INTRODUCTION — During fetal life, the ductus arteriosus (DA) diverts blood from the pulmonary artery into the aorta, thereby bypassing the lungs (figure 1). After birth, the DA undergoes active constriction and eventual obliteration. A patent ductus arteriosus (PDA) occurs when the DA fails to completely close after delivery.
Preterm infants with moderate to large left-to-right shunts have a greater mortality rate than those without a PDA. They also have an increased risk of developing pulmonary edema and hemorrhage, bronchopulmonary dysplasia, and potential end-organ damage due to a decrease in tissue perfusion and oxygen delivery. As a result, management of preterm infants with hemodynamically significant PDAs has been focused on PDA closure and prevention.
The management of PDA in preterm infants will be reviewed here. The pathophysiology, clinical manifestations, and diagnosis of PDA in preterm infants are reviewed separately. (See "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis".)
OVERVIEW — The management of PDA in preterm infants includes the following strategies:
●Conservative management with general supportive measures alone. (See 'Conservative management' below.)
●Pharmacologic closure using cyclooxygenase (COX) inhibitors (eg, indomethacin, ibuprofen) or acetaminophen (paracetamol). (See 'Directed pharmacologic therapy' below.)
●Prophylactic COX inhibitor therapy to prevent PDA.
●Surgical ligation. (See 'Surgical ligation' below.)
It remains unclear which approach is most advantageous for preterm infants, as there have not been large randomized controlled trials comparing these different strategies. Several unanswered questions regarding management include:
●Does a PDA need to be closed? There is increasing observational evidence that spontaneous closure occurs in a significant proportion of infants with symptomatic PDA, which may preclude the need for either pharmacologic or surgical intervention.
●What are the indications for PDA closure? Is timing of closure affected by the duration of exposure to shunting through a PDA, percent of shunting, and/or other clinical factors?
●When the decision is made for PDA closure, is one approach (medical versus surgical ligation) better than another? Does the choice change based on the specific clinical settings or subgroups of patients?
These uncertainties have led to variation in the management of PDA in preterm infants. In many neonatal intensive care units (NICUs), the management of PDA in preterm infants begins with conservative management with subsequent therapeutic interventions focused on PDA closure for symptomatic infants with persistent hemodynamically significant PDAs. Whereas, in other NICUs, early targeted therapy is given to infants with large hemodynamic shunts. (See 'Conservative management' below and 'Outcome' below and "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis", section on 'Consequences of a PDA' and "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis", section on 'Hemodynamically significant PDA'.)
MANAGEMENT APPROACH — There are no unequivocal data to determine the optimal management of PDA in preterm infants (see 'Overview' above and 'Comparison of management approaches' below). As a result, there is variability in the treatment of PDAs amongst neonatologists and pediatric cardiologists. Standardization of care within institutions has led to the development of clinical tools assessing the risk of a hemodynamically significant PDA based on clinical and echocardiographic scoring systems, such as one used by the neonatal division at the Brigham and Women’s hospital (algorithm 1).
The following is a consensus approach to the management of PDA in preterm infants based on the practice of the members of the neonatology and cardiology staff at our center that begins with supportive care, as spontaneous resolution will occur in many patients (see 'Conservative management' below). Subsequent directed pharmacologic therapy for PDA closure is administered to infants with persistent PDAs who remain dependent on mechanical ventilation after one week of age. Percutaneous transcatheter occlusion or surgical ligation is reserved for the infant who fails pharmacologic therapy and remains on maximal ventilator support. This moderate approach is supported by data from a multicenter study of 139 neonatal intensive care units showing that low and high treatment rates (pharmacologic or surgical intervention) were associated with increased death or severe neurologic outcome compared with a more moderate approach [1]. This U-shaped relationship suggests that low treatment rate was associated with a too high threshold for treatment with subsequent adverse effect of untreated significant hemodynamic impact, whereas the high treatment rate may have resulted in unnecessary treatment of infants with milder hemodynamic PDA and exposure to adverse effects of treatment. (algorithm 1).
In our center, this step-wise management approach for infants with hemodynamically significant PDA with initial conservative general measures has resulted in a decrease in the number of infants who receive directed pharmacologic treatment, and there is rarely a need for percutaneous or surgical intervention. In our practice, a hemodynamically significant PDA is diagnosed in infants who exhibit signs of pulmonary overcirculation and left ventricular overload and have an echocardiographic transductal diameter greater than 1.4 mm and evidence of left heart overload (eg, left atrial and ventricular enlargement). Throughout the course, ongoing monitoring is performed with clinical assessment and echocardiography. (See "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis", section on 'Hemodynamically significant PDA'.)
●Conservative management – For infants with a hemodynamically significant PDA, an initial conservative approach with supportive general measures is provided that includes (see 'General measures' below):
•Daily moderate fluid restriction between 120 and 130 mL/kg while providing adequate nutrition of at least 120 kcal/kg per day. The desired nutrients may need to be contained in a reduced volume of feeding in order to deliver the necessary daily caloric intake for growth and maintain fluid restriction. In these cases, a milk fortifier may be added to human milk to increase the caloric density. Mother's milk is preferred but if not available donor human milk should be used.
•Neutral thermal environment.
