Melatonin use for neuroprotection in perinatal asphyxia: a randomized controlled pilot study

Abstract

Objective:

Melatonin has been shown to be neuroprotective in animal models. The objective of this study is to examine the effect of melatonin on clinical, biochemical, neurophysiological and radiological outcomes of neonates with hypoxic–ischemic encephalopathy (HIE).

Study Design:

We conducted a prospective trial on 45 newborns, 30 with HIE and 15 healthy controls. HIE infants were randomized into: hypothermia group (N=15; received 72-h whole-body cooling) and melatonin/hypothermia group (N=15; received hypothermia and five daily enteral doses of melatonin 10 mg kg⁻¹). Serum melatonin, plasma superoxide dismutase (SOD) and serum nitric oxide (NO) were measured at enrollment for all infants (N=45) and at 5 days for the HIE groups (N=30). In addition to electroencephalography (EEG) at enrollment, all surviving HIE infants were studied with brain magnetic resonance imaging (MRI) and repeated EEG at 2 weeks of life. Neurologic evaluations and Denver Developmental Screening Test II were performed at 6 months.

Result:

Compared with healthy neonates, the two HIE groups had increased melatonin, SOD and NO. At enrollment, the two HIE groups did not differ in clinical, laboratory or EEG findings. At 5 days, the melatonin/hypothermia group had greater increase in melatonin (P<0.001) and decline in NO (P<0.001), but less decline in SOD (P=0.004). The melatonin/hypothermia group had fewer seizures on follow-up EEG and less white matter abnormalities on MRI. At 6 months, the melatonin/hypothermia group had improved survival without neurological or developmental abnormalities (P<0.001).

Conclusion:

Early administration of melatonin to asphyxiated term neonates is feasible and may ameliorate brain injury.

Introduction

Hypoxic–ischemic encephalopathy (HIE) affects 1.5 per 1000 live term births in developed countries.¹ Worldwide, asphyxia is the cause of death in almost one quarter of the 3.6 million neonates who die each year.² Following a hypoxic–ischemic insult, two phases of injury occur; an immediate phase of neuronal cell injury and exhaustion of energy stores is followed by a secondary phase largely mediated by oxidative stress, inflammatory cytokines and apoptosis. Therapeutic hypothermia is neuroprotective through attenuation of almost all phases of this injury.³ However, even with therapeutic hypothermia, 40 to 50% of those with moderate to severe encephalopathy will either die or have severe disability.⁴ Therefore, it is important to find adjuvant therapies that can be used in combination with hypothermia. Melatonin (N-acetyl-5-methoxytryptamine) is an endogenously produced indolamine, primarily produced by pineal body from serotonin. Melatonin has important physiological functions including control of circadian rhythm, sleep, visual, neuroendocrine, cerebrovascular, reproductive and immunological actions.⁵ Melatonin increases toward the end of pregnancy, freely crosses the placenta and crosses the fetal blood–brain barrier.⁶ Following birth its production has been suggested to have a role in brain development and pineal dysfunction was associated with sudden infant death.^7,8

Melatonin supplementation has been studied in adults, children and neonates. It was studied in neonates with respiratory distress syndrome,⁹ sepsis¹⁰ and as a preoperative antioxidant.¹¹

Melatonin has antioxidant, anti-inflammatory and anti-apoptotic properties.¹² As an antioxidant, it directly neutralizes reactive oxygen species and reactive nitrogen species including nitric oxide (NO). It also acts indirectly by stimulating both antioxidant enzyme activity and their cellular mRNA concentrations, including superoxide dismutase (SOD), glutathione peroxidase and glutathione reductase.^13,14 Melatonin has anti-inflammatory effect itself and via its metabolites.¹⁵ It prevents the translocation of the nuclear factor-kappa B of activated B cells, thus reducing the production of proinflammatory cytokines;¹⁶ reduces the inflammatory-derived activation of phospholipase A2, lipoxygenase and cyclooxygenases;¹⁷ reduces recruitment of polymorphonuclear leukocytes to inflammatory sites;¹⁸ and reduces vascular endothelial growth factor concentration.¹⁹ Finally, melatonin not only has significant anti-apoptotic activity which mainly targets the mitochondria, but also enhances several cell rescue pathways.¹²

Melatonin is an attractive agent for neuroprotection as it freely crosses the placenta and the blood–brain barrier. It protects the brain of asphyxiated animals when used alone²⁰ or in combination with therapeutic hypothermia.²¹ In this randomized trial, we planned to study the feasibility and efficacy of enterally administered melatonin to neonates with HIE who are receiving whole-body therapeutic hypothermia. The aim was to examine biochemical, neurophysiological, radiological and clinical outcomes associated with this combined treatment in neonates with moderate to severe HIE.

