Unknown risk factors contribute to risk for major bleeds during VKA therapy

The risk for major bleeding associated with prior minor bleeding during treatment with vitamin K antagonists appeared linked to fixed, still unknown, risk factors, according to study results.

Previous studies have shown that patients on vitamin K antagonists (VKAs) who experience minor bleeding are at increased risk for major bleeding; however, when researchers adjust for known risk factors — such as overanticoagulation, diabetes, diuretic use and malignancy — the increased risk persisted.

“This remaining increased risk indicates that not all risk factors for major bleeds were adjusted for, possibly because not all are known,” Nienke Van Rein, PharmD, clinical pharmacologist in the department of thrombosis and hemostasis of Einthoven Laboratory for Experimental Vascular Medicine at Leiden University Medical Center, and colleagues wrote. “More knowledge on the nature of these yet unknown risk factors may provide new leads on risk factors for major bleeds and hence improve prediction of who is at high risk of major bleeding during VKA treatment.”

Van Rein and colleagues used data from 26,130 patients (mean age at baseline, 70 years; 52% men) treated with VKAs at The Leiden Anticoagulation Clinic from 2003 to 2013. Researchers evaluated the starting date of VKA therapy, international normalized ratio (INR) target range (2.5-3.5 or 3-4), type of VKA and comedication.

Researchers classified a major bleed as bleeding symptoms in a critical area or organ that required a blood transfusion or led to hospitalization or death.

Patients were categorized into three groups based on their experience with minor bleeds: current exposure to minor bleeds — defined as bleeding in the past 1, 2 or 3 months — past exposure to minor bleeds and no exposure to minor bleeds.

In total, 913 major bleeds and 7,194 minor bleeds occurred during a mean follow-up of 2.1 years.

The incidence rate per 100 patient-years was 11.7 (95% CI, 1.4-11.9) for minor bleeds and 1.5 (95% CI, 1.4-1.6) for major bleeds.

Researchers used both a Cox proportional hazards model and case-crossover model to examine the major bleed incidence risk associated with preceding minor bleeding events.

Results from the Cox model showed an increased risk for major bleeds in patients who had current minor bleeding (HR = 2.4 to 2.7), as well as in patients with past minor bleeding (HR = 1.9 to 2).

In analyses adjusted for age, sex, INR target range, antidiabetic medication, antihypertensive medication, NSAIDs and antiplatelet drugs, the association between minor and major bleeding persisted.

“This suggests that the risk factors we and others adjusted for ... hardly affect the risk of major bleeding in patients who have had a prior minor bleed,” the researchers wrote.

The case-crossover study only included patients who had experienced a major bleed. Because patients were compared with themselves, this analysis corrected for all fixed risk factors, such as sex, chronic medication and socioeconomic status.

Results of this analysis showed minor bleeds were not associated with major bleeds (OR = 0.9; 95% CI, 0.5-1.5). Rather, minor bleeds were markers for an increased risk for major bleeds.

The self-reported nature of minor bleeds may be a limitation to this study. Further, researchers assumed skin hemorrhages were more likely to be caused by trauma than a bleeding event, which may not be true for all patients.

Although the risk factors remain unknown, the researchers hypothesized that ABO blood group, vessel wall damage and gene variants may be contributing factors.

“New predictors are necessary, considering that current prediction models for major bleeds do not perform well and cannot distinguish between patients who are at high risk for bleeds and thromboembolism,” the researchers wrote. – by Nick Andrews

Disclosure: The researchers report no relevant financial disclosures.