Deformation imaging to assess global and regional effects of cardiac regenerative therapy in ischaemic heart disease: A systematic review

1 INTRODUCTION

Ischaemic heart disease concerns an important health issue in modern-day medicine. Reperfusion of ischaemic myocardium by catheter-based and surgical treatment has proven to be successful in both treatment of acute myocardial infarction (AMI) in order to limit infarct size and chronic ischaemic cardiomyopathy (ICMP) in order to protect the heart from further deterioration (Windecker et al., 2014). Because the regenerative capacity of the human heart is intrinsically limited, ischaemic left ventricular (LV) dysfunction currently is a chronic condition. The ability to regenerate infarcted myocardium can be seen as the holy grail within cardiovascular research. Since 2002, when the first clinical trial (Strauer et al., 2002) on the effect of cell therapy on the failing heart was published, clinical research makes effort into regenerating the damaged myocardium (Fernández-Avilés et al., 2017). Unfortunately, myocardial regeneration has not found its way into the routine clinical setting. However, transfer of progenitor cells into infarcted myocardium has shown to sort a positive, albeit modest, effect on cardiac function (Afzal et al., 2015). The most commonly used functional outcome measure in cardiac regenerative research is the LV ejection fraction (LVEF) as measured by echo or magnetic resonance (MR) imaging, which is easy to perform and effective for measuring global cardiac function changes. In the clinical setting, LVEF has been shown to be a powerful predictor of outcome (Møller et al., 2006; Pocock et al., 2006; Solomon et al., 2005). Notwithstanding these advantages, the technique is hampered by high interoperator and intraoperator variability and is dependent on loading conditions (Dorosz, Lezotte, Weitzenkamp, Allen, & Salcedo, 2012; Wood, Choy, Nanda, & Becher, 2014). Furthermore, LVEF measurement may be less suitable to detect regional function changes potentially sorted by cardiac regenerative therapy. Keeping the high operator variability in mind, LVEF is arguably a suboptimal endpoint.

Cardiac deformation, expressed as strain (ε), is defined as the magnitude of deformation of a segment of myocardium during a cardiac cycle relative to the original length of this segment and is expressed as a percentage. It can be measured using tissue Doppler echocardiography (TDI), two-dimensional speckle tracking echocardiography (2DSTE), and MR tagging and tracking. Deformation is measured in either direction in which the myocardium deforms: longitudinal (shortening), circumferential (shortening), and radial (thickening; Teske et al., 2007). Peak systolic strain correlates well with LVEF but, unlike LVEF, is more likely to detect regional attenuations of cardiac function (Lima et al., 2017; Weidemann et al., 2002). Deformation imaging is a noninvasive, quantitative, largely operator independent, and therefore reproducible and objective means of measuring both global and regional cardiac function, which at least in the longitudinal direction seems to outperform standard LVEF measurement regarding feasibility, accuracy, and test–retest and interoperator variability (Barbier et al., 2015; Mirea et al., 2018). It is therefore a promising new technique, especially in the field of cardiac regenerative research. However, the additional value of this technique compared with current endpoint measurement within this field is currently inadequately understood.

We hypothesize that deformation imaging will be effective in detecting the potential effects in regional cardiac function sorted by cardiac stem cell therapy. Hence, efficacy of cell therapy can be assessed more precisely. The aim of this systematic review is to identify the added value of strain analysis over standard echocardiography in cardiac regenerative therapy for ischaemic heart disease by providing a comprehensive overview of current literature on cardiac stem cell therapy in which deformation endpoints were used as outcome measures.

