Indomethacin is used prophylactically to close the PDA before hemodynamic distress occurs, which has been associated with an increased morbidity and mortality, without the need for screening or echocardiographic surveillance. This approach exposes a large number of neonates, in whom the duct would close spontaneously, to a pharmacologic treatment that is not without risk.
Intravenous indomethacin, a non-selective cyclooxygenase inhibitor, is the standard pharmacological treatment for PDA in preterm neonates and has a reported effectiveness of 66–80%.21-23
Since the 1980s, several risk-benefit studies have been conducted to evaluate the intravenous administration of indomethacin for prophylactic treatment of PDA. The first small, controlled, randomized trials suggested that early prophylaxis (i.e., within the first 24 hours) with indomethacin reduced the incidence of serious IVH.
Interestingly, a significant increase in the use of indomethacin prophylaxis occurred after the Ment et al. trial was published; however, a significant decrease in the use of indomethacin prophylaxis followed the TIPP trial.24
In 1994, Ment et al.25
was the first to publish a prospective, randomized, placebo-controlled trial to investigate whether low-dose indomethacin (0.1 mg/kg intravenously at 6 to 12 postnatal hours and every 24 hours for two more doses) would lower the incidence and severity of IVH.
The authors enrolled 431 neonates with a birth weight of 600 to 1,250 g and no evidence of IVH at 6 to 11 hours of age. Serial cranial ultrasounds and echocardiograms were performed. Within the first five days, 25 (12%) indomethacin-treated and 40 (18%) placebo-treated neonates developed IVH (p
0.03, trend test). However, only one indomethacin-treated patient experienced grade 4 IVH versus 10 placebo-treated neonates (p
In 2001, Schmidt26
published the results of another large-scale, controlled, randomized trial, the Trial of Indomethacin Prophylaxis in Preterm (TIPP), that evaluated the long-term effects of indomethacin on motor, sensory and cognitive outcomes.
In this study, 1,202 ELBW infants (500–999 g) were randomly assigned to receive either indomethacin (0.1 mg per kilogram of body weight) or placebo intravenously once daily for three days soon after birth.
The primary outcomes investigated were death, cerebral palsy, cognitive delay, deafness, and blindness at a corrected age of 18 months. The secondary long-term outcomes were hydrocephalus necessitating placement of a shunt, seizure disorder, and microcephaly within the same time frame. Secondary short-term outcomes were PDA, pulmonary hemorrhage, chronic lung disease, ultrasonographic evidence of intracranial abnormalities, necrotizing enterocolitis, and retinopathy.
Of the 574 infants with primary outcome data who received indomethacin prophylaxis, 271 (47%) died or survived with impairments, as compared with 261 of the 569 infants (46%) assigned to placebo (odds ratio, 1.1; 95 percent confidence interval, 0.8 to 1.4; p
Indomethacin reduced the incidence of PDA (24% vs. 50% in the placebo group; odds ratio, 0.3; p
<0.001) and severe PVH and IVH (9%, vs. 13% in the placebo group; odds ratio, 0.6; p
0.02). No other outcomes were altered by the prophylactic administration of indomethacin. However, indomethacin prophylaxis increased the need for supplemental oxygen from day 3 to at least day 7 of life. Indomethacin also decreased the urine volume during the first four days of life and reduced weight loss by the end of the first week.
These papers concluded that in ELBW infants, indomethacin prophylaxis does not improve the survival rate without neurosensory impairment at 18 months, despite the fact that it reduces the frequency of PDA (NNT 4) and severe PVH and IVH.
The reduction in IVH can be explained by the maturation of the cerebral vascular basement membrane, improvement in cerebrovascular self-regulation and anti-inflammatory effects.27,28
More recent data detailing reduced germinal matrix hemorrhaging via inhibition of angiogenesis indirectly support this hypothesis.29
In a follow-up to the TIPP study, 999 ELBW were given indomethacin prophylaxis. The treatment did not prevent BPD, although it reduced the frequency of PDA.30
The frequency of CLD was higher, which may be associated with the reduced weight loss, the greater need for oxygen and perhaps the increase in extracellular liquid at the pulmonary level.
