to the radiologic resemblance, a high incidence of the coexistence of TB and anthracobrosis and a high prevalence of anthra cobrosis among women support this relationship [14]. For these reasons, Chung et al. [1] reported that empiric anti-TB medication may be helpful in the treatment of anthracobrosis patients without evidence of active TB. We evaluated whether CT can be used to differentiate anthracobrosis from endobron- chial TB. Distinctive CT features of anth- racobrosis can help radiologists diagnose anthraco brosis with ease before broncho- scopic evaluation and can be a rationale negating the hypothesis that endo bronchial TB is a major cause of anthra co brosis. To our knowledge, this study is the rst comparison of the CT ndings in these two diseases. Materials and Methods We reviewed the CT scans of 49 patients with anthracobrosis diagnosed on the basis of CT Differentiation of Anthracobrosis from Endobronchial Tuberculosis Hyun Jin Park 1
Seog Hee Park 1
Soo Ah Im 1
Young Kyoon Kim 2
Kyo-young Lee 3 Park HJ, Park SH, Im SA, Kim YK, Lee K 1 Department of Radiology, Kangnam St. Marys Hospital, College of Medicine, Catholic University of Korea, 505 Banpo-dong Seocho-gu, Seoul, 137-040, South Korea. Address correspondence to S. H. Park (parksh@catholic.ac.kr). 2 Department of Internal Medicine, Kangnam St. Marys Hospital, College of Medicine, Catholic University of Korea, Seoul, South Korea. 3 Department of Pathology, Kangnam St. Marys Hospital, College of Medicine, Catholic University of Korea, Seoul, South Korea. Chest I magi ng Ori gi nal Research AJR 2008; 191:247251 0361803X/08/1911247 American Roentgen Ray Society B ronchial anthracobrosis is a bron- choscopic nding, dened as bron- chostenosis associated with an- thracotic pigmentation without a relevant history of pneumoconiosis [1]. The CT ndings have been reported to include smooth bronchostenosis and peribronchial lymph nodes along with peripheral atelectasis [24]. Bronchogenic carcinoma and endo bronchial tuberculosis (TB) are two major causes of bronchostenosis that can be ac companied by peribronchial nodes and atelectasis [5, 6]. Bronchogenic carcinoma is known to be easily differentiated from anthra cobrosis, which usually manifests as an endobronchial mass in a single lobar or segmental bronchus [2]. In contrast to the study of carcinoma, there has been, to our knowledge, no comparison study of the relation between anthracobrosis and endo bronchial TB. Therefore, differentiation of anthracobrosis from TB has been difcult. Because of the similar CT ndings, endobron- chial TB has been suggested as a mech a nism Keywords: anthracobrosis, bronchostenosis, CT, tuberculosis DOI:10.2214/AJR.07.2161 Received March 2, 2007; accepted after revision January 7, 2008. OBJECTIVE. The purpose of this study was to use CT to differentiate anthracobrosis from endobronchial tuberculosis (TB), both of which are major causes of benign bronchostenosis. MATERIALS AND METHODS. We retrospectively reviewed the clinical and CT ndings of 49 patients with anthracobrosis and 35 patients with endobronchial TB diagnosed on the basis of bronchoscopic, microbiologic, and pathologic ndings. Forty-ve patients with anthracobrosis and 32 patients with endobronchial TB had bronchostenosis on CT and were enrolled in the analysis. Nine (20%) of 45 patients with anthracobrosis had coexistent active TB (two, endobronchial TB; six, pulmonary TB; one, TB pleurisy), and 13 (29%) had pulmonary infections other than TB. Two patients with anthracobrosis and coexistent endobronchial TB were excluded from the analysis. The CT ndings were analyzed with emphasis on the pattern, distribution, and location of bronchostenosis and the number of pulmonary lobes involved. RESULTS. Anthracobrosis was more common than endobronchial TB among elderly patients (p < 0.05). Statistically signicant ndings on CT were the pattern of bronchostenosis, presence of main bronchus involvement, and number of pulmonary lobes involved (p < 0.05). Bronchostenosis with anthracobrosis usually involves multiple lobar or segmental bronchi. The main bronchus, however, tends to be preserved in anthracobrosis. Most cases of endobronchial TB involve one lobar bronchus and the ipsilateral main bronchus with contiguity in extent. CONCLUSION. Anthracobrosis can be differentiated from endobronchial TB on CT. Furthermore, CT is helpful in the diagnosis of anthracobrosis before bronchoscopy is performed. Park et al. CT of Anthracofibrosis and Tuberculosis Chest Imaging Original Research D o w n l o a d e d
f r o m
w w w . a j r o n l i n e . o r g
b y
3 6 . 7 2 . 6 5 . 2 0 0
o n
0 1 / 2 6 / 1 4
f r o m
I P
a d d r e s s
3 6 . 7 2 . 6 5 . 2 0 0 .
