Title : Territory-resolved cardiac MRI feature-tracking strain enhances detection of obstructive coronary stenosis in hypertension
Abstract:
Background: In patients with essential hypertension, left ventricular (LV) global strain reduction often reflects diffuse remodeling rather than focal ischemia, making it difficult to differentiate functional impairment caused by obstructive coronary artery stenosis (OCAS). This study evaluated whether coronary-territory–resolved strain derived from cardiac MRI (CMR) feature tracking could improve the detection of OCAS compared with traditional global strain and angiographic Gensini scores.
Methods: We retrospectively analyzed 185 hypertensive patients who underwent both invasive coronary angiography and CMR, including 135 with coronary stenosis and 50 without. Fifty-two normotensive individuals served as controls. Global and segmental peak strains (radial, circumferential, and longitudinal) were measured from cine CMR using a 16-segment model. Each myocardial segment was assigned to its corresponding coronary territory (LAD, LCX, RCA) according to AHA recommendations. The Gensini score quantified stenosis severity. Logistic regression and ROC analyses were used to assess the incremental diagnostic value of territory-resolved strain for identifying OCAS (≥50% stenosis).
Results: Territory-resolved strain decreased progressively from controls to hypertensive patients without stenosis and further to those with OCAS (all P < 0.05). In the OCAS subgroup, ischemic territories showed significantly lower peak strain than non-ischemic regions (circumferential: −14.2 ± 3.8% vs. −18.9 ± 4.1%; longitudinal: −10.3 ± 2.9% vs. −15.1 ± 3.4%; both P < 0.001). The basal-to-apical strain gradient was blunted in patients with OCAS (P < 0.01). In multivariable logistic regression, territory-resolved circumferential strain remained independently associated with OCAS after adjusting for Gensini score and global strain (β = −0.271, P < 0.01). ROC analysis demonstrated improved diagnostic discrimination when adding territory-resolved strain to the model (AUC 0.94 vs. 0.87, P = 0.02).
Conclusion: Coronary-territory–based CMR strain analysis accurately matches the angiographic distribution of disease and significantly enhances the identification of hemodynamically significant coronary stenosis in hypertensive patients. Integrating anatomical (angiographic) and functional (strain-based) assessment offers a more comprehensive diagnostic strategy for hypertensive individuals at risk of ischemic heart disease. These findings highlight the clinical potential of CMR feature-tracking strain as a noninvasive tool to bridge structural and perfusion evaluation in routine hypertension management.
