PS01.160: PERIOPERATIVE PERFUSION ASSESSMENT WITH QUANTITATIVE FLUORESCENCE ANGIOGRAPHY

Abstract Background Anastomotic leakage (AL) occurs relatively frequent in oncologic esophageal surgery and is associated with increased morbidity and mortality, and risk of local recurrence of cancer. Poor anastomotic perfusion is probably an important factor of AL. Traditionally; perfusion has been assessed visually and manually by palpation, methods proven to be irreproducible and not possible in a minimal-invasive environment. Few studies on ICG-FA used in esophageal resection exists and some have shown promising results. However, most studies only use a visual interpretation of the angiography, leading to observer bias. In addition, this subjective visual assessment only allows a binary outcome; perfusion or no perfusion. A quantitative (q-ICG) approach enables a scaled, objective assessment, useful in lowering the risk of AL, however, also in exploring interventions and risk factors leading to improved or attenuated perfusion. The aim of this study was to establish a valid, reproducible perioperative q-ICG method. Methods Design and validation: A Q-ICG method for assessment of ICG-FA was developed in a porcine model, using seven pigs. To establish which metric of Q-ICG to use, neutron labeled microspheres were injected intraarterially as microspheral flow is considered the gold standard in microvascular flow assessment. Tissue samples from the stomach were excised and the regional flow was calculated, this was compared with Q-ICG metrics by correlation analysis. Attenuated perfusion: In seven pigs, the stomach and segments of small bowel were partly devascularized and local lactates were correlated to Q-ICG. Improved perfusion: As glucagon-like peptide-2(GLP-2) enhances fore- and midgut perfusion, 20 pigs underwent small bowel resection and anastomosis and randomized to GLP-2 or placebo, to test if Q-ICG could determine a difference in perfusion. Results We found excellent correlation of the q-ICG metric ‘normalized slope’ to regional microspheral flow (r = 0.92–0.96, P = 0.001). In addition, local lactate measurements of the devascularized stomach and small bowel correlated to the normalized slope of q-ICG (Spearman rho = 0.8). Q-ICG found significant improved anastomotic perfusion of pigs treated with GLP-2 (P < 0.05). Conclusion We established and validated a method of Q-ICG in normal, attenuated and improved perfusion, that may be useful in esophageal surgery, where construction of an anastomosis requires an optimal perfusion. Disclosure All authors have declared no conflicts of interest.