•Minimal respiratory support ensuring adequate oxygenation (target pulse oximetry saturation [SpO2] between 90 and 95 percent) and allowing permissive hypercapnia (partial pressure of carbon dioxide [PaCO2] 50 to 55 mmHg) as long as pH remains in the normal range (ie, 7.3 to 7.4). For infants who are mechanically ventilated, small tidal volume ventilation to minimize further pulmonary injury and positive end-expiratory pressure (PEEP) of 5 to 7 cm H2O to prevent atelectasis are recommended. (See "Bronchopulmonary dysplasia: Management", section on 'Respiratory support'.)
•A thiazide diuretic is used to treat infants who become fluid-overloaded or with signs of increased interstitial pulmonary fluid.
•Maintenance of a hematocrit above 35 percent.
●Pharmacologic intervention:
•For infants who remain on mechanical ventilation after one week who continue to have a hemodynamically significant PDA confirmed by echocardiography, pharmacologic intervention with a course of a cyclooxygenase (COX) inhibitor is given. The choice of COX inhibitor is center-dependent. In our institution, the preferred COX inhibitor is ibuprofen, which is administered intravenously as 10 mg/kg for the first dose, followed by two doses of 5 mg/kg given every 24 hours. However, other centers use indomethacin or acetaminophen. We do not routinely measure platelets prior to drug administration but do avoid COX inhibitor use in infants with coagulopathy. (See 'Nonselective cyclooxygenase inhibitors' below.)
•A second course of COX inhibitor therapy is given if follow-up echocardiograms show failure of PDA closure and the infant is still ventilator-dependent.
•For patients in whom nonselective COX inhibitor therapy is contraindicated, we use acetaminophen at a dose of 15 mg/kg every six hours. (See 'Acetaminophen (paracetamol)' below.)
●Procedural closure options include percutaneous transcatheter occlusion and surgical ligation. These procedures are rarely performed and are reserved for infants with large PDAs who fail to respond to pharmacologic therapy. In our center, percutaneous transcatheter occlusion is the preferred procedure and surgical ligation is reserved for those infants who are not a candidate for percutaneous closure, including those on maximal ventilator support defined as positive inspiratory pressure >25 mmHg and/or fractional inspired oxygen (FiO2) >70 percent. (See 'Percutaneous transcatheter occlusion' below and 'Surgical ligation' below.)
GENERAL MEASURES
General measures — All neonates with PDA are managed with moderate fluid restriction to limit excess interstitial pulmonary fluid and avoid pulmonary edema. In addition, the following general measures are provided:
●A neutral thermal environment that minimizes demands on left ventricular function. (See "Short-term complications of the preterm infant", section on 'Hypothermia'.)
●Minimal respiratory support that provides adequate oxygenation (target pulse oximetry saturation [SpO2] between 90 and 95 percent) and allows for permissive hypercapnia (partial pressure of carbon dioxide [PaCO2] 50 to 55 mmHg) as long as pH remains in the normal range (ie, 7.3 to 7.4). For mechanically ventilated infants, small tidal volume ventilation to minimize further pulmonary injury and positive end-expiratory pressure (PEEP) of 5 to 7 cm H2O to prevent atelectasis are recommended. (See "Approach to mechanical ventilation in very preterm neonates", section on 'Clinical approach' and "Overview of mechanical ventilation in neonates", section on 'Minimizing ventilator-induced lung injury'.)
●Maintenance of the hematocrit at 35 to 40 percent may increase pulmonary vascular resistance and reduce left-to-right shunting, however, no trials have evaluated the effect of blood transfusion on PDA closure [2]. (See "Red blood cell transfusions in the newborn".)
Fluid management — Excessive fluid administration (greater than 170 mL/kg per day) is associated with an increased incidence of PDA [3]. As a result, although evidence for efficacy is lacking, moderate daily fluid intake between 120 and 130 mL/kg is a reasonable approach to limit pulmonary edema in infants with hemodynamically significant PDA, especially in those with severe respiratory disease.
Diuretic therapy — Although diuretic therapy has been shown to improve short-term pulmonary mechanics, there is little evidence that the long-term use of diuretics improves clinical outcome of infants with PDAs and in fact, may be associated with adverse effects [4]. Nevertheless, we use diuretics to improve pulmonary function in patients who remain ventilator- or PEEP-dependent despite modest fluid restriction. When diuretic therapy is considered, we suggest the use of thiazide diuretics (eg, chlorothiazide). We do not routinely use furosemide or any other loop diuretic in the first week or two after birth, as this stimulates renal synthesis of prostaglandin E2, a potent vasodilator that maintains ductus arteriosus (DA) patency.
In a trial of diuretic therapy in preterm infants with respiratory distress syndrome (RDS), a PDA occurred more frequently in infants treated with furosemide compared with chlorothiazide (55 versus 24 percent) [5]. In a retrospective study of preterm infants below 32 weeks gestation, furosemide was associated with an increase in serum creatinine and hyponatremia, but not an increase in urine output [6].
DIRECTED PHARMACOLOGIC THERAPY — Pharmacologic therapy is directed towards inhibiting prostaglandin synthetase, as prostaglandin E2 (PGE2), which has two catalytic sites, promotes ductal patency [7]. The non-selective cyclooxygenase (COX) inhibitors (ibuprofen and indomethacin) inhibit the COX site, and acetaminophen (paracetamol) affects the peroxidase segment. However, other centers select alternative COX inhibitors based on their clinical experience, availability, and cost.
