Diagnostic value of endobronchial ultrasound elastography combined with rapid onsite cytological evaluation in endobronchial ultrasound-guided transbronchial needle aspiration

Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) technique was developed in 2002. Currently, EBUS-TBNA is highly recommended by the National Comprehensive Cancer Network (NCCN) and the American College of Chest Physicians (ACCP) for diagnosis and preoperative staging of patients with lung cancer. Ultrasonography is frequently used to identify malignant lymph node status and guiding lymph node aspiration. However, conventional ultrasonography has not shown particularly high accuracy in identifying malignant lymph nodes [1]. Ultrasound elastography is a real-time imaging technique to identify malignancy based on the tissue stiffness. The relative elasticity or stiffness of the tissue is determined by applying stress to the tissue, causing deformation; the resulting echo signals are converted to real-time images [2]. Red signals indicate a low elasticity coefficient in tissues, while blue signals demonstrate a high elasticity coefficient in tissues. Compared to benign lesions, malignant lesions are usually firmer, making them easier to detect with elastosonography [3]. Recently, ultrasound elastography has been widely applied for diagnosis of breast cancer and prostate cancer and in liver fibrosis staging [4‐6]. Since EBUS elastography can rapidly identify malignant mediastinal lymph nodes, it has also been used in EBUS-TBNA to better identify suspicious lesions. Mittal et al. reported that the sensitivity and specificity of EBUS elastography to identify malignant lymph nodes were 85.7% to 100% and 66.7% to 92.3%, respectively [7].

During EBUS-TBNA, rapid onsite cytological evaluation (ROSE) has been widely used to evaluate samples for example, guiding following passes and determining adequacy of samples for further immunohistochemistry or next generation sequencing [8]. However, some studies have shown that ROSE alone does not improve the diagnostic sensitivity of EBUS-TBNA [9].

Recently, ROSE has been used as an assistive diagnostic technique in some EBUS elastography related studies; however, the diagnostic value of combined EBUS elastography and ROSE remains undetermined [10, 11]. The aim of this study was to investigate the diagnostic value of combined EBUS elastography and ROSE in EBUS-TBNA.

EBUS-TBNA has become a mature interventional diagnostic procedure in clinical practice. Previous studies have found no statistically significant difference in diagnostic sensitivity between EBUS-TBNA and mediastinoscopy for malignant lymph nodes evaluations, in cases where previous imaging results have indicated enlargement of mediastinal and/or hilar lymph nodes or adjacent lesions in the lungs [16]. However, compared to mediastinoscopy, EBUS-TBNA is less invasive, less expensive, and more tolerable for patients [17, 18]. EBUS-TBNA has gradually replaced mediastinoscopy and is currently recommended in multiple guidelines as the first choice for mediastinal staging in lung cancer.

To further improve diagnostic sensitivity of EBUS-TBNA and decrease procedure-related complications, some studies have analyzed echographyc features such as echogenicity (homogeneous or heterogeneous), margin status, presence or absence of central hilar structure, short axis diameter, and coagulative necrosis under B-mode ultrasound to evaluate lesions [1, 19, 20]. Even when including such features with EBUS, discrepancies have persisted among different studies [21]. To date, no EBUS features have proven to be consistent with the diagnosis of malignant lymph nodes. Thus, combining real-time imaging technology and EBUS elastography could be a new technique to resolve these problems. Ultrasound elastography is increasingly widely applied in the diagnosis of breast cancer, prostate cancer, liver disease, myopathies, etc. [22‐25].

