MRI evaluation of post-infarct residual myocardial viability

Massimo De Filippo
MD
Researcher in Radiology

Maurizio Zompatori
MD
Professor of Radiology
University of Parma
Parma, Italy

Myocardial contractility and contractile reserve, regional perfusion, structural alterations and the capacity to predict functional reserve of viable but dysfunctional zones are some of the information provided by magnetic resonance imaging (MRI), which can be considered as a “one-stop-shop” study. Viable but dysfunctional myocardial tissue (hibernating and stunned ­myocardium) can be identified by MRI with high accuracy by use of specific techniques and sequences. The importance of early identification may be derived from two types of data:

• The mortality rate is four times higher in patients treated with medical therapy than in revascularised patients.
• Recognition of nonvital areas enables patients to avoid the risks associated with revascularisation.

Cine cardiac MRI lies at the basis of the study of myocardial vitality, based on the rationale that contractile dysfunction is linked to ischaemic suffering. One of the main advances brought about by cine cardiac MRI has been the introduction of steady-state free precession (SSFP), better known by its commercial names of true-FISP or FIESTA (depending on the manufacturer). In addition to classic cine sequences, cine-tagging is a method that utilises a sequence of radiofrequency impulses immediately after the echocardiogram (ECG) R-wave to obtain presaturation bands, oriented perpendicularly to the image plane. These bands or tags are visualised as intersecting hypointense lines overlying the ­myocardium during the cardiac cycle, and their movement (or lack of it) facilitates the detection of ­dyssynergic segments. However, in the evaluation of myocardial vitality, the study of wall ­kinetics ­utilising cine MRI showed only a 66% concordance compared with fluorodeoxyglucose-positron ­emission tomography (FDG-PET).

The combined use of perfusion sequences and delayed enhancement (DE) with paramagnetic ­contrast agent (CA) to evaluate myocardial vitality renders the diagnostic accuracy of MRI comparable to that of PET,(1,2) with the advantages of having better spatial resolution (a fundamental characteristic to evaluate the extension of wall damage, subendocardial, transmural infarct, etc) and of being a ­noninvasive and less costly investigation.

MRI perfusion
The high temporal resolution (<1 sec) needed for the correct study of myocardial perfusion requires the use of high-magnetic-field machines (1.5 T) correlated with dedicated software. In addition, the use of a cardiac coil leads to a high signal and elevated contrast. An optimal MRI study of cardiac perfusion employs sequences able to record the first passage of the bolus as a video clip through the respective reconstruction of multiple high-temporal-resolution images, acquired in at least three short-axis anatomical sections from the base to the apex of the left ventricle during an entire cardiac cycle. In this situation, ultrafast gradient echo sequences are at present utilised with techniques of parallel imaging such as SMASH (simultaneous acquisition of spatial harmonics) or SENSE (sensitive encoding technique). ECG sychronisation (trigger ECG) is the first technical aspect to be considered: it guarantees a reference for examination quality. Image resolution depends on heart rate and the MRI ­technology utilised: recent machines allow image acquisition of two- to six-section images for each cardiac cycle completed, guaranteeing the capacity to record the contrast agent (CA) distribution in the heart ­chambers and walls.

A further advantage of modern MRI technology is that it does not require prolonged breath-­holding for the patient. The most common CA is gadolinium (a heavy metal from the lanthanide family) chelated to diethylenetriaminepentaacetic acid (gadopentetate dimeglumine). To improve visualisation of perfusion gradients, a bolus containing a minimal quantity of paramagnetic contrast material (0.05–0.1 mmol/kg) should be administered at a rate of 3 ml/s with an automatic electronic injector. During ­contrast material administration, the viable myocardium will show uniform signal intensity increase, from the subendocardial to the subepicardial zones, followed by regular washout of the contrast ­medium. It is important to note that the perfusion gradients are related to both reversible and irreversible ischaemia (see Figure 1).

[[HHE07_fig1_Ca6]]

The sensitivity and specificity of the first-pass perfusion bolus of the CA to identify perfusion defects are 74–92% and 87–96%, respectively, when compared with conventional coronary angiography;(3) ­gadolinium perfusion MRI, when compared with nuclear medicine, has a sensitivity and specificity of 65–82% and 75–81%, respectively.(1) The use of perfusion sequences enables the study of the heart’s inotropic reserve, similar to echography under pharmacological stress (adenosine, dipyridamole, dobutamine).(4) Still, studies on diagnostic accuracy, generally carried out on limited and selected case groups, are not yet sufficient to justify the routine use of stress MRI rather than stress echocardiography. Stress MRI should be considered as a third-level method, to be used in selected cases only.

Pathophysiology of delayed enhancement
Five to 20 minutes after the injection of the CA (0.1–0.2 mmol/kg) into a peripheral vein, short-axis, four-chamber view axis images should be acquired by use of fast gradient echo inversion recovery (FGE-IR) sequences, for which the optimal inversion time (IT) will be determined. The IT is the time, in milliseconds, between application of the 180° prepulse and acquisition of the signal at the centre of the K space. IT ranges from 140 ms to 300 ms in the cardiac muscle. The optimal IT for the DE sequences is that at which myocardial tissue presents magnetisation equal to zero (absent or poor signal intensity) and the blood in the heart chamber is slightly hyperintense. The higher concentration of gadolinium in the infarcted area, due to delayed washout, determines delayed IT relaxation, causing the infarcted area to appear “bright”. The necrotic area reaches a signal intensity about 400 volts greater than vital cardiac tissue (bright is dead). Short-axis scans are sufficient to study the entire ventricle except for the cardiac apex, which requires two- and four-chamber-view scans. At present, extracellular gadolinium chelates have intravascular and interstitial distribution. The hypothesis that explains late impregnation of the CA (delayed enhancement, or DE) in acute infarction is related to the strong permeability of the necrotic myocytes due to the rupture of the cytoplasmic membranes. In chronic ischaemia, DE is the result of impaired wash-in and washout of the CA due to an increase interstitial space (fibrosis) secondary to the reduction of the number of myocytes.(5) In a canine model, DE correlates in size with histological slices, not only in the acute phase but also after eight weeks, suggesting that this technique may also be useful to identify myocardial injury in chronic coronaropathy.(6) The high spatial resolution of MRI makes it useful to evaluate the transmural extent of the scar (see Figure 2).

