Embed Size (px)
Citation preview
lable at ScienceDirect
Respiratory Medicine xxx (2015) 1e5
Contents lists avai
Respiratory Medicine
journal homepage: www.elsevier .com/locate/rmed
Diagnostic value and safety of medical thoracoscopy in tuberculouspleural effusion
Zhen Wang 1, Li-Li Xu 1, Yan-Bing Wu, Xiao-Juan Wang, Yuan Yang, Jun Zhang,Zhao-Hui Tong**, Huan-Zhong Shi*
Department of Respiratory and Critical Care Medicine, Beijing Institute of Respiratory Medicine and Beijing Chaoyang Hospital, Capital Medical University,Beijing, China
a r t i c l e i n f o
Article history:Received 4 March 2015Received in revised form12 May 2015Accepted 15 June 2015Available online xxx
Keywords:Medical thoracoscopyPleural biopsyTuberculous pleural effusion
* Corresponding author. Department of RespiratoryBeijing Chaoyang Hospital, Capital Medical UniversitDistrict, Beijing, 100020, China.** Corresponding author. Department of RespiratorBeijing Chaoyang Hospital, Capital Medical UniversitDistrict, Beijing, 100020, China.
E-mail addresses: [email protected] (Z.-H. To(H.-Z. Shi).
1 These authors contributed equally to the present
http://dx.doi.org/10.1016/j.rmed.2015.06.0080954-6111/© 2015 Published by Elsevier Ltd.
Please cite this article in press as: Z. WanRespiratory Medicine (2015), http://dx.doi.o
a b s t r a c t
Background: Differentiating tuberculous pleural effusion from other lymphocytic pleural effusions isoften challenging. This retrospective study aimed to assess the efficacy and safety of medical thoraco-scopy in patients with suspected tuberculous pleural effusion.Methods: Between July 2005 and June 2014, patients with pleural effusions of unknown etiologies un-derwent medical thoracoscopy in our institute after less invasive means of diagnosis had failed. De-mographic, radiographic, procedural, and histological data of patients with tuberculous pleural effusionwere analyzed.Results: During this 9-year study, 333 of 833 patients with pleural effusion were confirmed to havetuberculous pleurisy. Under thoracoscopy, we observed pleural nodules in 69.4%, pleural adhesion in66.7%, hyperemia in 60.7%, plaque-like lesions in 6.0%, ulceration in 1.5% of patients with tuberculouspleurisy. Pleural biopsy revealed the presence of Mycobacterium tuberculosis in the pleural tissue or/anddemonstration of caseating granulomas in 330 (99.1%) patients. No serious adverse events were recor-ded, and the most common minor complication was transient chest pain (43.2%) from the indwellingchest tube.Conclusions: Our data showed that medical thoracoscopy is a simple procedure with high diagnosticyield and excellent safety for the diagnosis of tuberculous pleural effusion.
© 2015 Published by Elsevier Ltd.
1. Introduction
Tuberculosis accounts for millions of active disease cases anddeaths in both developed and developing countries. Pulmonaryinfection with Mycobacterium tuberculosis is the most commonform of tuberculosis, tuberculous pleural effusion (TPE) remains afrequent form of extrapulmonary tuberculosis [1,2]. In addition,tuberculosis is the most common cause of exudative effusions inareas with a high prevalence of tuberculosis [2,3].
and Critical Care Medicine,y, 8 Gongti Nanlu, Chaoyang
y and Critical Care Medicine,y, 8 Gongti Nanlu, Chaoyang
ng), [email protected]
work.