Methods

Patients

We prospectively studied 45 neonates, 30 neonates diagnosed with HIE and 15 healthy controls. The 30 neonates with HIE were randomly assigned to one of two groups: hypothermia group (N=15) and melatonin/hypothermia group (N=15). Both HIE groups included newborns who fulfilled the following inclusion criteria: (1) inborn infants at term gestation (38 to 42 weeks); (2) Apgar scores ⩽3 at 5 min and/or delayed first breath (>5 min after birth); (3) profound metabolic or mixed acidosis with serum bicarbonate concentrations of <12 mmol l⁻¹ at initial blood gas analyses; and (4) evidence of moderate or moderately severe encephalopathy, such as lethargy, seizures, abnormal reflexes or hypotonia in the immediate neonatal period. Infants were excluded if they had any of the following: (1) twin gestation; (2) maternal neuro-endocrinal disturbances including diabetes mellitus; (3) chorioamnionitis or congenital infections; (4) low birth weight <2.5 kg; (5) congenital malformations of the central nervous system or gastrointestinal anomalies; (6) chromosomal abnormalities; or (7) postnatal age older than 6 h. The study also did not include infants presenting in extremis such as: (1) severe hypoxemia requiring full-oxygen supplement (FiO₂=100%) after delivery room resuscitation; (2) life-threatening coagulopathy; or (3) deep coma. Patients underwent full neurologic assessments at enrollment, performed by a single investigator (HE). The severity of HIE was graded according to a modification of the system described by Sarnat and Sarnat.²²

A group of healthy control (N=15) neonates were enrolled to compare different laboratory measurements (Figure 1). The study was conducted at Tanta University Hospital (Tanta, Egypt) and was approved by the Pediatrics Ethics Committee at Tanta University. Parental consents were obtained before enrollment of subjects.

Figure 1

Illustration of the study design.

Interventions

Whole-body hypothermia protocol

All HIE infants (N=30) received whole-body cooling introduced manually by turning off the heat source of the radiant warmer or incubator while exposing the infant to ambient temperature. Ice packs were introduced across the chest and under the head and shoulders until infant’s temperature decreased below 34 °C. Temperature was monitored by deep rectal thermometer every 30 min with goal temperature 33 to 34 °C. The radiant warmer heater output (or incubator temperature) was manually adjusted every 30 min if the temperature was below 33 °C. Ice packs were also applied when the temperature increased >34 °C. Hypothermia was allowed for a total of 72 h; infants were then re-warmed by increasing goal temperature by 0.5 °C every hour until reaching 36.5 °C.

This cooling protocol has been used for several years and therefore the staff is well experienced with it and the temperature has been reliably controlled in this population of infants. Similar protocol was previously reported to be safe and effective in a randomized controlled trial.²³ The use of automated cooling devices is a relatively new technology that was not available at the time of the study.

Melatonin administration

Melatonin was administered to the melatonin/hypothermia group (N=15) in a dose of 10 mg kg⁻¹ daily for a total of five doses. This dose was similar to doses used safely in previous neonatal studies.^9,11 Oral route was chosen because oral supplements are readily available over the counter. Although the half-life of administered melatonin is 45 to 60 min in adults,²⁴ it is up to 15 h in premature infants,²⁵ which justifies single daily dosing in neonates. Melatonin tablets (1 or 3 mg per tablet; Puritan's Pride, Oakdale, NY, USA) were crushed, then dissolved in 5 ml of distilled water and administered via an orogastric tube.

Laboratory evaluations

For all infants with HIE (N=30) and the healthy control (N=15), blood samples were obtained for measurement of NO, SOD and melatonin concentrations immediately at enrollment, before any intervention. This process was repeated on the fifth day of life only for the two HIE groups (N=30).