Nonselective cyclooxygenase inhibitors — In randomized trials, the nonselective COX inhibitors, ibuprofen and indomethacin are more effective than conservative management in closing of PDA in preterm infants [8-10]. However, the administration of COX inhibitors may not affect mortality or the risk of developing BPD. (See 'Comparison of management approaches' below.)
The choice of COX inhibitors is center-based. Ibuprofen is the preferred agent at our institution, as it is associated with a lower risk of necrotizing enterocolitis (NEC) and transient renal insufficiency [11,12]. However, other centers use indomethacin, and increasingly acetaminophen is being used.
Ibuprofen — In systematic reviews of randomized clinical trials, ibuprofen was as effective as indomethacin in closing PDA and was associated with a lower risk of NEC and transient renal insufficiency [12-14]. Risks were similar between ibuprofen and indomethacin for mortality, grade III intraventricular hemorrhage (IVH), and periventricular hemorrhage (grade IV IVH), pulmonary hemorrhage, intestinal hemorrhage, sepsis, and retinopathy of prematurity.
Dose — The standard dosing of ibuprofen for PDA closure for both oral and intravenous administration is an initial dose of 10 mg/kg followed by two additional doses of 5 mg/kg given at 24-hour intervals. Ibuprofen is typically given as an intravenous (IV) preparation in developed countries. However, the IV preparation is expensive and many nurseries in resource-limited countries use oral ibuprofen for PDA closure. In a systematic review, it appears that the oral administration of ibuprofen is equally as effective as IV administration [11].
High-dose ibuprofen administered either orally or intravenously as an initial dose of 15 to 20 mg/kg followed by two additional doses of 7.5 to 10 mg/kg administered every 12 to 24 hours is also used and appears to offer the highest likelihood of pharmacologic PDA closure [15,16]. (See 'Comparison of drug regimens' below.)
Indomethacin — It has been clearly established both by observational studies and randomized trials that indomethacin increases the rate of PDA closure compared with placebo or no therapy within 24 hours of its administration [17,18]. Less clearly established are the optimal timing, dose, and duration of treatment [19].
Dose — Indomethacin is usually given intravenously. Dosing varies amongst neonatal centers and ranges between 0.1 and 0.2 mg/kg per dose administered at 12- to 24-hour intervals. More than one dose is typically required for sustained constriction. Across published studies, the most common dosing schedule is three doses (0.2 mg/kg per dose) given at 12-hour intervals [15,20,21].
The pharmacokinetics of indomethacin vary among preterm infants, and serum half-life decreases with postnatal age. An alternative dosing schedule for the three administered doses used in several tertiary centers, including ours, is determined by the postnatal age of the patient and its effect on serum drug pharmacokinetics:
●Infants <48 hours of age – The first dose is 0.2 mg/kg followed by two additional doses of 0.1 mg/kg given at 12-hour intervals.
●Infants >48 hours and <7 days – Three total doses of 0.2 mg/kg per dose given at 12-hour intervals.
●Infants ≥7 days of age – Three total doses of 0.25 mg/kg per dose.
Although some studies suggest that obtaining indomethacin concentrations to adjust dosing may improve the rate of ductal closure, ductal constriction and adverse effects of indomethacin do not correlate strongly with plasma concentrations, and these measurements are not widely available [22-24]. As a result, we do not recommend obtaining indomethacin levels.
Indomethacin only transiently suppresses prostaglandin synthesis, so a prolonged course of indomethacin is not recommended for the routine treatment of PDA in preterm infants because of the increased risk of NEC with minimal additional benefit for PDA closure. This was illustrated in a systematic review of five randomized controlled trials that showed a prolonged (four or more doses) administration of indomethacin compared with a short course of therapy (three or less doses) did not yield statistically significant differences in PDA closure, retreatment, re-opening, or surgical ligation rates, but increased the risk of NEC [25]. As a result, we use a short course of three doses as noted above.
Complications — COX inhibitors decrease cerebral, gastrointestinal, and renal blood flow, the reduction is greater with indomethacin compared with ibuprofen [26-29]. Reported adverse effects associated with indomethacin include increased risk of bleeding, transient renal insufficiency, NEC, and spontaneous intestinal perforation [17,30,31]. Although ibuprofen is associated with a lower risk of gastrointestinal and renal side effects than indomethacin [11], these remain potential risks associated with both drugs.
It appears that both agents interfere with the binding of bilirubin to albumin, thus potentially increasing the risk of kernicterus. However, there are conflicting data on whether COX inhibitors increase bilirubin in the clinical setting [32-34]. Indomethacin affects bilirubin binding only at plasma concentrations that far exceed those used clinically [32,35], whereas an in vitro study reported that ibuprofen interferes with binding at serum concentrations achieved by the usual doses of the drug [36]. (See "Initial management of unconjugated hyperbilirubinemia in term and late preterm newborns".)
Contraindications — COX inhibitors are contraindicated in infants with the following:
●Proven or suspected infection that is untreated.
●Active bleeding, especially those with active intracranial hemorrhage or gastrointestinal bleeding.
●Thrombocytopenia and/or coagulation defects.
●NEC or suspicion of having NEC.
●Significant impairment of renal function (urine output <1 cc/kg per hour or serum creatinine >1 mg/dL [88.4 micromol/L]).
