Association between anticoagulant-related bleeding and mortality in patients with hematological malignancies and cancer-associated venous thromboembolism
Tzu-Fei Wang a), Suhong Luo b), c), Brian F. Gage c), Martin W. Schoen b), d), Amber Afzal c), Kenneth Carson e), Su-Hsin Chang c), Amir Mahmoud c), Kristen M. Sanfilippo b), c)
a) Department of Medicine, University of Ottawa at The Ottawa Hospital and Ottawa Hospital Research Institute, Ottawa, ON, Canada
b) Saint Louis Veterans Affairs Medical Center and Department of Medicine, Washington University School of Medicine, Saint Louis, MO, USA
c) Department of Medicine, Washington University School of Medicine, Saint Louis, MO, USA
d) Department of Medicine, St. Louis University Medical Center, St. Louis, MO, USA
e) Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, USA
Abstract
Introduction
Patients with hematological malignancies are at an increased risk of severe bleeding. Anticoagulant (AC) therapy further increases this risk. Mortality after these bleeds is unclear and may differ by bleeding site.
Aim
To evaluate the association between bleeding and mortality in patients with hematological malignancies prescribed AC therapy for cancer-associated venous thromboembolism (VTE).
Methods
In a nationwide cohort of US Veterans (2012−2020), we identified patients with hematological malignancies and cancer-associated VTE prescribed AC therapy. Bleeding events were identified by a previously validated algorithm using hospitalization International Classification of Disease (ICD) codes. Within 12 months of AC therapy initiation, we evaluated the association between bleeding and mortality using multivariate Cox regression models, with AC-related bleeding analyzed as a time-varying covariate.
Results
The cohort included 1825 patients. Within 12 months of starting AC therapy, 123 (6.7 %) had bleeding events and 162 (8.9 %) patients died. Patients with bleeding events were more likely to have anemia, history of bleeding, aspirin use, chemotherapy use, and frailty. A multivariable Cox proportional-hazard model showed that AC-related bleeding was associated with tripled mortality rate (aHR 3.26, 95 % CI 1.96–5.45). When stratified by bleeding site, intracranial bleeding was associated with the highest risk of death (aHR 13.41, 95 % CI 4.13–43.48), followed by gastrointestinal bleeding (aHR 4.55, 95 % CI 2.48–8.35).
Conclusion
In this cohort of patients with hematological malignancies and newly diagnosed VTE initiated on AC therapy, bleeding was associated with an increased risk of mortality within 12 months. Association was highest with intracranial and gastrointestinal bleeding.