Serum melatonin analysis

Melatonin was measured at enrollment as a baseline measurement of endogenous melanin production in study groups with and without HIE. It was repeated in both the HIE groups at day 5 to examine if melatonin was absorbed when administered enterally. Samples were analyzed for melatonin concentration using an enzyme immunoassay method (Melatonin ELISA, IBL International, Germany).

Superoxide dismutase activity

SOD activity was measured at enrollment in all study groups to detect baseline oxidative stress with and without HIE. It was repeated at day 5 in both HIE groups to examine the effect of melatonin on oxidative stress. SOD activity was determined in plasma supernatants after centrifugation of the chloroform–ethanol-treated plasma samples. Plasma was assayed by a well-defined spectrophotometric method²⁶ using Bioxytech SOD-525 commercial research kit (Oxis International, Portland, USA).

Serum nitric oxide concentration

NO concentration was measured at enrollment in all study groups to detect free radical production with and without HIE. It was repeated at day 5 in both HIE groups to examine the effect of melatonin on free radicals. NO concentration in serum was determined using the nitric oxide assay kit (Thermo Fisher scientific Inc., Rockford, USA).

Electroencephalography

Electroencephalography (EEG) was performed at enrollment and was repeated after 2 weeks of age with a digital computerized apparatus (Neurofax EEG-9000, Nihon Kohden, Tokyo, Japan). Recording was performed with a 16-channel, EEG polygraph system; bipolar montage was used, with electrodes placed on the basis of the 10 to 20 system, as modified for newborns. The single neurologist (TE-G) who interpreted all EEG tracings was not aware of the treatment group. Constantly discontinuous tracings was defined when there was constant alternation between relatively high amplitude bursts and low-voltage intervals that were further classified as exhibiting extreme discontinuity (maximal interval duration of >40 s), severe discontinuity (maximal interval duration of 20 to 40 s) or moderate discontinuity (maximal interval duration of <20 s).²⁷ Moreover, the incidences of paroxysmal abnormalities (abnormal EEG transients) and ictal EEG discharges were evaluated.

Brain magnetic resonance imaging

Infants with HIE who survived in both groups were transported to the magnetic resonance imaging (MRI) unit after 2 weeks and when they were clinically stable. Imaging of the brain was performed using the Siemens Magnetom Symphony 1.5-T system (Siemens, Munich, Germany). Images were obtained in the transverse plane, with T1-weighted spin echo, T2-weighted spin echo and age-related inversion recovery sequences. The posterior limb of the internal capsule was assessed as normal, equivocal or abnormal. The basal ganglia and thalami are assessed as normal or with minimal, moderate or severe abnormalities.²⁸ Abnormalities were classified as: (1) minimal, if focal lesions were seen but the posterior limb of the internal capsule was normal; (2) moderate, if focal abnormalities involving the posterior lentiform nuclei and ventrolateral nuclei of the thalami were seen, with equivocal or abnormal signal intensity within the posterior limb of the internal capsule; and (3) severe, if widespread abnormalities were seen in all regions of the basal ganglia and thalami and abnormal signal intensity within the posterior limb of the internal capsule. White matter abnormalities were documented according to which lobes of the brain were involved, whether there was a hemorrhagic element to the lesions and whether they were subcortical, periventicular or widespread. White matter abnormalities were graded as moderate or severe: (1) moderate, indicated small focal lesions with short T1 and short T2, consistent with hemorrhage and/or areas of exaggerated long T1 and long T2 but no loss of gray/white matter differentiation; and (2) severe, indicated more marked areas of abnormality, with larger areas of hemorrhage or exaggerated long T1 and T2 with loss of gray/white matter differentiation, consistent with infarction.²⁹ All images were assessed by an experienced radiologist (MH), who was masked to EEG readings and treatment groups.