Reducing the frequency of PDA may be important for those children requiring surgical closure, although twenty prophylactic treatments with indomethacin are required to avoid surgical treatment.
In another follow-up to the TIPP study that was continued up to 18 months of life, prophylactic indomethacin reduced the rate of early serious pulmonary hemorrhage because of its effects on PDA. However, prophylactic indomethacin was less effective in preventing serious pulmonary hemorrhages that occur after the first week of life.31
Currently, the main limitations of the TIPP trial are methodological concerns, such as the use of composite outcomes and a lack of statistical power.
Another important study was published by Vohr.32
Their cohort consisted of 328 VLBW (birth weight of 600–1,250 g) infants who were enrolled in a low-dose, prophylactic indomethacin prevention trial and IVH negative at six postnatal hours.
The cohort was divided into the following four subgroups for analysis: indomethacin plus IVH, indomethacin without IVH, saline plus IVH, and saline without IVH. The children were evaluated prospectively at eight years old. Children in both IVH groups had more cerebral palsy and hearing impairment as well as lower daily living skills scores, IQ scores, and vocabulary, reading and mathematics achievement test scores. Additionally, these children had greater educational resource needs. Logistic regression analyses showed that grade 3 to 4 IVH, periventricular leukomalacia and/or ventriculomegaly, male gender, maternal education, and the language spoken at home contributed to the outcomes. The authors concluded that although biological factors contribute significantly to school-related outcomes in VLBW survivors, social and environmental factors are also important contributors. Importantly, no effects of indomethacin or gestational age were identified in this study.
Cordero et al.33
compared the clinical responses of ELBW infants to indomethacin prophylaxis to that of other infants who were managed with indomethacin or surgical treatment only after a symptomatic PDA was detected. The study was a retrospective cohort investigation of 167 ELBW infants who received indomethacin prophylaxis (study) and 167 ELBW infants treated after detecting a symptomatic PDA (control) who were matched by year of birth (1999 to 2006), birth weight, gestational age (GA) and gender. Indomethacin prophylaxis did not show any advantages over early treatment for managing a symptomatic PDA in ELBW infants. The incidence of IVH was the same in the newborns receiving indomethacin prophylaxis and those in the control group.
The use of indomethacin to prevent PDA has been demonstrated to reduce the incidence of a symptomatic duct, the need for surgical closure and the occurrence of pulmonary hemorrhage.34,35
However, whether this reduction is significant has been controversial.36
A recent meta-analysis of 19 studies in Cochrane (involving 2,872 preterm infants) that also included the trials by Ment and Schmidt discussed above, confirmed that prophylactic indomethacin has short-term benefits for preterm infants, including a reduction in the incidence of symptomatic PDA, the need for surgical PDA ligation, and severe IVH. However, there was no evidence of its effect on mortality or neurodevelopment.5
Four neonates must be treated to close one DA (i.e., four is the number needed to treat) and twenty to avoid IVH. The same number 20 is needed to avoid surgical closure and pulmonary hemorrhage, and 25 are required to avoid a periventricular leukomalacia. Unfortunately, there was no effect on mortality or the long-term neurological outcomes.
