TF Yeh

Abstract

Patent ductus arteriosus (PDA) occurs frequently in premature infants. The overall incidence of clinical PDA reported in premature infants ranges from 18 to 80%. However, the incidence of clinical and subclinical PA detected by echocardiography can be as high as 90%. Significant PDA L→R shunt can complicate the clinical course of RDS, therefore significant PDA requires medical or surgical intervention. Although pharmacological or surgical closure of the ductus may be required, in some infants PDA may close spontaneously without specific therapy. Management of subclinical PDA remains controversial. Routine prophylactic use of indomethacin, either for prevention of PDA or IVH cannot be recommended at the present time until more follow up studies are available. Surgical ligation of the ductus may be indicated in infants who do not react to medical treatment, particularly infants of extreme low birth weight. This paper summarizes the diagnosis and the current management. Emphasis has been placed on pharmacological closure of the ductus using the prostaglandin inhibitor.

Patent ductus arteriosus (PDA) has become an important and challenging issue to neonatologists and pediatric cardiologists not only because of its high prevalence in premature infants but also as a complicating factor in the disease of such infants that may contribute to significant mortality and morbidity. This paper will summarize the update clinical information and focus on the pharmacological management of the ductus.

Incidence

The incidence of PDA is related to gestational age and birth weight of the infants; the younger the gestational age and the lower the birth weight, the higher the incidence of PDA.^1,2 The incidence of PDA is also related to the severity of the underlying pulmonary disease; infant with severe respiratory distress syndrome (RDS) who requires mechanical ventilation is often associated with a PDA.² Other factors, such as fluid overload, improper use of mechanical ventilation, hypoxia, acidosis and postnatal use of surfactant are reported to be associated with higher incidence of PDA.³

Based on aortic contrast echocardiography during the first day of life in premature infants with RDS, the incidence of PDA could be as high as 90%.⁴ Based on a color Doppler, the incidence of PDA in infants of 600-1250 gm during the first day is 86%; nearly 90% of these PDAs are moderate or large in size.⁵ However, based on a clinical prospective study, the incidence of clinically significant PDA in premature infants with RDS ranges from 6.7 to 51%, depending on the gestational age and birth weight of the infant (Table I).³ The presence of RDS is often associated with an increased frequency of significant PDA (Table II).³

Natural Course of PDA

The natural course of PDA has been prospectively studied by Reller et al.⁶ Using an echocardiographic, color-flow Doppler study, Reller demonstrated that spontaneous closure of the ductus arteriosus occurred in essentially all healthy premature infants, regardless of gestational age, by the fourth day of life.⁶ This timing of closure is similar to that in healthy term infants.⁶ In premature infants with uncomplicated RDS, majority of their ducti would close by the fourth day of life, and only about 11% of the ducti would remain open.⁷ The infants enrolled in Reller's study were large premature infants and had relatively mild RDS.⁷

Physiology and Pathophysiology

Functional closure of the ductus occurs by the fourth day of postnatal life is nearly complete in all healthy term and preterm infants. However, many factors may influence the constriction or relaxation of the ductus muscle. Among these, the alveolar and arterial oxygen tension and prostaglandin E (PGE2) appear to be the most important. It has been shown that during intrauterine life circulating PGE2 and PGI2 produced locally in the ductus maintain the patency of the ductus^8,9 and that following birth, the loss of the placenta, a source of PGE2 production, and the spontaneous decrease in local production of PGE2 and PGI2, allow factors that constrict the ductus, such as the increase in oxygen tension, to act relatively unopposed, and the ductus closes. The immature ductus is more sensitive to the dilating effects of prostaglandins and less sensitive to the constricting effects of oxygen,¹⁰ which accounts for the high incidence of PDA in preterm infants. Steroid may decrease the sensitivity of the ductus to PGE.¹¹ Prenatal steroids therapy may decrease the incidence of PDA.¹²

Circulating prostaglandins (PG) also may be important in maintaining the patency of the ductus arteriosus. During the acute stage of RDS, both PGE and PGF levels were elevated signiflcantly as compared with values in infants without RDS.¹³ Concomitant with the appearance of PDA there was a decrease in PGF and an increase in PGE.¹³ Because the lung is an important organ for synthesis and degradation of PG, it is possible that in premature infants with RDS, the synthesis and degradation of PG may be altered and that a resultant high circulating PGE may exert an effect on the ductus.

