This study is a prospective, single-center, randomized, non-inferiority study. The patients with PPN ≤ 3 cm in the longest diameter and had no evidence of endobronchial lesion who underwent bronchoscopy between April 2016 and January 2017 at Division of Respiratory Disease and Tuberculosis, Faculty of Medicine, Siriraj Hospital were enrolled. The patients who were pregnant or had contraindication for bronchoscopy or transbronchial lung biopsy (TBLB) were excluded. The primary objective is the diagnostic yield of thin bronchoscopy and ENB in diagnosis of PPN. The secondary objectives include factors affected the diagnostic yield and complication of the procedures. All chest radiograph and CT chest were reviewed. Baseline characteristic of patients and PPN, including the longest diameter, characteristics, location and the presence of CT bronchus sign were recorded. All patients provided their written informed consent.
Patients were randomly assigned in 1:1 ratio into 2 groups, TB group and ENB group. Randomization sequence was computer generated in block size of 4. In TB group, thin bronchoscope (BF-MP60, 4-mm diameter, 2.0-mm working channel diameter; Olympus, Tokyo, Japan) was used. In ENB group, conventional bronchoscope (BF type TE2, 5.9-mm diameter, 2.8-mm working channel diameter; Olympus, Tokyo, Japan) and electromagnetic navigation system (superDimension®; Medtronic, Minneapolis, Minnesota) were used. In both group, R-EBUS (UM-S20-20R, 20 MHz, 1.7-mm distal end diameter; Olympus, Tokyo, Japan) and fluoroscopy were used to confirm the location of the lesion and biopsy forceps before performing TBLB. The procedures in both groups were performed by one bronchoscopist.
Thin bronchoscopic method (TB group)
Bronchoscopic procedures were performed using local anesthesia with lidocaine and moderate conscious sedation with intravenous midazolam and fentanyl. When the target bronchus was located, the R-EBUS probe was inserted through the bronchoscopic working channel. When the EBUS image was obtained, TBLB were performed under fluoroscopic guidance followed by bronchoalveolar lavage.
The bronchus level reached with the bronchoscope, the location of the EBUS probe related to the lesion on an EBUS image, and procedure-related complications were recorded. The biopsy specimens were immersed in 10% formalin and analyzed by pulmonary pathologists. Pneumothorax were screened using fluoroscopy in all patients immediately after procedure.
The final diagnoses were established by cytology, histopathology and microbiology. Benign diseases were diagnosed if there was stability or improvement of the lesion on imaging for a minimum period of 24 months.
Electromagnetic navigation bronchoscopy method (ENB group)
Pre-procedural planning for identification of the target lesion, airway path and registration points was performed after importing the CT data into the superDimension software. Bronchoscopic techniques were similar to the TB group except in performing under real-time navigation. When the bronchoscope was located in the bronchus of interest, the locatable guide was withdrawn and the R-EBUS probe was inserted through the extended working channel (EWC). When the EBUS image was obtained, TBLB was performed under fluoroscopic guidance followed by bronchoalveolar lavage.
Comparative analysis on the diagnostic yield of ENB and TB in diagnosis of PPN for non-inferiority has been used. The difference of the diagnostic yield between 2 groups were 0.29 (calculated from diagnostic yield of TB and ENB in diagnosis of PPN was 59% and 88%, respectively). Noninferiority of the TB method was concluded if the lower border of the 95% confidence interval (CI) for the difference in the diagnostic yields exceeded the predetermined non-inferiority border of 5%. Demonstration of non-inferiority with a statistical power of 80% at a one-sided significance level of 0.05 would require 38 patients per group.
The continuous variables were presented as the mean or median and standard deviation. Pearson’s chi-square test or Fisher’s exact test were used to test the association between categorical variables. Unpaired t-test was used to test the difference in mean of normally distributed quantitative variables. Results were considered statistically significant when the p-value was less than 0.05. All statistical analyses were performed using statistical software (SPSS for windows, version 20.0; SPSS; Chicago, IL).
This study was approved by the ethics committee of our institution. Written informed consent was obtained in all patients prior to the bronchoscopic procedure.