The present study reports the differing impact of PS/PR induction in a porcine model based on state-of-the-art CMR imaging. PS but not PR resulted in deteriorated right cardiac function. First, whilst both increased afterload (PS) and preload (PR) resulted in increased RV volumes, RV deformation deteriorated following prompt increase in afterload caused by PS but improved following increased preload caused by PR. Secondly, volume overload and congestion following RV functional deterioration resulted in consecutive RA functional impairment in PS but not PR. Lastly, LV deformation significantly improved in PS but not PR potentially compensating for impaired LV preload to maintain LV SV.

Differing effect of pre- and afterload on RV deformation parameters

Myocardial deformation as appreciated from strain is load dependent. As demonstrated by the Frank-Starling mechanism for the LV, both increased preload and afterload lead to volume overload and subsequently increased deformation. Indeed, both PS and PR resulted in increased RV dilatation with subsequently higher SV. However, there were significant differences for PS and PR in RV strain. An increase in RV afterload in PS has been related to deteriorated RV strains, at least on a segmental level [35]. This has also been demonstrated in a population of HFpEF patients, where patients with raised pulmonary vascular resistance exhibited impaired longitudinal RV strain [36]. In line, we demonstrated significantly decreased RV GLS SR and a numerical but not significant decrease of overall RV GLS. However, increased RV SV has previously been attributed to increased RV end-systolic elastance in PS [24]. One the one hand side this seemingly contraintuitive finding may thus correspond to an increase in circumferential/radial RV contractility. On the other hand side overall dilatation with increased SV but not EF may merely represent limited forms of compensation and not overall improved contractility. Intriguingly, sole increase in preload in PR without the aspect of additionally increased afterload from PS led to a statistically significant increase in RV GLS. This can either be interpreted as (a) a consequence of the Frank-Starling mechanism or (b) as an increase in intrinsic contractility. Of note, RV GLS SR was largely unchanged in PR pigs compared to pre-intervention. Strain rate is generally considered to be a marker which more closely reflects contractile function and is less load-dependent than strain [37]. On balance, this suggests the Frank-Starling mechanism as the driving force behind the effect, as was discussed for mitral and aortic regurgitation previously [38]. Nonetheless, these results underline that the RV can handle volume overload better than pressure overload, at least in the timeframe studied here.

Development of PS increases RV pressure in a chronic manner, resulting in RV morphological and structural adaptations [39]. The effect of increased pressure load on myocardial tissue is a double-edged sword. On the one hand side concentric hypertrophy helps to overcome increased afterload, on the other hand side tissue remodelling includes fibrosis with subsequently reduced ventricular compliance [40]. Initially, as shown for the left ventricle, concentric hypertrophy enables the ventricle to maintain wall stress at lower levels throughout the cardiac cycle despite increases in pressure, and therefore was historically regarded as favourable adaptation [7, 41]. However, maladaptation to the increased afterload may ensue in the form of eccentric hypertrophy or dilatation, resulting in ventricular failure [40], calling into question the positive view of concentric hypertrophy. Adding to this, adverse histological and functional findings as well as reduced survival rates were noted in a murine model of pressure compared to volume overload [9]. In addition, an abruptly and severely increased pressure load can cause RV decompensation with life-threatening consequences. This can be the case with acute pulmonary embolism since the RV is not able to respond to the quick rise in afterload with remodelling [42], which might be best suited for comparison with acute induction of PS by surgery.

Unlike mild volume overload, which can be handled by dynamic modulation of SV [43], continuous severe volume overload may induce myocardial remodelling with subsequent heart failure [44]. In contrast to afterload-induced remodelling with thickening of the ventricle, preload mainly results in a dilated, thin-walled ventricle [45, 46]. However, the literature for the LV suggests that these effects only develop over time, with a proposed timeframe of up to 21 weeks of severe volume overload until systolic failure occurs [45]. Despite this potential danger of heart failure in volume overload in the long run, our study showed that at least in a timeframe of 10 to 12 weeks, the RV can better cope with volume overload than with pressure overload. This is in line with earlier findings based on four dimensional phase contrast magnetic resonance [25].

Right atrial involvement in PS and PR

Over the course of RV hypertrophy development, RV compliance may diminish, translating to hindered diastolic function, as previously shown in clinical and experimental settings of chronic RV pressure overload [47, 48]. Accordingly, we observed a statistically significant decrease of total RA strain, paralleled by a reduction in passive RA conduit strain of almost the same magnitude in the PS model. As conduit strain reflects passive ventricular filling [49], this finding is indicative of increased RV stiffness as an adverse by-product of hypertrophy [48]. Furthermore, RA functional failure in PS may indicate overall congestion introduced by RV functional failure in PS as opposed to PR. In PR, a previous study found an increased RV compliance based on invasive pressure-volume loops [24], translating to improved RV diastolic function. In the present PR population, there was no distinct impact on RA physiology except for a significant reduction for RA Ea strain rate and a numerical statistically non-significant trend for a decline in active RA Ea. This could imply that their RV filling is less dependent on active atrial contraction due to improved passive filling. Altogether, these RA results are in line with our findings for the RV, as they indicate significant right cardiac failure in PS corresponding to a more maladaptive response to pressure than volume overload.

