PDA in Prematurity: Rethinking a Decades-Old Debate in 2025

Echocardiographic evaluation in our case may reveal left-to-right shunting from the aorta into the pulmonary artery. Pulsed-wave Doppler interrogation of the ductus often demonstrates a pulsatile flow pattern, characterized by a systolic peak followed by a diastolic decline. This waveform indicates the absence of a significant constriction along the ductal lumen. In cases of a non-restrictive ductus, low-velocity flow reflects minimal pressure gradient between the aorta and pulmonary artery, suggesting near-equal pressures and free transmission across the ductal segment. This pressure equalization results in elevated pulmonary blood flow and pressure—physiologic conditions for which we are often worried that the immature pulmonary vasculature is ill-equipped. During fetal life, the large and patent ductus allows aortic and pulmonary arterial pressures to approximate, but pulmonary blood flow remains low due to the high pulmonary vascular resistance (PVR). After birth, in a preterm infant with a persistently large PDA, the pulmonary vasculature prematurely experiences systemic pressures and high flow—conditions that deviate from the expected developmental trajectory. Excessive pulmonary blood flow through a PDA may lead to pulmonary overcirculation, increasing the risk of alveolar flooding and pulmonary edema. The resultant ventilation-perfusion (V/Q) mismatch is thought to compromise gas exchange, often prompting clinicians to escalate ventilatory support. However, increased mechanical ventilation can induce further lung injury and inflammation, setting up a cycle of worsening respiratory failure. Inflammation on its own triggers mechanisms that may maintain the PDA open and unrestrictive, which is assumed by many to create a vicious cycle. This pathophysiologic sequence—initiated by ductal shunting and compounded by iatrogenic injury—has historically led to the belief that the PDA is a causative factor in the development of chronic lung disease. While the causality remains debated, the hemodynamic impact of a significant PDA in the fragile preterm lung warrants careful consideration.

As preterm infants progress through the first few days of life, pulmonary vascular resistance (PVR) gradually declines. In contrast, systemic vascular resistance (SVR) tends to remain relatively high. In the presence of a PDA, this imbalance facilitates left-to-right shunting, with augmented pulmonary blood flow occurring at the expense of systemic perfusion—a phenomenon often described as a "steal" effect. The pulmonary vasculature must accommodate this excess flow, resulting in increased pulmonary venous return to the left atrium (LA). If the LA is unable to decompress—typically due to the absence of a large interatrial communication such as an atrial septal defect (ASD)—it will begin to dilate. Due to the stiff immature left ventricle and its inability to accomodate progressively increased preload, there is a concern that this creates a potential for a post-capillary congestion phenomenon, further worsening lung compliance and alveolar fluid spillage. While most preterm infants possess a small patent foramen ovale (PFO), this is usually insufficient to alleviate the hemodynamic burden of a significant shunt. One of the hallmark echocardiographic markers of a echocardiographically-significant PDA is LA distension. The LA-to-aortic root ratio (LA/Ao), classically measured using M-mode in the parasternal long (or short) -axis view, was described by Dr. Norman Silverman. A ratio exceeding 1.4 is considered suggestive of LA dilation secondary to LA overload. However, it is important to recognize that this marker evolves over time. On day one of life, when PVR remains elevated, left-to-right shunting is limited, and LA size may still appear normal. LA distension typically becomes evident several days after birth as pulmonary blood flow progressively increases. As the LA dilates, a corresponding increase in preload occurs in the LV, leading to LV chamber enlargement. Chronic volume loading may result in annular dilation of the mitral valve and subsequent functional mitral regurgitation. Mitral insufficiency can decrease effective ejection fraction by diminishing the forward fraction of blood flow and possibly further worsening LA pressure and post-capillary hypertension. In rare cases, progressive aortic root dilation may occur, potentially contributing to aortic insufficiency—although this latter finding is less frequently emphasized in clinical practice. Severe aortic insufficiency can lead to decreased LV output and contribute to systemic steal. When interpreting an elevated LA/Ao ratio, clinicians should remember it is a ratio: an increase may result from LA enlargement, aortic root hypoplasia, or both. In the presence or absence of a PDA or other left-sided volume-loading lesion, differential diagnoses such as aortic arch hypoplasia, coarctation of the aorta, or a bicuspid aortic valve should also be considered, as these conditions may reduce aortic dimensions, artificially inflating the LA/Ao ratio.

Color Doppler interrogation across the inter-atrial shunt can reveal restrictive interatrial flow from left to right. Pulse wave Doppler placed through the inter-atrial shunt may demonstrate elevated velocities (1.2–1.4 m/s), consistent with a significant left-to-right atrial pressure gradient. This gradient typically reflects elevated LA pressure, especially during ventricular systole when the atrioventricular valves are closed and atrial filling is maximal. The LV internal diastolic diameter will increase with LV dilatation, as preload increases. In the apical four-chamber view, further signs of volume overload may be appreciated. Both the LA and LV appear dilated, and pulmonary veins returning to the LA may exhibit continuous high-velocity flow, consistent with increased pulmonary venous return. This is only true for those with compliant left atriums (or decompressed left atriums by an unrestrictive inter-atrial shunt), as high LA pressure may artificially decrease the pulmonary venous velocities, acting as an obstructive mechanism impeding pulmonary venous return. Indeed, pulmonary veins, while compliant, have limited capacity to accommodate extreme increases in venous return. When exposed to elevated flow—akin to "Niagara Falls through a straw"—Doppler recordings may demonstrate exaggerated D-wave velocities. These flow profiles typically include an S-wave (systolic forward flow), D-wave (diastolic forward flow), and a small A-wave (atrial contraction). An increase in the D-wave amplitude is a common finding in infants with significant left-to-right ductal shunting. If mitral annular dilation occurs, functional mitral regurgitation may be evident on color Doppler, often correlating with the severity of the shunt and chamber remodeling. The hemodynamic consequences of increased pulmonary venous return to the left heart are well explained by the Frank-Starling mechanism. As preload to the LV rises, the myocardial fibers stretch, enhancing contractile force and resulting in increased stroke volume. In preterm infants with significant left-to-right shunting, this manifests as elevated left ventricular output, frequently reflected in higher velocities across the left ventricular outflow tract (LVOT). In some cases, calculated cardiac outputs may exceed 300–400 mL/kg/min, a compensatory response driven by augmented preload. In the presence of a significant left-to-right PDA, excess pulmonary blood flow results in elevated LA pressure, reducing the LV-LA pressure gradient needed for mitral valve opening. When the mitral valve opens, the E-wave velocity increases, reflecting a rapid "gush" of blood into the LV due to high preload (high LA pressure). This augmented early diastolic filling can mimic restrictive physiology in severe cases, characterized by shortened IVRT, increased E/A ratio, and shortened deceleration time due to elevated LA pressures.

During the Middle Ages, bloodletting was universally regarded as an appropriate treatment for the plague. Those who questioned the practice—without knowledge of microbiology or the existence of bacteria—were dismissed as “plague nihilists.” Today, we know that the plague is caused by Yersinia pestis, a bacterium effectively treated with antibiotics. This historical analogy is strikingly relevant to how we continue to approach PDA management in preterm infants. Despite mounting evidence challenging the efficacy of NSAIDs and acetaminophen, neonatal literature still refers to their use for PDA as “treatment.” However, we must question whether it is scientifically sound to call this a treatment when the available pharmacologic agents are largely ineffective at closing ducts (even despite targeted interventions), and at improving outcomes. The PDA, even if considered pathological in extremely preterm infants, is not adequately addressed by these medications because they fail to consistently achieve closure. They are toxic. Their mechanisms of action do not reliably result in ductal closure and when they do, it is possibly in patients whose ducts would have closed spontaneously or in those at lower risk. Their inefficacy is particularly striking when considering their failure to mitigate the very adverse outcomes we seek to prevent, while introducing measurable and unmeasurable toxicity. If a medication existed that could achieve ductal closure universally, with minimal adverse effects, it would represent a true breakthrough. But what we have today resembles the bloodletting of the past—interventions universally labeled as “treatment” despite compelling evidence to the contrary. The insistence on framing NSAIDs and acetaminophen as viable treatments for PDA persists, even as their shortcomings become increasingly evident. We must transcend outdated paradigms and actively pursue novel strategies—whether through innovative pharmacologic agents, mechanical interventions, or conservative management approaches tailored to clinical context. Continuing to rely on ineffective or harmful interventions stifles progress and perpetuates an unscientific status quo. Neonates born <26 weeks GA unsurprisingly require the longest duration to achieve spontaneous ductal closure, with a median of 71 days, reflecting their embryological programming to close the ductus only upon reaching a particular level of maturity that takes longer to achieve. Additionally, infants born between 23 to 24 weeks GA are more likely to exhibit echocardiographic markers of left ventricular volume overload due to PDA, compared to their 25-28 weeks GA counterparts. 

