The researchers turned to eight male Saanen goats, animals chosen for their manageable size, low maintenance cost, and coronary anatomy that closely resembles the human heart with only sparse intrinsic collateral circulation. Rather than performing open-chest coronary ligation, the team used a fully interventional approach. Under fluoroscopic guidance, a balloon catheter was advanced into the left circumflex artery and inflated for 30 minutes to induce regional ischemia. Immediately afterward, a 1.7-F microcatheter was pushed distal to the occlusion site and roughly 5.8 million polystyrene microspheres, each just 20 micrometers in diameter, were slowly injected in divided boluses suspended in iohexol contrast. The design deliberately mimics the distal embolization of thrombotic and atheromatous debris that occurs during clinical percutaneous coronary intervention, reproducing embolization-induced microvascular obstruction rather than isolated coronary microvascular dysfunction.

Electrocardiographic monitoring tracked the model’s physiological fidelity. Coronary occlusion produced T-wave and ST-segment elevation, and after reperfusion the ST segments resolved by 30 percent or less, a pattern the authors interpreted as evidence of persistent microvascular perfusion impairment. One goat died during ischemia-reperfusion from ventricular fibrillation that could not be reversed, but the remaining seven survived to the planned endpoint without sustained malignant arrhythmias. That one-in-eight procedural mortality is notably lower than what traditional thoracotomy-based ligation models typically produce, underscoring the minimally invasive character of the technique.

Right after model establishment, each surviving goat underwent DSCT-MPI on a third-generation dual-source CT system. Iohexol contrast was injected intravenously at 3.0 milliliters per second, scan initiation was triggered at a region-of-interest threshold of 100 Hounsfield units, and imaging began seven seconds after peak temporal enhancement. The resulting iodine distribution maps quantify myocardial blood volume and regional perfusion status. Two blinded readers independently analyzed the maps, achieving an intraclass correlation coefficient of 0.88 for quantitative iodine measurement and a Cohen’s kappa of 0.82 for categorical perfusion classification. Any segment showing an iodine concentration drop greater than 20 percent relative to neighboring normal myocardium was defined as significantly hypoperfused.

One week later, the goats were euthanized and their hearts excised and sliced transversely from apex to base. Triphenyltetrazolium chloride staining distinguished viable bright-red myocardium from necrotic grayish-white regions, and hematoxylin-eosin staining revealed necrotic cardiomyocytes with inflammatory infiltration, interstitial edema, and nuclear pyknosis in infarcted areas. Using the American Heart Association’s 17-segment model, the team classified 119 myocardial segments across the seven goats: 39 were infarcted, 66 were normal, and 14 showed significant hypoperfusion on DSCT-MPI without TTC-confirmed necrosis. Those 14 segments, representing 11.8 percent of the total, were classified as non-infarcted hypoperfused regions consistent with distal microvascular obstruction rather than as coronary microvascular dysfunction, a distinction the authors emphasize as pathophysiologically critical.

The molecular signature of the infarcted tissue told a coherent story of reperfusion injury. Enzyme-linked immunosorbent assays quantified three apoptosis-related proteins in myocardial homogenates. Bcl-2, an anti-apoptotic survival factor, fell from 96.93 in normal myocardium to 52.20 in infarcted tissue. Bax, a pro-apoptotic protein, rose from 47.53 to 128.06, and caspase-3, the executioner enzyme of programmed cell death, surged from 29.44 to 123.16. All differences were highly significant at P less than 0.01. Iodine concentration dropped in parallel, from 3.47 milligrams per milliliter in normal myocardium to 1.17 in infarcted tissue, demonstrating concordance between the biochemical injury cascade and the imaging abnormalities.

Receiver operating characteristic analysis, performed on the 39 infarcted and 66 normal segments while excluding the 14 MVO-consistent segments, identified an optimal iodine concentration cutoff of 1.582 milligrams per cubic centimeter. At this threshold, DSCT-MPI achieved 89.7 percent sensitivity and 90.9 percent specificity for detecting infarcted myocardium, with an area under the curve of 0.848 and a 95 percent confidence interval of 0.736 to 0.959. Agreement with histopathological classification was substantial at a Cohen’s kappa of 0.798. Overall accuracy reached 90.5 percent, with 35 true positives, 60 true negatives, 6 false positives, and 4 false negatives. Because myocardial segments from the same goat are statistically correlated, the team used a generalized estimating equation framework with goat as the clustering variable rather than treating segments as fully independent observations.

Pairwise comparisons further quantified the perfusion gradient across the three segment categories. Infarcted segments averaged 1.37 milligrams per milliliter of iodine, MVO-consistent segments 1.69, and normal segments 2.12. The differences were all highly significant, with infarcted versus normal yielding a t-statistic of -20.35 and MVO-consistent versus normal a t-statistic of -10.23. Crucially, the MVO-consistent segments showed intermediate hypoperfusion that was significant but less severe than in infarcted tissue, confirming that DSCT-MPI can distinguish among normal, infarcted, and microvascular-obstructed myocardium in a single examination.

The authors are careful to frame these findings as exploratory pathology-based validation rather than a head-to-head comparison with cardiac magnetic resonance, which remains the reference noninvasive modality for myocardial tissue characterization. No simultaneous or sequential CMR examination was performed. DSCT-MPI offers complementary advantages, including higher spatial resolution, rapid image acquisition, and the ability to integrate with coronary CT angiography in a single scan, but it involves ionizing radiation and iodinated contrast. The team also acknowledges that the 20-micrometer monodisperse polystyrene microspheres represent an idealized embolic material that does not replicate the heterogeneous fibrin, platelet, erythrocyte, and lipid-rich plaque debris encountered clinically, a homogeneity that may have contributed to the relatively high imaging-pathology agreement observed.

Several other limitations temper the conclusions. DSCT-MPI was performed at a single time point immediately after model establishment, potentially capturing a period of pronounced perfusion impairment while missing the dynamic evolution of microvascular obstruction over subsequent days. The iodine threshold was derived and evaluated within the same exploratory dataset and may be optimistically estimated. With only seven analyzable animals, the effective independent sample size remained small, and the authors estimate that approximately 16 to 19 animals would be needed for a future confirmatory study. No echocardiographic assessment of regional wall motion or left ventricular function was performed. Nevertheless, the study demonstrates that a minimally invasive, reproducible goat model of embolization-induced myocardial reperfusion injury is feasible and that DSCT-MPI can descriptively identify both infarcted myocardium and non-infarcted hypoperfused regions consistent with distal microvascular obstruction, providing a translational platform for future investigations of post-reperfusion microvascular injury.

Subject of Research: Cardiovascular Medicine

Article Title: Assessment of myocardial perfusion using dual-source computed tomography myocardial perfusion imaging in a goat model of reperfusion injury induced by distal microembolization

Article References: Jin, H., Duan, C., Chen, X., Lu, Y., Liu, X., & Li, M. (2026). Assessment of myocardial perfusion using dual-source computed tomography myocardial perfusion imaging in a goat model of reperfusion injury induced by distal microembolization. Frontiers in Cardiovascular Medicine, 13, Article 1809725. https://doi.org/10.3389/fcvm.2026.1809725

Image Credits: AI Generated

DOI: 10.3389/fcvm.2026.1809725

Keywords: dual-source CT, myocardial perfusion imaging, microvascular obstruction, reperfusion injury, goat model, cardiac imaging, myocardial infarction, translational research, Assessment, myocardial, perfusion, dual-source