Beyond ductal closure: interpreting the TREOCAPA trial in the evolving management of patent ductus arteriosus in preterm infants

Beyond ductal closure

Few questions in neonatal medicine have generated as much debate as the management of patent ductus arteriosus (PDA) in extremely preterm infants. Despite more than five decades of pharmacologic intervention, uncertainty persists regarding whether closure of the ductus arteriosus improves clinically meaningful outcomes or merely changes a physiological marker of prematurity. During this period, the pendulum has swung repeatedly between aggressive early intervention and expectant management, reflecting a growing recognition that successful anatomical closure has not consistently translated into reductions in mortality or major neonatal morbidities (1-5). Rather than resolving the controversy, each successive randomized trial has refined the questions clinicians should be asking.

The recently published the prophylactic treatment of the ductus arteriosus in preterm infants by acetaminophen (TREOCAPA) trial by Rozé and colleagues represents one of the most rigorous investigations of prophylactic pharmacological treatment for PDA performed to date (6). In this multicentre, double-blind, placebo-controlled trial involving 778 extremely preterm infants across 43 neonatal intensive care units in 14 European countries, prophylactic intravenous acetaminophen initiated within the first 12 hours after birth significantly increased ductal closure by day seven and reduced the need for rescue treatment. Nevertheless, these anatomical benefits failed to translate into improvements in the primary outcome of survival without severe neonatal morbidity at 36 weeks’ postmenstrual age. Furthermore, acetaminophen was associated with a higher incidence of cholestasis, highlighting the need to reconsider the perception of acetaminophen as a relatively benign therapy.

Although these findings may initially appear disappointing for clinicians seeking a safer alternative to cyclooxygenase inhibitors, the true importance of TREOCAPA extends beyond acetaminophen itself. Rather, the trial challenges the longstanding assumption that earlier anatomical ductal closure necessarily improves clinically meaningful outcomes.

TREOCAPA adds to a growing body of randomized evidence demonstrating that increased ductal closure has not consistently improved clinically meaningful outcomes, including bronchopulmonary dysplasia (BPD), necrotizing enterocolitis (NEC), severe intraventricular hemorrhage (IVH), and death (1,3,7).

Importantly, this disconnect should not be interpreted as evidence that the ductus arteriosus is clinically irrelevant. The physiological consequences of a persistent ductal shunt remain biologically plausible and should not be dismissed simply because randomized trials have failed to demonstrate population-level benefit from routine treatment. Experimental and clinical studies have shown that prolonged left-to-right shunting can result in pulmonary overcirculation, ventricular volume overload, systemic hypoperfusion, and end-organ dysfunction. These mechanisms provide a strong physiological rationale for intervention in carefully selected infants while simultaneously illustrating why anatomical closure alone may be insufficient to overcome the multifactorial pathogenesis of BPD, NEC, IVH, or death.

Observational studies consistently associate persistent large ductal shunts with adverse neonatal outcomes. TREOCAPA reminds us, however, that association does not establish causation. PDA may represent both a contributor to hemodynamic instability and a marker of extreme physiological immaturity.

The evolution of neonatal hemodynamics provides an important framework for interpreting TREOCAPA. Increasingly, bedside cardiovascular assessment integrates ductal anatomy with ventricular function, pulmonary vascular resistance, systemic blood flow, and tissue perfusion, recognizing that the clinically important question is no longer whether the ductus is patent, but whether it is physiologically significant enough to justify intervention (8-11).

Rather than being viewed as a negative trial, TREOCAPA represents an important contemporary contribution to the evolution of PDA management. The study demonstrates that prophylactic acetaminophen accelerates ductal closure but does not improve the outcomes that matter most to infants and their families. Importantly, this dissociation between anatomical closure and clinical benefit is not novel to TREOCAPA but reinforces observations from preceding randomized trials and the longstanding movement away from ductal patency alone as a therapeutic target. TREOCAPA therefore strengthens, rather than initiates, the rationale for investigating which infants, if any, may benefit from intervention based on individualized cardiovascular physiology.

Why has anatomical success failed to improve clinical outcomes?

