Illuminating the Path Forward in Cardiac Regeneration Using Strain Magnetic Resonance Imaging.

In recent years, several important studies have shown advantages of strain imaging over imaging of ejection fraction (EF) for diagnostic and prognostic applications in heart disease. For example, global longitudinal strain assessed using echocardiography has demonstrated superior performance compared with echocardiographic assessment of EF for predicting major adverse cardiac events in heart failure, myocardial infarction, and valvular heart disease. Likewise, strain is now recommended in addition to EF to identify systolic dysfunction in chemotherapy patients. Using accurate and reproducible magnetic resonance strain imaging methods, such as myocardial tagging and displacement encoding with stimulated echoes, strain detects subclinical systolic dysfunction in diabetics and obese children at stages where changes in EF are not seen. Using regional analysis methods, strain imaging shows high potential for detecting late activating regions and optimizing the implementation of cardiac resynchronization therapy, providing more valuable information than cine imaging. This recent wave of impressive strain imaging successes has led to the suggestion that strain imaging could potentially replace or supercede EF. Within the field of cardiac regeneration, the efficacy of emerging therapies is often assessed by measuring longitudinal changes in EF in large groups of small animals. However, such measurements do not directly interrogate contractile function localized to specific regions that were damaged and treated. In this issue of Circulation: Cardiovascular Imaging, the article by Qin et al from the laboratory of Dr Joseph Wu extends the successful application of strain imaging into the realm of preclinical cardiac regenerative therapies, showing that sensitive and accurate magnetic resonance imaging (MRI) strain using myocardial tagging detects subtle beneficial effects of engineered heart muscle implanted at sites of postinfarct scar, whereas imaging of EF does not. See Article by Qin et al The experiments by Qin et al tested the hypothesis that engineered heart muscle could preserve contractile function in regions of the heart where scar formed 28 days after induction of myocardial infarction. The experimental group consisted of rats with postinfarct scar that were treated with engineered heart tissue, whereas the control group consisted of rats with postinfarct scar that underwent a sham operation with stitching, but were not treated with engineered heart muscle. Both groups of rats were studied 4 weeks after treatment or sham surgery using an imaging protocol that included multislice cine imaging, late gadolinium–enhanced imaging, and myocardial tagging. Ultrasound imaging was also performed. Although EF did not show any difference between the control and treatment groups and end diastolic volume and scar size only showed trends toward lower values in the treated group, circumferential strain derived from myocardial tagging localized to the scar region as defined by late gadolinium– enhanced imaging showed a statistically significant benefit of engineered heart muscle. Interestingly, ultrasound-based speckle tracking assessment of strain localized to the scar region did not show a benefit of engineered heart muscle. To resolve the apparent discrepancy between MRI-based and ultrasound-based strain results, the authors compared interand intraobserver variability of strain for the 2 modalities. In this regard, Bland—Altman analysis showed much larger limits of agreement for ultrasound-based strain compared with MRIbased strain. This analysis demonstrated that only the more accurate and reproducible MRI tagging method would be expected to detect the subtle benefits of engineered heart muscle with statistical significance using a sample size of 10 to 12 rats per group. An important lesson demonstrated by these data is that not all strain imaging methods are created equal—some are more accurate and reproducible than others and accordingly can support smaller sample sizes. This lesson may translate and indeed be amplified in clinical imaging, where high accuracy is needed when imaging is applied to individual patients (ie, a situation where the relevant group size has n=1). Although conventional myocardial tagging was used by Qin et al, newer MRI strain methods such as displacement encoding with stimulated echoes9,10 have been developed that maintain or even improve on the accuracy and reproducibility of tagging for strain quantification while requiring less manual intervention and maintaining high spatial resolution. Thus, accurate and reproducible high-resolution MRI strain imaging with rapid analysis has been demonstrated for both preclinical and clinical applications. (Circ Cardiovasc Imaging. 2016;9:e005687. DOI: 10.1161/CIRCIMAGING.116.005687.) © 2016 American Heart Association, Inc.