Abstract. The objective was to analyse the potential of CT to distinguish pneumonic-type bronchioloalveolar cell carcinoma (BAC) from infectious pneumonia.
The British Journal of Radiology, 74 (2001), 490–494
E
2001 The British Institute of Radiology
CT differentiation of pneumonic-type bronchioloalveolar cell carcinoma and infectious pneumonia J I JUNG, MD, H KIM, MD, S H PARK, MD, H H KIM, MD, M I AHN, MD, H S KIM, MD, K J KIM, MD, M H CHUNG, MD and B G CHOI, MD Department of Radiology, St Mary’s Hospital, College of Medicine, The Catholic University of Korea, #62 Yeouido-dong, Youngdungpo-gu, Seoul 150-010, Korea
Abstract. The objective was to analyse the potential of CT to distinguish pneumonic-type bronchioloalveolar cell carcinoma (BAC) from infectious pneumonia. The study consisted of 21 patients with pathologically proven BAC and 30 patients with infectious pneumonia. Both groups of patients had patchy or diffuse consolidation of more than half the area of a lobe or lobes on CT. CT findings in these two groups were compared with regard to morphological appearance, including CT angiogram, air bronchogram, mucous bronchogram, contrast enhancement pattern, pseudocavitation, cavity with air–fluid level, location, satellite lesion, ground-glass opacity and bulging of the interlobar fissure. Air-filled bronchi were morphologically analysed as dilatation, stretching, sweeping, widening of the branching angle, squeezing and crowding. Lymphadenopathy and pleural effusion were also analysed. CT findings favouring the diagnosis of BAC included an air-filled bronchus within the consolidation with stretching, squeezing, sweeping, widening of the branching angle and bulging of the interlobar fissure (p,0.05). It is concluded that CT may be helpful in differentiating pneumonic-type BAC from infectious pneumonia if the air-filled bronchus within the consolidation shows stretching, squeezing, widening of the branching angle or bulging of the interlobar fissure.
Bronchioloalveolar cell carcinoma (BAC) is a unique lung neoplasm with variable forms, i.e. single nodular, multifocal and lobar pneumonic types [1]. The pneumonic type of BAC shows lobar consolidation on plain radiography and, as it is often difficult to differentiate from pneumonia, diagnosis is therefore often delayed. CT is helpful in diagnosing pneumonic-type BAC and provides more information regarding carcinoma compared with plain radiography. CT findings of pneumonic-type BAC include low attenuating consolidation and the CT angiogram sign after contrast medium infusion [2]. The bronchus in the area of consolidation is often patent and has the appearance of a leafless tree [3, 4]. However, these findings overlap with those of pneumonia, especially necrotizing pneumonia. We have analysed various CT findings in the consolidative lesion of pneumonic-type BAC and in infectious pneumonia to determine differential points between the diagnoses.
Materials and methods Between 1992 and 1997, we identified 103 patients at our institution who were diagnosed as having BAC of the lung. These patients were Received 7 August 2000 and accepted 22 February 2001. 490
identified from the institution’s tumour registry. CT was available for review in 82 of these patients. CT in 21 patients showed infiltrates of more than half the area of a lobe or lobes and included the pneumonic-type BAC group. In addition, we reviewed 30 consecutive patients with pathologically and clinically proven pneumonia and obvious infiltrates of more than half the area of a lobe or lobes on CT, collected prospectively in 1997. Clinical charts and surgical or pathological findings were reviewed for each case. The 21 patients with pneumonic-type BAC were aged between 29 years and 75 years (mean 41 years); 6 were men and 15 were women. The diagnosis was established by pathological analysis of specimens obtained from transbronchial lung biopsy (n511), fine needle aspiration biopsy (n54) or surgery (n56). The pathological diagnosis was made by observing non-destructive growth of tumour along the alveolar wall. The 30 patients with infectious pneumonia were aged between 18 years and 87 years (mean 59 years); 19 were men and 11 were women. Identified causes of pneumonia were bacteria (n510), fungus (n51) and tuberculosis (n56); specific organisms were not identified in the remaining 13 patients who improved with medical therapy during follow-up. CT images were obtained at 8 or 10 mm The British Journal of Radiology, June 2001
CT of BAC and infectious pneumonia
intervals using a Somatom Plus Scanner (Siemens, Erlangen, Germany) or a GE Quick System (General Electric, Milwaukee, WI). The intravenous contrast medium iopromide (Ultravist 300; Schering, Berlin, Germany) was routinely administered to all patients as a bolus injection followed by rapid drip infusion. CT studies of both groups were reviewed independently by two experienced chest radiologists who were unaware of the final diagnoses. CT images in both groups were evaluated in random order. When interpretations differed, a third radiologist reviewed the cases and the majority opinion was used for the final decision. Interobserver variability among CT readers was assessed using the weighted kappa statistic. Each reviewer evaluated the CT images for evidence of the CT angiogram, air bronchogram, mucous bronchogram, marginal enhancement of the consolidation, pseudocavitation, cavity with air–fluid level, ground-glass opacity, bulging of the interlobar fissure, location of the consolidation and contrast enhancement pattern, i.e. homogeneous or heterogeneous. Air-filled bronchi within the consolidation were morphologically analysed as stretching, squeezing, sweeping, widening of the branching angle, crowding and dilatation. Stretching was defined as rigidity and elongation of the air-filled bronchus. Squeezing was defined as narrowing of the bronchus, sweeping as curving and extension of an unbroken air-filled bronchus, and crowding as gathering of air-filled bronchi. Evidence of lymph node, pleural effusion, satellite lesion and morphology of satellite lesion were evaluated. Statistical qualitative and quantitative differences between the CT findings of the two groups were analysed using the x2 test or Fisher’s exact test.