C o p y r i g h t
A R R S .
F o r
p e r s o n a l
u s e
o n l y ;
a l l
r i g h t s
r e s e r v e d
248 AJR:191, July 2008 Park et al. bronchoscopic criteria, including anthracotic pig- mentation and bronchostenosis, over 3 years (July 2003April 2006). Forty-one (91%) of the 45 patients had bronchostenosis on CT. The other four were excluded from analysis. None of the patients had an occupational history of exposure to mining or industry. Thirty-four (76%) of the patients did not smoke. Six (13%) of the patients had a history of pulmonary TB, and four of the six took anti-TB medication. All 45 patients with anthracobrosis underwent bronchoscopy with bronchial washing and brushing, and 23 (51%) of the 45 underwent biopsy. The biopsy specimens exhibited chronic nonspecic inammation (n = 22) and edema with mononuclear cell inltration (n = 1). Broncho scopic biopsy re- vealed coexistent endobronchial TB in two patients with anthracobrosis. These two patients were excluded from the statistical analysis. Two patients with anthracobrosis underwent pulmonary lobec- tomy. Four lymph nodes were dissected after lobec- tomy in one patient. All lobectomy specimens re- vealed anthra cotic pigmentation with focal bro sis of bronchi. Dissected lymph nodes proved to have reactive hyperplasia with anthracotic pigmentation. CT scans of 35 patients with endobronchial TB were available. Thirty-two (91%) of the 35 patients had CT ndings of bronchostenosis and were enrolled in the analysis. Fourteen (40%) of the 35 cases of endobronchial TB were in the active inammatory phase with acid-fast bacilli. The other cases were in the chronic state, and the patients had been treated for active endobronchial TB. The diagnoses of endobronchial TB were based on bronchoscopic, microbiologic, and patho- logic ndings. Both the anthracobrosis and endobronchial TB groups of patients underwent CT before broncho- scopy. Two CT scanners (Volume Zoom, Siemens Medical Solutions; LightSpeed, GE Healthcare) were used. All scans were reconstructed into axial images with 5-mm slice thickness at 5-mm intervals. Both unenhanced and contrast-enhanced images were obtained. Two chest radiologists re- trospectively reviewed the CT scans, and a con- sensus diagnosis was reached. CT ndings were analyzed with emphasis on bronchostenosis. The pattern of bronchostenosis was analyzed with a focus on a smooth or irregular endobronchial contour. The number of pulmonary lobes with broncho stenosis was recorded. Main bronchus involvement and bilaterality of bronchostenosis in the two groups were assessed. Concomitant lymph- adeno pathy and pulmonary lesions were included in the analysis. Enlarged nodes greater than 1 cm in short-axis diameter were dened as lympha deno- pathy. Students t and Pearsons chi-square tests were used for statistical analysis. Results Thirty-four women and nine men (median age, 76 years; range, 5690 years) had anthracobrosis; 24 women and eight men (median age, 51 years; range, 1980 years) had endobronchial TB. There were more women than men in both groups. The anthracobrosis group had a higher mean age than the endobronchial TB group (p = 0.000, Pearsons chi-square test). Twenty- seven (63%) of 43 patients with anthra co- brosis had concomitant disease at the time of diagnosis. Most of the concomitant disease was reactivated TB (nine cases, 33%) and pneumonia (13 cases, 48%). The CT features and location of disease are summarized in Tables 1 and 2. Most of the 43 cases of anthracobrosis were characterized by smooth luminal narrowing (41 cases, 95%) and peribronchial lymphadenopathy (39 cases, 91%) (Fig. 1). Sixteen of the 32 cases of endobronchial TB were characterized by irregular bronchial narrowing (Fig. 2). Peribronchial and mediastinal lymphadenopathy were more commonly associated with anthracobrosis than with endobronchial TB (p < 0.05, Students t test). The distribution of bronchostenosis was statistically different in comparisons of the CT ndings for the two groups (p < 0.05, Students t test and Pearsons chi-square test). More lobes were involved in anthracobrosis than in endobronchial TB (mean, 3.1 vs 1.3 lobes; p = 0.0001). Bilateral lung involvement was more common in anthracobrosis than in endobronchial TB (65% vs 6% of patients, p = 0.0001). The most common location of bronchostenosis in anthracobrosis was the right upper lobe (42 [98%] of 43 cases) and in endobronchial TB was the left upper lobe (22 [69%] of 32 cases). Bronchostenosis in TABLE 2: Location of Bronchostenosis Lesions on CT Location Anthracobrosis (n = 43) Tuberculosis (n = 32) p a Right upper lobe 42 (98) 7 (22) 0.0001 Right middle lobe 36 (84) 2 (6) 0.0001 Right lower lobe 