Quantitative analysis of colon perfusion pattern using indocyanine green (ICG) angiography in laparoscopic colorectal surgery

Abstract

Purpose

This study aimed to quantitatively evaluate colon perfusion patterns using indocyanine green (ICG) angiography to find the most reliable predictive factor of anastomotic complications after laparoscopic colorectal surgery.

Methods

Laparoscopic fluorescence imaging was applied to colorectal cancer patients ( n = 86) from July 2015 to December 2017. ICG (0.25 mg/kg) was slowly injected into peripheral blood vessels, and the fluorescence intensity of colonic flow was measured sequentially, producing perfusion graphs using a video analysis and modeling tool. Colon perfusion patterns were categorized as either fast, moderate, or slow based on their fluorescence slope, T _1/2MAX and time ratio (TR =  T _1/2MAX/ T _MAX). Clinical factors and quantitative perfusion factors were analyzed to identify predictors for anastomotic complications.

Results

The mean age of patients was 65.4 years, and the male-to-female ratio was 63:23. Their operations were laparoscopic low anterior resection (55 cases) and anterior resection (31 cases). The incidence of anastomotic complication was 7%, including colonic necrosis ( n = 1), anastomotic leak ( n = 3), delayed pelvic abscess ( n = 1), and delayed anastomotic dehiscence ( n = 1). Based on quantitative analysis, the fluorescence slope, T _1/2MAX, and TR were related with anastomotic complications. The cut-off value of TR to categorize the perfusion pattern was determined to be 0.6, as shown by ROC curve analysis (AUC 0.929, P < 0.001). Slow perfusion (TR > 0.6) was independent factor for anastomotic complications in a logistic regression model (OR 130.84; 95% CI 6.45–2654.75; P = 0.002). Anastomotic complications were significantly correlated with the novel factor TR (> 0.6) as the most reliable predictor of perfusion and anastomotic complications.

Conclusions

Quantitative analysis of ICG perfusion patterns using T _1/2MAX and TR can be applied to detect segments with poor perfusion, thereby reducing anastomotic complications during laparoscopic colorectal surgery.