis (arrowhead) in right middle lobe, cellular bronchiolitis (straight arrows) in both lungs, and lobular consolidation (curved arrows) in left upper and lower lobes. CT score was B, Annual follow-up CT scan obtained at level similar to A after treatment shows decreased extent of lung involvement with improvement of cellular bronchiolitis and of lobular consolidation. CT score at this time was 12. A B Fig. 4 Progressive Mycobacterium avium pulmonary disease with antibiotic treatment in 71-year-old man. A, Lung window CT scan (2.5-mm-section thickness) obtained at level of main bronchi shows bronchiectasis (arrowhead), mucus in medium-sized airway (thin arrows), and cellular bronchiolitis (thick arrow) in bottom of right upper lobe and in portion of right lower lobe. CT score was 10. B, Annual follow-up CT scan obtained at level similar to A after treatment shows increased extent of lung involvement with new appearance of lobular consolidation (curved arrow) in right upper lobe and patchy areas of cellular bronchiolitis (straight arrows) in both lungs. CT score at this time was 14. For each pattern of abnormality, the extent of improvement was proportional to the extent of the initial disease. In all three groups, cellular bronchiolitis showed the largest mean negative difference values ( 1.02, 1.07, 0.94 for MAC, M. avium, and M. intracellulare, respectively). Corrected p values of pairwise comparison between the difference value of each CT pattern in MAC, M. avium, and M. intracellulare diseases are shown in Table 5. In MAC disease, cellular bronchiolitis showed a significant difference compared with the other four variables (p < 0.001, p < 0.001, p < 0.001, p < for bronchiectasis, cavity, nodule, and consolidation, respectively). Likewise, cellular bronchiolitis also showed a significant difference compared with the remaining four variables (p = , p < 0.001, p < 0.001, p = for bronchiectasis, cavity, nodule, and consolidation, respectively) in M. avium disease. In M. intracellulare disease, there was a statistically significant difference between bronchiectasis versus cellular bronchiolitis (p = 0.023) and bronchiolitis versus nodule (p = 0.005). In other words, cellular bronchiolitis was the single CT pattern that was related to a decrease in the extent of disease at follow-up CT. Discussion Traditionally, pulmonary MAC infection occurs usually in patients with preexist- AJR:201, October

7 Lee et al. TABLE 4: Score and Adjusted p Values of Each Pattern of Parenchymal Abnormalities at CT Before and After Mycobacterium avium Complex Mycobacterium avium Mycobacterium intracellulare Score After Score Before Score After Score Before Score After Score Before CT Patterns Mean SD Mean SD Adjusted p a Mean SD Mean SD Adjusted p a Mean SD Mean SD Adjusted p a Bronchiectasis < < < Cellular bronchiolitis < < < Cavity < < Nodule < < < Consolidation < < < Total CT score < < < a Bonferroni adjustment. ing lung diseases such as chronic obstructive pulmonary disease, previous tuberculosis, or pneumoconiosis [12 14]. This type of MAC lung disease is still a main NTM disease and is a problem in many European countries [15, 16]. However, the nodular bronchiectatic form currently accounts for more than 50% of newly diagnosed cases of MAC pulmonary disease [6, 7, 17, 18]. This trend is thought to be because of the increased number of older adults in the population and perhaps because of a greater awareness about MAC infection as the pathogenesis of bronchiectasis [1, 2]. In the current study, all five findings of bronchiectasis, cellular bronchiolitis, cavity, nodules, and consolidation significantly decreased in patients with MAC disease after treatment even though cellular bronchiolitis showed the largest decrease in extent and was mainly related to disease improvement at follow-up CT. Kuroishi et al. [19] reported that the CT findings of a small nodule or nodules improved in 70% of cases of MAC pulmonary disease after treatment, which is consistent with the results of our study. Conversely, in that same study [19], neither pleural thickening nor bronchiectasis improved after treatment. In our study, we calculated the score of bronchiectasis as the sum of three subcategories: severity, extent, and mucus plugging. Mucus plugging and severity, which were evaluated by the bronchus diameter with respect to the accompanying pulmonary artery, were relatively