Fetal growth restriction (FGR) is a significant health care issue, affecting 20,000 Australian pregnancies every year. Undetected FGR is one of the key risk factors for stillbirth, but FGR can also cause significant impairments in short and long-term health outcomes for the child.

It is a major risk factor for preterm birth and is a recognised causal pathway to the neurodevelopmental injury underlying cognitive and behavioural impairment and cerebral palsy. Current obstetric care is focused on the detection of the growth restricted fetus and then ultrasound assessment of fetal wellbeing to guide timing of delivery. This approach seeks to maximize the gestational age of the fetus at delivery to minimise the risks of prematurity, while delivering the fetus in time to reduce the likelihood of stillbirth. Currently, no therapies exist that can maximize fetal wellbeing in the setting of growth restriction and minimise the frequency of antenatally acquired brain injury due to in-utero hypoxia.

This triple-blind, randomized, parallel group, placebo-controlled trial will administer maternal melatonin or placebo supplementation antenatally in the setting of early-onset severe FGR to determine whether melatonin can PROTECT the fetal brain and lead to improved neurodevelopmental outcomes.

Following detection of FGR, current goals in clinical care center on assessment of fetal wellbeing and evidence of a physiological adaption to placental insufficiency. This information guides the timing of steroids, if indicated, and planning of delivery to minimise the likelihood of stillbirth. Magnesium sulphate is the only available therapy shown to improve fetal brain development in the setting of placental insufficiency and hypoxia. Magnesium sulphate works through reducing glutamate release in a hypoxic environment, likely minimising hypoxic brain injury. It appears to reduce the risk of subsequent cerebral palsy by approximately 30%. However, magnesium sulphate is only used in the hours immediately before birth, while a significant proportion of underlying brain injury in FGR probably occurs over the preceding days to weeks. The use of a safe, maternally administered supplement commenced in the weeks prior to birth could provide further significant benefits in reducing the complications faced by premature infants in the setting of placental insufficiency.

Melatonin (5-methoxy-N-acetyltryptamine) is an endogenous lipid-soluble hormone produced primarily by the pineal gland in humans. It provides circadian and seasonal timing cues due to neuroendocrine control in response to daylight. As such, melatonin secretion is relatively low during the daytime, with an exponential increase in synthesis and secretion occurring from mid-afternoon and peaking at midnight.

In addition to timing cues, melatonin is a powerful antioxidant, acting both as a direct scavenger of oxygen free radicals, especially the highly damaging hydroxyl radical, and indirectly via up-regulation of antioxidant enzymes including glutathione peroxidase, glutathione-reductase, superoxide dismutase and catalase. The metabolites of melatonin provide further anti-oxidant effect.

Melatonin is an appealing treatment for use as a fetal neuroprotectant in pregnancy, as it freely crosses the placenta and blood-brain barrier. It also has an excellent safety profile with no known adverse effects. Placentae express receptors for melatonin, and thus melatonin may protect against oxidative stress generated by ischaemia-reperfusion injury of the placenta.

Melatonin has been studied in several clinical trials related to human reproduction and for different purposes. However, no randomized trial assessing the role of melatonin in fetal neuroprotection has been completed. Melatonin has been evaluated in assisted reproductive technology where the quality of oocytes is vital for the success of in-vitro fertilization (IVF). Melatonin and myo-inositol are two compounds found in the follicular fluid that are important for oocyte maturation and quality. Tamura et al. (in 2008) and Rizzo et al. (in 2010) conducted clinical studies where they co-treated patients with 2milligram (mg) and 3mg melatonin respectively. The patients in the Tamura et al. study were given melatonin from the fifth day of the previous menstrual cycle until the day of oocyte retrieval. Both studies revealed improved oocyte quality, but the tendency to increase pregnancy rates failed to reach statistical significance. A study conducted by Unfer et al. in 2011 administered 2g myo-inositol, 200µg folic acid plus 3mg melatonin per day for 3-months to women who failed to become pregnant in previous IVF cycles, at the commencement of a new IVF cycle. This treatment resulted in a total of 13 pregnancies, 9 of which were confirmed ultrasonographically and 4 undergoing spontaneous abortion. Treatment continued after completion of the IVF cycle, throughout pregnancy until delivery. Treatment was associated with better quality oocytes and more successful pregnancies. All babies that were born from melatonin-treated pregnancies were in healthy condition with no abnormalities.

