Efficacy of intraoperative ICG fluorescence imaging evaluation for preventing anastomotic leakage after left-sided colon or rectal cancer surgery: a propensity score-matched analysis

Abstract
Intestinal perfusion at the anastomotic site is thought to be one of the most influential risk factors for postoperative anastomotic leakage (AL). We evaluated the efficacy of indocyanine green (ICG) fluorescence imaging at the stump of the proximal colon in left-sided colectomy or rectal resection in terms of decreasing the incidence of AL. Prospectively collected data were retrospectively evaluated. Patients who underwent left-sided colectomy or rectal resection were enrolled (ICG group; n = 197), and patients who had undergone a similar procedure before the ICG group were enrolled from the charts as historical controls (HC group; n = 187). After ICG evaluation, anastomosis was performed where fluorescence was sufficient. The incidence of AL was compared between the ICG and HC groups. Propensity score (PS)-matched data were analyzed to clarify the risk of AL. AL occurred in 6 patients (3.3%) in the ICG group and 17 (10.7%) in the HC group. ICG evaluation revealed 179 patients with good fluorescence and 18 with poor/none perfusion (9.1%). The transection line was changed in all patients with poor/none fluorescence. Three of these 18 patients developed AL (16.7%), though transection line was changed at which is thought to be good. We hope AL in poor/none fluorescence can be prevented at the same rate of cases with good fluorescence. Actually, the rate of that was significantly higher compared with good fluorescence patients (P = 0.038). 93 patients in each group were compared by PS-matched data analysis, which showed the AL rate in the ICG group was significantly lower than that in the HC group (3.2% vs 10.8%, respectively; P = 0.046). Even though this study has limitations of comparison of data prospectively collected and retrospectively analyzed, intraoperative ICG fluorescence imaging evaluation could significantly decrease the incidence of AL.