The findings of TREOCAPA (6) reinforce a recurring observation from decades of randomized PDA trials: interventions that reliably promote ductal closure have consistently failed to improve clinically meaningful outcomes. While acetaminophen increased the rate of ductal closure by nearly 20 percentage points and reduced the need for rescue therapy, these anatomical successes did not translate into improved survival without severe neonatal morbidity. This discordance should prompt clinicians to reconsider whether ductal patency itself is the appropriate therapeutic target or whether it simply represents one component of a much more complex physiological process (6).

Interpretation of the anatomical endpoint also requires recognition that ductal closure in TREOCAPA was assessed at day 7 as an exploratory outcome. A closed ductus at this time point does not characterize subsequent ductal trajectory or cumulative exposure to ductal shunting, and the trial therefore provides limited insight into whether later persistence or reopening modifies the relationship between early closure and clinical outcome.

Historically, management strategies for PDA have been driven by the assumption that the ductus plays a direct causal role in major neonatal complications, including BPD, NEC, IVH, and death. This biological rationale is compelling. A large left-to-right ductal shunt increases pulmonary blood flow, reduces systemic perfusion through ductal steal, increases ventricular volume loading, and may compromise perfusion of vulnerable organs such as the brain, kidneys, and intestine (8,12). Observational studies have consistently demonstrated associations between persistent large PDAs and adverse neonatal outcomes (13,14). However, association should not be mistaken for causation. Infants with persistent PDAs are also among the most physiologically immature, frequently requiring prolonged respiratory support, vasoactive medications, and intensive cardiovascular management. Consequently, the PDA may function not only as a mediator of disease but also as a marker of the underlying vulnerability of the extremely preterm infant.

If adverse outcomes arise primarily from the biological consequences of prematurity rather than from the ductus itself, then pharmacologic closure alone would not be expected to substantially modify those outcomes. The TREOCAPA trial provides compelling evidence supporting this concept. Despite increased early ductal closure, no significant differences were observed in BPD, severe IVH, NEC, cystic periventricular leukomalacia, or survival without severe morbidity (6). Similar findings have emerged from trials evaluating prophylactic indomethacin, early ibuprofen treatment, and expectant management strategies, suggesting that the absence of clinical benefit is unlikely to reflect the pharmacologic properties of acetaminophen alone (1,7,15).

Similar conclusions have emerged from other randomized trials, including BeNeDuctus, in which expectant management was non-inferior to early ibuprofen treatment (1). More recently, Buvaneswarran et al. conducted a meta-analysis of 10 randomized trials involving 2,035 preterm infants with hemodynamically significant PDA (hsPDA); the authors found that active treatment during the first two postnatal weeks was associated with higher mortality and a higher composite outcome of death or moderate-to-severe BPD compared with expectant management (16). These findings are particularly important because they suggest that identifying a PDA as hemodynamically significant does not, by itself, establish that pharmacological treatment will confer clinical benefit and may, in some settings, be associated with net harm.

A complementary explanation for this persistent disconnect is that the pharmacologic therapy itself may offset some of the anticipated benefit (14). Cyclooxygenase inhibitors have long been associated with adverse effects on renal, gastrointestinal, and cerebral perfusion (2,13), while emerging experimental and clinical evidence suggests that both ibuprofen and acetaminophen may adversely affect the developing lung (17). Consequently, the absence of improved clinical outcomes may reflect not only the complex biology of PDA, but also the unintended consequences of therapies used to achieve ductal closure (17).

Importantly, these findings should not be interpreted as evidence that PDA treatment is never warranted. Rather, they highlight the limitations of a “one-size-fits-all” approach. PDA is a dynamic cardiovascular condition whose physiological impact varies markedly between infants and cannot be inferred from ductal diameter alone. Two infants with similar anatomical findings may have markedly different degrees of pulmonary overcirculation, systemic hypoperfusion, myocardial loading, and tissue oxygen delivery, making uniform prophylactic treatment unlikely to provide consistent benefit.

Advances in targeted neonatal echocardiography (TnECHO) have supported this shift by extending assessment beyond ductal anatomy to comprehensive physiological evaluation, recognizing that hemodynamic significance cannot be inferred from anatomy alone (8,11).