2 METHODS

3 RESULTS

3.1 Search results

The final search (Figure 1) was performed on December 13, 2018. A total of 126 papers were identified from PubMed database, 235 papers were identified from Embase database, and 26 papers were identified from Cochrane database. After removal of 154 duplicates, 233 titles and abstracts were screened. After exclusion of 163 papers based on title and abstract, 70 papers were selected for full-text screening. An additional 45 papers were excluded, after which five papers were included after checking cross-references. Finally, a total of 30 (Amado et al., 2006; Beitnes et al., 2011; Bhatti et al., 2013; Bonios et al., 2011; Cai et al., 2015; Chen et al., 2015; Duran et al., 2013; Heldman et al., 2014; Herbots et al., 2009; Hopp et al., 2011; Jaussaud et al., 2013; Karantalis et al., 2015; Karatasakis et al., 2010; Lebrun et al., 2006; Malliaras et al., 2014; Miao et al., 2014; Nasseri et al., 2009; Nasseri et al., 2014; Plewka et al., 2009; Qi et al., 2015; Quevedo et al., 2009; Rickers et al., 2004; Schneider et al., 2008; Schuleri et al., 2008; Schuleri et al., 2009; Sun et al., 2016; Van Ramshorst et al., 2009; Van Ramshorst et al., 2011; Williams et al., 2011; Yamada et al., 2013) papers were included, of which 15 papers were based on preclinical research and 15 papers were based on clinical research.

4 DISCUSSION

One of the main conclusions that can be drawn from this review is that the use of deformation outcome measures within cardiac cell therapy research is very heterogeneous. The included studies varied substantially in their choice of patient setting or animal model, type of cardiac stem cell, and method of administration. To add to the heterogeneity of our results, the method of deformation measurement (2DSTE, TDI, or MR feature tracking or tagging) and the outcome measure of choice (global or regional strain in the longitudinal, radial, or circumferential direction) varied distinctively between the different included studies. Comparing results of different studies or performing meta-analyses was therefore not feasible.

Notwithstanding the heterogeneity in deformation imaging within cardiac regenerative medicine, this article provides an insight into the value of strain analysis within this field of research. The results show that in 13 out of 15 preclinical papers and in nine out of 15 clinical papers deformation imaging results followed the results of LVEF outcome measurement, either both showing a significant effect of cardiac cell therapy on LV function or both showing a nonsignificant effect of cardiac cell therapy on LV function. There are no papers that reported a significant effect of cell therapy on LVEF measurement without deformation outcomes doing so. Importantly, and fitting with our main hypothesis, two out of 15 preclinical papers and seven out of 15 clinical papers showed a significant effect of cell therapy on cardiac deformation, whereas LVEF was not improved significantly in these studies. These results advocate a potential added value of deformation imaging within cardiac regenerative medicine for ischaemic heart disease.

Although experience with cardiac stem cell therapy shows a positive effect on cardiac function, underlying mechanisms are not fully elucidated. It is believed to be based on paracrine effects, leading to (a) neovascularization into regions affected by ischaemia (Schuleri et al., 2008), (b) modulation of inflammatory effects (Luger et al., 2017; Van Den Akker, Deddens, Doevendans, & Sluijter, 2013), (c) modulation of ventricular remodelling and extracellular matrix homeostasis (Laflamme, Zbinden, Epstein, & Murry, 2007), and (d) enhancement of cardiac repair by host resident cardiac progenitors (Etzion et al., 2001). Currently, effort is put into optimizing the effect of cardiac regenerative therapy. Indeed, this is reflected by the large heterogeneity in research strategies of the studies included in this article. In addition to these necessary efforts for improving cardiac cell therapy, however, we suggest assessing new outcome measures. This is because effects of cell therapy on global cardiac function as measured by LVEF are limited. In a meta-analysis assessing the effect of adult bone marrow cardiac cell therapy (Afzal et al., 2015), the authors presented a beneficial effect of bone marrow-derived cell therapy on LVEF of merely 2.92%. Of importance, the results of papers included in our systematic review are of similar magnitude. Keeping in mind the high interoperator and intraoperator variability and loading dependency of LVEF measurement (Dorosz et al., 2012; Wood et al., 2014), leading to variations of up to 18% for two-dimensional LVEF measurement by Simpson's rule (Jensen-Urstad et al., 1998), LVEF is in our opinion not suitable for endpoint measurement in cardiac regenerative medicine in individual cases.

