Predicting the risk of local recurrence in patients with breast cancer: an approach to a new computer-based predictive tool.

Objective: To develop a new web-based tool, designated IBTR!, which integrates prognostic factors for local recurrence (LR) into a model to predict the 10-year risk of LR after breast conserving surgery (BCS) with or without radiation therapy (RT) with the goal of assisting with patient counseling and medical decision-making.

Methods: All available randomized trials of BCS alone versus BCS plus RT, meta-analyses, and institutional reports were reviewed to identify the principal prognostic factors for LR after breast-conserving therapy. Patient age, margin status, lymphovascular invasion (LVI), tumor size, tumor grade, use of chemotherapy, and use of hormonal therapy were found to consistently and significantly impact LR across multiple studies. Based upon a composite analysis of the relevant published randomized and nonrandomized studies, relative risk (RR) ratios were estimated and assigned to each prognostic category. These RR ratios were entered into a mathematical model with the 10-year baseline rates of recurrence with and without RT, 7% and 24%, respectively, to predict patient-specific LR risk.

Results: Individual data entered into this computer model with regards to patient age, margin status, LVI, tumor size, tumor grade, use of chemotherapy, and use of hormonal therapy will generate patient-specific predicted 10-year LR risk with and without RT. A graphic representation of the relative risk reduction with RT will also be displayed alongside the numerical display.

Conclusion: IBTR! is a first attempt at a computer model incorporating LR prognostic factors in an evidence-based fashion to predict individual LR risk and the potential additional benefit from RT.