Original Article
895
Antioxidants and Biomarkers of Oxidative Stress in Preterm Infants with Symptomatic Patent Ductus Arteriosus Musaddaq Inayat, MD1 Fayez Bany-Mohammed, MD1 Arwin Valencia, MD2 Ching Tay, MS, RNC2 Josefina Jacinto, MD2 Jacob V. Aranda, MD, PhD3,4,5 Kay D. Beharry, BS3,4,5
University of California, Irvine, California 2 Division of Neonatal-Perinatal Medicine, Department of Pediatrics, Miller Children’s Hospital, Long Beach, California 3 Division of Neonatal-Perinatal Medicine, Department of Pediatrics, State University of New York, Downstate Medical Center, Brooklyn, New York 4 Department of Ophthalmology, State University of New York, Downstate Medical Center, Brooklyn, New York 5 State University of New York Eye Institute, NY, New York
Address for correspondence Kay D. Beharry, BS, Department of Pediatrics & Ophthalmology, State University of New York, Downstate Medical Center, 450 Clarkson Avenue Box 49, Room BSB 4-22, Brooklyn, NY 11203 (e-mail: [email protected]).
Am J Perinatol 2015;32:895–904.
Abstract
Keywords
► antioxidant ► oxidative stress ► patent ductus arteriosus ► premature infants
Objective Immature antioxidant and oxygen-sensing mechanisms are involved in the pathogenesis of the patent ductus arteriosus (PDA). We conducted a prospective, observational, pilot study to test the hypothesis that antioxidant activity is low at birth in preterm infants at risk for symptomatic PDA. Study Design Blood and urine samples were collected within 24 to 48 hours of life in 53 preterm infants (32 weeks’ gestation) who developed early PDA symptoms and in 30 term (37 weeks’ gestation) control infants. Thirty preterm infants developed hemodynamically significant PDA (hsPDA) and required pharmacologic treatment and/ or PDA ligation. For these infants, blood and urine samples were also collected at 24 hours posttreatment. Samples were analyzed for biomarkers of antioxidant activity, oxidative stress, and lipid peroxidation. Results At 24 to 48 hours after birth, plasma superoxide dismutase (SOD), urinary catalase, and plasma and urinary 8-isoPGF2α were significantly lower in preterm infants who developed hsPDA. Plasma 8-isoPGF2α levels rebounded post–PDA treatment, while urinary prostaglandin E2, plasma and urinary thromboxane B2, and plasma SOD declined. Conclusion The antioxidant status is low in preterm infants at risk for developing hsPDA. SOD may be a key antioxidant regulating functional ductus arteriosus closure. Therefore, low levels may result in persistence of a hsPDA.
For reasons not entirely understood, the ductus arteriosus (DA) in preterm infants often remains open for days and occasionally weeks after birth, thus presenting myriad problems related to
pulmonary overcirculation and systemic underperfusion.1 These include volume overload in the lungs which may result in pulmonary edema (requiring the use of diuretics), respiratory
received August 11, 2014 accepted after revision December 19, 2014 published online February 25, 2015
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0035-1544948. ISSN 0735-1631.
Downloaded by: University of Queensland. Copyrighted material.
1 Division of Neonatal-Perinatal Medicine, Department of Pediatrics,
Low SOD and Persistence of a Hemodynamically Significant PDA distress, increased ventilator support, and chronic lung disease.1,2 A relatively large patent ductus arteriosus (PDA) that offers no resistance of its own (nonrestrictive) “steals” significant systemic blood flow from organs distal to the duct and eventually cause congestive heart failure, renal failure, feeding intolerance, and necrotizing enterocolitis (NEC).3,4 Indomethacin and ibuprofen are pharmacologic agents commonly used to close the hemodynamically significant PDA (hsPDA) in preterm infants.5–7 Surgical ligation is an effective alternative for nonresponders to pharmacologic therapy.8,9 There has been recent debate about the risk–benefit ratio of treating most infants with PDA10,11; therefore, identifying a biomarker that predicts which infants will become most affected by persistent patency of the DA is of paramount importance. In 1971, Fay12 postulated the existence of oxygen sensors in the DA. Since then, numerous studies have confirmed that the DA responds to changes in PO2 with changes in the redox state to produce reactive oxygen species (ROS).13–17 ROS (such as superoxide anion, hydrogen peroxide, and peroxynitrite) are highly reactive molecules with unpaired electrons. To regain their stability, ROS react with nearby molecules to obtain the electrons they need, thus causing damage to the nearby molecules by changing their structure and function.18 ROS inhibit Kþ channels to cause membrane depolarization and influx of intracellular calcium ultimately restricting the DA. Hypoxia maintains the patency of the DA. However, at the time of birth, ROS increases when PaO2 changes from fetal to neonatal levels.19 In response, antioxidants (superoxide dismutase or SOD, glutathione peroxidase or GPX, and catalase) are overproduced to scavenge the ROS.20 These studies suggest a role for antioxidants in the O2 sensitivity of the DA. Given the role of ROS in functional closure of the DA,21 and given that preterm infants have immature antioxidant systems,22–24 we tested the hypothesis that preterm infants at risk for symptomatic PDA have low antioxidant status at birth. Low antioxidant status reduces the O2 responsiveness of the DA and results in failure of the PDA to close. To test our hypothesis, we conducted a prospective, observational, pilot study of the antioxidants and biomarkers of oxidative stress in preterm with early symptomatic PDA (24–48 hours of age) and term infants (as controls), and in those subset preterm infants who subsequently developed hsPDA 24 hours post–pharmacologic treatment and/or surgical ligation.
