- Email: [email protected]
Endobronchial Ultrasound Elastography Differentiates Intrathoracic Lymph Nodes: A Meta-Analysis
Endobronchial Ultrasound Elastography Differentiates Intrathoracic Lymph Nodes: A Meta-Analysis Ye-Feng Chen, MD,* Xiao-Wei Mao, MD,* Yu-Jun Zhang, MD, Chun-Yi Zhang, MD, Yue-Fang Yu, MD, E Qin, MD, Xing Chen, MD, and Ju-Xin Shen, MD Department of Pulmonary Medicine, Shaoxing People’s Hospital (Shaoxing Hospital, Zhejiang University School of Medicine), Shaoxing; and Department of Pulmonary Medicine, Shanghai Chest Hospital, Shanghai Jiao Tong University, Shanghai, China
Background. In this study, we investigated the clinical significance of endobronchial ultrasound elastography for differentiating malignant and benign intrathoracic lymph nodes. Methods. A meta-analysis was performed to evaluate the sensitivity and specificity of endobronchial ultrasound elastography in diagnosing intrathoracic lymph nodes. Publications before October 1, 2017, were included for analysis. Sensitivity, specificity, and other variables were pooled using the bivariate mixed-effects regression model. Results. Seven studies met the inclusion criteria and were included. The pooled sensitivity, specificity,
positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio was 0.93 (95% confidence interval [CI], 0.85 to 0.97), 0.85 (95% CI, 0.78 to 0.90), 6.3 (95% CI, 4.2 to 9.2), 0.08 (95% CI, 0.04 to 0.18), and 74 (95% CI, 33 to 168), respectively. The summary receiver operating characteristic curve was 0.93 (95% CI, 0.91 to 0.95). Conclusions. The results revealed endobronchial ultrasound elastography is a new technique with high sensitivity and specificity. It has a fine performance in diagnosing intrathoracic lymph nodes.
T
malignancy, make tissues stiffer, thus increasing the elastic coefficient, and the tissue displays in blue. This characteristic allows elastography to differentiate the malignant disease. In the field of pulmonary disease, endobronchial ultrasound (EBUS) guided-transbronchial needle aspiration (TBNA) provides a good performance, with specificity of 100% and sensitivity of approximately 90% [8]. Although the EBUS-TBNA has an excellent specificity, differentiating malignant diseases from benign diseases is sometimes hard due to the moderate sensitivity. Combining the advantages of elastography and EBUS, the newly developed technique of EBUS elastography can translate the tissue stiffness into a color histogram, thereby allowing differentiation of malignant and benign intrathoracic lymph nodes [9]. Several studies have investigated the clinical significance of EBUS elastography, but the sensitivity and specificity varied in these studies [10, 11]. In the present study, we conducted a meta-analysis to examine whether EBUS elastography is reliable for diagnosing intrathoracic lymph nodes.
he stiffness is a special character of tissue. The correlation between stiffness and tissue condition has been well analyzed in the thyroid, breast, and lymph nodes, among others [1–3]. Malignant tumors are typically stiffer than normal tissue partly due to contraction of collagen in the extracellular matrix of tumor cells [4]. This is the fundamental principle of clinical palpation. Elastography is a newly developing imaging technique that could assess tissue stiffness and visualize the distribution of stiffness in the target lesion. The technique of using tissue stiffness as a novel diagnostic tool began in breast cancer. Now it has been commonly used for the diagnosis of breast lesions, thyroid nodules, and others [5–7]. The technique estimates tissue stiffness through measuring the deformation of tissue in response to stress applied by a senor or heart beat. The bronchoscope sensor then gains the feedback of the target tissue, and the program embedded in the host computer translates the signal into a color-coded image. Tissues with a low elastic coefficient would display in red, yellow, or green, such as the blood vessel and adipose tissue. Pathophysiologic processes, such as
*Drs Ye-Feng Chen and Mao contributed equally to this work. Address correspondence to Dr Shen, Department of Pulmonary Medicine, Shaoxing People’s Hospital, 568 N Zhongxing North St, Shaoxing, Zhejiang, China 312000 (P.C.); email: michaelchan002@163. com.
Ó 2018 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2018;-:-–-) Ó 2018 by The Society of Thoracic Surgeons
The Supplemental Table and Figures can be viewed in the online version of this article [https://doi.org/10. 1016/j.athoracsur.2018.04.003] on http://www.annalstho racicsurgery.org.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.04.003
2
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Patients and Methods Literature Search We performed a literature search in the MEDLINE, Embase, and Web of Science databases and the Cochrane Library by using the key words “endobronchial” and “elastography.” Alternative spellings and abbreviations were also considered. Reference lists of the included studies and relevant reviews were manually examined. The latest publication date was October 1, 2017. Only publications in English were considered. Conference abstracts and letters were excluded.
Inclusion Criteria All potentially relevant studies meeting the following inclusion criteria were retrieved and assessed: (1) the study compared the clinical significance of EBUS elastography with other gold standards; (2) the study included sufficient data (true-positives, false-positives, true-negatives, and false-negatives) for calculating sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and diagnostic odds ratio (DOR); (3) the study enrolled the intrathoracic lymph nodes, and (4) a minimum of 30 lymph nodes were required per study. Case reports or technique introductions were not suitable for this study and were excluded. If several methods were applied in one cohort, only the best result was selected. Disagreements were resolved by discussion.
Data Extraction Data were extracted from all selected studies by 2 independent reviewers working with a standardized format to ensure that all relevant information was captured. Data extracted from each publication were author, publication year, study design (prospective or retrospective), malignant prevalence (0.5 or <0.5), method of analyzing elastography images (semiquantitative or quantitative), total number of patients enrolled, and the distribution of true positives, false positives, true negatives, and false negatives. Some of these items were listed as “not reported” if they were not described in the included studies. No minimum number of patients was required for a study to be included in our meta-analysis. Two reviewers assessed the trial methodology of diagnostic studies according to the quality assessment tool for diagnostic accuracy studies (QUADAS)-2 scoring system [12]. The third reviewer further examined the data and resolved any existing disagreement.
