Intraoperative Prediction of Mastectomy Skin Flap Necrosis using Standardized ICG Grading and Quantitative Perfusion Mapping.

BACKGROUND

Mastectomy skin flap necrosis (MSFN) remains a common complication after immediate breast reconstruction. Indocyanine green (ICG) angiography enables real-time perfusion assessment, but lack of standardized interpretation limits clinical utility. This study evaluates a standardized qualitative grading system combined with quantitative perfusion (QP) mapping for intraoperative risk stratification.

PATIENTS AND

METHODS

A single-center cohort study was conducted in patients undergoing skin- or nipple-sparing mastectomy with immediate reconstruction. Intraoperative ICG angiography was performed, with QP measurements obtained across breast quadrants and the nipple-areolar complex (NAC). A five-point qualitative "Yale Grade" was assigned. The primary outcome was MSFN within 30 days. Model performance was assessed via receiver operating characteristic analysis and multivariable logistic regression.

RESULTS

The cohort included 185 reconstructions (104 patients), with an MSFN rate of 24.3%. MSFN incidence and severity decreased stepwise with improving Yale Grade (p < 0.001). Breasts with MSFN demonstrated significantly lower QP across all quadrants (p ≤ 0.001), with progressive decline by necrosis severity. Average QP demonstrated the highest discrimination (AUC 0.792) with a threshold of 49.0 (63.6% sensitivity, 83.2% specificity). Minimum QP also performed well (AUC 0.759) at 33.5 (72.7% sensitivity, 73.0% specificity). Yale Grade ≤ 3 demonstrated high specificity (89.2%). For NAC necrosis, average QP demonstrated the highest accuracy (AUC 0.899). On multivariable analysis, perfusion metrics remained independently predictive, whereas clinical variables were not.

CONCLUSIONS

A standardized ICG grading system combined with QP mapping reliably predicts MSFN. Thresholds of average QP ≥ 49 and minimum QP < 33.5, integrated with Yale Grade ≤ 3, provide a practical, tiered framework for intraoperative decision-making.