Physiologic Consequences and Disease Associations

In extremely preterm infants (e.g., 23–24 weeks’ gestation), clinical interventions such as intubation and surfactant administration can precipitate a rapid drop in pulmonary vascular resistance (PVR), resulting in a surge of blood flow into the pulmonary circulation. Immature pulmonary capillaries, ill-equipped to accommodate such abrupt increases in flow, may rupture—culminating in pulmonary hemorrhage. This may particularly true if there is heterogenous vasodilation in the areas that the surfactant reached (ventilated), with some areas still vasoconstricted, limiting the restribution of blood flow to areas with already lower vascular resistances. Similarly, rapid hemodynamic transitions can adversely affect splanchnic circulation. Intestinal perfusion instability has been implicated in the pathogenesis of necrotizing enterocolitis (NEC) and spontaneous intestinal perforation (SIP). Furthermore, fluctuating cerebral blood flow, particularly in the context of impaired autoregulation, may increase the risk of intraventricular hemorrhage (IVH). Chronic underperfusion of the gastrointestinal tract during diastole, potentially driven by a significant PDA steal, may exacerbate gut injury, even in the presence of partial autoregulatory mechanisms. These are all speculative and arguments brought forward for decades to justify the search of strategies to accelerate ductal closure.  The cumulative impact of these hemodynamic perturbations—superimposed upon an already vulnerable developmental milieu— has been argued by many tenants of aggressive ductal closure strategies to contribute to the evolution of chronic pulmonary disease (bronchopulmonary dysplasia, BPD), retinopathy of prematurity (ROP), growth failure, pulmonary hypertension, and white matter injury. In the most severe cases, these complications can culminate in death. From a physiological standpoint, there is a compelling rationale that a persistently open duct may exacerbate or even precipitate these conditions.

Developmental Regulation of Ductal Patency and Closure

The high prevalence of PDA in extremely preterm infants is not coincidental; it reflects their developmental immaturity. In fetal life, ductal patency is maintained by a combination of elevated circulating prostaglandins—produced by both the placenta and the ductal endothelium—and relatively low oxygen tension. As gestation progresses, the ductal endothelium becomes less responsive to prostaglandins, facilitating closure. Additionally, the anatomical substrate required for constriction—namely, a well-developed medial muscle layer—matures with advancing gestational age. At 23 weeks’ gestation, this muscular component is often poorly developed, rendering the duct relatively unresponsive to stimuli for closure, even in the presence of elevated oxygen levels. 

The association between PDA and neonatal morbidities is further complicated by the multifactorial nature of many of these conditions. NEC, for example, arises from a confluence of factors including gut immaturity, dysbiosis, feeding practices, inflammatory signaling, placental pathology, and infection. Similarly, BPD results from the cumulative interplay between lung immaturity, oxygen toxicity, mechanical ventilation, and inflammation. Timing also introduces complexity. Nearly all preterm infants exhibit a PDA on the first day of life, yet a significant proportion—especially those at higher gestational ages—will experience spontaneous closure within the first several days. By day three, only approximately 70% of infants still have a PDA, with closure rates increasing as gestational maturity advances. This raises critical questions: If the pathogenesis of conditions like IVH occurs early, and that of BPD is cumulative over time, when—if at all—should we intervene on the ductus? Is there an optimal therapeutic window, or should management strategies be tailored based on specific risk profiles and timing of disease onset? These uncertainties underscore the challenges that researchers have been attempting to address.

Nonsteroidal anti-inflammatory drugs (NSAIDs) or acetaminophen have been used for their property in modulating the prostaglandins expression or through the effect on peroxydase. There may be even some speculation that these medications acutely increase pulmary vascular resistance, as they are vasoconstrictor, which mitigates the trans-ductal shunt and promotes its closure. However, evaluating the impact of these interventions requires careful consideration of their systemic effects, the multifactorial nature of neonatal complications, and the confounding variables inherent to extreme prematurity. More importantly, their efficacy in both achieving closure in the desired patients and their capacity to mitigate the assumed complications of the duct are equally important. 

Intraventricular Hemorrhage (IVH): IVH in preterm infants arises from a convergence of antenatal and postnatal risk factors, including lack of antenatal corticosteroid exposure, outborn status, suboptimal resuscitation, absence of delayed cord clamping, and environmental instability at delivery. Interestingly, prophylactic administration of indomethacin has been associated with a reduction in the incidence of severe IVH. However, the benefit is unlikely to stem from ductal closure alone. NSAIDs are potent systemic vasoconstrictors, exerting effects beyond the ductus arteriosus. Their influence extends to cerebral, renal, pulmonary, and mesenteric circulations. These drugs can significantly alter cerebral blood flow dynamics, potentially compromising perfusion to vulnerable brain regions (PMID 22222547 and 1972434). This may occur for a prolonged period of time after exposure. Studies dating back to the 1990s demonstrated a sustained decrease in cerebral oxygenation following indomethacin administration, reflecting impaired cerebrovascular autoregulation and oxygen delivery. Moreover, NSAIDs may reduce coronary artery flow (PMID: 21791936), raising concerns about myocardial perfusion in a population already vulnerable to ischemia. The risk of acute kidney injury, evidenced by rising serum creatinine, and mucosal ischemia resulting in gastrointestinal perforation, are additional complications tied to their systemic vasoconstrictive profile.

Pulmonary Hemorrhage: While there has been concern that the left-to-right shunt from a PDA could overwhelm the fragile pulmonary vasculature and contribute to pulmonary hemorrhage, clinical trials have not shown that you can prevent pulmonary hemorrhage throughout hospitalisation by aggressively attempting ductal closure. Rather, the hemorrhage is frequently the result of fragile pulmonary vasculature exposed to abrupt shifts in hemodynamic forces—particularly following interventions such as surfactant administration. Surfactant therapy, while essential for improving lung compliance and oxygenation, leads to a sudden drop in pulmonary vascular resistance (PVR). In the context of immature capillary networks, this rapid influx of blood can cause vascular rupture, even in infants without a PDA.

• • An Australian randomized controlled trial that employed early echocardiographic screening attempted to assess the benefit of early treatment. Although it planned to enroll over 400 infants, the study was terminated early due to low recruitment and was underpowered for definitive outcome analysis. The trial recruited infants less than 29 weeks gestation and did found that early indomethacin reduced pulmonary hemorrhage within the first 72 hours (1% vs. 21%). Notably, the trial reported no significant reduction in pulmonary hemorrhage rates with early targeted (by echocardiography) treatment in the first 24 hours of life (overall rate over the study period (9% vs. 23%, p=0.07)) - PMID: 24317704. This observation suggests that the early reduction in PH might be due to the direct pulmonary vasoconstrictive effect or anti-angiogenic properties of NSAIDs rather than ductal constriction itself. The strikingly high pulmonary hemorrhage rate in the control group also makes you wonder if this finding was by chance (23%). Also it raised questions regarding other contributing factors—such as intubation and surfactant administration (which can acutely drop pulmonary vascular resistances)—rather than the PDA itself. In comparative, in a cohort exposed to a strictly conservative management policy (PMID: 34815521), pulmonary hemorrhage occured in 21/280 infants (8%), an incidence below the reported one in the placebo group of the Australian trial and similar to their indomethacin group.

  • The Trial of Indomethacin Prophylaxis in Preterms (TIPP) found no difference in PH incidence between infants receiving prophylactic indomethacin and those receiving placebo (15% vs. 16%).

  • The TRIOCAPI trial reported that infants treated with ibuprofen had a lower incidence of severe pulmonary hemorrhage during the first 3 days compared to the placebo group (1.8% vs. 7.9%, p=.05). However, no significant difference between treatment groups was observed for overall pulmonary hemorrhage rates in the TRIOCAPI trial. Mortality rates were similar (mortality 20% (ibuprofen) vs 14% (placebo), with a trend to be higher in the ibuprofen group).

  • Similarly, the BeNeDuctus trial reported a low PH rate of 3% in the expectant management group versus 1% in the ibuprofen group, with no significant difference.

  • These findings challenge the hypothesis that the PDA is a primary driver of PH. Instead, PH appears to be influenced by a multitude of other factors, reflecting the complex immaturity and vulnerability of the preterm infant.

Necrotizing Enterocolitis (NEC): NEC is a multifactorial disease, and while systemic diastolic steal from a PDA has been hypothesized to reduce mesenteric perfusion, autoregulatory mechanisms in some preterm infants may mitigate this risk by augmenting systolic ejection and left ventricular output. Nonetheless, pharmacologic agents themselves may contribute to intestinal injury. NSAIDs and acetaminophen (paracetamol) may exert direct gastrointestinal toxicity, particularly via prostaglandin inhibition and vascular constriction during exposure to treaetment. Prostaglandins also play a protective role in maintaining mucosal integrity, and their suppression may predispose to ischemic or inflammatory injury. Additionally, the high osmolarity of oral formulations—especially acetaminophen syrup—may compromise the fragile intestinal mucosa in very low birth weight infants.

Bronchopulmonary Dysplasia (BPD): BPD represents a chronic, multifaceted lung disease shaped by prenatal factors (e.g., chorioamnionitis, IUGR, placental dysfunction) and postnatal insults (e.g., oxygen toxicity, mechanical ventilation, infections, NEC). Isolating the PDA as a singular causal factor in BPD pathogenesis is problematic. While excessive pulmonary blood flow from a significant PDA may theoretically induce pulmonary edema and inflammation, the underlying lung immaturity, impaired alveolar development, and inflammatory milieu are likely primary drivers. In fact, systemic inflammation itself can perpetuate ductal patency by stimulating prostaglandin production and endothelial activation. The impact of postnatal corticosteroids on BPD adds another dimension. Trials such as PREMILOC demonstrated that systemic low-dose hydrocortisone reduces BPD incidence, likely through anti-inflammatory effects. Interestingly, these improvements were accompanied by a decreased need for PDA intervention, highlighting the potential indirect benefit of reducing lung disease severity rather than directly addressing the PDA. The TIPP Trial on indomethacin prophylaxis also demonstrated that: "Rates of BPD in the 2 subgroups - placebo vs indomethacin - without a PDA were 43% (170/391) after indomethacin prophylaxis and 30% (78/257) after placebo (P [interaction] = .015). Logistic regression analysis with adjustment for prognostic baseline factors showed that adverse and independent effects of indomethacin prophylaxis on the need for supplemental oxygen and on weight loss by the end of the first week of life may increase the risk of BPD in infants without PDA." (PMID: 16769377).

