Radial probe endobronchial ultrasound using a guide sheath for peripheral lung lesions in beginners

From December 2015 to January 2017, a retrospective study was conducted to investigate the clinical outcomes of patients undergoing EBUS-GS performed by beginners. During the study period, two physicians, neither of whom had previously performed EBUS-GS or radial probe EBUS only, began EBUS-GS at Pusan National University Hospital, a university-affiliated, tertiary referral hospital in Busan, South Korea. Before starting EBUS-GS, the two beginners both had 4 years of experience with conventional bronchoscopy and 3 years of experience with convex probe EBUS (700 conventional bronchoscopies and 200 convex probe EBUS per year by each physician). All of the consecutive patients with a peripheral lung lesion, who underwent EBUS-GS performed by one of the physicians, were prospectively registered. For each physician, the first 100 consecutive patients who received EBUS-GS were included in the analyses. Prior to EBUS-GS, written informed consent was obtained from all of the patients. The Institutional Review Board of Pusan National University Hospital approved this study (No. E-2016084) and informed consent was waived due to the retrospective nature of this study and the anonymized personal information prior to analysis.

All of the chest computed tomography (CT) scans were performed within 2 weeks prior to EBUS-GS. The imaging parameters were 120 kVp and 100–250 mAs. The stored CT raw data were used to reconstruct images at a slice thickness of 0.625 mm and intervals of 0.625 mm. The size of each peripheral lung lesion was measured from the CT images, based on the mean diameter of the lesion on the axial lung window setting. A peripheral lung lesion was diagnosed when the location of the lesion was beyond the segmental bronchus [10]. The lesion was classified as ground-glass opacity, part-solid, or solid according to a visual assessment method based on CT attenuation and modified from a previous study [11].

All of the EBUS-GS procedures were performed during in-patient hospital stays. Before the procedure, a 20 MHz radial probe EBUS (UM-S20–17S; Olympus, Tokyo, Japan) and GS kit (K-201; Olympus, Tokyo, Japan) were prepared according to the standard method of Kurimoto [5]. Patients under conscious sedation with intravenous midazolam and fentanyl underwent conventional bronchoscopy with a 4.0 mm flexible bronchoscope (BF-P260F; Olympus, Tokyo, Japan) to inspect the large airway. Lidocaine (2%) was applied to the tracheobronchial tree via the working channel of the bronchoscope. Following conventional bronchoscopy, the bronchoscope was advanced into the bronchus of interest as far as possible under direct vision based on the CT image. Thereafter, the GS-covered radial probe EBUS was advanced through the working channel of the bronchoscope until resistance was met. Then the probe was pulled back slightly to allow ultrasound scanning under X-ray fluoroscopic guidance. When the location of the target lesion was identified using EBUS, the probe was removed while the GS was kept in place for subsequent brush cytology and forceps biopsy. According to the sonographic features of the target lesion, the relationship between the lung lesion and GS was classified into three patterns, as previously reported [2, 5, 12]: within, adjacent to, and outside the lesion (Additional file 1). Brush cytology and a forceps biopsy via the GS were performed under X-ray fluoroscopy for the histological examination. Endobronchial ultrasound guided transbronchial needle aspiration was not simultaneously performed for mediastinal lymph node sampling during EBUS-GS. All of the procedures were performed without the assistance of virtual bronchoscopy navigation or an electromagnetic navigation system [13, 14]. If the lesion was located outside the EBUS probe, the sampling approach, whether brush cytology, forceps biopsy, or bronchial washing, was selected at the discretion of the bronchoscopist. A representative case of EBUS-GS for a peripheral lung lesion is shown in Additional file 2. To determine if iatrogenic pneumothorax had developed, initial chest radiographs were obtained 4 h after the procedure, and follow-up chest X-rays the following day. Severe hemorrhage was defined as endobronchial bleeding requiring transfusion, intubation, or an interventional procedure. Respiratory failure requiring intubation, pulmonary infection, air embolism, or premature termination of the procedure due to another unexpected complication was also recorded. X-ray fluoroscopy was performed to detect whether the GS had broken during the procedure. To identify breakage of the radial probe EBUS, an ultrasound image of the withdrawn probe held in the air was taken after the procedure, and saved on a picture archiving and communication system (Additional file 2).

