This study used CMR and brachial pressure for non-invasive PV-loops for detailed ventricular function assessment in patients with Fontan circulation and healthy controls. We found that patients with Fontan circulation had lower stroke work and ventricular mechanical efficiency compared to controls. Fontan patients with a dominant RV had higher potential energy indexed to body surface area and lower contractility than controls. The ratio of arterial elastance to ventricular elastance (Ea/Ees) was increased in patients with Fontan circulation independent of ventricular morphology indicating suboptimal ventricular-arterial coupling. Energy per ejected volume was similar to controls, indicating that patients with Fontan circulation has the same ventricular oxygen consumption of the myocardium to eject blood to the systemic circulation as in healthy controls.
Stroke work and contractility
Stroke work was lower in patients than in controls, especially in patients with dominant left ventricle and did not correlate with EF. End-systolic elastance as a measure of contractility did not have any correlation with EF in our study and was reduced in patients with dominant right ventricles. These results are in line with earlier studies using echocardiography [21]. Likely the decreased contractility, defined this way, is at least partly caused by low preload and does not necessarily imply decreased intrinsic capacity of the myocardium to contract, as shown by Wong et al. By using PV loops they found that patients with hypoplastic left heart syndrome and Fontan circulation could increase their contractility under stress.
Ventricular-arterial coupling
The ventricular-arterial coupling, measured by Ea/Ees, has been suggested as a prognostic marker for patients with Fontan circulation [22]. When Ea/Ees is around 1 the cardiovascular system is thought to be most efficient [18, 23]. The ratio Ea/Ees has been reported to be in the range of 0.6-1.2 in healthy controls and animal models [24, 25]. In our study the healthy controls had a mean ratio of 0.8. Fontan patients, however, had a mean ratio of 1.5. Similar difference between Fontan patients and controls has been shown by Godfrey et al [22]. This suggests suboptimal ventricular-arterial coupling in Fontan patients, similar to patients with systolic heart failure [18]. However, the reason for increased Ea/Ees ratio differed between patients with dominant RV who had decreased contractility and patients with dominant LV who had increased arterial elastance. Patients with acquired heart failure with reduced EF have reduced contractility, Ees, and increased Ea and thus a high Ea/Ees ratio, which has been shown to be strongly associated with adverse clinical outcomes [19, 26] and patients with increased Ea after Fontan operation has been shown to have worse prognosis [27, 28]. The non-invasive method used in this study may facilitate further studies to better understand the pathophysiology of the Fontan circulation and the possible impact of ventricular morphology.
Energy expenditure
Mean potential energy indexed to BSA was higher in patients with dominant RV compared to controls. Potential energy is energy that at the end of systole will be converted into heat, thus wasted energy. Stroke work and potential energy adds up to the PV-loop area which has shown strong correlation with left ventricular oxygen consumption [29]. Thus, dividing the PV-loop area with SV results in how much mechanical energy is used to eject the SV and provides an estimation of the ventricular oxygen consumption per heartbeat. Interestingly, patients with Fontan circulation utilize same amount of energy per ejected volume as controls at rest. This might be explained by the low blood pressure in the patient group. Patients with acquired heart failure and low EF also have decreased contractility at rest, but to a higher degree than patients with Fontan circulation [11], but in contrast to Fontan patients, they use increased amount of energy per ejected volume at rest, with high potential energy where much energy is converted to heat instead of producing work.
Ventricular efficiency was decreased in patients with Fontan circulation in comparison with healthy controls. Ventricular efficiency correlated strongly with EF as shown earlier for patients with heart failure and healthy volunteers [11]. Studies of Fontan patients using echocardiography for volumetric assessment has also shown this relationship [21], however not as strong as in this study, which might be explained by more exact volume assessment with CMR.
Pathophysiological significance
The finding that patients with Fontan circulation has the same ventricular oxygen consumption of the myocardium to eject blood to the systemic circulation as in healthy controls is in contrast to patients with acquired heart failure and reduced EF where we have demonstrated increased energy per ejected volume and thus higher ventricular oxygen consumption [11]. This shows how PV loops can show differences in the pathophysiological mechanisms between patient groups. Moreover, stroke work and contractility showed no correlation with EF in patients, and this implies that the physiological parameters from non-invasive PV-loops may give further information in these patients compared to more commonly used clinical parameters. These measures from non-invasive PV loops may help understand heart failure and guide treatment in patients with Fontan circulation.
Clinical relevance
PV loops acquired non-invasively from CMR offers in depth physiological information of ventricular function that may detect subclinical changes in ventricular function over time. The benefits compared to invasively acquired PV-loops are that there is no need of radiation or general anesthesia, and the elimination of the risks associated with LV catheterizations.
Limitations
A limitation to the method is that it requires an unobstructed communication between the systemic ventricle and the brachial artery. The patients included in this study did however not have any signs of stenosis why this aspect is judged to have limited influence on the result. The method approximates V0 to be zero, although it should probably be a small but positive value. Validations of this method have, however, shown good agreement between in vivo measurements and model-derived parameters and thus the approximation of V0 is probably reasonable [11].
The non-invasive PV loop uses an estimation of the ventricular end-diastolic pressure. Within a range between 0 and 15 mmHg previous studies showed low influence on the hemodynamic parameters [11]. In the study cohort one patient had increased end-diastolic pressure to 22 mmHg during a catheterization done in near time to the MR scan which might have influenced of the stroke work and therefrom derived values. The results of a reanalysis of the examination with increased end-diastolic pressure showed only small changes (<10%) to the parameters. This shows that even if end-diastolic pressure were elevated in more patients the changes in the results would not be large enough to influence the difference between the Fontan patients and controls. The reanalysis also indicates that changes in end-diastolic pressure seems to have little effect on contractility, which is important when using this measure for assessing ventricular function.
The number of patients in the study is small, which should be taken into consideration. Also, a minority of the patients (29%) were under general anesthesia which might influence the result, but due to the small sample size statistical analysis between these patients and those not under general anesthesia is difficult. Due to ethical considerations, it was not possible to have anesthetized children as controls.