Early Detection of Concealed Cardiac Amyloidosis Using AI-ECG and CT-Derived Extracellular Volume in Patients With Atrial Fibrillation

N/ANot Yet RecruitingNCT07567508

Samsung Medical Center500 enrolled

Overview

This study investigates the clinical efficacy of a non-invasive screening protocol using AI-ECG and CT-ECV analysis for cardiac amyloidosis. The study targets on atrial fibrillation(AF) patients with "red-flag" indicators. Participants are randomized 1:1 into either an early screening or usual care group. * Early screening group : AI- ECG and/or CT-ECV analysis + AF treatment * Usual care group : AF treatment Both groups followed for 2 years to compare CA detection rates and clinical outcomes.

This study is a prospective randomized trial designed to evaluate the clinical utility of non-invasive early screening for cardiac amyloidosis (CA) using AI-ECG and CT-ECV in patients with atrial fibrillation (AF). The study primarily targets patients presenting with "red-flag" signs, such as heart failure symptoms, ECG findings of low voltage or pseudo-infarction, Troponin T \> 0.03 ng/L, NT-proBNP \> 332 pg/mL, LV wall thickness \> 12 mm, a history of carpal tunnel syndrome or spinal canal stenosis, or proteinuria (e.g., UACR ≥ 300 mg/g, 24-hour proteinuria ≥ 0.5-1.0 g/day, or nephrotic-range proteinuria \> 3.5 g/day). The objective is to assess whether early screening improves the detection rate of CA and whether such early detection and subsequent treatment can improve clinical outcomes for patients. Subjects who meet the inclusion and exclusion criteria and provide informed consent will be randomly assigned to either the Early Screening Group or the Usual Care Group in a 1:1 ratio for comparative analysis. Participants will be allocated using stratified block randomization (allocation ratio 1:1) with a computer-generated random number sequence. Stratification variables, including age (age ≥ 75 vs. \< 75 years) and sex. Upon enrollment, all patients will undergo a baseline evaluation encompassing demographic information, medical history, and medication status, along with blood tests (e.g., NT-proBNP, troponin, creatinine). Additionally, AI-ECG analysis using standard 12-lead ECG or ECV analysis using cardiac CT will be performed. For AI-ECG analysis, the Mayo Clinic AI-ECG algorithm will be utilized; raw ECG data will be transmitted, and results will be received only for those patients who have provided specific consent. Patients who test positive in the AI-ECG analysis or cardiac CT-ECV analysis will undergo further diagnostic workup, including SPEP, UPEP, IFE, Free Light Chain Assay, and 99mTc-DPD scintigraphy. For those who test positive in these subsequent evaluations, definitive diagnosis will be attempted through endomyocardial biopsy or genetic testing.

Study Type

INTERVENTIONAL

Interventions

AI-ECG analysisDIAGNOSTIC_TEST

Artificial Intelligence-enhanced Electrocardiogram (AI-ECG), developed by Mayo Clinic, is gaining attention as a non-invasive screening tool. AI-ECG predicts myocardial amyloid deposition based on a standard 12-lead ECG with high accuracy, AUC 0.84(95% CI 0.82-0.86) In particular, it demonstrated superior performance with an AUC of 0.9 or higher in ECGs exhibiting low voltage or pseudo-infarction patterns, suggesting its potential to detect the disease even before structural changes become apparent.

CT-ECV analysisDIAGNOSTIC_TEST

Myocardial Extracellular Volume (CT-ECV) analysis using cardiac CT is a tissue characterization technique that can be easily added to conventional cardiac CT protocols, enabling the quantification of myocardial fibrosis or infiltrative diseases. Among 874 subjects who underwent coronary CT, 12.4% exhibited a CT-ECV of ≥ 35%, and cardiac amyloidosis was incidentally discovered in 14.3% of these individuals.

Atrial fibrillation treatmentOTHER

Standard care and treatment in accordance with established AF guidelines.

Eligibility

Sex: ALLMin age: 19 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Adults aged 19 or older who have provided voluntary written informed consent. * Patients with a history of AF (paroxysmal or persistent) and one or more red-flag symptoms/signs. * Patients who can undergo at least one of the following: AI-ECG analysis or CT-ECV analysis. Exclusion Criteria: * Patients previously diagnosed with cardiac amyloidosis (AL or ATTR). * Patients with severe heart failure (NYHA class IV) or terminal illness with a life expectancy of less than 1 year. * Patients deemed inappropriate for participation by the investigator.

Outcomes

Primary Outcomes

Cardiac amyloidosis(CA) detection rate

CA detection rate in the early screening group based on AI-ECG and CT-ECV results The superiority hypothesis (detection rate of the early screening group \> control group) will be tested, utilizing the chi-squared test or Fisher's exact test as the primary analysis. Effect sizes will be reported as the difference in proportions, odds ratio (OR), and relative risk (RR), along with 95% CIs and p-values.

Time frame: From enrollment to the 2 year follow-up

Secondary Outcomes

Diagnostic Sensitivity and Specificity of AI-ECG alone, CT-ECV alone, and the combined Model

Compare the diagnostic sensitivity and specificity of each model, the difference in proportions between groups will be analyzed as a primary analysis using the chi-squared test or Fisher's exact test. Furthermore, since the sensitivity and specificity of the AI-ECG, CT-ECV, and combined models are paired indicators derived from the same subjects, the McNemar test will be used for the comparison of sensitivity and specificity between models. In cases where it is necessary to adjust for covariates (e.g., age, sex, NT-proBNP) or to consider repeated measures/correlation structures, a logistic regression analysis based on Generalized Estimating Equations (GEE) with subject IDs as clusters will be performed to compare model effects.

Time frame: From enrollment to 2 year follow-up

Correlation of AI-ECG/CT-ECV values with other screening results in patients with confirmed CA

In patients with a confirmed CA, the association between AI-ECG scores/probabilities, CT-ECV (continuous), and clinical findings-such as biopsy results (if available), degree of DPD uptake (e.g., grade), and SPEP/UPEP/IFE results (positive/negative or quantitative values)-will be evaluated using Spearman correlation analysis and, if necessary, logistic or linear regression analysis.