Prophylactic ibuprofen in premature infants: a multicentre, randomised, double-blind, placebo-controlled trial

Article Outline

• Copyright

Van Overmeire B, Allegaert K, Casaer A, Debauche C, Decaluwe W, Jespers A, et al. Lancet 2004;364:1945-9.

Context Ibuprofen is used for treatment and prevention of patent ductus arteriosus (PDA) in low-birthweight infants. Its effects on regional circulations differ from those of indomethacin.

Objectives To study the efficacy of early ibuprofen in reducing the frequency of severe intraventricular haemorrhage (IVH) and PDA.

Design Double-blind, randomized, multicenter trial.

Setting Neonatal intensive care units at seven hospitals in Belgium.

Participants 415 low-birthweight infants (gestational age <31 weeks).

Interventions Within 6 hours after birth, infants were randomly allocated ibuprofen-lysine (10 mg/kg then two doses of 5 mg/kg after 24 hours and 48 hours) or placebo intravenously.

Main outcome measures The primary outcome was occurrence of severe IVH; secondary outcomes were occurrence of PDA and possible adverse effects of ibuprofen.

Results 17 (8%) of 205 infants assigned ibuprofen and 18 (9%) of 210 assigned placebo developed severe IVH (relative risk 0.97 [95% CI 0.51–1.82]). In 172 (84%) infants in the ibuprofen group, the ductus was closed on day 3 compared with 126 (60%) of the placebo group (relative risk 1.40 [95% CI 1.23–1.59]). No important differences in other outcomes or side effects were noted; however, urine production was significantly lower on day 1 and concentration of creatinine in serum was significantly higher on day 3 after ibuprofen.

Conclusions Ibuprofen prophylaxis in preterm infants does not reduce the frequency of IVH, but it does decrease occurrence of PDA.

Comment This is a carefully designed and implemented study to determine whether prophylactic ibuprofen would prevent IVH in premature neonates. The authors found no protective IVH effect and are to be complimented on their reporting of negative results.

Their primary objective disproven, the authors describe a secondary effect of prophylactic ibuprofen on PDA closure. The crux of this effect is based on an increased ductal closure rate on day 3 of life in the treated group. However, there are several potential pitfalls inherent in this comparison. At this age, the infants in the ibuprofen group have already benefited from exposure to prostaglandin synthetase inhibitor therapy. On the other hand, infants in the placebo group have not yet been exposed to any PDA closure medication. Ductal closure rates can be validly compared only after all infants have been provided medical treatment (ie, following an attempt at therapeutic closure in those placebo-treated infants whose ductus did not close spontaneously). In the placebo group, it will generally be later than day 3 of life by the time clinically significant PDAs are diagnosed and treated therapeutically. If the definitive comparison, performed after such an attempt at therapeutic closure has been offered, reveals a comparable number of infants requiring surgical ductal ligation, and if long-term morbidity remains similar in the two groups, then essentially the same clinical result will have been achieved in both groups. Further, the observation that the PDA was closed by day 3 in 60% of the placebo group implies that an equivalent 60% of PDAs in the ibuprofen group would have been expected to close spontaneously; thus, these infants may have been unnecessarily exposed to a drug. Finally, even if treatment is required for a PDA, using ibuprofen therapeutically, rather than prophylactically, delays exposure to a drug whose toxicity (pulmonary hypertension, oliguria, increased creatinine) may be increased when given in the first hours of life.

In conclusion, before we consider routine implementation of prophylactic ibuprofen, we must consider whether it is justified to close PDAs prophylactically if treatment confers no long-term benefit and unnecessarily exposes infants to a drug with a small but documented potential toxicity. This may be one case in which an ounce of prevention is not worth a pound of cure.