The Utility Of EBUS In Evaluating The Mediastinal And Hilar Lymph Node In Lung Cancer

Background:Endobronchial ultrasound(EBUS) examination is a minimal technique with the application of convex probe-endobronchial ultrasound(CP-EBUS) or radial probe-endobronchial ultrasound(RP-EBUS). EBUS allows the acquisition of images beyond the airway lumen and mucosa, and evaluation of extraluminal structures-aiway wall structure, paratracheal, parabronchial and mediastinal tissues. It was demonstrated that paratracheal structures(including mediastinum)4cm around the EBUS probe could be visualized in optimum condition.The TNM-stage determines the treatment and prognosis in patients with lung cancer. The evaluation of mediastinal and hilar lymph node is a difficult point in clinical practice. It is crucial to determine the mediastinal stage accurately in preoperative patients or after induction therapy. Chest computered tomography(CT), Positron emission tomography combined with CT (PET/CT), mediastinoscopy and endobronchial ultrasonography EBUS are common used to evaluate the mediastinal and hilar lymoh node and the accuracy of these processes are different. CT and PET/CT are noninvasive methods, but the CT size criteria for metastasis has limitation. Due to the high false-positive rate, PET/CT using for mediastinal staging was controversial. However, there were several studies and meta-analyses revealed that PET is superior to CT for mediastinal LN staging in potentially operable NSCLC. In some guidelines, patients with resectable lung cancer are recommended to undergo mediastinoscopy or EBUS examinations and process pathological diagnosis. EBUS-TBNA is a preferred option because it is aminimally invasive procedure that is generally available and of the high sensitivity.Radial probe-endobronchial ultrasound(RP-EBUS) was used to evaluate the mediastinal and hilar lymph node in patients with confirmed or suspected lung cancer. The radial probe-EBUS provides360°images. The frequency of20MHz raises the resolution ratio to lmm and enhance the depth of penetration of4-5cm paratrachea adj acent structures.This study would assess four EBUS characteristics(diameter, shape, border and echogenicity) in predicting the malignant or benign mediastinal and hilar lymph node in patients with lung cancer.Objects and methods:1.Objects:(1)Inclusion criteria:Patients who were diagnosed or suspected lung cancer and underwent EBUS to detect the mediastinal and hilar lymph node in Guangdong General Hospital Bronchoscope Room were enrolled in the study. They were informed consent before the bronchoscope procedures.(2) Exclusion criteria: Patients with benign pulmonary lesions, no informed consent and/or no pathological diagnosis of the lymph node we recorded in EBUS examinations were excluded.2. Methods:Eligible patients underwent the convention bronchoscopy and EBUS examination in the Bronchoscope Room of Guangdong General Hospital.(1) EBUS examination availed of a radial probe((UM-BS20-26R, diameter2.0mm,20-MHz,Olympus, Japan),the ultrasound processor (ENDOECHOEU-M2000, Olympus, Japan) and the ultrasound probe driver ((MAJ-935, Olympus, Japan) for detecting the mediastinal and hilar lymph nodes. The probe with a water balloon sheath(MAJ-667, Olympus)was introduced to the airway through the work channel of the flexible bronchoscope (BF-1T20; Olympus, Japan).(2) EBUS examination:EBUS scanning was performed after the convention bronchoscopy. Two proficient endoscopic physicians processed the EBUS examinations. The parameter of the ultrasonic host were set as below:gain (15-17),the contrast degree (4-6) and the tested range was3-4cm. According to the the Mountain and Dresler classification map, we evaluate the station4,7,10,11lymph node. The operators observed the EBUS video images in real-time when the probe reached the target positions. Once the lymph node was/were detected, the operators would record the data of the lymph node:station, amount of the lymph node, size, shape, border definition and echogenicity. The size of each lymph node was measured by the survey meter of ultrasound processor in real-time video images by the operator. We documented the longest short-axis diameter.Clear images were output at the largest diameter and saved in the medical image system.-The depiction of other characteristics were conducted base on the agreement of two individuals after the EBUS procedures.(3) The EBUS characteristics criteria and data collection:Patients’clinical information were collected prior to EBUS procedures by the one of the investigators. Once the operator detected the lymph node, we documented the EBUS characteristics according to the criteria below:Size measurement(Diameter):The longest short-axis diameter was measured by the screen’s ruler markings in real-time.Shape:We defined the shape as round or other shapes(oval, triangular, draping).Border definition:Distinct border was defined as a sharp line(high echoic) encircle the lymph node. If the margin was fuzzy, we recorded it as indistinct.Echogenicity:The lymph node was visualized since the echo was different from the surrounding connective tissue structure. We recorded the echo intensity as homogeneous, heterogeneous echogenicity.(4) After EBUS examination, the investigators would collect the pathological findings of the lymph node detected by EBUS. Patients with no pathological diagnosis of the lymph node we recorded in EBUS examinations or were diagnosed benign lesions were excluded. All cases were proved by surgery or mediastinoscopy pathology.