The pathophysiology of PDA in infants with RDS is illustrated in the Figure 1. Following the decrease in pulmonary vascular resistance, the left to right shunt through the ductus increases, resulting in increased pulmonary flow and a congestive circulatory state. This may further compromise the poor lung compliance already present in infants with RDS.¹⁴ The consequent increases in ventilatory and oxygen therapy may further enhance the risk of developing chronic lung disease (CLD). Alterations in dynamics of cerebral circulation by the ductal shunt may increase the risk for intraventricular hemorrhage (IVH). Increased incidence of necrotizing enterocolitis has also been reported.

Fig. 1 Pathophysiology of respiratory distress syndrome (RDS) and patent ductus arteriosus (PDA). CL=Lung compliance; V/P=Ventilation: perfusion ratio. From Yeh and Carr (1991, Neonatal Therapeutics. Chicago: CV Mosby-Year Book).

Clinical Manifestation

Patent ductus arteriosus in premature infants can be subclinical (no heart murmur) or clinical (with heart murmur). Clinical PDA can be nonsignificant (no cardiovascular dysfunction) or significant (with cardiovascular dysfunction).

Subclinical PDA

In the absence of a ductus murmur, PDA can only be diagnosed noninvasively by contrast or Doppler echocardiography. However, there are factors that may be associated with high risk of developing ductus shunt. These include (a) birthweight < 1500g, (b) RDS, (c) acute perinatal stress, and (d) the need for assisted ventilation within 24 hours of birth.¹⁵

Clinical PDA without Significant Cardiopulmonary Dysfunction

A heart murmur is probably the only or the only prominent sign in these infants. In infants who have no underlying lung disease, the diagnosis of PDA can easily be made by the characteristic systolic murmur at the left sternal border in the second intercostal space. The systolic murmur often continues into the diastolic phase. In infants with RDS who require mechanical ventilation, the diagnosis of PDA can be confounded by signs associated with the underlying respiratory condition. Temporary disconnection of the respirator may make it easier to detect the cardiac murmur. Awareness of the high prevalence of PDA in infants with RDS during the first week of life and frequent careful physical examination are the initial steps leading to the diagnosis. Diagnosis can easily be made clinically and confirmed by doppler echocardiography. These infants usually have normal blood gases and acid-base balance, without apparent alteration in cardiopulmonary function.

Clinical PDA with Significant Cardiopulmonary Dysfunction

Besides the cardiac murmur, these infants may have resting tachycardia, bounding peripheral pulse, hyperactive precordium, cardiomegaly on chest X-ray and echocardiographic changes suggesting a large left-to-right shunt. Frequent apnea, carbon dioxide retention, unexplained deterioration of respiratory status or respirator dependence and systemic hypotension may be seen. A clinical score system designed by Yeh et al has been used to evaluate the signs of cardiovascular distress (cardiovascular distress or CVD score) (Table 111).¹⁶ This scoring system is essentially designed for a paediatrician in practice. There is a good correlation between this CVD score and blood gases and echocardiogram. A CVD score of 3 or greater is usually associated with LAIAO (left atrium/aortic root dimension) >= 1.3 on echocardiogram, indicating a significant ductus shunt (normal LAIAO values range from 0.66-1.06).

Echocardiography can be used not only for the diagnosis of PDA but also for estimating the degree of ductus shunt. An increase in LAIAO ratio (>= 1.3) suggests a significant left-to-right shunt. High left ventricular function index (as shown by left ventricular shortening fraction) indicates a left-to right shunt with high left-sided heart volume load and excludes causes of dilatation of the left side of the heart associated with decreased left ventricular function, such as cardiomyopathy. The shortening fraction is expressed as (EDD-ESD)/EDD x 100, where EDD is the end diastolic dimension and ESD is the end systolic dimension. The normal value for shortening fraction of left ventricle is 34±3%.

Other echocardiographic changes include decrease in the left ventricular pre-ejection period (LVPEP), decrease in the ratio of LVPEP to left ventricular ejection time, and diastolic pulmonary valve flutter. By using two-dimensional echocardiography, direct visualization of the PDA can be obtained, with few false negative or false positive studies. The contrast echocardiographic technique with injection of 2-5 ml of normal saline solution through the umbilical arterial catheter has been successful in demonstrating the left-to-right PDA shunt. Doppler colour-flow echocardiography can visualize a small PDA that cannot be imaged clearly on the two-dimensional echocardiogram, and it can obtain more information of the degree of ductus shunt and pulmonary artery pressure. With continuous wave Doppler techniques, the aortic to pulmonary artery pressure gradient can be measured with accuracy and the pulmonary artery pressure can be estimated using pulse Doppler techniques, by comparing (area x mean velocity of the aortic valve) to (area x mean velocity of the pulmonary valve).

