Parenchymal pseudotumoral tuberculosis: Case series and systematic review of literature Ritesh Agarwal , Rajagopala Srinivas, Ashutosh N. Aggarwal Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Sector-12, Chandigarh 160012, India Received 19 April 2007; accepted 19 October 2007 Available online 3 December 2007 KEYWORDS Tuberculosis; Pseudotumoral; Carcinoma lung; Pulmonary mass; Pulmonary nodule Summary Tuberculosis is a rampant infectious pandemic worldwide, with an annual incidence of 8 million active cases and 2 million deaths. Tuberculosis is also among the leading causes of solitary pulmonary nodule in the tropics. Although lymph nodal masses are common in tuberculosis, parenchymal masses are distinctly rare. We report eight immunocompetent adults admitted with large pulmonary masses, and ultimately diagnosed to have parenchymal pseudotumoral tuberculosis. Clinical and radiological resolution followed anti-tubercular therapy alone. We have also conducted a MEDLINE search and summarize the clinical features of all patients with parenchymal pseudotumoral tuberculosis. We also compare our findings with published literature and discuss the possible mechanisms leading to large inflammatory masses, clinical and radiologic clues to the same, common errors in clinical decision making and the therapy of this rare entity. & 2007 Elsevier Ltd. All rights reserved. Introduction Tuberculosis is a global emergency and is the leading cause of deaths due to infectious disease worldwide. The radiological manifestations are easily recognized. Primary tuberculosis presents as consolidation with or without lymphadenopathy. Atelectasis, pleural effusions and miliary pattern are other presenting features of primary tubercu- losis. Post-primary tuberculosis presents with non-homo- genous opacities or cavities in the apical or posterior segment of the right upper lobe or apical segment of lower lobe. Several unusual radiographic features are well described. 1–4 In primary disease, these include lym- phadenopathy without infiltrates; lower lobe infiltrates without adenopathy or effusion; solitary tuberculoma and occurrence of primary disease in patients aged more than 40 years. 1–3 In post-primary disease, non- homogeneous opacities in the lower lobes or anterior segment of upper lobes, mass lesions simulating neoplasms, isolated bronchopleural fistula and normal radiographs are atypical. 2,5 With decreasing disease burden and change in disease epidemiology, these ‘‘unusual manifestations’’ are generally forgotten, and thus may cause diagnostic dilemma. ARTICLE IN PRESS 0954-6111/$ - see front matter & 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2007.10.017 Corresponding author. Tel.: +911722756825; fax: +91 172 2745959. E-mail addresses: [email protected], [email protected](R. Agarwal). Respiratory Medicine (2008) 102, 382–389
Transcript
ARTICLE IN PRESS
Respiratory Medicine (2008) 102, 382–389
0954-6111/$ - see frdoi:10.1016/j.rmed.
�Corresponding aufax: +91 172 2745959
E-mail addresses(R. Agarwal).
Parenchymal pseudotumoral tuberculosis:Case series and systematic review of literature
Ritesh Agarwal�, Rajagopala Srinivas, Ashutosh N. Aggarwal
Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Sector-12,Chandigarh 160012, India
Received 19 April 2007; accepted 19 October 2007Available online 3 December 2007
SummaryTuberculosis is a rampant infectious pandemic worldwide, with an annual incidence of 8million active cases and 2 million deaths. Tuberculosis is also among the leading causes ofsolitary pulmonary nodule in the tropics. Although lymph nodal masses are common intuberculosis, parenchymal masses are distinctly rare. We report eight immunocompetentadults admitted with large pulmonary masses, and ultimately diagnosed to haveparenchymal pseudotumoral tuberculosis. Clinical and radiological resolution followedanti-tubercular therapy alone. We have also conducted a MEDLINE search and summarizethe clinical features of all patients with parenchymal pseudotumoral tuberculosis. We alsocompare our findings with published literature and discuss the possible mechanismsleading to large inflammatory masses, clinical and radiologic clues to the same, commonerrors in clinical decision making and the therapy of this rare entity.& 2007 Elsevier Ltd. All rights reserved.