Kidney injury: Injury to the developing kidneys may be long-lasting, possibly contributing to the association between BPD and hypertension. Indeed, non-steroidal anti-inflammatory drugs (NSAIDs) such as indomethacin and ibuprofen, which have been widely used to induce PDA closure, are associated with acute kidney injury, including reduced urine output and increased serum creatinine concentrations. Ibuprofen, specifically, can impair renal function for the first month of life in very preterm infants. These medications, known for their potent systemic vasoconstrictive effects, may contribute to local ischemia in organs like the kidneys due to their vasoconstrictive and microcirculatory impacts. While the potential for acute renal changes to lead to later renal impairment or a higher risk of systemic hypertension is not definitively known, NSAIDs can also cause systemic vascular remodeling. Some centers that actively treat the ductus have observed a phenotype of very significant systemic hypertension. This observation prompts the question of whether NSAID treatment might lead to increased renal injury and associated hypertension. Indeed, in centers actively attempting ductal closure, NSAIDs courses are often administered to infants with prolonged PDA patency (and often with increasing dosages. These are typically the infants of lower gestational age or with a higher inflammatory burden, whom are already at increased risk for BPD. These vulnerable infants may develop renal injury secondary to NSAIDs exposure, potentially increasing the risk of systemic hypertension. This outcome has not been well studied in randomized trials; however, if rigorously evaluated, NSAIDs use—particularly repeated courses—might reveal a further worsening of renal and systemic outcomes in this population. A similar concern applies to pulmonary vascular injury (pulmonary hypertension / PH), where NSAIDs-induced toxicity could plausibly exacerbate long-term vascular remodeling and dysfunction, spuriously creating an association between PDA and PH described in the observational studies often conducted in centers actively attempting ductal closure.

Pulmonary Hypertension (PH): Emerging data suggest NSAIDs may have a deleterious effect on pulmonary vascular development (PMID: 36732726). These agents not only vasoconstrict the pulmonary vasculature but also impair angiogenesis and vascular branching, leading to reduced pulmonary vessel density. Preclinical studies in neonatal rodents treated with ibuprofen have confirmed these adverse structural changes. Cyclooxygenase inhibition is known to cause pulmonary vascular constriction and induce pulmonary hypertension (PMID: 1508960, PMID: 27692860). Some experts even have postulated that the effect on ductal closure is possibly related to a transient but sustained rise in PVR, limiting the PVR/SVR ratio and leading to a drop in transductal shunt, allowing for closure. Importantly, observational associations between PDA and PH are subject to confounding (PMID: 36378949, PMID: 38015504). More immature infants are inherently more likely to have a PDA. They are also more inherently likely to develop lung injury andm consequently PH. Similarly, we could associate BPD or PH with other interventions such as total parenteral nutrition (TPN), peripherally inserted central catheters (PICC), or gavage feeding often as they reflect the degree of immaturity rather than causality leading to PH. One of the most striking aspects of studies linking PDA to pulmonary hypertension (PH) is that NSAIDs were routinely used as part of PDA management in these cohorts. As such, the observed increase in PH rates may reflect the vascular toxicity of the treatment itself, rather than the effect of the ductus. In fact, in settings where NSAIDs are not routinely administered, the incidence of PH is comparable—or even lower (PMID: 38898108). Furthermore, PH is frequently diagnosed in infants without a history of a significant PDA, reinforcing the view that bronchopulmonary dysplasia (BPD) and its associated vascular remodeling are more likely to be the primary contributors. One other study reported that "although months of prolonged PDA exposure can cause pulmonary vascular remodeling and pulmonary vascular disease (PVD) in term infants with PDA-congenital heart disease, (they) found no echocardiographic evidence for an association between the duration of PDA exposure and the incidence of late PVD during the neonatal hospitalization in preterm infants with BPD." (PMID: 36804505).

Reassessing Pharmacological Closure Strategies

Current pharmacologic approaches to PDA closure, while grounded in decades of practice, warrant critical re-evaluation. The non-specific effects of NSAIDs and acetaminophen raise important concerns regarding organ perfusion and developmental toxicity. Procedural interventions including surgical ligation and transcatheter closure, are employed in clinical practice. Yet, the overarching question remains: Are these interventions justified solely on the basis of achieving ductal closure, or should their primary goal be to improve meaningful neonatal outcomes? This distinction is critical. If the intervention's sole benefit is anatomical closure of the duct, without demonstrable improvement in morbidity or mortality—and if closure is accompanied by non-negligible adverse effects—then the rationale for intervention becomes tenuous. The true benchmark of success should be whether the intervention leads to a reduction in PDA-associated assumed or postulated complications such as IVH, NEC, BPD, or pulmonary hypertension, and whether it does so without incurring significant harm.

A comprehensive systematic review encompassing 68 randomized controlled trials and more than 4,800 neonates examined the effectiveness of various pharmacological agents—including indomethacin, ibuprofen, and acetaminophen—for PDA closure (PMID: 29584842). The overall closure rate was approximately 67%. However, this finding prompts several critical reflections. First, spontaneous closure of the ductus occurs in a substantial proportion of preterm infants, particularly as gestational age increases. Thus, it is uncertain how many of the 67% of pharmacologically closed PDAs would have closed without intervention. This introduces the risk of overtreatment: exposing a large population of infants to agents with systemic side effects who may not ultimately benefit from the intervention. NSAIDs are known to exert widespread vasoconstrictive effects, including on cerebral, renal, pulmonary, coronary and gastrointestinal vasculature. These effects can contribute to complications such as acute kidney injury, gastrointestinal perforation, pulmonary hypertension crisis and altered cerebral blood flow regulation. Additionally, NSAIDs impair platelet function, increasing the risk of bleeding. Long-term concerns remain regarding their impact on angiogenesis and organ development. There is some evidence suggesting a potential association with long-term hypertension and renal dysfunction, although these outcomes require further investigation. Meta-analytic data, such as the analysis conducted by Benitz and Bhombal (PMID: 28724506), have shown that while NSAIDs are effective in reducing the rate of ductal patency, they do not significantly affect key clinical outcomes. This finding further calls into question whether pharmacologic closure meaningfully alters the disease course of prematurity-related complications.

Results of recent RCTs

Multiple recent randomized controlled trials (RCTs) conducted across Europe and North America have failed to demonstrate meaningful differences in key clinical outcomes between early PDA intervention and conservative management. These findings raise critical questions: Are current PDA treatment strategies targeting the right physiological problem? Are we prioritizing ductal closure over patient-centered outcomes? Well-designed RCTs are structured to minimize biases and balance known and unknown confounders—factors that observational studies cannot fully control. Despite the substantial resources invested in these trials, the consistent lack of benefit across diverse populations suggests that early pharmacological administration for accelerating PDA closure with NSAIDs or acetaminophen may not confer the clinical advantages once assumed.

1.Baby OSCAR Trial

1. • Trial Design: A multi-centre, randomized, placebo-controlled trial investigating early selective treatment of PDA with intravenous ibuprofen in preterm infants. Infants born between 23+0 and 28+6 weeks' gestational age (GA) with a large PDA (defined as >1.5 mm diameter with unrestricted, pulsatile left-to-right shunting) were included. Randomization used minimisation to balance groups based on gestation, respiratory support, inotropes, and PDA size. Permission was included for open-label treatment if needed.

   • Primary Outcome: The composite outcome of death or moderate or severe bronchopulmonary dysplasia (BPD) at 36 weeks' postmenstrual age (PMA).

   • Primary Outcome Results: The trial found no significant difference in the primary outcome between the ibuprofen and placebo groups. The composite outcome occurred in 69.2% of the ibuprofen group versus 63.5% of the placebo group (adjusted Risk Ratio [aRR] 1.09; 95% CI 0.98–1.2). The results showed a trend towards a higher prevalence of BPD in the intervention arm, with adjusted odds ratios tending towards indicating harm.

   • Secondary Outcomes:

     1. Long-term neurodevelopmental and respiratory outcomes at 2 years corrected age showed no differences in survival without neurodevelopmental impairment and respiratory morbidity (based on abstract data presented at PAS 2025).

2. BeNeDuctus Trial

• Trial Design: An international, multicenter, randomized, controlled, noninferiority trial conducted across 17 sites in Europe. Extremely preterm infants (<28 weeks GA) with an echocardiographically confirmed PDA (>1.5 mm diameter and left-to-right shunting) were enrolled between 24 and 72 hours after birth and randomized to receive either expectant management or early ibuprofen treatment. A key feature was the use of clear thresholds for open-label pharmacologic treatment in the expectant management group, allowing for a more rigorous comparison with a true non-intervention control group. The trial aimed to demonstrate noninferiority of expectant management with a margin of 10 percentage points for the primary outcome. This trial had near-zero open label treatment with ibuprofen in the expectant management group. 

• Primary Outcome: A composite of necrotizing enterocolitis (NEC, Bell’s stage ≥ 2a), moderate-to-severe BPD, or death at 36 weeks' PMA.

