The term “myocardial stunning” describes “mechanical dysfunction that persists after reperfusion despite the absence of irreversible damage and despite restoration of normal or near-normal coronary flow”, that is the myocardium can return to normal function via revascularization (27). Experiments in a pig model (28) revealed that myocardial stunning resulted from increased interfilament lattice distance due to edema presence, and the severity of myocardial stunning primarily depends on the ischemia duration (29). In human study, acute dysfunction recovered after prompt reperfusion, accompanying with the regression of myocardial edema (30). Repetitive episodes of stunning (i.e., myocardial ischemia) will lead to development of hibernating myocardium which represents a more severe form of myocardial stunning and thus less likely to recover function after revascularization (31). Various imaging techniques can be used to distinguish stunned or hibernating myocardium from scar tissue, including echocardiography and MRI under stress or using contrast enhanced protocols, and nuclear medicine. Patients who present a substantial amount of potentially reversible myocardium will benefit more from revascularization.
No-reflow phenomenon refers to the impedance of microvascular blood flow despite restoration of the patency of infarct-related coronary artery (32). It is first described in dogs by Kloner et al. (33) in 1974. It is caused by multiple factors that obstruct the microcirculation, involving endothelial cellular swelling and protrusions, myocyte swelling and tissue edema, vasospasm and downstream embolization of plaque and thrombus in the microvasculature (34) (Figure 4). Thus, it is also termed microvascular obstruction (MVO). The existence of no-reflow tells us that restoration of epicardial artery patency does not necessarily indicate adequate perfusion at the tissue level. The new concept of reperfusion therefore has shifted to “open epicardial artery and open microvascular hypothesis” (35). No-reflow or MVO is an ominous sign, being associated with poor prognosis (36). The assessment of MVO using cardiac MRI and its prognostic value will be discussed in chapter 4.
Utilization of cardiac MRI in post-MI LV remodeling
In this chapter, we will talk about the utilization of MRI in the setting of myocardial infarction. First, we will present some basic concepts regarding cardiac MRI. Second, we will introduce commonly used CMR sequences in the assessment of MI. Third, we will provide the literature review about the utility of cardiac MRI in characterizing myocardial infarcts and in the prediction of LV remodeling. Cardiac MRI has the unique ability to assess MI and LV remodeling. It can provide a thorough analysis at a single exam (61): cine imaging, to assess ventricular size, morphology and function; contrast enhancement imaging, to quantify myocardial infarct size and the extent of microvasculature injury; and T2 imaging, to evaluate the area at risk. Emerging techniques such as parametric imaging (T1,T2 mapping) and diffusion imaging may provide more fine information on myocardial architecture (6). Thus, cardiac MRI is nowadays the gold standard imaging modality for post-MI assessments. Its increasing use in clinical routine helps clinicians to establish accurate diagnosis, risk stratification, therapeutic decision making, and monitoring therapeutic efficacy. Infarct size as well as no-reflow assessed by MRI is revealed strong predictors for LV remodeling (62–65).
T1 and T2 relaxation
MRI has the unique ability to generate intrinsic contrast between different soft tissues. T1 relaxation, T2 relaxation and proton density are intrinsic tissue properties that determine the native image contrast in MRI. The time it takes for longitudinal magnetization (Mz) to increase from zero to 63% of its initial maximum value (M0) is known as T1 (Figure 6). The time required for transverse magnetization (Mxy) to decrease to 37% of its initial value is known as T2 (Figure 7). Direct measurements of tissue T1 and T2 values (i.e., T1 and T2 mapping) can be used to detect diseased state of tissues. Besides, we can enhance image contrast to be weighted toward T1 (T1-weighted imaging, T1WI) or T2 (T2-weighted imaging, T2WI) by modifying the lengths of repetition time (TR) and echo time (TE). Usually, short TR and short TE results in T1-weighted contrast whereas long TR and long TE produce T2-weighted contrast. Image contrast can also be modified by exogenous materials, namely contrast agents. For example, gadolinium chelates are often used in scar imaging. Myocardial scar appears brighter than normal myocardium when T1-weighted pulse sequences are used because of significantly reduced T1 value in scar region by accumulated gadolinium.
