Preventing ischaemic stroke in patients with isolated sinus node disease: the value of evaluating atrial cardiomyopathy

This editorial refers to ‘Refining prediction of stroke in sinus node dysfunction patients without atrial fibrillation using the P-combined score: A multicenter study’, by H. Dong et al., https://doi.org/10.1093/eurjpc/zwad267. Over the next 25 years the incidence of ischaemic strokes is projected to increase by >27% in Europe, leading to higher rates of morbidity, mortality, and healthcare costs.1 Detecting patients at risk for ischaemic strokes and implementing preventive therapeutic strategies are crucial steps towards improving the prognosis of these individuals. In particular, patients with isolated sinus node dysfunction (ISND), even those without a history or current diagnosis of supraventricular tachyarrhythmias, face an elevated risk of cardiovascular events when compared with the general population.2 Isolated sinus node dysfunction, which comprises a wide spectrum of sinoatrial dysfunctions, is a frequently encountered condition in clinical practice and may be suspected through surface electrocardiogram (ECG), Holter-ECG, and internal loop recorder. Unlike atrial fibrillation (AF), the identification and management of the heightened thromboembolic risk in patients with ISND but no AF have yet to be established. To address this unmet need, in this issue of the European Journal of Preventive Cardiology, Dong et al.3 eloquently assess the prognostic value of 12-lead ECG to predict the occurrence of ischaemic strokes in individuals with ISND. Briefly, in >2100 patients with ISND but no AF at baseline or during follow-up, the authors determined P-wave indices including an increased terminal force of the P-wave in lead V1 (PTFV1), P-wave duration, abnormal P-wave axis, and the presence of an advanced interatrial block, as well as the occurrence of cardioembolic stroke. They found that incorporating a marker of an abnormal P-wave morphology—PTFV1 > 4000 µV*ms—to the traditional CHA2DS2-VASc score into a specific P-combined score may enhance the predictive capacity for cardioembolic events in this population [c-statistic of 0.75 (0.68–0.82) for the CHA2DS2-VASc score vs. 0.81 (0.75–0.87) for the P-combined score]. These findings suggest that various components of atrial remodelling, gathered around the concept of atrial cardiomyopathy (ACM), play a pivotal role in the incidence of ischaemic strokes in the specific population facing ISND. The epidemiological link between sinus node dysfunction and cardioembolic events had been previously explored. Patients with ISND exhibit an increased risk of ischaemic stroke compared with the general population, albeit lower than that of AF.2 Therefore, it is essential to identify those ISND patients at the highest risk of systemic embolism to implement effective prevention measures. Not surprisingly, the CHA2DS2-VASc score, originally designed to predict the thromboembolic risk in AF patients, has also proven useful for risk stratification within the population diagnosed with ISND.4 Such predictive capacity in non-AF patients may be partially attributed to the fact that the CHA2DS2-VASc score considers various risk factors for ACM and atherosclerosis, rather than specifically accounting for AF characteristics.5 Clinical data support that ACM may be directly involved in the origin of intracardiac emboli.5 On top of the different risk factors gathered within the CHA2DS2-VASc score, several direct non-invasive ACM markers such as atrial dilation and serum NT-proBNP levels seem to improve the prediction of thromboembolic risk in the ISND population.4 The study published in this issue demonstrates that incorporating another ACM marker, the PTFV1 derived from 12-lead surface ECG, enhances the prediction of ischaemic strokes.3 P-wave in lead V1 is a marker for P-wave abnormalities developed over 50 years ago that evaluates both electrical and structural atrial remodelling and emerges as a non-invasive estimate of atrial fibrosis. Considering that it also predicts AF occurrence in both the general population and in patients with ISND,6 PTFV1 appears to be a valuable diagnostic tool for ACM screening. Atrial cardiomyopathy is a recently described multidimensional notion accruing several components of a deleterious remodelling of both atria, eventually impairing their function (Figure 1).5 Despite the fact that ACM has been claimed to play a significant role in the pathology of AF and cardioembolic ischaemic strokes, its diagnosis remains challenging.5 Several non-invasive markers appear promising, but no single tool currently captures the various aspects of atrial remodelling, whether electrical, structural, morphological, or functional. In contrast, diagnosing ACM and quantifying its severity are crucial to design personalized medicine and preventive strategies for AF and cardioembolic ischaemic strokes. Certain indices extracted from 12-lead ECGs (e.g. P-wave indices as addressed by Dong et al.3) and echocardiograms are readily available as a part of routine clinical practice, but are limited in their