• Primary Outcome Results: Expectant management was shown to be noninferior to early ibuprofen treatment. A primary outcome event occurred in 46.3% of infants in the expectant management group compared to 63.5% in the early ibuprofen group (absolute risk difference [aRD] −17.2 percentage points; upper boundary of the one-sided 95% CI, −7.4; P<0.001 for noninferiority). The results suggest potential harm associated with early ibuprofen exposure, largely driven by a higher incidence of moderate-to-severe BPD in the ibuprofen group.

• Secondary Outcomes:

  • Moderate-to-severe BPD was significantly higher in the early ibuprofen group (50.9%) compared to the expectant management group (33.3%) (aRD −17.6 percentage points; 95% CI, −30.2 to −5.0).

  • Open-label acetaminophen was administered outside the trial protocol to 25% of the expectant management group and 38% of the early ibuprofen group.

  • A potential difference in the primary outcome according to sex was observed, suggesting a better outcome for expectant management in male infants, but this was a secondary analysis without adjustment for multiplicity.

  • The rate of PDA closure after a first course of ibuprofen was similar to earlier RCTs.

3. TRIOCAPI Trial

• Trial Design: A double-blind, multicenter, randomized, placebo-controlled clinical trial conducted in 11 French neonatal intensive care units (NICUs). Infants born at <28 weeks GA with a large PDA (defined by a specific diameter formula based on postnatal age) identified by echocardiography at 6–12 hours after birth were randomized to receive ibuprofen or placebo by 12 hours of age. Open-label ibuprofen rescue treatment was permitted for prespecified criteria of a hemodynamically significant PDA.

• Primary Outcome: Survival without cerebral palsy (CP) at 24 months corrected age.

• Primary Outcome Results: Early echocardiography-targeted ibuprofen treatment of a large PDA did not change the rate of survival without CP. Survival without CP was 71.3% in the ibuprofen group and 71.6% in the placebo group (aRR 0.98; 95% CI 0.83-1.16, P = .83). A reference group of infants with a small or closed ductus at enrollment had a survival without CP rate of 76.2%.

• Secondary Outcomes:

  • Open-label rescue treatment with ibuprofen was significantly more frequent in the placebo group (62.3%) compared to the ibuprofen group (17.5%). The median age of rescue treatment was similar (4 days in placebo vs 4 days in ibuprofen group).

  • Survival without morbidity at 36 weeks PMA was not significantly different.

  • There was no significant difference in the duration of mechanical ventilation, noninvasive ventilation, or oxygen delivery.

  • The trial did not find a difference in the risk of high-grade IVH. Non-significant trends towards increased rates of high-grade IVH and isolated gastrointestinal perforations were observed in the ibuprofen group (p=0.20 and 0.11 respectively).

4. El-Khuffash Trial (The PDA RCT)

• Trial Design: A pilot randomized controlled trial evaluating early targeted nonsteroidal anti-inflammatory drug (NSAID) treatment using a risk-based severity score. It included preterm infants (<29 weeks GA) with a PDA severity score of ≥ 5.0, indicating markers of pulmonary overcirculation and left ventricular diastolic dysfunction.

• Primary Outcome: A composite outcome of chronic lung disease and/or death before hospital discharge.

• Primary Outcome Results: The trial found no significant difference in the primary outcome between the ibuprofen and placebo groups. The composite outcome occurred in 53% of the ibuprofen group and 60% of the placebo group (OR 0.8; 95% CI 0.3–2.1). However, this was a pilot feasibility study non-powered to assess these outcomes. 

• Mortality occured in 8/30 (28%) in the ibuprofen group compared to 4/30 (13%) in the placebo, p=0.33

5. Sung Trial

• Trial Design: A single-center, randomized, double-blind, placebo-controlled, noninferiority clinical trial. It enrolled very preterm infants (GA 23-30 weeks) with hemodynamically significant PDA (ductal size >1.5 mm plus requiring assisted respiratory support) diagnosed between postnatal days 6 and 14. Infants were randomized to receive either oral ibuprofen or placebo. The trial used a per-protocol analysis and included few backup treatments. It aimed to verify the noninferiority of the nonintervention approach with a margin of 20% for the primary outcome.

• Primary Outcome: BPD or death.

• Primary Outcome Results: The nonintervention approach was found to be noninferior to oral ibuprofen treatment in terms of BPD incidence or death. The rates were 44% in the nonintervention group and 50% in the ibuprofen group (risk difference 6%; 95% CI, −0.11 to 0.22; noninferiority margin, −0.2; P = .51). Separate analysis of BPD incidence and mortality as secondary outcomes showed no significant difference between groups.

• Secondary Outcomes:

  • Ductal closure at 1 week after randomization was significantly higher with ibuprofen (20%) than nonintervention (4%) overall (P = .003), specifically in the GA 27-30 weeks subgroup (34% vs 7%, P = .007), but not in the GA 23-26 weeks subgroup (8% vs 2%, P = .34).

  • Ductal closure rates before hospital discharge were not significantly different (ibuprofen 89% vs nonintervention 82%, P = .27).

  • The incidence rates of other adverse outcomes, including severe IVH, ROP stage ≥3, and NEC stage ≥IIb, were not significantly different between the study groups.

6. Trial of Indomethacin Prophylaxis in Preterms (TIPP) Trial

• Study Population and Intervention: Multicenter, randomized, double-blind, placebo-controlled trial enrolled 1202 extremely low birth weight infants (500 to 999 g) soon after birth. Infants were randomly assigned to receive either indomethacin (0.1 mg/kg intravenously once daily for three days) or an equivalent volume of normal saline placebo.

• Primary Outcome: The primary outcome was a composite of death, cerebral palsy, cognitive delay, deafness, and blindness at a corrected age of 18 months.

  • Result: Indomethacin prophylaxis did not improve the rate of survival without neurosensory impairment. The rate of the composite primary outcome was similar between the indomethacin group (47%) and the placebo group (46%). This result held even after adjusting for prespecified baseline characteristics. There was also little evidence that indomethacin prophylaxis altered the rates of any of the individual components of the primary outcome.

• Secondary Outcomes: Numerous secondary outcomes were assessed, including long-term outcomes at 18 months corrected age (hydrocephalus, seizure disorder, microcephaly) and short-term outcomes during the initial hospitalization (patent ductus arteriosus, pulmonary hemorrhage, chronic lung disease, intraventricular hemorrhage, necrotizing enterocolitis, retinopathy of prematurity).

  • Indomethacin prophylaxis significantly reduced the incidence of PDA (24% vs. 50% in the placebo group; odds ratio, 0.3; P<0.001).

  • Severe Periventricular and Intraventricular Hemorrhage (Grade 3 or 4): The risk of grade 3 or grade 4 hemorrhages was lower in the indomethacin group (9% vs. 13%; odds ratio, 0.6; P=0.02). Indomethacin prophylaxis, but not treatment, is associated with a reduced risk of IVH or IVH > grade II, though this does not translate to better long-term neurodevelopmental outcomes. This may be by vasoconstrictive mechanism which may on its own create cerebral hypoxia, injury and long-term concerns. 

  • Chronic Lung Disease (defined as need for supplemental oxygen at 36 weeks PMA): Incidence rates were similar (52% in both groups).

  • NEC, ROP, Hydrocephalus with shunt, Seizure disorder, Microcephaly) were similar between groups. 

7. PDA-TOLERATE (PDA: TO Leave it alone or Respond And Treat Early) Trial

• Study Population and Intervention: This was an exploratory, multicenter, randomized controlled trial that enrolled preterm infants (<28 weeks gestational age) who had a moderate-to-large PDA shunt. Enrollment occurred between 6 and 14 days of age (mean age 8.1 ± 2.2 days) to allow for spontaneous ductal closure. Infants were randomized to receive either early routine pharmacologic treatment (ERT) or a conservative treatment (CT) approach where treatment was only given if prespecified respiratory and hemodynamic rescue criteria were met. Pharmacologic treatment in the ERT group included indomethacin, ibuprofen, or acetaminophen, with indomethacin backup if needed, based on center preference.

• Primary Outcome: The primary outcome was the composite of need for ligation or presence of a PDA at discharge.

• Result: There was no significant difference between the ERT group (32%) and the CT group (39%) in the primary outcome of ligation or presence of a PDA at discharge. There were no significant differences between the groups for BPD, death, or the combined outcome of BPD/death. Early routine treatment was associated with delayed achievement of full feeding. Early routine treatment may increase the risk of late-onset sepsis and death in infants ≥26 weeks gestational age. These findings were from post hoc analyses.

• Limitations and Context for PDA-TOLERATE: The trial was small and exploratory, limiting its power. A high rate of rescue treatment in the conservative arm (48% treated at a median of 12 days post-randomization) and failure of treatment in the ERT arm to close all PDAs minimized the exposure difference between groups, biasing results toward the null. The trial enrolled infants later (after the first week), so it could not assess potential benefits of earlier treatment.

Reflecting Real-World Practices and Entry Criteria

One criticism leveled against recent PDA trials, such as BeNeDuctus and Baby OSCAR, is their reliance on ductal diameter as a primary entry criterion. While diameter alone is acknowledged as an imperfect indicator of the full hemodynamic significance of a PDA (as outlined by the complexity of these evaluations in the sectios above), there are several reasons why the approach used in these trials does reflect common clinical real-world practice for initiating treatment:

• Bedside echocardiography is routinely used to define the hemodynamic significance of PDA. A ductal diameter greater than 1.5-2.0 mm is commonly considered a significant PDA in most centers. Recent studies show that clinical practices in NICUs managing extremely preterm infants frequently involve treating a significant percentage of these babies (e.g., 54-73% of infants <27 or <32 weeks, 62-66% of those <25 or <26 weeks in the Canadian Neonatal Network - PMID: 40450478). The criteria used in these trials yield to similar rates of exposure than those described of current practices in most NICUs managing extremely preterm infants who use pharmacological interventions.

