This study shows that PA size in growing Fontan patients is comparable to healthy controls and increases with age, appropriate to body size. Flow in the LPA increased significantly between both visits, but was still lower than in healthy controls. Pulsatility and distensibility were impaired and did not change over time, despite increase in flow. WSS was not significantly different between baseline and follow-up. Multivariable regression analysis revealed that age, mean flow and mean area are important predictors for WSS that explain most of the variability.
PA Size
Several studies have assessed the range of branch PA size in healthy children and the relation with age and/or body size. Using contrast-enhanced MRI, Knobel and colleagues have shown that there is a linear relation between BSA and PA size [
20]. Using echocardiography, an older study showed a similar trend in younger children [
28]. Our results indicate that LPA growth is proportionate to BSA in Fontan patients.
Only a few studies have investigated pulmonary artery size in Fontan patients using MRI. LPA size in our patients was comparable to that found in a previous smaller sample from our institutions [
26]. Bellsham-Revell and colleagues investigated growth of the LPA between the hemi-Fontan stage and the TCPC in patients with hypoplastic left heart syndrome (HLHS) [
4]. The (BSA-corrected) size of the LPA pre-TCPC in their study was comparable to the values that we have presented. In the study by Bellsham-Revell, it was found that, while there was an increase in area of the proximal LPA between stages, there was a relative decrease in size of the narrowest part of the LPA. The authors stated that the staged Fontan approach might be a risk factor itself for the narrowing of the LPA in these HLHS patients [
4]. Earlier studies have shown that LPA stenosis is a well-known problem, particularly in HLHS patients [
3]. In our population, there were only a few patients with HLHS. LPA areas in these patients (range 50–102 mm
2/m
2) were among the lowest sizes measured in our study. The previous studies and our results emphasize the need for close surveillance in this specific group.
Other studies have used X-ray catheter-assisted angiography to measure PA size, usually using the Nakata index [
16,
22]. Although this makes comparison challenging, a previous study in patients with congenital heart disease has shown excellent agreement between angiography and MRI measurements of the great arteries, including branch pulmonary arteries [
34].
In a recent study, Schmitt et al. [
29] found a decreased Nakata index of 150 mm
2/m
2 in 10 Fontan patients. Converting this number into a single PA branch area as we have measured would result in a PA size of 75 mm
2/m
2, which is lower than we found. This difference might be explained by the fact that patients in the study by Schmitt were older and all of their patients had an unfenestrated extracardiac conduit Fontan.
Few studies have assessed PA size longitudinally after Fontan completion. One study showed that Nakata index decreased after Fontan completion, but did not affect functional outcome [
1]. Another retrospective study also showed that PA growth failed to match the increase in BSA in Fontan patients [
32]. A recent study by Restrepo and colleagues, using MRI-derived 3D reconstructions of 25 TCPC patients, showed a similar trend. While there was an increase in absolute vessel diameters, normalized diameters decreased significantly with age [
24].
In a larger study, the same group confirmed their findings of PA size change during serial follow-up [
25]. Furthermore, mean flow rates remained unchanged. Using their extensive experience in this field, the authors showed that indexed power loss in the Fontan pathway increased significantly in the cohort, particularly in patients in whom the minimum normalized left pulmonary artery decreased [
25].
The finding on PA size in the recent Restrepo paper is in contrast to our measurements [
25]. Possible explanations include different methods for quantification of PA size. In our study, we have used 2D MRI measurement performed in a single plane, while calculations performed by Restrepo et al. were based on a 3D reconstruction of the entire Fontan pathway. Another possible explanation is the relatively high percentage of patients with HLHS in their cohort.
PA Function
Interestingly, there was a significant increase in flow normalized to BSA in our patients between baseline and follow-up. This is in contrast to a recent MRI study among 48 TCPC patients with follow-up duration comparable to our study, showing an increase in absolute flow over time, but no changes in flow corrected for BSA19. In the aforementioned study, it was shown that indexed power loss inside the Fontan pathway increases, despite BSA-corrected flow remaining constant [
25]. It remains unclear whether and how adjustment of the pulmonary circulation to the Fontan circulation occurs. WSS may be a crucial factor in that process. WSS is known to be lower in Fontan patients as compared to healthy controls. This is clear from the mean WSS, but even more so for the maximal WSS during the cardiac cycle. This is in accordance with an earlier study [
26]. Impairment of WSS in Fontan patients most likely results from a combination of reduced pulmonary flow and nearly absent pulsatility [
26].
