Factors Affecting Recurrence in Breast Cancer Using Cox Model

Background and The aim of this study was to estimate the disease-free survival rate, in female patients with breast cancer and determining the level of influencing factors. In a retrospective cohort study, the records of 377 patients attending Mashhad Omid hospital, Iran, spanning the years 2006 to 2011 were selected using convenience sampling. The patients were followed up until April 2014. Data analyses were done in SAS (ver.9.3) applying Cox’s semi-parametric regression method. The time after surgery until first recurrence or metastasis was considered as the dependent variable. From 377 patients (mean age: 49 years) 249 (0.66%) were found to have no any metastasis, while 128 (34%) had metastasis. Using Kaplan-Meier’s nonparametric method, the probability of the disease-free survival in 2920 days (8 years) follow-up was estimated to be 52.5% (mean time: 6 years and 3 months). There were significant differences between disease-free survivals in different tumor sizes and the involved lymph nodes and type of invasive breast cancer Also, increase of recur risk or metastasis was found to be associated with increases in size of tumor, in number of involved lymph nodes and in the level of cancer anti-gene CA 15-3 before first recur. The hazard ratio of the chance of metastasis occurrence while controlling other variables increased 19% against per unit increase in the size of tumor, 10% per involved number of lymph nodes, and 0.6 % per unit increase in CA 15-3 antigen. The CA 15-3 antigen was determined as the prognostic factor for occurrence of metastasis alongside tumor size and the number of involved nodes.