• Beyond Diameter Alone: Importantly, studies like Baby OSCAR did not rely solely on diameter. Baby OSCAR required a PDA diameter of at least 1.5 mm with unrestricted, pulsatile left-to-right shunting. This inclusion of flow pattern assessment makes the criteria more nuanced than just a simple diameter measurement. Source highlights Baby OSCAR as one of the only studies providing a clear definition of what they considered "unrestrictive flow".

• Magnitude of Shunt: The median PDA diameter in trials like BeNeDuctus was often well above the minimum 1.5 mm criterion, reported as a mean 2.1 mm. This indicates that the trials were enrolling infants with substantial shunts commonly targeted for treatment in clinical settings.

  1. BeNeDuctus: The trial reported a median PDA diameter of 2.1 mm at the time of the eligibility echocardiography. 

  2. Baby OSCAR Trial: The entry criteria for the Baby OSCAR trial required a PDA diameter of at least 1.5 mm with unrestricted, pulsatile left-to-right shunting. For the babies included in the Baby OSCAR trial, it is notably reported that 75% of patients had a diameter of 2 mm or greater. The median diameter in this trial was 2.2 mm. Interestingly, 480 infants had a diameter above 2 mm vs 166 infants with a diameter between 1.5 to 2 mm.

Representation of Extremely Immature Infants (<26 Weeks GA)

Another criticism is that PDA trials may not have included enough extremely immature infants, particularly those born before 26 weeks gestation. However, recent trials and subsequent analyses have made significant efforts to address this:

• Recent trials, including Baby OSCAR and BeNeDuctus, enrolled a substantial number of extremely preterm infants. Baby OSCAR, for example, specifically reported outcomes in over 300 babies born below 26 weeks gestation (about half of their cohorn were less than 26 weeks). Recent meta-analyses including trials conducted since 2010 found that these studies collectively included 707 infants born before 26 weeks gestation. According to the supplemental data from the BeNeDuctus trial, a total of 56 infants (<26 weeks GA) were in the Expectant Management group and 64 infants (<26 weeks GA) were in the Early Ibuprofen group at randomization, totaling 120 infants less than 26 completed weeks gestation. The subgroup analysis for infants born less than 26 weeks gestation at birth showed results for the primary outcome (composite of NEC, moderate to severe BPD, or death) that were consistent with the overall findings. The subgroup analysis for the primary outcome in infants born less than 26 weeks GA showed an absolute risk difference of -21.2 percentage points (95% CI -10.5). This indicates that in this subgroup, the primary outcome occurred 21.2 percentage points more frequently in the early ibuprofen group compared to the expectant management group, with the confidence interval strongly suggesting an increased risk (or harm) rather than benefit. 

• Recent analyses pooling data from multiple trials that include substantial numbers of extremely preterm infants also failed to show a benefit and suggested harm from pharmacological PDA treatment in this population. An analysis of seven randomized trials conducted since 2010 with low rates of open-label treatment, which collectively included 707 infants born before 26 weeks gestation, found that early treatment with NSAIDs to close persistent PDA increased mortality (pooled odds ratio 1.33, P = 0.05). This analysis also suggested increased risk for other morbidities like sepsis and PVL, and a trend towards increased BPD, pulmonary hemorrhage, and severe IVH. (PMID: 38918573)

• One review by Gupta S. et Donn SM. (PMID 39550255) described that data pooled from the El-Khuffash RCT, TRIOCAPI study, BeNeDuctus trial, and Baby OSCAR trial revealed a hazard ratio of 1.39 (95% CI: 1.05–1.78) for death by 36 weeks’ postmenstrual age in infants exposed to ibuprofen compared to those managed with placebo or expectant management.

• One meta-analysis from 2025 (PMID: 40423988) included 10 randomized clinical trials involving 2035 preterm infants born before 33 weeks of gestation comparing active treatment (pharmacologic or surgical, though primarily pharmacologic within the first 2 weeks of life in the included trials) with expectant management for hemodynamically significant PDA. Active attempt to accelerate closure of a PDA during the first 2 weeks of life was associated with potential harm. Indeed, there was significantly higher rates of mortality (15.5% vs 12.4%), with an odds ratio of 1.25 (95% CI, 1.01-1.56; P=.04). The meta-analysis also found a significantly worse composite outcome of death at 36 weeks’ postmenstrual age or at discharge or moderate to severe bronchopulmonary dysplasia (56.2% vs 50.8%) (P=.009). The results were consistent in subgroup analysis for more immature infants born at less than 29 weeks gestation, where death was also significantly higher in the active treatment group (16.2% vs 12.0%). The authors conclude that these findings suggest that an expectant management approach to PDA in preterm infants may be associated with a better morbidity and mortality profile. 

• One Bayesian Meta-Analysis (PMID 39815089) included five RCTs. The study found strong evidence in favor of the harmful effect of medications regarding the outcome of BPD and for the composite outcome of BPD or death. When pooling the two largest trials (Baby-OSCAR and BeNeDuctus), the analysis showed moderate evidence in favor of higher mortality in the medication group. For the subgroup of infants ≤26 weeks GA, pooling data from Sung et al. and BeNeDuctus showed moderate evidence (Bayes factor +/- = 0.14) in favor of a higher rate of BPD or death in the treatment group. Including Baby-OSCAR for this subgroup also showed moderate evidence for a higher rate of BPD or death with medications exposure. 

Long-term follow-up data presented at scientific meetings for both the Baby OSCAR (PAS 2025) and BeNeDuctus trials (ESPR) demonstrated no significant difference in survival without neurodevelopmental impairment (NDI) at 24 months of age. These findings are consistent with results from the TRIOCAPI trial and an observational study (PMID: 38278962) comparing outcomes between infants managed under a strict conservative policy and those treated with ibuprofen to accelerate ductal closure.

Rethinking the Strategy for PDA in Prematurity

The accumulating evidence from recent RCTs highlights a paradigm shift in PDA management. Early pharmacologic closure—once considered a standard preventive strategy—is increasingly unsupported by data. These trials suggest that routine early treatment may not improve—and may in some cases worsen—key outcomes such as mortality (arguably the most critical outcome) and bronchopulmonary dysplasia (BPD). While BPD has its own limitations as an outcome measure and may not fully reflect what is most meaningful to families and patients in the long term, an increase in its incidence is nonetheless undesirable. Importantly, these interventions have not demonstrated any consistent neurodevelopmental benefit. Given the potential for harm and the lack of demonstrated efficacy, the default approach should no longer be automatic early closure. Expectant and conservative management may offer a safer alternative for many preterm infants, sparing them from unnecessary exposure to drugs with systemic toxicity and unclear long-term consequences.

Experience with Conservative Management: Observational Evidence and Evolving Practice

A retrospective review of local data from 2015 to 2019 provides interesting information regarding outcomes in a center with a conservative PDA management policy in extremely preterm infants (PMID: 34815521). This cohort, which was not exposed to NSAIDs or acetaminophen for PDA treatment, demonstrated high rates of spontaneous ductal closure. Specifically, approximately 90% of infants—regardless of whether they were below or above 26 weeks’ gestation—achieved closure without pharmacologic or procedural intervention. These findings reinforce that even in the most immature infants, spontaneous closure is not only possible but common. Importantly, this conservative approach coincided with favorable clinical trends. Over the same period, a marked reduction in the composite outcome of death or bronchopulmonary dysplasia (BPD) was observed. Mortality declined significantly, while BPD rates showed a downward trend—from 85% to 59% in infants under 26 weeks’ gestation, and to 28% in those under 29 weeks. At 36 weeks’ corrected age, few infants remained ventilator-dependent. The vast majority required only low-flow nasal cannula or CPAP, with invasive ventilation generally limited to those referred with primary airway disease. The unit ceased routine use of NSAIDs in 2013, citing concerns over toxicity and uncertain efficacy. In 2014, bubble CPAP became the primary modality for noninvasive respiratory support (PMID 38887808). These practice changes correlated with a decrease in early intubation, BPD severity, and death from severe BPD. These positive and significant trends occurred alongside—and perhaps partly because of—the adoption of a strict conservative management policy of the PDA. Notably, they were observed despite a proportional increase in the number of infants born at 23–24 weeks’ gestation and the inclusion of more fragile infants with birth weights near 500 grams, who are typically at higher risk of inflating center-specific rates of adverse outcomes. While the data are observational and multifactorial influences—such as improved nursing protocols, updated surfactant strategies, and transition to a new NICU facility—likely contributed, the key insight is that cessation of NSAIDs use did not result in increased rates of BPD or mortality. On the contrary, outcomes improved.

