External validation of a predictive model for diagnosing transthyretin cardiac amyloidosis in suspected cases

Abstract

Introduction: Systemic amyloidosis is a disease caused by misfolded proteins forming amyloid fibrils, which deposit in tissues and organs, impairing their structure and function .When fibrillar material is deposited in the heart, it leads to cardiac amyloidosis (CA), an infiltrative cardiomyopathy typically presenting as heart failure. Non-invasive diagnosis is often guided by bone scintigraphy in patients with increased ventricular wall thickness and clinical red flags .However, these criteria may lack specificity, leading to unnecessary imaging. The T-Amylo model was developed to estimatepre-test probability of ATTR-CA using accessible clinical, electrocardiographic, and echocardiographic variables . As prediction models are developed in specific populations, they are subject to potential overfitting and may not generalize well. Therefore, external validation in independent cohorts is essential .This study aimed to evaluate the diagnostic performance of the T-Amylo model in an external cohort of patients with clinical suspicion of ATTR-CA in a Latinoamerican center.Methods: We conducted a retrospective, single-center, cross-sectional external validation study at Hospital Italiano de Buenos Aires. The study included consecutive adult patients who underwent bone scintigraphy with pyrophosphate for suspected ATTR-CA between 2016 and 2024.Results: During the study period, 622 patients were eligible. Of these, 214 were excluded: 33 due to a final diagnosis of AL amyloidosis, 157 because their septal thickness was <12 mm, and 24 for not meeting the red flag criteria. A total of 408 patients were included, of whom 123 (30 %) had positive DPD scintigraphy (Grade 2–3 uptake) and met the diagnostic criteria for ATTR-CA.Conclusion: The T-Amylo model demonstrated good overall performance in this external validation study and may assist clinicians in estimating the pre-test probability of ATTR-CA to guide appropriate diagnostic workup. Its application would allow prioritizing confirmatory studies and optimizing diagnostic resources.