Area at risk in acute myocardial infarction: oedema imaging and species-specific findings

Norlund et al.¹ from Lund University addressed two distinctly different aims in this edition of the European Heart Journal Cardiovascular Imaging. Their primary aim was to compare the diagnostic accuracy and image quality of T₂-STIR (T₂-weighted Short Tau Inversion Recovery) oedema imaging and CE-SSFP (contrast enhanced steady-state free precession cine MRI) with respect to determining myocardium at risk (MaR). Their second aim was to determine whether there is a bimodal temporal response to oedema or MaR imaging in humans as recently described in swine and possibly also present in dogs.

They used data from three separate studies to have multi-centre and multi-vendor representation of CMR MaR methodologies. They picked two clinical trials that used CMR to determine myocardial salvage: the CHILL-MI trial² and the MITOCARE study.³ These studies provided real world experience in clinical trials aimed at measuring myocardial salvage. To address the second aim of the study, they also included patients from a third study⁴ to ensure that they had data on Day 1 after myocardial infarction.

With respect to their first aim on image quality, Norlund et al.¹ found that T₂-STIR performs poorly for imaging MaR; only 65% of T₂-STIR datasets were considered of diagnostic quality. No clinical trial can afford to lose 35% of their population due to an imperfect imaging end point. Fortunately, CE-SSFP was of diagnostic quality in 97% of patients. That was even higher than the successful rate of completing myocardial infarction imaging with late gadolinium enhancement images, the critical second part of the CMR measurement of myocardial salvage.

The use of CE-SSFP cine MRI as a measurement of the MaR raised questions about how it works and how it might compare with other methods. Clearly CE-SSFP worked better than qualitative T₂-STIR images. Whether T₂ maps,⁵ T₁ maps,⁶ extracellular volume fraction maps, or other quantitative approaches might work better for quantifying MaR will require future study. A more fundamental but related question concerns exactly what is seen on CE-SSFP cine MRI. Why does CE-SSFP work for imaging MaR?

The paper does not provide enough details to delve deeply into the methods but do frame the general characteristics of why CE-SSFP cine MRI might highlight the MaR. First, the studies used multi-slice, multi-phase SSFP cine MRI methods available on each vendor's scanner. An SSFP sequence with TE in the range of 1.3–1.7 ms and a TR of 2.9–3.7 ms has intrinsic T₂/T₁ contrast.⁷ This means that signal intensity will increase as T₂ get longer, as expected with myocardial oedema in the MaR. Signal intensity also increases on an SSFP image as T₁ decreases since the T₁ is in the denominator of the contrast relationship. Since T₁ is longer in oedematous myocardium compared with normal myocardium, there will be a slight loss in signal intensity from this mechanism. Overall, MaR should be brighter on CE-SSFP 5 min after gadolinium-based contrast is injected due to both oedema and early gadolinium enhancement. Since gadolinium likely plays an important role in the success of CE-SSFP, timing issues become important since gadolinium is used in both the MaR and the myocardial infarction measurements of myocardial salvage.⁸

The second aim of the study highlights an important pathophysiological question concerning the dynamics of oedema after acute myocardial infarction. Fernández-Jiménez et al. from Madrid found that T₂ abnormalities and oedema followed a bimodal temporal response in a swine acute myocardial infarction model.⁹ T₂ abnormalities and post-mortem measurements of water content were prominent on Day 0, almost resolved by 24 h, were detectable on Day 4, but became prominent again by Day 7. In reperfused MI, neutrophils dominated ischaemic myocardium on Day 1 post-MI, macrophages dominated on Day 4, and collagen became prominent by Day 7.¹⁰ Reperfusion status and steroid treatment also modulated CMR and pathophysiological characteristics of the myocardium. They concluded that the first wave of pathologic changes were due to reperfusion while the second phase represented healing. Dogs may go through a similar bimodal pattern of oedema after acute myocardial infarction as evidenced by decreasing T₂ in the MaR from Day 0 to Day 2¹¹ and nearly complete resolution on Day 4.¹²

Human data do not support a conclusion that oedema fades in a few days or follows a bimodal pattern after acute myocardial infarction. Collectively, human studies from Lund,¹ Oxford,¹³ and the US National Institutes of Health¹⁴ do not show evidence of the bimodal temporal pattern of oedema during the first week after acute myocardial infarction. The study by Nordlund et al.¹ provides data in over 200 patients and spans Day 1–7 post-MI with data analysis at the level of current clinical trials. Dharmakumar concluded that the bimodal temporal course of oedema in the swine model will ‘… add to the understanding of tissue characteristics of acute MI, but are limited to the specific animal model investigated’.¹⁵ Perhaps this means that there is an overlap in the time period of post-ischaemic oedema and the start of the healing phase in humans.

Preclinical experiments clearly play a critical role in understanding pathophysiology, and the studies highlighted in this section have advanced our understanding of T₂ after acute MI. T₂ is not a simple and stable binary marker of the area at risk in pigs over the course of a week^9,10 and in dogs over the course of 4 days.^11,12 These studies highlight the need for caution in over-interpreting imaging results during a dynamic and complicated post-infarct phase of healing. By analogy, different mammalian species, including humans, develop myocardial necrosis during acute MI at markedly different rates.¹⁶ If the timing of intervention for acute MI was based solely on swine data, there would be no point to human acute percutaneous coronary interventions beyond ∼90 min after the onset of coronary occlusion let alone door-to-balloon time. Pigs and rats develop transmural MI much faster than dogs and humans. We are fortunate to have a little more time than predicted by swine models of acute myocardial infarction.

In conclusion, there is still much to learn about acute myocardial infarction. Imaging can play an important role in this learning process. It is important to keep an open mind to new findings. It is also important to recognize the limitations of both preclinical and human-level studies. Neither the level of research can answer all of the important questions. At some point, we need to synthesize the information gained and use what we have learned to guide patient care and to help develop new treatments and better diagnostic methods.

Conflict of interest: A.E.A. has research agreements with Siemens and Bayer Healthcare. These are institutional level research agreements.

Funding

This study was funded by the National Heart, Lung, and Blood Institute Intramural Research Program (HL006136-06).