A significant PDA can be diagnosed if there is clinical cardiopulmonary dysfunction or if there are echocardiographic changes suggesting a significant left to right shunt (Table IV).

Management

Although pharmacological or surgical closure of the ductus may be required, a PDA may close spontaneously without specific therapy. This is particularly true in large premature infants without severe underlying pulmonary disease.

Subclinical PDA

The management of a subclinical PDA remains controversial. Various studies by giving indomethacin within 24 hours of birth have shown a significant decrease in symptomatic PDA. This mode of prophylactic therapy has been routinely used in some intensive care units.¹⁷ A recent systematic review¹⁸ of 14 studies suggests that early prophylactic use of indomethacin also significantly reduced the incidence of symptomatic PDA but there is no evidence that treatment affects respiratory outcome. Prophylactic indomethacin significantly reduces the incidence of grades 3 and 4 intraventricular hemorrhage. However, there is no sound study assessing the longterm effect of prophylaxis on neurodevelopmental outcome. In view of the potential side effects of indomethacin in terms of reducing cerebral blood flow and the risk of cerebral hypoxia¹⁹ and the lack of long term study on neurodevelopmental outcome, routine prophylactic indomethacin can not be recommended at the present time.

Clinical PDA

The management of a clinical PDA is less controversial. For infants who have clinical PDA, with or without cardiovascular dysfunction, several clinical interventions may facilitate ductus closure or improve the cardiovascular status. These include the following³:

1. Fluid restriction to 80-100 ml/kg per day (100-120 ml/kg per day if under phototherapy).

2. Maintenance of haematocrit values >= 40% with packed-cell transfusion.

3. Ventilatory adjustment to keep arterial PaO₂ at 50-90 mmHg, PCO₂ <= 45 mmHg and pH > 7.25.

4. Maintenance of normal electrolytes and of serum Ca >= 1.8 mmol/l (7.5 mg/dl).

5. Treatment with furosemide (1 mg/kg i.v. every 12-24 hours for one or two doses) if there is evidence of fluid overloading or circulatory congestion.

6. Treatment with indomethacin or a combination of indomethacin and furosemide.

Furosemide

Furosemide (Lasix) is the diuretic most commonly used in neonates for various reasons. Furosemide produces diuresis and natriuresis by inhibition of active chloride reabsorption from the ascending limb of the loop of Henle', resulting in a decrease in passive reabsorption of sodium. Furosemide may stimulate renal prostaglandin synthesis.²⁰ The haemodynamic effect of furosemide may be mediated by renal PG. Furosemide also has systemic vascular effects that are independent of its diuretic action. The mechanism for the reduction of pulmonary edema in premature infants with PDA could be related to diuresis or to the vascular effect of the drug, or both. Furosemide can be recommended in infants with fluid overload, whether or not they have congestive heart failure. Furosemide also can be used simultaneously with indomethacin to prevent its renal side-effects.²¹ For rapid diuresis in infants with congestive heart failure, a dose of 1-3 mg/kg iv. is adequate. For infants with fluid overload and PDA, we recommend one or two doses of Furosemide (1 mg/kg) at 12-24 hour intervals.

The two potential side-effects of short-term furosemide therapy in neonates are ototoxicity and displacement of bilirubin from albumin-binding sites. Ototoxicity appears to be related to the serum concentration of furosemide, but this has not been reported in neonates. Similarly, the displacement of bilirubin from albumin binding sites has not been reported to cause any significant problem in neonates. Controversy remains as to whether furosemide promotes patency of the ductus. A randomized controlled study did not show an increase in PDA incidence with early use of furosemide.²²

Pharmacological Closure of PDA with PG Inhibitor

Prostaglandin E is synthesized in the wall of the ductus arteriosus, and circulating PGE may play an important role in maintaining patency of the ductus during fetal and early extrauterine life. Indomethacin, a PG synthetase (cyclo-oxygenase) inhibitor, has been used to promote ductus closure.^23,24

Indications

Indomethacin may be used for prophylaxis or for treatment of PDA. However, routine prophylactic therapy for very low birth weight infants has been controversial because of inadequate studies on the side-effects and long-term outcome. For infants who have clinical PDA, with or without significant cardiovascular dysfunction, administration of indomethacin will reduce the ductus shunt or close the ductus and improve cardiopulmonary status.