g, et al., Diagnostic value anrg/10.1016/j.rmed.2015.06.00
TPE can develop from either a primary M. tuberculosis infectionor disease reactivation [2,4]. TPE is curable, therefore, making anaccurate diagnosis as early as possible becomes essential. However,differentiating TPE from the many other causes of lymphocyticpleural effusions is often challenging. Diagnosis of TPE depends onthe demonstration of M. tuberculosis in pleural fluid or pleura tis-sue, or demonstration of epithelioid cell granulomas and/or case-ating granulomas in the pleura [1,5]. Sensitivity of pleural fluidsmear for acid-fast bacilli is very low (0e1%) [5e7], while myco-bacterial culture of pleural fluid and/or pleural biopsy specimens isrelatively sensitive (24e58%) [6e9], but is time consuming andrequires standardized laboratories. Even with advanced culturetechniques, the diagnostic yield is 63% for effusion culture [10].Medical thoracoscopy (MT) refers to the examination of the pleuralspace in a nonintubated patient under local anesthesia, and thisprocedure has been well documented to be highly sensitive andsafe for diagnosing exudative pleural effusions [11e13]. To our
d safety of medical thoracoscopy in tuberculous pleural effusion,8
Z. Wang et al. / Respiratory Medicine xxx (2015) 1e52
knowledge, this study is the first to evaluate usage and safety of MTin the diagnosis of TPE.
2. Methods
2.1. Study population
The study protocol was approved by the Institutional ReviewBoards for human studies of Beijing Chaoyang Hospital, China. Thedetailed medical history, clinical presentation, laboratory exami-nation results, and image data of all patients with undiagnosedexudative pleural effusions who underwent MT in our institutebetween July 2005 and June 2014 were recorded, and only thosedata of patients with definite diagnosis of TPE were included in thecurrent study. Before MT, all patients underwent the initial diag-nostic workup, which includes measurement of a pleural fluidmarker for tuberculosis, including adenosine deaminase, and theirpleural effusions remain undiagnosed. The characteristics of thestudy population are presented in Table 1.
All patients have had a thoracic computed tomography (CT) scanwithin 24 h before undergoing MT. According to the techniquesdescribed by Moy and colleagues, the size of a pleural effusion wasestimated as small, moderate, or large according to CT imagingwithanteroposterior quartile and maximum anteroposterior depthmeasured at the midclavicular line [14]: first anteroposterior-quartile effusions are small, second anteroposterior-quartile effu-sions are moderate, and third or fourth anteroposterior-quartileeffusions are large.
2.2. Thoracoscopic procedures
MT was performed by chest physicians in our pulmonary pro-cedural suite. After written informed consent was obtained, thepatient was positioned with the affected side up in the lateral de-cubitus position. The patient breathed spontaneously with sup-plemental oxygen via nasal cannula as needed. The patient wasconnected to blood pressure, cardiac, and pulse oximetry monitors.Aloka ultrasound system (Aloka Co. Ltd, Tokyo, Japan) was used toconfirm effusion presence and evaluate the best trocar entry site,generally located at the mid-to anterior axillary line, between thesixth to eighth intercostal space.
Under moderate sedation using fentanyl and midazolam, theskin, subcutaneous tissue, adjacent ribs, and parietal pleura wereanesthetized with 1% lidocaine (15e30 mL), and a small incisionwas made at the planned site of entry. The 8-mm disposable trocar
Table 1Characteristics of the study population (n ¼ 333).
Variables Values
Age, yr, mean ± SD 51.8 ± 18.5Sex, male, n (%) 216 (64.9)Smoking status, n (%)Current or ex-smoker 121 (36.3)Never smoker 148 (44.4)Unknown smoking status 64 (19.2)
History of tuberculosis 22 (6.6)Symptoms, n (%)Cough 215(64.6)Breathlessness 202 (60.7)Fever 160 (48.1)Chest pain 119 (35.7)Expectoration 108 (32.4)Weight loss 61 (18.3)Fatigue 34 (10.2)Night sweat 22 (6.6)Hemoptysis 4 (1.2)
Please cite this article in press as: Z. Wang, et al., Diagnostic value aRespiratory Medicine (2015), http://dx.doi.org/10.1016/j.rmed.2015.06.00
was then inserted, and a semi-rigid pleuroscope (Olympus LTF-240,Tokyo, Japan) was introduced into the pleural space right after allfluid was drained completely. A detailed inspection of the pleuralcavity was then performed, with documentation of any abnor-malities by photographic and/or video recordings. Biopsies wereperformed with flexible forceps under direct visual control in allsuspect areas, systematically in several parts of the parietal pleurafor mycobacterial, cytological, and histological examination. Inaddition, biopsy of the parietal pleura was performed over a rib toavoid the neurovascular bundle.