Cardiac imaging planes and LV segmentation model
Functional and anatomic evaluation of cardiac cavities requires multiple oblique planes along the axes of the heart itself (Figure 9). Vertical long-axis (VLA) view (or 2-chamber view) is used to evaluate the relationship between LA and LV. Horizontal long axis (HLA) view (or 4-chamber view) bisects all four cardiac chambers, providing assessment of chamber size, valve position. Short-axis view allows assessment of LV size, configuration and myocardial segments according to coronary artery territories.
The American Heart Association (AHA) proposed a 17-segment model in 2002 for regional assessment of left ventricle, in order to achieve a consensus by using different cardiac imaging modalities, including coronary angiography, echocardiography, single-photon emission computed tomography (SPECT), positron emission tomography (PET), cardiac MRI, and cardiac computed tomography (CT) (68). In this model, the LV is divided into three portions: basal (tips of the mitral valve leaflets), mid-cavity (papillary muscles), and apical (beyond papillary muscles but before cavity ends) (Figure 10a). The resulting distribution of myocardial mass for the basal, mid-cavity, and apical thirds of the heart are 35%, 35%, and 30%, which is close to autopsy data (69). Meanwhile, individual myocardial segments have been assigned to the three major coronary arteries (RCA, LAD, and LCX) (Figure 10b).
T2-weighted imaging, T2WI
Recently, T2-weighted imaging is used to depict AAR of acute MI (86) because it is sensitive to myocardial edema (87, 88). The amount of salvageable myocardium can be obtained when subtracting infarct size from the AAR (89), which is comparable to that measured by SPECT (90). Myocardial salvage index (MSI), intending to normalize myocardial salvage over the perfusion bed of different sizes, is often used: MSI = (AAR minus infarct size)/AAR. It possesses strong prognostic value and is a useful imaging biomarker to test new reperfusion therapies (16, 20, 91, 92). However, MSI value may vary depending on the measurement time frame after acute MI since edema gradually diminishes (93).
Although T2WI is useful in detecting myocardial edema, the technique per se is challenged by a set of problems (86, 94): (1) sensitive to motion artifacts and surface coil intensity variation; (2) subject to a low contrast between normal and affected myocardium; (3) bright signal artifact from stagnant blood flow, a problem probably encountered in patients with depressed LV function; (4) image interpretation variability may be caused due to insufficient signal-to-noise ratio (SNR).
Furthermore, the appropriateness of using T2WI to delineate myocardial edema in acute MI is still debated (95). First, robust validation experiments that compare T2WI against the true pathology are lacking. Second, apparent T2 findings are contradictory to physiologic basis. For instance, a homogeneously bright AAR is always reported using T2WI. Theoretically, however, the signal would have been heterogeneous because edema is not evenly present within the AAR: more edema in infarcted than in salvageable myocardium, thus the infarct area should appear brighter. Recently, Kim et al. (96) found that T2-weighted MRI closely represented infarcted regions rather than the area at risk.
T1 and T2 mapping seems promising alternative techniques to delineate the AAR (97–99). Langhans et al. (97) measured AAR in 14 acute MI patients using SPECT as reference technique. They found that at 1.5T T2 mapping (T2 threshold at 60 msec) showed the closest correlation with SPECT, intermediate with T1 mapping (T1 threshold at 1075 msec), and worst with T2WI which underestimated the AAR by 30%. Verhaert et al. (99) compared T2 mapping and T2-weighted short tau inversion recovery (T2-STIR) in 27 AMI patients (STEMI/non-STEMI: 16/11). T2 mapping was more sensitive than T2-STIR in detecting edema (96% vs. 74%). Typical examples are shown in Figure 13. Data, however, is still scant.