Introduction
Tuberculosis is a global emergency and is the leading causeof deaths due to infectious disease worldwide. Theradiological manifestations are easily recognized. Primarytuberculosis presents as consolidation with or withoutlymphadenopathy. Atelectasis, pleural effusions and miliarypattern are other presenting features of primary tubercu-losis. Post-primary tuberculosis presents with non-homo-
genous opacities or cavities in the apical or posteriorsegment of the right upper lobe or apical segment oflower lobe. Several unusual radiographic features arewell described.1–4 In primary disease, these include lym-phadenopathy without infiltrates; lower lobe infiltrateswithout adenopathy or effusion; solitary tuberculomaand occurrence of primary disease in patients agedmore than 40 years.1–3 In post-primary disease, non-homogeneous opacities in the lower lobes or anteriorsegment of upper lobes, mass lesions simulating neoplasms,isolated bronchopleural fistula and normal radiographs areatypical.2,5 With decreasing disease burden and changein disease epidemiology, these ‘‘unusual manifestations’’are generally forgotten, and thus may cause diagnosticdilemma.
Although lymph nodal masses are not uncommonlyencountered in tuberculosis, parenchymal masses aredistinctly rare in tuberculosis and have been variablyreported as occurring in primary and post-primary tubercu-losis. Herein, we report eight immunocompetent adults whopresented with large masses and were diagnosed to haveparenchymal post-primary pseudotumoral tuberculosis. Wehave also conducted a MEDLINE search on this rarepresentation and supplemented this by hand search ofliterature and our personal records.
Methods
Case series
The computer records of patients admitted from April 2002to December 2006 to the pulmonary medicine ward of thisinstitute with a diagnosis of tuberculosis were reviewed.Patients with the discharge diagnosis of pseudotumoraltuberculosis, tuberculosis and mass were selected. Theoriginal case records of the patients were then retrievedfrom the central registration department. Demographicinformation such as age and gender, clinical status atadmission were recorded. Details of the clinical manifesta-tions and investigations like computed tomography (CT),bronchoscopy, fine needle aspiration cytology (FNAC), trucutbiopsies were also recorded.
Patients were included if they met the following criteria:(1) one or more mass of at least 3 cm and one of thefollowing four findings on investigation—(a) histopathologyshowing acid-fast bacilli (AFB) and epithelioid granulomas or(b) granulomatous inflammation (AFB negative) and com-plete response to anti-tubercular therapy (ATT) alone or (c)biopsy specimen positive for Mycobacterium tuberculosis onculture or (d) sputum positive for AFB or sputum culturesgrowing M. tuberculosis.
Literature review
Two of the authors (RS and RA) independently conducted aMEDLINE search using the terms tuberculosis and mass,pseudotumoral tuberculosis, tuberculosis and carcinomalung in the English literature between 1964 and December2006 and supplemented this by hand search of literature andpersonal records. Parenchymal pseudotumoral tuberculosiswas defined by the presence of one or more masses of atleast 3 cm. The individual cases were reviewed and availabledata regarding clinical presentation, radiographic findings,management and outcomes were extracted. Lymph nodalpseudotumoral tuberculosis was not included for dataanalysis. Data of the current series was then compared withearlier published literature.
Statistics
The statistical package Stats Direct (Stats Direct version2.6.2 for MS-Windows, England, Stats Direct Ltd., 2005.http://www.statsdirect.com) was used to perform thestatistical analysis. Epidemiological and outcome para-
meters of the patients are presented in a descriptive fashion(mean7SD or median with range).