Patients and Methods Patients Infants with a birth weight (BW) of 1,500 g or 32 weeks’ gestation, who were suspected of having PDA, were eligible for enrollment in the study following informed and signed parental consent. Infants with major congenital anomalies, neuromuscular diseases, intrauterine growth restriction (IUGR), or circulatory failure were excluded. The study was approved by the Institutional Review Board of Miller Children’s Hospital, Long Beach, CA. Gestational age (GA) was determined from maternal last menstrual period and physical examination using modified Ballard assessment.25 All preterm infants in the study displayed signs suggestive of a PDA, and echocardiograms were American Journal of Perinatology
Vol. 32
No. 9/2015
Inayat et al.
obtained. The decision to obtain echocardiogram was made by the attending neonatologist. The presence of a hsPDA was determined by three or more clinical and physical findings (bounding pulses, wide pulse pressure, hyperdynamic precordium, systolic murmur, pulmonary edema, increased bronchovascular markings, and cardiac silhouette on chest radiographs) and confirmed with echocardiography. Two-dimensional echocardiography with color Doppler imaging were performed on days 3 to 6 of life and monitored by pediatric cardiologists. The echocardiographic criteria were as follows: ductal size, >1.5 mm; left atrial-to-aortic root ratio, >1.6; and retrograde diastolic flow in the descending aorta, >30% of anterograde flow. The ultimate decision to treat PDA was made by the attending physician and was not stratified by the study design. A total of 53 preterm infants and 30 term infants were enrolled in the study with 30 preterm infants subsequently developing hsPDA requiring treatment and 23 with nonsignificant PDA (nsPDA) that were followed clinically with the PDA ultimately closed spontaneously before discharge. The term infants served as controls for the studies of ROS and antioxidant systems; they did not undergo echocardiographic testing and did not show signs of PDA. Bronchopulmonary dysplasia (BPD) was defined as the need for supplemental oxygen and/or assisted ventilation at 36 weeks’ postmenstrual age. The presence and grade of intraventricular hemorrhage (IVH) was confirmed by cranial ultrasound examinations. Sepsis was confirmed by positive blood cultures. The diagnosis of NEC or small intestinal perforation was based on a combination of the surgeon’s and attending neonatologist’s physical examination, abdominal radiographic data, and intraoperative findings recorded by the surgeon. Retinopathy of prematurity (ROP) was confirmed following serial routine eye examination starting at 4 to 6 weeks after birth by a pediatric ophthalmologist and retina specialist. Evaluation of urine output (mL/kg/h), creatinine (mg/dL), and BUN (mg/dL) was done per standard of care.
Sample Collection Blood and urine samples were collected between 24 and 48 hours of life for all infants. For infants who developed hsPDA requiring treatment, blood and urine samples were also collected at 24 hours posttreatment (after the last dose of medical therapy or after PDA ligation surgery). Blood samples were collected in ethylenediaminetetraacetic acid (EDTA) tubes and centrifuged within 30 minutes at 3,000 rpm and 4°C for 20 minutes. Urine samples were collected in a sterile urine bag at the same time as the blood samples. Plasma and urine samples were stored in a 20°C freezer before transportation to the laboratory for processing.
Assay of Antioxidant Activity SOD, GPX, and catalase activities were determined in the plasma and urine using commercially available assay kits from Cayman Chemicals, Ann Arbor, MI, according to the manufacturer’s protocol.
Assay of 8-isoPGF2α Isoprostanes are prostanoid derivatives that are produced by the nonenzymatic peroxidation of arachidonic acid through
Downloaded by: University of Queensland. Copyrighted material.
896
Low SOD and Persistence of a Hemodynamically Significant PDA
Assay of NOx Levels of the nitric oxide stable metabolites (NOx), nitrate and nitrite, were determined in the plasma and urine using commercially available colorimetric kits from Assay Designs, Inc. according to the manufacturer’s protocol.
Assay of TxB2 and PGE2 Levels of thromboxane B2 (TxB2) and prostaglandin E2 (PGE2) were determined in the plasma and urine using commercially available enzyme immunoassay kits from Assay Designs, Inc. according to the manufacturer’s protocol.
Statistical Analysis Categorical data were analyzed using chi-square, Fisher exact, or Mann–Whitney U-tests, as appropriate. For numerical data, Student t-test was used to determine differences between the combined preterm (24–48 hours postbirth) and term groups, and hsPDA and nsPDA preterm groups. Paired t-test was used to determine differences between pre- and posttreatment data in the hsPDA preterm group. Levene test for equality of variances was used to determine normality. Alternate tests were used for nonnormal data. A subanalysis was performed using logistic regression with GA as the covariate. Significance was set at p < 0.05 and categorical data are reported as percentages while continuous data are reported as mean SD. All analyses were two-tailed and performed using SPSS Software, version 16.0 (SPSS, Inc., Chicago, IL).