Results The CT angiogram sign of consolidation was present in 6 of 21 patients with BAC. The attenuation of consolidation was heterogeneous in 15 of 21 patients. An air bronchogram sign was present in 18 of 21 patients. Dilatation of air-filled bronchi within the consolidation was present in 2 of these 18 patients, stretching was present in 16, sweeping was present in 8, widening of the branching angle was present in 8 and squeezing was present in 12 (Figure 1); crowding was not seen. None of the BAC patients had a mucous bronchogram. A bulging fissure abutting consolidation was present in 11 of 21 patients (Figure 2). A pseudocavitation was present in 13 patients and a cavity with an air–fluid level was present in 1 patient. Marginal enhancement of The British Journal of Radiology, June 2001
consolidation was present in six patients (Figure 1). The location of the mass was predominantly central in two patients and peripheral in five. BAC involved both the central and peripheral lung zones in the remaining 14 patients. In addition to consolidation, 10 patients had satellite lesions and 16 had associated groundglass opacities. The satellite lesions showed uniform micronodules in seven patients and a mixed pattern (nodule and ill defined patchy opacity) in three. Two patients had a pattern of ground-glass opacity with thickened interlobular fissures similar to the CT ‘‘crazy paving’’ pattern of pulmonary alveolar proteinosis (Figure 3). Seven patients had enlarged lymph nodes. There were fewer than three enlarged lymph nodes, which measured about 1–2 cm. Pleural effusions were present in five patients. The CT angiogram sign was present in 9 of the 30 patients with pneumonia. The attenuation of the consolidation was heterogeneous in 21 of 30 patients. The location of the consolidation was central in 9 patients, peripheral in 5, and central and peripheral in 24. An air bronchogram was present in 28 of the pneumonia patients. Among these 28 patients, dilatation of air-filled bronchi within the consolidation was present in 13, stretching was present in 14 and sweeping was present in 2. Widening of the branching angle was not seen in pneumonic consolidation, squeezing was present in nine patients and crowding was present in six. A mucous bronchogram was present in 7 of 30 patients, a pseudocavitation in 18, a cavity with air–fluid level in 6, marginal enhancement of the consolidation in 16, bulging of the interlobar fissure in 3, ground-glass opacities in 15 and pleural effusion in 13. A satellite lesion was present in 21 patients, 17 of whom showed a mixed pattern of two or more different morphologies, i.e. nodular, ill defined patchy, linear. Lymph nodes were present in eight patients. There were one or two enlarged lymph nodes, which measured 1–2.1 cm. Statistically significant differences (p,0.05) between the CT findings in patients with BAC and those with pneumonia included a different morphological pattern of air-filled bronchi within the consolidation, such as stretching, sweeping, widening of the branching angle, squeezing and bulging of the interlobar fissure (Table 1). Interobserver variability was low for two CT readers. The kappa statistic ranged from 0.736 to 1.0.
Discussion The consolidative form of BAC accounts for approximately 30% of all BAC tumours and 491
J I Jung, H Kim, S H Park et al
(a)
(b)
Figure 1. (a) CT of a 28-year-old woman with bronchioloalveolar carcinoma shows consolidation occupying the entire right hemithorax. Right upper lobe bronchus shows widening of the angle and sweeping (black arrows). The bronchus is also diffusely narrowed (squeezing). Metastases in the left upper lobe show as rounded nodules of variable size (white arrows). (b) The lower portion of the mass shows low attenuating consolidation with the CT angiogram sign. There is diffuse marginal enhancement of the tumour (arrows).
corresponds to a mucinous histological subtype [5–7]. Unfortunately, the radiographic similarity of consolidative BAC to pneumonia tends to delay the correct diagnosis [8]. Aquino et al [8] reported that CT findings that suggest BAC rather than pneumonia include nonresolving peripheral consolidative pneumonia, especially with associated nodules. Location differences between BAC and pneumonia were not seen in our study. Most of our cases with BAC and pneumonia involved both central and peripheral lung zones. These different results between Aquino et al [8] and our study were owing to the difference of definition of consolidation. Because we define consolidation as infiltrates of more than half of a lobe or lobes, there is more frequent consolidation involving both the central and peripheral lung zones in both the BAC and pneumonia groups.
