Observational study of coagulation activation in early breast cancer: Development of a prognostic model based on data from the real world setting

Abstract

© 2018 The Author(s). Background: Cancer and coagulation activation are tightly related. The extent to which factors related to both these pathologic conditions concur to patient prognosis intensely animates the inherent research areas. The study herein presented aimed to the development of a tool for the assessment and stratification of risk of death and disease recurrence in early breast cancer. Methods: Between 2008 and 2010, two hundreds thirty-five (N: 235) patients diagnosed with stage I-IIA breast cancer were included. Data on patient demographics and clinic-pathologic features were collected in course of face-to-face interviews or actively retrieved from clinical charts. Plasma levels of plasminogen activator inhibitor type 1 (PAI-1), fragment 1 + 2 (F1 + 2), thrombin antithrombin complex (TAT), factor VIII (FVIII), and D-dimer (DD) were measured at breast cancer diagnosis and prior to any therapeutic procedure, including breast surgery. The risk of death was computed in terms of overall survival (OS), which was the primary outcome. For a subset of patients (N = 62), disease free survival (DFS) was also assessed as a measure of risk of disease recurrence. Results: Median follow up was 95 months (range 6-112 months). Mean age at diagnosis was 60.3 ± 13.4 years. Cancer cases were more commonly intraductal carcinomas (N: 204; 86.8%), pT1 (131; 55.7%), pN0 (141; 60%) and G2 (126; 53.6%). Elevated levels of PAI-1 (113; 48.1%) represented the most frequent coagulation abnormality, followed by higher levels of F1 + 2 (97; 41.3%), DD (63; 27.0%), TAT (34; 40%), and FVIII (29; 12.3%). In univariate models of OS, age, pT, DD, FVIII were prognostically relevant. In multivariate models of OS, age (p = 0.043), pT (p = 0.001), levels of DD (p = 0.029) and FVIII (p = 0.087) were confirmed. In the smaller subgroup of 62 patients, lymph node involvement, percent expression of estrogen receptors and levels of FVIII impacted DFS significantly. Conclusions: We developed a risk assessment tool for OS including patient- and cancer-related features along with biomarkers of coagulation activation in a cohort of early BC patients. Further studies are warranted to validate our prognostic model in the early setting and eventually extend its application to risk evaluation in the advanced setting for breast and other cancers.