Background: The objective of this systematic review and meta-analysis was to pool randomized trials of patients undergoing left-sided colorectal anastomosis, comparing the use of fluorescence perfusion imaging versus visual inspection in reducing anastomotic leaks.
Methods: Databases searched included PubMed, Cochrane Library, Scopus, CINHAL (EBSCO), and Google Scholar based on the concepts: randomized, colorectal, anastomotic leak, and fluorescence imaging. The risk of bias was assessed using RoB2 and the certainty of the evidence with the GRADE Pro tool. The analysis used the log odds ratio for dichotomous data with 95% confidence intervals. Back-transformation of the log odds to odds ratio was performed for the summary of findings. All syntheses used the Random-effects model.
Results: Six randomized trials were included with 1949 patients and 204 events (leaks). Three trials included exclusively rectal cancer patients, while the other three involved benign and malignant pathologies of the sigmoid and rectum. The use of ostomy and preoperative radiation was variable. None of the studies had a high risk of bias. The pooled odds ratio for anastomotic leak reduction with Indocyanine Green (ICG) fluorescence was 0.586 (95% CI: 0.434-0.792). An absolute reduction of 4.7% in leak rates was observed, with no statistical heterogeneity (I2 = 0; p = 0.529). Due to clinical heterogeneity, the quality of evidence was rated moderate.
Conclusions: The use of ICG is associated with reduced leak rates following left-sided colorectal anastomosis with moderate confidence. ICG may be considered a standard of care given the clinically significant benefit in decreasing anastomotic leaks.
