Software-based perfusion assessment of the ileal J-pouch – a standardized and objective approach of interpreting intraoperative indocyanine green near-infrared fluorescence

Abstract

Abstract Objective For ileal J-pouch after restorative proctocolectomy, intraoperative perfusion assessment is vital to prevent complications such as anastomotic leak (AL) of the pouch. Indocyanine green derived near-infrared fluorescence (ICG-NIRF) is gaining popularity for intraoperative visualization, while its interpretation and relevance remain subjective. Standardized NIRF protocols with objective and quantifiable data are not yet established. This study aimed to evaluate a standardized approach of intraoperative ICG-NIRF imaging combined with a novel, software-based perfusion assessment for the detection of AL of the pouch. We hypothesized that an analysis of fluorescence intensity over time would result in novel, quantifiable and objective indicators of perfusion. Methods In this prospective single centre study patients undergoing restorative proctocolectomy with ileal J-pouch formation were included. Intraoperatively, ICG-NIRF visualization was performed at three time points in a strictly standardized setting and video recorded. Postoperatively, a special software was used to depict fluorescence intensity over time, generating perfusion graphs for each recorded pouch visualization. The characteristics of these perfusion graphs were correlated with postoperative outcome, including AL of the pouch within 30 days precisely defined and screened for by pouchoscopy. Results 18 patients were included, of which 4 developed AL of the ileal pouch-anal anastomosis (IPAA). Distinctive perfusion curves were generated for regions of interest (ROIs) across each pouch. Novel characteristics of perfusion were identified (ingress, maximum slope, egress and maximum egress). Differences in the shape of the perfusion graphs became evident when comparing pouches with a normal to those with an AL. In cases of AL, the pouch apex displayed lower ingress, maximum slope and egress values, indicating impaired perfusion. This was confirmed by comparing the maximum slope of the perfusion curve versus the egress on the pouch apex. Conclusion We present a novel methodology to assess anastomotic perfusion in pouch surgery with ICG-NIRF in real-time. By dynamically analyzing fluorescence intensity over time in the form of perfusion graphs and correlation with outcome, novel parameters of perfusion were identified. These have the potential to improve the objectivity and quantifiability of intraoperative perfusion assessment.