Evaluation of thoracoscopic ultrasonography for localization of pulmonary nodules in resected ex-vivo human lungs
Hideki Ujiie1, Tatsuya Kato1, Suhaib Hasan1, Hsin-pei Hu1, Priya Patel1, Hironobu Wada1, Daiyoon Lee1, Kosuke Fujino1, David Hwang2, Marcelo Cypel1, Marc de perrot1, Andrew Pierre1, Gail Darling1, Tom Waddell1, Shaf Keshavjee1, Kazuhiro Yasufuku1.
1Division of Thoracic Surgery, Toronto General Hospital, University Health Network, University of Toronto, Toronto, ON, Canada, 2Department of Pathology, Toronto General Hospital, University Health Network, University of Toronto, Toronto, ON, Canada.
OBJECTIVE: Localization of small, non-visible and non-palpable nodules is challenging during video-assisted thoracoscopic surgery. The purpose of this study is to evaluate the feasibility of thoracoscopic ultrasonography for nodule localization in resected ex-vivo human lungs.
METHODS: Resected ex-vivo specimens from twenty five patients with malignant lung tumors who underwent either lobectomy or segmentectomy were used. The lung was positioned in its appropriate anatomical orientation on a surgical table and deflated or semi-inflated using a syringe and catheter. The tumor was localized and measured in its greatest dimension with a prototype convex probe ultrasound thoracoscope (XLTF-UC180, OLYMPUS CORPORATION) in different frequencies (5.0-12.0 MHz). Ultrasound measurements of tumor size and depth from lung surface were compared and correlated to the true diameter and depth from lung surface acquired from pathological morphology.
RESULTS: Ex-vivo evaluation was performed in sixteen solid tumors and nine part solid ground-glass opacity (GGO) nodules. All tumors were successfully localized in the deflated lung specimens (average size: 13.73 ± 5.22 SD mm) (Figure). Solid tumors were more easily visualized than GGO’s. Part solid GGO tumors were not easily detected in the semi-inflated specimen due to peri-tumoral air surrounding the tumor. In all cases, the prototype convex probe ultrasound thoracoscope produced better image quality than the commercially availabe laparoscopic ultrasound device. The tumor boundaries were best evaluated with an ultrasound frequency of 10 MHz. Tumor boundaries were less defined on the ultrasound images of tumors with greater depth and lungs with underlying disease. A strong positive correlation exists between the ultrasound measurement and true measurement of tumor size (R2 = 0.89, P < 0.001).
CONCLUSIONS: The newly developed convex probe ultrasound thoracoscope can be used to localize nodules in resected ex-vivo human lungs. The clarity of the tumor boundaries is influenced by the tumor type, its depth from lung surface, and the lung’s underlying pulmonary disease. Complete lung deflation and the use of 10 MHz ultrasound frequency optimizes the visualization of target tumors.
Back to 2016 Annual Meeting Thoracic Track