Results Infants’ Characteristics A total of 83 infants were enrolled in the study. Of these, 30 were term infants (BW: 3440.8 422.5 g; GA: 39.2 1.0 weeks); 30 were preterm infants diagnosed with hsPDA (BW: 888.4 332.9 g; GA: 25.9 2.0 weeks); and 23 were preterm infants diagnosed with nsPDA (BW: 1080.7 380.4 g; GA: 27.5 2.3 weeks). Blood and urine samples were collected on all patients between 24 and 48 hours of birth for a total of 83 blood and 83 urine samples. In the group that developed hsPDA, blood and urine samples were collected at 24 hours posttreatment completion (indomethacin, ibuprofen, and/or surgical ligation) for a total of 30 blood and 30 urine samples. Infant characteristics are presented in ►Table 1. Infants who developed hsPDA had lower mean GA (25.9 2.0 vs. 27.5 2.3 weeks, p ¼ 0.012), and tended to have lower BW (888.4 332.9 vs.1080.7 380.4 g, p ¼ 0.069). The rate of cesarean delivery was comparable among the groups, as were infant gender, PROM, preeclampsia, multiple gestations, and prenatal steroids. However, there were significantly more Caucasian infants in the group that
897
did not develop hsPDA (p ¼ 0.041). Use of postnatal steroids (p ¼ 0.002), surfactant (p ¼ 0.012), and caffeine (p ¼ 0.0002) was more frequent in the group with hsPDA. Length of O2 therapy and/or mechanical ventilation was comparable; however, length of hospital stay was greater in the hsPDA group (p ¼ 0.048). Of the infants with hsPDA, 11 (37%) received surgical ligation (8 were nonresponders and 3 did not receive prior pharmacologic treatment). ►Table 2 lists the major neonatal outcomes for both preterm groups. Infants with hsPDA had higher incidences of apnea (p ¼ 0.0002) and proven sepsis (p ¼ 0.01) than infants with nsPDA. The infants with hsPDA also tended to have higher BPD (p ¼ 0.066). There were no differences in urinary output, creatinine, and BUN in the hsPDA group between the 24 to 48 hours postbirth and 24 hours posttreatment data not shown.
Antioxidant Status To examine whether the antioxidant status of preterm infants at birth is different from term infants, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 1. At 24 to 48 hours postbirth, plasma catalase activity (μM) did not differ among the groups. In contrast, urinary catalase activity at 24 to 48 hours after birth was significantly lower in the preterm group (18.6 20.6, p < 0.001) compared with term infants (68.2 38.4) (►Fig. 1). There were no differences in plasma or urinary SOD activity (U/mL) between preterm and term infants. Plasma SOD activity at 24 to 48 hours after birth was significantly lower in infants with a hsPDA (2.7 1.1, p < 0.001) than those infants with a nsPDA (8.5 2.2) (►Fig. 2A). Plasma SOD activity declined posttreatment (2.2 0.7, p < 0.05) compared with 24- to 48-hour levels (2.7 1.1) in the infants with hsPDA (►Fig. 2B). Urinary SOD activity (U/mL) was higher in the preterm infants with a hsPDA (9.4 3.8, p < 0.01) compared with preterm infants with a nsPDA (3.3 4.5) (►Fig. 2C), and did not change in response to treatment. Plasma and urinary GPX remained comparable among the preterm and term groups and did not differ in response to treatment in the hsPDA group.
Oxidative Stress To examine differences in the biomarker for oxidative stress between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 3. Plasma 8-isoPGF2α levels (pg/mL) at 24 to 48 hours of age were significantly lower in preterm infants (9317.2 7119.6, p < 0.05) than term infants (12659.6 6529) (►Fig. 3A). Plasma 8-isoPGF2α levels were also lower in the preterm infants who later developed a hsPDA (6060.9 5302.5, p < 0.01) compared with preterm infants who did not (13281.5 9161.7) (►Fig. 3B). Plasma 8-isoPGF2α levels increased posttreatment (41581.7 16573.5, p < 0.001) (►Fig. 3C). Urinary 8-isoPGF2α levels (pg/mL) were lower in preterm infants (10749.5 11907.1, p < 0.001) compared with term infants (40476.4 13175.7) (►Fig. 3D). Urinary 8-isoPGF2α levels were comparable between the nsPDA and hsPDA groups at 24 to American Journal of Perinatology
Vol. 32
No. 9/2015
Downloaded by: University of Queensland. Copyrighted material.
ROS. 8-isoPGF2α or 8-isoprostane is commonly studied and is abundantly generated in vivo during oxidative stress and lipid peroxidation. To establish degree of oxidative stress, levels of 8-isoPGF2α were determined in the plasma and urine using commercially available enzyme immunoassay kits from Assay Designs, Inc., Ann Arbor, MI, according to the manufacturer’s protocol.
Inayat et al.
Low SOD and Persistence of a Hemodynamically Significant PDA
Inayat et al.
Table 1 Infant characteristics Variable
hsPDA (n ¼ 30)
nsPDA (n ¼ 23)
p-Value
Birth weight (g)
888.4 332.9
1080.7 380.4
0.069
GA (wk)
25.9 2.0
27.5 2.3
0.012
Apgar scores 1 min (< 7.0)
13 (43%)
6 (26%)
0.25
5 min (< 7.0)
2 (7%)
2 (9%)
1.0
Male
18 (60%)
14 (61%)
1.0
Female
12 (40%)
9 (39%)
1.0
6 (20%)
11 (48%)
0.041
Gender
Race Caucasian African American
6 (20%)
2 (9%)
0.44
Hispanic
13 (43%)
8 (35%)
0.58
Asian
2 (7%)
1 (4%)
1.0
Other
3 (10%)
1 (4%)
0.62
Multiple gestations
10 (33%)
5 (17%)
0.54
C-section
17 (57%)
16 (70%)
0.40
PPROM
9 (30%)
3 (13%)
0.19
Preeclampsia
7 (23%)
6 (26%)
1.0
Prenatal steroids
13 (43%)
9 (39%)
0.78
Postnatal steroids
20 (67%)
5 (22%)
0.002
Surfactant treatment
19 (63%)
6 (26%)
0.012
Caffeine
30 (100%)
14 (61%)
0.0002
74.9 33.1
45.3 29.4
0.009
Days on caffeine
18 (60%)
N/A
–
6.0 1.2
N/A
–
11 (37%)
N/A
–
24.6 6.6
N/A
–
Length of O2/ventilator therapy
54.9 42.9
29.8 31.4
0.11
Length of hospital stay
94.9 44.3
68.6 39.4
0.048
Indomethacin/Ibuprofen treatment Age at treatment (d) Surgical ligation Age at ligation (d)
48 hours postbirth and did not change in response to treatment in the hsPDA group.
