Associazione Nazionale Medici Cardiologi Ospedalieri

CONGRESS ABSTRACT

CONGRESS ABSTRACT

CORRELATION BETWEEN CARDIAC MRI AND SPECKLE TRACKING ECHOCARDIOGRAPHY IN PATIENTS WITH CARDIAC AMYLOIDOSIS

Capolongo Antonio Caserta (Caserta) – Ospedale Sant’Anna E San Sebastiano | Susini Francesca Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Rosi Sara Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Pastore Maria Concetta Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Ragazzoni Gian Luca Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Giacomin Elisa Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | D’Ascenzi Flavio Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Cavigli Luna Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Mazzei Maria Antonietta Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Calabrò Paolo Caserta (Caserta) – Sant’Anna E San Sebastiano Di Caserta | Cameli Matteo Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​ | Focardi Marta Siena (Siena) – Azienda Ospedaliero-Universitaria Senese​

Background : Cardiac amyloidosis (CA) causes myocardial involvement that can be characterized by cardiac magnetic resonance (CMR) late gadolinium enhancement (LGE). Whether left atrial strain is associated with LGE segment burden after accounting for left ventricular (LV) global longitudinal strain (GLS) remains unclear. Methods : We retrospectively included 56 consecutive CA patients undergoing transthoracic echocardiography with speckle-tracking and contrast-enhanced CMR. Myocardial involvement was quantified as LGE segment burden (LGEseg; number of LGE-positive LV segments, range 0-17). Associations were assessed using Spearman correlation and negative binomial regression. Prespecified negative binomial models adjusted for age, sex, and amyloid type evaluated bi-plane left atrial reservoir strain (LASr) as the primary predictor with adjustment for GLS (primary model), LASr without GLS (secondary model, descriptive), and GLS as a standalone predictor (secondary analysis). Results : LGEseg ranged from 0 to 17 (median 5.5; IQR 2-12.5). Global GLS was -14.66±4.03% and worsened across LGEseg categories (Welch ANOVA p=0.019; 0-5 vs 12-17 p=0.01). GLS correlated with LGEseg (ρ=0.43, p=0.001) and remained associated with higher LGEseg in the secondary analysis (IRR 1.11 per 1-unit GLS worsening; 95% CI 1.04-1.18; p=0.001). LASr correlated inversely with LGEseg (ρ=-0.49, p=3.38×10⁻⁴). In the primary model including both LASr and GLS, LASr remained significantly associated with lower LGEseg (IRR 0.955 per 1-unit increase; 95% CI 0.921-0.991; p=0.014; n=46), whereas GLS was attenuated (p=0.075). Conclusions : In CA, LASr remains associated with CMR LGE segment burden after adjustment for GLS, supporting left atrial strain as a sensitive marker of disease burden complementing LV mechanics.