To evaluate the maternal-fetal transfer of melatonin a study by Okatani et al. in 1998 administered a single oral dose of 3mg melatonin to 33 women at term (37-40 weeks gestation) 1- to 4-hours before a planned caesarean section. Levels of melatonin were evaluated in maternal venous blood and umbilical venous and arterial blood. A total of 12 healthy pregnant women delivered by vaginal birth served as controls. Administration of melatonin led to a rapid (<120 minutes) and marked (>20-fold) increase in the fetal serum levels. There were no differences between maternal and fetal serum levels of melatonin, suggesting a rapid and unrestricted transfer of melatonin from mother to fetus.

The same investigators tested whether melatonin could up-regulate antioxidant enzymes. No longer than 12 hours before voluntary termination of pregnancy (between 7- and 9-weeks gestation), an oral dose of 6mg melatonin was administered to 47 pregnant women. A significant increase of the antioxidant enzyme glutathione peroxidase was observed in chorionic homogenates derived after the procedure, leading to the conclusion that melatonin might provide an indirect protection against injury caused by reactive oxygen species as seen in preeclampsia, FGR and fetal hypoxia.

The dose used in this trial is based on data from a clinical trial of melatonin for preeclampsia showing that 30mg per day was safe for mother and baby without any apparent adverse effects. Venous cord blood concentrations of melatonin achieved were unchanged between a mother receiving 8mg and 30mg per day of melatonin (melatonin concentration ~2100pg/mL). This cord blood concentration would appear sufficient for neuroprotection according to information in sheep models. However, the degree of oxidative stress reduction achieved within the placental bed was less in mothers receiving 8mg melatonin per day. As such, it was felt that the higher dose of 30mg per day was more likely to achieve a clinically significant result.

The investigating team has shown that melatonin supplementation exerts multiple anti-oxidant and anti-inflammatory effects, leading to a significant reduction in oxidative stress and lipid peroxidation within the fetal brain in an ovine model of FGR. In the absence of melatonin, this study showed that lipid peroxidation within the fetal brain led to significant white matter hypomyelination and axonal injury, causing impaired neurological performance in the lambs. Injury was ameliorated entirely in those exposed to melatonin supplementation, with no structural brain injury seen and neurodevelopmental outcomes normalised.

As a result, a small (n=12) phase 1 trial was conducted at Monash Health supplementing pregnancies affected by severe FGR with 8mg of melatonin per day. Melatonin use was well tolerated with no adverse effects seen. A reduction in the degree of placental lipid peroxidation was seen (n=6).

Early-onset FGR carries significant fetal risks of premature birth. Following diagnosis, those babies requiring delivery <32 weeks gestation carry approximately an 8% risk of stillbirth or neonatal death, with those born <28 weeks gestation having a significantly higher perinatal mortality rate. Around 30% of survivors will suffer serious neonatal morbidity. Furthermore, 8% are found to have neurodevelopmental impairment at two years of life. These numbers are likely to be an underrepresentation as they are from a trial population, which was closely surveyed compared to the general population.

With approximately 97% of FGR infants born <32 weeks delivered by caesarean section, the mother of a preterm FGR fetus faces the risks associated with morbidity and mortality relating to caesarean birth. Furthermore, the mother also faces a significant risk of morbidity and mortality from pre-eclampsia, which develops among 15 - 40% of women who have a growth-restricted fetus.

The most common side effects of melatonin are headache, dizziness, nausea and sleepiness. Melatonin does not have any acute pharmacological effects on the nervous or vascular systems, apart from its benign but active impact on sleep mechanisms. Extremely high doses of up to 800mg/kg of melatonin were safely administered to animals without deaths, meaning a median lethal dose could not be established. In humans, long-term treatment with high, daily doses of up to 10g melatonin did not cause any toxicity except for isolated cases of cutaneous flushing, abdominal cramps, diarrhoea, scotoma lucidum and migraine.

Prolonged ingestion of 1g melatonin per day caused only subjective drowsiness but did not provoke any toxicity in the eyes, liver, kidneys and bone marrow. In a phase II clinical trial conducted in the Netherlands, 1400 women were given 75mg melatonin nightly over 4-years, with no side effects reported.