Results
During the four-and-half-year period, 1286 (822 male and464 female patients) patients were admitted which included234 admissions of patients with tuberculosis. Of the 234admissions with tuberculosis, we identified 8 (3.4%) patientswith histopathologically confirmed diagnosis of parenchymalpseudotumoral tuberculosis (Table 1). The mean (SD) age was36 (713.6) with a range of 20–55 years (Table 2). Theseinclude five males and three female patients (M:F ¼ 5:3). Allpatients were immunocompetent and non-reactive for hu-man immunodeficiency virus by enzyme linked immunoassay;anti-nuclear and anti-neutrophil cytoplasmic antibody(ANCA) were also negative by indirect immunofluorescence.Serum biochemistry, complete blood counts and urinalysiswere normal in all patients. All patients were positive ontuberculin testing (Mantoux test) at onset (median15� 17mm, range 12� 10mm to 25� 18mm). Most werenon-smokers (5/8) and presented after a mean (SD) delay of16 (711.9) weeks from onset of symptoms. Non-productivecough (100%), constitutional symptoms (75%) and dyspnea(50%) were the most common presenting symptoms (Table 1).Fever and hemoptysis were present in two patients only(25%). There were a total of eight masses, six on the rightand two on the left side (R:L ¼ 3:1). Most masses were largeand involved more than one segment (Figures 1–4; theradiology of individual patients is provided as an onlinesupplement file). The anterior segment and the apicalsegment of the right upper lobe were the most common sitefor pulmonary mass (62.5%, 5/8 each), followed by the apicalsegment of right lower lobe (50%, 4/8) and the lingula(Table 1). Centrilobular nodules, often with ‘‘tree-in-bud’’sign, were seen along with the mass on high-resolutioncomputed tomography (37.5%, 3/8). Cavitation was seen in50% (4/8) of the masses. Lymphadenopathy (2/8, 25%) andwas infrequent and calcification was not observed. Sponta-neous pneumothorax was seen in one patient. PercutaneousFNAC (2/5, 40%) and trucut biopsy (2/3, 66%) were usedmore frequently to achieve diagnosis. Bronchoscopy wasabnormal in 75% (6/8) but yielded the diagnosis only in twopatients (2/8, 25%). Sputum was positive for AFB in twopatients (25%). More than one modality was used often fordiagnosis (6/8, 75%). One patient was recognized to have thisrare presentation only after the surgical biopsy specimenshowed granulomatous inflammation with AFB (broncho-scopy, FNAC and trucut biopsy did not yield a diagnosis). Allpatients showed resolution with anti-tubercular drugs (iso-niazid 600mg, rifampicin 600mg, pyrazinamide 1500mg andethambutol 1200mg thrice a week for 2 months followed byisoniazid 600mg and rifampicin 600mg thrice a week for 4months) administered as directly observed therapy under therevised national tuberculosis control program and nonereceived steroids. In three patients, therapy was extendedto 8 months due to clinical and radiological non-resolution.The mean duration of therapy administered was 27 weeks(range 24–32 weeks). Residual fibrosis was observed in threepatients (37.5%, 3/8). All patients were asymptomatic atfollow-up with no significant sequelae.
Abbreviations: AFB: acid-fast bacilli; ATT: anti-tubercular therapy; BALF: bronchoalveolar lavage fluid; EBBx: endobronchial biopsy;FNA: fine needle aspiration; LUL: left upper lobe; N: normal; PCR: polymerase chain reaction; RLL: right lower lobe; RML: right middlelobe; RUL: right upper lobe; �: negative; +: positive.
R. Agarwal et al.384
Review of reported literature3,6–10 showed marked het-erogeneity in the site, size of parenchymal masses anddiagnostic method employed (Table 2). Of the total of 36masses in 35 patients, the right side was preferentiallyinvolved (18:7, N ¼ 25) and the upper lobe was the mostcommon site involved 52.7% (19/36 masses). The mostcommon segment involved in the upper lobe was the rightapical segment. Atelectasis and centrilobular nodules(27.8%, 10/36) were variably described but cavitation,calcification and adenopathy were distinctly rare. Themethod of diagnosis was available for review in 31 cases.Bronchoscopic biopsy and FNAC were the most commondiagnostic modalities but sputum AFB, bronchoalveolarlavage fluid (BALF) AFB (total 29.1%, 9/31 cases) andsurgical biopsy (22.5%, 7/31) were also used. Most patientsresponded to ATT alone and data on steroids or otheradjunctive therapy was not available. Data on follow-up was
lacking (14 cases only), and residual fibrosis developed infive of these (5/14, 35.7%) (Table 2).
Discussion
Tuberculosis is a global emergency and is the leading causeof deaths due to infectious disease worldwide.11 Indiaaccounts for 1.8 million new cases and 370,000 deaths eachyear.12 With such a huge burden, the pre-test probability ofpulmonary tuberculosis for any thoracic lesion is sufficientlyhigh. On the other hand, lung cancer is among the mostcommon adult cancers and is increasing in incidence. Mostpatients in developing countries present with advanceddisease and large masses.13 In fact, the admitting diagnosiswas not tuberculosis in any of our patients and four patientswere admitted with a provisional diagnosis of lung cancer.
ARTICLE IN PRESS
Table 2 Clinical features, investigations and outcome of all patients with parenchymal pseudotumoral tuberculosis reportedin English literature.
Parameter Observedvalue
Cherian et al.7,8 Miller et al.3 Woodring et al.6 Others9,10
Figure 1 Chest radiograph shows a large mass occupyingalmost the entire right hemithorax. Also seen are two air fluidlevels due to secondary pneumothorax which resulted after aultrasound guided fine needle aspiration which resolvedspontaneously.