Contraindications

The tentative contraindications for indomethacin therapy are (a) hyper-bilirubinaemia, (b) blood urea nitrogen levels exceeding 3.3 mmol/l (20 mg/dl), (c) shock, (d) intracranial haemorrhage, (e) necrotizing enterocolitis, (f) haemorrhagic disease, and (g) a platelet count less than 50000/mm.³

Efficacy of Indomethacin Closure of PDA

Based on 7 control trials,¹⁸ prophylactic use of intravenous indomethacin given within 24 hours of birth significantly reduced the incidence of symptomatic PDA (PDA associated with signs/symptoms of congestive heart failure). (Pooled event rate ratio, ERR=0.309, 95% CI 0.215, 0.443; event rate difference, ERD= -0.217, 95% CI -0.275, -0.160). The incidence of echo-diagnosed PDA-that is, the sum of clinical and subclinical PDA is reduced even further by prophylactic indomethacin (Pooled ERR= -0.304, 95% CI 0.220, 0.417; pooled ERD=-0.289, 95% CI -0.353, -0.225). The prophylactic dosage of indomethacin ranges from 0.1 to 0.2 mg/kg every 24 hours for 3 doses.

Reports on the use of indomethacin in premature infants with clinically significant PDA have been favourable, but indomethacin has not been shown to be beneficial in all cases. Indomethacin may have been less effective in (a) premature infants of very low birth weight and very low gestational age, (b) infants of high postnatal age (more than 6 weeks), and (c) infants of high postconceptional age. Because of the poor bioavailability, indomethacin given by the oral route is less effective than that given intravenously. Based on a double-blind controlled study in which indomethacin was given at a mean postnatal age of 8.9 days and with dosage of 0.3 mg/kg per day for a maximum of three doses, 89.2% of infants would show ductus closure and/or improvement in cardiovascular status (Figure 2).² From this study, 22% of the infants would show spontaneous closure of the ductus. When adjustments were made for the number of infants expected to improve spontaneously, indomethacin was successful in closing the ductus in 86% of the infants. This success rate was consistent with the report from a national collaborative study.²⁵

Fig. 2 Clinical, echocardiographic and murmur changes evaluated approximately 24h after the last dose of indomethacin or placebo. These changes were divided into four categories. "Success" was achieved if both positive clinical and echocardiographic responses were observed. Modified from Yeh et al.2

When indomethacin is given to infants within 2 weeks of birth, after one dose (0.3 mg/kg), about 50% of the infants show ductus response, after two doses (in 24 hours) 75%, and after three doses (in another 24 hours) 89%. After one dose of indomethacin, 14% of the responding infants had return of the ductus murmur, after two doses 9.5%, and after three doses, 8%. The smaller the infant, the higher the incidence of returning of the ductus murmur.²

It is not completely understood why some infants did not respond to indomethacin and some infants had their murmur returning following initial ductus closure. Some explanations could be derived from the pharmacokinetic data of indomethacin.^26,27 Series studies suggested that the ductus response to indomethacin correlated significantly with plasma concentration.^28,29 Brash et al28 have shown that the ductus response to indomethacin is related to plasma concentration at 24 hours after the dose is given; failure of the ductus to respond was often associated with a plasma level of less than 250 ng/ml. A study by Yeh et a129 demonstrated that ductus response to indomethacin correlated significantly with plasma level at 12 hours after intravenous dosing: when the blood levels reach 600 ng/ml, there is a 50% chance that ductus will close, and when the blood levels reach 1400 ng/ml, there is a 75% chance that this will happen (Figure 3).²⁹ Low plasma levels could be seen in infants of advanced postnatal age, in whom the half-life of indomethacin is short, or in infants of very low birthweight, in whom the volume of distribution of indomethacin is large.²⁷

Fig 3. Incidence of ductus closure, decreased urine output (U/O), hyponatraemia and hyperkalaemia, as a function of plasma indomethacin concentration at 12h after a single dose.

The pharmacokinetic data of intravenously administered indomethacin in infants >= 2 weeks postnatal age²⁶ showed a mean plasma half-life ranging from 15.4-20.7 hours, which is about three times longer than that reported in adults. The plasma clearance of indomethacin was positively correlated with postnatal age: the older the infant, the faster the clearance. Indomethacin has been shown in adults to undergo demethylation, deacylation and conjugation with glucuronic acid. Variability in hepatic metabolism, renal excretion and enterohepatic recirculation could contribute to differences in pharmacokinetics between newborns and adults.²⁶

Side-Effects

The possible side-effects being reported following indomethacin therapy are³:

1. Transient renal dysfunction;
2. Hyponatraemia;
3. Reduction in cerebral blood flow;
4. Transient decrease in plasma glucose level;
5. Decreased platelet aggregation;
6. Increased risk for gastrointestinal haemorrhage;
7. Decreased mesenteric blood flow;
8. Increased risk of necrotizing enterocolitis;
9. Gastric perforation;
10. Increased risk of retinopathy of prematurity;
11. Displacement of bilirubin from binding sites.

