Confirmation of double-peaked time distribution of mortality for breast cancer in a population-based study

Abstract

593 Background: Double-peaked time distributions of mortality hazard function have been reported for breast cancer patients from Western populations treated with mastectomy alone and is thought to reflect accelerated tumor growth at metastatic sites mediated by angiogenesis after primary tumor removal. Similar data is not available for Asian populations. We sought to investigate if differences exist in the pattern of mortality hazard function between Western breast cancer patients and Asian patients in Singapore treated with mastectomy alone which may suggest underlying differences in tumor biology between the two populations. Methods: We performed a retrospective cohort study of 3809 female unilateral breast cancer patients registered in the Singapore Breast Cancer Registry between October 1994 to June 1999. Data regarding patient demographics, tumor characteristics and death data were available. Overall survival curves were calculated using the Kaplan-Meier method and compared by the log-rank test. Death-specific hazard rate was calculated as the conditional probability of dying in a time interval, given that the patient was alive at the beginning of the interval. The life table method was used to calculate the yearly hazard rates. All tests were two-sided and significant at the 5% level. Results: In this cohort, 1169 patients (45.9%) had mastectomy alone without further adjuvant therapy. Demographics were: median age 55; 87.4% Chinese; 62.9% postmenopausal; 51.6% hormone receptor positive; 70.3% node negative; 22.8% high histological grade; 76.9% stage I or II. A double-peaked mortality hazard pattern, with a first major mortality rise reaching the maximum at the 2nd - 4th year after diagnosis and a second small mortality rise near the 9th year was observed. Analyses by subgroups showed a similar pattern in node positive or high grade tumors but not in node negative or well to moderately differentiated tumors. A double peaked pattern was seen regardless of menopausal status. The double-peaked pattern observed in our patients was quantitatively and qualitatively similar to that reported by Western series. Conclusions: Our study confirms a double-peaked process in Asian patients and gives further support to the tumor dormancy hypothesis. No significant financial relationships to disclose.