Abstract 8: Breast Reconstruction Pre-operative Risk Assessment: Applying the Risk Calculator to Define High Risk Patients

Abstract

PURPOSE: In the last decade an increasing number of patients with breast cancer have undergone breast reconstruction, including high risk ones. Patients with multiple comorbidity factors have been associated with increased complications. However, how should we define a high risk patient and should their reconstruction plan follow the same of a lower risk patient? The BRA Score has been proposed to calculate the pre-operative risk. Our aim was to correlate the validity of BRA Score as a pre-operative risk calculator for its practical use. METHODS: Patients that underwent different types of breast reconstruction had their pre-operative risk retrospectively calculated per breast using the BRA Score, and the actual complications they developed were collected. From the BRA Score we calculated risk of overall complication based on MROC (Risk-MROC) and TOPS (Risk-TOPS), surgical site infection risk (SSI-Risk), and 30 day reoperation risk (Reop-Risk). Data gathered from patient charts included post-operative overall complications (PO-Comp), surgical site infection (SSI), and reoperations due to complications. The following groups were considered: group 1, reconstructed breasts that had the analyzed complication; group 2, breasts without the complication. The ROC curve was used to evaluate the calculated risk as a complication predictor test. RESULTS: Charts of 389 breast reconstructions from 255 patients were evaluated. Compared to Group 2, Group 1 had a significantly higher Risk-MROC (20.8 ± 11.12 vs 15.24 ± 9.16, p≤0.01), Risk-TOPS (19.7 ± 7.28 vs 15.5 ± 6.56, p≤0.01), and Reop-Risk (7.48 ± 3.27 vs 6.22 ± 5.22, p≤0.01); and similar SSI-Risk (3.75 ± 2.3 vs 3.94 ± 2.38, p=0.96). As tests for predicting the PO-Comp, Risk-MROC and Risk-TOPS were adequate, with areas under the ROC curve of 0.662 and 0.669, respectively. For predicting the reoperations, Risk-MROC, Risk-TOPS, and Reop-Risk presented areas of 0.666, 0.691, and 0.652, respectively. A predicted risk of 25.5% using Risk-MROC and Risk-TOPS would provide a specificity of 79% and 89%, respectively. CONCLUSIONS: In this patient population, the BRA Score was a helpful tool to predict overall complications and reoperations. The calculator was not found to be useful in predicting surgical site infection. An overall risk of 25.5% derived from either the MROC or TOPS database would provide high specificity in determining a very high risk breast reconstruction patient. Patients with such high pre-operative risk may benefit from modifications in the breast reconstruction treatment plan to lower the complication rate. By using BRA Score, we can reliably predict the possible outcome of the reconstructions which can be used to better counsel the patient.