Patent ductus arteriosus in extremely preterm infants : characteristics, risk factors and treatment decisions

Abstract

Survival rates in infants born extremely preterm, before 28 weeks of gestation, are increasing and the focus has shifted towards decreasing morbidites after extremely preterm birth and promoting life-long health for the survivors.1-3Approximately 50-60% of extremely preterm infants have the past decade been treated for patent ductus arteriosus (PDA) to induce ductal closure.4,5In the abscence of ductal closure, hemodynamic changes follow which can result in significant systemic hypoperfusion and excessive pulmonary perfusion which has been associated with an increased risk of the neonatal morbidities such as intraventricular hemorrhage (IVH), necrotising enterocolitis (NEC), renal failure and and bronchopulmonary dysplasia (BPD).6 Cyclooxygenase (COX) inhibitors, ibuprofen or indomethacin, are used as first line of treatment to promote ductal closure.7,8 Surgical closure may be indicated if pharmacological therapy fails, or when contraindications to COX inhibitors are present.9 The theoretical rationale for PDA treatment is unquestionable, but trials showing long-term benefits from PDA treatment are scarce.10-12 In this thesis, paper I-III are cohort studies conducted in Sweden and in Europe. The study in paper IV is a hospital-cohort study in Stockholm. The overall aim of the studies is to investigate PDA incidence, neonatal characteristics associated with PDA closure or treatment, variation in treatment strategies, and association with neonatal outcomes. Furthermore, to evaluate the role of cardiac biomarkers in predicting PDA closure.

In paper I, the aim was to investigate if timing of pharmacological PDA treatment (at 0-2 days, 2-6 days or ≥7 days of age) was associated with risk of later PDA surgery or death; or risk for BPD at 36 weeks postmenstrual age (PMA). This was investigated in the population-based prospective Extremely Preterm Infants in Sweden (EXPRESS) cohort (infants born at <27 weeks of gestation during 2004-2007).13 Two hundred ninety of 585 children were treated pharmacologically, of whom 102 later underwent PDA surgery. In a model stratified on GA and adjusted for clustering on region, hazard ratios (HR) for late and intermediate vs early start were 1.10 [CI 0.53–2.28] and 0.89 [CI 0.57–1.39] respectively. Compared to early start, the risk of BPD after late start of PDA treatment was associated with a significantly lower risk of BPD odds ratio (OR) 0.29 [CI 0.13-0.61] in a model stratified on gestational age (GA) and adjusted for sex and small for gestational age (SGA).

In paper II, the aim was to investigate incidence and variation in PDA treatment and association with BPD at 36 weeks PMA or death; and survival without major neonatal morbidity. This was performed in a large European cohort (the EPICE study) of infants born at <32 weeks of gestation during 2011-2012, including 6898 infants.14 The results show that there is significant variation of 10% to 39% in PDA treatment between the different regions (p<0.001). The regions were categorized according to low (<15%, n = 6), medium (15–25%, n = 9), or high (>25%, n = 4) proportion of PDA treatment. The difference in PDA treatment could not be explained by differences in perinatal characteristics between these regions. Infants treated for PDA, compared to those not treated, were at higher risk of BPD or death in all regions, with an overall propensity score adjusted risk ratio of 1.33 [95% confidence interval 1.18–1.51]. Survival without major neonatal morbidity was not related to PDA treatment.

In paper III, the neurodevelopmental outcome after PDA treatment was studied. In the EXPRESS cohort (see paper I), the survivors at 6.5 years of age had an extensive neurodevelopmental follow-up.15 Four hundred and thirty five of 486 children had available data on both PDA treatment and neurodevelopmental outcome. PDA treatment as an exposure was categorized as no PDA treatment; pharmacological PDA treatment; PDA surgery after prior pharmacological treatment; and primary PDA surgery. The outcomes studied were NDI (by the definition of Moore16) and the full-scale intelligent quotient (FSIQ) as measured by Wechsler Intelligent Scale for Children (WISC-IV17). No increased risks of adverse neurodevelopment were found among children treated pharmacologically for PDA, regardless of whether they later had surgical PDA closure or not. The risk of moderate to severe NDI was higher among children treated with primary PDA surgery in the adjusted model than in extremely preterm children not receiving PDA treatment, IRR 1.62 [95% CI 1.28-2.06] p<0.001 and a lower FSIQ, adjusted mean difference -7.1 [95% CI -11 till -3.2] p<0.001. Timing of PDA surgery was investigated as an exposure in the children undergoing PDA surgery. Children having PDA surgery at <10 days of age (irrespective if primary or after prior pharmacological treatment), compared to at >20 days of age, had increased risk for moderate to severe NDI, IRR 3.26 [95% CI 2.40 to 4.42] p<0.001; and adjusted mean difference of FSIQ -15 [95% CI-19 till -12] p<0.001.

In paper IV, the perinatal characteristics associated with spontaneous PDA closure were investigated in a hospital-based cohort of extremely preterm infants in Stockholm. The association of the biomarkers N-Terminal fragment-pro-Brain Natriuretic Peptide (NT-proBNP) and cardiac Troponin T (cTnT) with spontaneous closure and all types of PDA treatment was investigated. Fifty-eight of 98 infants were treated for PDA with a median age at start of treatment of 8 days (interquartile range, IQR 5-11). Six (6%) infants closed their PDA at ≤7 days of age. All infants who closed their duct spontaneously were born at ≥25 weeks of gestation. Higher NT-proBNP values on day 3 were associated with later need of PDA surgery and lower levels were associated with PDA closure without any PDA treatment.

In conclusion, there is large variation in PDA treatment across Europe which is not associated with perinatal characteristics. This indicates need of standardization of diagnostics and treatment. In all four studies, the strongest predictor for PDA treatment is GA and the spontaneous closure rate in extremely preterm infants born at ≥25 weeks is relatively high. A more conservative approach with later start of PDA treatment is not associated with increased risks of morbidity and PDA treatment is not associated with decreased risk of survival without major morbidities. PDA pharmacological treatment with or without later surgery is not associated with adverse neurodevelopment. PDA surgery at <10 days of age and primary PDA surgery are associated with increased risk of adverse neurodevelopment. More precise PDA diagnostic criteria and are needed in future studies. NT-proBNP may be useful as an additional parameter in combined scores of clinical and echocardiographic markers of ductal severity. Optimally designed, blinded placebo-controlled studies with a clear definition of the PDA exposure and related outcomes are needed to understand the PDA in extremely preterm population