Factors Associated with Reconstruction Failure and Major Complications After Postmastectomy Radiation to a Reconstructed Breast

Abstract

Purpose: Postmastectomy radiation therapy is known to increase risk of complications in the reconstruction setting. We aim to identify the variables associated with reconstruction failure and other major complications.

Methods and materials: A prospectively collected institutional database was queried for patients with up to stage IIIC breast cancer treated from 2000 to 2017, undergoing mastectomy, immediate implant or autologous tissue reconstruction, and radiation to the reconstructed breast within 1 year of surgery. Reconstruction failure was defined as complication requiring surgical revision or implant removal. Additional major complications were defined as any infection, contracture, necrosis, or fibrosis. Covariates of interest included age, body mass index, smoking status, stage, hormone receptor and HER2 status, systemic therapy timing, radiation technique, nodal irradiation, and interval between surgery and start of postmastectomy radiation therapy. Differences in complication rates were assessed with χ² or Fisher exact tests. Competing risk regression was used to estimate hazard ratios; covariates were included one at a time to avoid over adjustment.

Results: A total of 206 reconstructed breasts in 202 patients resulted from our initial query, with 139 treated with intensity-modulated radiation therapy (IMRT) and 67 treated with conventional radiation therapy (CRT). Median follow-up was 45 months (range, 4-210 months); patient cohorts were generally similar. Eight patients were excluded from toxicity analysis for insufficient follow-up (<2 years). Overall, reconstruction failure and major complication rates were significantly lower in the IMRT group. Reconstruction failure rates were 3.0% for IMRT versus 16.4% for CRT (P = .002), and major complication rates were 6.8% for IMRT versus 24.6% for CRT (P < .001). On univariate analysis, CRT was significantly predictive of implant failure (hazard ratio, 5.54; P = .003) and increased complication rates (hazard ratio, 3.83; P = .001). Significance persisted on multivariable analysis. Survival outcomes were similar, with no difference in 2 year overall survival (P = .12) and local recurrence (P = .41).

Conclusions: Using IMRT may improve reconstruction outcomes over CRT, with significantly lower reconstruction failure and complication rates without compromising local control or survival.