Statistical Analysis The statistical analysis was conducted with STATA 12.0 software (StataCorp, College Station, TX) in the “midas” program. All the analyses were conducted following the standard methods recommended for a meta-analysis of diagnostic test evaluations [13]. For each study, we analyzed five variables: sensitivity, specificity, PLR, NLR, and DOR. The bivariate regression model was used to calculate the pooled sensitivity, specificity, and the other variables [14]. All analysis was based on a summary
Ann Thorac Surg 2018;-:-–-
receiver operating characteristic curve [13, 15]. The likelihood c2 test and I2 statistics were also used to determine statistical heterogeneity across the publications. An I2 value that exceeds 50% is considered to be not statistically heterogeneous across publications. The likelihood ratio c2 test is considered to be statistically heterogeneous when p is less than 0.05. When heterogeneity was detected, the Spearman correlation coefficient was used to determine the threshold effect. The meta-regression and subgroup analysis were further performed to explore other sources of across-study heterogeneity. Covariates included malignant prevalence of lymph nodes (>0.5 or <0.5), study design (prospective, retrospective, or not mentioned), and method (semiquantitative or quantitative). The potential publication bias was estimated by Deeks funnel plots (statistically significant publication bias when p < 0.1) [16].
Results Study Selection Our literature search yielded 36 publications for consideration, 29 of which were excluded for the following reasons: 10 were review articles or comments, two were case reports, three were not written in English, seven were a letter to editor or meeting abstract, and the seven were not specified (Fig 1). The remaining seven publications met all of the inclusion criteria and were included for this meta-analysis [10, 11, 17–21].
Study Descriptions and Quality Assessment The characteristics of the included publications are listed in Table 1 and Supplemental Table 1. A total of 504 patients were enrolled, ranging from 49 to 120 patients per study. Three of the seven publications were prospective studies, three were retrospective studies, and the last study did not report. The malignant prevalence exceeded 0.5 in three publications and was less than 0.5 in the other four. Four publications applied quantitative methods in analyzing elastography image, whereas the other three had used semiquantitative methods. The QUADAS-2 summary plot is presented in Supplemental Figure 1. The pooled sensitivity was 0.93 (95% confidence interval [CI], 0.85 to 0.97), and the specificity was 0.85 (95% CI, 0.78 to 0.90). The results showed a PLR of 6.3 (95% CI, 4.2 to 9.2) and a NLR of 0.08 (95% CI, 0.04 to 0.18). The DOR was 74 (95% CI, 33 to 168), and the area under the curve was 0.93 (95% CI, 0.91 to 0.95). Those results demonstrated that EBUS elastography provided a fine diagnostic performance. The detailed sensitivity and specificity with 95% CI for each study is presented in a forest plot (Figs 2 and 3). No complications were reported in the included publications.
Heterogeneity Assessment and Meta-Regression Analysis The p value for the Spearman correlation coefficient was less than 0.5 (p ¼ 0.49), confirming that the threshold
Ann Thorac Surg 2018;-:-–-
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
3
Fig 1. The flow chart of the study selection.
to 1.00) and studies with quantitative methods (0.93; 95% CI, 0.87 to 1.00). No statistically significant difference in the pooled sensitivity or the pooled specificity was found in different malignant prevalence subgroups. The results of the subgroup analysis are provided in Table 2 and Figure 4.
effect was not significant. The meta-regression analysis revealed that the method of analyzing elastography and the study design were the major source of heterogeneity (p < 0.001; Table 2).
Subgroup Analysis Covariates including malignant prevalence of lymph nodes (>0.5 or <0.5), study design (prospective, retrospective, or not specified), and method (semiquantitative or quantitative) were further analyzed. The subgroups with a prospective design showed a lower specificity (0.80; 95% CI, 0.71 to 0.89) than subgroups with a retrospective design or not specified (0.89; 95% CI, 0.83 to 0.95; p ¼ 0.001). No statistically significant difference (p ¼ 1.00) was found between the pooled sensitivities of the studies with a prospective design (0.95; 95% CI, 0.89 to 1.00) and studies with a retrospective design or not specified (0.91; 95% CI, 0.83 to 1.00). The subgroups with semiquantitative methods had a higher specificity (0.90; 95% CI, 0.83 to 0.96) than subgroups with quantitative methods (0.81; 95% CI, 0.73 to 0.89; p ¼ 0.001). No statistically significant difference (p ¼ 0.74) was found between the pooled sensitivity of the studies with semiquantitative methods (0.92; 95% CI, 0.83
Publication Bias As shown in Supplemental Figure 2, the p value of 0.62 (p > 0.10) suggested there was no publication bias.
Comment Elastography, which is sometimes called “virtual biopsy” [22], is a valuable tool in differentiating malignant and benign disease [23, 24]. The blue, green, and red colors of elastography images correspond to hard, intermediate, and soft tissues, respectively. We can differentiate malignant and benign lesions by analyzing the elastography images [25]. Previous reports have shown that applying elastography alone can obtain good performance in diagnosing some lesions such as thyroid, breast, and gastrointestinal lymph nodes [5–7]. Similar to endoscopic ultrasonography elastography, EBUS elastography
Table 1. Characteristics of the Included Studies
First Author [Reference]
Year
Number of Lymph Nodes
Izumo [11] Nakajima [10] He [17] Rozman [18] Sun [19] Korrungruang [20] Huang [21]
2014 2015 2015 2015 2017 2017 2017
61 49 68 80 68 120 58
NA ¼ not available.
Malignant Prevalence of Lymph Nodes
Design
Quantitative or Not
Method
0.43 0.33 0.62 0.44 0.51 0.80 0.42
Retrospective Retrospective NA Prospective Prospective Prospective Retrospective
Not Quantitative Quantitative Quantitative Not Quantitative Not
3 classifications Stiff area ratio Strain ratio Strain ratio 5 classifications Strain ratio 3 classifications
4
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Ann Thorac Surg 2018;-:-–-
Fig 2. The forest plots of pooled sensitivity and specificity. The solid squares denote the sensitivity/specificity of each study and the horizontal lines represent the 95% confidence intervals (CI). The diamond denotes the combined sensitivity/specificity and the 95% CI.