reversible variables that probably accounted for the decrease in the score of bronchiectasis. Fujiuchi et al. [20] analyzed the chest CT scans of 30 patients with MAC pulmonary disease before and after therapy; they found that bronchial wall thickening and small nodules showed improvement, whereas pleural thickening and most cavitary lesions did not. Histopathologically, cellular bronchiolitis represents NTM granulomas, which are composed of epithelioid macrophages, giant cells, and perhaps some solid and sticky caseous materials within the respiratory bronchioles; therefore, cellular bronchiolitis may indicate a reversible status [21]. Several reports have described the changes in CT findings with treatment in patients with NTM pulmonary disease [19, 22, 23]. However, the novelty of our study is that we performed a comparative study of the two species that make up MAC lung disease. Moreover, we enrolled a relatively large study population of more than 100 patients who underwent standardized antibiotic therapy and who had follow-up CT studies at 12 months after the treatment. In terms of overall disease extent, the total CT score was statistically different between the two species (total CT score for M. avium vs M. intracellulare, vs 13.31, respectively; p = 0.025); in other words, patients with M. intracellulare disease showed more extensive disease than those with M. avium lung disease before treatment. Although the differences lacked significance, M. intracellulare disease showed a higher score in all five patterns of lung abnormalities compared with M. avium lung disease at initial CT scans. In addition, the disease extent was still greater in patients with M. intracellulare disease (total CT score, 10.63) than in those with M. avium disease (8.68) after treatment. Our results are consistent with those of Koh et al. [24]; in their study of 295 patients with MAC lung disease, Koh et al. found M. intracellulare disease manifests a more severe clinical presentation at the time of diagnosis than M. avium disease. Also, they reported that treatment response rates are lower in patients with M. intracellulare disease than in those with M. avium disease [24]. Thus, the results of our study provide further imaging evidence that M. intracellulare is a more virulent pathogen than M. avium. NTM disease is identified to the species level in general, but the diagnosis of the disease does not imply prompt treatment. MAC pulmonary disease is chronic in nature and responses to treatment are relatively slow. According to one study, without treatment, 60% of patients with the nodular bronchiectatic form of MAC lung disease showed disease progression and 40% of patients had stable disease during a mean observation period of 28 months at CT [25]; none showed spontaneous improvement. In brief, without treatment, most patients with MAC pulmonary disease have progressive disease. In another study [26], among 164 patients with a diagnosis of pulmonary MAC disease, the mortality rate was 33% for untreated patients with chronic MAC versus 22% for treated patients; these findings indicate that treatment to sputum negative conversion is warranted in patients with MAC disease and can improve the chance of survival [26]. Moreover, the high rate of adverse reactions to the regimen makes the achievement of clinical cure of MAC pulmonary disease difficult [27]. NTM disease including MAC disease differs from tuberculosis in that NTM disease is usually indolent and takes time to develop pulmonary lesions; nevertheless, 770 AJR:201, October 2013

8 Serial CT of MAC Pulmonary Disease TABLE 5: Significant Corrected p Values of Pairwise Comparisons in Terms of Difference Values of Each CT Pattern Among Patients With Mycobacterium avium Complex (MAC), Mycobacterium avium Disease, and Mycobacterium intracellulare Disease Variable 1 Variable 2 there are many patients with NTM disease in whom pulmonary structures are destroyed extensively. According to guidelines of the ATS and the Infectious Diseases Society of America, a favorable outcome is noted in patients who tolerate therapy [1]. The importance of our study lies in the fact that CT shows improvement in disease extent after treatment, with bronchiolitis being the CT pattern that is related to the improvement at follow-up CT, thus further confirming the necessity of treatment of patients with this disease. The