Nomogram Predicting Locoregional Recurrence to Assist Decision-Making of Postmastectomy Radiation Therapy in Patients With T1-2N1 Breast Cancer.

Abstract

Postmastectomy radiation therapy (PMRT) had heterogeneous effects on survival outcome of patients with T1-2N1 breast cancer. A reliable model to estimate individuals' risk of locoregional recurrence (LRR) and the potential benefit derived from PMRT is needed. We retrospectively analyzed 1141 patients with T1-2N1 breast cancer who underwent mastectomy between January 2001 and December 2012. Based on the Fine and Gray competing risks regression in 623 unirradiated patients, a nomogram predicting LRR was conducted for risk quantification. Decision tree analysis was performed for patient grouping. The impact of PMRT was evaluated among 3 subgroups. With a median follow-up of 74.9 months, the 5-year cumulative incidence of LRR, distant recurrence (DR) and breast cancer mortality (BCM) were 3.9%, 8.8%, and 6.0%, respectively, for the entire cohort. Based on nomogram scores, patients were classified into 3 risk groups in decision tree analysis. In the high-risk group, PMRT was found to be associated with a 12.7% risk reduction of 5-year LRR, 9.2% risk reduction of 5-year DR, and 7.0% risk reduction of 5-year BCM, whereas it was not significantly associated with LRR, DR, or BCM in low- and intermediate-risk groups. The nomogram performed individualized risk quantification of LRR in patients with T1-2N1 breast cancer. A newly identified patient subgroup with high risk of LRR were found to derive survival benefit from PMRT.