At the end of the procedure, a 24F chest drain was inserted fordrainage. Chest radiographs were routinely obtained after theprocedure and re-check until the removal of the chest tube.
2.3. Diagnostic criteria for TPE
The diagnosis of TPE was established by the presence of M.tuberculosis in biopsy specimen, or by demonstration of caseatinggranulomas or epithelioid cell granulomas in pleural tissue with noevidence of other granulomatous diseases.
2.4. Statistical analysis
Data are presented as mean ± standard deviation (SD) ornumber with percentage. Descriptive statistical methods were usedfor data analysis.
3. Results
Between July 2005 and June 2014, 833 patients with undiag-nosed pleural effusions successfully underwent MT, and pleuralbiopsy samples were obtained for diagnostic evaluation. Eventu-ally, TPE was the final diagnosis in 333 (40.0%) patients with lym-phocytic exudates; the mean age of TPE patients was 51.8 ± 18.5years.
As shown in Table 1, the most common symptoms of TPE werecough (64.6%), breathlessness (60.7%), fever (48.1%), chest pain(35.7%), expectoration (32.4%), and weight loss (18.3%).
In 127 (38.1%) TPE patients, pleural fluid occurred only on theleft side, in 161 (48.4%) only on the right, and in 45 (13.5%) bothsides were affected (Table 2). In either unilateral or bilateral effu-sion, the percentages of small, moderate, and large size of pleuraleffusions were 20.4%, 19.2%, and 60.4%, respectively. Overall, in
Table 2Characteristics of CT findings and pleural effusions (n ¼ 333).
Characteristics n (%)
CT imagingPulmonary consolidation or infiltration 178 (53.5)Pulmonary atelectasis 145 (43.5)Mediastinal lymphopathy 137 (41.1)Pleural thickening 111 (33.3)Pulmonary mass or nodules 73 (21.9)Pleural nodularity 13 (3.9)
Side of effusionRight 161 (48.4)Left 127 (38.1)Bilateral 45 (13.5)
Size of effusionSmall 68 (20.4)Moderate 64 (19.2)Large 201 (60.4)
Effusion appearanceYellow 281 (84.4)Blood-stained 51 (15.3)Chylous 1 (0.3)
nd safety of medical thoracoscopy in tuberculous pleural effusion,8
Table 3Thoracoscopic findings (n ¼ 333).
Variables Value
Pleural fluid removed, mL 1005.9 ± 594.4Parietal pleural biopsies, n 10 ± 1Thoracoscopic findings, n (%)Pleural nodules 231 (69.4)Pleural adhesion 222 (66.7)Hyperemia 202 (60.7)Plaque-like lesions 20 (6.0)Ulcer 5 (1.5)
Z. Wang et al. / Respiratory Medicine xxx (2015) 1e5 3
15.3% of patients, the appearance of pleural effusion was blood-stained, in 84.4% was yellow, and in 0.3% was chylous.
In addition to pleural effusion, CT imaging revealed pulmonaryconsolidation or infiltration (53.5%), pulmonary atelectasis (43.5%),mediastinal lymphopathy (41.1%), pleural thickening (33.3%), pul-monary mass or nodules (21.9%), and pleural nodularity 13 (3.9%)(Table 2).
Under MT, one or more abnormalities could be observed on thesurface of parietal or/and visceral pleura in all patients studied. Asshown in Fig. 1 and Table 3, we observed pleural nodules in 231(69.4%) patients, pleural adhesion in 222 (66.7%), hyperemia in 202(60.7%), plaque-like lesions in 20 (6.0%), and ulcer in 5 (1.5%).Generally, the above pleural abnormalities due to TPE distributeddiffusely on the surface of the involved pleura.