The safety of melatonin use in pregnancy was explored in early pregnant Sprague-Dawley rats, at doses ranging from 1 to 200mg/kg/day and did not affect antenatal mortality, fetal body weight or other measures of fetal wellbeing. Maternal adverse effects seen at high doses, included mild sedation, reduced maternal weight gain and reduced food intake. This study sought to determine the maternal and fetal no adverse effect level (NOAEL). The NOAEL is the exposure level where a particular substance does not statistically or biologically significantly increase the frequency or severity of adverse effects in an exposed population compared to a suitable control population. The maternal NOAEL in this study was found to be 100mg/kg/day, the fetal NOAEL was established at ≥200mg/kg/day when administered to the mother. The maternal lowest observed adverse effect level toxicity was 200mg/kg/day. With the above information taken in context, the Australian Therapeutic Goods Administration (TGA) has assigned melatonin a Pregnancy Category B3 classification.

The investigators have recently completed a phase 1 trial (NCT01695070) using melatonin supplementation in pregnancy, as well as a clinical trial in women with pre-eclampsia (ACTRN12613000476730) using the same dose as proposed for this trial, and to date no adverse effects have been identified in the mother, fetus or neonate.

PROTECT Me aims to be a multicentre, triple-blinded, randomized, parallel group, placebo controlled trial. This trial will be undertaken and co-ordinated by Monash Health.

Other perinatal hospitals across Australia and New Zealand have agreed to join the trial so far. Each centre will nominate a local investigator +/- a researcher to oversee local recruitment.

The required sample size has been calculated to detect if melatonin supplementation affords a clinically relevant difference in neurodevelopmental outcomes among survivors. An increase of 4-5 quotient points in the Bayley-IV Cognitive scale has been deemed sufficiently clinically meaningful to drive changes in health policy previously. Power analysis shows that 69 participants per group will allow the detection of a difference in the Bayley-IV cognitive score of 5 points between the two groups, with a power of 90% and an alpha level of 0.05, using 2 sided T test for comparison. This assumes a standard deviation of 9 and that, on average, the growth restricted infant has been shown to have a cognitive score 5 points lower than the healthy preterm infant and 8 points lower than the healthy term infant. Typically, the Bayley IV score has a standard deviation of 15, however reduced variability has been seen in the FGR population and this has informed the standard deviation used here. Among pregnancies complicated by early onset FGR a perinatal loss rate of ~15% is commonly observed. Allowing for a perinatal loss rate of 15%, an extra 44 women will be recruited. Assuming an additional 5% loss to follow-up rate, the investigators will aim to recruit an extra 14 participants.

This trial also aims to assess whether the impact of melatonin is different at different gestational ages. Therefore, a sub-analysis will be undertaken to compare those with early onset FGR identified <28 weeks' gestation to those with late-onset FGR identified between 28-31+6 weeks gestation. To ensure that this sub-analysis is adequately powered, participants recruited will be randomized to either melatonin or placebo based on their gestational age at diagnosis. Therefore, recruiting 84 participants per group will see the overall trial aiming to recruit 336 participants.

Inclusion Criteria:

1. Singleton Pregnancy

2. Severe fetal growth restriction, defined as:

   • Abdominal circumference ≤3rd centile for gestational age according to charts supplied that have been adapted from Westerway et al; or

   • Abdominal circumference <10th centile in combination with at least one abnormal fetoplacental Doppler study, being:

     • Uterine artery (raised pulsatility index ≥95th centile)
     • Umbilical artery (pulsatility index ≥95th centile or absent/reversed end-diastolic flow)

3. Confirmed 23+0 - 31+6 weeks' gestation

4. Age ≥18 years

5. Understand English

Exclusion Criteria:

1. A fetus with a known chromosomal, major structural anomaly or non-placental cause of fetal growth restriction
2. Pregnancies requiring immediate delivery (e.g. absent A wave in ductus venosus, preterminal CTG or biophysical profile)
3. Co-recruitment in another clinical trial where a pharmaceutical product or nutritional supplement impacting on oxidative stress is the trial intervention.
4. Currently prescribed Fluvoxamine

Triple-blinded, parallel group, placebo controlled trial