Nitric Oxide Stable Metabolites To examine differences in the NOx between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 4. Plasma NOx levels (μmol/L) were higher in the preterm infants (343.4 279.4, p < 0.001) than term infants (162.8 92.8) (►Fig. 4A). Plasma NOx levels were lower in infants with a hsPDA (240.2 157.2, p < 0.01) than infants with a nsPDA (468.9 116.6) (►Fig. 4B). Conversely, urinary NOx levels were higher in the preterm infants with a hsPDA (221.0 67, p < 0.01) than those with a nsPDA (165.5 68.3) (►Fig. 4C). Plasma and urinary NOx levels did not changed in response to treatment. American Journal of Perinatology
Vol. 32
No. 9/2015
PGE2 and TxB2 To examine differences in the prostanoids between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Figs. 5 and 6. There were no differences in plasma or urinary PGE2 levels (pg/mL) between the preterm and term groups, or between the nsPDA and hsPDA groups. However, plasma PGE2 levels increased slightly posttreatment (2007.4 1390.9, p < 0.05) in the preterm infants with hsPDA compared with levels at 24 to 48 hours after birth (1335.1 1126.6) (►Fig. 5A). Conversely, urinary PGE2 levels declined posttreatment (3265.5 1454.3, p < 0.05) compared with levels at 24 to 48 hours after birth (4157.0 1174.7) (►Fig. 5B). At 24 to 48 hours, plasma TxB2 levels (pg/mL) did not differ among the groups. In contrast, urinary TxB2 levels (pg/mL) were significantly lower in the preterm (1073.5 776.8, p < 0.001) than
Downloaded by: University of Queensland. Copyrighted material.
898
Low SOD and Persistence of a Hemodynamically Significant PDA
Inayat et al.
899
Table 2 Major neonatal complications Variable
Symptomatic PDA (n ¼ 30)
Nonsymptomatic PDA (n ¼ 23)
p-Value
BPD
11 (37%)
3 (13%)
0.066
IVH (grades 3 and 4)
4 (13%)
3 (13%)
1.0
Apnea
30 (100%)
14 (61%)
0.0002
Proven Sepsis
12 (40%)
2 (9%)
0.01
NEC
4 (13%)
2 (9%)
0.69
SIP
1 (3%)
1 (4%)
1.0
ROP (grades 3 and 4)
4 (13%)
0 (0%)
0.12
Laser therapy
2 (7%)
0 (0%)
0.50
Death
3 (10%)
3 (13%)
1.0
in the term (2847.7 805.8) group (►Fig. 6A). Posttreatment, both plasma (932.5 1547.7, p < 0.001) and urine (683.3 559.4, p < 0.01) TxB2 levels were lower than levels at 24 to 48 hours (plasma: 2628.1 3284; urine: 1223.8 847.5) (►Fig. 6B, C).
Subanalysis to Control for Gestational Age Logistic regression was performed with GA as the covariate to determine whether the oxidant/antioxidant biomarkers were different between nsPDA and hsPDA groups. For the hsPDA group, only the pretreatment data were used. Data were also analyzed to examine whether urinary output, creatinine, and BUN correlated with oxidant/antioxidant biomarkers at 24 to 48 hours postbirth and at 24 hours posttreatment in the hsPDA group. Data showed significant differences in plasma and urinary SOD and NOx, and plasma 8-isoprostane between the groups. Pre- and posttreatment urinary output, creatinine, and BUN levels did not correlate with any of the pre- and posttreatment oxidant/antioxidants data.
Fig. 1 Urinary catalase activity in term infants (n ¼ 30) and preterm infants (n ¼ 53). Samples were taken between 24 and 48 hours after birth. Data from the nsPDA and hsPDA preterm infants at 24 to 48 hours postbirth are combined for the preterm group. Data are mean SD.
Discussion Functional closure of the DA occurs in 50% of all infants by 24 hours of birth, in 90% by 48 hours, and in 100% by 72 hours in term infants.11 For this reason, we examined the baseline levels of antioxidant activity and biomarkers of ROS between 24 and 48 hours of birth in preterm infants with early symptomatic PDA and compared them with that of term infants. Our data showed that plasma 8-isoPGF2α and SOD were lower while urinary catalase was higher in infants with hsPDA. At 24 hours posttreatment, plasma SOD further declined, but plasma 8-isoPGF2α increased substantially. These data may suggest a predictive value of low plasma SOD and/or 8-isoPGF2α as biomarkers for hsPDA. To our knowledge, this is the first study to examine and compare these biomarkers at 24 to 48 hours of age and in response to treatment for hsPDA. Our study contrasts with previous findings of higher urinary 8-OHdG (a biomarker for DNA damage and oxidative stress) in preterm infants.23 The difference is most likely due to the older GAs of the preterm infants in that study (34.5 0.5 weeks) compared with infants with hsPDA (25.9 2.0 weeks) and nsPDA (27.5 2.3 weeks) in our study. Nevertheless, the lower SOD status of preterm infants in our study was consistent with the findings of Nassie et al.23 It is possible that lower SOD was related to developmental immaturity, as infants with hsPDA had lower BW and GA. It is also known that ROS upregulates various antioxidant systems26; it is conceivable that low ROS in infants with hsPDA may have resulted in low antioxidant status (low SOD). Persistent PDA increases cardiopulmonary complications in preterm infants
895
Antioxidants and Biomarkers of Oxidative Stress in Preterm Infants with Symptomatic Patent Ductus Arteriosus Musaddaq Inayat, MD1 Fayez Bany-Mohammed, MD1 Arwin Valencia, MD2 Ching Tay, MS, RNC2 Josefina Jacinto, MD2 Jacob V. Aranda, MD, PhD3,4,5 Kay D. Beharry, BS3,4,5
University of California, Irvine, California 2 Division of Neonatal-Perinatal Medicine, Department of Pediatrics, Miller Children’s Hospital, Long Beach, California 3 Division of Neonatal-Perinatal Medicine, Department of Pediatrics, State University of New York, Downstate Medical Center, Brooklyn, New York 4 Department of Ophthalmology, State University of New York, Downstate Medical Center, Brooklyn, New York 5 State University of New York Eye Institute, NY, New York
Address for correspondence Kay D. Beharry, BS, Department of Pediatrics & Ophthalmology, State University of New York, Downstate Medical Center, 450 Clarkson Avenue Box 49, Room BSB 4-22, Brooklyn, NY 11203 (e-mail: [email protected]).
