P148

H. Furusea (Dr), Y. Matsumoto*b (Dr), K. Uchimuraa (Dr), T. Imabayashia (Dr), T. Tsuchidaa (Dr)

a Department of Endoscopy, Respiratory Endoscopy Division, National Cancer Center Hospital, Tokyo, JAPAN ; b Department of Endoscopy, Respiratory Endoscopy Division/Department of Thoracic Oncology, National Cancer Center Hospital, Tokyo, JAPAN

* yumatsum@ncc.go.jp

Background: It is essential to visualize target peripheral pulmonary lesions (PPLs) sufficiently by radial endobronchial ultrasound (R-EBUS) in diagnostic bronchoscopy using forceps. Otherwise, transbronchial needle aspiration (TBNA) and cryobiopsy have been reported effective for improving the diagnostic outcomes, but the appropriate boundary of the decision is unclear. Therefore, we aimed to identify the boundary based on R-EBUS images.

Methods: Consecutive patients who underwent forceps biopsy for PPLs using R-EBUS between June 2015 and May 2017 were retrospectively reviewed. Cases in which R-EBUS showed invisible or blizzard sign and those who underwent TBNA/cryobiopsy were excluded. The angle where the lesion covered the R-EBUS probe was defined as “contact angle”, and it was measured using an ImageJ based on captured R-EBUS images. Factors affecting the diagnostic yield were statistically analyzed, including the contact angle. Moreover, the relationship between the angle changes and the diagnostic yields was evaluated.

Results: We analyzed a total of 835 lesions, with the median size of 26.8 (range 7.1-121.0) mm. Entirely and partially circumferential R-EBUS images were observed in 471 (56.4%) and 364 (43.6%) cases, respectively; the corresponding diagnostic yields were significantly higher in the former (92.1% vs. 60.4%, p<0.001). Univariable analyses revealed that the increased contact angle was significantly associated with the successful diagnosis (p<0.001) in addition to the larger lesion size, positive bronchus sign, and visible on chest radiography. In the multivariable analysis, only the increased contact angle was the significant factor (10 degrees change, odds ratio 1.095 [95% confidence interval 1.076-1.115], p<0.001). The area under the receiver operating curve of the contact angle to the diagnostic yield was 0.770, and the optimal cut-off value was 310 degrees.

Conclusion: The contact angle by R-EBUS was significantly associated with the diagnostic yield of PPLs using forceps, with the cut-off value of 310 degrees.

Disclosure of funding source(s): none