Figure 2. CT of a 75-year-old man with bronchioloalveolar carcinoma shows low density consolidation with an air bronchogram. There is bulging of the interlobar fissure (arrows). 492
Our study shows that CT findings favouring the diagnosis of BAC rather than pneumonia include stretching, squeezing, sweeping and widening of the branching angle of the air-filled bronchus within the area of consolidation. These bronchial changes of BAC are due to the unique lepidic growth; BAC arises beyond the recognizable bronchus and spreads locally in the peripheral air spaces, using the alveolar walls as its stroma. The tumour seldom obliterates the original
Figure 3. CT of a 65-year-old man with mucinous bronchioloalveolar carcinoma shows diffuse groundglass opacities and consolidation. Note the associated thickened interlobular septa producing a ‘‘crazy paving’’ pattern. The bronchus in the involved lesion is stretched and squeezed (arrows). The British Journal of Radiology, June 2001
CT of BAC and infectious pneumonia Table 1. Summary of the results CT pattern
Number of patients (%) with BAC (n521)
Number of patients (%) with pneumonia (n530)
CT angiogram sign Air bronchogram Dilatation Stretching Sweeping Widening of angle Squeezing Crowding Mucus-filled bronchus Location Central Peripheral Both Enhancement (.20 HU) Marginal enhancement Pleural effusion Pseudocavitation Cavity with air–fluid level Lymph node Satellite lesions Protean pattern of satellite lesion Bulging of fissure
6 18 2 16 8 8 12 0 0
(29) (86) (11) (89)* (44)* (44)* (67)* (0) (0)
9 28 13 14 2 0 9 6 7
(30) (93) (46)* (50) (7) (0) (32) (21)* (23)*
2 5 14 19 6 5 13 1 7 10 3 11
(10) (24) (67) (90) (29) (24) (62) (5) (33) (48) (30) (52)*
1 5 24 30 16 13 18 6 8 21 17 3
(3) (17) (80) (100) (53) (43) (60) (20)* (27) (70) (81)* (10)
BAC, bronchioloalveolar cell carcinoma. *p,0.05.
architecture of the lung, including the bronchus within the tumour. As the tumour fills the alveolar spaces and infiltrates the alveolar septa and bronchial walls, the bronchus becomes narrowed, stretched and rigid [4]. Our study confirmed these findings as a key feature in differentiating BAC from pneumonia. Bulging of the interlobar fissure was present in 11 of 21 BAC patients in our study. This is one of the characteristic findings of BAC and can be caused by mucin production in the tumour, resulting in swelling of the lobe [3]. Satellite lesions within the same lobe or in other lobes were found in BAC and pneumonia at the same frequency in our study. Most of the satellite lesions in BAC showed micronodules, whereas most of those in pneumonia showed a protean pattern in our study. A satellite lesion in consolidative BAC also occurs frequently through bronchogenic or lymphohaematogeneous spread [9]. Morphology of the satellite lesions is relatively uniform and showed micronodules in BAC in our cases as well as those of Aquino et al [8]. Akira et al [10] reported that diffuse BAC frequently showed nodules on high resolution CT. Nodules of diffuse BAC were often distributed in a centrilobular or bronchocentric location. Many centilobular or bronchocentric nodules were found at autopsy to be caused by aerated bronchioli or bronchi surrounded by alveoli filled with mucus and tumour. Satellite lesions in pneumonia are frequent owing to the bronchogenic or direct spread of infection; they show a The British Journal of Radiology, June 2001
poorly defined centrilobular nodular (air space nodule) or lobular consolidation. Because bronchopneumonia involves the airways, it frequently results in atelectasis. In addition, linear fibrosis occurred during recovery from pneumonia [11]. Our results show that other findings reported as key features of BAC, such as multiple cysts or bubble like radiolucencies [12–15], air bronchogram [3, 16] and CT angiogram [2], are not helpful in differentiating BAC from pneumonia, as the rate of occurrence was not significantly different in either group. A cavity with an air–fluid level was more frequent in infectious pneumonia in our study and was present in only one patient with BAC. There was pathologically central necrosis on microscopic examination, while overt cavitation occurred in 7% of BAC patients [17]. Two of our patients with BAC had a CT pattern similar to the crazy paving pattern of pulmonary alveolar proteinosis, as in a case reported by Tan and Kuzo [18] where an air bronchogram with squeezing also provided a clue to the correct diagnosis. Bronchial dilatation and mucous bronchogram were more frequent in infectious pneumonia than in BAC in our study (p,0.05). The air-filled bronchi of chronic pneumonic consolidation usually show some degree of tortousity and ectatic change due to fibrosis and atelectasis of the involved lobe [3]. Interestingly, bronchial dilatation was found in two cases of BAC in our study. Akira et al [10] also reported bronchial 493
J I Jung, H Kim, S H Park et al
dilatation in 4 of 22 cases with diffuse BAC on high resolution CT. In the present study, marginal enhancement of BAC was a unique finding that had not been previously reported. Marginal enhancement of BAC was either visceral pleural enhancement or marginal lobular atelectasis of the remaining lung parenchyma, in contrast to the low attenuating tumour consolidation. However, we saw several patients with necrotizing pneumonia who also had this pattern, which could be explained by the same mechanism as BAC. In conclusion, although CT findings overlap between pneumonic-type BAC and infectious pneumonia, CT signs that favour the diagnosis of BAC include an air-filled bronchus within consolidation with stretching, squeezing, widening of the branching angle and bulging of the interlobar fissure.
