Many times in medicine, we employ a treatment that accomplishes some beneficial therapeutic goal, only to later find that an unanticipated action of the therapy has caused a significant problem in patients. This doesn't necessarily mean that the employed treatment is bad or should be discarded, only that all effects must be considered, and balanced in the decision about using that therapy.

Indomethacin therapy for PDA is efficacious in closing a patent ductus arteriosus (PDA), but it is not without risks or side effects. Indomethacin, acting by the same mechanisms that lead to ductal closure, alters renal and gastrointestinal perfusion. Its use can result in transient or permanent renal dysfunction (1,2,9). Oliguria is related to the inhibition of two cyclooxygenase enzymes (COX-1 and COX-2), which are involved in the physiologic functioning of the kidney and other tissues (3). Indomethacin also has been associated with the development of necrotizing enterocolitis and gastrointestinal hemorrhage (4).

These acute side effects are important, but the risks are generally thought to be acceptable because of the many benefits of improved cardiovascular status and clinical stability. These benefits are thought to relate to the fact that indomethacin treatment will decrease the incidence of IVH in preterm infants. A large multicenter trial (5) demonstrated that indomethacin decreased both the incidence and severity of IVH, and early low-dose indomethacin improved survival better than placebo.

The long-term effects of this therapy have been the subjects of concern, because indomethacin has been shown to transiently lower cerebral blood flow in preterm infants. The worry has been that while preventing IVH, the risk of brain ischemia could be increased (6). Two important recent studies have examined the long-term effects of indomethacin therapy. The first study involved further follow-up of children enrolled in the large early indomethacin trial (5), followed to school age. These children had been evaluated at age 36 months, at which time there was no evidence of increased cognitive impairment in indomethacin-treated children (7).

The neurodevelopmental assessments of the infants demonstrated the indomethacin-treated children had no adverse neurodevelopmental sequelae as they approached school age. Only 7% of the children suffered cerebral palsy, in both the indomethacin and placebo treated groups. The study did find that 28% of the children did have measured IQ values below the normal, and other measures of subnormal development. It was true that the proportion of children with minor or no developmental handicap was significantly smaller in the indomethacin group, which may translate into better overall development and social skills.

In the second study (8), infants of birth weights between 500 and 999g were randomized to receive either indomethacin or placebo daily for three days. The risk of death or survival with impairments was equal in both groups. The indomethacin treated group had a significantly lower incidence of patent ductus arteriosus and of severe periventricular and intraventricular hemorrhage (IVH/PVH), but the incidence of white matter injury was not changed. It appears that this white matter injury correlates with the observed neurologic impairments. Because the overall incidence of severe IVH/PVH was low in both groups (9% in the indomethacin group and 13% in the placebo group), the small absolute reduction in incidence did not significantly alter or lower the overall risk of neurologic impairment. The authors caution that while prophylactic indomethacin can be done safely and in an efficacious manner, there is no apparent protection from neurologic impairments.