From anatomy to physiology: the emergence of precision neonatal hemodynamics

If TREOCAPA represents another step away from routine prophylactic ductal closure, an equally important question emerges around what should replace anatomy-based decision-making. Increasingly, the answer lies not in identifying more effective pharmacologic agents but in improving our understanding of neonatal cardiovascular physiology. This evolution reflects a broader shift from anatomy-based treatment strategies toward individualized physiological assessment, with therapeutic decisions guided by comprehensive neonatal hemodynamic evaluation rather than ductal patency alone.

Importantly, TREOCAPA evaluated very early prophylactic treatment in an unselected population rather than a strategy in which intervention was directed toward infants with an established hemodynamic phenotype. Its findings therefore demonstrate the limitations of routine early pharmacological ductal closure at the population level but do not determine whether a subgroup of infants with specific physiological characteristics might benefit from intervention. Whether such treatment-responsive phenotypes can be reliably identified remains an important question for future prospective trials.

Historically, treatment decisions and many randomized trials have relied on criteria such as ductal diameter, shunt direction, and chronological age to define hsPDA. However, ductal diameter correlates only modestly with true shunt volume and provides limited insight into myocardial performance, pulmonary vascular resistance, systemic perfusion, or end-organ oxygen delivery (8,11). Modern hemodynamic assessment instead integrates pulmonary overcirculation, myocardial loading, systemic blood flow, ductal flow characteristics, and evidence of end-organ hypoperfusion, recognizing that anatomy alone may lead to both overtreatment of physiologically insignificant PDAs and undertreatment of infants with substantial cardiovascular compromise.

The concept of hsPDA has evolved from a fixed echocardiographic diagnosis to a dynamic physiological state reflecting shunt volume, myocardial adaptation, pulmonary vascular transition, systemic vascular resistance, and cardiovascular maturity (9). Because these variables change rapidly after birth, PDA significance cannot be defined by a single echocardiographic study or predetermined treatment window. Accordingly, hemodynamic significance may depend not only on the magnitude of the ductal shunt at a single time point, but also on its duration and trajectory over time. Cumulative exposure to a physiologically significant shunt may therefore be an important dimension of disease burden, supporting serial rather than isolated assessment and warranting consideration when defining patient eligibility and timing of intervention in future physiology-based trials.

Importantly, however, physiological significance should not be equated with treatment responsiveness. Demonstrating a substantial ductal shunt and its hemodynamic consequences does not establish that pharmacological closure will improve outcome; the potential benefits of reducing shunt exposure must be weighed against treatment-related effects and the broader cardiovascular consequences of altering transitional physiology.

TnECHO has fundamentally changed how clinicians evaluate these infants. It extends well beyond confirming ductal patency to provide a comprehensive assessment of cardiovascular function, including ventricular outputs, myocardial systolic and diastolic performance, estimates of pulmonary artery pressure, superior vena cava flow, left atrial and ventricular volume loading, ductal flow characteristics, and markers of systemic hypoperfusion (18,19). Additional physiological variables—including pulmonary vein Doppler, mitral inflow velocities, left ventricular isovolumic relaxation time, left atrial-to-aortic ratio, ventricular output ratios, peripheral Doppler flow patterns, and atrial-level shunting—may further refine assessment of ductal significance and improve physiological phenotyping of affected infants. These physiological measurements allow clinicians to distinguish infants with similar anatomical lesions but markedly different cardiovascular phenotypes.

In extremely preterm infants, cardiovascular instability reflects interactions among myocardial adaptation, pulmonary vascular transition, respiratory disease, and systemic inflammation; PDA is therefore often only one component of a broader hemodynamic syndrome.

Emerging technologies, including advanced echocardiography, near-infrared spectroscopy (NIRS), lung ultrasound, and circulating biomarkers, may further refine physiological phenotyping beyond anatomy alone. Integrating these complementary tools into future trials may help characterize physiological heterogeneity and test whether clinically meaningful treatment-responsive phenotypes can be identified (19-24).

Recent observational data provide further support for the potential relevance of physiological phenotyping. Toyoshima et al. reported that serial echocardiographic measures of PDA-related hemodynamic burden were associated with neurodevelopmental outcomes at 3 years, with patterns differing across gestational-age strata; however, whether such physiological markers can identify infants who derive benefit from PDA intervention remains to be established prospectively (25).