Among these side-effects, the most clinically significant and recognizable complication is transient renal dysfunction. A transient decrease in urine output, fractional excretion of sodium and chloride and osmolar and free water clearance is observed following indomethacin administration.^30,31 Renal function may return to normal, usually within 72 hours after discontinuation of the medication. Indomethacin may decrease the glomerular filtration rate. Since the decrease in free water clearance is usually greater than that of osmolar clearance, a dilutional hyponatraemia may occur.³¹ There was no apparent linear correlation between the plasma indomethacin levels and renal side-effects. Seyberth et al³² have demonstrated altered renal function when the plasma indomethacin levels were maintained between 150 and 750 ng/ml. A study by Yeh et al²⁹ indicated that when plasma indomethacin levels reach 170 ng/ml, the majority of infants have decreased urine output. This plasma level is far below the therapeutic levels for ductus closure. Thus, the safe therapeutic range of plasma levels within which the ductus will close and side-effects be minimal appears to be narrow (Figure 3).²⁹ In other words, to reach the therapeutic levels for ductus closure, majority of the infants would experience some degree of renal dysfunction. Fortunately, the renal side-effects are often transient and usually do not last for more than 72 hours after discontinuation of the drug.

The renal side-effects of indomethacin may be prevented by simultaneous administration of furosemide.²¹ Furosemide may induce renal PG synthesis. Simultaneous administration of furosemide (1 mg/kg) promotes free water clearance and osmolar clearance, which otherwise would be reduced if indomethacin was used alone (Figure 4). Although furosemide has been shown to enhance ductus patency, we did not observe any significant change in the efficacy of indomethacin on the closure of PDA with the simultaneous administration of furosemide. The simultaneous administration of indomethacin and furosemide is particularly useful in premature infants who have PDA, congestive heart failure and oliguria.³³

Fig. 4 Comparison of fractional excretion of sodium (FENa) and chloride (FECl), glomerular filtration rate (GFR), and urine output in infants who received indomethacin alone (Gr I) with those who received indomethacin and furosemide simultaneously (Gr I + F). Infants in Gr I + F had significantly higher FENa. FECl, GFR and urine output than infants in GR I. Modified froom Yeh, et al.

Therapeutic Regimen and Recommended Dosage

Current dosing schedules are somewhat variable, depending on the therapeutic purposes and the postnatal age of the infant. In general, an i.v. dose of 0.2-0.3 mg/ kg every 12-24 hours, for a total of three doses if needed, is adequate for one course of therapy. Infants may be given a second course of therapy, usually 48-72 hours following the last dose of first course. Ductus response to second course of therapy is usually poor in infants who did not respond to first course or whose ductus reopens following initial closure. The following recommendations are based on the published literature.

1. For infants who have symptomatic PDA, dosage should be adjusted, based on postnatal age. For infants of 48 hours of age or younger, an initial dose of 0.2 mg/kg followed by 0.1 mg/kg every 24 hours for a total of three doses can be recommended. For infants between 48 hours to 2 weeks, a dose of 0.2-0.3 mg/kg every 12-24 hours for 3 doses is adequate. For infants over 4 weeks of age, 0.3 mg/kg every 12 hours may be given. Infants whose postnatal age is 6 weeks or older are usually not responsive to indomethacin.³⁴

2. To prevent reopening of the ductus, prolonged indomethacin therapy has been given to infants weighing less than 1500 g. Doses were given as follows: 0.15 mg/kg every 12 hours for two doses initially, then 0.1 mg/kg per day as maintainance for 5 days.³⁵ Hammerman et al³⁶ conducted a study of prolong therapy and has demonstrated that a 5 to 7 days treatment significantly minimized PDA recurrences and decreased the need for surgical ligation. The prolonged therapeutic regimen, however, does not improve mortality or morbidity. Further studies are needed before it can be generally recommended.

