Visualization of lymphatic flow in laparoscopic colon cancer surgery using indocyanine green fluorescence imaging

Intraoperative visualization of lymphatic flow could guide surgeons performing laparoscopic colon cancer surgery on the extent of intestinal resection required. The purpose of this study was to investigate indocyanine green fluorescence imaging for intraoperative detection of lymphatic flow and nodes in such patients. All patients undergoing elective laparoscopic surgery for colorectal cancer from October 2016 to July 2017 were included in this study. Indocyanine green was injected submucosally around the tumors via a colonoscope and lymphatic flow assessed with a laparoscopic near-infrared camera system intraoperatively. Lymphatic flow was visualized perioperatively in 43 of 57 patients (75.4%). The rate of visualized lymphatic flow was significantly higher in patients with a lower clinical stage than in those with a higher clinical stage (p = 0.0103). Among the 14 patients in whom lymphatic flow was not visualized, 10 (71.4%) had cStage III or IV cancer. Our results indicate the potential role of intraoperative navigation in colon cancer surgery in early-stage colon cancers. This method allows the surgeon to clearly identify lymphatic flow during surgery and allows the determination and individualization of the lymph node dissection range.

All eligible patients who were diagnosed with CRC and underwent elective laparoscopic surgery at Kindai University Hospital from October 2016 to July 2017 were enrolled in this study. The cancer location was diagnosed by colonoscopy, and the clinical stage was diagnosed by enhanced computed tomography, positron emission tomography-computed tomography, and/or magnetic resonance imaging.
Surgical procedures. The intestinal resection areas and lymph node dissection areas were determined in accordance with the tumor staging described in the Japanese Society for Cancer of the Colon and Rectum guidelines 5 . Lymph node dissection was performed regardless of the fluorescence results. When lymphatic flow by ICG-FI was observed beyond the D3 dissection range, lymph node dissection in the area along the lymphatic flow was added if clinically feasible. For example, if lymphatic flow to the right gastroepiploic artery region was observed in a patient with right-sided transverse colon cancer, lymph node dissection in the same region was added to the treatment protocol. ICG fluorescence imaging. For the detection of lymphatic flow, 0.2 to 0.3 mL of ICG solution (2.5 mg/ mL) was injected into the submucosal layer just beneath the tumor at a single site preoperatively, using an injection needle under colonoscopy. ICG injection was performed 1 or 2 days before surgery. Lymphatic flow was confirmed using a PINPOINT laparoscopic NIR camera system (NOVADAQ, Ontario, Canada) (Fig. 1). After removal of the specimen, the lymph nodes were extracted and observed with the same NIR camera to confirm the existence of fluorescence (Fig. 2).
The fluorescence of lymphatic flow was confirmed by all surgeons [operator, assistant, and camera assistant including the first author (H.U.)] at the time of the first intra-abdominal observation. In terms of lymph nodes, the fluorescence of them was confirmed ex vivo after specimen removal in a nodule-by-nodule fashion.
Statistical analysis. Data are presented as median and range. Qualitative data were reported as the number of patients (percentage of patients) and were compared with either the Pearson χ 2 test or Fisher's exact test, as deemed appropriate. Statistical analyses were performed using JMP 13 software (SAS Institute Inc., Cary, NC, USA).

Results
In total, 57 patients with CRC were analyzed. The patients' characteristics are shown in Table 1. Among the 57 patients, 33 (57.9%) were male and 24 (42.1%) were female. The median age was 71.5 years (range 45-94 years), and the median body mass index (BMI) was 22.9 kg/m 2 (range 15.7-30.4 kg/m 2 ). The tumor locations included the cecum in 5 (8.8%) patients, ascending colon in 16 (28.1%), transverse colon in 10 (17.5%), descending colon in 5 (8.8%), and sigmoid colon in 21 (36.8%) ( Table 1). None of the patients underwent conversion to an open procedure. All procedures were safely performed without any complications. Preoperative chemotherapy was www.nature.com/scientificreports/ performed in two (3.5%) patients (modified FOLFOX6 plus bevacizumab, n = 1; XELOX, n = 1). In four patients, simultaneous laparoscopic hepatectomy was performed for metastatic tumors. No adverse reactions to the ICG injection were observed in any of the patients. The lymphatic flow was observed at the time of the first intra-abdominal exploration. The lymphatic flow was visualized perioperatively in 43 (75.4%) patients and was not visualized in 14 (24.6%) patients. The relationship between the visualized lymphatic flow and the clinicopathological factors is shown in Table 2. The rate of visualized lymphatic flow was significantly higher in patients with a lower clinical stage than in those with a higher clinical stage (p = 0.0103). Among the 14 patients in whom lymphatic flow was not visualized, 10 (71.4%) had cStage III or IV cancer as shown in Table 3. There were no significant differences in the rate of visualized lymphatic flow between patients with high BMI (≥ 22.9 kg/m 2 ) and low BMI (< 22.9 kg/m 2 ), right-and left-sided tumors, and injection time in day 1 or day 2 before surgery ( Table 2).
Seventeen patients presented pathological lymph node metastasis (Table 4). Of these, lymphatic flow was visualized in 11 (64.7%) patients. All metastatic lymph nodes were identified in the fluorescent-marked lymphatic area. In the case of patient No. 6, who underwent right hemicolectomy for ascending colon cancer, the lymphatic flow was observed in the right branch of the middle colic artery, and lymph node metastasis was identified within the same area. This finding indicated that fluorescent lymphatic flow was consistent with lymph node metastasis. However, in patients No. 8 and No. 10 who presented left-sided transverse colon cancers at approximately at the same site and had the same pathological stage, the lymphatic flows differed. This phenomenon indicated that the lymphatic flow at the tumor may change in cases of metastatic lymph nodes.
The correlation between ICG fluorescent positivity and cancer cell positivity for lymph nodes is shown in Table 5