Acetaminophen (Paracetamol): A Limited and Possibly Harmful Alternative

Acetaminophen has emerged as a pharmacologic alternative to NSAIDs for PDA closure in many centeres. However, its efficacy remains limited. The PDA Tolerate study, a randomized controlled trial conducted primarily in the United States, found that acetaminophen achieved ductal closure in only 27% of patients—despite multiple days of therapy. The effectiveness was particularly poor in the most immature infants. Safety concerns surrounding acetaminophen are significant and less well understood than for NSAIDs. In adults, hepatic toxicity is well characterized and results from saturation of hepatic metabolism via cytochrome P450 enzymes. In neonates, however, hepatic cytochrome activity is minimal, and high P450 expression occurs in the lungs (PMID 37169914, PMID 33617854, PMID 39836061). This unique enzymatic profile raises the possibility that high-dose acetaminophen may exert localized pulmonary toxicity by overwhelming detoxification pathways and generating oxidative stress. Indeed, some experimental and observational studies have raised concerns that acetaminophen may suppress protective anti-inflammatory responses in the lung, potentially promoting BPD-like pathology. Additional safety concerns include the risk for cholestasis, and osmolarity of oral formulations. Acetaminophen syrup is highly hyperosmolar, and its use in preterm infants has been associated with gastrointestinal injury, including spontaneous intestinal perforation—although this remains speculative and largely based on case reports and animal models.

Procedural Closure: Surgical and Transcatheter Techniques

Procedural closure options include surgical ligation and catheter-based interventions. While these methods offer definitive anatomical closure, they carry substantial risks, and the timing of intervention is critical. Procedural closure, often performed late (e.g., at three weeks of life), and are unlikely to alter early complications such as IVH or NEC. Conversely, performing these procedure early (e.g., at two days of life in a 23-week infant) involves significant risks. Post-ligation syndrome, acute cardiorespiratory instability, and inflammatory responses are well-documented. Long-term complications may include recurrent laryngeal nerve injury, diaphragmatic dysfunction, scoliosis, vocal cord paralysis, and (possibly) adverse neurodevelopmental outcomes (although there might be a counfounder by indication). Associations with BPD and retinopathy of prematurity (ROP) have also been reported, though causality is difficult to establish. One RCT (PMID 2498657) investigated a very aggressive approach of prophylactic ligation of the PDA in infants born at 1000 grams or less. The trial enrolled 84 infants. Infants in the intervention group underwent surgical ligation of the ductus arteriosus on the day of birth. Infants in the control group received standard medical therapy and no medical or surgical intervention to close the ductus unless congestive heart failure or persistent hypotension attributable to the PDA developed. Rescue ligation was considered only if infants manifested cardiopulmonary compromise. The trial found no significant differences between the ligation group and the control group in the primary combined outcome of death or neurodevelopmental impairment. However, the study reported a significantly lower incidence of necrotizing enterocolitis (NEC) in the ligation group (8%) compared to the control group (30%). This is an odly high number for NEC (30%). The NEC (stage 2 or more) rate in the previously mentioend conservative cohort of extremely preterm patients <29 weeks was only 8% (n=17/214), corresponding to the rate in the ligated group of the Cassidy trial (PMID: 34815521). Further, no significant differences were observed in other major pulmonary or neurologic problems in the original analysis of the Cassidy trial. Despite the original finding of no significant neurologic differences, later analyses and discussions referencing this trial have suggested concerning associations. A re-examination by Clyman and Cassady (PMID 19324366), among others, suggested that surgical ligation compared with medical therapy alone was associated with more neurosensory impairment, BPD, and ROP .   These later interpretations, while acknowledging the reduction in NEC, highlighted the potential harms associated with early ligation regarding respiratory and long-term neurodevelopmental outcomes. In essence, the Cassady trial was one of the first rigorous attempts to evaluate a highly aggressive strategy (prophylactic ligation) for PDA in extremely preterm infants.

Catheter-based PDA closure offers a less invasive alternative but is not without risk—particularly in infants under 1000 grams. Potential complications include vascular injury (e.g., limb ischemia, limb loss, hematoma, vessel rupture), atrial perforation, valvular trauma, device embolization, and mechanical obstruction of adjacent structures (aorta or pulmonary artery). Procedural anesthesia, bleeding, and the requirement for anticoagulation further increase risk. Patient may die during the procedure if there is a ruptured vascular structure or cardiac structure (which has been reported). Critically, no study to date has demonstrated that transcatheter or surgical PDA closure improves clinically relevant outcomes such as survival, BPD, NEC, or neurodevelopment. Given these risks and the current absence of studies demonstrating benefit, the position articulated here is that procedural closure—particularly in fragile, extremely low birth weight infants—should only be undertaken within the context of a research protocol such as the PIVOTAL trial, following thorough parental counseling and informed consent. Offering invasive therapies without evidence of benefit and outside a trial framework is ethically questionable and potentially harmful.

Prioritizing Outcomes Over Closure

Collectively, these data support a more conservative, physiology-driven approach to PDA management in preterm infants. Pharmacological closure—whether via NSAIDs or acetaminophen—carries systemic risks and has not convincingly improved outcomes in randomized trials. Procedural options, while definitive in anatomic terms, introduce significant procedural morbidity with unclear benefit. Spontaneous closure remains the dominant trajectory for the majority of infants, even among those born at the limits of viability (PMID: 34815521, PMID 28701390, PMID: 36672586, PMID: 30759169, PMID 25169243). When conservative management is accompanied by modern respiratory support strategies and meticulous neonatal care, outcomes can improve rather than deteriorate. Fundamentally, the physiology of ductal closure in preterm infants mirrors the developmental processes observed in utero. While more mature infants typically achieve closure earlier, even infants born at 23–24 weeks’ gestation can close their PDA spontaneously over time. This delayed but natural closure trajectory has been documented across multiple cohorts. The future of PDA management should prioritize long-term infant health over short-term anatomical correction, reserving intervention for cases with clear hemodynamic compromise or within well-designed clinical trials.

From: (1) Sung SI, Chang YS, Kim J, Choi JH, Ahn SY, Park WS. Natural evolution of ductus arteriosus with noninterventional conservative management in extremely preterm infants born at 23-28 weeks of gestation. PLoS One. 2019 Feb 13;14(2):e0212256. doi: 10.1371/journal.pone.0212256. PMID: 30759169; PMCID: PMC6374019. (2) Semberova J, Sirc J, Miletin J, Kucera J, Berka I, Sebkova S, O'Sullivan S, Franklin O, Stranak Z. Spontaneous Closure of Patent Ductus Arteriosus in Infants ≤1500 g. Pediatrics. 2017 Aug;140(2):e20164258. doi: 10.1542/peds.2016-4258. Epub 2017 Jul 12. PMID: 28701390.

Other Modalities: Postnatal Steroids and Conservative Management

Supportive, non-interventional management remains a cornerstone of PDA care in many centers, particularly where the emphasis has shifted from routine closure to physiologically guided observation. One area of growing interest is the potential role of postnatal corticosteroids in promoting ductal constriction, either directly or indirectly through modulation of systemic inflammation. A study demonstrated that administration of dexamethasone around day 20 of life in preterm infants was associated with a reduction in ductal diameter, with some infants achieving complete closure (PMID 36068678). Another study found that (PMID 39604779) post-natal betamethasone was associated with ductal constriction in extremely preterm infants and concluded: "In extremely preterm infants with a severe respiratory condition at 3 weeks of life, oral betamehtasone treatment can help wean invasive ventilation and is associated with PDA closure. It could reduce the need for surgical or endovascular treatment that are associated with serious adverse effects."

• Animal Studies: Studies in fetal rats and lambs (PMID: 15731544) have shown that steroid hormones, including dexamethasone, can cause ductal constriction in a dose-dependent manner. Potential mechanisms for how glucocorticoids like dexamethasone influence ductal closure include direct vasoconstriction, interference with prostaglandin (PG) synthesis (as PGE2 maintains patency), or a reduced sensitivity of the ductal muscle to the dilating effects of PGE2. Improvements in lung disease, increased diuresis, and improved oxygen tension may also indirectly facilitate ductal closure. Steroids can modulate inflammation, which may participate in keeping the duct patent. 

• Human Observational Data and Trials:

  • Early dexamethasone therapy reportedly reduced the incidence of clinically detectable PDA in preterm infants weighing less than 1000g at birth who had severe respiratory distress syndrome (RDS). One study reported an incidence of PDA of 23% in the dexamethasone group versus 59% in the placebo group (p=0.05) in infants less than 1000g. This study also found no recurrence of ductal patency after indomethacin therapy in the Dex group, compared to some recurrences in the placebo group. 

  • Some older studies noted a temporal association between commencing dexamethasone therapy and PDA closure. However, one older study on Dex found no consistent closing effect on PDA in the four infants with PDA included in that particular analysis, noting that closure, when seen, often occurred at postnatal ages where spontaneous closure is common. 

  • More recently, a prospective echocardiographic study evaluating low-dose dexamethasone (a 10-day tapering or 'DART' regimen with a cumulative dose of 0.89 mg/kg) administered at a median age of 20 days to 30 extremely preterm infants with lung disease found significant ductal constriction and closure in 35% (7 out of 20 infants with PDA at baseline) (PMID 36068678). It is often noted that dexamethasone is used to facilitate extubation or reduce respiratory support in infants with chronic lung disease. The observed effect on PDA may be secondary to improvements in respiratory status or reductions in inflammation. 