●Congenital heart disease (CHD) in which patency of the DA is necessary for satisfactory pulmonary or systemic blood flow (eg, pulmonary atresia, severe tetralogy of Fallot, severe coarctation of the aorta).
For these infants, we administer a course of acetaminophen (paracetamol) when it is decided that directed pharmacologic therapy should be given. (See 'Acetaminophen (paracetamol)' below.)
Response to therapy — An echocardiogram performed 24 to 48 hours after completion of the treatment course showing PDA closure confirms a positive response to therapy.
However, a significant proportion of patients will fail to respond to an initial dose of COX inhibitors [37,38]. Limited data suggest a second course of COX inhibitor therapy is associated with a 40 percent rate of ductal closure in patients who fail to respond to initial therapy [39-41]. Therefore, administration of a second course of COX inhibitor is reasonable in infants who fail an initial course of therapy. But, if the patient fails to respond to two courses of therapy, a positive response to additional drug treatment is unlikely and no further medical therapy is given. Risk factors associated with a persistent PDA after one course of COX inhibitor therapy include lower gestational age (GA), lack of exposure to antenatal corticosteroid therapy, increased severity of respiratory distress, and intrauterine inflammation [38,42]. In addition, closure is unlikely in infants who are older than one month of age.
Surgical ligation should be considered in infants with persistent significant PDA who continue to require maximum mechanical ventilator support, whereas no further intervention is indicated in those with small persistent PDA. (See 'Surgical ligation' below.)
Feeding during treatment — Because of adverse effects of indomethacin or ibuprofen on the gastrointestinal tract, there have been concerns about initiation or continuation of enteral feeds during administration of either drug. However, several studies have shown that enteral feeds during COX inhibitor therapy appear to be safe [43-45]. As a result, we continue enteral feedings while COX inhibitor therapy is administered.
Acetaminophen (paracetamol) — Limited data report that both oral and intravenous (IV) preparations of acetaminophen, which inhibits prostaglandin synthetase, successfully result in PDA closure [46-48]. However, the administered dose of acetaminophen was 60 mg/kg per day over a two- to seven-day period, which is considerably higher than recommended doses for pain and fever control in neonates. As a result, there remain concerns regarding hepatotoxicity in the use of this medication at this higher dose as well as long-term effect on neurodevelopment. In addition, data suggest that acetaminophen is less effective for PDA closure than either indomethacin or ibuprofen [12,49,50].
We use acetaminophen for patients in whom nonselective COX inhibitor therapy is contraindicated as a dose of 15 mg/kg every six hours for three to seven days. After the third day of therapy, an echocardiogram is performed to see if there has been a response to therapy. If the PDA is closed, therapy is discontinued. However, if it remains open, acetaminophen is continued till a full seven days of treatment. Laboratory monitoring includes following liver function tests. Further conclusive data are needed to demonstrate whether this medication should be preferred over the nonselective COX inhibitors.
Comparison of drug regimens — Direct head-to-head comparison of pharmacologic agents (indomethacin, ibuprofen, and acetaminophen) is challenging due to variations of treatment protocols across published studies. These variations include differences in the route of administration, dosing, method of administration (enteral versus intravenous bolus or continuous), use of co-interventions (eg, furosemide or dopamine administration), and criteria used to define hemodynamically significant PDA [15].
However, a network meta-analysis, which included 17 direct and 45 indirect comparisons of 14 variations, concluded that high-dose oral ibuprofen was the most effective regimen for PDA closure [15]. In this analysis, based on the premise that included trials were sufficiently similar (patient population, types of intervention, and study methodology), the overall PDA closure rate was 67 percent and included the following findings:
●High-dose oral ibuprofen (initial dose of 15 to 20 mg/kg followed by 7.5 to 10 mg/kg every 12 to 24 hours for total of three doses) was more likely to be associated with closure of a hemodynamically significant PDA and prevent surgical ligation compared with standard intravenous dosing of ibuprofen (OR, 3.59, 95% CI 1.64-8.17; absolute risk difference 199 per 1000 infants, 95% CI, 95-258) or indomethacin (OR 2.35, 95% CI, 1.08-5.31; absolute risk difference 124 per 1000 infants, 95% CI, 14-188). (See 'Dose' above.)
●High-dose oral ibuprofen had the lowest risk for repeat pharmacotherapy, followed by oral acetaminophen.
●Standard IV dosing of ibuprofen was associated with a lower likelihood of closure compared with high IV dose of ibuprofen and standard oral dose of ibuprofen or acetaminophen.
●There was no statistically significant difference amongst the treatment regimens, including placebo or no treatment for mortality, NEC, or grade III or IV IVH (periventricular hemorrhage).
These results need to be carefully considered, particularly based on the assumption of transitivity (ie, studies were sufficiently similar for network analysis). Nevertheless, this analysis provides a good basis for future clinical trials that can elucidate the optimal choice of drug, including dosing and route of administration, when pharmacologic therapy is used for PDA closure.
Based on subsequent data, it appears that acetaminophen is less effective than either indomethacin or ibuprofen.
●In clinical trials, PDA closure rate was lower in the infants who received intravenous acetaminophen compared with those who received intravenous indomethacin [50,51].