Eventually, MT biopsies yielded tuberculosis pathology in 330(99.1%) patients by demonstration of one or more of followinghistological abnormalities in pleural biopsy specimen: 1) acid-fastbacillus; 2) caseating granulomas; 3) epithelioid cell granulomawith no evidence of other granuloma diseases (Fig. 2 and Table 4).We noted that the growth of M. tuberculosis in pleural fluid wasnegative in all patients, while positive culture results could be seenin pleural biopsy specimen from 3 patients with TPE.
In the 3 patients for whom pleural biopsies were inconclusive, adiagnosis of TPE was secured after isolation of M. tuberculosis insputum cultured during the follow up period; pleural effusionresolution and clinical improvement was seen with anti-tuberculous chemotherapy.
During this 9-year study, no serious adverse events wererecorded in any one patient. As shown in Table 5, among the minoradverse events were 144 (43.2%) patients of local pain requiringadditional analgesic. 9.0% of patients had subcutaneous emphy-sema who recovered after chest-tube drainage. No major bleeding,and minor bleeding was seen in 7.5% of patients. 4.5% of patientshad transient self-limited fever (38 �C ormore). 0.3% of patients had
Fig. 1. Thoracoscopic images taken from patients with tuberculous pleurisy showing A) tubon parietal (upper) and visceral (bottom) pleura; C) multiple military tuberculous nodules opleural hyperemia with necrosis of the pleural node; F) parietal pleural hyperemia with w
Please cite this article in press as: Z. Wang, et al., Diagnostic value anRespiratory Medicine (2015), http://dx.doi.org/10.1016/j.rmed.2015.06.00
empyema caused by methicillin-sensitive Staphylococcus aureuswho recovered after chest-tube drainage and antibiotic treatment.
4. Discussion
Differentiating diagnosis of TPE and the other exudative pleuraleffusions is always difficult, since sometimes theymay have similarclinical manifestations and laboratory characteristics in the absenceof pathological or etiological evidence. Since June 2005, MT hasbeen being performed routinely in our institute for the patientswith exudative pleural effusion not having any diagnosis by eitherclinical, radiologic, laboratory, or cytologic investigation. Duringthis 9-year period, a total of 833 patients with undiagnosed pleuraleffusions underwent MT successfully, and 333 of them wereeventually confirmed to suffer from TPE.
More than 90% of all pleural effusions in the developed coun-tries are caused by congestive heart failure, malignancy, pneu-monia, and pulmonary embolism, while tuberculosis is a commoncause of pleural effusions in the developing countries [2,15]. In2011, there were 8.7 million new cases of active tuberculosisworldwide and 1.4 million deaths [3]. From 1990 to 2010, theprevalence of smear-positive tuberculosis decreased from 170 (95%
erculous nodules with irregular distribution on parietal pleura; B) tuberculous nodulesn parietal pleura; D) diffuse parietal pleura nodules and pleural adhesions; E) parietalhite pleural plaques.
d safety of medical thoracoscopy in tuberculous pleural effusion,8
Fig. 2. Representative tuberculosis pathology seen in pleural biopsy specimen from two patients with tuberculous pleural effusion. A) epithelioid cell granulomas are seen on theparietal pleura; B) a acid-fast bacillus (arrow) is seen within the granulomas; C) caseating necrosis is present in the parietal pleura with scattered multinucleated giant cells; D) aacid-fast bacillus (arrow) is seen within the caseating necrosis. Panels A and C, hematoxylin and eosin staining; original magnification: �200. Panels B and D, Ziehl-Neelsen staining;original magnification: �400.
Table 4Abnormalities under microscopy (n ¼ 333).
Variables Value, n (%)
Caseating granulomasCaseating granulomas 5 (1.5)Caseating granulomas þ Acid-fast bacillus 7 (2.1)
Epithelioid cell granulomaEpithelioid cell granuloma 112 (33.7)Epithelioid cell granuloma þ Acid-fast bacillus 64 (19.2)
Caseating granulomas þ Epithelioid cell granulomaCaseating granulomas þ Epithelioid cell granuloma 68 (20.4)Caseating granulomas þ Epithelioid cell granuloma þ Acid-fast bacillus 74 (22.2)
Table 5Complications of thoracoscopy (n ¼ 333).