3. To avoid the vascular side effects of indomethacin, particularly the reduction of cerebral blood flow, continuous infusion of indomethacin therapy has been proposed. Indomethacin may cause concurrent vasoconstriction of other vasoactive vascular beds, most notably the cerebral and renal vasculature, which, in turn, can produce potentially adverse effects. Indomethacin, when administered as a bolus to premature infants, may cause a 40% to 50% reduction in cerebral blood flow in human subjects.^37-40 It may also cause periventricular leucomalacia and cystic lesions in infants exposed in utero to indomethacin as a tocolytic agent.⁴¹ Since slowing the indomethacin infusion rate may alleviate the indomethacin mediated reduction in cerebral blood flow velocity,^42,43 a continuous infusion of indomethacin may be a better way of drug administration⁴⁴ in order to minimize its side effect. More studies are needed to confirm this.

4. Because of the high prevalence of PDA in very low birth weight infants, indomethacin has been given during the first day of extrauterine life for prevention of subsequent development of symptomatic PDA. Three doses were given: 0.1-0.2 mg/kg given in 24 hours interval. Again, there was no apparent improvement in morbidity or mortality with this therapeutic regimen.

Outcome and Long-Term Follow-up

There has been no solid evidence to demonstrate that the presence of PDA contributes significantly to the development of bronchopulmonary dysplasia (BPD). Similarly, there has been no solid evidence of a decrease in the incidence of BPD following surgical or indomethacin closure of the ductus arteriosus. However, indomethacin closure of the ductus does decrease the incidence of surgical ligation and the need for assisted ventilation.

The use of indomethacin did not show long-term adverse effects, e.g. increased incidence of recurrent respiratory infection, neurologic defects (major and minor) or abnormal electroencephalogram when the infants were examined at 1 year and at preschool age.^45,46

Digoxin

The use of digoxin for PDA has been questioned by many paediatric cardiologists and neonatologists. A control study by McGrath⁴⁷ failed to show any advantage of using digoxin in premature infants with PDA. Furthermore, premature infants treated with digoxin have a high incidence of arrhythmia and subendocardial ischaemia. The poor elimination of the drug may result in high levels in the serum, leading to toxicity. Thus, digoxin should be reserved only for infants in whom indomethacin is contraindicated, such as those with necrotizing enterocolitis or renal failure. Because of the prolonged half-life when indomethacin is added to digoxin therapy, the digoxin dosage should be reduced by 50% until urine output and digoxin serum level can be better assessed.⁴⁸ The relevant action of digoxin is the positive inotropic action. This action is applicable only when there is ventricular myocardial dysfunction. Unless echocardiography demonstrates dysfunction, digoxin is not indicated.

Surgical Closure of PDA

It is generally acknowledged that closure of a haemodynamically significant PDA is of benefit in premature infants. However, controversy still exists on whether surgical ligation or indomethacin closure of the ductus represents the optimum method of management. This is particularly true in very low birthweight infants, since this group of infants is often refractory to indomethacin therapy.²⁷ Palder et al⁴⁹ showed that surgical ligation of a haemodynamically significant PDA yields a more predictable result, with low morbidity and mortality, and suggested that this mode of therapy may be a preferred treatment for premature infants of birthweight less than 800 g. Surgical ligation can be accomplished by an experienced surgeon in a neonatal intensive care unit (NICU) with minimal mortality and acceptable morbidity. Surgical ligation of PDA in NICU prevents the complications often observed in the operation room, such as hypothermia, inadvertent extubation and interruption of vascular access and monitoring.

However, surgical ligation is not without risks. There are complications directly or indirectly related to operative precedure. These include pneumothorax, pleural effusion (serous or chylous), recurrent nerve and phrenic nerve injury and excessive operative blood loss. We feel that for premature infants of birthweight greater than 1000 g, surgical ligation should be reserved only for those who fail to respond to two courses of indomethacin therapy. However, for premature infants of extremely low birthweight (less than 1000g), prompt surgical ligation may be considered if they fail to respond to one course of indomethacin therapy or there are contra-indications to indomethacin therapy.

Acknowledgement

This paper is revised from T F Yeh, et al "Patent ductus arteriosus" in Bailliere's Clinical Pediatrics, 3(1):131-46.

References

1. Siassi B, et al. Incidence and clinical features of patent ductus arteriosus in low birth weight infants: a prospective analysis of 150 consecutively born infants. Pediatrics 1976;57:347-51.

2. Yeh TF, Luken J, Thalji A, et al. Intravenous indomethacin therapy in premature infants with persistent ductus arteriosus: a double-blind controlled study. J Pediatr 1981a;98:137-45.

3. Yeh TF, Can I. Pharmacological closure of PDA. In: Yeh TF (ed.) Neonatal Therapeutics 1991:123-38. Chicago CV Mosby and Year Book.