importance of serial CT studies in MAC pulmonary disease should be elaborated. Findings of extensive bronchiolitis on CT may indicate the disease is a reversible state and may help us predict which patients have more potential for radiologic improvement after treatment. Because there is a relatively weak correlation between CT findings and pulmonary function test results [10], extensive disease on CT may indicate the necessity of antibiotic treatment and an improvement in disease extent may suggest treatment response such as sputum negative conversion. In patients with MAC pulmonary disease, reported sputum negative conversion rates are variable, ranging from 25% to 90% [2, 19, 27 30]. In the current study, the sputum negative conversion rate of 95% is relatively high. This high rate is presumably because we excluded patients with the upper lobe fibrocavitary or unclassifiable form of MAC disease and because a culture result was considered to be negative when patients could not expectorate sputum. Also, we excluded patients who did not fulfill the standard treatment period (e.g., because of gastrointestinal side effects); thus, treatment response rates tended to be higher. According to another study [31], the sputum culture response rate for patients with nodular bronchiectatic MAC lung disease was 83%. Adjusted p a MAC M. avium M. intracellulare Cellular bronchiolitis Bronchiectasis < Cellular bronchiolitis Cavity < < Cellular bronchiolitis Nodule < < Cellular bronchiolitis Consolidation < a Bonferroni adjustment. Our study has several limitations. First, it is retrospective; thus, selection bias might have been present. Moreover, because a high percentage (95%) of patients had negative sputum conversion rates, we could not compare CT findings of patients with and those without negative sputum conversion. Second, because of our inclusion criteria, only patients with the nodular bronchiectatic form of MAC pulmonary disease were studied selectively. Patients were excluded from the study if they had upper lobe fibrocavitary MAC pulmonary disease or an unclassifiable type, mainly because accurate evaluation of serial CT findings is difficult in patients with the fibrocavitary form of MAC lung disease. In that condition, patients frequently have a concurrent tuberculous or fungal infection or an emphysema-associated malignancy [14, 32]. Thus, it might have been difficult to compare the initial and follow-up CT studies of this population because concurrent infection or malignancy and the related treatment changes (surgical removal or radiation fibrosis) would have also contributed to imaging finding changes. In conclusion, the lung parenchymal abnormalities in the nodular bronchiectatic form of MAC pulmonary disease showed significant decreases in extent at CT with antibiotic treatment, which is related mainly to the improvement in the extent of cellular bronchiolitis at follow-up CT. Lung involvement at CT is more extensive in M. intracellulare disease than M. avium disease before and even after treatment. References 1. Griffith DE, Aksamit T, Brown-Elliott BA, et al.; ATS Mycobacterial Diseases Subcommittee; American Thoracic Society; Infectious Disease Society of America. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med 2007; 175: [Erratum in Am J Respir Crit Care Med 2007; 175: ] 2. Kasperbauer SH, Daley CL. Diagnosis and treatment of infections due to Mycobacterium avium complex. Semin Respir Crit Care Med 2008; 29: Field SK, Fisher D, Cowie RL. Mycobacterium avium complex pulmonary disease in patients without HIV infection. Chest 2004; 126: Han XY, Tarrand JJ, Infante R, Jacobson KL, Truong M. Clinical significance and epidemiologic analyses of Mycobacterium avium and Mycobacterium intracellulare among patients without AIDS. J Clin Microbiol 2005; 43: Sim YS, Park HY, Jeon K, Suh GY, Kwon OJ, Koh WJ. Standardized combination antibiotic treatment of Mycobacterium avium complex lung disease. Yonsei Med J 2010; 51: Koh WJ, Kwon OJ, Jeon K, et al. Clinical significance of nontuberculous mycobacteria isolated from respiratory specimens in Korea. Chest 2006; 129: Okumura M, Iwai K, Ogata H, et al. Clinical factors on cavitary and nodular bronchiectatic types in pulmonary Mycobacterium avium complex disease. Intern Med 2008; 47: Prevots DR, Shaw PA, Strickland D, et al. Nontuberculous mycobacterial lung disease prevalence at four