Although not significant, there was a trend toward a further decrease of WSS with longer follow-up. Plotting WSS against age showed a strong decrease of WSS at early teenage age with a stabilization after the age of 15 years. In multivariable regression analysis, (the logarithm of) age was an important predictor of WSS, independent from both body size and PA area.
Loss of pulsatility of blood flow and therefore lower mean and maximal WSS values have been associated with endothelial dysfunction of the PA’s [
19,
21].
In a study including 10 young patients reduced pulsatility after bidirectional Glenn correlated to impaired endothelial relaxation [
21]. Another study has demonstrated abnormal response of the PA’s to exogenous nitrous oxide (NO) in Fontan patients [
18]. Supplemental NO led to a fall in PVR, suggesting an elevated basal PVR, possibly related to endothelial dysfunction [
18].
A further decrease of the WSS with age could also result in deterioration of endothelial function, mediated by endothelin, a potent vasoconstrictor. A study comparing endothelin receptor expression in failed and non-failed Fontan patients has shown an overexpression of these receptors in the failed Fontan group [
15].
WSS is also decreased in patients with pulmonary arterial hypertension (PAH). In contrast to Fontan patients, PAH patients have dilated pulmonary arteries, due to a longstanding elevated mean PA pressure [
33]. Studies have shown a negative correlation between vessel size and WSS. It has been hypothesized that the decreased WSS in PAH patients leads to an increased arterial stiffness and reduced distensibility of the PA’s [
31,
33].
More direct evidence that the reduced WSS not only influences the function of the PA’s but also has effect on the structure of the vessel wall comes from a case report of a 35-year-old Fontan patient (APC). Immunohistological analysis revealed serious changes in the composition of the (main) pulmonary artery wall [
2]. There was a profound reduction of muscular component and fragmentation of elastic fibers, which might influence distensibility and the vasodilatory ability. It is likely that this is also true for younger Fontan patients, operated upon according to current techniques, but this should be further investigated.
Clinical Implications
It has been demonstrated that exercise capacity is reduced in Fontan patients and that it reduces further with age [
5]. In the current study, we have shown that WSS also decreases with age in Fontan patients. The decrease in WSS might influence endothelial function in the PA’s. In healthy subjects, there is an increase of distensibility during exercise with an increase in the release of NO [
12]. This induces vasodilatation and enhances pulmonary blood flow by a decrease in pulmonary vascular resistance. In Fontan patients, there often is an increase in pulmonary vascular resistance during exercise, indicating that this mechanism is impaired. This contributes to impaired ventricular filling [
10]. A previous study from our institution showed that Fontan patients are not able to increase stroke volume during exercise [
27]. Other studies have shown similar results [
35]. The abnormal function of the PA’s thus has direct consequences for exercise function of the patients, and may contribute to decline of exercise capacity.
Considering these observations, it is of utmost importance to be able to influence pulmonary vascular resistance in the Fontan circulation. Several studies using bosentan, an endothelin receptor antagonist, have not shown significant improvement [
23,
30]. Another study using sildenafil, a phosphodiesterase-5 inhibitor, to assess the influence during exercise in Fontan patients, showed an increase in stroke volume and cardiac index and a decrease in PVR during exercise [
35]. Exercise capacity improved after sildenafil administration, but mainly in those patients with a poor baseline exercise capacity. This indicates that the reduced endothelial function could be attenuated to affect exercise capacity. In another study, sildenafil was administered for a period of 6 weeks, but failed to show a significant improvement in exercise capacity [
11]. It has been speculated that this result was caused by the fact that relatively fit Fontan patients were included [
14]. Further studies are necessary to identify those patients that benefit the most from this potential therapy or to uncover other potential targets and means for possible therapeutical intervention.
Limitations
Sample size was relatively small and the Fontan population is heterogeneous with respect to different cardiac diagnoses. This may be one of the reasons for differences between the outcome of this and other recent studies [
25]. The subgroups were too small for group-to-group comparison.
This study assumes laminary flow in the PA’s for the calculation of WSS. Although care was taken not to measure flow too close to the caval connection point, flow disturbances may have been present, depending on the individual anatomy and intravascular flow pattern in patients.
Since measurements were performed in one of the branch pulmonary arteries, this study does not provide direct knowledge on the smaller pulmonary vasculature further downstream.