Longitudinal strain of left ventricular basal segments and E / e ′ ratio differentiate primary cardiac amyloidosis at presentation from hypertensive hypertrophy: an automated function imaging study

Background Longitudinal strain is an early marker of left ventricular ( LV ) dysfunction in several cardiac diseases. Our aim was to differentiate cardiac amyloidosis ( CA ) at diagnosis from hypertensive LV hypertrophy ( LVH ) by analysis of longitudinal myocardial deformation. Methods Thirty healthy controls (C), 30 newly diagnosed, never treated hypertensives with LVH ( H ‐ LVH ), and 33 patients with CA at diagnosis underwent echo Doppler including speckle tracking–based automated function imaging ( AFI ). Averaged peak systolic global longitudinal strain ( GLS , 18 segments) and basal, middle, and apical longitudinal strain ( BLS , MLS , and ALS , respectively, six segments each) were calculated. Results Left ventricular mass index, relative wall thickness, and ejection fraction did not differ between H‐ LVH and CA . E / e ′ ratio was higher in CA than in H‐ LVH (P<.001) and C (P<.0001). GLS was lower in CA than in C (P<.0001), without difference with H‐ LVH . ALS did not differ among the three groups, MLS was significantly lower in both CA and H‐ LVH than in C but BLS was lower in CA compared to both H‐ LVH and C (both P<.0001). In the pooled population, E / e ′ was independently associated with BLS (β=−0.54, P<.0001). At receiver operating curve analysis, CA was predicted by BLS ≤−11.3% (sensitivity=63.3%, specificity=100%) and E / e ′≥12.3 (sensitivity=69.7%, specificity=83.3%). The best AUC (=0.819) was obtained by the combination E / e ′+ BLS . Conclusions Our findings highlight a real difference of E / e ′ ratio and longitudinal strain of LV basal segments between hypertensive LVH and CA , which could be used to differentiate the two diseases.