underwent rapid development in pulmonary medicine. The EBUS elastography was easy to do when performing EBUS-TBNA, only taking a few minutes before TBNA. Varela-Lema and colleagues [26] reported the first study for applying EBUS elastography to the diagnosis of intrathoracic disease. This pioneering study, together with the studies by Andreo and colleagues [27] and He and colleagues [28] that followed, suggested that it was possible to estimate the lesion by analyzing the elastography images. Inage and colleagues [29] successfully used EBUS elastography to determine whether a tumor had invaded the surrounding structures. The diagnostic yield, however, varied from 83.7% to 96.7% in different studies [10, 11]. The clinical significance of EBUS elastography was therefore far from clear before the current study. In the present study, we pooled 7 published studies and analyzed 504 patients. Our meta-analysis revealed that EBUS elastography technique had a high sensitivity (0.93) and specificity (0.85) for diagnosing malignant and benign intrathoracic lymph nodes. In addition, the overall test performance assessed by the area under the curve was quite remarkable (area under summary receiver operating characteristic curve ¼ 0.93). In this meta-analysis, the mean DOR of 74 suggested that EBUS elastography
was reliable in detecting intrathoracic lymph nodes. Likelihood ratios are more clinically meaningful than sensitivity or specificity values. The NLR value was 0.09, which was low enough for clinical purposes. That is to say, if the EBUS elastography showed a negative result of malignant disease, then the lesion was more likely to be benign. However, the value of PLR was 6.3, which was not high enough to exclude the benign lymph nodes when EBUS elastography was positive. This is easy to understand. The fiber formation and tissue calcification in chronic inflammation lead to increased stiffness of lymph nodes [30], which easily leads to the false-positive results of EBUS elastography. That there was a highly significant heterogeneity among the studies analyzed in this study is noteworthy. No threshold effect was observed. Furthermore, we conducted meta-regression analysis to explore the source of the heterogeneity. We found that the method of analyzing elastography and the study design were the major sources of heterogeneity. In the subgroup analysis, studies with retrospective design had a higher specificity than studies with prospective design. Also, the studies with quantitative method had a less specificity than studies with semiquantitative method. The other subgroups
Ann Thorac Surg 2018;-:-–-
5
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Fig 3. (A) The summary receiver operator characteristic (SROC) curve of the included studies. (B) The likelihood ratio (LR) of the combined results. (AUC ¼ area under the curve; SENS ¼ sensitivity; SPEC ¼ specificity.)
(malignant prevalence) did not show significantly different sensitivities or specificities. Izumo and colleagues [11] were the first to conduct a systematic analysis for the elastography images. They classified the lymph nodes elastography images into three types and obtained a fine diagnostic yield. However, the semiquantitative methods largely relied on operators’ objective judgments and might cause interobserver disagreement [31]. Huang and colleagues [21] and Korrungruang and colleagues [20] adopted the similar classification method in their studies. However, He [17] and Sun [19] and their colleagues chose four or five types for classification [17, 19]. In contrast to the semiquantitative methods, quantitative methods mostly relied on computer-aided calculation, and thereby overcame the bias of artificial judgments. The quantitative methods, however, did not seem to be stable. For example, the strain ratio was chosen in the studies of He [17], Rozman [18], and Korrungruang [20] and their colleagues to describe the elastography images. Although their results were similar, the cutoff values for
these three studies were 32.07, 8, and 2.5, respectively. Sun and colleagues [19] selected a new method, the “mean grey value,” to describe the stiffness of target lesions. A new algorithm in which the image was first quantized and then clustered in the quantized color space in prior segmentation part was introduced by Nakajima and colleagues [10]. Whether the semiquantitative method or quantitative method is a better approach is still not clear. In the study by He and colleagues [17], the quantitative method of strain ratio showed a better diagnostic yield than the semiquantitative method of the elastography score. However, in the studies by Sun [19] and Huang [21] and their colleagues, the elastography score showed better performance than the quantitative method of mean grey value. In our meta-analysis, the semiquantitative method showed a good specificity compared with the quantitative method, and there was no significant difference for the sensibility between those subgroups. Subjective categorizations of B-mode images are useful for evaluating the presence of malignant lymph nodes.
Table 2. Subgroup Analysis Subgroup Malignant prevalence >0.5 <0.5 Design Prospective Retrospective or NA Method Semiquantitative Quantitative CI ¼ confidence interval;
Studies, No.
Summary Sensitivity (95% CI)
3 4
0.96 (0.91–1.00) 0.89 (0.80–0.98)
3 4
0.95 (0.89–1.00) 0.91 (0.83–1.00)
3 4
0.92 (0.83–1.00) 0.93 (0.87–1.00)
p
Summary Specificity (95% CI)
0.77
0.11 0.86 (0.77–0.95) 0.85 (0.77–0.92)
1.00
0.001 0.80 (0.71–0.89) 0.89 (0.83–0.95)
0.74
NA ¼ not available;
No. ¼ number.
p
0.001 0.90 (0.83–0.96) 0.81 (0.73–0.89)
6
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Ann Thorac Surg 2018;-:-–-
technique, has been widely applied in clinical practice. Some researcher regarded it as being more stable than strain ratio. Up to now, however, the shear wave sensor has not been embedded in the EBUS machine. We hope the technique of EBUS combined with shear wave could be a reality in the near future. Some limitations should not be neglected. First, all of the studies included were single-center and small pilot were adopted (range, 49 to 120), and four of these were retrospective studies. Selection bias could not be avoided. In addition, all of the studies except one were conducted at centers in East Asia. The granulomatous diseases (sarcoidosis, tuberculosis, etc) that are common in Asian countries are relatedly rare in developed countries. Although no publication bias was observed in our study, publications with positive or significant results were more likely to be published; therefore, a certain degree of publication bias was inevitable. Our results suggest that the EBUS elastography is a reliable technique for diagnosing intrathoracic lymph nodes with a high sensitivity and specificity and could be applied more widely in diagnosing pulmonary disease. The authors wish to thank Guoxiang Ruan for his reasonable suggestions for this paper. Fig 4. The subgroup analysis of the study. (CI ¼ confidence interval.)
