Introduction Elastography was first proposed by Ophir of Texas State University in the United States in 1991 ，which can objectively reflect the elastic information of the tissue through the deformation of the external force. Then, the elastic information is converted to RGB (red,green,blue) mode image, where hard tissue is shown in blue, medium tissue in green and soft tissue in red, overlaid on the B-mode image. Different image colors reflect the difference of tissue hardness, and then help identify the difference between benign and malignant tissue hardness. UE has become a hot spot in medical ultrasound imaging shortly after it was proposed. Now UE is widely used in the diagnosis of diseases in the breast, thyroid, prostate and other related tissues. The common clinical UE includes quasi-static elasiticity imaging, acoustic radiation force pulse (ARFI) and shear wave elastography (SWE).
Lung cancer is one of the most common malignancies with high morbidity and mortality，the 5-year survival rate being only 16%. Lymphatic metastasis, the most common mode of metastasis, is of great significance for staging of lung cancer. Therefore, the correct diagnosis of benign and malignant intrathoracic lymph nodes becomes the key to definite diagnosis and accurate treatment. EBUS-TBNA has shown its certain diagnostic value in lung cancer, lymphoma, sarcoidosis and lymph node tuberculosis. However, researches related to real-time endobronchial UE (EBUS-RTE) are scarce, EBUS-RTE combined with EBUS B-mode ultrasound and enhanced CT is even fewer.
This study aimed to evaluate the value of endobronchial UE in differentiating benign and malignant intrathoracic lymph nodes qualitatively and quantitatively, and the combined features were evaluated as well. Here we show that endobronchial UE has significant value in the differential diagnosis of benign and malignant intrathoracic lymph nodes.
Materials and Methods Patients Patients undergoing EBUS-TBNA examination in the Department of Respiratory Diseases, Nanjing Hospital Affiliated to Nanjing Medical University from June 2016 to April 2017 were recruited in this study. All patients were evaluated by chest enhanced CT and/or 18-FDG PET-CT. Those who had enlarged intrathoracic lymph nodes(≥1cm) and/or 18-FDG high uptake (SUV Max \> 2.5) without bleeding tendency, abnormal coagulation function and serious cardiac dysfunction were finally selected. The study was approved by the Ethics Committee of Nanjing Hospital (YL 20160713-020) and supported by Nanjing Science and Technology Commission (201505002). All the patients gave written informed consent.
EBUS B-Mode ultrasound The patients were sedated by local anesthesia (lidocaine, China Otsuka Pharmaceutical Incorporated Company) and conscious sedation (midazolam, Jiangsu Nhwa Pharmaceutical incorporated company; fentanyl, Yichang Humanwell Pharmaceutical Incorporated Company; and propyl chloride, Xi'an Libang Pharmacertical Company) .The patients received an ultrasonic bronchoscopy (the Olympus 290 electronic bronchoscopy system of Japanese Olympus Company) from the mouth, the glottis, through the trachea. The locations of the lymph nodes were initially determined according to preoperative imaging examination. The B-mode ultrasound was started and the image focus and depth of observation was adjusted after the ultrasonic water sac was filled, so that the whole target lymph nodes and the surrounding normal tissues were clearly displayed on the screen. the investigators observed multi sections, selected the largest layer of the target lymph node, and retained pictures.The ultrasound image characteristics were independently recorded by three skilled respiratory physicians. The results were discussed and judged until consensus was reached.
EBUS UE The B-mode was switched to the elastography mode. When the image was stable, it would be frozen and photographed. The observer judged the image type according to the color ratio of the image. Type 1: mainly non-blue (green and red); type 2: partly blue, partly non-blue (green and red); and type 3: mainly blue. Then, the strain rate ratios of normal tissue (green and red) and the hardest region (the deepest blue) in the target lymph node region were measured and recorded. In contrast to B-mode ultrasound lymph node images, the elastography images were taken out as region of interest (ROI) in Image J software. The ROI was converted into RGB encoding mode, and the blue part was extracted from the Metlab software by setting the difference between the B component and the R and G components. After that, the blue area to ROI area ratio was calculated. According to the blue area ratio, the following grading standard was used to evaluate each lymph node as follows. 1 point: over 80% of the section was non blue (green, yellow and red); 2 point: more than 50%, but less than 80% of the section was non-blue (green, yellow and red); 3 point: more than 50%, but less than 80% of the section was blue; and 4 point: over 80% of the section was blue.
EBUS-TBNA and Final Diagnosis After elastography, switch to the blood flow pattern, avoid the blood vessels, confirm the safe needle path to the target node and adjust the depth of the needle (Cook Ireland Limited Liability Company, Ireland). EBUS-TBNA was performed under the guidance of real-time ultrasound. Each target lymph node was punctured 3\~4 times. All the histological and cytological specimens were analyzed by pathologists who were blinded to the elastography values. A positive diagnosis was confirmed by pathology and/or cytology, and all patients with negative results underwent thoracoscopic or open chest surgery and follow-up to confirm the benign and malignant lymph nodes.
Statistical Analysis All data statistics were analyzed by SPSS 22 (IBM, New York, USA) statistical software. The normal distribution data were described by mean and standard deviation, and the quantitative data were described by median (maximum and minimum), and the qualitative data were described by percentage. All the numerical variables were analyzed by one-way ANOVA, and the chi-square test was used for categorical variables. The difference was statistically significant when P\<0.05. If the benign and malignant group values conformed to normal distribution and the variance was homogeneous, the independent sample t test was used, otherwise the nonparametric test (Mann-Whitney test) was utilized. ROC analysis was conducted to evaluate the diagnostic value of single characteristics and combined characteristics. The maximum Youden index was calculated to determine the optimal cut-off point of differential diagnosis of benign and malignant lymph nodes. The diagnostic accuracy, sensitivity, specificity, PPV and NPV of each meaningful characteristic and combined characteristics were calculated respectively.
