Farah Ladak, MD, MPH, Jerry Dang, MD, Noah Switzer, MD, MPH, Chunhong Tian, PhD, Daniel Birch, MSc, MD, FRCSC, FACS, Simon Turner, MD, FRCSC, Shahzeer Karmali, MD, FRCSC, FACS. University of Alberta
Introduction: The present study aims to evaluate the predictive value of indocyanine green (ICG) for the detection and prevention of anastomotic leak following esophagectomy. Anastomotic leak is a highly morbid and potentially fatal complication of esophagectomy. Ensuring adequate perfusion of the gastric conduit can minimize the risk of postoperative leak. Intraoperative evaluation with fluorescence angiography using ICG offers a dynamic assessment of gastric conduit perfusion, and can guide anastomotic site selection.
Methods: A search of electronic databases MEDLINE, EMBASE, SCOPUS, Web of Science and the Cochrane Library using the search terms “indocyanine/fluorescence” AND esophagectomy was completed to include all english articles published between 1946 and August 2017. Articles were selected by two independent reviewers based on the following major inclusion criteria: (1) Esophagectomy with gastric conduit reconstruction; (2) use of fluorescence angiography with indocyanine green to assess perfusion; (3) age ≥ 18 years; (4) sufficient outcome data for the calculation of leak rates and (5) sample size ≥ 5. The quality of included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2.
Results: Our literature search yielded 146 potential studies, of which 14 studies were included for meta-analysis after screening and exclusions. There were eleven prospective and three retrospective studies. The pooled anastomotic leak rate when ICG was used was found to be 10%. Pooled sensitivity and specificity for leak detection were 0.83 (0.70 to 0.93) and 0.60 (0.55 to 0.66), respectively. When studies involving intraoperative modifications were removed, pooled sensitivity and specificity were only marginally changed to 0.75 (0.51 to 0.91) and 0.67 (0.55 to 0.77), respectively. The diagnostic odds ratio was found to be 5.68 (2.29 to 14.10) across all studies and 5.06 (0.93 to 27.55) when intraoperative interventions were excluded. Only three trials included a control group, giving a sample size of 251. In studies with a comparator group, ICG was associated with an 87% reduction in the risk of anastomotic leak [OR: 0.13 (0.03-0.50)].
Conclusions: In non-randomized trials, the use of ICG as an intraoperative tool for visualizing vascular perfusion and conduit site selection, is promising. However, poor data quality and heterogeneity in reported variables limits cross-study comparisons and generalizability of findings. Randomized, multi-center trials are needed to account for independent risk factors for leak rates and to better elucidate the impact of ICG in predicting and preventing anastomotic leaks.
Presented at the SAGES 2017 Annual Meeting in Houston, TX.
Abstract ID: 86448
Program Number: P416
Presentation Session: iPoster Session (Non CME)
Presentation Type: Poster