Published ahead of print on August 17, 2006, doi:10.1164/rccm.200605-694OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200605-694OC
Sildenafil Prevents Rebound Pulmonary Hypertension after Withdrawal of Nitric Oxide in ChildrenDepartments of Intensive Care and Cardiology, The Royal Children's Hospital; Department of Pediatrics, University of Melbourne; and Murdoch Children's Research Institute, Melbourne, Australia Correspondence and requests for reprints should be addressed to Dr. Lara Shekerdemian, M.D., M.R.C.P., F.R.A.C.P., Paediatric Intensive Care Unit, The Royal Children's Hospital, Flemington Road, Parkville, Victoria 3052, Australia. E-mail: lara.shekerdemian{at}rch.org.au
Rationale: Rebound pulmonary hypertension (PHT) can complicate the weaning of nitric oxide (NO), and is in part related to transient depletion of intrinsic cyclic guanosine monophosphate. Rebound is characterized by increased pulmonary arterial (PA) pressure, cardiopulmonary instability, and in some cases, the need to continue NO beyond the intended period of use. There is anecdotal evidence that sildenafil, a phosphodiesterase-5 inhibitor, may prevent recurrence of rebound. Objectives: We investigated the role of sildenafil in preventing rebound (an increase in PA pressure of 20% or greater, or failure to discontinue NO) in patients in whom previous attempts had not been made to wean from NO. Methods: Thirty ventilated infants and children, receiving 10 ppm or greater inhaled NO, were randomized to receive 0.4 mg/kg of sildenafil, or placebo, 1 h before discontinuing NO. Twenty-nine patients completed the study. Measurements: PA pressures and blood gases were measured before the study drug, and 1 and 4 h after stopping NO. Main Results: Rebound occurred in 10 of 14 placebo patients, and 0 of 15 sildenafil patients (p < 0.001). PA pressure increased by 25% (14–67) in placebo patients, and by 1%(–9–5) in sildenafil patients (p < 0.001). Four placebo patients could not be weaned from NO due to severe cardiovascular instability, whereas all sildenafil patients were weaned (p = 0.042). Duration of ventilation after study was 98.0 (47.0–223.5) h for placebo patients and 28.2 (15.7–54.6) h for sildenafil patients (p = 0.024). Conclusion: A single dose of sildenafil prevented rebound after withdrawal of NO, and reduced the duration of mechanical ventilation. Prophylaxis with sildenafil should be considered when weaning patients from inhaled NO.
Key Words: children • pulmonary hypertension • nitric oxide • rebound • sildenafil Pulmonary hypertension (PHT) is an important cause of morbidity and mortality in pediatric intensive care. PHT can complicate the intensive care unit (ICU) course of infants and children with congenital heart disease (CHD) and patients with acute lung injury (1, 2). Nitric oxide (NO), a highly selective pulmonary vasodilator, was first introduced as a therapeutic agent in the early 1990s (3). Inhaled NO produces pulmonary vasodilatation in ventilated lung regions through activation of endothelial guanylyl cyclase, resulting in increased levels of cyclic guanosine monophosphate (cGMP), and this in turn relaxes pulmonary vascular smooth muscle. Inhaled NO has been shown to reduce the need for extracorporeal life support in term and near-term infants with hypoxic respiratory failure (4, 5). In infants and children with PHT early after surgery for CHD, inhaled NO reduces pulmonary arterial (PA) pressure, without altering systemic hemodynamics (6, 7). Inhaled NO may also improve oxygenation in infants and children with acute lung injury (8). Although a beneficial effect on survival has never been demonstrated in any prospective study of inhaled NO therapy, it continues to warrant consideration as adjunctive therapy for some of our sickest patients in the ICU. An important complication that is associated with the use of inhaled NO is the development of rebound PHT on its withdrawal. This phenomenon of rebound PHT was first described in a cohort of young infants receiving inhaled NO after surgery for CHD (9). Rebound PHT typically occurs during the final phase of NO weaning (typically the final 5 ppm). Rebound PHT is associated with elevation of PA pressure (10), difficulty in ventilation (11), and, in some instances, severe hypoxia and cardiovascular instability (12). Furthermore, the development of rebound PHT does not appear to depend on the presence of preexisting PHT (12) and also occurs in apparent nonresponders to NO therapy (13). Rebound PHT responds, in the short term, to reinstitution of NO. However, this does not offer a complete solution to the problem, as rebound may recur during subsequent attempts to wean from NO. The NO–cGMP pathway of the pulmonary vascular endothelium and smooth muscle plays an important role in the pathophysiology of rebound PHT. Inhaled NO results in the down-regulation of endogenous NO synthase in the pulmonary vascular endothelium (12) and a reduction in guanylyl cyclase activity (14), and its cessation results in acute reduction in plasma and lung tissue cGMP levels (12, 14). Stopping inhaled NO can therefore be associated with pulmonary vasoconstriction until cGMP levels become naturally replete over subsequent hours. In theory, if levels of intrinsic cGMP can be pharmacologically maintained during the final stage of weaning inhaled NO, and for up to 4 h after its discontinuation, then rebound PHT might be avoidable. Sildenafil, a selective inhibitor of phosphodiesterase type 5 (PDE-5), has been widely explored in the treatment of acute and chronic PHT of various etiologies (15–17). Sildenafil prevents the hydrolytic breakdown of cGMP in the pulmonary vascular smooth muscle. In addition to its pulmonary vasodilator effects in the presence of PHT, there is recent anecdotal evidence that it may prevent the recurrence of rebound in infants who have experienced rebound PHT after withdrawal of NO (18). In this prospective, randomized, placebo-controlled trial, we aimed to establish whether a single dose of sildenafil prevents rebound PHT after withdrawal of inhaled NO in infants and children.