Importantly, adopting a physiology-based approach does not imply delaying treatment until irreversible organ injury has occurred. Rather, individualized assessment may offer an opportunity to identify infants with objective evidence of cardiovascular compromise while avoiding unnecessary treatment in those without substantial hemodynamic burden. Such enrichment strategies could potentially improve the efficiency of future randomized trials by selecting patients with the greatest potential to derive benefit from intervention.

TREOCAPA also challenges investigators to rethink how future PDA trials are designed. Previous randomized trials have largely enrolled infants according to gestational age, chronological age, or anatomical ductal characteristics. However, these enrollment strategies assume that all infants with similar anatomical findings share comparable cardiovascular physiology. Hemodynamic research suggests precisely the opposite. Physiological heterogeneity among extremely preterm infants is considerable, and this heterogeneity may explain why treatments demonstrating consistent anatomical efficacy repeatedly fail to improve population-level clinical outcomes.

Where do we go from here?

The TREOCAPA trial represents an important contribution to the evolution of PDA management. Rather than identifying a new standard pharmacologic therapy, it compels the neonatal community to reconsider the fundamental objectives of treatment. This distinction is particularly important because neonatal medicine has gradually moved beyond the concept that structural abnormalities should be corrected simply because they can be corrected. Similar shifts have occurred in other areas of neonatal cardiovascular care. Therapeutic decisions for persistent pulmonary hypertension, hypotension, and myocardial dysfunction are increasingly guided by physiological assessment rather than isolated numerical thresholds. PDA research and management may increasingly evolve along the same trajectory. The objective should therefore move beyond universal ductal closure toward assessment of physiological significance and optimization of cardiovascular physiology for the individual infant, an approach also emphasized in a recent Commentary advocating physiology-guided patient selection in future PDA trials (8,9,26).

Future clinical trials should consider enrichment strategies targeting infants at greatest risk of sustained, physiologically important ductal shunting and incorporate serial hemodynamic phenotyping using TnECHO, validated cardiovascular severity scores, biomarkers of myocardial stress, and objective measures of systemic and regional perfusion. Extreme gestational immaturity may represent an important component of such enrichment strategies but should be integrated with objective evidence of hemodynamic burden and biological susceptibility rather than used as an isolated criterion for treatment.

Technological advances provide additional opportunities to refine this individualized approach. Artificial intelligence-assisted echocardiographic analysis, automated quantification of ventricular function, continuous cerebral and somatic NIRS, lung ultrasound, and integration of multimodal physiological monitoring may allow clinicians to characterize cardiovascular phenotypes with greater precision than has previously been possible (20,21).

For practicing neonatologists, TREOCAPA reinforces that routine prophylactic pharmacological treatment of all extremely preterm infants cannot be recommended solely to improve clinically meaningful neonatal outcomes and that successful ductal closure should not be regarded as a surrogate for clinical benefit. When treatment is being considered, therapeutic decisions should incorporate careful clinical and physiological assessment, recognizing that the significance of a PDA and the potential benefit of intervention vary between infants.

Perhaps the greatest contribution of TREOCAPA is that it changes the question. TREOCAPA suggests that whether the ductus should be closed may no longer be the most important question. Instead, future research should ask which infant, at what stage of transitional adaptation, and with which physiological characteristics is most likely to benefit from intervention. At present, however, no validated physiological phenotype, echocardiographic threshold, or combination of clinical and hemodynamic variables can reliably identify such infants or predict who will benefit from ductal intervention. Indeed, the large number of interacting and rapidly evolving determinants of cardiovascular adaptation may preclude a simple or universally applicable definition. Comprehensive cardiovascular phenotyping and precision neonatal hemodynamics should therefore be viewed not as an established solution, but as a framework through which these uncertainties can be investigated and, potentially, clinically meaningful treatment-responsive phenotypes identified.

In this sense, TREOCAPA should not be viewed as the end of the PDA story but as the beginning of its next chapter. The future is unlikely to be defined by discovering another drug capable of achieving higher closure rates, but rather by determining whether we can identify the right patient, at the right time, with the right physiological indication for treatment. As neonatal cardiovascular care embraces precision medicine, success should ultimately be measured not by whether the ductus closes, but by whether the infant fares better because it did.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.