These features include size exceeding 1 cm, circular shape, heterogeneity, lack of central hilar structure, and lack absence of coagulation necrosis, among others [32]. However, none of those features seems to be as reliable as the EBUS elastography according to those studies [10, 11, 17, 18]. Besides, proper training and experience were required before the B-mode images could be distinguished. EBUS elastography has fine sensitivity and diagnostic yield in diagnosing intrathoracic lymph nodes [8]. But that is not to say that a biopsy specimen can be replaced by elastography. Different from the lymph nodes of the digestive tract, intrathoracic lymph nodes register more benign diseases, such as tuberculosis and sarcoidosis, among others. EBUS elastography distinguishes the malignant and benign lesion through the stiffness of the tissue [33, 34]. EBUS elastography was good at NLR, but not for PLR, which easily led to the false-positive results of EBUS elastography. That is to say, PLR was not high enough to exclude the benign lymph nodes when EBUS elastography was positive. However, EBUS-TBNA is a fine technique with high PLR, which may compensate the shortcomings of EBUS elastography [35, 36]. In addition, the metastatic area was localized within the lymph node, which may lead to a false-negative result of EBUS-TBNA in some cases. However, the operator can visualize the suspicious area and control the puncture area to improve the diagnostic value of TBNA [10]. In addition, all of the publications included in this study were based on the EBUS real-time elastography technique. Shear wave, another finer elastography
References 1. Lyshchik A, Higashi T, Asato R, et al. Elastic moduli of thyroid tissues under compression. Ultrason Imaging 2005;27: 101–10. 2. Samani A, Zubovits J, Plewes D. Elastic moduli of normal and pathological human breast tissues: an inversiontechnique-based investigation of 169 samples. Phys Med Biol 2007;52:1565–76. 3. Bhatia KS, Cho CC, Tong CS, Yuen EH, Ahuja AT. Shear wave elasticity imaging of cervical lymph nodes. Ultrasound Med Biol 2012;38:195–201. 4. Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic moduli of breast and prostate tissues under compression. Ultrason Imaging 1998;20:260–74. 5. Bojunga J, Herrmann E, Meyer G, Weber S, Zeuzem S, Friedrich-Rust M. Real-time elastography for the differentiation of benign and malignant thyroid nodules: a metaanalysis. Thyroid 2010;20:1145–50. 6. Gong X, Xu Q, Xu Z, Xiong P, Yan W, Chen Y. Real-time elastography for the differentiation of benign and malignant breast lesions: a meta-analysis. Breast Cancer Res Treat 2011;130:11–8. 7. Xu W, Shi J, Zeng X, et al. EUS elastography for the differentiation of benign and malignant lymph nodes: a metaanalysis. Gasterointest Endosc 2011;74:1001–9; quiz 1115.e1-4. 8. Gu P, Zhao Y, Jiang L, Zhang W, Xin Y, Han B. Endobronchial ultrasound-guided transbronchial needle aspiration for staging of lung cancer: a systematic review and meta-analysis. Eur J Cancer 2009;45:1389–96. 9. Ophir J, Cespedes I, Ponnekanti H, Yazdi Y, Li X. Elastography: a quantitative method for imaging the elasticity of biological tissues. Ultrason Imaging 1991;13:111–34. 10. Nakajima T, Inage T, Sata Y, et al. Elastography for predicting and localizing nodal metastases during endobronchial ultrasound. Respiration 2015;90:499–506. 11. Izumo T, Sasada S, Chavez C, Matsumoto Y, Tsuchida T. Endobronchial ultrasound elastography in the diagnosis of
Ann Thorac Surg 2018;-:-–-
12.
13. 14.
15.
16.
17.
18.
19. 20.
21.
22.
mediastinal and hilar lymph nodes. Jpn J Clin Oncol 2014;44: 956–62. Whiting PF, Weswood ME, Rutjes AW, Reitsma JB, Bossuyt PN, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Method 2006;6:9. Deville WL, Buntinx F, Bouter LM, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Method 2002;2:9. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982–90. Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a summary ROC curve: dataanalytic approaches and some additional considerations. Stat Med 1993;12:1293–316. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 2005;58:882–93. He HY, Huang M, Zhu J, Ma H, Lu XD. Endobronchial ultrasound elastography for diagnosing mediastinal and hilar lymph nodes. Chin Med J (Engl) 2015;128: 2720–5. Rozman A, Malovrh MM, Adamic K, Subic T, Kovac V, Flezar M. Endobronchial ultrasound elastography strain ratio for mediastinal lymph node diagnosis. Radiol Oncol 2015;49: 334–40. Sun J, Zheng X, Mao X, et al. Endobronchial ultrasound elastography for evaluation of intrathoracic lymph nodes: a pilot study. Respiration 2017;93:327–38. Korrungruang P, Boonsarngsuk V. Diagnostic value of endobronchial ultrasound elastography for the differentiation of benign and malignant intrathoracic lymph nodes. Respirology 2017;22:972–7. Huang H, Huang Z, Wang Q, et al. Effectiveness of the benign and malignant diagnosis of mediastinal and hilar lymph nodes by endobronchial ultrasound elastography. J Cancer 2017;8:1843–8. Giovannini M, Hookey LC, Bories E, Pesenti C, Monges G, Delpero JR. Endoscopic ultrasound elastography: the first step towards virtual biopsy? Preliminary results in 49 patients. Endoscopy 2006;38:344–8.
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
7
23. Dietrich CF, Jenssen C, Herth FJ. Endobronchial ultrasound elastography. Endosc Ultrasound 2016;5:233–8. 24. Jiang JH, Turner JJ, Huang JA. Endobronchial ultrasound elastography: a new method in endobronchial ultrasoundguided transbronchial needle aspiration. J Thorac Dis 2015;7(Suppl 4):S272–8. 25. Saftoiu A, Vilman P. Endoscopic ultrasound elastography–a new imaging technique for the visualization of tissue elasticity distribution. J Gastrointestin Liver Dis 2006;15:161–5. 26. Varela-Lema L, Fernandez-Villar A, Ruano-Ravina A. Effectiveness and safety of endobronchial ultrasoundtransbronchial needle aspiration: a systematic review. Eur Respir J 2009;33:1156–64. 27. Andreo GF, Centeno CC, Sanz SJ, Barturen BA, Hernandez GA, Ruiz MJ. Initial experience with real-time elastography using an ultrasound bronchoscope for the evaluation of mediastinal lymph nodes. Arch Bronconrumol 2015;5:e8–11. 28. He HY, Chen JL, Ma H, Zhu J, Wu DD, Lv XD. Value of endobronchial ultrasound elastography in diagnosis of central lung lesions. Med Sci Monit 2017;23:3269–75. 29. Inage T, Nakajima T, Yoshida S, Yoshino I. Endobronchial elastography in the evaluation of esophageal invasion. J Thorac Cardiovasc Surg 2015;149:576–7. 30. Coussens LM, Werb Z. Inflammation and cancer. Nature 2002;420:860–7. 31. Calvete AC, Rodriguez JM, de Dios BJ, Rios A, AbellanRivero D, Reus M. Interobserver agreement for thyroid elastography: value of the quality factor. J Ultrasound Med 2013;32:495–504. 32. Wang L, Wu W, Hu Y, Teng J, Zhong R, Han B, Sun J. Sonographic features of endobronchial ultrasonography predict intrathoracic lymph node metastasis in lung cancer patients. Ann Thorac Surg 2015;100:1203–9. 33. Mittal S, Mohan A, Madan K. Endobronchial ultrasound elastography for the differentiation of benign and malignant lymph nodes. Respirology 2017;22:1037–8. 34. Steinfort DP. Elastography in endobronchial ultrasound: Stretching the boundary of minimally invasive staging further. Respirology 2017;22:843–4. 35. Sheski FD, Mathur PN. Endobronchial ultrasound. Chest 2008;133:264–70. 36. Belanger AR, Akulian JA. An update on the role of advanced diagnostic bronchoscopy in the evaluation and staging of lung cancer. Ther Adv Respir Dis 2017;11:211–21.
