Trial of Selective Early Treatment of Patent Ductus Arteriosus with Ibuprofen

Extreme preterm infant survival has improved over recent decades, yet bronchopulmonary dysplasia (BPD) rates remain high. A large patent ductus arteriosus (PDA, ≥1.5 mm diameter) persisting beyond 3 days is associated with increased odds of death, severe morbidity, and BPD, and PDA persistence is inversely related to gestational age. While prophylactic cyclooxygenase inhibitors can reduce severe intraventricular and pulmonary hemorrhage, they do not improve long-term neurodevelopment and expose many infants to unnecessary treatment because spontaneous ductal closure is common. Data on treating asymptomatic PDA are limited and have not shown improved clinical outcomes, leading to reduced pharmacologic treatment rates. Bedside functional echocardiography enables targeted identification of large PDAs with unrestricted shunt unlikely to close spontaneously, potentially allowing selective early treatment and avoidance of unnecessary exposure. Indomethacin and ibuprofen have similar efficacy for early targeted PDA treatment, with ibuprofen having a more favorable side-effect profile. The study tested the hypothesis that early (≤72 hours), echocardiography-targeted ibuprofen for large PDA would reduce mortality and short-term morbidity, including BPD, compared with placebo.

Design: Multicenter, randomized, double-blind, placebo-controlled trial conducted in 32 UK neonatal intensive care units according to a published protocol, with outcome assessors, clinicians, and families blinded. Oversight by a Trial Steering Committee with recommendations from an independent Data Monitoring Committee; coordinated by the NPEU Clinical Trials Unit, University of Oxford. Participants: Infants born between 23+0 and 28+6 weeks’ gestation, <72 hours old, with echocardiographically confirmed large PDA (≥1.5 mm diameter) and unrestricted left-to-right pulsatile flow, without clinical concerns for acute pulmonary hypertension. Written parental consent obtained. Randomization: Central secure web-based dynamic allocation using probabilistic minimization (1:1), balancing PDA size, gestational age, age, sex, site, multiple birth, mode of respiratory support, and inotrope use; multiple births randomized individually. Intervention: Ibuprofen sodium (intravenous) at 10 mg/kg loading dose followed by two 5 mg/kg doses at least 24 hours apart (single course). Placebo: equal volume 0.9% sodium chloride. Only one course permitted. Echocardiography performed for eligibility within 72 hours of birth and repeated at 18–24 days to assess ductal patency. Predefined criteria guided open-label medical or surgical PDA treatment after enrollment. Quality control: Independent blinded review of a sample of echocardiograms from 65 infants to verify eligibility criteria; 93.8% met predefined echo eligibility. Outcomes: Primary—composite of death or moderate/severe BPD at 36 weeks’ post-menstrual age, with physiologic oxygen reduction testing to differentiate BPD severity. Secondary—individual components of the primary outcome; severity of BPD; severe intraventricular hemorrhage (grade III/IV); cystic periventricular leukomalacia; retinopathy of prematurity requiring treatment; significant pulmonary hemorrhage; acute pulmonary hypertension treated with vasodilators; necrotizing enterocolitis (Bell stage II+); PDA status at ~3 weeks; open-label symptomatic PDA treatment; surgical ligation; discharge home on oxygen; weight gain (change in z-score). Process outcomes included dosing adherence and timing, and completion of oxygen reduction testing and echocardiography. Statistical analysis: Intention-to-treat, excluding only those with missing data; no imputation. Models adjusted for minimization factors and correlation between multiples; center as a random effect. Binary outcomes analyzed using mixed-effects Poisson regression with robust variance to estimate risk ratios and 95% CIs; continuous outcomes via linear regression with mean differences and 95% CIs. A predefined shortlist of secondary outcomes guided inference without multiplicity adjustment. Stata/SE v15 used. Sample size: Planned n=730 to detect an absolute risk reduction of 12% (60% to 48%) in the primary outcome with 90% power and 5% alpha, assuming 1% loss to follow-up. Recruitment achieved n=653 (326 ibuprofen; 327 placebo).

