Rethinking Thromboembolic Risk in ATTR-CM Care

In transthyretin amyloid cardiomyopathy (ATTR-CM), thromboembolic risk doesn’t behave the way we’d expect. Events occur despite therapeutic anticoagulation, in patients with sinus rhythm, and without clear stratification by CHA₂DS₂-VASc score.

In a comprehensive multimodality imaging study, Ozbay and colleagues provide compelling evidence that the left atrium itself is intrinsically diseased in ATTR-CM—and that this primary atrial cardiopathy independently contributes to thrombotic risk.

Examining Atrial Dysfunction

In a propensity-matched cohort of 104 patients with ATTR-CM and 104 patients with non-amyloid left ventricular hypertrophy (LVH), the atrial phenotype diverged in clinically meaningful ways. Left atrial reservoir strain (LASr) was markedly reduced in ATTR-CM (approximately 10% vs 17% in LVH), while left atrial stiffness was more than doubled (median ~2.1 vs 0.9 1/%). These abnormalities were present despite smaller or only mildly enlarged left atrial volumes in ATTR-CM.

Unlike the LVH cohort, atrial size in ATTR-CM did not increase proportionally with worsening diastolic dysfunction and showed poor correlation with filling pressures or strain parameters. This dissociation between chamber size and mechanical function supports the concept of direct amyloid infiltration producing structural and electromechanical impairment rather than secondary remodeling from pressure overload alone.

Cardiac magnetic resonance further reinforced this distinction; atrial late gadolinium enhancement (LGE) was more intense in ATTR-CM, yet mechanical dysfunction was already severe even at lower LGE grades, suggesting a diffuse infiltrative substrate rather than a purely fibrosis-mediated process.

Beyond Atrial Fibrillation

Although atrial fibrillation was more prevalent in ATTR-CM, the central observation here is that atrial dysfunction tracked with the amyloid phenotype itself—not merely with rhythm status. In multivariable models adjusted for atrial fibrillation and diastolic dysfunction, both LASr and atrial stiffness remained independently associated with ATTR-CM. In contrast, atrial fibrillation predicted thrombotic events in the LVH group but not in the ATTR-CM cohort.

Clinical outcomes reinforce this distinction. Over a median follow-up of 31 months, 20 thrombotic events occurred in the ATTR-CM group. 65% of these events developed despite anticoagulation, and 40% occurred in patients who were in sinus rhythm at baseline. Neither baseline atrial fibrillation nor CHA₂DS₂-VASc score discriminated risk within the ATTR-CM cohort.

Instead, mechanical parameters carried prognostic weight. An LASr threshold of 9% or lower was associated with a 7.8-fold increased risk of thrombotic events after adjustment for atrial fibrillation. An atrial stiffness value of 1.7 or greater conferred roughly a fourfold increased risk. Left atrial volume index did not predict events, reinforcing that deformation and stiffness—not size—capture the clinically relevant substrate.

Electrocardiographic findings align with this structural narrative. Patients with ATTR-CM demonstrated prolonged P-wave duration and more frequent advanced interatrial block despite less atrial enlargement, consistent with infiltrative conduction delay rather than stretch-mediated remodeling.

Implications for Clinical Strategy

Taken together, these findings support a shift in conceptual framework. In ATTR-CM, thromboembolic risk appears to arise from an infiltrative atrial myopathy characterized by impaired reservoir function, increased stiffness, conduction abnormalities, and probable electromechanical dissociation. Atrial fibrillation may amplify risk, but it is not required for thrombogenesis in this population.

For clinicians, the implications are pragmatic. Reliance on CHA₂DS₂-VASc scoring alone is insufficient in cardiac amyloidosis. Incorporating left atrial strain and stiffness—obtainable through speckle-tracking echocardiography—may improve risk stratification. While prospective randomized data are needed before recommending anticoagulation solely on the basis of impaired atrial mechanics, this study provides a strong physiologic rationale for rethinking how thromboembolic prevention is approached in ATTR-CM.

Reference:
Ozbay B, Rearick C, Satyavolu BS, et al. Primary left atrial cardiopathy in transthyretin amyloidosis cardiomyopathy by multimodality imaging: implications for thrombotic events. JACC Cardiovasc Imaging. 2025;18(8):867-881. doi:10.1016/j.jcmg.2025.04.007.