●In the multicenter PDA-TOLERATE trial of infants <28 weeks GA, secondary analysis demonstrated that constriction rates compared with conservative management (observation) were highest for indomethacin (relative risk [RR], 3.21, 95% CI 2.05-5.01), followed by ibuprofen (RR 2.03, 95% CI 1.05-3.91), and then acetaminophen (RR 1.33, 95% CI 0.55-3.24) [49].
●In contrast, a study of 161 preterm infants (gestational age <32 weeks) reported oral paracetamol (administered as 15 mg/kg per dose in six hourly intervals for three consecutive days) to be noninferior to oral ibuprofen (administered as an initial 10 mg/kg/dose, followed by 5 mg/kg/dose at 24 and 48 hours after the first dose) in successful closure of a hemodynamically significant PDA (95 versus 94 percent [RR 1.01, 95% CI 0.94-1.1]) [52]. However, there were several limitations that raised concerns of the validity of the finding; more than half of the patients were greater than 28 weeks GA and had a greater likelihood of spontaneous closure, the selected relatively wide range of noninferiority of 15 percent, and the corresponding small numbers of patients [53].
SURGICAL LIGATION — Although surgical ligation is increasingly uncommon, we perform surgical ligation for symptomatic patients who fail to respond to one or two courses of pharmacologic therapy and require maximal ventilator support, or if drug treatment is contraindicated in an infant on maximal ventilator support. Surgical ligation is associated with risks of blood pressure fluctuations, respiratory compromise, infection, intraventricular hemorrhage (IVH), chylothorax, recurrent laryngeal nerve paralysis, bronchopulmonary dysplasia (BPD), and death [54-56].
Because data are observational, it remains uncertain whether surgical ligation is a major contributor to morbidity and mortality, or if patients who undergo surgical ligation are more severely compromised to begin with (see 'Outcome' below). It has been proposed that surgical ligation may contribute to brain injury due to intraoperative compromise of cerebral oxygen saturation and postoperative hemodynamic instability. However, a retrospective study of infants who failed to respond to pharmacologic closure from a single center reported that after adjustment for confounding factors, survivors who underwent ligation compared with those without ligation had similar risks for a composite outcome of death or neurodevelopmental impairment, neurodevelopmental impairment alone, chronic lung disease, or retinopathy of prematurity [57]. In a follow-up report from the same center, 157 of 166 extremely preterm infants (gestational age <28 weeks) treated with surgical ligation survived and were successfully extubated (median 12 days after ligation) [58]. Of note, earlier extubation was associated with larger PDA diameter and moderate to severe left ventricular dilatation, suggesting that these infants had respiratory compromise from a clinically significant PDA.
Nevertheless, in patients in whom surgical ligation is performed, immediate postoperative care must include continuous cardiovascular monitoring, and the use of volume support and inotropic agents to maintain adequate blood pressure and perfusion [59]. Targeted echocardiography may be used to assess postoperative myocardial function and guide management.
PERCUTANEOUS TRANSCATHETER OCCLUSION — Percutaneous PDA closure has been performed in both term and preterm infants, including some patients who weigh <1000 g [60-64]. In 2019, the US Food and Drug Administration approved the Amplatzer Piccolo Occluder for use in infants >700 g and >3 days of life. However, it remains unknown whether this intervention is as effective and safe as surgical ligation, especially in very preterm infants [65]. In a systematic review of 28 studies that included 373 infants who weighed ≤1.5 kg, closure was successfully completed in 96 percent of cases, but there was an 8 percent rate of major adverse events, including five deaths related to the procedure and an overall adverse incidence of 27 percent [66]. There was also a concern for publication bias and results of the meta-analysis may not reflect the true incidence of major adverse events, mortality, and technical success in clinical practice.
At present, the transcatheter approach should be confined to institutions with requisite expertise and experience with interventional techniques in preterm infants until further data are available that show it is equivalent to surgical ligation for both efficacy and safety. In addition, patient selection for percutaneous PDA closure should be based on clear evidence of a hemodynamically significant PDA that has failed medical management.
CONSERVATIVE MANAGEMENT — Observational evidence suggests that conservative management using supportive therapy alone in preterm infants with PDA may be a reasonable option versus pharmacologic or surgical treatment with ongoing assessment and intervention if needed [67-72]. Several experts in the field advocate for conservative management of PDA in preterm infants, based on the data that demonstrates a high rate of spontaneous closure and established adverse effects of both pharmacologic and surgical intervention and evidence that treatment results in similar rates of mortality and morbidity [15,69,73,74].
However, data are limited and include the following supportive evidence [15,68,70,72]:
●In a network meta-analysis that compared pharmacologic agents' effectiveness for PDA closure, placebo or no treatment had the poorest rate for PDA closure. However, all treatment options (including placebo and no therapy) had similar outcomes of mortality, necrotizing enterocolitis (NEC), or intraventricular hemorrhage (IVH) [15]. (See 'Comparison of drug regimens' above.)
●In a retrospective study from two European centers of very low birth weight (VLBW) infants from 2012 to 2013, spontaneous PDA closure was observed in 237 of 280 infants (85 percent) treated conservatively [72]. PDA closure intervention was provided to 17 of the 297 eligible infants (14 with initial medical therapy and four who underwent surgical ligation), although criteria for surgical treatment were not presented. The median time to spontaneous PDA closure increased with decreasing gestational age (range from 71 days for infants born at <26 weeks gestation to 6 days for those at ≥30 weeks gestation) and decreasing BW (range from 48 to 8 days for infants with a BW <750 g and those with a BW between 1250 and 1500 g).
