Patent ductus arteriosus (PDA) is one of the commonest congenital heart defects caused by persistent communication between the descending aorta and the pulmonary artery that results from failure of normal physiologic closure of the fetal ductus [1]. In preterm newborns the ductus frequently fails to close and as a result, 70% of babies born before 28 weeks postmenstrual age (PMA) have PDA that require medical or surgical closure [1, 2].

The immature ductus arteriosus has higher sensitivity to the vasodilating effects of prostaglandins and nitric oxide [3]. PDA can be treated surgically or medically with one of prostaglandin (PG) inhibitors like Indomethacin or Ibuprofen[4]. Inhibition of PG synthesis by non-selective blockers of both cyclooxygenase (COX) 1 and 2 is effective for the non-surgical closure [5].

Indomethacin was discredited due to association with transient or permanent derangement of renal function, NEC, gastrointestinal haemorrhage or perforation, alteration of platelet function and impairment of cerebral blood flow [6, 7]. Ibuprofen, a non-selective COX inhibitor, is as effective as Indomethacin in closing a PDA with reduced risk of NEC and transient renal insufficiency [8].

Another NSAID (a non-steroidal anti-inflammatory drug), Mefenamic acid, has also been reported to close a PDA[9], but no reliable randomised controlled trials have been reported. Various complications associated with Indomethacin and possibly Ibuprofen have encouraged the search for alternative drugs to treat a PDA non-surgically.

Since the late nineties only, the role of Paracetamol against prostacyclin-synthesis was known [10, 11], but it was suggested as an acceptable alternative drug for PDA closure in the current decade only after certain observations verified its dose dependent effectiveness [2, 12]. Paracetamol seems to act at the peroxidase segment of PG-H2 synthetase while Ibuprofen acts on the cyclooxygenase enzyme.

Peroxidase is activated at 10 fold lower peroxide concentration than is cyclooxygenase, therefore paracetamol mediated inhibition is mediated at reduced local peroxide level concentrations (as the condition with hypoxia). Many studies mainly including small randomized trials or case series used

Final inclusion in data-analysis was considered for Echo-proven hsPDA cases only, even though the initiation of the drug was based on the above clinical criteria most of the time.

Exclusion criteria: PDA suspects in term >37weeks, structural PDA-dependent congenital heart disease, prior treatment with prophylactic Indomethacin, significant hyperbilirubinemia requiring exchange transfusion

Therapeutic drug regimens (primary interventions used for PDA closure):

Group A - Preterm newborns kept nil per oral (not on oral / tube feeding)

• IV Paracetamol: 15 mg/kg/day 6 hourly for 3days

Group B - Preterm newborns orally allowed (on oral / tube feeding)

• Oral Ibuprofen: 10mg/kg/day on day 1, 5 mg/kg/day for day 2 and day 3

Secondary interventions: After the failure of the first course of initial intervention drug (IV PCM or oral Ibuprofen), for the second course, the drug was crossed over if no enteral route contraindication:

1. Intravenous Paracetamol after oral Ibuprofen
2. Oral Ibuprofen after Intravenous Paracetamol (if patient stabilized and orally allowed)

Intravenous Paracetamol (as the second course) after Intravenous Paracetamol (if the first course failed and the baby was yet sick, kept NPO)

**If the patient was kept NPO, repeat the course of 3days i/v Paracetamol as the secondary intervention was started PCM group.

Out of 1337 preterm newborns admitted during the study period of 1.5 years in the NICU of study setting, 43 cases had clinically suspected PDA and were initially enrolled for echocardiography confirmation of PDA, of which 30 cases (with clinically symptomatic or hs-PDA by 2D-echo) were selected for study drug-treatment allocation after excluding 13 patients due to various reasons (e.g. echo could not be done or it could not confirm PDA before and/or during first 3-day course of therapy, echo s/o CHDs with duct dependent lesion or patient died before the completion of the first course of treatment.

Table 1 below shows that amongst 30 PDA cases, female newborns were 16 (53.33%) in number comparable to male preterms (n=14, 46.67%) and outborn babies were more than half (n=17, 56.67%) of total PDA cases.

Table 1: Gender-distribution and place of birth of the study subjects

Table 2 shows that 16 newborns (53.33%) amongst preterms presented with PDA had extreme prematurity with < 28 weeks of gestational age, while 20% were of 28-32 week, 16.67% of 32-34 week and only 10% were 34-36 weeks late preterms. Table 2 also shows that 56.67% of cases were ELBW (extremely low birth weight, < 1000 g), 33.33% cases were Very Low BW (between birth weight 1000 -1499 g), remaining 10% cases also had low birth weight between 1500-2500 g. Mean gestational age and birth weight of these preterm subjects was 30.54 week and 1137g respectively. Thus, the incidence of PDA is much higher in lower gestational age and ELBW groups.

Table 2: Gestational age and birth weight of study subjects

But as table 5 reveals, out of those 5 preterm patients, only one patient was benefited with PDA closure by i/v PCM (after failed Ibuprofen therapy); while the rest four cases had certain complicating co-morbidities like sepsis, shock, pulmonary bleed, NEC (necrotizing enterocolitis, only in an Ibuprofen recipient preterm) etc. and actual PDA closure could not be documented as they succumbed before completion of second-line treatment intervention.