Endoscopic elastography has been well developed recently and has demonstrated its superiority in lesion evaluation compared with conventional ultrasound, especially in esophageal endoscopic ultrasound. A previous study showed that the sensitivity, specificity, PPV, NPV, and overall accuracy of endoscopic elastography for diagnosing pancreatic solid tumors were 93%, 66%, 92%, 69%, and 85%, respectively [26], while the sensitivity and specificity in differentiating benign versus malignant peripancreatic lymph nodes were 91.8% and 82.5%, respectively [27]. Several studies have demonstrated that EBUS elastography can differentiate between benign and malign lymph nodes.[13, 28, 29]. In a study by Izumo et al., that retrospectively included the evaluation of 75 lymph nodes, EBUS elastography showed a sensitivity, specificity, PPV, NPC and overall diagnostic accuracy of 100%, 92.3%, 94.6%, 100% and 96.7% respectively for diagnosing malignancy [12]. A NPV of 100% demonstrated that EBUS elastography could decrease the number of unnecessary aspirations and the total examination time. Controversially, other studies have indicated approximately 75% to 87% sensitivity and 65% to 68% specificity of EBUS elastography in diagnosing malignant lymph nodes [30, 31]. This discrepancy could be partly explained by different interpretations of the EBUS elastography images. For example, Izumo et al. did not include lymph nodes that appeared partially blue or nonblue on elastography images in the statistical analysis. More recently, some modifications have been applied to the observation method of elastography, i.e., a more objective quantification method. Subjective quantification should be replaced for the objective quantification method, which includes the stiffness ratio, strain histogram, stiffness area ratio, or strain ratio to predict lesion elasticity characteristics [32‐35]. Fujiwara et al. [33]. achieved a sensitivity of 83.0% and a specificity of 96.2% by combining B-mode ultrasound and elastography in the prediction of nodal metastasis. However, quantitative measurement requires additional software for image processing and calculations, which are time consuming. Furthermore, additional training and experience are required to evaluate B-mode ultrasound results. Therefore, Izumo's score is currently the most widely used method in clinical practice and was applied in this study. Consistent with previous studies that found EBUS elastography sensitivity ranging from 71% to 90.6%, this study showed a sensitivity of 90.5%. However, the specificity was only 57.3%, which was less than that in other studies (67% to 82.6%) [13, 30, 36]. This discrepancy may result from multiple factors related to the final results during EBUS elastography, e.g., calcification or necrosis in the lymph nodes [3]. Previous studies have reported false positive results, mostly due to increased stiffness in cases of tuberculosis, pneumoconiosis, or sarcoidosis [37‐40]. In this study, 12 lymph nodes had blue images on elastography, but the final pathological diagnosis was calcification or fibrosis. Thus, false positives did occur in this study. Conversely, necrosis, hemorrhage, or liquefaction in malignant lesions can cause false negatives. Therefore, the accuracy of the EBUS elastography score could have been affected by the structure of the lymph nodes. In addition, EBUS elastography is a subjective method and is largely related to the operator and the physiological situation (heart rate and respiratory rate) of the patient [41].

ROSE can be used to evaluate the sample during EBUS-TBNA. However, some studies have found no significant difference in the diagnostic value of EBUS-TBNA with or without ROSE [9]. ROSE can provide rapid feedback regarding sample adequacy, increasing aspiration efficiency. ROSE can also guide the operator in identifying the sampling site, determining the adequacy of the sample, and decreasing procedure duration [8]. Meena et al. showed that pulmonologists with cytopathology training could perform onsite cytological evaluation of EBUS-TBNA samples, and no significant difference in accuracy of the sample was identified when compared with cytopathologists [42, 43], benefiting clinical practice. In this study, ROSE was performed by a well-trained pulmonologist; the results demonstrate that the sensitivity and specificity of ROSE in diagnosing malignant lymph nodes were 95.73% and 79.05% respectively, consistent with findings in previous studies of 88.5% and 83.0% [43]

In previous studies, ROSE was only applied as an ancillary technique to evaluate sufficiency of the sampling in EBUS-TBNA [10, 11]. To better analyze the effects of elastography or ROSE, elastography and ROSE were combined for statistical analysis. In the present study, the sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio of elastography and ROSE in diagnosing malignant lymph nodes were 85.84%, 92.65%, 11.67, and 0.15, respectively, and the AUC was 0.940; each figure was greater in the combination than in the only elastography or only ROSE group. Since EBUS elastography or ROSE during EBUS-TBNA have been applied in multiple institutions, it is possible that performing EBUS-elastography and ROSE together during EBUS-TBNA could improve its diagnostic ability.

This study had some limitations. First, it was conducted in a single institution with limited sample sizes. Further studies must be performed in larger populations. Second, this study only compared the value of the elastography score and ROSE, without additional information regarding lymph node size, integrity, or vascularity. Combining these factors could increase the diagnostic value of elastography and ROSE.

The combination of elastography and ROSE during EBUS-TBNA for patients with enlarged mediastinal lymph nodes, could increase the clinical diagnostic value compared with ROSE or elastography alone. Combining elastography and ROSE during EBUS-TBNA in clinical practice is recommended.