Am J Perinatol 2015;32:895–904.
Abstract
Keywords
► antioxidant ► oxidative stress ► patent ductus arteriosus ► premature infants
Objective Immature antioxidant and oxygen-sensing mechanisms are involved in the pathogenesis of the patent ductus arteriosus (PDA). We conducted a prospective, observational, pilot study to test the hypothesis that antioxidant activity is low at birth in preterm infants at risk for symptomatic PDA. Study Design Blood and urine samples were collected within 24 to 48 hours of life in 53 preterm infants (32 weeks’ gestation) who developed early PDA symptoms and in 30 term (37 weeks’ gestation) control infants. Thirty preterm infants developed hemodynamically significant PDA (hsPDA) and required pharmacologic treatment and/ or PDA ligation. For these infants, blood and urine samples were also collected at 24 hours posttreatment. Samples were analyzed for biomarkers of antioxidant activity, oxidative stress, and lipid peroxidation. Results At 24 to 48 hours after birth, plasma superoxide dismutase (SOD), urinary catalase, and plasma and urinary 8-isoPGF2α were significantly lower in preterm infants who developed hsPDA. Plasma 8-isoPGF2α levels rebounded post–PDA treatment, while urinary prostaglandin E2, plasma and urinary thromboxane B2, and plasma SOD declined. Conclusion The antioxidant status is low in preterm infants at risk for developing hsPDA. SOD may be a key antioxidant regulating functional ductus arteriosus closure. Therefore, low levels may result in persistence of a hsPDA.
For reasons not entirely understood, the ductus arteriosus (DA) in preterm infants often remains open for days and occasionally weeks after birth, thus presenting myriad problems related to
pulmonary overcirculation and systemic underperfusion.1 These include volume overload in the lungs which may result in pulmonary edema (requiring the use of diuretics), respiratory
received August 11, 2014 accepted after revision December 19, 2014 published online February 25, 2015
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0035-1544948. ISSN 0735-1631.
Downloaded by: University of Queensland. Copyrighted material.
1 Division of Neonatal-Perinatal Medicine, Department of Pediatrics,
Low SOD and Persistence of a Hemodynamically Significant PDA distress, increased ventilator support, and chronic lung disease.1,2 A relatively large patent ductus arteriosus (PDA) that offers no resistance of its own (nonrestrictive) “steals” significant systemic blood flow from organs distal to the duct and eventually cause congestive heart failure, renal failure, feeding intolerance, and necrotizing enterocolitis (NEC).3,4 Indomethacin and ibuprofen are pharmacologic agents commonly used to close the hemodynamically significant PDA (hsPDA) in preterm infants.5–7 Surgical ligation is an effective alternative for nonresponders to pharmacologic therapy.8,9 There has been recent debate about the risk–benefit ratio of treating most infants with PDA10,11; therefore, identifying a biomarker that predicts which infants will become most affected by persistent patency of the DA is of paramount importance. In 1971, Fay12 postulated the existence of oxygen sensors in the DA. Since then, numerous studies have confirmed that the DA responds to changes in PO2 with changes in the redox state to produce reactive oxygen species (ROS).13–17 ROS (such as superoxide anion, hydrogen peroxide, and peroxynitrite) are highly reactive molecules with unpaired electrons. To regain their stability, ROS react with nearby molecules to obtain the electrons they need, thus causing damage to the nearby molecules by changing their structure and function.18 ROS inhibit Kþ channels to cause membrane depolarization and influx of intracellular calcium ultimately restricting the DA. Hypoxia maintains the patency of the DA. However, at the time of birth, ROS increases when PaO2 changes from fetal to neonatal levels.19 In response, antioxidants (superoxide dismutase or SOD, glutathione peroxidase or GPX, and catalase) are overproduced to scavenge the ROS.20 These studies suggest a role for antioxidants in the O2 sensitivity of the DA. Given the role of ROS in functional closure of the DA,21 and given that preterm infants have immature antioxidant systems,22–24 we tested the hypothesis that preterm infants at risk for symptomatic PDA have low antioxidant status at birth. Low antioxidant status reduces the O2 responsiveness of the DA and results in failure of the PDA to close. To test our hypothesis, we conducted a prospective, observational, pilot study of the antioxidants and biomarkers of oxidative stress in preterm with early symptomatic PDA (24–48 hours of age) and term infants (as controls), and in those subset preterm infants who subsequently developed hsPDA 24 hours post–pharmacologic treatment and/or surgical ligation.
