Key result
Clinically relevant bleeding linked to higher mortality in cancer patients, with intracranial hemorrhage raising risk ~317%.
Why the study?
While cancer-associated venous thromboembolism is widely studied, clinically relevant bleeding remains an underestimated threat with limited large-scale data.
What is the incidence of clinically relevant bleeding and its association with mortality in individuals with cancer?
Cohort (n=2,455,332)
Yes
What is the incidence of clinically relevant bleeding and its association with mortality in individuals with cancer?
Clinically relevant bleeding is a frequent complication in cancer patients, particularly those on anticoagulants, and is strongly associated with increased mortality.
Answers are grounded in this paper and cite their sources.
CRB warrants vigilance in cancer patients; leaves open whether prevention improves survival in prospective trials.
Cancer care is often complicated by coagulopathy leading to thrombosis and bleeding. While venous thromboembolism (VTE) has been extensively studied, bleeding remains an underestimated threat. To address this knowledge gap, we leveraged the Epic Cosmos database to determine the impact of cancer-associated clinically relevant bleeding (CRB) in 2 455 332 individuals with incident cancer encounters from 215 US health systems between 2018 and 2024. In the 12 months leading up to the cancer diagnosis, the period prevalence of CRB was 5.3% in the overall cohort, highlighting CRB as a common early presentation preceding cancer diagnosis. In the 12-month window after cancer diagnosis, the cumulative incidence of CRB was 6.3% in the overall cohort, 7.0% in those receiving systemic antineoplastic therapy, and up to 15.0% among individuals on certain baseline anticoagulants. In addition to age, sex, and race, factors appreciably associated with CRB on multivariable analysis included cancer types, metastatic disease, systemic therapy, anticoagulant or antiplatelet use, higher comorbidity index, recent hospitalization, recent history of bleeding or VTE, and selective laboratory features (hemoglobin, platelets, albumin, eGFR). Furthermore, the onset of CRB as a time varying covariate was independently associated with mortality in multivariable adjusted analysis: intracranial hemorrhage (HR 4.17, 95% CI 4.11-4.24); gastrointestinal bleeding (HR 2.52, 95% CI 2.50-2.55); other bleeding (HR 2.46, 95% CI 2.44-2.49). Together, these findings reveal a substantial and overlooked risk of bleeding in cancer care and underscore a need for validated bleeding risk models that can be integrated alongside VTE prediction tools to guide personalized thromboprophylaxis decisions.
No takes yet. Share an insight, caveat, or question.
Lim et al. (2026) conducted a cohort in Cancer (n=2,455,332). Clinically relevant bleeding vs. No clinically relevant bleeding was evaluated on Mortality (HR 4.17, 95% CI 4.11-4.24). Clinically relevant bleeding was independently associated with mortality in cancer patients, including intracranial hemorrhage and gastrointestinal bleeding.
Synapse has enriched 10 related papers. Consider them for comparative context: