Takeuchi, Masaaki

Abstract Background and Aims Accurate differentiation of cardiac amyloidosis (CA) from phenotypic mimics remains challenging using current clinical and echocardiographic techniques. The accuracy of a novel artificial intelligence (AI) screening algorithm for echocardiography-based CA detection was assessed. Methods Utilizing a multisite, multiethnic dataset (n = 2612, 52% CA), a convolutional neural network was trained to differentiate CA from phenotypic controls using transthoracic apical four-chamber video clips. External validation was conducted globally across 18 sites including 597 CA cases and 2122 controls. Classification accuracy was assessed on the entire external validation dataset, and subgroup analyses were performed both on technetium pyrophosphate scintigraphy referrals, and individuals matched for age, sex, and wall thickness. Model accuracy was also compared with the transthyretin CA score and the increased wall thickness score within a subset of older heart failure with preserved ejection fraction patients with increased wall thickness. Results Cardiac amyloidosis patients and controls displayed similar age, sex, race, and comorbidities. After the removal of uncertain AI predictions (13%), model discrimination and classification were excellent for the entire external validation dataset [area under the receiver operating characteristic curve (AUROC) 0.93, sensitivity 85%, specificity 93%], irrespective of CA subtype (sensitivity: light-chain = 84%, wild-type transthyretin = 85%, and hereditary transthyretin = 86%). Performance was maintained in subgroup analysis in patients clinically referred for technetium pyrophosphate scintigraphy imaging (AUROC 0.86, sensitivity 77%, specificity 86%) and matched patients (AUROC 0.92, sensitivity 84%, specificity 91%). The AI model (AUROC 0.93) also outperformed transthyretin CA score (AUROC 0.73) and increased wall thickness (AUROC 0.80) scores. Conclusions This AI screening model—using only an apical four-chamber view—effectively differentiated CA from other causes of increased left ventricular wall thickness.

ImportanceChronic hemodynamically significant aortic regurgitation (AR) is associated with excess risk of death, yet data for Asian patients are lacking, and whether Asian patients can abide by Western guidelines as to when aortic valve surgery should be performed is unknown.ObjectiveTo assess AR presentation and cutoffs of left ventricular ejection fraction (LVEF), LV end-systolic dimension index (LVESDi), and LV end-systolic volume index (LVESVi) that are associated with risk of death in Asian patients with AR.Design, Setting, and ParticipantsThis retrospective cohort study included consecutive patients with chronic, moderately severe to severe AR from 3 tertiary referral centers (2 in Japan and 1 in Taiwan) from June 11, 2008, through November 19, 2020, with follow-up through November 11, 2021.ExposuresAortic regurgitation severity, graded by a comprehensive integrated approach.Main Outcomes and MeasuresThe primary outcome was the association between volume-derived LVEF, LVESDi, and LVESVi and all-cause death (ACD). The secondary outcome was the association of these LV indexes with cardiovascular death (CVD). Clinical and echocardiographic data were analyzed retrospectively. A de novo disk-summation method was used to derive LV volumes and volume-derived LVEF.ResultsOf 1259 patients (mean [SD] age, 64 [17] years; 934 [74%] male), 515 (41%) were Japanese and 744 (59%) were Taiwanese. The median follow-up was 4.1 years (IQR, 1.56-7.24 years). The mean (SD) body surface area was 1.67 (0.21) m2; LVEF, 55% (11%); LVESDi, 24.7 (5.7) mm/m2; LVESVi, 50.1 (28.0) mL/m2; and indexed mid–ascending aorta size, 24.7 (5.5) mm/m2. Aortic valve surgery occurred in 483 patients (38%); 240 patients (19%) died during follow-up. Overall mean (SD) 8-year survival was 74% (2%). Separate multivariate models adjusted for covariates demonstrated independent associations of LVEF, LVESDi, and LVESVi with ACD (LVEF: hazard ratio [HR] per 10%, 0.80; 95% CI, 0.70-0.92; P = .002; LVESDi: HR, 1.04; 95% CI, 1.01-1.06; P = .002; LVESVi: HR per 10 mL/m2, 1.11; 95% CI, 1.05-1.17; P < .001) and CVD (LVEF: HR per 10%, 0.69; 95% CI, 0.56-0.85; P < .001; LVESDi: HR, 1.05; 95% CI, 1.01-1.09; P = .01; LVESVi per 10 mL/m2: HR, 1.15; 95% CI, 1.06-1.24; P < .001). In the total cohort, spline curves showed that mortality started to increase for an LVEF of 53% or less, LVESDi of 22 mm/m2 or greater, and LVESVi of 46 mL/m2 or greater for both ACD and CVD. Early surgery was beneficial in 3 strata of LVESDi (<20, 20 to <25, and ≥25 mm/m2) and 2 strata of LVESVi (<46 and ≥46 mL/m2).Conclusions and RelevanceThis multicenter cohort study of Asian patients with hemodynamically significant AR found cutoff values of LVEF, LVESDi, and LVESVi that were associated with increased risk of death. These findings suggest that Western guidelines seem applicable in Asian patients and, most importantly, that indexed LV parameters with a lower cutoff could be used in discriminating patients with excess mortality risk.

Abstract Background Despite over a decade of evidence, uptake of GLS into guidelines and practice has been slow. Age, BP, and software have been reported to influence GLS, and have led to uncertainty about defining the normal range (NR). As there have been software changes since the last meta-analysis, we sought to define NR in the current era. Methods Pubmed, Cochrane, and EMBASE were searched using the keywords “Left Ventricle”, “Normal global longitudinal strain”, and “Speckle tracking Echocardiography” and relevant synonyms from 2011. Studies were included if they reported GLS based on speckle tracking techniques, included at least 20 healthy individuals, and had up to date contact details available. Healthy was defined as lack of known disease. Platforms used included General Electric, Philips, Canon and Tomtec. The authors of 8 of the 12 studies (67% response) provided individual patient data. Linear regression was used to determine predictors of GLS. Results There were 2396 pts, mean age 42 years (range 18–92), weight 66±12kg, height 169±9cm, BSA 1.7±0.2m2, and SBP 120±13 mmHg. Normal range for GLS was 21.0±2.6%. In multivariable analysis age (β=-0.02, p<0.01), weight (β=-0.03, p<0.01), SBP (β=-0.01, p<0.01) and platform were associated with GLS. GLS tends to vary with extremes of age and BP (see Figure 1 and Figure 2). GLS vs Age group Conclusion Extremes of age and BP may lead to low GLS, but generally, GLS is <16% is abnormal. Acknowledgement/Funding None