Statistical analyses were performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA). The results are presented as numbers (percentages) or medians (interquartile ranges [IQRs]), as appropriate. Pearson’s chi-square test or Fisher’s exact test was used for categorical variables and the Mann–Whitney U-test was used for continuous variables. A P-value < 0.05 was considered statistically significant. To assess the learning curve of the procedure, cumulative sum (CUSUM) analyses were used to produce a learning curve for each physician. The definition of CUSUM analysis applied in this study was that of Bolsin and Colson (Additional file 3) [15]. A detailed description of the CUSUM analysis in this study is provided in Additional file 4.

Two hundred patients with peripheral lung lesions were included in the study (100 patients per physician). Their baseline characteristics are shown in Table 1. The median mean lesion diameter was 26 mm (IQR, 20–37 mm). Using the radial probe EBUS, 162 (81.0%) of the lesions were identified as being ‘within’ image and 24 (12.0%) ‘adjacent to’ image. However, 14 lung lesions (7.0%) were invisible. According to the appearance of the peripheral lung lesions on CT, there were 170 solid (85.0%), 26 part-solid (13.0%), and 4 ground-glass opacity (2.0%) lesions. The median number of brush cytology tests and forceps biopsies, performed via the GS, was 3 (IQR, 3–3) and 6 (IQR, 6–7), respectively. The overall EBUS-GS time was 20 min (IQR, 14–25 min). In addition, no significant difference in baseline characteristics was observed between the 100 study patients in which EBUS-GS was performed by one of the physicians (Additional file 5).

Table 2 lists the clinical diagnoses of the study patients. The overall diagnostic yield of EBUS-GS was 73.0%. Histological and cytological diagnoses were established in 146 (73.0%) and 42 (21.0%) of the 200 peripheral lung lesions, respectively. Diagnostic yields were significantly different among patients whose lesions had a mean diameter <  20 mm, 20–30 mm, and >  30 mm (46.8% vs. 80.8% vs. 81.3%, respectively, P < 0.001) (Table 3). No significant difference was observed in the diagnostic yield between solid and mixed lesions (75% vs. 69%, P = 0.553). However, the diagnostic yield of ground-glass opacity nodules was only 25%. Diagnostic yield “within the lesion” on EBUS findings was significantly higher than that of “adjacent and outside the lesion” on EBUS (80% vs. 58% vs. 14%, respectively, P < 0.001). In addition, the diagnostic yield obtained by the two physicians did not differ significantly (74.0% vs. 72.0%, P = 0.750).

The results of the CUSUM analysis are presented as learning curves, in which a positive deflection represents false results and a negative deflection represents true results (Fig. 1). The curves show that the two physicians attained competence immediately and the curves remained below the predetermined decision interval throughout the study period (H1 = 4.97). In addition, the graphs of the two physicians crossed the lower decision boundary during the study period.

Additional CUSUM analyses were performed for 50 consecutive patients with peripheral lung lesions < 30 mm. The respective curves remained between the predetermined decision interval (H0 = − 5.15 and H1 = 5.15), again indicating that the physicians attained competence immediately, even when performing procedures involving small lung lesions.

Overall, complications related to EBUS-GS during the learning curve occurred in three patients (1.5%): pneumothorax developed in two patients (1.0%) but resolved spontaneously without the need for chest tube drainage (Fig. 2), and one patient (0.5%) suffered pulmonary infection after the procedure (Fig. 3). Within the total group of study patients, none developed pneumothorax requiring chest tube drainage, severe hemorrhage, air embolism, or respiratory failure. There were no premature terminations of the procedure and none of the patients died due to the procedure.

During the study period, two radial probes EBUS were used by the two physicians and one probe broke. During EBUS-GS, breakage of the GS, observed fluoroscopically, only occurred in one patient (0.5%) (Fig. 4).