4. Dudell GG, Gersony WM. Patent ductus arteriosus in neonates with severe respiratory distress. J Pediatr 1984;4:915-20.

5. Couser RJ, Ferrara TB, Wright GR, et al. Prophylactic indomethacin therapy in the first twenty four hours of life for the prevention of patent ductus arteriosus in preterm infants treated prophylactically with surfactant in the delivery room. J Pediatr 1996;128:631-7.

6. Reller MD, Ziegler ML, Rice MJ, et al. Duration of ductal shunting In healthy premature infants: an echocardiographic color flow Doppler study. J Pediatr 1988;22:1171-4.

7. Reller MD, Colasurdo MA, Rice MC, McDouall RW. The timing of spontaneous closure of the ductus arteriosus in infants with respiratory distress syndrome. Am J Cardiol 1990;66:75-8.

8. Clyman Rl. Ontogeny of the ductus arteriosus. J Pediatr 1979;95:583-7.

9. Coceani F, Olley PM. Role of prostaglandins, prostacyclin and thromboxanes in the control of prenatal patency and postnatal patency and postnatal closure of the ductus arteriosus. Seminars in Perinatology 1980;4:109-13.

10. Clyman Rl. Heymann MA. Pharmacology of the ductus arteriosus. Pediatr Clin North Am 1981;28:77-93.

11. Clyman Rl, Mauray F, Roman C, et al. Glucocorticoids alter the sensitivity of the lamb ductus arteriosus to prostaglandin. Eur J Pediatr 1981;98:126-8.

12. Clyman Rl, Ballard PL, Sinderman S, et al. Prenatal administration of betamethasone for prevention of patent ductus arteriosus. J Pediatr 1981b;98:123-6.

13. Friedman WE. In Session II: Physiology and pharmacology of the ductus arteriosus: studies of the responses of the ductus arteriosus Presented at the seventy-fifth Ross conference on Pediatric Research: The Ductus Arteriosus 1977:4-7 December Florida.

14. Yeh TF, Thalji A, Luken L, et al: Improved lung compliance following indomethacin therapy in premature infants with persistent ductus arteriosus. Chest 198 lb,80: 698-700.

15. Cotton RB, Lindstrom DP, Strahlman M. Early prediction of symptomatic patent ductus arteriosus from perinatal risk factors: discriminant analysis model. Acta Paediatr Scand 1981 ;70:723-7.

16. Yeh TF, Raval D, Luken J, et al. Clinical evaluation of premature infants with patent ductus arteriosus: a scoring system with echocardiogram. Acid-base, and blood gas correlations. Crit Care Med 1981c;9:655-7.

17. Cotton RB, Haywood JL, Fitzgerald GA. Symptomatic patent ductus arteriosus following prophylactic indomethacin. Biol Neonate 1991;60:273-82.

18. Fowile PW. Prophylactic indomethacin: systematic review and meta-analysis. Arch Dis Child 1996;74:F81-7.

19. Edwards AD, Wyatt JS, Richardson C, et al. Effects of indomethacin on cerebral hemodynamics in very preterm infants. Lancet 1990;335:1491-5.

20. Attallah M. Interaction of prostaglandins with diuretics. Prostaglandins 1979;18:369-75.

21. Yeh TF, Wilks A, Singh J, et al. Furosemide prevents the renal side effects of indomethacin therapy in premature infants with patent ductus arteriosus. J Pediatr 1982a;101:433-7.

22. Yeh TF, Shibli A, Len ST, et al. Early furosemide therapy in premature infants with RDS-a randomized study. J Pediatr 1984;105:603-9.

23. Friedman WE, Hirschklau MJ, Printz MP, et al. Pharmacologic closure of patent ductus arteriosus in the premature infant. N Engl J Med 1976;195:526-9.

24. Heymann MA, Rudolph AM, Silverman NH. Closure of the ductus arteriosus in premature Infants by inhibition of prostaglandin synthesis. N Engl J Med 1976;295:530-3.

25. Gersony Win, Peckham GJ, Ellison RC, et al. Effects of indomethacin in premature infants with patent ductus arteriosus: results of national collaborative study. J Pediatr 1983;102:395-906.

26. Thalji AA, Carr I, Yeh TF, et al. Pharmacokinetics of intravenously administered indomethacin in premature infants. J Pediatr 1980;97:995-1000.

27. Yeh TF, Luken J, Raval D, et al. Indomethacin treatment in small versus large premature infants with ductus arteriosus comparison of plasma indomethacin concentration and clinical response. Br Heart J 1983;50:27-30.