Background. In this study, we investigated the clinical significance of endobronchial ultrasound elastography for differentiating malignant and benign intrathoracic lymph nodes. Methods. A meta-analysis was performed to evaluate the sensitivity and specificity of endobronchial ultrasound elastography in diagnosing intrathoracic lymph nodes. Publications before October 1, 2017, were included for analysis. Sensitivity, specificity, and other variables were pooled using the bivariate mixed-effects regression model. Results. Seven studies met the inclusion criteria and were included. The pooled sensitivity, specificity,
positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio was 0.93 (95% confidence interval [CI], 0.85 to 0.97), 0.85 (95% CI, 0.78 to 0.90), 6.3 (95% CI, 4.2 to 9.2), 0.08 (95% CI, 0.04 to 0.18), and 74 (95% CI, 33 to 168), respectively. The summary receiver operating characteristic curve was 0.93 (95% CI, 0.91 to 0.95). Conclusions. The results revealed endobronchial ultrasound elastography is a new technique with high sensitivity and specificity. It has a fine performance in diagnosing intrathoracic lymph nodes.
T
malignancy, make tissues stiffer, thus increasing the elastic coefficient, and the tissue displays in blue. This characteristic allows elastography to differentiate the malignant disease. In the field of pulmonary disease, endobronchial ultrasound (EBUS) guided-transbronchial needle aspiration (TBNA) provides a good performance, with specificity of 100% and sensitivity of approximately 90% [8]. Although the EBUS-TBNA has an excellent specificity, differentiating malignant diseases from benign diseases is sometimes hard due to the moderate sensitivity. Combining the advantages of elastography and EBUS, the newly developed technique of EBUS elastography can translate the tissue stiffness into a color histogram, thereby allowing differentiation of malignant and benign intrathoracic lymph nodes [9]. Several studies have investigated the clinical significance of EBUS elastography, but the sensitivity and specificity varied in these studies [10, 11]. In the present study, we conducted a meta-analysis to examine whether EBUS elastography is reliable for diagnosing intrathoracic lymph nodes.
he stiffness is a special character of tissue. The correlation between stiffness and tissue condition has been well analyzed in the thyroid, breast, and lymph nodes, among others [1–3]. Malignant tumors are typically stiffer than normal tissue partly due to contraction of collagen in the extracellular matrix of tumor cells [4]. This is the fundamental principle of clinical palpation. Elastography is a newly developing imaging technique that could assess tissue stiffness and visualize the distribution of stiffness in the target lesion. The technique of using tissue stiffness as a novel diagnostic tool began in breast cancer. Now it has been commonly used for the diagnosis of breast lesions, thyroid nodules, and others [5–7]. The technique estimates tissue stiffness through measuring the deformation of tissue in response to stress applied by a senor or heart beat. The bronchoscope sensor then gains the feedback of the target tissue, and the program embedded in the host computer translates the signal into a color-coded image. Tissues with a low elastic coefficient would display in red, yellow, or green, such as the blood vessel and adipose tissue. Pathophysiologic processes, such as
*Drs Ye-Feng Chen and Mao contributed equally to this work. Address correspondence to Dr Shen, Department of Pulmonary Medicine, Shaoxing People’s Hospital, 568 N Zhongxing North St, Shaoxing, Zhejiang, China 312000 (P.C.); email: michaelchan002@163. com.
Ó 2018 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2018;-:-–-) Ó 2018 by The Society of Thoracic Surgeons
The Supplemental Table and Figures can be viewed in the online version of this article [https://doi.org/10. 1016/j.athoracsur.2018.04.003] on http://www.annalstho racicsurgery.org.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.04.003
2
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Patients and Methods Literature Search We performed a literature search in the MEDLINE, Embase, and Web of Science databases and the Cochrane Library by using the key words “endobronchial” and “elastography.” Alternative spellings and abbreviations were also considered. Reference lists of the included studies and relevant reviews were manually examined. The latest publication date was October 1, 2017. Only publications in English were considered. Conference abstracts and letters were excluded.
Inclusion Criteria All potentially relevant studies meeting the following inclusion criteria were retrieved and assessed: (1) the study compared the clinical significance of EBUS elastography with other gold standards; (2) the study included sufficient data (true-positives, false-positives, true-negatives, and false-negatives) for calculating sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and diagnostic odds ratio (DOR); (3) the study enrolled the intrathoracic lymph nodes, and (4) a minimum of 30 lymph nodes were required per study. Case reports or technique introductions were not suitable for this study and were excluded. If several methods were applied in one cohort, only the best result was selected. Disagreements were resolved by discussion.
Data Extraction Data were extracted from all selected studies by 2 independent reviewers working with a standardized format to ensure that all relevant information was captured. Data extracted from each publication were author, publication year, study design (prospective or retrospective), malignant prevalence (0.5 or <0.5), method of analyzing elastography images (semiquantitative or quantitative), total number of patients enrolled, and the distribution of true positives, false positives, true negatives, and false negatives. Some of these items were listed as “not reported” if they were not described in the included studies. No minimum number of patients was required for a study to be included in our meta-analysis. Two reviewers assessed the trial methodology of diagnostic studies according to the quality assessment tool for diagnostic accuracy studies (QUADAS)-2 scoring system [12]. The third reviewer further examined the data and resolved any existing disagreement.
Statistical Analysis The statistical analysis was conducted with STATA 12.0 software (StataCorp, College Station, TX) in the “midas” program. All the analyses were conducted following the standard methods recommended for a meta-analysis of diagnostic test evaluations [13]. For each study, we analyzed five variables: sensitivity, specificity, PLR, NLR, and DOR. The bivariate regression model was used to calculate the pooled sensitivity, specificity, and the other variables [14]. All analysis was based on a summary
Ann Thorac Surg 2018;-:-–-
receiver operating characteristic curve [13, 15]. The likelihood c2 test and I2 statistics were also used to determine statistical heterogeneity across the publications. An I2 value that exceeds 50% is considered to be not statistically heterogeneous across publications. The likelihood ratio c2 test is considered to be statistically heterogeneous when p is less than 0.05. When heterogeneity was detected, the Spearman correlation coefficient was used to determine the threshold effect. The meta-regression and subgroup analysis were further performed to explore other sources of across-study heterogeneity. Covariates included malignant prevalence of lymph nodes (>0.5 or <0.5), study design (prospective, retrospective, or not mentioned), and method (semiquantitative or quantitative). The potential publication bias was estimated by Deeks funnel plots (statistically significant publication bias when p < 0.1) [16].
