Melatonin as a Treatment after Traumatic Brain Injury: A Systematic Review and Meta-Analysis of the Pre-Clinical and Clinical Literature

Abstract

Traumatic brain injury (TBI) is common; however, effective treatments of the secondary brain injury are scarce. Melatonin is a potent, nonselective neuroprotective and anti-inflammatory agent that is showing promising results in neonatal brain injury. The aim of this study was to systematically evaluate the pre-clinical and clinical literature on the effectiveness of melatonin in improving outcome after TBI. Using the systematic review protocol for animal intervention studies (SYRCLE) and Cochrane methodology for clinical studies, a search of English-language articles was performed. Eligible studies were identified and data were extracted. Quality assessment was performed using the SYRCLE risk of bias tool. Meta-analyses were performed using standardized mean differences (SMD). Seventeen studies (15 pre-clinical, 2 clinical) met inclusion criteria. There was heterogeneity in the studies, and all had moderate-to-low risk of bias. Meta-analysis of pre-clinical data revealed an overall positive effect on neurobehavioural outcome with SMD of 1.51 (95% CI: 1.06–1.96). Melatonin treatment had a favorable effect on neurological status, by an SMD of 1.35 (95% CI: 0.83–1.88), and on cognition by an SMD of 1.16 (95% CI: 0.4–1.92). Melatonin decreased the size of the contusion by an SMD of 2.22 (95% CI: 0.8–-3.59) and of cerebral edema by an SMD of 1.91 (95% CI: 1.08–2.74). Only two clinical studies were identified. They were of low quality, were used for symptom management, and were of uncertain significance. In conclusion, there is evidence that melatonin treatment after TBI significantly improves both behavioral outcomes and pathological outcomes; however, significant research gaps exist, especially in clinical populations.

Introduction

Traumatic brain injury (TBI) is one of the commonest causes of neurological morbidity and death worldwide.¹ Recent estimations suggests an incidence of 790/100,000 person years.² Despite being common, evidence supporting its management and specific treatments are lacking.³ After the acute injury, a torrent of complex and varied pathophysiological processes ensue that unfortunately result in further significant secondary brain injury, but that also potentially offers a therapeutic window.⁴ Part of the problem in making treatment advances is that the injuries themselves are diverse, and they occur in markedly varied biopsychosocial settings, which influences outcome. As many treatments that have focused on a specific pathway or symptom have failed to show efficacy in human studies, it becomes attractive to consider a nonselective agent as a therapeutic candidate.^5,6

Melatonin (MEL) could be a promising neuroprotective agent in TBI. Although MEL's role in the chronoregulation of major physiological processes (e.g., the sleep–wake cycle) is well accepted,^7,8 more recently its therapeutic potential is being explored in acquired brain injury, most notably neonatal hypoxic-ischemic encephalopathy.^9
–11 MEL has pluripotent antioxidant and anti-inflammatory properties,^12

–15 which are both receptor mediated (at physiological levels) and non-receptor mediated (especially at supra-physiological levels).^16

–19 Further, its lipophilic properties allow it to cross cell membranes easily and reach subcellular compartments.²⁰ The latter is a useful property, considering that TBI results in widespread cellular process disruption such as metabolic cascades, indiscriminate neurotransmitter release, oxidative stress, and mitochondrial dysfunction. Indeed, a recent Neurotrauma Pharmacology working group identified a need for pharmacotherapies that promote neurorepair, neuroregeneration, and neuroprotection.⁶

MEL affords neuroprotection through a wide variety of mechanisms,¹⁸ including acting as a direct free radical scavenger and antioxidant²¹ through reduction in oxidative stress by decreasing oxidative/nitrosative species and by increasing antioxidant enzymes.^22

–25 MEL also improves mitochondrial function by increasing electron transport and the function of complexes I and IV and by decreasing direct mitochondrial oxidative damage.^26
–28 As an indirect and direct consequence of these actions, MEL has also been shown to inhibit programmed cell death (apoptosis).^29
–31

In addition to providing protection from toxic metabolic intermediaries, MEL has also been shown to modulate neurotransmitter effects in TBI and other neurological conditions. Excessive release of the excitatory neurotransmitter glutamate occurs immediately after TBI.³² MEL has the potential to “balance” this because of its action at inhibitory gamma-aminobutyric acid (GABA) receptors, especially GABA_A receptors.^33