(5) Statistical Analysis:The EBUS characteristics were compared with the final pathologic result of the lymph node. We used SPSS13.0data package for statistics analysis. χ2-test was used to inspect the relationship between the EBUS characteristics and the properties of the lymph node. The logistic regression model was established to assess the relevance between the pathology findings and these four significant factors. We evaluated the discrimination value of the significant characteristics with odds ratio(OR) and95%-confidence interval (95%CI). The sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy rate were calculated by standard definitions.Results:1.74patients (48men and26women) aged from21to78years old were collected from the Pulmonary Medicine Department, Respiratory Medicine Department, Thoracic Surgery between December2011and October2012. The mean age was58.66±11.70years old. Patients(n=15) who lacked the pathology diagnosis of the lymph node or with the pathology diagnosis of inflammatory pseudotumor and other noncancerous lesion were excluded. The histological diagnosis of the pulmonary lesions were concluded below:53were adenocarcinoma,17were squamous cell carcinoma,1were small cell carcinoma,1were large cell carcinoma,1were adenosquamous carcinoma and1were metastatic carcinoma, respectively. Surgery pathology were obtained in69patients and5patients were diagnosed by mediastinoscopy.24patients were found malignant lymph node by histological examination, the mediastinal lymph node metastatic prevalence was32.4%. No obvious complications were observed during the process of EBUS examination in this study.2.EBUS characteristics:76lymph node(25.41%) presented malignant, indicating mediastinal metastasis, and223lymph node(74.58%) were benign. Most of the lymph node were observed in the station7(n=123,41.14%). The EBUS features were summarized according to four characteristics:The size measured by EBUS was recorded as the longest short-axis diameter ranged from3.00mm to16.00mm and the median was7.80mm. Diameter of the malignant lymph node(n=76) ranged from 7.20-16.00mm and the benign node were3.00-13.30mm.91lymph node(30.43%) were recorded as round for shape,143lymph node (47.83%) were oval,26(8.70%) were triangular and39(13.04%) were draping. For the definition of the border, we recorded103(34.45%) lymph node as distinct and196(65.56%) as indistinct.91lymph node(30.40%) were estimated as heterogeneous and208(69.60%) nodes were homogeneous for echogenicity. The accuracy of prediction:Receiver operator characteristic (ROC)curve determined the optimal cutoff of the longest short-axis diameter as8.75mm. The area under the ROC curve (AUC) values was0.937(95%CI:0.910-0.964).The sensitivity was89.5%and the specificity was80.7%. For round shape, distinct border and heterogeneous echo, the sensitivity and the specificity in predicting the malignant lymph node were88.2%and84.2%,84.2%and82.5%,69.7%and83.0%, respectively.3.Predictive value:It is revealed in the univariate analysis that these four EBUS characteristics of the mediastinal and hilar lymph node (the longest short-axis diameter, shape, border and echogenicity) were significant in discriminating the malignant and benign node(P<0.001). We established the logistic regression model to evaluate the relevance between the pathology findings and these four significant factors. The longest short-axis diameter, the round shape, distinct border and heterogeneous echogenicity of the lymph node detected by EBUS were independent factors in predicting malignant lymph node, with the odds ratio(OR) were2.054,(95%CI=1.510,2.793)14.39(95%CI=4.30,47.172),13.30(95%CI=3.638,48.638),16.00(95%CI=4.283,59.793) respectively. There was a positive correlation between the diameter and the malignant lymph. The roung shape,distinct border and the heterogeneous echo were significant predictors for discriminating the malignant and benign lymph node. The theoretical predictive value of heterogeneous echo was highest among four characteristics, round shape, distinct border, and larger diameter were weakened change in order.4. The diagnoses model of EBUS characteristics predicting malignant lymph node:According to the logistic regression analysis, the regression equation was established:P=1/[1+e-(-11.548+2-667Diameter+2-667Shape+2-588Border+2-773Echogenicity)-](e was for natural constant.Malignance, round shape, distinct border and heterogeneous echo were recorded as1, otherwise recorded as0.) Prediction probability0.5was the cut off. when P<0.5were predicted benign, P≥0.5were predicted malignance. The sensitivity, specificity and accuracy were88.16%,96.85%and94.6%, respectively.5. Exploration of the operation skill of EBUS:The parameter of the ultrasonic host were set as below:gain (15-17),the contrast degree (4-6) and the tested range was3-4cm.Combining other operative skills of the water ballon,distinct EBUS images could be obtained.Conclusion:1. Radial probe-EBUS were qualified to evaluated the malignant or benign mediastinal and hilar lymph node in patients with lung cancer.2. The longest short-axis diameter, shape, boarder and echogenicity were significant predictors of malignant mediastinal and hilar lymph node.3. Diameter≥8.75mm,round shape,distinct border and heterogeneous echo were associated with malignant lymph node. The predictive value of heterogeneous echo was highest among four characteristics, round shape, distinct border, and larger diameter were weakened change in order.4. If the characteristics indicate malignant probabilities, further examination should be performed to confirm the metastasis.5. The parameter of the ultrasonic host were set as below:gain (15-17),the contrast degree (4-6) and the tested range was3-4cm.Combining other operative skills of the water ballon,distinct EBUS images could be obtained.6. EBUS was an effective technique of mediastinal and hilar evaluation.It is worth further exploration and popularization.