LV compensation of RV dysfunction

Pigs from both PR and PS interventions showed increased LV mass and EDV. With ESV remaining similar, this resulted in increased SV and EF. However, only pigs following PS induction displayed a significant increase in LV GCS compared to pre-intervention as well as post-intervention PR pigs. First, this may suggest a compensatory effect of the LV to uphold cardiac output despite reduced preload caused by RV functional failure in PS. Second, this hint at circumferential compensation echoes findings from patients with pulmonary hypertension: in spite of diminished longitudinal LV function, their overall systolic LV function stayed the same, owing to an increase in lateral movement [50]. From a mechanistic standpoint, the LV and RV are intricately linked on several layers, such as their shared septum and common myocardial fibres. This allows for functional interplay [51,52,53,54]. Estimates for contribution of LV contraction to RV systolic function range as high as > 50% [54]. Consequently, improved LV contractility as appreciated from increased LV GCS may explain the previously reported increases of RV end-systolic elastance as an indication of improved contractility in PS [24], despite a tendency for RV GLS to deteriorate. Accordingly, in the previously published analysis of pigs studied here, LV Ees was numerically increased in PS, albeit not reaching statistical significance [24].

Advantages of the employed animal model

Using large animal models helps to close the gap between small animal laboratory studies and clinical trials. In a cellular and mouse model adverse effects of afterload as opposed to preload on left ventricular remodelling and function have been demonstrated [9]. These findings are expanded to RV physiology incorporating a model more closely resembling human heart physiology [56]. In addition, the advantages of using large animal models for cardiovascular disease exploration lies in their body size allowing for easier surgical manipulation and more detailed pathophysiological assessment in the setting of imaging studies [57]. The utilized animal model also allowed for the induction of specific changes in pre- and afterload, respectively, which enabled us to observe alterations of myocardial deformation over time attributed to the artificially created hemodynamic differences. During the original experiments, echocardiographic control was established during the process of model creation to achieve maximum pulmonary artery stenosis possible [25] Presence of pulmonary regurgitation was also confirmed using echo. Therefore, loading conditions (increase of pre-/afterload) in the respective groups at time of model creation were experimentally ensured.

Furthermore, the use of CMR-FT for deformation analysis in landrace pigs has already been demonstrated to show low intra- and interobserver variability for the assessment of LV GLS and GCS [58]. CMR FT technology is already available for patient care and needs no specific transfer from the animal model to the human.

Outlook for further research and possible clinical applications

Our findings underline the potential utility of CMR-FT in a porcine model to track changes caused by PS and PR induction. Based on this, the model might be used in the future to assess the long-term effect of PS vs. PR as well as targeted therapeutic interventions respectively, as part of a longitudinal study. In line, assessment of the clinical outcome within the porcine model would be of interest to evaluate the possible prognostic implications.

Our results show the capability of deformation analyses to detect changes in all four chambers in PS and PR models. Indeed, current imaging guidelines relating to repaired TOF patients hint at the option of ventricular assessment via deformation analysis, whilst to date volumetric analyses remain the reference standard [59]. Based on our results, the assessment of deformation imaging should be evaluated further in the future, as they offer additional pathophysiological information over established imaging parameters. If proven useful in valvular disease, it could facilitate early detection of ventricular dysfunction to monitor patients and potentially guide early treatment decisions. Indeed, strain analysis is already a relevant research focus in the realm of aortic stenosis. For instance, published literature shows the value of CMR-FT for identification of LV dysfunction, risk stratification and assessment of treatment effect after transcatheter valve implantation [60].

Limitations

Our analysis offers insights into the pathophysiology of volume vs. pressure overload in the right ventricle. Due to this study’s nature as a subanalysis, its limitations overlap with those of previous analyses based on the same study [24, 25]: The small population of pigs likely renders it underpowered for the identification of subtle, within and between-group differences. However, the fact that in several key parameters, significant differences were detected despite the small sample size, underscores the robustness of CMR-FT to pick up change in myocardial function. Also, as the two CMR scans were conducted at the relatively fixed interval of 10 to 12 weeks, any acute or subacute changes in between or long-term effects potentially occurring after the study could not be assessed using this design. Additionally, due to the primary study design, clinical outcome of pigs was not recorded in detail [25].

Furthermore, as can be appreciated from Table 1, there were striking differences in median values in several key parameters before surgical interventions, although these differences did not reach statistical significance. As shown in Table 5, after accounting for these non-significant differences in baseline values, there was still a significant association of the type of intervention (PS/PR) with most of the parameters studied. This underlines the strong effect of the surgical procedure on clinical and imaging parameters studied.

The most obvious clinical example to which our animal study might be applied is TOF patients before (PS group) and after repair, once regurgitation has set in (PR group). This translation however is somewhat limited, as repaired TOF patients experience the cumulative effect of already present pressure overload (from their baseline disease) and volume overload (from newly developed pulmonary regurgitation). In contrast, in our study, the pigs undergoing artificial creation of PR were healthy before the intervention. As such, a previously published experimental design focused on the induction of PR in an already pressure-overloaded RV [61].

Lastly, circumferential or radial motion might compensate for EF in RV physiology in case of impaired longitudinal contraction. However, we did neither analyze GCS nor GRS for the RV and as such cannot assess this potential compensation mechanism.