**3. Discussion**

The aim of this study was to compare cardiac alterations in a transgenic rat model for PE with the human post-PE situation. In order to describe cardiovascular changes after pregnancy in both species, STE was performed by blinded observers. The reported data reflect equivalent changes in the transgenic rat model and the human situation. Functional parameters such as EF and GLS were reduced in both species post-PE. We observed an increase in relative wall thickness and an increase in the posterior wall thickness of the LV as signs of structural remodeling. Melchiorre et al. were among the first who described permanent cardiovascular changes concerning relaxation abnormalities, diastolic dysfunction, and alterations according to LV geometry during and after PE in humans [4,18,19]. They evaluated LV dysfunction and geometry according to the European Association and American Society of Echocardiography guidelines, not mentioning the individual parameters that were altered. The novelty of our study is based on the very sensitive method of measuring myocardial function using STE. Former preeclamptic women often su ffer from asymptomatic cardiac abnormalities with prevalence of asymptomatic heart failure stage B including concentric remodeling and mildly impaired EF [6]. STE o ffers more sensitive parameters by describing the deformation of the myocardium and is even able to distinguish between individually contracting muscle layers of the heart. Previous publications demonstrated the extensive ability of STE to di fferentiate between physiological and pathological hypertrophic changes of the heart [20]. With the unique possibility of targeted myocardial specification, STE is being used more frequently in animal models [21–23]. The GLS, as the most sensitive parameter, was induced much earlier than EF [24] and showed a reduction of cardiac functionality post-PE in both species of our translational comparison. To our knowledge, this is the only animal model investigating cardiovascular changes postpartum by STE. An important open and unacknowledged question is whether the changes seen in humans are reversible [25] or permanent after delivery [26], and, if this is going to be an irreversible remodeling, whether replacement fibrosis is also present. We observed perivascular and interstitial fibrosis in the transgenic animal model [16], suggesting that a persisting structural remodeling post-PE and thus permanent cardiovascular abnormalities after preeclamptic pregnancy might be present after human preeclamptic pregnancy. Cardiovascular magnetic resonance (CMR) displays a high potential in the detection of myocardial fibrosis. One of our latest publications about CMR in formerly preeclamptic women describes the structural remodeling postpartum and underlines the burden of increased cardiovascular risk in later life. In this four year postpartum study, di ffuse injury of the myocardium was assessed by parametric mapping, focal injury by late gadolinium enhancement imaging, and cardiac function was evaluated by cine imaging and tissue tracking (strain). The post-preeclamptic group showed increased left-atrial end-diastolic volume stroke volume with a slight increase in left ventricular hypertrophy. We could not detect di fferences in focal or di ffuse myocardial tissue composition between the groups. Follow up studies are needed to identify the critical window when morphologic tissue di fferentiation, such as progression to myocardial fibrosis or inflammation, occurs in post-preeclamptic patients. Beside replacement fibrosis, another indication for permanent alterations in the maternal heart is the concentric remodeling. LV posterior wall and the extrapolated relative wall thickness increased in this study in both post-PE species. Intriguingly, the LV mass only increased significantly in the animal model. However, already published human studies show a permanent hypertrophy of the maternal heart [4,18,27].

The importance of PE as a risk factor in the development of chronic heart disease has been reinforced, and the future cardiovascular health of formerly preeclamptic women is getting more and more attention. Since 2011, the American Heart Association concludes that a pathological pregnancy is an important novel risk factor for later CVD [1]. While California was the only state in the USA to reduce maternal mortality due to hemorrhage using disease-specific toolkits and evidence-based protocol [28], death rates from maternal CVD are rising in the rest of the country. That demonstrates the urgen<sup>t</sup> need for a cardiovascular follow-up after preeclamptic pregnancy and a preventive strategy worldwide. New guidelines from the European Society of Cardiology for the managemen<sup>t</sup> of CVD during pregnancy [29] recommend lifestyle modifications and annual visits to a primary care physician to check blood pressure and metabolic factors. A su fficient strategy for cardiovascular prevention and postpartum treatment is still missing. To investigate the disease related remodeling process of the maternal heart, animal models are essential. The transgenic rat model mimics a preeclamptic pregnancy in several important aspects [11–15] and shows characteristic features of cardiac dysfunction after a preeclamptic pregnancy, as shown in this study.

In conclusion, our data indicate that the transgenic rat model and the use of STE depicts the cardiovascular abnormalities in the human situation well. That provides the basis investigating the causes of structural changes and functional disorders and gives the opportunity to test promising interventions. To underpin the methodical safety, we used one single observer for STE analysis in both species and demonstrated excellent intraobserver comparison. Interobserver agreements di ffered between EF and GLS and underlined the important aspect of fixed observers within a study. However, it should be mentioned that preeclampsia is not a monocausal disease, and there is no model that completely reflects the heterogeneous clinical picture of this disorder. Another limitation is the postpartum observation time point of four weeks in the animal model. Further studies with longer postpartum periods corresponding to 10–20 human years will be necessary in the future to be able to make further statements about the cardiovascular risk after a preeclamptic pregnancy.