Acknowledgment We thank Bonnie Hami (Department of Radiology, University Hospitals of Cleveland, Cleveland, OH) for her editorial assistance in the preparation of the manuscript.
References 1. Akata S, Fukushima A, Kakizaki D, Ase K, Amino S. CT scanning of bronchioloalveolar carcinoma: specific appearances. Lung Cancer 1995;12:221–30. 2. Im J, Han MC, Yu EJ, Han JK, Park JM, Kim C, et al. Lobar bronchioloalveolar carcinoma: ‘‘angiogram sign’’ on CT scans. Radiology 1990;176: 749–53. 3. Im J, Cho BI, Park JH, Chang KH, Yeon KM, Han MC, et al. CT findings of lobar bronchioloalveolar carcinoma. J Comput Assist Tomogr 1986;10:320–2. 4. Zheutlin N, Lasser EC, Rigler LG. Bronchographic abnormalities in alveolar cell carcinoma of the lung. Dis Chest 1954;25:542–9.
494
5. Barsky SH, Grossman DA, Ho J, Holmes EC. The multifocality of bronchioloalveolar lung carcinoma: evidence and implications of a multiclonal origin. Mod Pathol 1994;7:633–40. 6. Manning JT Jr, Spjut HJ, Tschen JA. Bronchioloalveolar carcinoma: the significance of two histopathologic types. Cancer 1984;54:525–34. 7. Metzger RA, Muhern CB Jr, Arger PH, Coleman BG, Epstein DJ, Gefter WB. CT differentiation of solitary from diffuse bronchioloalveolar carcinoma. J Comput Assist Tomogr 1981;5:830–3. 8. Aquino SL, Chiles C, Halford P. Distinction of consolidative bronchioloalveolar carcinoma from pneumonia: do CT criteria work? AJR 1998;171: 359–63. 9. Liebow AA. Bronchiolo-alveolar carcinoma. Adv Intern Med 1960;10:329–58. 10. Akira M, Atagi S, Kawahara M, Iuchi K, Johkoh T. High-resolution CT findings of diffuse bronchioloalveolar carcinoma in 38 patients. AJR 1999; 173:1623–9. 11. Fraser RS, Muller NL, Colman N, Pare´ PD. Pulmonary infection. In: Fraser and Pare´’s diagnosis of diseases of the chest (4th Edn). London: WB Saunders Co, 1999:695–1066. 12. Weisbrod GL, Chamberlain D, Herman SJ. Cystic change (pseudocavitation) associated with bronchioloalveolar carcinoma: a report of four patients. J Thorac Imaging 1995;10:106–11. 13. Adler B, Padley S, Miller RR, Mu¨ller NL. High resolution CT of bronchioloalveolar carcinoma. AJR 1992;159:275–7. 14. Lee KS, Kim Y, Han J, Ko EJ, Park C, Primack SL. Bronchioloalveolar carcinoma: clinical, histopathologic, and radiologic findings. Radiographics 1997;17:1345–57. 15. Ohba S, Takashima T, Hamada S, Kitagawa M. Multiple cystic cavitary alveolar cell carcinoma. Radiology 1972;104:65–6. 16. Hill CA. Bronchioloalveolar carcinoma: a review. Radiology 1984;150:15–20. 17. Wong JS, Weisbrod GL, Chamberlain D, Herman SJ. Bronchioloalveolar carcinoma and the air bronchogram sign: a new pathologic explanation. J Thorac Imaging 1994;9:141–4. 18. Tan RT, Kuzo RS. High-resolution CT findings of mucinous bronchioloalveolar carcinoma: a case of pseudopulmonary alveolar proteinosis. AJR 1997; 168:99–100.
The British Journal of Radiology, June 2001
