The role of endobronchial ultrasound‐guided transbronchial needle aspiration liquid‐based cytology in the diagnosis of mediastinal lymphadenopathy

Abstract Background Endobronchial ultrasound‐guided transbronchial needle aspiration (EBUS‐TBNA) is a minimally invasive, reliable technique for sampling mediastinal lymph nodes (LNs). Liquid‐based cytology (LBC) is widely used for cervical cancer screening because it provides reliable and feasible results. The present study aimed to evaluate effectiveness of the combination of EBUS‐TBNA and LBC in the diagnosis of mediastinal lymphadenopathy. Methods A total of 602 LNs that were retrospectively analyzed were sampled in 442 patients who underwent EBUS‐TBNA between January 2014 and December 2016. The histopathological result of TBNA tissue or cell blocks was considered as the gold standard to evaluate diagnostic utility of LBC and conventional smears (CS) for the diagnosis of mediastinal lymphadenopathy. Results Of the 602 LNs, 265 were mediastinal LN metastases from lung cancer, four were lymphoma, and 333 were benign. The sensitivity of LBC and CS in the diagnosis of mediastinal LN metastases from lung cancer was 72.8% and 63%, respectively, and the specificity was 98.5% and 97%, respectively. The positive predictive values for LBC and CS were 97.5% and 94.4%, respectively, whereas the negative predictive values were 82.2% and 76.9%, respectively. The accuracy of LBC and CS was 88% and 83.7%, respectively. The diagnostic value of LBC was significantly higher than that of CS (P = .001). Conclusions The combination of EBUS‐TBNA and LBC is a highly reliable and feasible procedure that optimizes diagnostic utility for the diagnosis of lung cancer and mediastinal LN staging.


| INTRODUCTION
Recently, endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has emerged as a first-line, minimally invasive procedure for the diagnosis of patients with mediastinal lymphadenopathy. 1,2 Under guidance of an ultrasound probe, mediastinal lymph node (LN) biopsies can be performed effectively with TBNA. Moreover, EBUS-TBNA has been demonstrated high sensitivity (92%) and excellent diagnostic accuracy (98%) in the diagnosis and staging of lung cancer. 3 Hence, current guideline-recommended EBUS-TBNA as the first-line approach for mediastinal LN staging of lung cancer. 4 Liquid-based cytology (LBC), as an extensively used cytopathologic technique, is a widely used method for screening cervical cancer because it significantly improved detection rates of cervical cancer and precancerous lesions. 5 Recently, LBC for nongynecological specimens has gained increasing importance due to good fixation and well-preserved nuclear details, 6 which got more reliable and feasible results compared with conventional smears (CS). Therefore, the present study aimed to retrospectively evaluate diagnostic utility of EBUS-TBNA combined with LBC in mediastinal lymphadenopathy. The patients fasted for at least 4 hours prior to the procedure.

| PATIENTS AND METHODS
Before commencing the bronchoscopy, local anesthesia was achieved with 2% lidocaine. Some patients also received an intravenous administration of midazolam and (or) fentanyl for conscious sedation. Blood pressure, heart rate, oxygen saturation, electrocardiogram, and consciousness level of the patients were monitored during the procedure. automated slide processor. Part of the aspirates were uniformly spread onto glass slides and fixed with 95% ethanol to make CS slides. The aspirates of next puncture were taken to make CS and LBC in the reversed order. Specimens from both CS and LBC were stained with hematoxylin and eosin (H&E), and an optical microscope was utilized to examine the slides ( Figure 1). A positive cytologic result of malignancy was accepted as evidence of cancer. It was considered a cytological negative outcome that no malignancy was seen in the cytologic preparations. In addition, the staining result was considered benign if a few atypical cells were identified but were insufficient for a malignant diagnosis.
After obtaining histology samples, the specimens were immediately fixed with 10% formalin, embedded in paraffin, sliced, stained with H&E. If adequate tissue was not obtained during the TBNA procedure, we used the aspirates to make cell blocks for a histopathological diagnosis. The final pathology diagnosis of histology samples (including cell blocks) of EBUS-TBNA was considered as the gold standard. Immunohistochemistry was performed on histology samples or cell blocks to further clarify the classification of lung cancer.
The present study was approved by the Ethics Committee of the First Affiliated Hospital of Soochow University. Signed informed consent was obtained from the patients or guardians.

| Statistical analysis
The sensitivity, specificity, accuracy, positive predictive values and negative predictive values of CS and LBC were calculated by using standard formulas. Diagnostic utility of these techniques for adenocarcinoma, small cell lung cancer, and total lung cancer cases was compared using McNemar's test. Fisher's exact test was used in comparing diagnostic utility of LBC with that of CS for squamous cell carcinoma and non-small cell lung cancer not otherwise specified (NSCLC-NOS). All statistical analyses were performed using the SPSS statistical software package (SPSS version 19.0, Chicago, USA). P < .05 was considered statistically significant.  Table 4).  11 moreover, LBC has been widely used for respiratory tract specimens. 7 In addition, LBC specimens can be efficiently used for molecular tests when the available material is insufficient to prepare cell blocks. 12 Kobayashi et al analyzed the difference in cell number, cell morphology, and slide background between CS and LBC samples obtained by TBNA, and then concluded that LBC could be reliably and routinely used in specimens obtained via TBNA. 13 Fan et al reported the sensitivity of LBC (71.6%) was significantly higher than that of traditional cytology (57.8%) in the diagnosis of lung cancer, and LBC was also found to be more valuable in the diagnosis of small cell lung cancer. 14  False negatives in the present study occurred probably due to the following reasons: malignant cells were not aspirated or not seen, or misinterpreted as benign. There were preparation artifacts which obscured cytological detail. 16 Several limitations of our study must also be acknowledged.

| RESULTS
First, this study was a retrospective analysis and subject to the limitations of the study design, so an information bias was inevitable.
Second, in this study, the patients included were all from one single hospital.
In conclusion, the combination of EBUS-TBNA and LBC is a highly reliable, safe, and feasible procedure that optimizes diagnostic utility for the diagnosis of lung cancer and mediastinal LN staging.