Complications n (%)
Paina 144 (43.2)Subcutaneous emphysema 30 (9.0)Minor bleeding 25 (7.5)Fever 15 (4.5)Empyema 1 (0.3)
a Pain requiring analgesics such as ibuprofen.
Z. Wang et al. / Respiratory Medicine xxx (2015) 1e54
confidence interval, 166e174) patients to 59 (49e72) patients per100,000 Chinese population, however, China still had an estimated1 million new tuberculosis cases in 2010, accounting for 11% ofglobal tuberculosis incidence [16]. Moreover, 34.2% (30.9e37.6%) ofthe new cases of tuberculosis and 54.5% (49.6e59.4%) of the pre-viously treated cases were resistant to at least one of the first-lineanti-tuberculosis drugs; 5.7% (4.5e7.0%) of new cases and 25.6%(21.5e29.8%) of previously treated cases were multidrug-resistanttuberculosis [17]. As a matter of fact, tuberculosis continues to beone of the most common causes of mortality and morbidity due to
Please cite this article in press as: Z. Wang, et al., Diagnostic value aRespiratory Medicine (2015), http://dx.doi.org/10.1016/j.rmed.2015.06.00
infectious cause in China. Our recent unpublished data showed thatduring the past 3-year period, the cause of pleural effusions in10.7% (165/1541) of patients admitted to our hospital wastuberculosis.
TPE develops after rupture into the pleural space of a subpleurallung parenchymal caseous focus followed by a delayed hypersen-sitivity reaction to the mycobacterial protein. The mycobacterialload is often low and explains why microscopic examination ofpleural fluid is rarely positive [5e7]. Culture of pleural fluid takesseveral weeks are required to grow M. tuberculosis with very lowsensitivity (24e58%) [6e9]. Where diagnostic difficulty exists, anumber of biomarkers in pleural fluid have been evaluated exten-sively for the diagnostic purpose of TPE, and our meta-analysesdemonstrated that INF-g and adenosine deaminase are two valu-able indicators for diagnosing TPE [18,19]. More recently, we andthe others have shown that IL-27 in pleural fluid is a sensitive andspecific biomarker for the differential diagnosing TPE from pleuraleffusions with the other causes [20e22]. One key criticism of usingthese biomarkers rather than culture or biopsy examinations forthe diagnosis of TPE is that none of the biomarkers provide definiteetiologic diagnosis. It can only be accepted that a definite diagnosis
nd safety of medical thoracoscopy in tuberculous pleural effusion,8
Z. Wang et al. / Respiratory Medicine xxx (2015) 1e5 5
of TPE is achieved when M. tuberculosis is demonstrated in sputumor pleural specimens, or when epithelioid cell granulomas and/orcaseating granulomas are found in pleural biopsies. Therefore,differential diagnosis of TPE sometimes mandates more invasiveprocedures like MT or thoracostomy when one or more thor-acenteses fail to reach definite diagnosis.
In all patients studied, we found one or more abnormalities onthe surface of parietal or/and visceral pleura, and these abnor-malities included pleural nodules, pleural adhesion, hyperemia,pleural plaques, and ulcer. The most important finding of our cur-rent study was that almost all patients with TPE (99.1%) could bediagnosed by performing pleural biopsy under MT. Our data indi-cated that MT is a very efficacious diagnostic procedure for themanagement of TPE in China, a developing country with hightuberculosis prevalence.
It should be mentioned that CT- or ultrasound-guided pleuralbiopsies have been demonstrated to be helpful for the investigationof pleural effusions [23,24]; and it has been suggested the image-guided pleural needle biopsy can be used as the primary methodof diagnosis in patients with pleural thickening identified by CTwhich can be targeted [23,25,26]. However, for those patients withonly pleural fluid appearance on CT and in those who may havebenign pleural pathologies, excluding conditions such as hepatichydrothorax and congestive heart failure, the primary method ofdiagnosis should be MT [25]. As a matter of fact, neither blindneedle biopsies nor image-assisted biopsies have being widespreadfor diagnosing TPE in most hospitals in China.
