Introduction
In fetal circulation, the ductus arteriosus allows blood to transfer from the right ventricle to the aorta, bypassing the non-functioning lungs. Within the lumen of the ductus, endogenous prostaglandins, namely prostaglandin E2 (PGE2) and prostaglandin I2 (PGI2), maintain ductal patency. The patency of the ductus allows pulmonary blood flow to be maintained. Reference Bouma and Mulder1 At birth, the decrease in endogenous prostaglandins and the increase in arterial oxygen saturations facilitate the closure of the ductus. In CHD, a neonate is determined to have a ductal-dependent lesion when the pulmonary or systemic blood flow is dependent on the patency of the ductus arteriosus. Reference Bouma and Mulder1
Alprostadil, a synthetic analogue of prostaglandin E1 and a potent vasodilator, has been utilised for infants with ductal-dependent CHDs to maintain the patency of the ductus arteriosus until palliative or corrective surgery becomes possible. Reference Olley, Coceani and Bodach2–Reference Heymann, Berman, Rudolph and Whitman4 Alprostadil received approval from the Food and Drug Administration in 1981 for temporary maintenance of patency of the ductus arteriosus in neonates with ductal-dependent CHD until surgery can be performed. Reference Roehl and Townsend5 The effectiveness of alprostadil was established through studies conducted in the late 1970s, which reported successful maintenance of ductal patency in different forms of CHDs utilising doses ranging from 0.025 to 0.75 mcg/kg/min. Reference Olley, Coceani and Bodach2–Reference Heymann, Berman, Rudolph and Whitman4 However, subsequent safety studies prompted by observations made by various investigators indicated that the approved dosing strategy was associated with side effects, including episodes of respiratory depression, apnoea, and peripheral vasodilation. Reference Lewis, Freed, Heymann, Roehl and Kensey6–Reference Host, Halken, Kamper and Ullquist10
Hallide-Smith et al. conducted a case series study in 1984, exploring lower dose ranges than those previously reported, and demonstrated that effective clinical improvements could be achieved with dosages of 0.005 to 0.01 mcg/kg/min without any notable serious side effects. Reference Hallidie-Smith11 Subsequently, numerous case series and reviews have examined the clinical efficacy of dosing strategies lower than the initial recommended dose of 0.1 mcg/kg/min. Reference Silove, Roberts and De Giovanni12–Reference Yucel, Cevik and Bulut17 These studies have demonstrated the clinical efficacy of administering alprostadil at lower doses, observing a decrease in dose-related adverse events commonly experienced by patients in this population. Despite documented adverse events, the recommended dosing guidance for alprostadil in medical references has remained unaltered since its initial approval.
At our institution, the starting dose of alprostadil is currently 0.05 mcg/kg/min. Historically, a lower maintenance dose of alprostadil of 0.006 mcg/kg/min was utilised. Alprostadil infusions are titrated to maintain patency of the ductus arteriosus and monitored utilising conventional clinical parameters and echocardiograms. Given the lack of data on an ideal dose regimen and prior reports of adverse events even at manufacturer-recommended starting doses, we sought to describe alprostadil dosing at our institution using two different dose strategies and associated adverse events.
Methods
This study was motivated by an observation of adverse events following changes to computer order entry for alprostadil, leading to a higher initial starting dose. For patients in cohort 1, the standard of care was to administer alprostadil at a dose of 0.006 mcg/kg/min. For patients in cohort 2, the standard of care was to initiate alprostadil at a dose ranging from 0.01–0.05 mcg/kg/min with little to no decrease in dose over time. This observed change in dosing strategy accompanied an institutional change in the electronic medical record and system harmonisation.
Study design
This was a single-centre retrospective study at Johns Hopkins All Children’s Hospital evaluating patients who received alprostadil from January 2015 to December 2015 (cohort 1) and January 2021 to December 2021 (cohort 2). Patients were eligible for inclusion if they were neonates diagnosed with ductal-dependent CHD, admitted to the cardiovascular ICU, and received alprostadil within the first seven days of life. Patients who were initiated on alprostadil after day 7 of life or those who received alprostadil for greater than 24 hours at a referring institution were excluded from analysis.