  • The Canadian Paediatric Society recommends considering low-dose dexamethasone for infants who remain ventilated after the first week of life with worsening lung disease. This recommendation reflects the understanding that such infants are likely in a heightened inflammatory state, which may contribute both to persistent lung disease and delayed ductal closure. The DART trial demonstrated that low-dose dexamethasone (cumulative dose ~0.89 mg/kg) may aid in extubation, but this regimen may be insufficient in the context of severe inflammation. In contrast, the Syracuse study, a prospective randomized controlled trial, enrolled 59 infants under 27 weeks with severe lung disease between days 10–20 of life and compared two dexamethasone regimens: a 42-day tapering course (7.6 mg/kg cumulative dose) versus a shorter, 9-day course (4 mg/kg). The longer course was associated with significantly increased intact survival (survival without major brain injury) at school age (PMID 31349916). These findings suggest that more robust anti-inflammatory therapy can have long-term neurodevelopmental benefits, potentially by improving cerebral oxygenation and limiting white matter injury. In one prospective cohort of 392 preterm infants (≤32 weeks GA), low-dose dexamethasone (0.89 mg/kg) was administered after the first postnatal month for evolving BPD (PMID 40360237). MRI at term-equivalent age showed significantly larger cerebellar and subcortical grey matter volumes in treated infants. Dexamethasone was also associated with improved motor outcomes at 2 years corrected age, suggesting a potential neuroprotective effect.  While NSAIDs and acetaminophen target ductal smooth muscle tone directly, corticosteroids may influence PDA physiology through systemic anti-inflammatory effects, potentially restoring the infant’s endogenous mechanisms for ductal closure. In carefully selected cases—particularly those with evolving BPD, ventilator dependence, and inflammatory markers—postnatal dexamethasone may serve a therapeutic role beyond respiratory support, potentially aiding ductal closure and improving long-term outcomes. Nevertheless, spontaneous closure remains a reliable outcome in most infants when supported by evidence-based conservative care strategies. 

  • Betamethasone (BTM): Betamethasone has a vasoconstrictor effect on the PDA. Studies in fetal rats showed increased constriction of the ductus arteriosus with combined administration of indomethacin and betamethasone. Antenatal exposure to betamethasone has been associated with less frequent PDA in infants. The observed effectiveness of betamethasone could be related to its contractile effect on the ductus arteriosus. A retrospective, single-centre study of 51 extremely preterm infants (median GA 25.7 weeks, median age at treatment 28 days) who received oral betamethasone to help wean from invasive ventilation found that BTM treatment was associated with PDA closure in almost all infants (98.0% had a closed or non-haemodynamically significant PDA after treatment). The percentage reduction in PDA diameter was significantly greater in the most immature infants (< 26 weeks GA). The authors of this study suggest that this late oral BTM treatment could reduce the need for surgical or endovascular treatment, which are associated with serious adverse effects.

Declining Treatment Rates and Evolving Conclusions

Over the past decade, PDA treatment rates—encompassing pharmacologic, surgical, and catheter-based approaches—have markedly declined across North America and the world. This shift reflects a growing consensus that current interventions, despite extensive investigation, have failed to demonstrate consistent or meaningful improvements in clinical outcomes (and efficacy in achieving closure). At the same time, they are associated with a nontrivial risk of adverse effects. This evolving practice landscape prompts a fundamental ethical and clinical question: If an intervention for a physiologic condition fails to improve outcomes and introduces harm, should it continue to be used? The weight of current evidence strongly suggests that pharmacologic strategies using NSAIDs, acetaminophen, or early procedural closure do not confer a survival or morbidity benefit in preterm infants. Numerous randomized controlled trials—despite involving thousands of patients and millions of dollars in research funding—have failed to show improvement in critical outcomes such as BPD, IVH, NEC, or neurodevelopmental impairment. On the contrary, these treatments carry significant risks, including renal dysfunction, gastrointestinal perforation, pulmonary toxicity, and impaired vascular development. Given this context, the routine administration of NSAIDs for the sole purpose of ductal closure—especially without explicit parental consent—raises serious ethical concerns. With overwhelming evidence pointing to a lack of benefit and the presence of harm, continuing to use these drugs indiscriminately is difficult to justify. Furthermore, additional randomized trials focused on these same interventions may also be ethically problematic unless they offer new hypotheses or alternative therapeutic frameworks.

A Shift Toward Conservative, Individualized Management

Emerging data increasingly support a conservative, physiology-guided approach centered on supportive care and the avoidance of routine closure. Key elements of this strategy include:

• Non-invasive respiratory support, such as bubble CPAP, which reduces the need for intubation and mitigates lung injury and inflammation;

• Standardized, high-quality neonatal care, including optimized nutrition, infection control, and family-centered support;

• Judicious use of postnatal corticosteroids, particularly dexamethasone, in infants with evolving lung disease and inflammatory markers, where inflammation may play a role in both BPD and PDA persistence.

This approach is aligned with the foundational principle of neonatal care: less is often more. For these fragile and developmentally vulnerable infants, minimizing unnecessary interventions may result in better long-term outcomes. The burden of proof lies not anymore in demonstrating the presence of a PDA, but in showing that closing it improves outcomes without causing harm. At present, the most promising path forward appears to lie not in closing every duct, but in supporting every infant with a thoughtful, individualized approach combining advanced respiratory care and mitagation strategies to limit the shunt (such as mainly permissive hypercapnia which increased the PVR/SVR ratio, and maintaining adequate non-invasve positive end expiratory pressure to allow lung growth and adequate functional residual capacity).

References:

• • Matthew Laughon, Carl Bose, William E. Benitz. 1-s2.0-S0022347610004257-main.pdf [Internet]. [cited 2025 Jun 7]. Available from: https://www.sciencedirect.com/science/article/pii/S0022347610004257

  • M. Ellen Kuenzig, Eric I. Benchimol. 1-s2.0-S0022347620313585-main.pdf [Internet]. [cited 2025 Jun 7]. Available from: https://www.sciencedirect.com/science/article/pii/S0022347620313585

  • William E. Benitz, Valerie Y. Chock. 1-s2.0-S0022347620313585-main.pdf [Internet]. [cited 2025 Jun 7]. Available from: https://www.sciencedirect.com/science/article/pii/S0022347620313585

  • William E. Benitz, Richard A. Polin. Recommitting to Rules of Evidence and Laws of Inference. J Pediatr [Internet]. 2024 [cited 2025 Jun 7]. Available from: https://doi.org/10.1016/j.jpeds.2024.114199

  • S. Spenard, C. Backes, D.A. Fitzgerald, G. Sant’Anna, G. Altit. Current approaches to the patent ductus arteriosus: Implications for pulmonary morbidities. Paediatr Respir Rev [Internet]. 2025 [cited 2025 Jun 7]. Available from: https://doi.org/10.1016/j.prrv.2025.05.001

  • W.E. Benitz, S. Bhombal. The use of non-steroidal anti-inflammatory drugs for patent ductus arteriosus closure in preterm infants. Semin Fetal Neonatal Med [Internet]. 2017 Oct [cited 2025 Jun 7];22(5):302–7. Available from: http://dx.doi.org/10.1016/j.siny.2017.07.004

  • S. Gupta. 1-s2.0-S1744165X24000477-main.pdf [Internet]. 2024 Nov 12 [cited 2025 Jun 7]. Available from: https://doi.org/10.1016/j.siny.2024.101565

  • Chen et al. Vascular and pulmonary effects of ibuprofen on neonatal lung development. Respir Res [Internet]. 2023 Feb 2 [cited 2025 Jun 7];24(1):39. Available from: https://doi.org/10.1186/s12931-023-02342-4

  • Sankar MN, Bhombal S, Benitz WE. PDA: To treat or not to treat. Congenit Heart Dis [Internet]. 2019 [cited 2025 Jun 7];14(46–51). Available from: https://doi.org/10.1111/chd.12708

  • Remy A, Vincent M, Pastor-Diez B, et al. Late postnatal steroid treatment using oral betamethasone can help to close ductus arteriosus in extremely preterm infants who cannot be weaned from ventilation. Eur J Pediatr [Internet]. 2024 Jan 1 [cited 2025 Jun 7];184(1):50. Available from: https://doi.org/10.1007/s00431-024-05840-9

  • Colbert AJ. Medical manipulation of the ductus arteriosus. In: Garson A Jr, Bricker JT, Fisher DJ, editors. The science and practice of pediatric cardiology. 2nd edn. Baltimore: Williams & Wilkins;

  • Maya R. Grayck, Bradford J Smith, Alexander Sosa, Emma Golden, William C. McCarthy, Mack Solar, Natarajan Balasubramaniyan, Lijun Zheng, Evgenia Dobrinskikh, Mercedes Rincon, David J. McCulley. A Single Early Life Acetaminophen Exposure Causes Persistent Abnormalities in the Murine Lung. Am J Respir Cell Mol Biol [Internet]. [cited 2025 Jun 7]. Available from: https://doi-org.proxy3.library.mcgill.ca/10.1165/rcmb.2024-0452OC

  • Heyman E, Ohlsson A, Shennan AT, Heilbut M, Coceani F. Closure of patent ductus arteriosus after treatment with dexamethasone. Acta Paediatr Scand [Internet]. 1990 Jun [cited 2025 Jun 7];79(6–7):698–700. Available from: https://onlinelibrary.wiley.com/doi/10.1111/j.1651-2227.1990.tb11538.x

  • Gupta S, Subhedar NV, Bell JL, et al. Trial of selective early treatment of patent ductus arteriosus with ibuprofen. N Engl J Med [Internet]. 2024 [cited 2025 Jun 7];390(314–25). Available from: https://www.nejm.org/doi/full/10.1056/NEJMoa2305582