Late gadolinium-enhancement, LGE
LGE technique is routinely performed to assess myocardial viability. The presence of viability is associated with greater benefits of revascularization in chronic MI (100). LGE imaging is performed using T1-weighted GRE sequences after intravenous injection of gadolinium-based contrast agents. It has been extensively validated in animals (101) and in humans (102, 103). The technique is premised on a combination of increased gadolinium concentration and delayed washout in infarcted area compared to viable myocardium (104, 105). Normally, gadolinium diffuses only into extracellular space. In acute MI, the loss of cellular membrane integrity allows gadolinium to enter into the intracellular space, leading to increased gadolinium in regions of acute infarcts as well as delayed clearance (wash-out) due to decreased capillary density (106, 107). In chronic MI, discrete collagen fibre meshwork and loss of cellularity in scar region retains gadolinium (108) (Figure 14, 15). When T1-weighted sequence is applied after a delay of 10-30 minutes following contrast injection, MI regions that contain more gadolinium show hyperenhancement due to significantly shortened T1, namely late enhancement. The identification of MI depends on regional signal differences between normal and affected myocardial tissue. The in-plane resolution of LGE images is typically 1.4×1.4mm, which yields 5 to 10 pixels within LV wall, making it superior to SPECT technique in detecting subendocardial infarcts (102, 103). However, imaging time delay after contrast agent administration and the dose of contrast may influence the accuracy of LGE imaging (109–113).
Clinically, the transmural extent of late enhancement is usually assessed at a five-point scale: 0=0%, 1=1-25%, 2=26-50%, 3=51-75%, 4=76-100% (76). Segments with LGE transmural extent greater than 75% are considered nonviable because they have little likelihood to improve function at follow-up (114).
In addition to hyperenhancement pattern, a central hypoenhancement (dark zone) is frequently found in acute MI, namely no-reflow or microvascular obstruction, MVO (104, 105) (Figure 16). It is caused by reduced wash-in of gadolinium within that region due to severely injured microvasculature (Figure 17) and often occurs in transmural infarcts. The presence of MVO is associated with poor prognosis (115).
Table of contents :
I. INTRO DUCTION
I.1. The heart
I.2. Myocardial infarction
I.2.2 Reperfusion injury
2.1. Myocardial stunning
I.2.2.2. No No-reflow phenomenon
I.2.3 Infarct healing
Post-infarction left ventricular remodeling
I.3.2. Temporal evolution
I.3.3. Clinical importance
I.4. Utilization of cardiac MRI in post post-MI LV remodeling
I.4.1. Basic concepts of cardiac MRI
I.4.1.1. T1 and T2 relaxation
I.4.1.2. Synchronization in cardiac MRI
I.4.1.3. Cardiac imaging planes and LV segmentation model
I.4.2. Common cardiac MRI sequences used in myocardial infarction
I.4.2.1. Cine MRI
I.4.2.2. T2-weighted imaging, T2WIweighted imaging, T2WI
I.4.2.3. Late gadolinium-enhancement, LGEenhancement, LGE
I.4.3. Cardiac MRI findings and prognostic significance
I.4.3.1. Prognostic significance of infarct sizerct size
I.4.3.2. Infarct shrinkage
I.4.3.3. Infarct heterogeneity
I.4.3.4. No-reflow or microvascular obstruction, MVOreflow or microvascular obstruction, MVO
I.5. Synopsis of REMI study
I.5.2. Patient inclusion
I.5.3. Exams conducted
I.5.3.1. Blood tests
I.5.3.3. Cardiac MRI
II. Work accomplished
II.1. Standardization of methods for LGE quantification
II.2. Quantitative characterization of LGE region
II.3. Cardiac MRI predictors for LV remodeling
II.4. Impact of MVO on LV wall and local remodeling
III. Overall conclusion and perspectives