The institutional ethics committee at the Royal Children's Hospital, Melbourne, approved this prospective, randomized, double-blinded, placebo-controlled study.
Setting
Eligibility Thirty children were recruited for the study between July 2003 and September 2005. All participants were intubated, sedated, and ventilated in the ICU, and receiving inhaled NO at 5 ppm or greater at the time of consent. They all had measurable PA pressures, either by echocardiographic estimation of right ventricular systolic pressure from the Doppler-derived tricuspid regurgitant jet velocity (in 14 patients) or from a direct PA line (in 16 patients). Inhaled NO concentration was measured using a bedside chemiluminescence analyzer. The study protocol commenced when participants were receiving 5 ppm of NO; the weaning of NO until this level was entirely at the discretion of the attending ICU physician.
Randomization
Protocol In patients without a PA catheter, systolic PA pressure was measured using standard equations: from the peak instantaneous echocardiographic Doppler-derived pressure difference between the right ventricle and right atrium, and the simultaneous central venous pressure. None of these patients had obstruction to the right ventricular outflow tract, or intracardiac shunts, which would influence the accuracy of PA pressure measurement.
Outcome Measures
Statistics
Sample Size Calculation
Interim analysis was performed after the first 30 patients had been randomized, and the study was then stopped based on the significant findings discussed below. The enrollment flow chart is given in Figure 1. One infant who had been randomized to receive placebo died after randomization, but before the weaning process had started (Figure 1). Patient details and indications for NO therapy are given in Tables 1 and 2. Baseline demographic and cardiopulmonary data were similar for the two groups. There were no adverse events associated with the study drug administration.
The Incidence of Rebound PHT No patient who received sildenafil experienced rebound. In contrast, 10 of the 14 patients who received placebo experienced an acute elevation of PA pressure (by 20% or greater) during the latter part of the weaning process (p < 0.001; Fisher's exact test; Table 3). Of these, four who developed severe rebound, with major hemodynamic compromise, could not be weaned from NO.
Ability to Successfully Wean from Inhaled NO No patient given sildenafil required reinstitution of inhaled NO after discontinuing therapy, whereas four patients in the placebo group failed to wean from inhaled NO therapy (p = 0.042; Fisher's exact test). The indications for reinstituting therapy with NO were one or more of the following: acute PHT with pressures reaching or exceeding systemic levels (three patients), severe desaturation (three patients), or acute PHT and systemic hypotension (two patients). The timing of reinstituting NO therapy was at 1 ppm (one patient) within 10 min of stopping inhaled NO (two patients) and 2 h after stopping NO in one patient with progressive desaturation since attempted discontinuation, despite increasing inspired oxygen fraction. All patients who failed to wean were electively given sildenafil during a subsequent weaning attempt (more than 24 h later), and did so successfully. No patient in whom NO was discontinued during the study required reinstitution of NO therapy more than 2 h after stopping NO therapy.
PA Pressure
Changes in PA pressure. PA pressure was unchanged between baseline and 1 h after discontinuing inhaled NO in the patients who received sildenafil (absolute range from –17 to +12%; p = 0.62). All but one patient receiving placebo had a rise in PA pressure after discontinuing inhaled NO (absolute range, 0 to +125%; p < 0.001). The changes in PA pressure are illustrated in Figure 2.
Systemic Oxygenation Oxygenation was unchanged for both groups after stopping NO therapy. However, the interpretation of these data is limited by the need in the placebo group for acute increases in inspired oxygen fraction after stopping NO (three patients) and the reinstitution of NO therapy before repeat blood gas analysis (two patients).
