Virtual mediastinoscopy: a new diagnostic tool for thoracic surgeons

Objectives: Cervical mediastinoscopy is the gold standard for staging of patients with suspected or proved lung cancer with enlarged and/or positive mediastinal lymph-nodes at enhanced computed tomography (CT) and position emission tomography (PET) scan. We conduct a pilot study to determine technical feasibility of virtual mediastinoscopy (VM) and to examine clinical utility of this technique. Methods: VM was based on images obtained by integrated PET scan and helicoidal CT images with a slice thickness of 2.5 mm. Virtual endoscopic images (VI) and virtual VM movies obtained by a specific computer software were used in the cranio-caudal direction to reveal locations of lymph-nodes and great vessel during cervical video mediastinoscopy (CVM). Results: Virtual images from 28 patients with lung cancer and enlarged mediastinal nodes positive at PET were obtained and accurately evaluated before CVM. Patient group included 18 men and 10 women, aged 48–76 years (median, 67 years). The maximum standard uptake value (SUVmax) of the mediastinal lymph-nodes ranged from 1.8 to 6.5. In all patients the VM enabled better understanding of the spatial and positional interrelationship between the PET positive mediastinal nodes and mediastinal anatomy allowing in all the cases but one a systemic biopsy in mediastinal stations. Based on the detailed virtual simulation on the nodal location, mediastinoscopic biopsy was performed by a Chamberlain’s procedure in one case. A nodal biopsy was performed in a single representative station in 13 patients, in two stations in 10, and in 3 stations in 5. N2 disease was mediastinoscopically identified in 10 patients. Eighteen patients had a benign disease (5 sarcoidosis). Conclusions: This study showed the feasibility of generating 3-D virtual mediastinoscopic images offering a best preoperative knowledge of the mediastinal anatomy providing the surgeon to make this invasive procedure safer and resulting more accurate.