integrated health care delivery systems. Am J Respir Crit Care Med 2010; 182: Kim HS, Lee KS, Koh WJ, et al. Serial CT findings of Mycobacterium massiliense pulmonary disease compared with Mycobacterium abscessus disease after treatment with antibiotic therapy. Radiology 2012; 263: Song JW, Koh WJ, Lee KS, et al. High-resolution CT findings of Mycobacterium avium-intracellulare complex pulmonary disease: correlation with pulmonary function test results. AJR 2008; 191: Hansell DM, Bankier AA, MacMahon H, McLoud TC, Müller NL, Remy J. Fleischner Society: glossary of terms for thoracic imaging. Radiology 2008; 246: Christensen EE, Dietz GW, Ahn CH, et al. Pulmonary manifestations of Mycobacterium intracellularis. AJR 1979; 133: Christensen EE, Dietz GW, Ahn CH, et al. Initial roentgenographic manifestations of pulmonary Mycobacterium tuberculosis, M. kansasii, and M. intracellularis infections. Chest 1981; 80: Rosenzweig DY. Pulmonary mycobacterial infections due to Mycobacterium intracellulare-avium complex: clinical features and course in 100 consecutive cases. Chest 1979; 75: Maugein J, Dailloux M, Carbonnelle B, Vincent V, Grosset J; French Mycobacteria Study Group. Sentinel-site surveillance of Mycobacterium avium complex pulmonary disease. Eur Respir J AJR:201, October

9 Lee et al. 2005; 26: Kim JS, Tanaka N, Newell JD, et al. Nontubercu- Chest 1999; 115: van Ingen J, Hoefsloot W, Dekhuijzen PN, Boeree lous mycobacterial infection: CT scan findings, 28. Ahn CH, Ahn SS, Anderson RA, Murphy DT, MJ, van Soolingen D. The changing pattern of genotype, and treatment responsiveness. Chest Mammo A. A four-drug regimen for initial treat- clinical Mycobacterium avium isolation in the 2005; 128: ment of cavitary disease caused by Mycobacteri- Netherlands. Int J Tuberc Lung Dis 2010; 23. Obayashi Y, Fujita J, Suemitsu I, Kamei T, Nii M, um avium complex. Am Rev Respir Dis 1986; 14: Iseman MD. Mycobacterium avium complex and the normal host: the other side of the coin. N Engl J Med 1989; 321: Prince DS, Peterson DD, Steiner RM, et al. Infection with Mycobacterium avium complex in patients without predisposing conditions. N Engl J Med 1989; 321: Kuroishi S, Nakamura Y, Hayakawa H, et al. Mycobacterium avium complex disease: prognostic implication of high-resolution computed tomography findings. Eur Respir J 2008; 32: Fujiuchi S, Matsumoto H, Yamazaki Y, et al. Analysis of chest CT in patients with Mycobacterium avium complex pulmonary disease. Respiration 2003; 70: Jeong YJ, Lee KS, Koh WJ, Han J, Kim TS, Kwon OJ. Nontuberculous mycobacterial pulmonary infection in immunocompetent patients: comparison of thin-section CT and histopathologic findings. Radiology 2004; 231: Takahara J. Successive follow-up of chest computed tomography in patients with Mycobacterium aviumintracellulare complex. Respir Med 1999; 93: Koh WJ, Jeong BH, Jeon K, et al. Clinical significance of the differentiation between Mycobacterium avium and Mycobacterium intracellulare in M. avium complex lung disease. Chest 2012; 142: Yamazaki Y, Kubo K, Takamizawa A, Yamamoto H, Honda T, Sone S. Markers indicating deterioration of pulmonary Mycobacterium avium-intracellulare infection. Am J Respir Crit Care Med 1999; 160: Ito Y, Hirai T, Maekawa K, et al. Predictors of 5-year mortality in pulmonary Mycobacterium avium-intracellulare complex disease. Int J Tuberc Lung Dis 2012; 16: Huang JH, Kao PN, Adi V, Ruoss SJ. Mycobacterium avium-intracellulare pulmonary infection in HIV-negative patients without preexisting lung disease: diagnostic and management limitations. 134: Tanaka E, Kimoto T, Tsuyuguchi K, et al. Effect of clarithromycin regimen for Mycobacterium avium complex pulmonary disease. Am J Respir Crit Care Med 1999; 160: Horsburgh CR Jr, Mason UG 3rd, Heifets LB, Southwick K, Labrecque J, Iseman MD. Response to therapy of pulmonary Mycobacterium aviumintracellulare infection correlates with results of in vitro susceptibility testing. Am Rev Respir Dis 1987; 135: Wallace R, Brown-Elliot B, McNulty S, et al. Outcome of standard multidrug therapies for nodular/ bronchiectatic (NB) Mycobacterium avium complex (MAC) lung disease. Am J Respir Crit Care Med 2010; 2010:A Kobashi Y, Fukuda M, Yoshida K, Miyashita N, Niki Y, Oka M. Chronic necrotizing pulmonary aspergillosis as a complication of pulmonary Mycobacterium avium complex disease. Respirology 2006; 11: AJR:201, October 2013