Duration of Mechanical Ventilation
We have shown that rebound PHT is a common problem during NO withdrawal, which can be safely and effectively prevented with a single dose of enteral sildenafil, given at the final stage of the weaning process. This study is unique in a number of ways. It is the first study to investigate pharmacologic prophylaxis of rebound PHT during the primary attempt to wean from inhaled NO. Also, this is the first prospective trial of sildenafil in the prevention of rebound PHT. Finally, this study is the first that defines the extent of the problem of rebound PHT in infants and children weaning from inhaled NO in the pediatric ICU. In 1995, Miller described significant hemodynamic instability during weaning from inhaled NO in a cohort of young infants early after cardiac surgery. This necessitated continued inhaled NO treatment beyond the intended period of therapeutic benefit, and a significant increase in PA pressure after ceasing therapy. Thus, the term "rebound PHT" was coined (9). In addition to rebound PHT, the prolonged use of inhaled NO can be associated with methemoglobinemia (19), the need for continued mechanical ventilation, potentially increased bleeding risk (20), and significant expense (21). Effective and successful weaning from inhaled NO, as and when clinically indicated, is clearly pivotal. Before commencing this study, our NO weaning protocol was already aimed at minimizing the likelihood of rebound PHT, as described in METHODS. This process included a stepwise reduction in the dose of NO, and an increase in inspired oxygen fraction during the final part of the weaning process, while paying particular attention to acid-base balance, and avoiding pain or agitation. Despite these measures, rebound PHT occurred in 10 of 14 control subjects, of whom four became hemodynamically unstable and could not be weaned from NO on that occasion. By contrast, rebound PHT did not occur in any of the patients who received sildenafil. Furthermore, the presence of rebound PHT was subsequently associated with prolonged duration of ventilation and ICU stay. Sildenafil has been well explored by us (15, 22) and others (23) as an acute pulmonary vasodilator for patients with PHT. A potential limitation for its use in patients with acute lung disease has been an observation of arterial hypoxemia secondary to increased intrapulmonary shunt (24, 25). This may be in part dose related, and is clearly more marked in the presence of acute parenchymal lung disease. In the present study of children without acute lung disease, we were encouraged that a single dose of enteral sildenafil was not associated with any deterioration in oxygenation, though the final stage of our weaning protocol included a deliberate increase in inspired oxygen fraction. PDE-5 inhibitors have been used in the treatment of recurrent rebound PHT after previous unsuccessful attempts to wean from inhaled NO in children after surgery for CHD (18, 26), and in newborn infants with PHT (27). However, ours is the first study to explore the "prophylactic" use of a PDE-5 inhibitor agent during the primary weaning attempt. As well as achieving the primary endpoint of enabling the weaning from inhaled NO without features of rebound PHT, our study clearly shows that patients in whom this process was facilitated with the use of sildenafil had a smoother subsequent ICU course, with substantially shorter duration of ventilation and ICU stay. The exact mechanism underlying this important observation is unclear, but this was not due to baseline differences between the groups. Inhaled NO is currently unrivalled in its efficacy as a selective pulmonary vasodilator. Although there are no clinical trials in the literature that have demonstrated improved survival or reduced duration of ventilation with inhaled NO, it remains within our therapeutic spectrum for the management of our sickest patients. If future clinical trials of NO therapy are to be performed, given the association between rebound with increased ICU stay, it may be of interest to include sildenafil in the final stage of the weaning process. A potential limitation of this study is that we did not directly measure endogenous NO synthase activity or cGMP levels. These measurements would have complemented our data. However, our study was aimed at addressing a clinical problem, the pathophysiology of which has previously been extensively investigated, and was based on robust data that have examined the endogenous NO system in the presence of inhaled NO, after discontinuing it, and in response to sildenafil. A second limitation of the study is that not all patients had PA pressure catheters. PA catheters are rarely used in the pediatric ICU outside of the cardiac surgical setting, and in adult practice, their popularity has fallen over recent years. The echocardiographic estimation of right ventricular systolic pressure from the tricuspid regurgitant Doppler jet is an established means of assessing right ventricular (and PA) pressures. The patients in our study who did not have direct PA catheters had right ventricular systolic pressures that were easily measurable using echocardiography. Finally, the timing of the study drug was a critical component of our protocol design. There are currently no pharmacokinetic data regarding the profile of sildenafil in critically ill children. Therefore, the timing of the dose of sildenafil was based on pharmacokinetic data from healthy adults, which show that enteral sildenafil is very well absorbed, and maximum plasma concentrations are reached within 1 h of its administration. Its subsequent elimination half-life is between 3 and 4 h (28). This pharmacokinetic profile balances well the changes in endogenous NO synthase activity and intrinsic cGMP production that occur when weaning from inhaled NO (12).
Conclusions
The authors thank Ms. Kay Hynes, and her colleagues in the Drug Information Department, The Royal Children's Hospital, for their help in the preparation of the study drugs, and assistance with the randomization process.
Originally Published in Press as DOI: 10.1164/rccm.200605-694OC on August 17, 2006 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form May 23, 2006; accepted in final form August 17, 2006
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
level of 0.05, then a total sample size of 40 participants (20 participants per study arm) would be necessary. However, it was agreed that interim analysis would be permitted after the first 30 patients were recruited. 
