Left atrial deformation discloses atrial amyloid infiltration among patients with increased myocardial thickness

Abstract

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Swedish research council Swedish heart and lung foundation Background Transthyretin cardiac amyloidosis (ATTR-CA) is an infiltrative disease with thickening and stiffening of the heart walls as a result. Amyloid deposits are found not only in the ventricular walls, but in the atriums as well. Atrial function is known to play an integral part in the overall function of the heart. Purpose We aimed to evaluate left atrial deformation including left ventricular-atrial strain loop (LV-LA loop). The latter to test LA deformation and its dependence of LV deformation in ATTR CA and compare it to left ventricular hypertrophy (LVH). Methods Retrospective analysis was performed on available echocardiographic data, obtained between years 2004–2022. The study population was divided into three groups: healthy controls with no identified cardiac disease; ATTR-CA patients with verified ATTR-CA diagnosis and interventricular septal thickness end diastole (IVSd) of ≥14mm; LVH including IVSd ≥14mm and no CA diagnosis. Left heart strain measurements, including LV global longitudinal strain (LV GLS) and LA peak atrial longitudinal strain (LA PALS), were acquired from apical 4-chamber view during same single heart cycle. Strain coordinates throughout the heart cycle were plotted using MATHLAB, combining LV GLS and LA PALS to compute a strain-strain loop. A slope was determined using regression line to index relative function of the left ventricle to left atrium (GLS-PALS-slope). Results Data was gathered in total of 90 patients (67.8±1.2 years, 55,6% male) divided into three groups containing 30 patients each: ATTR-CA including 24 hereditary ATTR (including mutations; 19 Val30Met, one The60Ala, one Val122Ile, one Ala45Glys, one Ala97Ser and one His88Arg) and six wild type ATTR, LVH with increased myocardial thickness believed caused by aortic stenosis (n=13; one grade I, three grade II-III and eight grade III), essential hypertension (n=10) and hypertrophic cardiomyopathy (n=7, four of which had outflow obstruction). The lowest PALS and PALS-GLS slope was detected in ATTR-CA (p-value 0.005 respectively 0.003 between ATTR-CA compared to LVH). No significant statistical difference could be verified in GLS between ATTR-CA and LVH (p-value 0.114). A ROC curve of ATTR-CA demonstrated similar results between PALS and GLS-PALS slope (AUC 0.79 resp. 0.78) which was slightly higher than GLS (AUC 0.69). Slightly higher AUC was generated when slope was indexed with LAVI (AUC 0.80). Conclusions LA deformation demonstrates a strong ability to differentiate ATTR-CA from LV hypertrophy. Adding LV strain in relation to LA deformation showcases the mechanical dissociation of LA to LV in ATTR-CA, potentially unmasking atrial infiltration. Its use in prognostic evaluation needs further studies.