Abstract 54: Does Fluorescent Angiography Reduce Postoperative Complications After Breast Reconstruction? A Propensity-score Matched, Cohort Analysis

Abstract

Purpose: There remains a paucity of randomized clinical trials or high-grade evidence substantiating the use of florescent angiography following reconstructive breast surgery. We utilize a national, multi-institutional, surgical outcomes database to design a pseudorandomization study that evaluates the comparative effectiveness of florescent angiography during breast surgery. Methods: Using the American College of Surgeon National Quality Improvement Program database, we performed a retrospective, exposure-matched, cohort analysis over a 13-year period (2005-2017). International Classification of Disease-9/10 codes along with Current Procedure Terminology codes were used to identify all patients undergoing mastectomy with immediate implant-based, expander-based, or autologous breast reconstruction. We stratified patients into two cohorts: those receiving and not receiving fluorescent angiography (FA & nFA groups). We used propensity-score matching to mitigate selection bias and reduce baseline heterogeneity between >50 demographic, clinical, and perioperative variables. Using a caliper distance of 0.2 and nearest-neighbor matching technique, we identified two identical matches from nFA for each patient in FA (1:2 matching). Our outcomes of interest were 30-day rates of postoperative morbidity and mortality. Results: 113,504 patients met our initial inclusion criteria; 1.6% (1,809) of patients underwent concurrent FA. After matching we were left with 5,421 patients; 1,809 (33.3%) patients in FA and 3,612 (66.6%) patients in the nFA groups. 21.7% (1,174) patients underwent bilateral mastectomies. 66.3% (3,592) had a simple-total mastectomy, 14.1% (765) had modified-radical mastectomies and 5.9% (321) had skin sparing mastectomies. 47.1% (2,554) of reconstructions were expander-based, 14.4% (779) were implant-based, and 43.3% (2341) were autologous reconstructions (latissimus dorsi, transverse rectus abdominus flaps and free flaps). Before matching, the FA group on baseline had a higher proportion of co-morbidities and higher proportion of autologous reconstructions. Post-match diagnostics demonstrated successful matching on all baseline variables. Overall, 30-day postoperative wound complication rates were similar between both cohorts (FA: 4.8% [87], nFA: 4.8% [174], P =0.990). Rates of superficial surgical site infections (FA: 2.4% [88], nFA: 2.0% [37], P =0.366), deep surgical site infections (FA: 1.1% [19], nFA: 0.9% [33], P =0.626), organ space infections (FA: 1.0% [18], nFA: 0.8% [29], P =0.472) and wound/flap dehiscence (FA: 1.4% [25], nFA: 1.0% [36]%, P =0.205) were also similar between both groups. Other 30-day postoperative complication rates (all P >0.05) and reoperation rates (FA: 8.9% [161], nFA: 8.5%, P = 0.645) remained similar between both groups. There was no postoperative mortality in either group. Conclusion: This national appraisal of surgical outcomes demonstrates that fluorescent angiography did not decrease postoperative morbidity following reconstructive breast surgery. Well-powered, prospective, randomized trials may help elucidate the effect of such technology within specific surgical cohorts or high-risk groups.