Abstract
BACKGROUND: Wild-type transthyretin cardiac amyloidosis (ATTRwt-CM) is a frequent but underdiagnosed cause of heart failure with preserved ejection fraction (HFpEF) and left ventricular hypertrophy (LVH). Early identification is essential given the availability of disease-modifying therapies. The T-Amylo and Davies scores are non-invasive tools for estimating ATTR CM probability, but their comparative performance in the same real-world population is not well defined. OBJECTIVES: To compare the diagnostic accuracy of T-Amylo and Davies scores in consecutive patients referred for suspected cardiac amyloidosis. METHODS: We retrospectively analyzed 81 patients (mean age 76.8 ± 8.3 years, 74% male) who underwent a standardized work-up: ECG, echocardiography with strain, NT-proBNP and troponin, bone scintigraphy, and immunofixation. ATTR CM was diagnosed according to established non-biopsy criteria. Both scores were calculated retrospectively, and sensitivity, specificity, predictive values, accuracy, and agreement were assessed. RESULTS: ATTR CM was confirmed in 28 patients (34.5%). T-Amylo showed higher sensitivity (91.2% vs. 73.5%) and NPV (89.7% vs. 79.1%), while Davies had greater specificity (85.0% vs. 65.0%) and PPV (80.5% vs. 70.8%). Overall accuracy was comparable (T-Amylo 77.0% vs. Davies 79.7%). Agreement between scores was moderate (κ = 0.59). CONCLUSIONS: T-Amylo is best suited as a screening tool for suspected ATTR CM, while Davies offers confirmatory value in high-probability cases. Combining these tools in a sequential strategy may optimize diagnostic efficiency, reduce unnecessary testing, and expedite initiation of disease-modifying therapy.