Total cavopulmonary connection (TCPC) is the final palliative surgical procedure for children with single ventricle physiology. Adequate pulmonary blood flow in these patients is dependent on a low pulmonary vascular resistance. High resistance may hinder surgical progression to total cavopulmonary connection and is known to increase short-term mortality after this surgery [1‐3].

Clinical signs and secondary effects of elevated pulmonary vascular resistance include cyanosis, reduced exercise tolerance, protein-losing enteropathy, and plastic bronchitis. Mechanisms such as passive non-pulsatile flow and cyanosis/hypoxemia caused by systemic-to-pulmonary venous collaterals have been suggested to contribute to vascular remodeling and increased resistance over time in a patient with single ventricle physiology [4‐6].

In children with pulmonary arterial hypertension, pulmonary vasodilator therapy has been shown to reduce pulmonary vascular resistance and alleviate symptoms [7‐9]. In clinical guidelines, children with single ventricle physiology are identified as a group who can suffer from clinically significant pulmonary hypertensive vascular disease even when their mean pulmonary arterial pressure is below 20 mmHg, the threshold of pulmonary hypertension [10‐13]. The recently revised pediatric pulmonary hypertension guidelines added patients with single ventricle physiology to group 5 (5.4), pulmonary hypertension with unclear and/or multifactorial mechanism. It also acknowledges that there is insufficient data to show that targeted therapies are safe and efficient in this population and that further studies are required [11].

The two most commonly used oral medications for pulmonary vasodilatation in children are sildenafil and bosentan. Dysfunction of the endothelial nitric oxide pathway has been demonstrated in single ventricle physiology patients, and the lack of pulsatile blood flow has been hypothesized to cause an increase in endothelin-1 [14, 15]. This has provided the rationale for attempts to target these pathways to reduce pulmonary vascular resistance in patients with single ventricle physiology.

Improved exercise tolerance following sildenafil treatment has been reported in patients with total cavopulmonary circulation [16]. Studies have also shown that pulmonary vasodilators lowers mean pulmonary arterial pressure and vascular resistance [17‐20]. Sildenafil treatment for up to two years has been reported in patients with single ventricle physiology with no major side effects [21]. Still, data are limited to few studies with small cohorts and/or short-term follow-up, and it is not yet clearly determined if pulmonary vasodilators can provide sustained benefit to patients in TCPC or patients in earlier stages of this surgical pathway.

The aim of this analysis was to investigate the effects of oral pulmonary vasodilators, used as single or combination therapy, in children at different surgical stages of single ventricle physiology. Focus was on potential impact on saturation, pulmonary arterial pressure, and progression to next surgical stage. We also report the incidence of serious adverse events during treatment and at weaning.

In single ventricle physiology low pulmonary vascular resistance is one of the important factors which will influence functional ability, quality of life, and long-term survival [1]. It is established that a mean pulmonary arterial pressure > 15 mmHg or impaired ventricular function will increase the risk of single ventricle physiology failure after total cavopulmonary connection surgery [2, 3, 22, 23]. This has provided the rationale for increased use of pulmonary vasodilators in this patient group.

This retrospective analysis describes the effects of vasodilator therapy in 36 children with single ventricle physiology and pulmonary vascular disease. To our knowledge this represents one of the larger pediatric cohorts and the total follow-up time of 4.7 (0–12.8) years is longer than for previous publications.

Treatment with a single vasodilator, either sildenafil or bosentan, was associated with a significant improvement in saturation within 12 months after treatment initiation. The mean increase from 80 to 85% is moderate, but an indirect indicator of reduced pulmonary vascular resistance and increased pulmonary blood flow. We also observed a reduction in the mean pulmonary arterial pressure from 19 to 14 mmHg within a mean of 1.3 (range 0.3–2.6) years after treatment start. The clinical effect of the increase in saturation and reduction in pulmonary arterial pressure was harder to determine. Although the majority of patients reported improved well-being on treatment these reports were subjective.

There remains debate as to the mechanisms of pulmonary hypertensive vascular disease in patients with palliated univentricular hearts. On the one hand, there are mechanisms that induce high pulmonary vascular resistance. It has previously been shown that lack of pulsatile flow is associated with a decreased expression of important vasodilator substances, decreased vasorelaxation, and a vascular remodeling characterized by altered apoptosis of vascular smooth muscle cells [5, 24]. There is also an overexpression of the potent vasoconstrictor endothelin-1 in lung tissue of single ventricle patients [15]. Increased medial thickness and extension of the smooth muscle cells to the walls of distal intra-acinar pulmonary arteries in biopsies taken prior to total cavopulmonary connection have been shown to correlate with poor prognosis [25, 26]. On the contrary, Ridderboos et al.[6] found intimal fibrosis and reduction of muscular smooth cells in lung tissue obtained from autopsy in long-standing single ventricle physiology patients. A clear understanding of these mechanisms should deliver new treatment that may further reduce pulmonary vascular resistance.

