Overview: DOACs, NVAF, and the role of cardiac hemodynamics
Non-valvular atrial fibrillation (NVAF) requires careful stroke prevention, with direct oral anticoagulants (DOACs) such as apixaban, rivaroxaban, and dabigatran replacing warfarin in many patients. While DOACs have demonstrated non-inferiority to warfarin in large trials, real-world data reveal nuances in bleeding and stroke risk that may hinge on cardiac hemodynamics. A growing body of evidence suggests that transthoracic echocardiographic measures—particularly right ventricular systolic pressure (RVSP) and left ventricular ejection fraction (LVEF)—modulate DOAC-related outcomes. This article distills the key findings from a retrospective real-world cohort that linked echocardiographic parameters with bleeding and stroke rates across DOAC types.
Study design and population
The analysis drew on NVAF patients treated with a DOAC at Mayo Clinic sites and affiliated systems between 2001 and 2017. Exclusion criteria eliminated those with prior anticoagulation for venous thromboembolism, significant valvular disease, end-stage renal disease, DOAC switching, or insufficient follow-up. The primary outcome was time to first major bleeding (per ISTH criteria) after DOAC initiation; secondary outcomes included intracranial and gastrointestinal bleeding and time to first ischemic stroke among those without prior stroke. Echocardiographic data—especially LV ejection fraction (LVEF) and RVSP—were integrated to stratify patients and explore biologic gradients in outcomes.
Key findings: bleeding and stroke by DOAC and hemodynamics
<strongBleeding risks: Across 2,308 participants, there were 235 major bleeding events over a follow-up of about 2.1 years. In adjusted analyses, rivaroxaban showed a meaningful increase in major bleeding versus apixaban (HR 1.41; 95% CI 1.04–1.90). Dabigatran showed a non-significant trend toward higher bleeding (HR 1.19). Importantly, a strong interaction emerged with RVSP: higher RVSP amplified bleeding for dabigatran and rivaroxaban, while apixaban’s bleeding rate rose less with increasing RVSP. In patients with RVSP ≥ 45 mmHg, rivaroxaban and dabigatran were associated with markedly higher bleeding compared with apixaban (nearly double).
<strongGastrointestinal bleeding: Similar interaction patterns appeared—dabigatran and rivaroxaban carried higher GI bleed risk than apixaban, particularly as RVSP rose. In patients with reduced LVEF, rivaroxaban or dabigatran yielded more than twice as many GI bleeds as those on apixaban, underscoring hemodynamic influence on DOAC pharmacodynamics and bleeding risk.
<strongIntracranial hemorrhage: Intracranial bleeding events were rare and did not show a clear pattern by DOAC and RVSP strata, suggesting other factors may predominate in this outcome in real-world settings.
Stroke outcomes and hemodynamic context
Among participants without prior stroke, 68 incident strokes occurred over roughly 3,977 person-years. After adjusting for confounders, rivaroxaban trended toward lower stroke risk (HR 0.63) relative to apixaban, though this did not reach conventional statistical significance. Apixaban and dabigatran generally did not differ substantially in stroke rates overall. However, with higher RVSP, stroke risk patterns shifted: apixaban and dabigatran showed higher stroke rates at elevated RVSP, whereas rivaroxaban did not follow this linear trend. In practical terms, elevated RVSP appeared to preserve the expected stroke-prevention benefit of DOACs, but with varying degrees of risk depending on the agent used.
Biologic explanations and implications for practice
The authors propose several mechanisms for the observed interactions between hemodynamics and DOAC outcomes. Heart failure and pulmonary hypertension can alter drug absorption, distribution, and clearance. For example, RV pressure elevation and hepatic/renal congestion may affect DOAC bioavailability and metabolism, potentially leading to greater fluctuations in drug levels. Pharmacokinetic differences among DOACs—such as apixaban’s relatively favorable GI absorption profile and lower renal excretion compared with dabigatran—may explain why apixaban repeatedly showed a more favorable bleeding profile in patients with higher RVSP. Conversely, conditions that accompany high right-heart pressures could exaggerate bleeding risks with dabigatran and rivaroxaban while leaving apixaban comparatively safer. These findings align with the concept of a biologic gradient: as RVSP rises, DOAC choice materially shifts bleeding risk, with more modest effects on intracranial bleeding due to its overall lower baseline risk across DOACs.
Clinical takeaways
- Consider incorporating echocardiographic hemodynamics, especially RVSP and LVEF, when selecting a DOAC for NVAF patients, as these factors appear to moderate bleeding risk and, to a degree, stroke risk.
- In patients with elevated RVSP, apixaban may offer a more favorable bleeding profile compared with dabigatran or rivaroxaban, potentially guiding anticoagulant choice in higher-risk groups.
- Ongoing real-world data and prospective studies should further clarify the causal pathways linking hemodynamics with DOAC pharmacokinetics and clinical outcomes.
Limitations and next steps
As a retrospective observational study, causality cannot be established. Echocardiography was used to estimate RVSP rather than invasive right-heart catheterization, which, while more precise, could introduce sampling bias. Nevertheless, the study’s comprehensive adjustment for confounders and stratified analyses by RVSP and LVEF strengthen the plausibility of the reported associations. Future research should validate findings in diverse populations and examine potential mechanisms to optimize DOAC therapy in patients with pulmonary hypertension and heart failure.