Results Study Selection Our literature search yielded 36 publications for consideration, 29 of which were excluded for the following reasons: 10 were review articles or comments, two were case reports, three were not written in English, seven were a letter to editor or meeting abstract, and the seven were not specified (Fig 1). The remaining seven publications met all of the inclusion criteria and were included for this meta-analysis [10, 11, 17–21].
Study Descriptions and Quality Assessment The characteristics of the included publications are listed in Table 1 and Supplemental Table 1. A total of 504 patients were enrolled, ranging from 49 to 120 patients per study. Three of the seven publications were prospective studies, three were retrospective studies, and the last study did not report. The malignant prevalence exceeded 0.5 in three publications and was less than 0.5 in the other four. Four publications applied quantitative methods in analyzing elastography image, whereas the other three had used semiquantitative methods. The QUADAS-2 summary plot is presented in Supplemental Figure 1. The pooled sensitivity was 0.93 (95% confidence interval [CI], 0.85 to 0.97), and the specificity was 0.85 (95% CI, 0.78 to 0.90). The results showed a PLR of 6.3 (95% CI, 4.2 to 9.2) and a NLR of 0.08 (95% CI, 0.04 to 0.18). The DOR was 74 (95% CI, 33 to 168), and the area under the curve was 0.93 (95% CI, 0.91 to 0.95). Those results demonstrated that EBUS elastography provided a fine diagnostic performance. The detailed sensitivity and specificity with 95% CI for each study is presented in a forest plot (Figs 2 and 3). No complications were reported in the included publications.
Heterogeneity Assessment and Meta-Regression Analysis The p value for the Spearman correlation coefficient was less than 0.5 (p ¼ 0.49), confirming that the threshold
Ann Thorac Surg 2018;-:-–-
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
3
Fig 1. The flow chart of the study selection.
to 1.00) and studies with quantitative methods (0.93; 95% CI, 0.87 to 1.00). No statistically significant difference in the pooled sensitivity or the pooled specificity was found in different malignant prevalence subgroups. The results of the subgroup analysis are provided in Table 2 and Figure 4.
effect was not significant. The meta-regression analysis revealed that the method of analyzing elastography and the study design were the major source of heterogeneity (p < 0.001; Table 2).
Subgroup Analysis Covariates including malignant prevalence of lymph nodes (>0.5 or <0.5), study design (prospective, retrospective, or not specified), and method (semiquantitative or quantitative) were further analyzed. The subgroups with a prospective design showed a lower specificity (0.80; 95% CI, 0.71 to 0.89) than subgroups with a retrospective design or not specified (0.89; 95% CI, 0.83 to 0.95; p ¼ 0.001). No statistically significant difference (p ¼ 1.00) was found between the pooled sensitivities of the studies with a prospective design (0.95; 95% CI, 0.89 to 1.00) and studies with a retrospective design or not specified (0.91; 95% CI, 0.83 to 1.00). The subgroups with semiquantitative methods had a higher specificity (0.90; 95% CI, 0.83 to 0.96) than subgroups with quantitative methods (0.81; 95% CI, 0.73 to 0.89; p ¼ 0.001). No statistically significant difference (p ¼ 0.74) was found between the pooled sensitivity of the studies with semiquantitative methods (0.92; 95% CI, 0.83
Publication Bias As shown in Supplemental Figure 2, the p value of 0.62 (p > 0.10) suggested there was no publication bias.
Comment Elastography, which is sometimes called “virtual biopsy” [22], is a valuable tool in differentiating malignant and benign disease [23, 24]. The blue, green, and red colors of elastography images correspond to hard, intermediate, and soft tissues, respectively. We can differentiate malignant and benign lesions by analyzing the elastography images [25]. Previous reports have shown that applying elastography alone can obtain good performance in diagnosing some lesions such as thyroid, breast, and gastrointestinal lymph nodes [5–7]. Similar to endoscopic ultrasonography elastography, EBUS elastography
Table 1. Characteristics of the Included Studies
First Author [Reference]
Year
Number of Lymph Nodes
Izumo [11] Nakajima [10] He [17] Rozman [18] Sun [19] Korrungruang [20] Huang [21]
2014 2015 2015 2015 2017 2017 2017
61 49 68 80 68 120 58
NA ¼ not available.
Malignant Prevalence of Lymph Nodes
Design
Quantitative or Not
Method
0.43 0.33 0.62 0.44 0.51 0.80 0.42
Retrospective Retrospective NA Prospective Prospective Prospective Retrospective
Not Quantitative Quantitative Quantitative Not Quantitative Not
3 classifications Stiff area ratio Strain ratio Strain ratio 5 classifications Strain ratio 3 classifications
4
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Ann Thorac Surg 2018;-:-–-
Fig 2. The forest plots of pooled sensitivity and specificity. The solid squares denote the sensitivity/specificity of each study and the horizontal lines represent the 95% confidence intervals (CI). The diamond denotes the combined sensitivity/specificity and the 95% CI.
underwent rapid development in pulmonary medicine. The EBUS elastography was easy to do when performing EBUS-TBNA, only taking a few minutes before TBNA. Varela-Lema and colleagues [26] reported the first study for applying EBUS elastography to the diagnosis of intrathoracic disease. This pioneering study, together with the studies by Andreo and colleagues [27] and He and colleagues [28] that followed, suggested that it was possible to estimate the lesion by analyzing the elastography images. Inage and colleagues [29] successfully used EBUS elastography to determine whether a tumor had invaded the surrounding structures. The diagnostic yield, however, varied from 83.7% to 96.7% in different studies [10, 11]. The clinical significance of EBUS elastography was therefore far from clear before the current study. In the present study, we pooled 7 published studies and analyzed 504 patients. Our meta-analysis revealed that EBUS elastography technique had a high sensitivity (0.93) and specificity (0.85) for diagnosing malignant and benign intrathoracic lymph nodes. In addition, the overall test performance assessed by the area under the curve was quite remarkable (area under summary receiver operating characteristic curve ¼ 0.93). In this meta-analysis, the mean DOR of 74 suggested that EBUS elastography
was reliable in detecting intrathoracic lymph nodes. Likelihood ratios are more clinically meaningful than sensitivity or specificity values. The NLR value was 0.09, which was low enough for clinical purposes. That is to say, if the EBUS elastography showed a negative result of malignant disease, then the lesion was more likely to be benign. However, the value of PLR was 6.3, which was not high enough to exclude the benign lymph nodes when EBUS elastography was positive. This is easy to understand. The fiber formation and tissue calcification in chronic inflammation lead to increased stiffness of lymph nodes [30], which easily leads to the false-positive results of EBUS elastography. That there was a highly significant heterogeneity among the studies analyzed in this study is noteworthy. No threshold effect was observed. Furthermore, we conducted meta-regression analysis to explore the source of the heterogeneity. We found that the method of analyzing elastography and the study design were the major sources of heterogeneity. In the subgroup analysis, studies with retrospective design had a higher specificity than studies with prospective design. Also, the studies with quantitative method had a less specificity than studies with semiquantitative method. The other subgroups
Ann Thorac Surg 2018;-:-–-
5
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Fig 3. (A) The summary receiver operator characteristic (SROC) curve of the included studies. (B) The likelihood ratio (LR) of the combined results. (AUC ¼ area under the curve; SENS ¼ sensitivity; SPEC ¼ specificity.)