Deformation imaging has already proven its value in, for example, hypertrophic cardiomyopathy (Fuster, Van Der Zee, & Miller, 2009) and valvular heart disease (Casas-Rojo et al., 2016; Gunjan, Kurien, & Tyagi, 2012), in which deterioration of deformation parameters can be observed before LVEF decreases. This apparent superior sensitivity for changes in cardiac function makes deformation imaging of great potential use within cardiac regenerative medicine. Besides being less operator dependent compared with LVEF, strain analysis allows for assessment of regional cardiac function, making this technique especially suitable for therapies in which a regional effect is expected to be sorted. Regional strain is indeed analysed in all but two included studies in this article, of which seven studies suggest an improved sensitivity to changes in LV function compared with LVEF measurement. There are however no studies reporting on significantly increased global strain, whereas LVEF was not altered. Furthermore, from the results, it is not clear which type of regional strain parameter is best used. As an additional advantage, 2DSTE seems more accurate, also in direct comparison with LVEF measurement (Barbier et al., 2015; Mirea et al., 2018).

The effect of stem cell therapy varies strikingly between preclinical and clinical research: Preclinical papers show a positive effect on deformation parameters more often than clinical studies do. This “translational failure” is a known phenomenon, due to which preclinical results may not always be applicable to the clinical setting. A possible explanation can be found in the risk of bias analysis, showing a higher risk for performance bias, detection bias, and reporting bias in preclinical studies. Also, study models might be more heterogeneous in the patient setting.

5 LIMITATIONS

This aim of this study was to investigate whether deformation imaging could provide added value over conventional endpoint measurement in stem cell research. Although evidence is not scarce, and data indeed do suggest added value of deformation imaging, articles that were related to in this study were too heterogeneous in their study design for pooled analysis. Therefore, we could not provide a statistical analysis on deformation imaging as a diagnostic tool over conventional functional imaging. In addition, deformation imaging was not compared with other methods of endpoint measurement, such as myocardial perfusion imaging, so the value compared with these remains to be elucidated. Moreover, the relevance of so-called surrogate endpoints in stem cell research needs to be addressed. We aimed to investigate deformation imaging as a new surrogate endpoint for stem cell research. However, regardless of the probable additional value of deformation imaging over conventional endpoint measurement, we are unsure how this relates to clinical endpoints such as mortality, hospital admissions, quality of life, or recovery from myocardial infarction. This remains to be studied. To conclude, the conventional echocardiography endpoint used in the included papers was two-dimensional LVEF, of which diagnostic accuracy is inferior compared with three-dimensional assessment. Therefore, results might not be applicable to the current situation in which three-dimensional LVEF assessment is common practice.

6 FUTURE PERSPECTIVE

In order to be able to compare deformation outcomes within cardiac regenerative medicine, the methods of performing this should be standardized. We propose to use the “Definitions for a common standard for 2D speckle tracking echocardiography: consensus document of the EACVI/ASE/Industry Task Force to standardize deformation imaging” (Voigt et al., 2015) as a guide for standardized deformation imaging within cardiac regenerative research. This provides guidance into both image acquisition and image processing. Preferably, deformation should be measured using 2DSTE or MR imaging tagging or feature tracking; TDI-derived strain is arguably too labourious and complicated for common clinical use. Future efforts of this task force might further enhance diagnostic performance of 2DSTE and allow for more interstudy comparison. Future research might provide insight into which the deformation parameters are most sensitive to LV functional change after regenerative therapy. Also, newer strategies such as three-dimensional speckle tracking echocardiography might provide additional diagnostic accuracy.

7 CONCLUSION

Deformation imaging shows to be a promising new endpoint measure in cardiac stem cell research, which is at least equally, and possibly more, effective than LVEF assessment in the detection of changes in cardiac function and may provide research with a more objective measure of response to therapy. However, current use is diverse and needs standardization in order to compare results between studies.

CONFLICT OF INTEREST

None declared.