Searching for biomarkers of atrial cardiomyopathy at high risk of cardioembolism: What are the missing pieces of the puzzle?

The ARCADIA trial compared the efficacy of apixaban and aspirin in cryptogenic stroke associated with markers of atrial cardiopathy (AC), based on the hypothesis that anticoagulant therapy could prevent cardioembolism related to the atrial disease, independent of atrial fibrillation (AF) occurrence. The absence of apixaban’s superiority over aspirin raised questions about the appropriateness of the selected markers for AC. Consequently, the current sub-analysis of the ARCADIA trial by Kamel et al. ¹ was aimed at evaluating the validity of these markers, by examining their relationship with subsequent occurrence of AF, an established hallmark of AC. This objective is particularly relevant given that AF predicts cardioembolic events better than other AC markers. However, the study revealed that the markers used in the ARCADIA trial were only moderately predictive of AF; furthermore, when combined with classic clinical proarrhythmic factors (i.e. age, sex, hypertension . . .), the only electrocardiographic marker used in this study (P terminal force V1 (PTFV1)) did not significantly enhance the predictive power for AF occurrence.

These findings raise questions about the most appropriate biomarker (or combination of biomarkers) to identify both AC and cardioembolic risk, independent of AF episodes.

Previous studies have shown various degrees of AF predictiveness among multimodal biomarkers, including several electrocardiographic markers (premature atrial contractions, maximum P-wave duration, PR interval, P-wave dispersion, P-wave index/P-wave standard deviation), echocardiographic markers, other biomarkers and clinical characteristics. These were evaluated at different times relative to cerebral ischemic events, demonstrating a varying effectiveness in identifying presence and duration of AF episodes. ² However, the lack of direct comparisons of these markers in predicting AF and thromboembolic events limits the ability to select the most appropriate group of markers. Furthermore, in the ARCADIA trial, few patients met the AC criterion of severe left atrial enlargement, a known significant predictor of both AF and stroke. Notably, a sub-analysis of the NAVIGATE trial suggested that in patients with severe left atrial enlargement rivaroxaban could reduce the risk of recurrent stroke. ³

The relationship between AC markers and thromboembolic risk is complex, especially considering the broad definition of AC, which includes “any structural, architectural, contractile, or electrophysiological changes affecting the atria that could produce clinically relevant manifestations.” ⁴ According to the EHRA/HRS/APHRS/SOLAECE expert consensus, the histological characterization of AC includes four types (type I: principal cardiomyocyte changes; type II: principally fibrotic changes; type III: combined cardiomyocyte-pathology/fibrosis; type IV: primarily non-collagen infiltration). These types may vary over time and differ at atrial sites in certain patients, so biomarkers might identify distinct AC subtypes with varying sensitivity and specificity. This also suggests that specific AC subtypes could involve unique alterations (structural, electrical, mechanical) impacting thrombotic risk to varying degrees, indicating that not all AC types are associated with thrombogenic mechanisms requiring anticoagulation. Additionally, different severity grades of AC could be identified by specific electrocardiographic markers and associated with distinct cardioembolic risks. We do not yet know if a specific group of markers could accurately identify different stages of AC.

A further neglected component of this intricate puzzle is the impact of systemic factors on thromboembolic risk in AC. Specific factors such as age, sex, and renal failure can impact thromboembolic risk in the context of underlying atrial alterations. Notably, systemic inflammation, often present in ischemic stroke patients but largely ignored, can promote both local thrombogenicity and atrial structural remodeling via potentially reversible alterations, which significantly impact AC occurrence and severity. Specifically, interleukin-6, that can favor atrial thrombosis, has been shown to modulate cardiac connexins influencing atrial conduction and modifying specific electrocardiographic markers (P wave dispersion/P-wave standard deviation) used to identify AC. These alterations appear to be rapidly reversible after anti-inflammatory therapy. ⁵

Such considerations highlight current knowledge gaps and underscore the need to better study AC to select a group of biomarkers capable of identifying not just AC per se, but specifically AC associated with a high risk of thromboembolic events.