Long-Term Outcomes After Autologous or Tissue Expander/Implant-Based Breast Reconstruction and Postmastectomy Radiation for Breast Cancer.

PURPOSE

The toxicity profile of breast reconstruction with postmastectomy radiation therapy (PMRT) varies by technique and timing, and long-term data are limited. We compared rates of complications requiring reoperation (CRR) and reconstruction failure (RF) between immediate autologous reconstruction (I-AR), immediate tissue expander/implant reconstruction (I-TE/I), delayed autologous reconstruction (D-AR), and delayed tissue expander/implant reconstruction (D-TE/I) in patients receiving PMRT.

METHODS AND MATERIALS

Patients who received autologous reconstruction (AR) or tissue expander/implant reconstruction (TE/I) and PMRT between 2000 to 2008 were included. Reconstruction was immediate if performed on the same day as mastectomy followed by PMRT (I-AR or I-TE/I) or delayed if after PMRT (D-AR and D-TE/I). CRR was defined as an unplanned return to the operating room for infection, dehiscence, necrosis, hematoma, or hernia (with AR) and extrusion, leak, or contracture (with TE/I). RF was defined as unplanned conversion to another reconstruction technique or to flat chest wall. Cumulative incidence of CRR and RF was calculated using Kaplan-Meier and compared using the log-rank test. Logistic regression was used to identify variables associated with CRR and RF.

RESULTS

Two hundred four patients were included. Median follow-up was 8 years. There were 127 AR cases (63%) and 77 TE/I cases (38%). At 5 years, CRR was 18%, 38%, 34%, and 70% (P = .010) and RF was 4%, 22%, 7%, and 56% (P < .0001) for I-AR, I-TE/I, D-AR, and D-TE/I, respectively. On multivariate analysis, TE/I (hazard ratio [HR] 2.0; P = .011), body mass index ≥30 (HR 3.9; P = .002), and smoking (HR 2.7; P = .001) were significant predictors for CRR, and TE/I (HR 6.6; P < .0001), diabetes (HR 4.1; P = .044), and hypertension (HR 3.5; P = .005) were significant for RF. When excluding RF because of infection, the rate of RF was not significantly different among the 4 groups (P = .23).

CONCLUSIONS

With PMRT, TE/I reconstruction in the immediate and delayed setting is associated with higher CRR and RF compared with AR. Patient factors should guide selection of technique. Efforts to reduce rates of RF with TE/I should focus on minimizing risks for infection.