17 (40) 1 (3) 0.0001 Left upper lobe 23 (53) 22 (69) 0.236 Left lower lobe 15 (35) 10 (31) 0.808 Mean no. of lobes involved 3.1 1.3 0.0001 Involvement Right lung only 15 7 0.80 Left lung only 0 23 Both lungs 28 2 0.0001 NoteUnless otherwise indicated, values are number of patients with percentage in parentheses. a Statistical analysis with Pearsons chi-square test and Students t test. TABLE 1: CT Features in Anthracofibrosis and Tuberculosis Groups CT Finding Anthracobrosis (n = 43) Tuberculosis (n = 32) p a Bronchostenosis Pattern 0.0001 Smooth 41 16 Irregular 2 16 Mean no. of lobes involved 3.1 (range, 15) 1.3 (range, 15) 0.0001 Bilaterality 28 (65) 2 (6) 0.0001 Involvement of main bronchus 3 (7) 28 (88) 0.0001 Lymphadenopathy Peribronchial 39 (91) 20 (63) 0.0001 Mediastinal 38 (88) 22 (69) 0.005 NoteUnless otherwise indicated, values are number of patients with percentage in parentheses. a Pearsons chi-square test and Students t test. D o w n l o a d e d
f r o m
w w w . a j r o n l i n e . o r g
b y
3 6 . 7 2 . 6 5 . 2 0 0
o n
0 1 / 2 6 / 1 4
f r o m
I P
a d d r e s s
3 6 . 7 2 . 6 5 . 2 0 0 .
C o p y r i g h t
A R R S .
F o r
p e r s o n a l
u s e
o n l y ;
a l l
r i g h t s
r e s e r v e d
AJR:191, July 2008 249 CT of Anthracofibrosis and Tuberculosis the right upper lobe, or in any lobe of the right lung, was more common in anthra co brosis than in TB (p < 0.05). Both anthra co brosis and TB commonly involved the left upper lobe, but the difference was not statistically signicant (53% vs 69%, p = 0.236). Only left lung involvement was absent in anthraco- brosis. In anthracobrosis, the main bron- chus and trachea were usually unaffected (three of 43 cases, 7%) (Fig. 3). Endobronchial TB extended contiguously along the ipsilat- eral main bronchus and distal trachea (28 [88%] of 32 cases) (Fig. 4). Discussion The diagnosis of anthracobrosis is based on two major bronchoscopic ndings: bron- chostenosis and anthracotic pigmentation of overlying mucosa. This condition occurs commonly among Asian and black persons and is especially common among older women who present with a chronic cough, sputum, and dyspnea [14]. The pathogenesis of anthracobrosis remains unknown [14]. Two major hypotheses have been suggested. One is related to wood smoke inhalation. Many patients with anthracobrosis are elderly women from rural regions. They are commonly described as homemakers who have handled rewood or soft coals in a closed cooking area. Ciliary movement removes most inhaled anthracotic particles, but residual particles can accumulate at the branching points of the airway. Anthracotic pigmentation itself does not induce focal bronchial narrowing because carbon is inert. However, various changes in the integrity and immunity of the bronchial mucosa can occur in association with the presence of anthracotic pigmentation, and infection and other forms of air pollution can induce brosis [2, 79]. The other hypothesis on the pathogenesis of anthracobrosis is the association of anthracobrosis with endobronchial TB. There is a high incidence (4161%) of co- existence of TB and anthracobrosis, and the CT ndings are similar for the two conditions. There also is evidence of development of anthracotic pigmentation during the course of endobronchial TB [14]. Some authors have reported bronchostenosis with anthraco- brosis in patients responding to anti-TB medication [3, 10]. Endobronchial TB is common among young women, and this majority is a common nding in both endo- bronchial TB and anthracobrosis. According to the second hypothesis, bronchostenosis is thought to be caused by an exaggerated immunologic response to TB antigens in the lymphatic vessels or contiguous lung, to intra- luminal infection from bacilli originating in upstream cavities, or to extrinsic compression from proximal intrathoracic lymph nodes [14]. On the basis of this hypothesis, Chung et al. [1] reported that ant-TB medication was empirically taken by anthracobrosis patients without evidence of active TB, and empiric therapy may be necessary. Our anthracobrosis patients were most commonly women older than 70 years. Co- existent TB was present in 20% of these patients. The rate of coinfection with TB in our study, however, was lower than that re- ported in a previous study [2]. Non tuberculous pneumonia was present in 30% of our anthra- cobrosis patients. The pathologic specimens from patients with anthracobrosis who had undergone lobe ctomy and node dissection ex- hibited only chronic in ammation and anthra- cotic pigmentation without evidence of TB. The CT ndings of anthracobrosis are bronchostenosis and peribronchial lympha- denopathy [14]. Bronchostenosis and