●In a retrospective Korean study that reviewed outcomes from two time periods from 2009 to 2011 (mandatory closure either with medical or surgical treatment) and 2012 to 2014 (supportive therapy alone), there were no differences in mortality, NEC, or IVH [70]. Of note, the incidence of PDA was lower in the second time period of nonintervention.
Although these data suggest that conservative treatment is effective in most VLBW infants, it remains unclear whether there is a subset of patients (eg, more immature infants) who will benefit from PDA closure.
In a study that compared outcomes from two time periods from 2005 to 2011 (prophylactic indomethacin therapy) and 2011 to 2016 (conservative management) of extremely preterm infants (gestational age <28 weeks) with moderate to large PDA (based on echocardiography) at seven days of age, multivariate analysis demonstrated prophylactic indomethacin was associated with a lower risk of BPD or the combined outcome of BPD or death [75]. There was no difference in the incidence of IVH grades III and IV, NEC, retinopathy of prematurity, or death between the two time periods.
Further research efforts are in progress to determine if there are subsets of preterm infants who will benefit from selected interventions and others who can be managed conservatively [76].
COMPARISON OF MANAGEMENT APPROACHES — There have not been large randomized controlled trials comparing pharmacologic management of PDA versus surgical ligation or conservative treatment in preterm infants. Large randomized controlled trials are required to truly compare the different treatment approaches in regards to mortality and complications and to determine if there is a subset of patients (eg, extremely preterm [EPT] infants [gestational age (GA) <28 weeks] and extremely low birth weight [ELBW] infants [BW <1000 g]) that require more aggressive management [67,77]. However, significant challenges regarding study design must be addressed, including identifying the at-risk population based on clinically significant hemodynamic PDA, avoidance of open label treatment in the placebo arm, evaluation of the physiologic impact of each treatment option, and setting realistic clinically important short-term and long-term outcomes [78].
Several clinical trials of preterm infants with a moderate to large PDA and requiring respiratory support have demonstrated similar outcome between early routine medical therapy (indomethacin, ibuprofen, or acetaminophen) and conservative management. However, these studies are limited due to concerns regarding study design (adequate blindness, open label use of drug in placebo group, and the size of the trial).
●In an exploratory trial of 202 EPT infants who at the end of one week of life had an echocardiographic confirmation of a moderate to large PDA and required respiratory support, early routine medical therapy (indomethacin, ibuprofen, or acetaminophen) and conservative management had similar outcomes in the combined primary outcome of ligation or presence of a PDA at discharge (32 versus 39 percent) [79]. There were no differences in the secondary outcome of necrotizing enterocolitis (NEC), bronchopulmonary dysplasia (BPD), BPD and death, death, and need for ongoing respiratory support. Subsequently published secondary analyses reported vascular constriction was greatest for infants treated with indomethacin, followed by ibuprofen and then acetaminophen [49], and ≥ 10 days intubation was associated with an increased risk of moderate to severe BPD [80]. However, this study had several limitations, including unblinded treatment allocation by the medical team and study coordinators who were recording outcome data, and almost 20 percent of eligible infants were not enrolled [79,81]. The group of eligible but not enrolled infants were more immature; appeared to be sicker, as they were more likely to have needed surfactant and intubation than the enrolled patients; and received treatment at an earlier postnatal age (mean age 5.4 versus 8.2 days). This excluded group of infants had similar morbidity and mortality compared to the infants enrolled in the trial.
●In a second non-inferiority trial of 146 preterm infants (GA between 23 and 30 weeks) with a PDA diameter >1.5 mm based on echocardiography and who required respiratory support, patients stratified based on GA (23 to 26 weeks and 27 to 30 weeks) and randomly assigned to nonintervention (normal saline) or treatment with ibuprofen had similar primary outcomes of BPD or death (44 versus 50 percent, risk difference 6 percent, 95% CI -11 to 22 percent) [9]. PDA closure at 1 week after randomization was higher in the treated group versus nonintervention (20 versus 4 percent, adjusted odds ratio 8.77, 95% CI 2.08-36.87).
●In a small pilot study of 60 EPT infants at risk for BPD based on a severity risk score, PDA closure rates were higher for infants who received medical therapy compared with those who received placebo (57 versus 17 percent) [10]. However, there was no difference in the primary outcome of death or BPD between the two groups, but this may be due to the small number of outcome events.
●In a trial of 337 EPT infants, 228 infants with a large PDA identified by echocardiography 6 to 12 hours after birth were randomly assigned to either ibuprofen or placebo by 12 hours of age [82]. The primary outcome of survival at 2 years of age without evidence of cerebral palsy was similar between the two treatment groups (71.3 versus 71.6 percent, adjusted relative risk [RR] 0.98, 95% CI 0.83-1.16). Closure of the PDA by 3 days of age was higher in the ibuprofen group (69.5 versus 19.5 percent). However, this study was limited by the open-label rescue treatment of ibuprofen in almost two-thirds of the placebo group (62.3 percent) and 17.5 percent of the ibuprofen group by day 3 of life and the rate of spontaneous closure in the placebo group.
Observational data are also available with varying results, which may be due to study design due to differences in gestational age (GA), birth weight (BW), time period of care, severity of illness, and criteria for interventions [83,84].