Data elements
Data elements of interest include patient demographics (e.g. age, gender, weight), diagnosis of ductal-dependent CHD, alprostadil dosing information (e.g. dose, dose changes, duration), and clinical outcomes (e.g. mortality, complications, interventions). The alprostadil dose was evaluated at the initiation of therapy and the discontinuation of therapy. The duration of alprostadil therapy and the number of dose adjustments that occurred were also evaluated.
Outcomes
The primary objective of this study is to describe the dosing of alprostadil in neonates with ductal-dependent CHD during two different periods. The secondary objectives were to determine clinical outcomes and adverse events associated with different alprostadil dosing strategies.
Definitions
Fever was defined as a temperature measured ≥38°C. Sepsis workup in this study was defined as a fever resulting in blood cultures being drawn and the initiation of antibiotics, with or without a C-reactive protein or procalcitonin level being obtained.
Statistics
Descriptive statistics were used to summarise patient demographics, alprostadil dosing, and clinical outcomes. Continuous variables were reported as a median value with ranges. Categorical variables were presented as frequencies and percentages. Comparative analysis between the two study periods (cohort 1 vs. cohort 2) was performed using the Mann–Whitney U tests for continuous variables and Chi-square for categorical variables.
Results
Eighty-two patients were initially screened for this study. Seventeen patients were excluded due to receiving alprostadil for more than 24 hours at an outside institution before admission to our study site (n = 14) or having insufficient data (n = 3). For this study, insufficient data are defined as data collection variables needed for collection that were unable to be collected from historical documentation (some of which included scanned paper charts). Of the sixty-five remaining patients included in the study, thirty-eight were in cohort 1 and twenty-seven in cohort 2.
Baseline demographics for each cohort are reported in Table 1 and were similar between the two cohorts. Systemically obstructive lesions were most commonly reported, with cohort 1 having a lower incidence of pulmonary obstructive lesions compared to cohort 2 (18.4% versus 29.6%, p = 0.447) but a higher incidence of mixed obstructive lesions compared to cohort 2 (26.3% versus 14.8%, p = 0.435). The median doses of alprostadil at initiation were significantly lower in cohort 1 compared to cohort 2 (0.006 [0.006–0.01] vs. 0.025 [0.0175–0.05], p < .001). In addition, the median doses of alprostadil at discontinuation were significantly lower in cohort 1 compared to cohort 2 (0.006 [0.006–0.0064] vs. 0.01 [0.01–0.03], p < 0.001). The median duration of therapy was longer in cohort 1 compared to cohort 2, and caffeine appeared to be less frequently utilised in cohort 1 compared to cohort 2 (5.3% vs. 18.5%, p = 0.089) (Table 2). With regard to clinical outcomes, over 70% of subjects in both cohorts underwent surgical intervention, and time to intervention (days) was similar between the cohorts (Table 2). Of note, 3 (7.9%) patients in cohort 1 required an emergent intervention compared to no patients in cohort 2, and overall survival was similar between the two cohorts (Table 2).
Baseline patient demographics
Clinical outcomes
Apneic events and necrotising enterocolitis were defined by documentation of such in the electronic medical record. Apneic events were less prevalent in cohort 1 than in cohort 2 (26.3% vs. 51.9%, p = 0.037). Similarly, apneic events requiring higher-level respiratory support occurred less frequently in cohort 1 compared to cohort 2 (18.4% vs. 33.3%, p = 0.168). Of note, more patients in cohort 2 (25.9%) developed fevers while on alprostadil compared to patients in cohort 1 (13.2%). The incidence of sepsis workup was similar between the two cohorts, and there were no cases of necrotising enterocolitis reported in either cohort.