lymph- a denopathy, however, are nonspecic ndings and can be manifestations of a number of diseases, both benign and malignant. Among all of the possible causes of bronchostenosis, carcinoma and TB are two major diseases that must be differentiated from anthra- cobrosis. Bronchogenic carci noma usually forms endobronchial nodular protrusion and commonly involves focal areas of a single lobe or segment of bronchus. With these A D Fig. 181-year-old woman with anthracobrosis who presented with cough and sputum. A, Axial CT scan obtained with lung window setting at level of right main pulmonary artery shows smooth luminal narrowing (arrows) at segmental bronchi of right upper lobe. B, Contrast-enhanced CT scan at same level as A shows enlarged nodes around right upper lobe bronchus (black arrows), in subcarinal region (arrowhead), and around left upper lobe bronchus (white arrow). C, Unenhanced CT scan at same level as A and B shows tiny hyperdense foci (arrows) within enlarged nodes. D, Bronchoscopic image shows luminal narrowing of right upper lobe bronchus with focal deposition of black pigmentation (arrows). Pathologic specimen obtained at bronchoscopic biopsy revealed chronic inammation without granuloma. C B D o w n l o a d e d
f r o m
w w w . a j r o n l i n e . o r g
b y
3 6 . 7 2 . 6 5 . 2 0 0
o n
0 1 / 2 6 / 1 4
f r o m
I P
a d d r e s s
3 6 . 7 2 . 6 5 . 2 0 0 .
C o p y r i g h t
A R R S .
F o r
p e r s o n a l
u s e
o n l y ;
a l l
r i g h t s
r e s e r v e d
250 AJR:191, July 2008 Park et al. endobronchial TB, which are major causes of benign bronch ostenosis in Korea. Bronchostenosis was a common CT nding among our patients with anthracobrosis and those with TB. Our study showed that bronch- ostenosis from anthracobrosis could be differentiated from bronchostenosis from TB with CT alone. Bronchostenosis with anthra- cobrosis is multifocal and tends to involve segmental or lobar bronchi in both lungs. The main bronchus and trachea are preserved, and there is no contiguity in extent of disease. On the other hand, bronchostenosis with TB involves a single lobar bronchus in a contiguous spreading pattern along the bronchus. The presence of luminal narrowing of the ipsi lateral main bronchus and distal trachea is common. These differences be tween anthra cobrosis and endobronchial TB may be evidence against the earlier hypo thesis that TB is a major causative factor in anthra cobrosis. The major limitations of our study were small sample size and selection bias. Broncho- genic carcinoma, another major cause of considerations, malignant broncho stenosis can be easily differentiated from anthra- cobrosis [2]. Differentiation, however, is more difcult when endobronchial TB is considered. Because of the similarity be tween the radiologic features of endo bronchial TB and those of anthra cobrosis, anthracobrosis can be dia- gnosed only with bronchoscopy, not with CT, and a relation between the two diseases has been suggested [14]. To our knowledge, there has been no report of a comparison of the differences between anthracobrosis and A Fig. 234-year-old woman with endobronchial tuberculosis who presented with dyspnea. Sputum acid-fast stain result was positive for acid-fast bacilli. A, Axial CT scan obtained at lung window setting shows uneven contiguous luminal narrowing (black arrow) of left main and upper lobe bronchi. Multiple small nodules are clustered with linear branching opacities (white arrows). B, Contrast-enhanced CT scan at same level as A shows focal soft-tissue attenuation around stenotic airway (arrow). C, Bronchoscopic image shows diffuse stenosis of left main bronchus with luminal irregularity. Left main bronchus is hyperemic with purulent secretions. Pathologic specimen obtained at bronchoscopic biopsy revealed chronic inammation and epithelioid granuloma with caseous necrosis. C B A D Fig. 382-year-old woman with anthracobrosis who presented with dyspnea. AC, Axial CT scans obtained at lung window setting (main bronchus level, A; right middle lobe bronchus level, B; right inferior pulmonary vein level, C) show multifocal smooth bronchial narrowing of segmental bronchi of three lobes of right lung and left upper lobe (straight arrows, AC). Both main bronchi are preserved. There is no contiguity in disease extent. Volume loss of right middle lobe (curved arrow, C) is evident. D, Contrast-enhanced CT scan at same level as A shows multiple enlarged nodes around stenotic bronchi and high-attenuation foci (arrowheads) in subcarinal region. C B D o w n l o a d e d
f r o m
w w w . a j r o n l i n e . o r g
b y
3 6 . 7 2 . 6 5 . 2 0 0
o n
0 1 / 2 6 / 1 4
f r o m
I P
a d d r e s s
3 6 . 7 2 . 6 5 . 2 0 0 .