A number of potential risk factors for high pulmonary vascular resistance were identified in our cohort. Late pulmonary arterial banding with excess circulation to the lungs was found in 11 patients and 10 patients had a restrictive atrial communication that required an intervention. Highly complex heart defects with isomerism and anomalous drainage of venous return were also common in the cohort. Prior to vasodilator therapy 79% (22/28 of known pressure values) of the patients in our study had elevated pulmonary arterial pressures.

Our results corroborate the findings of previous studies. Park et al. [18] retrospectively studied 34 patients and reported a significant reduction in mean pulmonary arterial pressure following vasodilator treatment. Mori et al. [17] followed 42 patients, whereof the majority were children, who were in different stages of single ventricle physiology and treated with sildenafil for 3 months. They too reported a significant reduction in mean pulmonary arterial pressure. Recently, Castaldi et al. studied effects of pulmonary vasodilators and could among other hemodynamic measures also show a reduction in pulmonary artery resistance [27]. Studies focused on total cavopulmonary connection in smaller cohorts have also shown positive early effects of sildenafil including increased saturation, reduced pulmonary arterial pressure, and reduced pleural effusion [20, 28]. Ovaert et al. [29] treated 10 single ventricle patients for 16 weeks with bosentan. They did not observe any significant improvement, but a trend towards an increase in saturations and reasoned that small numbers was a limitation for their study. Other smaller studies have reported favorable effects of bosentan on mean pulmonary arterial pressure, functional class, and 6-min-walk test [30, 31]. Handler et al. [32] showed significant decrease in pulmonary vascular resistance and increase in saturation with subcutaneous treprostinil for 17 children with single ventricle. A recently published large randomized controlled trial by Goldberg et al. demonstrated that udenafil compared to placebo had a modest positive effect on oxygen saturations and measures of submaximal exercise performance in adolescents after total cavopulmonary connection palliation, but udenafil was not shown to improve the primary outcome measure of peak oxygen consumption [33]. The patients in our cohort by comparison were far more symptomatic at baseline with a high proportion of risk factors for pulmonary vascular dysfunction and therefore may have a greater potential for benefit. In their initial report Goldberg et al. have not reported whether a subgroup of more symptomatic patients sustained greater benefit.

In our study, sildenafil was the first choice in most cases when a child was started on pulmonary vasodilator therapy. Bosentan was in most cases chosen as an add-on when a second vasodilator was considered necessary. This practice mirrors those reported from other registries [34, 35].

We did not observe any consistent additional changes in saturation and mean pulmonary arterial pressure when a second vasodilator was added. This lack of response with add-on therapy was also seen in the COMPASS-2 study where added bosentan was not superior to sildenafil monotherapy in delaying the time for the first morbidity/mortality event in patients with pulmonary arterial hypertension and biventricular physiology [36]. On the other hand, initial combination therapy has shown reduced risk of clinical failure events in the recent published study by White et al.[37]. In our study, a few patients with single ventricle physiology were given initial combination therapy, but the numbers were too small for comparison with monotherapy.

Few adverse effects were observed during treatment or after discontinuation. With a total follow-up time of 4.7 years (0–12.8), this represent one of the longer observations. In this cohort both sildenafil and bosentan were well tolerated.

There are limitations to this observational study. The United Kingdom Service for Pulmonary Hypertension in Children is a national referral center. The findings in this study are therefore vulnerable to referral bias. Many of the patients were already treated with sildenafil when first seen by the service. They may also represent a selection of more complex single ventricle physiology compared to those seen at a local center. Despite a large center the number of children with single ventricle physiology treated with these drugs were small. Some patients had to be excluded from the analysis because of confounding factors (see methods). This study has no control group, as all patients followed by the United Kingdom Service for Pulmonary Hypertension in Children had an indication for vasodilator therapy and all were treated. The heterogeneity of the cardiac diagnoses in this group made it difficult to predict who would benefit from vasodilator treatment.

In conclusion, vasodilator therapy can achieve and sustain improvements in both saturation and mean pulmonary arterial pressure in a group of severely affected children with single ventricle physiology. The total mean follow-up time was 4.7 years. There were no major adverse reactions with long time vasodilator therapy and vasodilators could be discontinued safely when indicated. 7/12 patients progressed in surgical stage during treatment. This study adds useful information on long-term effects of pulmonary vasodilators in a group of complex single ventricle patients with suspected pulmonary vascular disease. The cohort is also one of the largest so far, with longer total follow-up time compared to previous publications. The mechanisms for pulmonary vascular disease in these children are not fully understood and further studies, ideally experimental studies and prospective multicenter randomized controlled trials, are needed.