- Enrollment: 653 infants randomized (326 ibuprofen; 327 placebo). Primary outcome assessable in 318 vs 318. - Primary outcome: Death or moderate/severe BPD at 36 weeks’ PMA occurred in 69.2% (220/318) ibuprofen vs 63.5% (202/318) placebo; aRR 1.09 (95% CI 0.98–1.20), p=0.10. - Death by 36 weeks’ PMA: 13.6% (44/323) ibuprofen vs 10.3% (33/321) placebo; aRR 1.32 (95% CI 0.92–1.90). - Among survivors to 36 weeks’ PMA: Moderate/severe BPD in 64.2% (176/274) ibuprofen vs 59.3% (169/285) placebo; aRR 1.09 (95% CI 0.96–1.23). - PDA status at ~3 weeks: Closed or <1.5 mm with restricted flow in 55.5% (176/317) ibuprofen vs 37.0% (117/316) placebo; aRR 1.50 (95% CI 1.30–1.74). - Open-label symptomatic PDA treatment: Medical treatment in 13.3% ibuprofen vs 25.5% placebo; surgical ligation 2.8% vs 9.6%. - Selected secondary outcomes (adjusted RRs): Severe IVH (grade III/IV) 1.30 (0.93–1.82); cystic PVL 1.62 (0.69–3.83); treated ROP 0.98 (0.68–1.42); significant pulmonary hemorrhage 1.39 (0.70–2.77); acute pulmonary hypertension requiring vasodilator 1.04 (0.51–2.13); NEC Bell stage II+ 1.01 (0.67–1.51); discharge home on oxygen 1.06 (0.92–1.22); weight gain (z-score change) mean difference 0.1 (−0.1 to 0.2). - Treatment adherence: 87.3% vs 89.4% received all 3 doses; median age at first dose 61 hours in both groups. - Safety: Two serious adverse events possibly related to ibuprofen; overall no clear excess in serious complications attributable to ibuprofen.

Early, echocardiography-targeted ibuprofen treatment for large PDA did not improve the composite of death or moderate/severe BPD at 36 weeks’ post-menstrual age compared with placebo, nor did it clearly reduce death or BPD considered separately. Although ibuprofen increased the likelihood of ductal closure or constriction by ~18% at ~3 weeks and reduced subsequent open-label medical and surgical PDA treatments, these effects did not translate into improved short-term clinical outcomes. Several factors may explain the lack of clinical benefit: only 55.5% of ibuprofen-treated infants had closed or small PDAs at ~3 weeks; open-label treatment occurred relatively frequently, especially in the placebo group (approximately 30%), diminishing between-group differences in ductal patency; and open-label rescue therapy tended to occur late (median 11–12 days post-randomization), potentially prolonging exposure to hemodynamically significant shunting in many infants. The absence of serial echocardiographic data limited assessment of shunt duration and its relationship to outcomes. Findings align with prior trials of early targeted ibuprofen that showed reduced pulmonary hemorrhage and PDA-related symptoms without improvements in mortality, BPD, or neurodisability. No clear safety signal emerged in this cohort, and differences from other trials (e.g., patient respiratory support profiles and higher cumulative ibuprofen exposure in some studies) may account for varying associations with BPD across the literature.

In extremely preterm infants with a large PDA identified by early echocardiography, a single early course of ibuprofen did not improve the composite outcome of death or moderate/severe BPD at 36 weeks’ post-menstrual age compared with placebo. While ibuprofen increased the rate of ductal closure/constriction and reduced subsequent open-label medical and surgical PDA treatments, these effects did not yield better short-term clinical outcomes. Future research could evaluate earlier initiation, alternative dosing strategies or repeated courses, comparative effectiveness of different agents, and strategies incorporating serial echocardiography to better characterize shunt duration and guide timing of intervention.

- Open-label PDA therapy occurred in 29.8% of infants in the placebo arm, likely increasing closure rates and reducing between-group separation in ductal patency and outcomes. - Recruitment shortfall: 653 enrolled versus 730 planned, due to drug non-availability, evolving clinical practice, competing trials, and COVID-19, potentially reducing power to detect modest effects. - Timing of intervention: Median age at first dose was 61 hours, with enrollment allowed up to 72 hours; earlier treatment might achieve higher closure rates. - Lack of serial echocardiography limited analysis of shunt duration and its impact on outcomes. - A single ibuprofen course was used; higher or repeated dosing regimens were not assessed. - Some outcome assessments and echocardiograms were missing in a minority of infants, though primary outcome ascertainment was high and analyses were intention-to-treat without imputation. - No multiplicity adjustment for numerous secondary outcomes limits definitive inference for those endpoints.