C o p y r i g h t
A R R S .
F o r
p e r s o n a l
u s e
o n l y ;
a l l
r i g h t s
r e s e r v e d
AJR:191, July 2008 251 CT of Anthracofibrosis and Tuberculosis bronchostenosis that should be differentiated from anthracobrosis, was not included in this comparison study, and anthracobrosis patients with a history of pulmonary TB were included. Thus a few patients with old endobronchial TB might have been included in the anthracobrosis group because TB history among anthraco- brosis patients depended only on medical records and the patients memory. In addition, we did not include follow-up data, so we did not know how many cases of anthracobrosis would later prove to be endobronchial TB. Anthracobrosis can be differentiated from endobronchial TB on CT, so CT is a useful technique for the diagnosis of anthra cobrosis before bronchoscopic evaluation. Endobron- chial TB seems not to be a major causative factor in anthracobrosis, and empiric anti-TB medication pending conrm ation of an or- ganism is unnecessary for anthra cobrosis patients. Further study is needed to disclose the pathogenesis of anthra cobrosis. References 1. Chung MP, Lee KS, Han J, et al. Bronchial stenosis due to anthracobrosis. Chest 1998; 113: 344350 2. Kim HY, Im JG, Goo JM, et al. Bronchial anthraco- brosis (inammatory bronchial stenosis with an- thracotic pigmentation): CT ndings. AJR 2000; 174:523527 3. Long R, Wong E, Barrie J. Bronchial anthraco- brosis and tuberculosis: CT features before and after treatment. AJR 2005; 184:[suppl]S33S36 4. Choe HS, Lee IJ, Lee Y. The CT ndings of bron- chial anthracobrosis: comparison of cases with or without active tuberculosis. J Korean Radiol Soc 2004; 50:109114 5. Moon WK, Im JG, Yeon KM, Han MC. Tubercu- losis of the central airways: CT ndings of active and brotic disease. AJR 1997; 169:649653 6. Kim Y, Lee KS, Yoon JH, et al. Tuberculosis of the trachea and main bronchi: CT ndings in 17 patients. AJR 1997; 168:10511056 7. No TM, Kim IS, Kim SW, et al. The clinical inves- tigation for determining the etiology of bronchial anthracobrosis. Korean J Med 2003; 65: 665674 8. Sandoval J, Salas J, Martinez-Guerra ML, et al. Pulmonary arterial hypertension and cor pulmon- ale associated with chronic domestic woodsmoke inhalation. Chest 1993; 103:1220 9. Gold JA, Jagirdar J, Hay JG, Addrizzo-Harris DJ, Naidich DP, Rom WN. Hut lung: a domestically acquired particulate lung disease. Medicine (Bal- timore) 2000; 79:310317 10. Han SH, Cha GY, Lee YM, et al. Study of antitu- berculous medications in anthracobrosis. Tuberc Respir Dis 2001; 50:224231 A Fig. 456-year-old man with history of endobronchial tuberculosis 10 years in past who presented with dyspnea. A, Contrast-enhanced CT scan at level of carina shows smooth luminal narrowing of right main bronchus (straight arrow). Right upper lobe bronchus is completely obstructed, resulting in peripheral atelectasis (curved arrow). B, Contrast-enhanced CT scan at level of aortic arch shows luminal narrowing of distal trachea with evenly thickened tracheal wall (straight arrow). Collapse of right upper lobe (curved arrow) is evident. C, Bronchoscopic image shows bronchostenosis of right main bronchus. Opening of right upper lobe bronchus is obstructed without evidence of mucosal change. C B D o w n l o a d e d