# Development and validation of an echocardiographic nomogram for identifying cardiac amyloidosis in patients with left ventricular hypertrophy

## TL;DR

This study creates a new echocardiogram-based model to help doctors identify cardiac amyloidosis in patients with thickened heart walls.

## Contribution

A novel nomogram model using common echocardiographic parameters to accurately predict cardiac amyloidosis in patients with left ventricular hypertrophy.

## Key findings

- The nomogram model achieved high accuracy (0.91–0.92) and sensitivity (0.90–0.91) in predicting cardiac amyloidosis.
- Key predictive factors included asymmetric hypertrophy, AMYLI score, and valvular regurgitation.
- The model's high negative predictive value (0.93–0.98) suggests it can effectively rule out the disease.

## Abstract

Echocardiography is the principal non-invasive imaging modality for screening cardiac amyloidosis (CA). This study aimed to establish a cohort of CA-associated left ventricular hypertrophy (CA-LVH) within a hospital-based population and to develop an echocardiographic identification model for CA using readily available echocardiographic parameters.

This retrospective nested cohort study involved the collection of clinical and echocardiographic data from three hospitals affiliated with the West China Medical Center, Sichuan University, between January 1, 2008, and December 31, 2023. The relative wall thickness (RWT) was calculated as twice the left ventricular posterior wall thickness (LVPW) divided by the left ventricular internal diameter (LVID). Asymmetric hypertrophy was defined as a ratio of interventricular septal thickness (IVS) to LVPW greater than 1.3. The AMYLI score was computed as the product of RWT and E/e’ ratio.

A total of 185 CA patients (183 AL-CA and 2 ATTR-CA) who underwent 309 echocardiography examinations from different time periods with 1,213 echocardiographic data points from in-hospital non-CA-LVH cases matched for age, gender, and body surface area were included. Multivariable logistic regression analysis identified a history of hypertension [odds ratio (OR): 0.04, 95% confidence interval (CI): 0.021–0.073], LVID [OR: 0.927, 95%CI: 0.878–0.977], left ventricular ejection fraction (LVEF) [OR: 0.95, 95%CI: 0.908–0.993], AMYLI score [OR: 1.088, 95%CI: 1.024–1.161], asymmetric hypertrophy [OR: 3.729, 95%CI: 1.884–7.441], granular sparkling [OR: 3.111, 95%CI: 1.355–7.431], small pericardial effusion [OR: 2.77, 95%CI: 1.563–4.937], mild aortic regurgitation [OR: 2.353, 95%CI: 1.278–4.361], mild mitral regurgitation [OR: 4.331, 95%CI: 2.347–8.141], and mild tricuspid regurgitation [OR: 3.837, 95%CI: 2.026–7.358] as independent predictive factors for CA in LVH patients. The predictive factors were used to construct a nomogram model, which demonstrated high accuracy (0.91–0.92), specificity (0.91–0.92), sensitivity (0.90–0.91), positive predictive value (0.73), negative predictive value (0.93–0.98), and Youden index (0.81–0.83).

The developed nomogram displayed remarkable predictive accuracy, which has the potential to enhance CA screening via routine echocardiography and strategically guide subsequent diagnostic evaluations.

The online version contains supplementary material available at 10.1186/s12872-025-04973-5.