Neurologic and developmental outcomes

Detailed neurologic examinations at the age of 6 months were performed by a single pediatric neurologist, not aware of treatment group allocation. The patients were further evaluated with the Denver Developmental Screening Test II (DDST-II).³⁰ The test includes multiple items to examine four major categories (gross motor, language, fine motor-adaptive and personal-social). Infants were scored for each test item as advanced, normal, caution or delayed. The overall developmental assessment of an infant was considered failed if there were ⩾2 delays, questionable if there were one delay and/or ⩾2 cautions and normal if there were no delays and a maximum of one caution. DDST-II is a simple, easily administered screening test that was shown to accurately predict severe adverse outcomes.³¹

Statistical analyses

Data were analyzed using SPSS (Statistical Package for the Social Science, Chicago, IL, USA) version 18. Data were summarized using mean, s.d. for continuous variables and percentage for categorical variables. Comparisons between groups were done using t-test for continuous variables and chi square/Fisher’s exact test for categorical variables. Repeated continuous measures were compared using paired t-test. Correlation was measured using Spearman's rank correlation coefficient. Differences were considered significant when P values were <0.05. A convenient sample of 45 patients was chosen; because of the pilot nature of the study, power analysis was not feasible.

Results

Thirty newborns with HIE were enrolled and randomized to the hypothermia group (N=15) and the melatonin/hypothermia group (N=15). The two HIE groups were similar in demographic and clinical characteristics (Table 1). Baseline metabolic panel, complete blood count and coagulation profile did not differ between groups. At day 5, follow-up labs were all similar except that the melatonin/hypothermia group had greater hemoglobin concentration and less ionized calcium when compared to hypothermia group (Table 2).

Table 1 Demographic data and clinical characteristics for infants with HIE

Table 2 General laboratory investigations in study groups

Baseline SOD, NO and melatonin concentrations were significantly higher in both HIE groups when compared to the healthy control group (P<0.001). The hypothermia and the melatonin/hypothermia groups did not differ in their baseline concentrations (Figure 2).

Figure 2

Baseline serum melatonin, plasma superoxide dismutase (SOD) and serum nitric oxide (NO) concentrations in the three study groups: healthy control (white), hypothermia group (gray) and melatonin/hypothermia group (black).

At day of life 5, melatonin in the hypothermia group increased from 20.6±2.5 to 32.1±3.5 pg ml⁻¹ (P<0.001) while in the melatonin/hypothermia group it increased from 21±2.4 to 42.7±5.1 pg ml⁻¹ (P<0.001). The difference between baseline and follow-up melatonin was significantly more in the melatonin/hypothermia group (P<0.001; Figure 3).

Figure 3

Changes in superoxide dismutase (SOD), nitric oxide (NO) and melatonin concentration in the hypothermia group (white) and melatonin/hypothermia group (black).

SOD in the hypothermia group decreased from 308.2±58.7 to 235.3±36.7 U ml⁻¹ (P<0.001), while in the melatonin/hypothermia group it decreased from 312.9±51.6 to 278.7±41.5 U ml⁻¹ (P<0.001). The difference between baseline and follow-up SOD was significantly less in the melatonin/hypothermia group (P=0.004; Figure 3).

NO in the hypothermia group decreased from 169.5±35.9 to146.2±35.7 μmol l⁻¹ (P<0.001), while in the melatonin/hypothermia group it decreased from 170.5±18.5 to 112.2±19 μmol l⁻¹ (P<0.001). The difference between baseline and follow-up NO was significantly more in the melatonin/hypothermia group (P<0.001; Figure 3).

Of the hypothermia group, four died while only one patient died in the melatonin/hypothermia group (P=0.33). MRI in survivors (N=25) did not differ in the incidence of basal ganglia and thalamus abnormalities between groups. However, white matter abnormalities were noted in four in the hypothermia group and none in the melatonin/hypothermia group (P=0.014; Table 3).

Table 3 EEG and MRI findings in the study groups

Baseline EEG did not differ between the two HIE groups (Table 3). On follow-up EEG, background abnormalities did not differ between groups; however, the melatonin/hypothermia group had less seizure activity (P=0.032; Table 3). Normal follow-up EEG was reported in 10 of the 14 survivors of the melatonin/hypothermia group; all of whom had normal MRI except one who showed mild abnormalities. In the hypothermia group, 4 out of 11 had normal EEGs, all of whom had normal MRI. EEG abnormalities correlated significantly with MRI abnormalities (r=0.92, P<0.001).

