Predicting red breast syndrome after ADM-assisted breast reconstruction: An exploratory scoring model using routine hematologic markers.

BACKGROUND

Red breast syndrome (RBS) is a challenging complication of acellular dermal matrix (ADM)-assisted breast reconstruction. Its incidence and hematologic predictors in patients receiving neoadjuvant chemotherapy (NAC) remain poorly characterized.

METHODS

This retrospective cohort study included 70 patients who underwent skin-sparing mastectomy with immediate implant-based reconstruction following NAC (January 2011-May 2025) at a single tertiary institution. Patients were stratified into ADM (n = 47) and Non-ADM (n = 23) groups. Within the ADM group, Firth penalized logistic regression with backward elimination identified preoperative variables independently associated with RBS. Optimal cutoffs were determined by the Youden Index. A composite hematologic scoring model (WBC + PLT + Ki-67 High, cutoff ≥2) was developed and internally validated using bootstrap resampling (2,000 iterations).

RESULTS

RBS occurred in 31.9% (15/47) of ADM patients versus 0% in Non-ADM patients (p = 0.001). The final model identified WBC (OR 1.235, 95% CI 1.066-1.502, p = 0.004), PLT (OR 1.013, 95% CI 1.001-1.028, p = 0.037), and Ki-67 High (OR 8.975, 95% CI 1.356-155.173, p = 0.020) as independently associated with RBS. Individual biomarkers showed modest discrimination (AUC 0.605-0.641, all p > 0.10), while the composite score achieved AUC = 0.816 (p < 0.001; sensitivity 86.7%, specificity 71.9%, NPV 92.0%). Bootstrap-corrected AUC was 0.817 (95% CI 0.636-0.959; optimism=-0.001). No significant temporal trend in RBS incidence was observed (Cochran-Armitage, p = 0.527).

CONCLUSIONS

RBS incidence is substantially elevated in post-NAC ADM patients. This exploratory, single-center study develops a hypothesis-generating combination score for preoperative RBS risk stratification using routine laboratory and pathology assessment. External validation in prospective multi-institutional cohorts is required before clinical application.