Purpose: Up to 19% of all colorectal resections develop clinically apparent insufficiencies. Insufficient perfusion of the anastomosis is recognized as an important risk factor. As tissue perfusion can be objectified intraoperatively using laser fluorescence angiography (LFA), its effect on the rate of anastomotic complications was evaluated in a retrospective matched-pairs analysis.
Methods: Between 2003 and 2008, all anastomosis or resection margins in colorectal cancer resections were investigated intraoperatively using LFA (LFA group). Patients with colorectal cancer resections between 1998 and 2003 without LFA served as the control group. Four hundred two patients were matched for age, T-stage, type of resection and anastomosis, defunctioning stoma, administration of blood, emergency conditions, and body mass index. Statistical analysis was performed using the Fisher and the Wilcoxon tests.
Results: Twenty-two surgical revisions were necessary due to anastomotic leakage, seven (3.5%) in the LFA group and 15 (7.5%) in the control group. Subgroup analysis revealed that in elective resections the rate of revision was 3.1% (LFA group) and 7.7% (control group) (p = 0.04, risk of revision (ROR) reduced by 60%). In patients older than 70 years, the rate of revision was 4.3% (LFA group) compared to 11.9% (control group) (p = 0.04, ROR reduced by 64%). After hand-sewn anastomosis, the rate of revision was 1.2% (LFA group) and 8.5% (control group) (p = 0.03, ROR reduced by 84%). Hospital stay was significantly reduced in the LFA group (Wilcoxon test; p = 0.01).
Conclusion: There was an overall reduction in the absolute revision rate of 4% in the LFA group and a significantly reduced rate of revision in the subgroup analysis of patients undergoing elective colorectal resections, in patients older than 70 years and in patients with hand-sewn anastomosis. This demonstrates that LFA is a method that may significantly reduce not only the rate of severe complications in colorectal surgery but also the hospital length of stay.