However, prophylactic indomethacin does not reduce the incidence of pneumothorax, the duration of ventilation, the duration of oxygen therapy or the incidence of CLD (at 28 days or 36 weeks). CLD, brain injury and ROP were predictive factors of late death or neurosensory impairment in the follow-up to the TIPP study.37
There are recent data on the long-term alterations that are observed at school age in preterm children who were treated with indomethacin or saline.38,39
At 12 years of age, those children who had received indomethacin prophylaxis had an increased amount of gray matter in the left inferior parietal lobe, which is responsible for phonologic processes. Males had better phonological scores than saline-treated males.40
Again, newborns given indomethacin prophylaxis had greater parenchyma in the left lingual lobe at 8 years, but there was no effect on reading. Finally, newborns given indomethacin prophylaxis at birth had better connectivity between specific areas of the two hemispheres (from the right BA 40 and BA 44-45 to the left BA 22). However, indomethacin did not affect the intellectual function of children who were born preterm.41,42
The only major side effect of indomethacin reported in the 2010 Cochrane review was the increased incidence of oliguria, but it was not associated with any documented kidney damage. In the same review, no differences in the incidence of NEC or excessive bleeding were demonstrated. However, there were several side effects following the use of indomethacin demonstrated that raise serious concerns, including a reduction in cerebral blood flow,27,43,44
the volume of blood delivered to the brain and the release of brain tissue oxygen;45
oliguria and transitory kidney insufficiency;46
, necrotizing enterocolitis; isolated intestinal perforation; and gastrointestinal hemorrhage.21,47
It is important to note that administering furosemide before each indomethacin dose resulted in a significant increase in the serum creatinine level and hyponatremia without increasing urine output.48
Although most of the included studies were high quality, the drug dosage varied from one study to another, and the patient population was not homogeneous in terms of weight or gestational age. Some studies included case histories that occurred before the use of prenatal steroids or endotracheal surfactants was widespread. Furthermore, in each study, some patients in the control group were contaminated (crossed-over) because they had received indomethacin outside the limits of the study.5
Interestingly, some authors found that the timing of the first dose of indomethacin was significantly associated with the closure rate and that early administration reduced the need for surgical ligation. Up to 85.2% of DAs closed if the first dose of indomethacin was administered within 24 hours of birth; however, this rate decreased to 48.1% when treatment was started 72 hours or later after birth. The corresponding rates for surgical ligation were 3.7% and 25.9%, respectively.49
In conclusion, various clinical trials have demonstrated the effectiveness of indomethacin prophylaxis in closing Botallo's duct, but none have answered the fundamental question of whether the prophylactic closure improved the outcome. Prophylaxis does not appear to influence the development of CLD, septicemia, ROP or mortality. Although prophylaxis is associated with a reduction in serious IVHs, an important predictor of long-term neurological outcome, why prophylaxis does not influence long-term neuromotor outcomes is not known.
In an observational study, a longer duration of indomethacin exposure was associated with less white matter injury in infants delivered before 28 weeks of gestation.50
In a study by Madan,51
29% of 881 ELBW patients received indomethacin prophylaxis within the first 24 hours of life. The authors found that prophylaxis had no effect on outcomes except for a borderline increase in NEC frequency in those subjects who received prophylaxis and indomethacin therapy. In a recent study, the indomethacin prophylaxis was directed by echocardiography.52
Recently, the prophylactic administration of indomethacin in extremely premature infants (i.e., those born between 23 and 24 weeks of gestation) decreased the incidence of symptomatic PDA without increasing the incidence of adverse effects.53
In a study conducted in Asia, the incidence of IVH and other episodes of bleeding were significantly higher than in the controls, and the study was stopped early.54
Lower platelet counts have been associated with a higher failure rate of indomethacin-induced PDA closure in human newborns, which is pertinent to clinical practice.55
Additionally, administering cortisol to immature fetal lambs in utero results in a ductus that responds to oxygen and prostaglandin inhibition similar to that of a mature fetus, which explains the decreased incidence of PDA in preterm humans who are born to mothers who received antenatal corticosteroids.56,14
Two commentary articles in the Journal of Pediatrics
emphasized that the consequences of treatment for PDA, including the pharmacological side effects and surgery, might be more harmful to the infant than the PDA itself.57,19
Moreover, there has been concern about the use of NSAIDs and their long-term renal effects in ELBW infants. In fact, ELBW infants treated with NSAIDs and aminoglycosides have been found to have a renal volume less than the 10th percentile at 7 years of age in 40% of cases, which is associated with α1 microglobinuria.58
Recently, Evans has adopted a new, targeted refinement of prophylactic indomethacin, which consists of giving indomethacin for early post-natal duct constriction assessed echocardiographically. The author has also cited the ongoing Australian DETECT trial, which includes infants born before 29 weeks of gestation who, before being treated (within the first 12 hours and, ideally, in the first 6 hours), were examined using an echocardiogram to assess the duct diameter. In this trial, preterm infants with well-constricted ducts were not treated.59