Left ventricular ejection fraction evaluation by Tc-DPD cardiac gated SPECT in patients with ATTR amyloidosis: feasibility and limitations.

Fiche publication

Date publication

juillet 2025

Journal

The international journal of cardiovascular imaging

Auteurs

Membres identifiés du Cancéropôle Est :
Pr IMPERIALE Alessio

Tous les auteurs :
Dramé B, Martz M, El Ghannudi S, Hubele F, Hien MS, Bessac DM, Zeyons F, Hunolstein JV, Bathily EHAL, Mertz L, Blondet C, Schindler TH, Imperiale A

Lien Pubmed

https://www.ncbi.nlm.nih.gov/pubmed/40663288

Résumé

Cardiac function assessment using Tc-DPD ECG-gated SPECT (G-SPECT) in ATTR amyloidosis is underexplored. This study evaluates left ventricular ejection fraction (LVEF) and left ventricular end-diastolic volume (LVEDV) with G-SPECT, comparing results to transthoracic echocardiogram (TTE) and cardiac magnetic resonance imaging (CMR). A regression model obtained comparing G-SPECT and CMR was used to independently predict CMR LVEF from G-SPECT. In this retrospective study, 38 patients with cardiac ATTR amyloidosis (Perugini 2-3) underwent TTE, CMR, and G-SPECT within six months, excluding acute cardiac events or therapeutic changes. G-SPECT were evaluated using QPS-QGS software. A wall-model of polar map presentation was obtained from LV short-axis slices. In the first 30 consecutive patients, mean LVEF was 52.5%±11.3% by TTE, 57.9%±12.4% by CMR, and lower by G-SPECT (42.7%±10.1%, p < 0.001). LVEF from G-SPECT correlated significantly with CMR (r = 0.71, p < 0.001). The regression formula predicting CMR LVEF was: LVEF=0.875 x LVEFG + 20.612. Mean LVEDV was 74.4 ± 21.0 ml/m² by CMR and 61.4 ± 22.9 ml/m² by G-SPECT (p < 0.001), with strong correlation (r = 0.76, p < 0.001). In the last 8 patients, predicted CMR LVEF was comparable to measured CMR LVEF (p = 0.87), with significant correlation (r = 0.73, p = 0.039). This study reveals that Tc-DPD G-SPECT underestimates LVEF and LVEDV compared to CMR, though a consistent correlation was observed between the two modalities. As one of the first structured attempts to explore G-SPECT functional parameters in this setting, our work should be considered a proof of concept, highlighting both the feasibility and the current limitations of this approach. Further studies are needed to refine its accuracy and clarify its place within multimodal clinical imaging strategies.