Intramyocardial hemorrhage commonly occurs in large reperfused myocardial infarctions. However, its long-term fate remains unexplored. We hypothesized that acute reperfusion intramyocardial hemorrhage leads to chronic iron deposition.
Fifteen patients (mean age = 58±8 years; 3 women), who underwent successful angioplasty for first STEMI, were recruited following informed consent. Cardiovascular Magnetic Resonance (CMR) imaging (1.5T) was performed on day 3 and month 6 post-angioplasty. 2D T2* maps (6 TEs = 2.6-13.7 ms; ΔTE=2.2ms) and Late Gadolinium Enhancement (LGE) images of the entire left ventricle (LV) were acquired. Threshold-based image analysis was performed to identify remote, hemorrhagic (Hemo+) and non-hemorrhagic (Hemo-) myocardium.
Fourteen canines, subjected to ischemia-reperfusion (I-R) injury (3 hours of LAD occlusion followed by reperfusion), underwent CMR (1.5T) on days 3 and 56 post-I-R injury. Three sham-operated animals (Shams) were also studied using CMR at similar time points. 2D T2* maps (6 TEs = 3.4-18.4 ms; ΔTE=3.0ms) and LGE images of the entire LV were acquired. Threshold-based image analysis was performed to identify remote, Hemo+ and Hemo- myocardium. Subsequently, animals were euthanized (day 56), hearts were excised and imaged ex-vivo. Sections of Hemo+, Hemo-, remote and Sham myocardium were isolated and histology was performed. The concentration of iron ([Fe]) within each type of tissue was measured using mass spectrometry.
Six months post-angioplasty, mean T2* of the scar tissue in patients with Hemo+ infarctions (n=11 as determined by T2* losses within the infarct on day 3 CMR; Figure 1) was 40% lower than that of remote myocardium, suggesting chronic iron deposition (p<0.001). In contrast, mean T2* of Hemo- infarctions (n=4) was not significantly different from that of remote myocardium at both 3 days and 6 months post-angioplasty (p=0.51).
Figure 1. Patient Studies - Representative CMR images (A; acquired from a 42-year old patient following successful angioplasty for first STEMI) with significant T2* loss (arrows) at the site of acute and chronic myocardial infarction (identified by LGE imaging, arrows) are shown. Mean T2* of Hemo+ (B) was 40% lower than that of Hemo- and remote myocardium (p<0.001) on both acute and chronic CMR studies.
In canines, in-vivo mean T2* of Hemo+ myocardium was 40% lower than those of Sham, remote and Hemo- myocardium (p<0.001) at both 3 days and 56 days post-I-R injury (Figure 2B). Similarly, mean ex-vivo T2* of Hemo+ myocardium was 40% lower than those of Sham, remote and Hemo- myocardium (p<0.001; Figure 2C). Perl's stain confirmed localized chronic iron deposition only within Hemo+ infarctions. Mean [Fe] of Hemo+ infarctions was nearly 10-fold higher than those of Sham, remote and Hemo- myocardium (p<0.001; Figure 2D). A strong linear relationship was observed between log(ex-vivo T2*) and -log([Fe]) (R2=0.7; p<0.001; Figure 2E).
Figure 2. Animal Studies -Representative CMR images (A; T2* in the top row and LGE in the bottom row) acquired from an animal with hemorrhagic myocardial infarction in acute and chronic phases along the long- and short-axis (along the dashed red line in the long-axis images), together with corresponding ex-vivo images are shown (A). In-vivo T2* images (both acute and chronic phases) clearly demonstrate the evidence of signal loss in the LAD territory. Both mean in-vivo T2* (B) and ex-vivo T2* (C) of Hemo+ sections were 40% lower than those of Shams, Remote, and Hemo- in both acute and chronic phases (p<0.001). Mass spectrometric analysis (D) showed that iron content of Hemo+ tissue was 10-fold higher than that of other tissues (p<0.001). Linear regression analysis (E) between log(ex-vivo T2*) and -log([Fe]) showed a strong correlation (R2 = 0.74; p<0.001).
Acute reperfusion intramyocardial hemorrhage leads to regional chronic iron deposition within the infarct zones. T2* CMR can accurately characterize localized chronic iron deposition following reperfusion-induced myocardial hemorrhage. The clinical significance of this finding remains to be investigated.
This work was supported in part by grants from American Heart Association (SDG 0735099N) and National Heart, Lung, And Blood Institute (HL091989).