(malignant prevalence) did not show significantly different sensitivities or specificities. Izumo and colleagues [11] were the first to conduct a systematic analysis for the elastography images. They classified the lymph nodes elastography images into three types and obtained a fine diagnostic yield. However, the semiquantitative methods largely relied on operators’ objective judgments and might cause interobserver disagreement [31]. Huang and colleagues [21] and Korrungruang and colleagues [20] adopted the similar classification method in their studies. However, He [17] and Sun [19] and their colleagues chose four or five types for classification [17, 19]. In contrast to the semiquantitative methods, quantitative methods mostly relied on computer-aided calculation, and thereby overcame the bias of artificial judgments. The quantitative methods, however, did not seem to be stable. For example, the strain ratio was chosen in the studies of He [17], Rozman [18], and Korrungruang [20] and their colleagues to describe the elastography images. Although their results were similar, the cutoff values for
these three studies were 32.07, 8, and 2.5, respectively. Sun and colleagues [19] selected a new method, the “mean grey value,” to describe the stiffness of target lesions. A new algorithm in which the image was first quantized and then clustered in the quantized color space in prior segmentation part was introduced by Nakajima and colleagues [10]. Whether the semiquantitative method or quantitative method is a better approach is still not clear. In the study by He and colleagues [17], the quantitative method of strain ratio showed a better diagnostic yield than the semiquantitative method of the elastography score. However, in the studies by Sun [19] and Huang [21] and their colleagues, the elastography score showed better performance than the quantitative method of mean grey value. In our meta-analysis, the semiquantitative method showed a good specificity compared with the quantitative method, and there was no significant difference for the sensibility between those subgroups. Subjective categorizations of B-mode images are useful for evaluating the presence of malignant lymph nodes.
Table 2. Subgroup Analysis Subgroup Malignant prevalence >0.5 <0.5 Design Prospective Retrospective or NA Method Semiquantitative Quantitative CI ¼ confidence interval;
Studies, No.
Summary Sensitivity (95% CI)
3 4
0.96 (0.91–1.00) 0.89 (0.80–0.98)
3 4
0.95 (0.89–1.00) 0.91 (0.83–1.00)
3 4
0.92 (0.83–1.00) 0.93 (0.87–1.00)
p
Summary Specificity (95% CI)
0.77
0.11 0.86 (0.77–0.95) 0.85 (0.77–0.92)
1.00
0.001 0.80 (0.71–0.89) 0.89 (0.83–0.95)
0.74
NA ¼ not available;
No. ¼ number.
p
0.001 0.90 (0.83–0.96) 0.81 (0.73–0.89)
6
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
Ann Thorac Surg 2018;-:-–-
technique, has been widely applied in clinical practice. Some researcher regarded it as being more stable than strain ratio. Up to now, however, the shear wave sensor has not been embedded in the EBUS machine. We hope the technique of EBUS combined with shear wave could be a reality in the near future. Some limitations should not be neglected. First, all of the studies included were single-center and small pilot were adopted (range, 49 to 120), and four of these were retrospective studies. Selection bias could not be avoided. In addition, all of the studies except one were conducted at centers in East Asia. The granulomatous diseases (sarcoidosis, tuberculosis, etc) that are common in Asian countries are relatedly rare in developed countries. Although no publication bias was observed in our study, publications with positive or significant results were more likely to be published; therefore, a certain degree of publication bias was inevitable. Our results suggest that the EBUS elastography is a reliable technique for diagnosing intrathoracic lymph nodes with a high sensitivity and specificity and could be applied more widely in diagnosing pulmonary disease. The authors wish to thank Guoxiang Ruan for his reasonable suggestions for this paper. Fig 4. The subgroup analysis of the study. (CI ¼ confidence interval.)