scope and ability to explore the atrial substrate. New diagnostic approaches may provide valuable insights into ACM characterization and assessment. Late gadolinium enhancement cardiovascular magnetic resonance (LGE-CMR) comprehensively assesses functional, morphological, and ultrastructural atrial remodelling. The significant correlation between thromboembolic risk and left atrial fibrosis and sphericity assessed by LGE-CMR supports its value in identifying clinically relevant ACM.7 However, while helpful in selected, small populations,8 notable limitations related to cost and accessibility jeopardize its potential use for large-scale screening. Electrocardiographic imaging (ECGi), a non-invasive diagnostic technique developed over the last decade, appears capable of providing an integrated evaluation of electrical and structural atrial remodelling based on non-invasive activation and propagation mapping.9 Its potential value in ACM diagnosis is, however, still being investigated. The potential value of atrial cardiomyopathy assessment for initiating preventive therapy against ischaemic stroke. ACM, atrial cardiomyopathy; ECG, electrocardiogram; ECGi, electrocardiographic imaging; LA, left atrium; LGE-CMR, late gadolinium enhancement cardiovascular magnetic resonance. While academically interesting, the authors appropriately emphasize the need to subject their findings to future interventional trials.3 It is crucial to assess whether long-term anticoagulation tailored to the presence of ACM offers relevant benefits over conventional management. Controlled randomized trials evaluating the potential benefit of long-term anticoagulation to prevent the onset of systemic embolism and ischaemic stroke in individuals diagnosed with ISND and a significant ACM, but without a history of AF, are warranted. However, these studies may present significant challenges. Despite a sound physiological rationale and numerous positive observational studies, the ARCADIA trial failed to prove a benefit of anticoagulation driven by ACM assessment in patients with an embolic stroke of undetermined source (ESUS).1 ARCADIA was the first, yet negative, randomized controlled trial evaluating the benefit of managing thromboembolic risk prevention by anticoagulant agents depending on the presence of ACM in a specific population. Several explanations may account for this neutral outcome. The excess risk detected by the assessment of ACM might not be high enough high to yield a favourable risk–benefit balance of chronic anticoagulation. In such a scenario, establishing a thromboembolic risk prevention strategy based on ACM screening may not align with the World Health Organization’s screening recommendations. At least partially, we believe that the lack of benefit owes to a suboptimal strategy for diagnosing and quantifying ACM. In ARCADIA, the ACM was defined by the presence of at least one of the following: PTFV1 exceeding 5000 µV*ms in 12-lead ECG, a serum NT-proBNP level surpassing 250 pg/mL, or a left atrial indexed diameter ≥30 mm/m2 on echocardiogram.10 While still statistically significant, each marker by itself may be insufficiently specific or sensitive to capture all the atrial remodelling components and reliably establish a diagnosis of ACM. Moreover, it is also possible that diagnostic criteria for every marker should be refined. This is illustrated by the discrepancy in PTFV1 threshold values between ARCADIA10 and the study presented by Dong et al.,3 where the latter proposed a threshold of 4000 µV*ms instead of 5000 µV*ms. Novel CMR or ECGi techniques and parameters evolve as powerful tools to enhance our understanding, assessment, and risk stratification of the ACM. In summary, despite significantly increasing the risk of stroke by one-third, ISND is most likely a risk marker for ischaemic stroke than a direct risk factor (Figure 1). Screening for ACM in this population may enhance the identification of a subpopulation at a particularly high risk and facilitate the implementation of personalized and preventive medical strategies. Nonetheless, it remains imperative to enhance diagnostic methods for ACM in the general population by employing non-invasive and scalable diagnostic tools and to demonstrate the substantial benefits of thromboembolic risk prevention based on ACM assessment. While the literature on evaluating ACM, its connection to cardiovascular diseases, and the adoption of treatment strategies based on its presence has greatly expanded over the past decade, there is still a lengthy and challenging journey ahead to integrate the concept of ACM into routine clinical practice. This integration should aim to improve both ACM diagnostic management and ACM-based therapeutic approaches to prevent cardioembolic ischaemic strokes. This editorial received partial funding from the Instituto de Salud Carlos III (PI22/00953). J.-B.G. also acknowledges support from an institutional grant, the ‘Contractes d’Investigació Avançada Fundació BBVA—Hospital Clínic Barcelona Joan Rodés—Josep Baselga 2022’ (HCB_BIO_001/2). No original or reused data is presented in this manuscript.