Background: Indocyanine green (ICG) is commonly used to assess perfusion, but quality defining features are lacking. We sought to establish qualitative features of esophageal ICG perfusion assessments, and develop an esophageal anastomotic scorecard to risk-stratify anastomotic outcomes.
Methods: Single institution, retrospective analysis of children with an intraoperative ICG perfusion assessment of an esophageal anastomosis. Qualitative perfusion features were defined and a perfusion score developed. Associations between perfusion and clinical features with poor anastomotic outcomes (PAO, leak or refractory stricture) were evaluated with logistic and time-to-event analyses. Combining significant features, we developed and tested an esophageal anastomotic scorecard to stratify PAO risk.
Results: From 2019 to 2021, 53 children (median age 7.4 months) underwent 55 esophageal anastomoses. Median (IQR) follow-up was 14 (10-19.9) months; mean (SD) perfusion score was 13.2 (3.4). Fifteen (27.3%) anastomoses experienced a PAO and had significantly lower mean perfusion scores (11.3 (3.3) vs 14.0 (3.2), p = 0.007). Unique ICG perfusion features, severe tension, and primary or rescue traction-induced esophageal lengthening [Foker] procedures were significantly associated with PAO on both logistic and Cox regression. The scorecard (range 0-7) included any Foker (+2), severe tension (+1), no arborization on either segment (+1), suture line hypoperfusion >twice expected width (+2), and segmental or global areas of hypoperfusion (+1). A scorecard cut-off >3 yielded a sensitivity of 73% and specificity of 93% (AUC 0.878 [95%CI 0.777 to 0.978]) in identifying a PAO.
Conclusions: A scoring system comprised of qualitative ICG perfusion features, tissue quality, and anastomotic tension can help risk-stratify esophageal anastomotic outcomes accurately.