Background

Transthyretin amyloid cardiomyopathy (ATTR-CM) is an underrated form of heart failure (HF) with preserved ejection fraction (HFpEF). The clinical work-up for ATTR-CM includes definite diagnosis with scintigraphy or endomyocardial biopsy (EBM) at specialized institutions. Older age and male predominance have been described for ATTR-CM as compared to a broader HFpEF population. However, ATTR-CM features independent of these covariates remain uncertain. Identification of simple clinical variables specific for ATTR-CM might aid in selecting patients for further work-up during initial cardiologic screening. We therefore sought to compare ATTR-CM patients with an age and sex matched cohort of HFpEF patients without ATTR-CM as well as to develop and validate a score for clinical ATTR-CM prediction.

Methods and Results

We here describe a contemporary cohort of 60 consecutive patients diagnosed w/ ATTR-CM by EBM at a specialized tertiary care center. Patients were of advanced age (78±6 years) and 92% were male. Overall, 32% of patients presented with NYHA class ≥3; 45% and 55% were in Gilmore-stages I and II, respectively. Mean left ventricular ejection fraction (LVEF) was 50±15% and 31% patients demonstrated a reduced LVEF. Patients with preserved LVEF (n=41) were age- and sex matched to a cohort of patients with HFpEF (n=41) from a large observational study (LIFE-Heart) using a propensity score.

Patients with HFpEF demonstrated less severe dyspnea (NYHA class ≥3: 8% vs. 32%, p<0.01), more obesity (49% vs. 20%, P=0.01), more arterial hypertension (100% vs. 70%, p<0.01), numerically higher rates of diabetes (51% vs. 34%, p=0.10) and lower pacemaker rates (2% vs. 41%, p<0.01). Rates of coronary artery diseases and the presence of atrial fibrillation were similar between groups. ATTR-CM patients showed more severe myocardial hypertrophy (IVSd 17±3 vs. 13±2mm, p<0.01) and diastolic dysfunction (E/e’ 20±7 vs. 11±3, p<0.01) as well as higher laboratory markers of cardiac injury and renal dysfunction (p<0.01 for troponin, NT-proBNP and creatinine, respectively).

After dichotomization of the above-mentioned variables according to receiver operating curve (ROC) analyses based Youden indices and including all variables with significant differences between groups, independent multivariate predictors for ATTR-CM on logistic regression were IVSd >14mm, E/e’ >14 and obesity (p>0.01 for all, r²=0.63). A weighted score was derived from associated odds ratios with IVSd >14mm=1point; absence of obesity =2 points; and E/e’ >14=3 points. Area under the ROC (AUC) for the summation score was 0.91 (CI 0.82 – 0.96, p<0.01) and a score of more than 3 points predicted an ATTR-CM diagnosis with high sensitivity (90%) and specificity (70%, see Figure). In an external validation cohort with 68 ATTR-CM and 145 HFpEF patients the score performed well overall (AUC 0.85) and within male patients only (AUC 0.83, p<0.01 for both).

Conclusion

ATTR-CM presents with reduced LVEF in 1/3 of cases. In the remainder, ATTR-CM shares common features with age- and sex matched HFpEF patients. However, higher degrees of LV hypertrophy, more severe diastolic dysfunction and the absence of obesity are independently predictive of an ATTR-CM diagnosis. A simple score based on these variables might facilitate the selection of patients for further ATTR-CM work-up.