●In a study from the National Institute of Child Health and Human Development (NICHD), the 18-month outcome based on the management of PDA in infants with BW below 1000 g born between 2000 and 2004 was evaluated for 403 infants who were treated conservatively, 1525 treated with indomethacin alone, 135 with primary surgery, and 775 who received indomethacin initially followed by surgery [83]. The following findings were noted:
•Infants who underwent either primary or secondary surgical closure had lower GA, BW, and Apgar scores, but were less likely to be small for gestational age (SGA).
•There were no differences in the measured demographics between infants treated conservatively and those who only received indomethacin.
•There also were no differences between the conservatively and medically managed groups in mortality, neurodevelopmental outcome at 18 months, the composite outcome of mortality and neurodevelopmental outcome, or the risk of bronchopulmonary dysplasia (BPD) or necrotizing enterocolitis (NEC).
•There were no differences between the primary or secondary surgical groups and indomethacin groups in the composite outcome of mortality and neurodevelopmental outcome or rate of NEC. Patients who underwent secondary surgical closure compared with those treated with indomethacin alone were more likely to have BPD and neurodevelopmental impairment, but had a better survival rate.
The study design, however, has a bias for increased survival in patients treated surgically because surgical deaths were only counted after the procedure, whereas deaths in the medically treated group were reported any time after the diagnosis of PDA.
●Similar concerns for residual survival bias were noted in another retrospective cohort study of 754 EPT infants born during a later period (2006 to 2012) with a lower mortality rate for infants who underwent ligation compared with those managed medically [85]. In this study, there was no difference in the composite outcome of neurodevelopmental outcome or death.
●In an observational study from the Canadian Neonatal Network of 3556 infants with PDA from 2004 to 2008, 2026 were treated with indomethacin alone (57 percent), 626 with both surgical ligation and indomethacin (18 percent), 577 with conservative treatment (16 percent), and 327 with surgical ligation alone (9 percent) [56]. Multivariate analysis showed a higher mortality and morbidity (ie, BPD, grade 3 or 4 intraventricular hemorrhage [IVH], and stages 3 to 5 retinopathy of prematurity [ROP]) for patients treated with surgical ligation, irrespective of previous indomethacin therapy, compared with infants treated conservatively or with indomethacin alone. There were no differences in mortality or morbidity between the groups who were treated conservatively or with indomethacin alone.
●In a population-based cohort Australian study of EPT infants (GA <25 weeks) born between 1998 and 2004, multivariate analysis showed greater risk of moderate-to-severe neurodevelopmental disability at two to three years of age for survivors with medically treated PDA (adjusted odds ratio [aOR] 1.62, 95% CI 1.20-2.20) or with surgically treated PDA (aOR 2.0, 95% CI 1.13-3.56) compared with those who did not have a diagnosis of PDA or did not receive intervention for PDA [86].
●Two single-center studies report poorer outcome associated with changes in management to a more conservative regimen [87,88].
PROPHYLACTIC/EARLY THERAPY — We do not recommend the routine use of prophylactic pharmacologic therapy to reduce the incidence of PDA. Although the evidence is of low to moderate quality, early routine treatment does not appear to be beneficial in reducing significant disease and unnecessarily exposes infants without significant PDA to the potential adverse effects of the administered drug [89].
Cyclooxygenase inhibitors — Although the use of prophylactic COX inhibitors had been proposed to reduce the incidence of PDA and improve neonatal outcome, prophylactic COX inhibitors appear not to be more effective at improving mortality, pulmonary outcome, or reducing the risk of necrotizing enterocolitis (NEC) than early treatment of a symptomatic PDA [90-93]. In addition, prophylactic treatment would unnecessarily expose many infants who would not develop significant PDAs to COX inhibitors with potentially serious adverse effects. As a result, we recommend not to administer prophylactic treatment with COX inhibitors to reduce PDAs in preterm infants. However, in some centers, prophylactic therapy is provided to extremely preterm (EPT) infants [94]. Further research is needed to see if there is a specific population of preterm infants that would benefit from prophylactic intervention. (See 'Complications' above.)
A systematic review of the literature reported that prophylactic use of ibuprofen compared with placebo or no intervention probably decreased the incidence of PDA and the need for rescue therapy or surgical ductal closure [95]. Although there was a reduced risk for severe IVH, there was no evidence of benefit in reducing mortality, BPD, NEC or the time to reach full feeds. In the control group, PDA closure occurred spontaneously by day 3 or 4 in 58 percent of neonates. There were significant adverse effects associated with ibuprofen, including oliguria and increased risk of gastrointestinal hemorrhage. The conclusion from the authors of the review was the use of prophylactic ibuprofen exposed a large proportion of infants unnecessarily to a drug that has important side effects without conferring any important short-term benefits.
Acetaminophen (paracetamol) — Prophylactic treatment to reduce the incidence of PDA with acetaminophen (paracetamol) is not recommended, because infants who would not develop significant PDAs are unnecessarily exposed to drugs, and there is no evidence of long-term benefit. In addition, the optimal dosing and the long-term safety of acetaminophen remain unknown.
Data are limited to a small trial of 48 patients that reported intravenous administration of acetaminophen prevented symptomatic PDA without adverse side effects [96]. In this trial, patients randomly selected to receive acetaminophen were given a loading dose of 20 mg/kg within 24 hours of birth, followed by 7.5 mg/kg every six hours for four days.
