Adding budesonide to surfactant did not improve lung outcomes in extremely preterm babies

Bronchopulmonary dysplasia (BPD) — chronic lung disease of prematurity — is one of the most common serious complications in babies born extremely early, before 28 weeks' gestation [s1]. Systemic corticosteroids reduce the risk of BPD but carry their own hazards, which has driven a long search for a way to deliver the steroid to the lungs alone [s1]. The PLUSS trial, published in JAMA, tested exactly that idea and found little to show for it.

What the trial tested

PLUSS was a double-blind randomised clinical trial run in 21 neonatal units across four countries — Australia, New Zealand, Canada and Singapore [s1]. It enrolled infants born at less than 28 weeks' gestation and less than 48 hours old who were either mechanically ventilated or on noninvasive support with a clinical decision to give surfactant [s1]. Recruitment ran from January 2018 to March 2023, with the last infant discharged in August 2023 [s1].

Infants were randomly allocated 1:1 to one of two treatments: budesonide at 0.25 mg/kg mixed with the surfactant poractant alfa and delivered through an endotracheal tube or thin catheter, or the same surfactant with no steroid added [s1]. The primary outcome was survival free of BPD at 36 weeks' postmenstrual age — a single composite that counts both staying alive and avoiding the diagnosis [s1]. The trial specified 15 secondary outcomes, including the two components of that composite and nine predefined safety outcomes [s1].

What it found

The primary analysis included 1,059 infants: 524 in the budesonide-and-surfactant group and 535 in the surfactant-only group [s1]. These were very small, very sick babies — mean gestational age 25.6 weeks (SD 1.3) and mean birth weight 775 g (SD 197), with 586 (55.3%) male [s1].

Survival free of BPD occurred in 134 infants (25.6%) given budesonide and surfactant, against 121 (22.6%) given surfactant alone — an adjusted risk difference of 2.7% (95% CI −2.1% to 7.4%) [s1]. Because that interval comfortably spans zero, the difference is statistically compatible with no benefit at all.

Breaking the composite apart tells the same story. At 36 weeks' postmenstrual age, 83.2% of infants were alive in the budesonide group and 80.6% in the surfactant-only group; of the survivors, 69.3% and 71.9% respectively were diagnosed with BPD [s1]. In other words, both the survival and the lung-disease components moved only fractionally, and in the modest directions the composite already implied [s1]. The study's own conclusion is blunt: in extremely preterm infants receiving surfactant for respiratory distress syndrome, early intratracheal budesonide may have little to no effect on survival free of BPD [s1].

How to read it

This is the kind of result that is easy to misread from a distance. A 25.6%-versus-22.6% split looks, at a glance, like a win for the steroid. But the confidence interval around the difference runs from a 2.1% harm to a 7.4% benefit, which means the trial cannot rule out that budesonide did nothing — and cannot rule out that it helped a little [s1]. A large, carefully conducted trial that lands on a wide interval straddling zero is telling clinicians that the hoped-for effect, if it exists, is smaller than it would need to be to change practice on this evidence alone [s1].

It is worth being precise about what PLUSS does and does not settle. It tested one steroid, at one dose, mixed with one surfactant, given early, with a primary endpoint at 36 weeks [s1]. It does not speak to later respiratory outcomes, to other steroids or delivery methods, or to longer-term neurodevelopment — questions the trial's secondary and follow-up analyses are designed to probe [s1]. The statistical plan was published in advance, which strengthens confidence that the primary result was not chosen after the fact [s2].

What to watch

The immediate question is whether the trial's longer-term and safety outcomes, and pooled analyses combining PLUSS with similar trials, shift the picture at all [s1][s2]. PLUSS was large by the standards of neonatal trials — 1,059 infants across 21 units in four countries — which is precisely what makes its flat result informative rather than merely inconclusive: a smaller study could have missed a real effect, but a trial this size landing on a 2.7% difference with a confidence interval spanning zero is a stronger signal that there is little here to find [s1]. For now, the central finding is a cautionary one: delivering a corticosteroid straight into the lungs, an approach with real theoretical appeal, did not deliver a clear improvement in the outcome that matters most to families of extremely preterm babies.

This article describes research and is not medical advice. Decisions about the care of preterm infants are for families and their treating clinicians.