Sobrevida de pacientes com câncer de mama tratadas no hospital das clínicas da UFMG entre 2001 e 2008

Abstract

INTRODUCTION: Several characteristics are associated with the prognosis of breast cancer, such as staging, positivity for hormone receptors, histological grade, race and socioeconomic factors. It is debatable whether age is an independent prognostic factor for the disease. OBJECTIVE: This study evaluated the survival of patients with breast cancer surgically treated at the Hospital das Clínicas (HC)-UFMG between 2001 and 2008. METHODS: Breast cancer cases stages I to III which underwent treatment in that period were identified in the Breast Pathology Laboratory of the Faculty of Medicine. A probabilistic record linkage with the database of the State Department of Health Mortality Information System (MIS) was performed in the period from 01/01/2001 to 31/12/2011, with the RecLink software. Then, medical records were consulted. Kaplan- Meier curves and Cox models were built for univariate and multivariate analysis. Hazard Ratios (HR) and 95% Confidence Intervals (CI) were calculated. The significance level was 0.05. Initially, 897 patients treated by the Unified Health System (SUS) or through private funding were analyzed. In a second step, we analyzed only the patients treated by the SUS, focusing on age as a prognostic factor for the disease (n=783). RESULTS: In the first analysis, there were 282 deaths, 228 of them from breast cancer (81%). The cause-specific survival at five years was 95.5% for stage I, 85.1% for stage II and 62.1% for stage III disease. Patients treated by the SUS had more advanced tumors at diagnosis and poorer survival in univariate analysis. In multivariate analysis, higher stage, higher histological grade and age over 70 years were associated with poorer survival. In the second step, changes in the relationship between age and survival were observed depending on the inclusion of different variables to the Cox models. In cause-specific analysis, patients aged 70 years and older and up to 35 years had higher mortality compared with patients aged 36 to 69 years (70 and older, HR=1.42 , CI=1.02-1.97; up to 35 years, HR=1.77, CI=1.06- 2.96). When tumor and patients characteristics were added to the model, the disadvantage of survival became non-significant for those aged up to 35 years but remained significant for the patients 70 years old and older (up to 35, HR=1.58, CI=0.90-2.75; 70 and older, HR=1.57, CI=1.07- 2.31). In other causes survival analysis, only being 70 years of age and older, being white, and having at least one comorbidity were associated with an increased risk of death. CONCLUSIONS: Patients treated by the SUS had worse survival, possible due to higher stage of the disease at diagnoses. Being 70 years old and older was an independent predictor of shorter survival. The association between age and survival is mediated by factors related to the patient and to the tumor. In future research, other factors must be taken into account, such as socioeconomic variables and treatments employed.