  • Benitz WE. Treatment of persistent patent ductus arteriosus in preterm infants: time to accept the null hypothesis? Arch Dis Child Fetal Neonatal Ed [Internet]. 2012 Jan [cited 2025 Jun 7];97(1):F80–2. Available from: http://fn.bmj.com/

  • Altit G. The Controversy Surrounding PDA in Prematurity [Internet]. YouTube; 2024 [cited 2025 Jun 7]. Available from: https://www.youtube.com/watch?v=p50_D2d0P_8

  • Schmidt B, Davis P, Moddemann D, Ohlsson A, Roberts RS, Saigal S, et al. Long-term effects of indomethacin prophylaxis in extremely-low-birth-weight infants. N Engl J Med [Internet]. 2001 Jun 28 [cited 2025 Jun 7];344(26):1966–72. Available from: https://www.nejm.org/doi/full/10.1056/NEJM200106283442602

  • Backes CH, Slaughter JL. Expectant Management of Patent Ductus Arteriosus in Premature Infants. N Engl J Med [Internet]. 2023 Mar 16 [cited 2025 Jun 7];388(11):1033–5. Available from: https://www.nejm.org/doi/full/10.1056/NEJMe2214764

  • Hundscheid T, Onland W, Kooi EMW, et al. Expectant Management or Early Ibuprofen for Patent Ductus Arteriosus. N Engl J Med [Internet]. 2023 Mar 16 [cited 2025 Jun 7];388(11):980–90. Available from: https://www.nejm.org/doi/10.1056/NEJMoa2207418

  • NeoCardio Lab - Thoughts on NSAIDs for PDA [Internet]. [cited 2025 Jun 7]. Available from: https://www.neocardiolab.com/thoughts-on-nsaids-for-pda

  • S. Gupta. PIIS1744165X24000477.pdf [Internet]. 2024 Nov 12 [cited 2025 Jun 7]. Available from: https://doi.org/10.1016/j.siny.2024.101565

  • Lapointe A, Villeneuve A, Hebert A, Nouraeyan N, Lakshminrusimha S, Singh Y, et al. SS May 2025-all_Pulmonary Hemorrhage_GS.pdf [Internet]. [cited 2025 Jun 7]. Available from: https://connects.mcgill.ca/

  • Hundscheid T, Onland W, Kooi EMW, et al. Supplementary Appendix. Expectant management or early ibuprofen for patent ductus arteriosus [Internet]. 2023 [cited 2025 Jun 7]. Available from: https://www.nejm.org/doi/suppl/10.1056/NEJMoa2207418/suppl_file/nejmoa2207418_appendix-data-sharing.pdf

  • A. Sehgal. Cardiorespiratory adaptation to low‐dose dexamethasone for lung disease in. J Physiol [Internet]. 2022 Oct [cited 2025 Jun 7];600(19):4362–73. Available from: https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP282973

  • Benitz WE. Unwinding old habits: deimplementation of treatment regimens for patent ductus arteriosus in preterm infants. J Pediatr (Rio J) [Internet]. 2020 Feb [cited 2025 Jun 7];96(2):138–41. Available from: https://doi.org/10.1016/j.jped.2018.12.003

  • Evans NJ, Archer LNJ. Postnatal circulatory adaptation in healthy term and preterm neonates. Arch Dis Child [Internet]. 1990 [cited 2025 Jun 7];65:24–6. Available from: https://fn.bmj.com/content/archdischild/65/1/24.full.pdf

  • Rozé JC, Cambonie G, Le Thuaut A, et al. Effect of Early Targeted Treatment of Ductus Arteriosus with Ibuprofen on Survival Without Cerebral Palsy at 2 Years in Infants with Extreme Prematurity: A Randomized Clinical Trial. J Pediatr [Internet]. 2021 Jun 1 [cited 2025 Jun 7];233:113–20.e1. Available from: https://escholarship.org/uc/item/3pr8j6v2

  • Buvaneswarran S, Wong YL, Liang S, et al. Active Treatment vs Expectant Management of Patent Ductus Arteriosus in Preterm Infants: A Meta-Analysis. JAMA Pediatr [Internet]. 2025 May 27 [cited 2025 Jun 7]. Available from: https://jamanetwork.com/journals/jamapediatrics/fullarticle/2819631

  • Sung SI, Lee JH, Ahn SY, Chang YS, Park WS. Comparison of Nonintervention vs. Targeted Closure of Patent Ductus Arteriosus in Very Preterm Infants. JAMA Pediatr [Internet]. 2020 Oct 1 [cited 2025 Jun 7];174(10):953–60. Available from: https://jamanetwork.com/journals/jamapediatrics/fullarticle/2767862

  • Benitz WE. Treatment of persistent patent ductus arteriosus in preterm infants: time to accept the null hypothesis? J Perinatol [Internet]. 2010 Apr [cited 2025 Jun 7];30(4):241–52. Available from: https://www.nature.com/articles/jp20103

  • Clyman RI, Liebowitz M, Kaempf J, et al. An exploratory randomized controlled trial of treatment of moderate-to-large patent ductus arteriosus at 1 week of age. J Pediatr [Internet]. 2019 Feb [cited 2025 Jun 7];205:41–48.e6. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6487511/

  • Wright CJ. Acetaminophen and the Developing Lung: Could There Be Lifelong Consequences? J Pediatr [Internet]. 2021 Aug [cited 2025 Jun 7];235:264–276.e1. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9704024/

  • Wright CJ, McCulley DJ, Mitra S, Jensen EA. Acetaminophen for the Patent Ductus Arteriosus: Has Safety Been Adequately Demonstrated? J Perinatol [Internet]. 2023 Oct [cited 2025 Jun 7];43(10):1230–7. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10628079/

  • Benitz WE, Committee on Fetus and Newborn. Patent Ductus Arteriosus in Preterm Infants. Pediatrics [Internet]. 2016 Jan [cited 2025 Jun 7];137(1):e20153730. Available from: https://publications.aap.org/pediatrics/article-full/137/1/e20153730/52459/Patent-Ductus-Arteriosus-in-Preterm-Infants

  • Semberova J, Sirc J, Miletin J, et al. Spontaneous Closure of Patent Ductus Arteriosus in Infants ≤1500 g. Pediatrics [Internet]. 2017 Aug [cited 2025 Jun 7];140(2):e20164258. Available from: https://publications.aap.org/pediatrics/article-full/140/2/e20164258/64902/Spontaneous-Closure-of-Patent-Ductus-Arteriosus

  • Rastogi A, Akintorin S, Bez M, Morales P, Pildes RS. A controlled trial of dexamethasone to prevent bronchopulmonary dysplasia in surfactant treated infants. Pediatrics [Internet]. 1996 Aug [cited 2025 Jun 7];98(2 Pt 1):204–10. Available from: https://pediatrics.aappublications.org/content/98/2/204.long

  • Matsushita FY, Krebs VLJ, de Carvalho WB. Heterogeneity in treatment response for patent ductus arteriosus: a meta-analysis. Eur J Pediatr [Internet]. 2025 Feb [cited 2025 Jun 7];184(2):603–16. Available from: https://doi.org/10.1007/s00431-025-06130-8

  • Benitz WE. Commentary on Matsushita et al. Eur J Pediatr [Internet]. 2025 Feb [cited 2025 Jun 7];184(2):617–8. Available from: https://doi.org/10.1007/s00431-025-06192-8

  • Chen et al. Vascular and pulmonary effects of ibuprofen on neonatal lung development. Respir Res [Internet]. 2023 Feb 2 [cited 2025 Jun 7];24(1):39. Available from: https://doi.org/10.1186/s12931-023-02342-4

  • Clyman RI, Benitz WE. Patent ductus arteriosus and bronchopulmonary dysplasia. J Perinatol [Internet]. 2019 Oct [cited 2025 Jun 7];39(10):1435–41. Available from: https://doi.org/10.1038/s41372-019-0483-x

  • Sankar MN, Benitz WE. Does crossover treatment of control subjects invalidate results of randomized trials of patent ductus arteriosus treatment? J Perinatol [Internet]. 2020 Dec [cited 2025 Jun 7];40(12):1863–70. Available from: https://doi.org/10.1038/s41372-020-00848-z

  • de Carvalho Nunes G, Wutthigate P, Simoneau J, et al. Natural evolution of the patent ductus arteriosus in the extremely premature newborn and respiratory outcomes. J Perinatol [Internet]. 2022 May [cited 2025 Jun 7];42(5):642–8. Available from: https://doi.org/10.1038/s41372-021-01277-2

  • W.E. Benitz, C.H. Backes. At a crossroads for early medical treatment of persistent patent ductus arteriosus in preterm infants. J Perinatol [Internet]. 2024 Nov [cited 2025 Jun 7];44(11):1534–7. Available from: https://doi.org/10.1038/s41372-024-02022-1

  • Villamor E, Gairing-Haslinger G, Gamper G, Bartoš F, Hinterbuchner B, Kiechl-Kohlendorfer U, et al. Should we treat patent ductus arteriosus in very preterm infants? A Bayesian meta-analysis of mortality and bronchopulmonary dysplasia. Pediatr Res [Internet]. 2025 Apr 1 [cited 2025 Jun 7];97(5):1427–36. Available from: https://doi.org/10.1038/s41390-025-03820-9

  • Evans N. Cardiovascular effects of dexamethasone in the preterm infant. Arch Dis Child [Internet]. 1994 [cited 2025 Jun 7];70(1 Spec No):F25–30. Available from: https://fn.bmj.com/content/archdischild/70/1_Spec_No/F25.full.pdf