R. Agarwal et al.386
The distinction of nodules from masses on the basis of sizeis primarily because size is a predictor of malignancy.14 Theaim of evaluation is to recognize malignancy while avoiding
morbidity in the majority. With larger sizes (masses morethan 3 cm), the probability of malignancy is so high thatabsence of smoking, young age, calcification may still notpredict benign etiology. However, tuberculosis may presentwith protean manifestations and has been reported to mimicboth benign and malignant tumors.15
A pseudotumor of the lung is composed of inflammatorycells and showing complete maturity of fibroblastic compo-nents with a striking lack of mitosis. The distinction of thesemasses from tuberculomas has been detailed previously.8
Mechanistically, tubercular pseudotumoral masses may bedue to conglomerate nodes (lymph nodal pseudotumor) orparenchymal granulomatous mass. These are pathogeneti-cally distinct and evolution may depend on whether thedisease is primary or post-primary tuberculosis. The distinc-tion between primary and post-primary disease is oftenarbitrary and is based on age, tuberculin testing andradiological features. Lymph nodal masses denote primaryinfection invariably, in the absence of significant immuno-deficiency. In post-primary tuberculosis, the interplay ofhost immunity and mycobacterial virulence determines theradiological features. Tubercular caseous material drainsinto a bronchus, forms a cavity, reaching a new host.16 In theabsence of bronchial rupture and decompression, enlargingnecrotic areas may form large parenchymal masses. This hasbeen variably reported to occur in progressive primarytuberculosis7,8 and post-primary tuberculosis as well.6
Predominant endobronchial involvement may simulatemalignancy endoscopically and is commonly reported. Largeparenchymal masses are, however, extremely uncommon
ARTICLE IN PRESS
Figure 2 Computed tomography shows a large mass 11� 9 cm with areas of cavitation occupying the right upper and middle lobeand the apical segment of the right lower lobe.
Figure 3 Computed tomography shows a heterogeneously attenuating mass measuring 9� 6 cm occupying the right upper lobeapical and anterior segment.
Parenchymal pseudotumoral tuberculosis 387
and may mimic malignancy clinically, endoscopically andradiologically.8
Data on such parenchymal masses is rare. In several largeseries of patients with post-primary tuberculosis, no mass-like opacities were seen.2 From our search we identified1389 citations out of which 72 were related to tuberculosiscomplicating bronchogenic carcinoma; 13 citations relatedto pseudotumoral forms at other sites17–29 and 18 related topseudotumoral pulmonary tuberculosis. Of the 18 refer-
ences, only 6 citations (including 27 cases) reportedparenchymal pseudotumoral tuberculosis3,6–10 (Tables 1and 2). Thus, presentation of tuberculosis as large parench-ymal masses in healthy immunocompetent adults is dis-tinctly rare.
The location of parenchymal masses has been variablyreported as upper lobe3,6 or lower lobe predominant8
earlier. This disparity is likely to represent masses develop-ing as a complication of progressive primary or post-primary
ARTICLE IN PRESS
Figure 4 Computed tomography lung windows show the same mass as in Figure 3 in the right upper lobe apical and anterior segment(left). Also seen are centrilobular nodules and occasional tree-in-bud pattern (right, arrow).
