Non-contrast T₁ and T₂ relaxometry characterizes reperfusion injury of acute MI in swine

Background

Reperfusion injury in acute myocardial infarction (MI) results in edema, necrosis, microvascular obstruction (MVO), and intramyocardial hemorrhage (IMH), the latter presents an interesting clinical target. [1] Cardiovascular MRI has been shown capable of characterizing all of these tissue components. Other than MVO, which is currently detected by flow-deficient regions in contrast enhanced imaging, all other tissue components can be identified by T₁ and T₂ (T₂*). Theoretically, the byproducts of blood breakdown observed with IMH lead to decreased T₁ and T₂ (T₂*). [2] Conversely, free water accumulation (edema) and necrosis lead to increased T₁ and T₂. [2] Hence, direct and quantitative measurement of relaxation rates is promising in myocardial tissue characterization, avoiding ambiguity typical of weighted images (i.e. T₂-weighted spin-echo), undesired signal loss from T₂* (weighted) images or the uncertainty introduced by contrast agent kinetics. Hypothesis: Combined T₁ and T₂ mapping can characterize reperfused MI without contrast agents.

Methods

MI was induced in swine by 1 (N=3) or 2 (N=3) hr balloon occlusion of the LAD after the first diagonal, with MRI 7-9 days post MI (Achieva TX, Philips). Relaxometry: 3D respiratory navigator-gated T₂-mapping [3]; 2D Breath-hold T₁-mapping (MOLLI) [4]. Clinical standard: breath-hold black-blood T₂W TSE (BB-T₂-STIR) [5]; early (3 min post) gadolinium-enhanced images (EGE) using PSIR and 0.2 mmol/kg Magnevist. [6]. IMH was identified in T₂W images/T₁/T₂ maps as areas of hypointensity surrounded by hyperintense signal/T₁/T₂ representing edema. MVO was defined in EGE images as hypointense areas surrounded by enhanced MI. The co-localization of tissue types among techniques was examined.

Results

IMH was detected in all animals with 2 hr occlusions, identified by decreased T₁ and T₂, and was spatially consistent with the hypoenhanced core in BB-T₂-STIR and with MVO in EGE. Edema was observed in all animals (elevated T₁and T₂). (Fig. 1)

Figure 1

Matched representative SAX images from swine without hemorrhage after 1 hr LAD occlusion (Case #1) (a-d) and with hemorrhage after a 2 hr occlusion (Case #2) (e-h). In Case #1, significant T₁ and T₂ elevation are present in both maps (c,d), though no MVO was observed with EGE (b) nor was a hypointense core present in T₂W images (a). In comparison, Case #2 shows clear IMH, demarked by decrease in T₁ and T₂, surrounded by edema, shown by increased T₁ and T₂ (g and h), all of which are excellently co-localized with that in T₂W (e) and EGE (f). Green arrowhead indicates edema; red arrowhead indicates the core of IMH. Note that T₁ mapping is influenced by off-resonance and lower image resolution. As the result of competing effects on T₁ and T₂ from IMH and edema, partial volume averaging makes the T₂ in IMH close to that of normal myocardium.

Planimetry showed that relative to remote myocardium, T₁ and T₂ of edema were significantly higher (p < 0.001 and p <1e-5, respectively), while within IMH T₁ was lower (p = 0.001) and T₂ the same (p = 0.28). (Fig. 2)

Figure 2

a. ROI-based T₁ & T₂ for edema, IMH and remote myocardium from matched T₁ and T₂ maps. b. Changes in T₁ & T₂ after subtraction of the reference values of remote myocardium. Edema had higher T₂ than both remote myocardium or IMH. T₁ can be used to discriminate between edema, IMH and normal myocardium, though the distributions may overlap. Edema and IMH can be classified with higher specificity using both T₁ and T₂.

Conclusions

Though either T₁ or T₂ can be used to separate tissues, combined T₁ and T₂ mapping may allow for more accurate detection of IMH in reperfusion injury, without variability from contrast kinetics, or BB-T₂-STIR artifacts. [7] Based on a small number of animals, T₂ was superior in edema detection, while T₁ performed better in IMH detection. Combined relaxometry may identify tissues with better specificity than individual and may help clarify the link between MVO and IMH. High-resolution relaxometry may be necessary to avoid partial volume.

Funding

Funded in part by the American Heart Association - 11SDG5280025.