[[HHE07_fig2_Ca6]]

Wide transmural scar extension predicts a low probability of functional recovery after ­revascularisation in patients with chronic coronaropathy: a transmural extension above 50% seems to be the cut-off point for revascularisation.(7) Hyperenhancement of 25% or less of myocardial wall thickness on DE-MRI after revascularisation seems to be correlated to complete recovery of thickness and the ­restoration of contractile function.(7) The DE-MRI technique is highly reproducible and shows good correlation with SPECT(8,9) and PET.(1,2) Areas of DE in the hearts of patients with dilatative cardiomyopathy or in healthy volunteers have been observed. Thanks to the high spatial resolution of MRI, small areas of scar ­tissue can be identified in patients with coronary arterial disease (CAD) who have normal findings on SPECT(10) or in non-Q-wave chronic infarction.(11) DE is therefore specific for regional myocardial ­damage, which, in any case, is not exclusively caused by CAD. In fact, it can also be observed in patients with obstructive hypertrophic cardiopathy, with concurrent ischaemia secondary to hypertrophy,(12) in acute perimyocarditis,(13) in right ventricular arrhythmogenic dysplasia or in sarcoidosis with myocardial involvement (see Figures 3–6). The extent of injury can be quantified by mapping the regions of hyperenhancement and indicating the percentage of damage within a single region (trans/subendocardial extent). Extensive infarctions present as areas with predominant transmural hyperenhancement in the territory supplied by the occluded artery. Moreover, it is possible to identify minute areas of myocardial necrosis such as those observed after percutaneous coronary intervention, in which a mild elevation of serum creatine kinase-MB or elevated levels of troponin 1 can be present.(14) In this regard, it should be noted that the topographic distribution of DE takes on major diagnostic significance. In fact, the presence of DE confined to the subepicardium cannot, for haemodynamic reasons, assume ­ischaemic significance, although it may instead be related to irreversible myocardial injury, as already noted (for example, in myocarditis–pericarditis or sarcoidosis) and when consistent with the clinical picture. High concentrations of gadolinium are found only in the areas with irreversible myocardial damage. If microcirculation is intact, the penetration of gadolinium in the infarcted area is ­comparable to what is observed in normal or ischaemic but vital tissue. Still, gadolinium clearance is considered prolonged in the infarcted areas with respect to both functional and dysfunctional tissue. Microvascular damage after a recent myocardial infarction, indicated as the no-reflow phenomenon by many authors, is indicative of an unfavourable prognosis, such as ventricular remodelling, cardiac decompensation, severe arrhythmias and sudden death.(15)

[[HHE07_fig3_Ca7]]

[[HHE07_fig4_Ca7]]

[[HHE07_fig5_Ca8]]

[[HHE07_fig6_Ca8]]

Although the cause of no-reflow is not well understood, it is likely that microvascular obstruction is due to endothelial swelling, interstitial oedema and capillary obstruction.(16) Penetration of the CA in the area with microcirculatory obstruction is notably slow, evidencing a low gadolinium concentration with respect to the adjacent infarcted tissue that presents conserved microcirculation.(17) In the quantitative analysis of irreversible injury, the presence of nonenhancing subendocardial areas (no-reflow) within the nonenhancing segment must be included in the overall assessment of the infarcted tissue (see Figure 5). Microvascular damage, characterised by perfusion sequences and low-signal DE, both before and after the administration of CA, is present only in patients with acute myocardial infarction; DE is a nonspecific marker of irreversible myocardial damage.

References

1. Klein C, et al. Circulation 2002;105:162-7.
2. Kuhl HP, et al. J Am Coll Cardiol 2003;41:1341-8.
3. Plein S, et al. Radiology 2005;235(2):423-30.
4. Nagel E, et al. Z Kardiol 1999;88:622-30.
5. Lima JA, et al. Circulation 1995;92:1117-25.
6. Kim RJ, et al. Circulation 1999;100:1992-2002.
7. Kim RJ, et al. N Engl J Med 2000;343:1445-523.
8. Ramani K, et al. Circulation 1998;98:2687-94.
9. Mahrholdt H, et al. Circulation 2000;106:2322-7.
10. Wagner A, et al. Lancet 2003;361:374-9.
11. Wu E, et al. Lancet 2001;357:21-8.
12. Choudhury L, et al. J Am Coll Cardiol 2002;40:2156-64.
13. Klein C, et al. Circulation 2003;107:82.
14. Ricciardi MJ, et al. Circulation 2001;103:2780-3.
15. Wu KC, et al. Circulation 1998;97:765-72.
16. Reffelmann T, et al. Heart 2002;87:162-8.
17. Judd RM, et al. Circulation 1995;92:1902-10.