Patients and Methods Patients Infants with a birth weight (BW) of 1,500 g or 32 weeks’ gestation, who were suspected of having PDA, were eligible for enrollment in the study following informed and signed parental consent. Infants with major congenital anomalies, neuromuscular diseases, intrauterine growth restriction (IUGR), or circulatory failure were excluded. The study was approved by the Institutional Review Board of Miller Children’s Hospital, Long Beach, CA. Gestational age (GA) was determined from maternal last menstrual period and physical examination using modified Ballard assessment.25 All preterm infants in the study displayed signs suggestive of a PDA, and echocardiograms were American Journal of Perinatology
Vol. 32
No. 9/2015
Inayat et al.
obtained. The decision to obtain echocardiogram was made by the attending neonatologist. The presence of a hsPDA was determined by three or more clinical and physical findings (bounding pulses, wide pulse pressure, hyperdynamic precordium, systolic murmur, pulmonary edema, increased bronchovascular markings, and cardiac silhouette on chest radiographs) and confirmed with echocardiography. Two-dimensional echocardiography with color Doppler imaging were performed on days 3 to 6 of life and monitored by pediatric cardiologists. The echocardiographic criteria were as follows: ductal size, >1.5 mm; left atrial-to-aortic root ratio, >1.6; and retrograde diastolic flow in the descending aorta, >30% of anterograde flow. The ultimate decision to treat PDA was made by the attending physician and was not stratified by the study design. A total of 53 preterm infants and 30 term infants were enrolled in the study with 30 preterm infants subsequently developing hsPDA requiring treatment and 23 with nonsignificant PDA (nsPDA) that were followed clinically with the PDA ultimately closed spontaneously before discharge. The term infants served as controls for the studies of ROS and antioxidant systems; they did not undergo echocardiographic testing and did not show signs of PDA. Bronchopulmonary dysplasia (BPD) was defined as the need for supplemental oxygen and/or assisted ventilation at 36 weeks’ postmenstrual age. The presence and grade of intraventricular hemorrhage (IVH) was confirmed by cranial ultrasound examinations. Sepsis was confirmed by positive blood cultures. The diagnosis of NEC or small intestinal perforation was based on a combination of the surgeon’s and attending neonatologist’s physical examination, abdominal radiographic data, and intraoperative findings recorded by the surgeon. Retinopathy of prematurity (ROP) was confirmed following serial routine eye examination starting at 4 to 6 weeks after birth by a pediatric ophthalmologist and retina specialist. Evaluation of urine output (mL/kg/h), creatinine (mg/dL), and BUN (mg/dL) was done per standard of care.
Sample Collection Blood and urine samples were collected between 24 and 48 hours of life for all infants. For infants who developed hsPDA requiring treatment, blood and urine samples were also collected at 24 hours posttreatment (after the last dose of medical therapy or after PDA ligation surgery). Blood samples were collected in ethylenediaminetetraacetic acid (EDTA) tubes and centrifuged within 30 minutes at 3,000 rpm and 4°C for 20 minutes. Urine samples were collected in a sterile urine bag at the same time as the blood samples. Plasma and urine samples were stored in a 20°C freezer before transportation to the laboratory for processing.
Assay of Antioxidant Activity SOD, GPX, and catalase activities were determined in the plasma and urine using commercially available assay kits from Cayman Chemicals, Ann Arbor, MI, according to the manufacturer’s protocol.
Assay of 8-isoPGF2α Isoprostanes are prostanoid derivatives that are produced by the nonenzymatic peroxidation of arachidonic acid through
Downloaded by: University of Queensland. Copyrighted material.
896
Low SOD and Persistence of a Hemodynamically Significant PDA
Assay of NOx Levels of the nitric oxide stable metabolites (NOx), nitrate and nitrite, were determined in the plasma and urine using commercially available colorimetric kits from Assay Designs, Inc. according to the manufacturer’s protocol.
Assay of TxB2 and PGE2 Levels of thromboxane B2 (TxB2) and prostaglandin E2 (PGE2) were determined in the plasma and urine using commercially available enzyme immunoassay kits from Assay Designs, Inc. according to the manufacturer’s protocol.
Statistical Analysis Categorical data were analyzed using chi-square, Fisher exact, or Mann–Whitney U-tests, as appropriate. For numerical data, Student t-test was used to determine differences between the combined preterm (24–48 hours postbirth) and term groups, and hsPDA and nsPDA preterm groups. Paired t-test was used to determine differences between pre- and posttreatment data in the hsPDA preterm group. Levene test for equality of variances was used to determine normality. Alternate tests were used for nonnormal data. A subanalysis was performed using logistic regression with GA as the covariate. Significance was set at p < 0.05 and categorical data are reported as percentages while continuous data are reported as mean SD. All analyses were two-tailed and performed using SPSS Software, version 16.0 (SPSS, Inc., Chicago, IL).
