Risk Factors for Local Relapse and Inferior Disease-free Survival After Breast-conserving Management of Breast Cancer: Recursive Partitioning Analysis of 2161 Patients.

BACKGROUND: The purpose of this study was to analyze risk factors for ipsilateral in-breast relapse and inferior disease-free survival (DFS) after standard adjuvant whole-breast radiotherapy (± boost and systemic treatment) as part of a multimodal breast-conserving approach. PATIENTS AND METHODS: Decision trees were built through recursive partitioning analysis (RPA). The median follow-up for all 2161 patients was 114 months (9.5 years). RESULTS: Local relapse in the treated breast was uncommon (actuarial rates after 5 and 10 years were 2.7% and 5.8%, respectively). In RPA, the first split was related to age (52 years), with younger patients having a significantly higher risk of local relapse. The younger patients were stratified further by lymph node ratio (LNR). In patients older than 52 years, lack of endocrine treatment was associated with significantly higher risk. DFS was 80.7% at 10 years. The first split was caused by LNR, and the group with unfavorable LNR (> 0.20) could not be stratified further. Ten-year DFS in this group was as low as 50.6%. Patients with favorable LNR (0-0.20) could be stratified by additional risk factors, in particular primary tumor size. CONCLUSION: RPA is a suitable method to assign patients with early stage breast cancer to different risk groups, both regarding local relapse and DFS. Although age was a major risk factor for local relapse after breast-conserving management, LNR was associated with both endpoints. The systemic treatment approaches used in this study failed to provide satisfactory DFS in patients with LNR > 0.20 and 2 other subgroups.

Local relapse after breast-conserving surgery and radiotherapy: effects on survival parameters.

PURPOSE: This retrospective analysis of 1,610 women treated for breast cancer and 88 patients with local relapse aims to show the poor survival parameters after local failure and to evaluate risk factors and compare them with other studies and analyses published. PATIENTS AND METHODS: Between 1984 and 1997, 1,610 patients presenting with a total of 1,635 pT1-2 invasive and noninvasive carcinomas of the breast were treated at the authors' institution. The mean age was 57.1 years (range 25-85 years). Treatment protocols involved breast-conserving surgery with or without systemic therapy and whole-breast radiotherapy in all women, followed by a boost dose to the tumor bed according to risk factors for local recurrence. All axillary node-positive patients underwent systemic therapy (six cycles of classic CMF or tamoxifen 20 mg/day for 2-5 years). The time of diagnosis of local relapse was defined as time 0 for the survival curves after local failure. The association of clinicopathologic factors was studied using uni- and multivariate analyses. Survival and local control were calculated by the Kaplan-Meier actuarial method and significance by the log-rank test. RESULTS: After a mean follow-up of 104 months, 88 local failures were recorded (5.4%). Calculated from the time of diagnosis of local relapse, 5-year overall survival (OS) was 62.8%, metastasis-free survival 60.6%, and disease-specific survival 64.2%. In patients with failure during the first 5 years after treatment, the survival parameters were worse (OS 50.6%) compared to those who relapsed after 5 years (OS 78.8%; p < 0.028). Significances were also found for initial T- and N-stage and type of failure (solid tumor vs. diffuse spread). CONCLUSION: This analysis again shows that the survival parameters are worsening after local relapse, especially in case of early occurrence. In breast cancer treatment, therefore, the goal remains to avoid local failure.