Thrombotic and hemorrhagic events and their impact on mortality among adults with newly diagnosed acute myeloid leukemia Free

Introduction

Patients with acute myeloid leukemia (AML) are at increased risk of both thrombosis and bleeding and both are seen as presenting events or early complications. AML diagnosis is usually established during hospitalization so these events affect early hospital course and initial inpatient treatment. Data regarding the risk and consequences of in-hospital thrombosis and bleeding in patients with newly diagnosed AML are limited. Our objective was to define the occurrence of symptomatic venous thrombotic and hemorrhagic events during initial inpatient admission of patients with newly diagnosed AML, and determine their impact on short-term mortality.

Methods

Adult patients admitted to medical inpatient services for >1 midnight with a new AML diagnosis during hospitalization were identified between 2016-2022 from 6 hospital systems and 15 individual hospitals in the United States located in Michigan, Minnesota, North Carolina, Pennsylvania, Texas, and Vermont. Symptomatic venous thromboembolism (VTE, including deep vein thrombosis [DVT] and pulmonary embolism [PE]) and major and clinically relevant bleeding events were identified using previously validated computable phenotypes for events present on admission and during hospitalization (occurring ≥24 hours after admission), following ISTH definitions (PMC10277582). Subjects could bleed at more than 1 anatomical site. The primary outcome was inpatient mortality during the index hospitalization. Multivariable logistic regression models were used to estimate the adjusted odds ratio (aOR) and 95% confidence interval (95%CI) for the association of VTE and bleeding events with in-hospital mortality.

Results

We identified 1,663 patients with a new diagnosis of AML. Median age was 65 (IQR 56-73), 57.8% were male, and 77% were non-Hispanic White. Overall, 127 (7.6%) of the 1,663 patients died during their first hospitalization.

VTE occurred in 44 of 1,663 patients (2.6%), including 29 (1.7%) events identified at admission and 15 (0.9%) identified during hospitalization. The 44 VTE events occurred at the following sites: 15 (34.1%) lower extremity DVT, 14 (31.8%) upper extremity DVT, and 15 (34.1%) PE. The absolute risk of death was 15.9% for patients with VTE events versus 7.4% for those without VTE. Adjusting for age, sex, race, and Elixhauser comorbidity index, patients with VTE events had 97% higher odds of hospital mortality (aOR=1.97, 95%CI 0.75-4.49).

Bleeding occurred in 399 of 1,663 patients (24%), including 217 (13%) bleeding identified at admission and 201 (12.1%) identified during hospitalization. The 399 bleeding events included 105 (26.3%) gastrointestinal, 97 (24.3%) nasal, 76 (19%) genitourinary, 60 (15%) central nervous system (CNS), and 30 (7.5%) pulmonary bleeding. The absolute risk of dying during the admission was 16.8% in those with bleeding versus 4.7% for those without bleeding. Adjusting for age, sex, race, Elixhauser comorbidity index, and anticoagulation and antiplatelet therapy on admission, patients with bleeding had a 3.5-fold increase in hospital mortality (aOR=3.51, 95%CI 2.38-5.20). The risk of death was highest in patients with CNS (36.7%, aOR=9.16, 95%CI 4.79-17.27), pulmonary (36.7%, aOR=11.20, 95%CI 4.74-25.62), and gastrointestinal bleeding (24.8%, aOR=5.07, 95%CI 2.88-8.77), and lowest for nasal bleeding (12.4%, aOR=2.66, 95%CI 1.29-5.12).

Conclusion

The initial diagnosis and hospitalization of patients with AML is associated with a higher risk of bleeding as compared to thrombosis. Greatly under-appreciated is the burden and impact of bleeding in this population, with a quarter of patients experiencing bleeding severe enough to warrant medical attention. Remarkably, CNS bleeding was more common than VTE. In addition, bleeding is associated with greater than 3-fold increased risk of in-hospital mortality, and the increased risk of death persisted across bleeding sites, though was particularly high for patients experiencing a gastrointestinal or CNS bleed. This highlights the need to develop and implement proactive measures to prevent bleeding in this population and to carefully consider the risks and benefits of interventions which potentially increase bleeding such as pharmacologic VTE prophylaxis.