Discussion
ICG fluorescence imaging has been used in innovative surgical techniques including perioperative blood flow assessment in coronary artery bypass grafting, intraoperative imaging during flap plasty, and detection of sentinel lymph nodes 11,12 . Additionally, ICG fluorescence imaging has been used to assess the blood flow of the anastomotic sites in patients with CRC 10 . A few studies have described the visualization of lymphatic flow by ICG fluorescence imaging in patients with CRC 9,10 . In this study, lymphatic flow was visualized in approximately 75%  www.nature.com/scientificreports/ of all patients. The rate of visualized lymphatic flow was significantly higher in patients with lower clinical stage than in those with higher clinical stage. Patients with cStage III and IV had a higher incidence of non-visualized lymphatic flow. These results suggested that the ICG fluorescent imaging technique had insufficient detectability for lymph flow in patients at higher clinical stage. However, our results indicate a potential role for ICG for intraoperative navigation during colon cancer surgery in select patients with clinically node-negative colon cancers. When lymph node metastasis occurs, the lymph node can become filled with cancer cells, and the lymphatic flow is likely to disappear. In fact, we reviewed the status of lymphatic invasion for all cases. Among the 43 patients with visualized lymphatic flow, 27 exhibited no lymphatic invasion (62.8%) and 16 patients presented with positive lymphatic invasion (37.2%) ( Table 2), while of the 14 patients in whom lymphatic flow was not visualized, 5 presented with no lymphatic invasion (35.7%), while 9 (64.3%) exhibited positive lymphatic invasion. Therefore, whether the fluorescent lymphatic flow could be visualized or not was associated with the degree of microscopic lymphatic invasion. The same non-significant trend was observed among groups with or without pathological lymph node metastasis. These results indicated that the most suitable patients for visualization of lymphatic flow are likely those who have a clinically node-negative cancer.
Some studies have shown high sensitivity and specificity using ICG fluorescent imaging, although neither sensitivity nor specificity was markedly high in our study (Table 5). ICG fluorescent lymph nodes did not always indicate cancer-positive nodes and even ICG non-fluorescent lymph nodes could be cancer-positive nodes because ICG is not a cancer-specific fluorophore suitable for the visualization of lymphatic flow during laparoscopic surgery for colon cancer. Our main purpose was to detect the lymphatic flow using the technique to determine which cancer cells could potentially metastasize to the lymph nodes. An expected role for this technique might be the optimization of the extent of lymph node dissection based on the lymphatic flow observed around the tumor in an individual patient.
In a previous study, ICG was injected into the subserosa through the trocars used during the operation or into the submucosa through colonoscopy 9 . In the present study, we preferred to inject ICG using a colonoscope 1 or Table 4. Clinicopathologic characteristics of patients with cancer cell positive lymph nodes. C cecal cancer, A ascending colon cancer, T transverse colon cancer, D descending colon cancer, S sigmoid colon cancer, lap-C laparoscopic ileocecal resection, lap-Rt laparoscopic right hemicolectomy, lap-T laparoscopic transverse colon resection, lap-Lt laparoscopic left hemicolectomy, lap-S laparoscopic sigmoid colon resection, RCA right colic artery, MCA middle colic artery, IMA inferior mesenteric artery.  www.nature.com/scientificreports/ 2 days before surgery. We believe that ICG injection using colonoscopy enables accurate injection just beneath the tumor site and leads to more precise diagnosis of the lymphatic regions. This study has some limitations. First, the number of patients was small, and the follow-up period was short. Second, whether lymphatic flow could be observed in other ICG fluorescence imaging devices should be verified in further studies. Third, some technical errors and biases could not be ignored, particularly, in terms of the insufficient detectability of lymph flow. There may also be potential biases in our study due to the failure of ICG injection, the presence of a localized thick layer of mesorectal fat, or insufficient observation of fluorescence.

conclusions
We visualized intraoperative lymphatic flow in patients undergoing CRC surgery. Our results indicate the potential role of intraoperative navigation using ICG in colon cancer surgery in early-stage colon cancers. This method allows the surgeon to clearly identify lymphatic flow during surgery and allows the determination and individualization of the lymph node dissection range. However, further research is needed for selected patients with CRC.