Discussion
This single-centre study reports that the higher doses of alprostadil used more recently in our institution in neonates with ductal-dependent CHD were associated with an increased incidence of adverse events but no observed difference in clinical outcomes as compared to the lower doses of alprostadil used previously. An optimal dose of alprostadil to maintain ductal patency remains inconclusive at this time. The initial dose of alprostadil utilised at our institution is currently 0.05 mcg/kg/min. Alprostadil infusions are titrated to maintain patency of the ductus arteriosus and monitored utilising conventional clinical parameters and echocardiograms. Prior to 2021, within the health system, a lower initial alprostadil dose of 0.006 mcg/kg/min was utilised as compared to the current higher initial dose of 0.05 mcg/kg/min. In 1976, Olley et al. conducted one of the first case series that described the use of alprostadil in ductal-dependent neonates. Reference Neutze, Starling, Elliott and Barratt-Boyes3 They reported success maintaining the ductus arteriosus with initial alprostadil doses of 0.11–0.14 mcg/kg/min. Freed et al. further evaluated 492 neonates across 52 centres with cyanotic and acyanotic CHD. These researchers found that doses as low as 0.05 mcg/kg/min of alprostadil were efficacious. However, they could not identify the lowest effective dose of alprostadil due to limitations in monitoring. Reference Freed, Heymann, Lewis, Roehl and Kensey18 Hallide-Smith et al. conducted a dose-finding study and reported similar clinical efficacy to Freed et al. with doses ranging from 0.005 mcg/kg/min to 0.01 mcg/kg/min. Reference Hallidie-Smith11 More recent studies support the findings of Hallide-Smith and have also demonstrated clinical success with initial alprostadil doses of 0.003–0.01 mcg/kg/min. Reference Huang, Lin and Huang16,Reference Yucel, Cevik and Bulut17 This study reported that lower initial doses of alprostadil at 0.006 mcg/kg/min were clinically efficacious. These findings are consistent with published literature and support the use of lower doses of alprostadil than what is currently recommended in compendia and the package insert. Clinical efficacy in previous studies had variable definitions, including echocardiographic measurement of the ductus arteriosus or measurement of clinical parameters such as systemic arterial blood pressure and peripheral capillary oxygen saturation. Reference Roehl and Townsend5,Reference Hastreiter, Van Der Horst, Sepehri, Dubrow, Fisher and Levitsky7,Reference Silove, Roberts and De Giovanni12,Reference Kramer, Sommer, Rammos and Krogmann14,Reference Huang, Lin and Huang16,Reference Yucel, Cevik and Bulut17 In this study, clinical efficacy was determined by patient survival until corrective intervention (cardiac catheterisation or surgery). Similar survival rates between the two cohorts were noted and are similar to previous studies. Reference Silove, Roberts and De Giovanni12,Reference Yucel, Cevik and Bulut17,Reference Tálosi, Katona, Rácz, Kertész, Onozó and Túri19 Emergent interventions were more common in cohort 1, with two of the three patients unable to tolerate low-dose alprostadil due to significant apneic and bradycardic events, and one patient required an intervention within hours of delivery due to severe cyanosis. Based on this information, we believe it is unlikely that a higher dose of alprostadil would have benefited these patients, but with limitations of a retrospective review, this could not be determined with certainty.
Common adverse effects of alprostadil described in the literature include necrotising enterocolitis, apnoea, bradycardia, and fever. Reference Lewis, Freed, Heymann, Roehl and Kensey6,Reference Huang, Lin and Huang16,Reference Freed, Heymann, Lewis, Roehl and Kensey18–Reference Coceani and Olley20 We also observed similar rates of apneic events, necrotising enterocolitis, and fever when using alprostadil as compared to other published literature, and similarly found many of these adverse events to be dose-dependent. Reference Lewis, Freed, Heymann, Roehl and Kensey6,Reference Host, Halken, Kamper and Ullquist10–Reference Yucel, Cevik and Bulut17 No patients in either cohort developed necrotising enterocolitis. Patients in cohort 1 were less likely to have an apneic event and were also less likely to require increased respiratory support due to an apneic event. Patients in cohort 1 were also half as likely to develop a fever. The incidence of a sepsis workup was similar between the two cohorts.
Potential limitations of this study include its retrospective nature, which may be subject to inherent biases and missing data due to the inability to obtain data variables from scanned-in paper charts, due to an institutional change in the electronic medical record. The findings from this single-centre study may not be generalisable to other populations or healthcare settings. In addition, the study’s observational design limits the ability to establish causality between alprostadil dosing and clinical outcomes. Nonetheless, our findings confirm prior reports indicating that lower doses of alprostadil may be well tolerated.
Conclusion
Higher doses of alprostadil in neonates with ductal-dependent CHD were associated with an increased incidence of adverse events, but no observed difference in clinical outcomes as compared to lower doses of alprostadil. Further studies to evaluate the dosing, efficacy, and safety of the lowest dose of alprostadil would be beneficial to confirm these findings.
Acknowledgements
None.
Financial support
No funding to disclose.
Competing interests
The authors have nothing to disclose.
Ethical standards
This study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board at the Johns Hopkins School of Medicine. Patient confidentiality was ensured by de-identifying the collected data and securely storing the information.
Open access