Results Infants’ Characteristics A total of 83 infants were enrolled in the study. Of these, 30 were term infants (BW: 3440.8 422.5 g; GA: 39.2 1.0 weeks); 30 were preterm infants diagnosed with hsPDA (BW: 888.4 332.9 g; GA: 25.9 2.0 weeks); and 23 were preterm infants diagnosed with nsPDA (BW: 1080.7 380.4 g; GA: 27.5 2.3 weeks). Blood and urine samples were collected on all patients between 24 and 48 hours of birth for a total of 83 blood and 83 urine samples. In the group that developed hsPDA, blood and urine samples were collected at 24 hours posttreatment completion (indomethacin, ibuprofen, and/or surgical ligation) for a total of 30 blood and 30 urine samples. Infant characteristics are presented in ►Table 1. Infants who developed hsPDA had lower mean GA (25.9 2.0 vs. 27.5 2.3 weeks, p ¼ 0.012), and tended to have lower BW (888.4 332.9 vs.1080.7 380.4 g, p ¼ 0.069). The rate of cesarean delivery was comparable among the groups, as were infant gender, PROM, preeclampsia, multiple gestations, and prenatal steroids. However, there were significantly more Caucasian infants in the group that
897
did not develop hsPDA (p ¼ 0.041). Use of postnatal steroids (p ¼ 0.002), surfactant (p ¼ 0.012), and caffeine (p ¼ 0.0002) was more frequent in the group with hsPDA. Length of O2 therapy and/or mechanical ventilation was comparable; however, length of hospital stay was greater in the hsPDA group (p ¼ 0.048). Of the infants with hsPDA, 11 (37%) received surgical ligation (8 were nonresponders and 3 did not receive prior pharmacologic treatment). ►Table 2 lists the major neonatal outcomes for both preterm groups. Infants with hsPDA had higher incidences of apnea (p ¼ 0.0002) and proven sepsis (p ¼ 0.01) than infants with nsPDA. The infants with hsPDA also tended to have higher BPD (p ¼ 0.066). There were no differences in urinary output, creatinine, and BUN in the hsPDA group between the 24 to 48 hours postbirth and 24 hours posttreatment data not shown.
Antioxidant Status To examine whether the antioxidant status of preterm infants at birth is different from term infants, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 1. At 24 to 48 hours postbirth, plasma catalase activity (μM) did not differ among the groups. In contrast, urinary catalase activity at 24 to 48 hours after birth was significantly lower in the preterm group (18.6 20.6, p < 0.001) compared with term infants (68.2 38.4) (►Fig. 1). There were no differences in plasma or urinary SOD activity (U/mL) between preterm and term infants. Plasma SOD activity at 24 to 48 hours after birth was significantly lower in infants with a hsPDA (2.7 1.1, p < 0.001) than those infants with a nsPDA (8.5 2.2) (►Fig. 2A). Plasma SOD activity declined posttreatment (2.2 0.7, p < 0.05) compared with 24- to 48-hour levels (2.7 1.1) in the infants with hsPDA (►Fig. 2B). Urinary SOD activity (U/mL) was higher in the preterm infants with a hsPDA (9.4 3.8, p < 0.01) compared with preterm infants with a nsPDA (3.3 4.5) (►Fig. 2C), and did not change in response to treatment. Plasma and urinary GPX remained comparable among the preterm and term groups and did not differ in response to treatment in the hsPDA group.
Oxidative Stress To examine differences in the biomarker for oxidative stress between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 3. Plasma 8-isoPGF2α levels (pg/mL) at 24 to 48 hours of age were significantly lower in preterm infants (9317.2 7119.6, p < 0.05) than term infants (12659.6 6529) (►Fig. 3A). Plasma 8-isoPGF2α levels were also lower in the preterm infants who later developed a hsPDA (6060.9 5302.5, p < 0.01) compared with preterm infants who did not (13281.5 9161.7) (►Fig. 3B). Plasma 8-isoPGF2α levels increased posttreatment (41581.7 16573.5, p < 0.001) (►Fig. 3C). Urinary 8-isoPGF2α levels (pg/mL) were lower in preterm infants (10749.5 11907.1, p < 0.001) compared with term infants (40476.4 13175.7) (►Fig. 3D). Urinary 8-isoPGF2α levels were comparable between the nsPDA and hsPDA groups at 24 to American Journal of Perinatology
Vol. 32
No. 9/2015
Downloaded by: University of Queensland. Copyrighted material.
ROS. 8-isoPGF2α or 8-isoprostane is commonly studied and is abundantly generated in vivo during oxidative stress and lipid peroxidation. To establish degree of oxidative stress, levels of 8-isoPGF2α were determined in the plasma and urine using commercially available enzyme immunoassay kits from Assay Designs, Inc., Ann Arbor, MI, according to the manufacturer’s protocol.
Inayat et al.
Low SOD and Persistence of a Hemodynamically Significant PDA
Inayat et al.
Table 1 Infant characteristics Variable
hsPDA (n ¼ 30)
nsPDA (n ¼ 23)
p-Value
Birth weight (g)
888.4 332.9
1080.7 380.4
0.069
GA (wk)
25.9 2.0
27.5 2.3
0.012
Apgar scores 1 min (< 7.0)
13 (43%)
6 (26%)
0.25
5 min (< 7.0)
2 (7%)
2 (9%)
1.0
Male
18 (60%)
14 (61%)
1.0
Female
12 (40%)
9 (39%)
1.0
6 (20%)
11 (48%)
0.041
Gender
Race Caucasian African American
6 (20%)
2 (9%)
0.44
Hispanic
13 (43%)
8 (35%)
0.58
Asian
2 (7%)
1 (4%)
1.0
Other
3 (10%)
1 (4%)
0.62
Multiple gestations
10 (33%)
5 (17%)
0.54
C-section
17 (57%)
16 (70%)
0.40
PPROM
9 (30%)
3 (13%)
0.19
Preeclampsia
7 (23%)
6 (26%)
1.0
Prenatal steroids
13 (43%)
9 (39%)
0.78
Postnatal steroids
20 (67%)
5 (22%)
0.002
Surfactant treatment
19 (63%)
6 (26%)
0.012
Caffeine
30 (100%)
14 (61%)
0.0002
74.9 33.1
45.3 29.4
0.009
Days on caffeine
18 (60%)
N/A
–
6.0 1.2
N/A
–
11 (37%)
N/A
–
24.6 6.6
N/A
–
Length of O2/ventilator therapy
54.9 42.9
29.8 31.4
0.11
Length of hospital stay
94.9 44.3
68.6 39.4
0.048
Indomethacin/Ibuprofen treatment Age at treatment (d) Surgical ligation Age at ligation (d)
48 hours postbirth and did not change in response to treatment in the hsPDA group.