R. Agarwal et al.388
tuberculosis. The preference of primary tubercular infectionfor the lower lobes is because of greater ventilation therein;impaired lymph drainage and better oxygenation in theupper lobes favor re-activation tuberculosis. In fact, giventhe high rates of infection in the tropics, the mean age ofthe patients and the positive tuberculin testing at onset, theupper lobe predominance of most cases is obvious. While theapical segment of the right upper lobe is the most commonsite, the anterior segment is also involved3,8 and theanterior segment of the upper lobe was involved in 46% ofpost-primary tuberculosis.30 Observed clinical features,including hemoptysis, centrilobular nodules and cavitationcould also vary depending on primary or post-primaryinfection. The presence of centrilobular nodules, thoughuncommon, implies airway spread, suggesting tuberculosisand predicting a greater yield of sputum AFB and broncho-scopy. Absence of AFB in smear, centrilobular nodules onimaging, pulmonary masses, upper respiratory symptoms orabnormal urinalysis should also prompt exclusion of ANCA-related vasculitis, given the reported response of anoccasional case of Wegener’s granulomatosis to anti-tuber-cular therapy alone.31 Data on therapy of pseudotumoraltuberculosis is scant. All our patients showed completeresolution with short-course anti-tubercular chemotherapy.While therapy is often prolonged on clinical groundsempirically in smear-negative and extra-pulmonary tuber-culosis, data on optimum duration of therapy of pseudotu-moral parenchymal tuberculosis is unavailable from thepublished literature because of its rarity. The role ofsteroids and surgery is unclear. In the event of radiologicalnon-resolution, surgical lung biopsy is indicated to rule outassociated bronchogenic carcinoma.32
Heuristics are learned shortcuts that all physicians use tosolve complex clinical problems.33 Despite being quiteuseful, they may contribute to diagnostic errors. Cliniciansmay fall a prey to a common type of cognitive bias—repre-sentativeness heuristics. The probability of carcinoma lungin a middle aged smoker presenting with hemoptysis,anorexia, weight loss and lung mass is high. This bias isstrengthened when fever is absent and radiology issuggestive; an alternative competing diagnosis is thus notconsidered. Availability heuristic states that new events are
easily grouped cognitively with memorable previous experi-ences that are similar in some detail or pattern to thepresent event.33 While it may contribute to recall bias, itmay contribute to evaluation by suggesting an alternativecompeting diagnosis; tuberculosis, in this case. Clinicianscan reduce errors in clinical decision making by balancingcompeting biases.34 Wider appreciation of parenchymalpseudotumoral tuberculosis is necessary to prevent over-evaluation, prevent ‘‘premature closure’’ and administertimely therapy.
The limitations of this study includes the retrospectivenature of the study and the fact that only admitted patientswere evaluated and it may be possible that some cases thathave undergone diagnostic evaluation in the outpatientclinic might have been missed. The strength of this study isthe fact that the index data have been compared with theavailable evidence using a systematic search of theliterature.
Conclusions
In conclusion, the myriad presentations of tuberculosiscontinue to challenge physicians’ diagnostic skills. Parench-ymal pseudotumoral tuberculosis is a rare entity, with about27 cases reported in English literature. Most patientspresent with non-productive cough and constitutionalsymptoms. Diagnosis may be difficult, given the pauci-bacillary nature of the disease and high index of suspicionwith use of multiple diagnostic modalities is often needed.Clinicians can reduce errors in clinical decision making bybalancing competing biases in decision making. Currentmedical literature often contributes to this, by dissemina-tion of knowledge of rarer presentations of common clinicalproblems.
Conflict of interest
The authors declare that they have no conflict of interest inthe manuscript or any financial disclosure.
ARTICLE IN PRESS
Parenchymal pseudotumoral tuberculosis 389
Appendix A. Supplementary materials
Supplementary data associated with this article can befound in the online version at doi:10.1016/j.rmed.2007.10.017.
References
1. Woodring JH, Vandiviere HM, Fried AM, Dillon ML, Williams TD,Melvin IG. Update: the radiologic manifestations of pulmonarytuberculosis. AJR Am J Roentgenol 1986;146:497–506.
2. Hadlock FP, Park SK, Awe RJ, Rivera M. Unusual radiographicfindings in adult pulmonary tuberculosis. AJR Am J Roentgenol1980;134:1015–8.
3. Miller WT, Gregor RRM. Tuberculosis: frequency of unusualradographic findings. AJR Am J Roentgenol 1978;130:867–75.
4. Agarwal R, Kumar V, Jindal SK. Obstructive granulomatousbronchiolitis obliterans due to Mycobacterium tuberculosis.Monaldi Arch Chest Dis 2005;63(2):108–10.
5. Agarwal S, Hong DK, Soslow J, Chang KW. Not your routineforeign body: endobronchial tuberculosis in an infant. Pedia-trics 2005;116(1):246–8.
6. Woodring JH, Vandiviere HM, Fried AM, Dillon ML, Williams TD,Melvin IG. Update: the radiographic features of pulmonarytuberculosis. AJR Am J Roentgenol 1986;146(3):497–506.
7. Cherian MJ, Dahniya MH, Al-Marzouk N, Osmanagich E, Abul A,Haider A. Primary pulmonary tuberculosis presenting as masslesions and simulating tumours in children. Australas Radiol1998;42(4):309–12.
8. Cherian MJ, Dahniya MH, Al-Marzouk NF, Abel A, Bader S, BuerkiK, et al. Pulmonary tuberculosis presenting as mass lesions andsimulating neoplasms in adults. Australas Radiol 1998;42(4):303–8.