These features include size exceeding 1 cm, circular shape, heterogeneity, lack of central hilar structure, and lack absence of coagulation necrosis, among others [32]. However, none of those features seems to be as reliable as the EBUS elastography according to those studies [10, 11, 17, 18]. Besides, proper training and experience were required before the B-mode images could be distinguished. EBUS elastography has fine sensitivity and diagnostic yield in diagnosing intrathoracic lymph nodes [8]. But that is not to say that a biopsy specimen can be replaced by elastography. Different from the lymph nodes of the digestive tract, intrathoracic lymph nodes register more benign diseases, such as tuberculosis and sarcoidosis, among others. EBUS elastography distinguishes the malignant and benign lesion through the stiffness of the tissue [33, 34]. EBUS elastography was good at NLR, but not for PLR, which easily led to the false-positive results of EBUS elastography. That is to say, PLR was not high enough to exclude the benign lymph nodes when EBUS elastography was positive. However, EBUS-TBNA is a fine technique with high PLR, which may compensate the shortcomings of EBUS elastography [35, 36]. In addition, the metastatic area was localized within the lymph node, which may lead to a false-negative result of EBUS-TBNA in some cases. However, the operator can visualize the suspicious area and control the puncture area to improve the diagnostic value of TBNA [10]. In addition, all of the publications included in this study were based on the EBUS real-time elastography technique. Shear wave, another finer elastography
References 1. Lyshchik A, Higashi T, Asato R, et al. Elastic moduli of thyroid tissues under compression. Ultrason Imaging 2005;27: 101–10. 2. Samani A, Zubovits J, Plewes D. Elastic moduli of normal and pathological human breast tissues: an inversiontechnique-based investigation of 169 samples. Phys Med Biol 2007;52:1565–76. 3. Bhatia KS, Cho CC, Tong CS, Yuen EH, Ahuja AT. Shear wave elasticity imaging of cervical lymph nodes. Ultrasound Med Biol 2012;38:195–201. 4. Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic moduli of breast and prostate tissues under compression. Ultrason Imaging 1998;20:260–74. 5. Bojunga J, Herrmann E, Meyer G, Weber S, Zeuzem S, Friedrich-Rust M. Real-time elastography for the differentiation of benign and malignant thyroid nodules: a metaanalysis. Thyroid 2010;20:1145–50. 6. Gong X, Xu Q, Xu Z, Xiong P, Yan W, Chen Y. Real-time elastography for the differentiation of benign and malignant breast lesions: a meta-analysis. Breast Cancer Res Treat 2011;130:11–8. 7. Xu W, Shi J, Zeng X, et al. EUS elastography for the differentiation of benign and malignant lymph nodes: a metaanalysis. Gasterointest Endosc 2011;74:1001–9; quiz 1115.e1-4. 8. Gu P, Zhao Y, Jiang L, Zhang W, Xin Y, Han B. Endobronchial ultrasound-guided transbronchial needle aspiration for staging of lung cancer: a systematic review and meta-analysis. Eur J Cancer 2009;45:1389–96. 9. Ophir J, Cespedes I, Ponnekanti H, Yazdi Y, Li X. Elastography: a quantitative method for imaging the elasticity of biological tissues. Ultrason Imaging 1991;13:111–34. 10. Nakajima T, Inage T, Sata Y, et al. Elastography for predicting and localizing nodal metastases during endobronchial ultrasound. Respiration 2015;90:499–506. 11. Izumo T, Sasada S, Chavez C, Matsumoto Y, Tsuchida T. Endobronchial ultrasound elastography in the diagnosis of
Ann Thorac Surg 2018;-:-–-
12.
13. 14.
15.
16.
17.
18.
19. 20.
21.
22.
mediastinal and hilar lymph nodes. Jpn J Clin Oncol 2014;44: 956–62. Whiting PF, Weswood ME, Rutjes AW, Reitsma JB, Bossuyt PN, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Method 2006;6:9. Deville WL, Buntinx F, Bouter LM, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Method 2002;2:9. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982–90. Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a summary ROC curve: dataanalytic approaches and some additional considerations. Stat Med 1993;12:1293–316. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 2005;58:882–93. He HY, Huang M, Zhu J, Ma H, Lu XD. Endobronchial ultrasound elastography for diagnosing mediastinal and hilar lymph nodes. Chin Med J (Engl) 2015;128: 2720–5. Rozman A, Malovrh MM, Adamic K, Subic T, Kovac V, Flezar M. Endobronchial ultrasound elastography strain ratio for mediastinal lymph node diagnosis. Radiol Oncol 2015;49: 334–40. Sun J, Zheng X, Mao X, et al. Endobronchial ultrasound elastography for evaluation of intrathoracic lymph nodes: a pilot study. Respiration 2017;93:327–38. Korrungruang P, Boonsarngsuk V. Diagnostic value of endobronchial ultrasound elastography for the differentiation of benign and malignant intrathoracic lymph nodes. Respirology 2017;22:972–7. Huang H, Huang Z, Wang Q, et al. Effectiveness of the benign and malignant diagnosis of mediastinal and hilar lymph nodes by endobronchial ultrasound elastography. J Cancer 2017;8:1843–8. Giovannini M, Hookey LC, Bories E, Pesenti C, Monges G, Delpero JR. Endoscopic ultrasound elastography: the first step towards virtual biopsy? Preliminary results in 49 patients. Endoscopy 2006;38:344–8.
REVIEW CHEN ET AL EBUS-ELASTOGRAPHY DIFFERENTIATES LNS
7
23. Dietrich CF, Jenssen C, Herth FJ. Endobronchial ultrasound elastography. Endosc Ultrasound 2016;5:233–8. 24. Jiang JH, Turner JJ, Huang JA. Endobronchial ultrasound elastography: a new method in endobronchial ultrasoundguided transbronchial needle aspiration. J Thorac Dis 2015;7(Suppl 4):S272–8. 25. Saftoiu A, Vilman P. Endoscopic ultrasound elastography–a new imaging technique for the visualization of tissue elasticity distribution. J Gastrointestin Liver Dis 2006;15:161–5. 26. Varela-Lema L, Fernandez-Villar A, Ruano-Ravina A. Effectiveness and safety of endobronchial ultrasoundtransbronchial needle aspiration: a systematic review. Eur Respir J 2009;33:1156–64. 27. Andreo GF, Centeno CC, Sanz SJ, Barturen BA, Hernandez GA, Ruiz MJ. Initial experience with real-time elastography using an ultrasound bronchoscope for the evaluation of mediastinal lymph nodes. Arch Bronconrumol 2015;5:e8–11. 28. He HY, Chen JL, Ma H, Zhu J, Wu DD, Lv XD. Value of endobronchial ultrasound elastography in diagnosis of central lung lesions. Med Sci Monit 2017;23:3269–75. 29. Inage T, Nakajima T, Yoshida S, Yoshino I. Endobronchial elastography in the evaluation of esophageal invasion. J Thorac Cardiovasc Surg 2015;149:576–7. 30. Coussens LM, Werb Z. Inflammation and cancer. Nature 2002;420:860–7. 31. Calvete AC, Rodriguez JM, de Dios BJ, Rios A, AbellanRivero D, Reus M. Interobserver agreement for thyroid elastography: value of the quality factor. J Ultrasound Med 2013;32:495–504. 32. Wang L, Wu W, Hu Y, Teng J, Zhong R, Han B, Sun J. Sonographic features of endobronchial ultrasonography predict intrathoracic lymph node metastasis in lung cancer patients. Ann Thorac Surg 2015;100:1203–9. 33. Mittal S, Mohan A, Madan K. Endobronchial ultrasound elastography for the differentiation of benign and malignant lymph nodes. Respirology 2017;22:1037–8. 34. Steinfort DP. Elastography in endobronchial ultrasound: Stretching the boundary of minimally invasive staging further. Respirology 2017;22:843–4. 35. Sheski FD, Mathur PN. Endobronchial ultrasound. Chest 2008;133:264–70. 36. Belanger AR, Akulian JA. An update on the role of advanced diagnostic bronchoscopy in the evaluation and staging of lung cancer. Ther Adv Respir Dis 2017;11:211–21.