Time-Varying Prognostic Impact of Hormone Receptors for Breast Cancer Patients

W. Yin,J. Lu,G. Liu,G. Di,J. Wu,Z. Shen,Z. Shao

DOI: https://doi.org/10.1200/jco.2008.26.15_suppl.22131

IF: 45.3

2008-01-01

Journal of Clinical Oncology

Abstract:22131 Background: In most studies, the effect of steroid hormone receptor (HR) on prognosis was illustrated by survival analysis, which may obscure the changes of event probability over time. However, such information is highlighted using the hazard function. We sought to investigate the time-varying prognostic impact of HR for breast cancer patients by use of hazard function. Similar data is still not available at present. Methods: We performed a retrospective study of 2143 female unilateral breast cancer patients undergoing surgery in our hospital from 1992 to 2003. The primary endpoint was recurrence-free survival (RFS), defined as the time from surgery to the earliest occurrence of local-regional or distant relapse (including contralateral breast cancer), or death from any cause. Results: According to different HR status, patients were categorized into the three subgroups as follows: double-positive (ER+/PR+), single-positive (ER+/PR− and ER−/PR+) and double-negative (ER−/PR−). Both univariate and multivariate survival analysis indicated that tumor size, axillary nodal status, HR status were prognostic factors for RFS (P <0.05). Annual recurrence hazard curve showed double-peaked pattern irrespective of HR status. In different subgroups, however, discrepancies existed regarding the time of hazard peaks. Double-negative patients showed an early major surge peaking at the 2nd year after surgery and a slight increase at the 8.5th year. As to double-positive patients, bimodal pattern was indicated by a gradual increase in risk until 2.5 years after surgery, followed by a modest decline until year 7.5, and then a second rise to a maximum at 9.5 years. The corresponding peaks for the single-negative patients were at the 1.5th year and at the 10th year. In other words, double-negative breast cancers were more likely to recur early compared with double-positive ones, while the recurrence pattern for single-negative tumors was intermediate between double-negative and double-positive counterparts. Conclusions: According to the prognostic impact of HR on time distribution, we could formulate different strategies for different patients in the treatment and follow-up, which provides the possibility of delaying the recurrence hazard to improve the therapeutic effect. No significant financial relationships to disclose.