At 6 months of age, all surviving infants (n=25) presented for follow-up neurological examination and developmental screening (11 in the hypothermia group and 14 in the melatonin/hypothermia group). Only 3 infants in the hypothermia group had normal neurological evaluation and normal DDST-II, while 10 infants in the melatonin/hypothermia group had normal examination and DDST-II. Survival without abnormalities at 6 months was significantly increased in the melatonin/hypothermia groups (P<0.001).

Discussion

This is the first clinical trial to demonstrate the feasibility and potential efficacy of melatonin when administered in combination with hypothermia to infants with HIE. Melatonin was associated with favorable biochemical and clinical outcomes without noticeable side effects.

Melatonin concentration in the serum was greater in infants with HIE at baseline than healthy control. This can be attributed to endogenous production of melatonin in response to an HIE-related oxidative stress. Following melatonin administration, the melatonin/hypothermia group had a greater increase in serum melatonin concentration, which confirms the adequate absorption of the enterally administered melatonin in HIE neonates.

Plasma SOD was increased in both HIE groups when compared to the healthy controls. Previous studies correlated the greater SOD activities with the more severe cases of HIE.³² Following both hypothermia and melatonin/hypothermia, there was a significant decline in SOD concentrations denoting an amelioration of the HIE-related oxidative stress. Unexpectedly, the decline in SOD was less significant in the melatonin/hypothermia group as compared to the hypothermia group, which could be explained by the complex and multi-mechanisms of actions for melatonin. Although it directly scavenges oxidative radicals, thereby reducing SOD activity, melatonin upregulates SOD via enzyme mRNA synthesis and eventually enzyme stimulation.³³

Both HIE groups displayed greater serum NO concentrations, compared with healthy controls. These findings are consistent with previous studies that reported a pattern of increased NO concentrations after perinatal asphyxia and was correlated with the severity of brain damage.³⁴ In this study, NO concentrations decreased within 5 days in all HIE infants, however, melatonin was associated with a steeper decline. Previous studies reported favorable effects for hypothermia on oxidative stress,³⁵ melatonin was also shown to decrease lipid peroxidation and nitrite/nitrate concentrations in the serum.³⁶ This study demonstrates a synergistic effect of melatonin and hypothermia on reducing NO load in infants with HIE.

Nine infants had normal brain MRI and normal EEG in the melatonin group, whereas only four infants in the hypothermia group had normal MRI and EEG; the difference between the two groups was not significant (P=0.08). However, there were fewer incidences of white matter abnormalities and seizure activity in the melatonin/hypothermia group.

Survival with normal neurological and developmental assessments at 6 months of age improved with melatonin. DDST-II is a very reliable screening tool to accurately detect severe developmental abnormalities. Combining neurological examination with DDST-II is therefore a clinically acceptable approach for screening at 6 months of age. Finally, combining melatonin to therapeutic hypothermia was not associated with noticeable morbidities.

This is the first study to use melatonin as an adjuvant therapy in neonates receiving therapeutic hypothermia for the management of HIE. This combination was reported only in a piglet model of perinatal asphyxia, where melatonin augmented neuroprotective role of hypothermia with improved cerebral energy metabolism and reduced brain damage.²¹ The improved outcomes observed with melatonin treatment could not be related to other biases. Investigators performing laboratory evaluations, EEG and MRI interpretations and developmental screening were all masked to group assignments. The study was randomized to eliminate the chance of any selection bias. We could not find a specific difference in management between groups to which we could attribute the improvement in the melatonin group.

A limitation of this study is the small number of patients. Four patients had severe encephalopathy in the hypothermia group whereas only two in the melatonin/hypothermia group were diagnosed with severe encephalopathy. Although that was not statistically significant, it can be a source of bias in the presence of a small sample size. We should also be cautious with the interpretation of the neurodevelopmental follow-up data since it was limited to only 6 months of age. Larger studies with extended follow-up are needed to examine long-term protective effects. In addition, studies utilizing different dosing regimen would be helpful to establish the most efficacious dose of this simple, yet powerful supplement.

Conclusion

Combination of melatonin to therapeutic hypothermia in infants with moderate to severe HIE was efficacious in reducing oxidative stress and improving survival with favorable neurodevelopmental outcomes at 6 months of age.