28. Brash AR, Hickey DE, Graham TP, et al. Pharmacolinetics of indomethacin in the neonate. Relation of plasma indomethacin levels to response of ductus arteriosus. N Engl J Med l981;305:67-71.

29. Yeh TF, Achanti B, Patel H, et al. Indomethacin therapy in premature infants with patent ductus arteriosus-determination of therapeutic levels. Dev Pharmacol Ther 1989;12:169-78.

30. Cifuentes RF, Olley PM, Balfe JW, et al. Indomethacin and renal function in premature infants with persistent patent ductus arteriosus. J Pediatr 1979;4:115-24.

31. Betkerur MV, Yeh TF, Miller K, et al. Indomethacin and its effect on renal function and urinary kallikrein excretion in premature infants with PDA. Pediatrics 1981;68:99-102.

32. Seyberth HW, Rascher W, Hackenthal R, et al. Effect of prolonged indomethacin therapy on renal function and selected vasoactive hormone in very low birth weight infants with symptomatic patent ductus arteriosus. J Pediatr 1983;103:979-84.

33. Yeh TF, Wilks A, Luken J. et al. Indomethacin therapy in premature infants with patent arteriosus and oliguria. Dev Pharmacol Ther 1986;9:369-74.

34. Achanti B. Yeh TF, Pildes RS. Indomethacin therapy in infants with advanced postnatal age and patent ductus arteriosus. Clin Invest Med 1986;9:250-3.

35. Rhdes PG, Ferguson MG, Reddy NS, et al. Effects of prolonged versus acute indomethacin therapy in very low birth-weight infants with patent ductus arteriosus. European J Pediatr 1988;147:481-4.

36. Hammerman C, Aramburo MJ. Prolonged indomethacin therapy for the prevention of recurrences of patent ductus arteriosus. J Pediatr 1990;117:771-6.

37. Cowan F. Indomethacin, patent ductus arteriosus and cerebral blood flow. J Pediatr 1986;106:341-4.

38. Evans D, Levenc M, Archer L. The effect of indomethacin on cerebral blood flow velocity in premature infants. Dcv Med Child Neurol 1987;29:776-82.

39. Hanigan W, Kennedy G, Roemisch F, et al. Administration of indomethacin for the prevention of periventricular-intraventricular hemorrhage in very low birth weight neonates. J Pcdiatr 1988;112:941-7.

40.Van Bel F, Van De Bor M, Stijnen T, et al. Cerebral blood flow velocity changes in preterm infants after a single dose of indomethacin: duration of its effects. Pediatrics 1989;84:802-7.

41. Baerts W, Fetter WP, Hop WC, et al. Cerebral lesions in preterm infants after tocolytic indomethacin. Dev Med Child Neurol 1990;32:910-8.

42. Edwards A, Wyatt J, Ridrardson C, et al. Effects of indomethacin of cerebral hemdynamics in very pretcrin infants. Lancet 1990;335:1491-5.

43. Simdo A, Mardoum R, Merritt TA, et al. Effects on cerebral blood flow velocity of show and rapid infusion of indomethacin. J Perinatol 1994;14:29-35.

44. Hainincrman C, Strates E, Koinan K, et al. Continuous vs multipcr rapid infusions of indomethacin: Effects on cerebral blood flow velocity. Pediatrics 1995;95:244-8.

45. Merritt TA, White CL, Cocn RW. et al. Pre-school assessment of infants with a patent arteriosus. Comparison of ligation and indomethacin therapy. Am J Dis Child 1982;136:507-2.

46. Yeh TF, Goldbarg HR, Henck T, et al. Intravenous indomethacin therapy in premature infants with patent ductus arteriosus: Causes of death andone;year follow-up. Am J Dis Child 1982b;135:803-7.

47. Mc Grath RL. In General Discussion, Session III: Persistent patency of ductus artcriosus in premature infants. In heymann MA & Rudolph AIVI(cds). The Ductus Arteriosus: Report of the Seventy-fifth Ross Conference on Pediatric Research, p92. Columbus, Ohio: Ross Laboratories.

48. Koren G, Zarfin Y, Perlman M, et al. Effects of indomethacin on digoxin pharmaco-kinetics in premature infants. Pediatr Pharmacol 1984;4:25-30.

49. Palder SB, Schwartz MZ, Tyson KRT, Man CC. Management of PD: a comparision of operative V pharmacologic treatment. J Pediatr Surg 1987;22:1171-4.