Nitric Oxide Stable Metabolites To examine differences in the NOx between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Fig. 4. Plasma NOx levels (μmol/L) were higher in the preterm infants (343.4 279.4, p < 0.001) than term infants (162.8 92.8) (►Fig. 4A). Plasma NOx levels were lower in infants with a hsPDA (240.2 157.2, p < 0.01) than infants with a nsPDA (468.9 116.6) (►Fig. 4B). Conversely, urinary NOx levels were higher in the preterm infants with a hsPDA (221.0 67, p < 0.01) than those with a nsPDA (165.5 68.3) (►Fig. 4C). Plasma and urinary NOx levels did not changed in response to treatment. American Journal of Perinatology
Vol. 32
No. 9/2015
PGE2 and TxB2 To examine differences in the prostanoids between preterm and term, data from nsPDA and hsPDA infants at 24 to 48 hours postbirth were combined to form one preterm group. Only significant data are presented in ►Figs. 5 and 6. There were no differences in plasma or urinary PGE2 levels (pg/mL) between the preterm and term groups, or between the nsPDA and hsPDA groups. However, plasma PGE2 levels increased slightly posttreatment (2007.4 1390.9, p < 0.05) in the preterm infants with hsPDA compared with levels at 24 to 48 hours after birth (1335.1 1126.6) (►Fig. 5A). Conversely, urinary PGE2 levels declined posttreatment (3265.5 1454.3, p < 0.05) compared with levels at 24 to 48 hours after birth (4157.0 1174.7) (►Fig. 5B). At 24 to 48 hours, plasma TxB2 levels (pg/mL) did not differ among the groups. In contrast, urinary TxB2 levels (pg/mL) were significantly lower in the preterm (1073.5 776.8, p < 0.001) than
Downloaded by: University of Queensland. Copyrighted material.
898
Low SOD and Persistence of a Hemodynamically Significant PDA
Inayat et al.
899
Table 2 Major neonatal complications Variable
Symptomatic PDA (n ¼ 30)
Nonsymptomatic PDA (n ¼ 23)
p-Value
BPD
11 (37%)
3 (13%)
0.066
IVH (grades 3 and 4)
4 (13%)
3 (13%)
1.0
Apnea
30 (100%)
14 (61%)
0.0002
Proven Sepsis
12 (40%)
2 (9%)
0.01
NEC
4 (13%)
2 (9%)
0.69
SIP
1 (3%)
1 (4%)
1.0
ROP (grades 3 and 4)
4 (13%)
0 (0%)
0.12
Laser therapy
2 (7%)
0 (0%)
0.50
Death
3 (10%)
3 (13%)
1.0
in the term (2847.7 805.8) group (►Fig. 6A). Posttreatment, both plasma (932.5 1547.7, p < 0.001) and urine (683.3 559.4, p < 0.01) TxB2 levels were lower than levels at 24 to 48 hours (plasma: 2628.1 3284; urine: 1223.8 847.5) (►Fig. 6B, C).
Subanalysis to Control for Gestational Age Logistic regression was performed with GA as the covariate to determine whether the oxidant/antioxidant biomarkers were different between nsPDA and hsPDA groups. For the hsPDA group, only the pretreatment data were used. Data were also analyzed to examine whether urinary output, creatinine, and BUN correlated with oxidant/antioxidant biomarkers at 24 to 48 hours postbirth and at 24 hours posttreatment in the hsPDA group. Data showed significant differences in plasma and urinary SOD and NOx, and plasma 8-isoprostane between the groups. Pre- and posttreatment urinary output, creatinine, and BUN levels did not correlate with any of the pre- and posttreatment oxidant/antioxidants data.
Fig. 1 Urinary catalase activity in term infants (n ¼ 30) and preterm infants (n ¼ 53). Samples were taken between 24 and 48 hours after birth. Data from the nsPDA and hsPDA preterm infants at 24 to 48 hours postbirth are combined for the preterm group. Data are mean SD.
Discussion Functional closure of the DA occurs in 50% of all infants by 24 hours of birth, in 90% by 48 hours, and in 100% by 72 hours in term infants.11 For this reason, we examined the baseline levels of antioxidant activity and biomarkers of ROS between 24 and 48 hours of birth in preterm infants with early symptomatic PDA and compared them with that of term infants. Our data showed that plasma 8-isoPGF2α and SOD were lower while urinary catalase was higher in infants with hsPDA. At 24 hours posttreatment, plasma SOD further declined, but plasma 8-isoPGF2α increased substantially. These data may suggest a predictive value of low plasma SOD and/or 8-isoPGF2α as biomarkers for hsPDA. To our knowledge, this is the first study to examine and compare these biomarkers at 24 to 48 hours of age and in response to treatment for hsPDA. Our study contrasts with previous findings of higher urinary 8-OHdG (a biomarker for DNA damage and oxidative stress) in preterm infants.23 The difference is most likely due to the older GAs of the preterm infants in that study (34.5 0.5 weeks) compared with infants with hsPDA (25.9 2.0 weeks) and nsPDA (27.5 2.3 weeks) in our study. Nevertheless, the lower SOD status of preterm infants in our study was consistent with the findings of Nassie et al.23 It is possible that lower SOD was related to developmental immaturity, as infants with hsPDA had lower BW and GA. It is also known that ROS upregulates various antioxidant systems26; it is conceivable that low ROS in infants with hsPDA may have resulted in low antioxidant status (low SOD). Persistent PDA increases cardiopulmonary complications in preterm infants