The fact that MT is a safe procedure in the diagnosis of exudativepleural effusions has been documented in the previous reviews[11,12]. A recent meta-analysis also revealed that there was nomortality with the procedure, and that the rates of major andminorcomplications are 1.5% and 10.5%, respectively [13]. Our currentdata also showed that no serious adverse events were recorded inany one patient in this 9-year study, and that the most commonminor complication was transient chest pain from the indwellingchest tube. Therefore, MT is a very safe procedure in the diagnosisof TPE.
One strengthen of the current study was that a large size ofstudy population was investigated, in which 333 patients with TPEwere included. On the other hand, our study also had limitations.First, this was not a prospective study, all required data werecollected and analyzed retrospectively. Second, few patients in ourseries underwent blind needle biopsies or image-assisted biopsiesbefore undergoing MT. This was an in part explanation why therewere so many patients with pleural effusion (833 cases) undergo-ing MT in our institute during a period of 9 years.
In conclusion, MT is an efficacious procedure in the diagnosis ofTPE. It is a simple method with high diagnostic yield and excellentsafety. In the developing countries (i.e. China) with high tubercu-losis prevalence, MT should be particularly helpful for the man-agement of patients with clinically suspected TPE.
Conflict of interest statement
None of the authors have a financial relationship with a com-mercial entity that has an interest in the subject of this manuscript.
Acknowledgments
The authors are grateful to Department of Pathology at BeijingChaoyang Hospital, Capital Medical University, Beijing, China and in
Please cite this article in press as: Z. Wang, et al., Diagnostic value anRespiratory Medicine (2015), http://dx.doi.org/10.1016/j.rmed.2015.06.00
particular Dr. Mu-Lan Jin and Dr. Xiao-Li Diao for their assistance inevaluating and validating of the histological abnormalities inpleural biopsy specimen of this study. This work was supported inpart by grants from National Natural Science Foundation of China(No. 91442109, No. 31470883, and No. 81270149), and in part bykey project of the department of science and technology(D141107005214003), Beijing, China.
References
[1] A. Gopi, S.M. Madhavan, S.K. Sharma, S.A. Sahn, Diagnosis and treatment oftuberculous pleural effusion in 2006, Chest 131 (2007) 880e889.
[2] R.W. Light, Update on tuberculous pleural effusion, Respirology 15 (2010)451e458.
[3] A. Zumla, M. Raviglione, R. Hafner, C.F. von Reyn, Tuberculosis, N. Engl. J. Med.368 (2013) 745e755.
[4] R. Thomas, Y.C. Lee, Causes and management of common benign pleural ef-fusions, Thorac. Surg. Clin. 23 (2013) 25e42.
[5] A.H. Diacon, B.W. Van de Wal, C. Wyser, et al., Diagnostic tools in tuberculouspleurisy: a direct comparative study, Eur. Respir. J. 22 (2003) 589e591.
[6] C. Escudero Bueno, M. Garcia Clemente, B. Cuesta Castro, et al., Cytologic andbacteriologic analysis of fluid and pleural biopsy specimens with Cope'sneedle. Study of 414 patients, Arch. Intern. Med. 150 (1990) 1190e1194.
[7] L. Valdes, D. Alvarez, E. San Jose, et al., Tuberculous pleurisy: a study of 254patients, Arch. Intern. Med. 158 (1998) 2017e2021.
[8] D.M. Epstein, L.R. Kline, S.M. Albelda, W.T. Miller, Tuberculous pleural effu-sions, Chest 91 (1987) 106e109.
[9] A.F. Seibert, J. Haynes Jr., R. Middleton, J.B. Bass Jr., Tuberculous pleural effu-sion. Twenty-year experience, Chest 99 (1991) 883e886.
