Texture Analysis of SPECT-MPI Provides Prognostic Value in Improving Cardiac Resynchronization Therapy Response

Texture analysis (TA) is a powerful tool for extracting quantitative information, assessing myocardial heterogeneity, evaluating therapeutic efficacy, and predicting outcomes in heart disease. This study investigated whether TA based on gated single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) can enhance the prediction of response to cardiac resynchronization therapy (CRT).

A total of 165 patients who underwent gated SPECT MPI and received CRT were enrolled in the study. Quantitative analysis of SPECT imaging generated 1225 TA features. Phase analysis of resting gated short-axis SPECT myocardial perfusion images was utilized to assess left ventricular (LV) systolic and diastolic mechanical dyssynchrony (LVMD), including phase standard deviation (PSD), phase bandwidth (PBW), and entropy. Patients were categorized into CRT response and non-response groups based on a ≥5% improvement in LV ejection fraction (LVEF) measured by echocardiography at the 6-month follow-up. Variables with a p-value <0.05 in the univariate logistic regression analysis were incorporated into a backward stepwise multivariate logistic regression model for further analysis.

During follow-up, 60.0% (99 of 165 patients) demonstrated a response to CRT. Univariate logistic regression analysis revealed that CRT response was significantly associated with N-terminal pro-brain natriuretic peptide (NT-proBNP), non-sustained ventricular tachycardia (NS-VT), LV end-diastolic diameter (LVEDD), LV end-systolic diameter (LVESD), scar burden, systolic and diastolic PSD, PBW, entropy, and 51 TA parameters. In the backward stepwise multivariate regression analysis, inverse difference moment normalized (IDMN), NS-VT, NT-proBNP, diastolic PSD, and LVEDD emerged as independent predictors of CRT response.

TA based on gated SPECT MPI provides independent prognostic predictor for CRT response in medically treated Heart failure patients.