Page 27 - Fluorescence-guided cancer surgery

After opening of the abdominal cavity the tumor was exposed without causing damage to lymphatic vessels around the tumor as much as possible. When no metastasized disease was found, 1.6ml ICG:Nanocoll was administered subserosally in 4 quadrants around the tumor. Directly after injection NIR uorescence images of lymphatic pathways were acquired using the Mini-FLARETM NIR uorescence imaging system18. Fluorescence imaging was performed on multiple time points during surgery. A SLN was de ned as uorescent hotspot that appeared after injection of the tracer. When multiple uorescent hotspots appeared in the same LN basin all uorescent LNs were de ned as SLNs. The anatomical location of the uorescent hotspots was determined using the lymph node stations as de ned by the Japanese Research Society for the Study of Gastric Cancer19. Patients underwent a standard-of-care partial or total gastrectomy with modi ed D2 resection, consisting of resection of the peri-gastric LNs and LN station 7,8 and 9. After resection, the specimen was analyzed ex vivo using the FLARETM NIR uorescence imaging system at the Pathology Department. The marked uorescent hotspots were resected from the specimen, transected, xed in formalin and embedded in para n for routine hematoxylin and eosin staining, and analyzed for tumor status.
The in vivo signal-to-background ratio (SBR) of the SLN was calculated by dividing the uorescence intensity of the SLN by the uorescence intensity of the directly surrounding fatty tissue. Accuracy rate was de ned by the number of patients in which tumor-negative SLNs were found when no tumor-positive lymph nodes were found in the entire specimen and the number of patients in which tumor-positive SLNs were found when tumor-positive lymph nodes were found in the whole specimen divided by the total number of patients. Accuracy rate was expressed as percentage with a 95% con dence interval. A false-negative patient was de ned as a patient in whom tumor-negative SLNs were found, while tumor-positive LNs were found in the resection specimen.
Con dence intervals for the binomial proportions were calculated using exact binomial con dence intervals. Numerical data were summarized with median (range).
Intraoperative near-infrared uorescence imaging
Intraoperative imaging procedures were performed using the Mini- Fluorescence-Assisted Resection and Exploration (Mini-FLARETM) image- guided surgery system, as described earlier18. Brie y, the system consists of
SLN detection in gastric cancer 25