In a closed-chest porcine model, mechanical dyssynchrony, as evidenced by the standard deviation in TTP, occurs early after MI and does not significantly worsen in the near term. Dyssynchrony originates primarily from delayed electrical and mechanical activation of the infarcted region.
Myocardial infarction and its common consequence, heart failure, present a significant health problem in the United States and the world . Despite strong clinical gains from CRT in the overall heart failure population, results in ischemic CM have been underwhelming . Areas of infarction have delayed mechanical activation due to local conduction abnormalities, delays in electro-mechanical coupling, and myocardial dysfunction. However, the mechanical relationship between infarct areas and peri-infarct myocardium is unclear . Moreover, the mechanical behavior of the peri-infarct zone with respect to dyssynchrony is unclear. To address this knowledge gap we studied regional mechanics early and late post-MI in an extensively characterized porcine model of MI [15, 24, 25]. We selected a closed-chest approach to avoid the confounding influence of sternotomy and pericardiectomy that are known to affect myocardial mechanics. This model exhibited all the classic features of morphologic and functional remodeling seen in clinical and experimental MI. Additionally, using high resolution ex vivo CMR in this animal model, Ashikaga et al  have demonstrated a complex 3D structure of the scar: they found a thin rim of viable myocardium on the endocardial aspects of the scar (endocardial border-zone) and islands of viable myocardium within transmural-appearing scar that would result in fragmented electrograms and delayed activation of the infarcted region (by endocardial mapping).
Our model allows us to reliably characterize the distribution and extent of infarct by LGE. Similarly, we were able to evaluate the time course of changes in regional contractility and dyssynchrony following MI using CMR tagging. Unlike echocardiography, tagged CMR allows evaluation of myocardial strain at high spatial resolution and reproducibility. Strain, which evaluates regional myocardial deformation, is more reflective of myocardial mechanics than displacement mapping using parameters such as tissue velocity which are prone to artifacts from translational motion and tethering . These artifacts may lead to high variability in tissue velocity-based indices of dyssynchrony . This advantage of strain over velocity mapping is more pronounced in regional pathologies such as myocardial infarction . The dyssynchrony index used in this study has been previously validated in CMR based studies of dyssynchrony [19, 20]. It offers the best snapshot of the mechanical behavior of individual segments relative to the entire heart provides a numerically expression for the temporal dispersion in mechanical activity in the heart. One potential advantage of CMR based zHARP assessment of dyssynchrony is that the multiple peaks, often noted in echo-based tissue velocity traces, were not observed in this study. However, this observation may require a wider population to be confirmed.
Our study demonstrates significant mechanical dyssynchrony within days of an MI, which is in concordance with previously published work using echocardiography in clinical populations . Echocardiographic tissue velocity-based dyssynchrony indices suggest a standard deviation of time to peak displacement of approximately 30 ms represents significant mechanical dyssynchrony [28, 29]. Although we cannot directly extrapolate these echo-based results, it is noteworthy that our index of global dyssynchrony (SD16) was significantly abnormal at 1 week post-MI (50 ms). Our data show that dyssynchrony occurs early and provide insights into why echo-based evaluation of dyssynchrony days after MI was highly predictive of long term outcomes [8, 10]. However, since different imaging methodologies and indices were used in our study versus the previous echocardiography-based studies, wider, direct comparisons between our data and previous echo-based data are difficult. Furthermore, the subjects are not exactly the same, since patients may have more extensive disease and larger MIs, including acute on chronic ischemia and multi-vessel disease, compared to the well-circumscribed, relatively small apical MI in our model. We did not evaluate longitudinal displacement as done by Zhang et al  and decided not to use radial strain, as done by Mollema et al. , since we were unable to obtain adequate quality εR tracings by CMR. Recent data questioning the reliability and therefore usefulness of echo-based dyssynchrony indices also need to be considered when comparing our results to these previous publications 
Another important finding of our study is that delayed mechanical activation is not linked to reduced regional function per se. Regional contractility and conduction velocity were lowest in MI segments and intermediate in peri-MI segments, unlike the study by Klemm et al.  in patients with ischemic cardiomyopathy, which found viability and increased CV in areas with slow wall motion. We did not find mechanical delays in the peri-MI zones despite reduced regional function at 1 week post MI. Whether this is due to mechanical tethering of the peri-MI segments to the normal segment or a true lack of regional dyssynchrony could not be assessed by our study.
Conduction velocity was significantly lower in the infarct and infarct border-zone (IBZ) when compared to the remote myocardium. The infarct was the latest activated region because of very slow conduction, indicating that reduction in wave propagation velocity is the most important contributor to the time delay in regional contraction that we observed. Impulse propagation in the heart is dependent on active membrane properties determined by the ion channel composition, cell size, gap junction function and distribution . Previous work  has demonstrated that surviving myocytes in the healed IBZ have normal resting membrane potential and normal action potential morphologies. However, CV can be reduced in the IBZ due to distortion of myocyte alignment (non-uniform anisotropy), interstitial fibrosis, and/or gap junction remodeling [32, 33]. This may explain the reduction in CV without significant change in TTP in the IBZ. Alterations in Ca2+handling and Ca2+ transients that have been previously reported in infarct border-zone myocytes , in combination with non-uniform anisotropy in the IBZ could manifest as reduced peak systolic strain.
Another possible explanation for lack of a relationship between delayed mechanical activation and reduced regional function is infarct expansion, although we did not see an increase in the number of delayed-enhancement segments late post-MI, suggesting this was not a dominant mechanism in our study. Lastly, differences may be due to segment definition, compared to previous studies: a peri-MI segment in our study was a segment adjacent to an MI segment (Figure 1) and was not a partial thickness MI. We used this definition as in our model of a well-circumscribed MI, there were few if any segments with partial thickness LGE.
There are several clinical implications of our findings for CRT in patients with ischemic CM. Our data indicate that 1) delayed electrical and mechanical activation of the infarct is the main cause of dyssynchrony; 2) despite adverse remodeling of the left ventricle post-MI, further worsening of mechanical dyssynchrony does not occur. Hence, assessment of dyssynchrony one week post-MI or before discharge from the hospital, should be adequate to assess the consequences of MI on mechanical synchrony. The one week time point was chosen for logistical reasons in this animal study. However, based on the evolution of myocardial infarction, imaging any time in the first week post-MI should suffice. While the relationship between scar burden and lack of response to CRT in patients with ischemic CM has been reported before [35, 36], the underlying mechanism has not been completely elucidated. Based on our results, the current standard practice of pacing the lateral wall is unlikely to substantially change global dyssynchrony unless the infarct is in the paced region. Additionally, pacing a normal region may in fact worsen cardiac mechanics in ischemic CM by bypassing the His-Purkinje system and relying on cell-cell electrical propagation. Simply placing the LV lead in an infarcted territory may also not be the best option  since these segments would be unable to respond mechanically because of inadequate viable myocardium. Hence, despite the presence of mechanical dyssynchrony, patients with transmural infarcts may not respond to traditional CRT post-MI when the late activated region corresponds to the infarct. Also, beta blockers that reduce adverse remodeling and improve mechanical dyssynchrony in non-ischemic cardiomyopathy may not be effective in reducing dyssynchrony post-MI, because dyssynchrony  was caused by massive loss of myocytes in the infarcted region and was not the result of adverse remodeling. Based on our results, this type of dyssynchrony would be most amenable to strategies that promote regeneration of viable myocardium and improvement of conduction in the infarcted region .