136 | Chapter 7 the various genetic cardiomyopathies, possibly reflecting early cardiomyocyte loss in DCM, myocardial disarray and interstitial fibrosis in HCM, and desmosomal dysfunction in ACM. Multiple deformation parameters are available, but the main findings reproduced by the different studies are: (1) reduced GLS in relatives at risk for DCM, (2) reduced basal (antero) septal strain in relatives at risk for HCM and (3) reduced RV free wall strain in relatives at risk for ACM (Central Illustration). With regard to risk stratification, the prognostic value of deformation imaging has most extensively been investigated in DCM and ACM, showing a benign prognosis when deformation abnormalities are absent. Prognostic data for deformation imaging in HCM relatives are limited, making its role in clinical risk stratification uncertain. Since life-threatening arrhythmias rarely occur in early stages of HCM58, follow-up of septal thickness may be sufficient for arrhythmic risk stratification. At this point, the presence or absence of deformation abnormalities may be helpful for determining follow-up strategies in DCM and ACM relatives. Considering the high negative prognostic value of deformation imaging in the available studies, relatives who have normal findings by deformation imaging may be offered lower follow-up intensity than relatives who have subclinical deformation abnormalities. The exact follow-up strategies, including the required intervals, remain to be investigated in longer longitudinal studies. In addition, more studies in younger subjects are needed to gain further knowledge about the penetrance in the pediatric population and the required age to start screening, since our systematic search only yielded one pediatric study. Identification of an early disease substrate has the most potential to lead to therapeutic strategies in relatives at risk for DCM or ACM. Thereby, these relatives can potentially be offered early heart failure medication to prevent end-stage heart failure, or antiarrhythmic medication and implantable cardioverter defibrillator implantation to prevent sudden cardiac death.35,46 However, since the numbers of hard end-points are low in the published studies, the prognostic value of deformation imaging should be confirmed in larger studies, preferably with longer follow-up intervals. Deformation imaging may also soon become relevant for patient selection in HCM, given the promising trials on medical treatment of early stage HCM.59,60 We encourage the authors of reported studies to investigate and publish the long term ( >10 years) outcomes of their cohorts whenever possible. Moreover, we strongly encourage collaborations between different research groups to create larger cohorts and to create a platform for external validation. We would like to emphasize that deformation imaging should not be used as stand-alone parameter, but interpreted in the context of other clinical variables, and in conjunction with other examinations such as electrocardiography and CMR. Studies implementing relevant deformation imaging parameters into multi-modality risk prediction models are therefore of great interest. Finally, we would like to encourage the use of machine learning approaches to improve the classification of deformation curves of relatives.61 Machine learning models may detect hidden patterns in deformation curves and improve the classification of these curves, potentially leading to earlier identification of high-risk relatives. Besides using machine learning merely for classification purposes, specific techniques can be applied to visualize the features that are detected by the machine learning model.62 The application of such techniques will enrich our knowledge of hidden features in the deformation curves of relatives, which will enhance the clinical utility of deformation imaging in this population.
RkJQdWJsaXNoZXIy MTk4NDMw