12. Revised national tuberculosis control programme—technicaland operational guidelines for tuberculosis control. [2005; cited10.4.2007]. Available from: /http://www.tbcindia.org/pdfs/Technical%20&%20Operational%20guidelines%20for%20TB%20Control.pdfS.
13. Behera D. Managing lung cancer in developing countries:difficulties and solutions. Ind J Chest Dis 2006;48:243–4.
14. Lillington GA, Caskey CI. Evaluation and management of solitarymultiple pulmonary nodules. Clin Chest Med 1993;14:111–9.
15. Saleemi S, Khalid M, Zeitouni M, Al-Dammas S. Tuberculosispresenting as endobronchial tumor. Saudi Med J 2004;25(8):1103–5.
16. Fraser RG, Colman N, Muller NL, Pare PD. Infectious diseases ofthe lungs. Synopsis of diseases of the chest, 3rd ed.Phladelphia: Elsevier Saunders; 2005. p. 249–63.
17. Lesnik A, Bolivar J, Morel J, Taourel P. On the AJR viewbox.Pseudotumoral colonic tuberculosis complicating rheumatoidarthritis treated with a tumor necrosis factor antagonist. AJRAm J Roentgenol 2006;186(6):1799–800.
18. del Agua C, Felipo F, Paricio J, Equizabal C, Delgado M.Tuberculosis of the breast as a pseudotumoral image. BreastJ 2006;12(2):180.
19. Adsay NV, Basturk O, Klimstra DS, Kloppel G. Pancreaticpseudotumors: non-neoplastic solid lesions of the pancreas thatclinically mimic pancreas cancer. Semin Diagn Pathol 2004;21(4):260–7.
20. Culafic D, Boricic I, Vojinovic-Culafic V, Zdrnja M. Hepatictuberculomas. A case report. Rom J Gastroenterol 2005;14(1):71–4.
21. Varela M, Fernandez J, Navasa M, Bruix J. Pseudotumoralhepatic tuberculosis. J Hepatol 2003;39(4):654.
22. Gonzalez Herranz J, Farrington DM, Angulo Gutierrez J,Rodriguez Ferrol P. Peripheral osteoarticular tuberculosis inchildren: tumor-like bone lesions. J Pediatr Orthop B 1997;6(4):274–82.
23. Buxi TB, Vohra RB, Sujatha Y, Chawla D, Byotra SP, Gupta PS,et al. CT appearances in macronodular hepatosplenic tubercu-losis: a review with five additional new cases. Comput MedImaging Graph 1992;16(6):381–7.
24. Achem SR, Kolts BE, Grisnik J, MacMath T, Monteiro CB,Goldstein J. Pseudotumoral hepatic tuberculosis. Atypicalpresentation and comprehensive review of the literature.J Clin Gastroenterol 1992;14(1):72–7.
25. Vuong PN, Houissa-Vuong S, Bleuse B, Schoonaert MF. Pseudo-tumoral tuberculosis of the uterine cervix. Cytologic presenta-tion. Acta Cytol 1989;33(3):305–8.
27. Lesser TH, Jefferis AF, Shaw HJ. Pseudotumoral mycobacterialinfection in the head and neck: a clinical study. J Laryngol Otol1985;99(3):277–83.
28. Laajam MA. Primary tuberculosis of the esophagus: pseudotu-moral presentation. Am J Gastroenterol 1984;79(11):839–41.
29. Rubies-Prat J, Soler-Amigo J, Plans C. Pseudotumoral tubercu-losis of the esophagus. Thorax 1979;34(6):824–5.
30. Woodring JH, Fried AM, Chuang VP. Solitary cavities of the lung:diagnostic implications of cavity wall thickness. AJR Am JRoentgenol 1980;135(6):1269–71.
31. Khilnani GC, Banga A, Sharma SC, Gupta SD. Wegener’sgranulomatosis: an isolated lung mass responding to antituber-culosis therapy and atypical course. J Assoc Physicians India2003;51:731–3.
32. Kim Y, Goo JM, KIm HY, Song JW, Im J. Coexisting bronchogeniccarcinoma and pulmonary tuberculosis in the same lobe:radiologic findings and clinical significance. Korean J Radiol2001(2):138–44.
33. Tversky A, Kahneman D. Judgment under uncertainty: heur-istics and biases. Science 1974;185:1124–31.
34. Kassirer JP, Kopelman RI. Cognitive errors in diagnosis:instantiation, classification and consequences. Am J Med 1989;86:433–41.