Finally, no definitive conclusion could be obtained on the actual closure rate with two study drugs after the secondary intervention, as actual cross-over of drugs was not feasible and ultimately all 05 non-responders from both study groups received i/v PCM only. Most of these cases (4 of total 5) either succumbed to critical illness or left against advice and could not complete the second course of 3-days therapy, so we had to analyze the primary outcome of PDA closure rate in two study arms after the first course of the drugs (as shown in table 4).

Table 6 shows the side-effect profile in the two treatment groups. In the i/v Paracetamol group, 01 cases each of necrotizing enterocolitis and pulmonary hemorrhage was observed, and in 07 newborns, significant hyperbilirubinemia requiring phototherapy was documented. In the oral Ibuprofen treatment group, 01 cases each of gastrointestinal bleeding, necrotizing enterocolitis (NEC), and pulmonary hemorrhage was observed along with 03 cases of hyperbilirubinemia.

Table-6: Side-effects profile of iv PCM and oral Ibuprofen treatment group

Thus, although the incidence of clinical PDA in this hospital was around 11 per 1000 live preterm births, 2D-echo confirmed PDA cases finally analyzed as inborn cases were 13 out of 1878 (i.e. 7 per 1000 inborn preterm births) and outborn admissions contributed more than half of study subjects (17 out of total 30 patients). The actual incidence can only be calculated by a timely screening of all intramural deliveries and here in this study, it was very lower even amongst preterm newborns, when compared to other studies [19-21], which would have included all symptomatic and asymptomatic PDA detected on bedside echocardiographic screening of all preterms.

El Hajjar M et al. concluded that the incidence of PDA in preterm neonates is far greater, with reports ranging from 20%–60% (depending on the study population and diagnostic criteria)[19], but such high data might have been extracted as PDA cases amongst the extreme preterms which required admission to NICUs.

As per Hoffman JI et al. the incidence of PDA in preterm infants was estimated to be 55% if screened at birth in the first 24 hours [20] and it might persist in as low as 57 per 100 000 live births after the second week of life due to functional followed by structural closure [21]. Schneider DJ M et al. reported that the incidence of PDA in term neonates is only 1 in 2,000 births, and accounting for 5%–10% of all congenital heart disease [22].

Table 8 representing patient characteristics of few previous studies indicates that most of the earlier studies published were case series based observational studies or RCTs with smaller sample size and mainly including more premature and lower birth weight babies, such selection criteria may be the reason for the higher prevalence of PDA in their settings.

Table-8: The sample size and patient characteristics in various studies

paracetamol and the ibuprofen groups in terms of a failure of PDA closure (the quality of evidence was moderate grade). [13,24,31,32,33]

In our index study, the all-cause mortality rate even amongst preterms with open ductus arteriosus was not very high (10%), as only 3 patients died (out of all 30 study subjects) and 1 of clinically improving patient in primarily i/v PCM group, left against medical advice for personal reasons before completing 2^nd course.

Thus, 9.52% of deaths in the i/v PCM group and 11.11% in the oral Ibuprofen group suggested a statistically insignificant difference in mortality between the two groups (p= 0.291). Similarly comparable mortality outcome has been reported by a metanalysis (n = 272 infants) of other three old studies (Al Lawama 2017; Dang 2013; Oncel 2014) as mentioned in table 9 [13,24,31].

Another recent study by Sunil B et al (2018) showed that paracetamol is effective on ductal closure with minimal side effects. Out of 36 preterm infants with hemodynamically significant PDA (hs-PDA) treated with intravenous paracetamol, PDA closure was achieved in 27 babies (75%) [16]. Adriansyah R et al. had also evaluated the effect of intravenous paracetamol and found a 65.5% PDA closure rate at day 14 of evaluation [17].

Tekgunduz KS et al (2015) investigated the use of safe and effective dosing (15 mg/kg/dose) of i/v

Paracetamol appears to be a promisingly effective alternative to Ibuprofen for medical closure of PDA with fewer documented adverse effects over and above theoretical benefit of better safety and tolerability profile of i/v PCM.

For early medical closure of PDA in preterm newborns, intravenous Paracetamol is a cheap, easily available and equally effective alternate to oral Ibuprofen, with a relatively better safety profile.

The intravenous route seems to be more feasible and preferable especially amongst extremely premature and clinically sick babies, who usually can’t tolerate oral feeds as well as any drugs.

This study supports and strengthens the favourable role of intravenous paracetamol as a first-line drug for medical closure of PDA as a potentially effective, cheap and safe alternative to oral Ibuprofen, which itself is most widely used due to easy availability and less toxicity than Indomethacin.

Rathia SK conceptualized and helped in protocol writing, analyzed data, prepared and finalized the manuscript. Kurrey VK helped in conceptualization (as a guide), protocol writing, data analysis and manuscript writing, and will be the principal corresponding author.

Shrivastava S helped during protocol writing, concept build-up, and patient enrolment by screening & confirmatory echocardiography. Gupta AK wrote the protocol, recruited patients and helped in data analysis and manuscript writing.

Phuljhele S supervised the study and helped in protocol writing and finalizing the manuscript. The final manuscript was approved by all authors.

We acknowledge the institute’s neonatal intensive care unit, its medical and nursing staffs for special help on patient screening, monitoring, and routine care.