OUTCOME — Mortality is increased in infants with a persistent PDA especially in extremely preterm infants with a gestational age (GA) 28 weeks or less [97]. The higher mortality rate is ameliorated with successful closure but persists in those with failed attempted closure.
In a retrospective study of 252 infants born at or below 28 weeks gestation, survival outcomes were compared among infants who never had a significant PDA, infants whose significant PDA had been successfully closed medically, and those who had a persistent significant PDA after unsuccessful medical closure [98]. Infants with a persistent PDA had a fourfold increased risk of death compared with infants who never had a significant PDA. There was no difference in mortality rate between the group of infants who had successful medical closure of their PDA and those who never had a significant PDA.
In another retrospective review from a single tertiary center in the United States, 41 very low birth weight (VLBW) infants with a persistent PDA (failed indomethacin therapy and not surgically ligated) had an eightfold increased risk of death compared with 260 infants with a closed PDA after adjustment for confounding factors that included perinatal factors, level of maturity, and disease severity [99].
In infants who receive intervention, it is unclear whether there is a difference in outcome between pharmacologic and surgical therapy (see 'Comparison of management approaches' above). Information is limited as data are primarily based on observational studies and there are concerns for bias based on study design [56,83,100,101]. For example, if the clinical status of the patients (severity of illness) is not controlled, outcome for surgical patients may be poorer as patients who were surgically repaired were more likely to be critically ill and are at greater risk for long-term complications (eg, bronchopulmonary dysplasia and severe retinopathy of prematurity) [100]. In contrast, in several studies there was a bias due to the lack of accounting for confounding factors, as surgical deaths were only counted after the procedure, whereas deaths in the pharmacologically treated group were reported any time after the diagnosis of PDA [83,100].
Randomized trials that compare the clinical outcome of the different management approaches are required to decide upon the most efficacious and safe therapeutic intervention for preterm infants with PDA. Until these data are available, we administer pharmacologic treatment to preterm infants who continue to have persistent symptomatic hemodynamically significant PDA requiring mechanical ventilator support in our care as discussed above [102]. (See 'Management approach' above.)
SUMMARY AND RECOMMENDATIONS — Preterm infants with clinically significant patent ductus arteriosus (PDA) have a greater mortality rate than those without a PDA. Hemodynamically significant PDAs also are associated with an increased risk of pulmonary edema and hemorrhage, bronchopulmonary dysplasia (BPD), and potential end-organ damage due to decrease in tissue perfusion and oxygen delivery. As a result, management has been focused on PDA closure and minimizing these poor outcomes. (See 'Outcome' above and "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis".)
●The optimal management approach to management of PDA in preterm infants is uncertain and practice varies considerably regarding management approaches and, if intervention is chosen, the timing of the intervention. (See 'Overview' above and 'Comparison of management approaches' above.)
The different strategies include:
•Conservative management with supportive care alone (see 'General measures' above)
•Pharmacologic closure using cyclooxygenase (COX) inhibitors including indomethacin, ibuprofen, and acetaminophen (paracetamol) (see 'Directed pharmacologic therapy' above)
•Percutaneous transcatheter occlusion
•Surgical ligation (see 'Surgical ligation' above)
●In our center, a step-wise management approach for infants with hemodynamically significant PDA typically begins with conservative supportive care. Pharmacologic therapy is administered to infants with persistent PDAs who remain dependent on mechanical ventilation after one week of age.
Percutaneous transcatheter occlusion and surgical ligation are reserved for the rare infant who fails pharmacologic therapy and remains on maximal ventilator support. (See 'Management approach' above and "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis", section on 'Hemodynamically significant PDA'.)
•General measures provided to all preterm infants include a neutral thermal environment, moderate fluid restriction between 110 and 130 mL/kg per day, optimizing respiratory support that provides adequate oxygenation (target pulse oximetry saturation [SpO2] between 90 and 95 percent) and allows for permissive hypercapnia (partial pressure of carbon dioxide [PaCO2] between 55 and 65 mmHg as long as pH remains in the normal range [7.3 to 7.4]) at minimal ventilatory settings, and maintenance of a hematocrit above 35 percent. (See 'General measures' above.)
•In infants with a PDA who remain dependent on mechanical ventilation after one week, we suggest a course of COX inhibitors versus supportive care alone (Grade 2B). A second course of COX inhibitor is administered if follow-up echocardiography demonstrates a persistent PDA and the infant remains ventilator dependent. (See 'Directed pharmacologic therapy' above.)
•For infants with large PDAs who remain on maximal ventilator settings and have failed to respond to COX inhibitor therapy, percutaneous catheter occlusion or surgical ligation are options for PDA closure. In our center, percutaneous catheter occlusion is preferred as the requisite expertise and experience are available. In other centers, surgical ligation may be the preferred procedure. (See 'Percutaneous transcatheter occlusion' above and 'Surgical ligation' above.)
●We recommend not to administer prophylactic COX inhibitors versus early pharmacologic treatment of a significant PDA (Grade 1B). Prophylactic therapy would unnecessarily expose infants who would not develop significant PDAs to drugs with potentially serious adverse effects. (See 'Prophylactic/early therapy' above.)
●Mortality is increased in infants with a persistent PDA, especially in extremely preterm infants with a gestational age (GA) 28 weeks or less. (See 'Outcome' above.)