[10] S.Y. Ruan, Y.C. Chuang, J.Y. Wang, et al., Revisiting tuberculous pleurisy:pleural fluid characteristics and diagnostic yield of mycobacterial culture in anendemic area, Thorax 67 (2012) 822e827.
[11] R.F. Casal, G.A. Eapen, R.C. Morice, C.A. Jimenez, Medical thoracoscopy, Curr.Opin. Pulm. Med. 15 (2009) 313e320.
[12] A.R. Medford, J.A. Bennett, C.M. Free, S. Agrawal, Current status of medicalpleuroscopy, Clin. Chest Med. 31 (2010) 165e172.
[13] R. Agarwal, A.N. Aggarwal, D. Gupta, Diagnostic accuracy and safety ofsemirigid thoracoscopy in exudative pleural effusions: a meta-analysis, Chest144 (2013) 1857e1867.
[14] M.P. Moy, J.M. Levsky, N.S. Berko, A. Godelman, V.R. Jain, L.B. Haramati, A new,simple method for estimating pleural effusion size on CT scans, Chest 143(2013) 1054e1059.
[15] R.W. Light, Clinical practice. Pleural effusion, N. Engl. J. Med. 346 (2002)1971e1977.
[16] L. Wang, H. Zhang, Y. Ruan, et al., Tuberculosis prevalence in China, 1990-2010; a longitudinal analysis of national survey data, Lancet 383 (2014)2057e2064.
[17] Y. Zhao, S. Xu, L. Wang, et al., National survey of drug-resistant tuberculosis inChina, N. Engl. J. Med. 366 (2012) 2161e2170.
[18] J. Jiang, H.Z. Shi, Q.L. Liang, S.M. Qin, X.J. Qin, Diagnostic value of interferon-gamma in tuberculous pleurisy: a metaanalysis, Chest 131 (2007) 1133e1141.
[19] Q.L. Liang, H.Z. Shi, K. Wang, S.M. Qin, X.J. Qin, Diagnostic accuracy of aden-osine deaminase in tuberculous pleurisy: a meta-analysis, Respir. Med. 102(2008) 744e754.
[20] W.B. Yang, Q.L. Liang, Z.J. Ye, et al., Cell origins and diagnostic accuracy ofinterleukin 27 in pleural effusions, PLoS One 7 (2012) e40450.
[21] Y.B. Wu, Z.J. Ye, S.M. Qin, C. Wu, Y.Q. Chen, H.Z. Shi, Combined detections ofinterleukin 27, interferon-gamma, and adenosine deaminase in pleural effu-sion for diagnosis of tuberculous pleurisy, Chin. Med. J. Engl. 126 (2013)3215e3221.
[22] L. Valdes, E. San Jose, L. Ferreiro, et al., Interleukin 27 could be useful in thediagnosis of tuberculous pleural effusions, Respir. Care 59 (2014) 399e405.
[23] C.F. Koegelenberg, C.T. Bolliger, J. Theron, et al., Direct comparison of thediagnostic yield of ultrasound-assisted Abrams and Tru-Cut needle biopsiesfor pleural tuberculosis, Thorax 65 (2010) 857e862.
[24] C. Hooper, Y.C. Lee, N. Maskell, Investigation of a unilateral pleural effusion inadults: British thoracic society pleural disease guideline 2010, Thorax 65(Suppl 2) (2010) ii4eii17.
[25] M. Metintas, G. Ak, E. Dundar, et al., Medical thoracoscopy vs CT scan-guidedAbrams pleural needle biopsy for diagnosis of patients with pleural effusions:a randomized, controlled trial, Chest 137 (2010) 1362e1368.
[26] N.M. Rahman, N.J. Ali, G. Brown, et al., Local anaesthetic thoracoscopy: BritishThoracic Society Pleural Disease Guideline 2010, Thorax 65 (Suppl 2) (2010)ii54eii60.
d safety of medical thoracoscopy in tuberculous pleural effusion,8