–36 It has been also shown to decrease the neurotoxicity associated with beta-amyloid, which accumulates in several neurodegenerative diseases including chronic traumatic encephalopathy (CTE).³⁷ Neuroinflammation is also thought to play a role in TBI and CTE.³⁸ Although inflammation and glial cell activation after TBI can be beneficial, if it is excessive, it can lead to significant damage and impaired function. MEL is a potent anti-inflammatory agent.^13,39,40 It achieves this partly through cytokine signaling and also indirectly by decreasing the inflammatory mediators nitric oxide and malondialdehyde production.⁴¹

Not only is MEL attractive because of its neuroprotective properties, it also offers therapeutic potential for many of the common post-TBI symptoms such as sleep disruption, pain, mood disturbance and increased anxiety.^42

–47 MEL can help the initiation of sleep via its action on MEL receptors,⁴⁸ and can also decrease pain both at the tissue level and by modulating the opioid and GABAergic systems.^49,50 MEL may be useful in migraine,^51,52 and has efficacy in disorders of chronic pain and anxiety.^53,54 Importantly, MEL has an excellent side effect profile, with good tolerability in high does and even in children.^55
–57

In summary, MEL may offer safe nonselective neuroprotection and symptomatic treatment following TBI. The aim of this study was determine the effect of treatment with MEL (or other melatonergic agent) on the outcome (anatomical, cognitive, physical, or behavioral) after TBI compared with control or usual care in either humans or animals.

Methods

The following databases were comprehensively searched between July and December 2017 by investigator K.M.B.: The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library), PubMed (1951 to December 2017), CINAHL (1982 to December 2017), PsycINFO (1966 to December 2017), MEDLINE^® (Ovid) (1948 to December 2017), Embase (1988 to December 2017), and the National Institute of Health Clinical Trials Database, 2017. The following search strategy was used using the MeSH headings or keywords: (1) human OR animal AND brain injuries AND melatonin OR N-acetyl-5-methoxytryptamine OR melatonergic agent AND child development OR infant development OR cognition OR intellectual disability OR developmental disabilities OR psychomotor performance OR psychomotor disorders OR sleep OR psychological OR behavior OR language OR outcome OR mortality OR morbidity OR brain damage, chronic. Studies were included if they were reported in English or if translations could be obtained from the author.

Selection criteria

Studies had to meet the following selection criteria for inclusion. (1) The study had to be a randomized controlled trial (RCT) or comparative trials using a control group or case series (> 10 participants) satisfying inclusion criteria for types of participants, interventions, and outcomes. (2) The study population included adults and children (0–19 years of age, including neonates) or animals with a TBI. (3) Treatment with MEL or N-acetyl-5-methoxytryptamine or MEL analog or melatonergic agent. The intervention group had to have been compared with a group receiving sham, placebo, or other nonexperimental control in animal studies. In human studies, in order to be inclusive, the same criteria were used as for animal studies, and also included case series. (4) Treatment could begin either before the injury or at any time point post-injury and could be provided in the laboratory, at home, or in hospital. (5) There could be any validated outcome measure assessing anatomical, physical, cognitive, and/or behavioral outcomes. A wide variety of measures tend to be employed across studies; therefore, we did not require that each one had been validated in the TBI population, as this was anticipated to limit the number of studies that could be included. Functional outcome assessed by objective, validated, reliable scales were included. Studies using global outcome scores such as the Glasgow Outcome Score (GOS), and mortality and morbidity rates were also included. (6) There had to be full text available within a peer-reviewed journal, published in English. Articles that reported on the same sample were treated as a single study.

Two of the authors independently (K.M.B. and M.V.) assessed the eligibility of studies for inclusion in the review by first reviewing articles by title and abstract to exclude those not meeting inclusion criteria. Articles that appeared to meet the inclusion criteria were further evaluated by reading the full text. Consensus for article inclusion was reached by discussion among the authors (K.M.B., M.V., and M.J.E.).

Data extraction

Details about the study design and population demographics were extracted from the included studies (Table 1). The theoretical structure, content, and dosage of the intervention programs were tabulated (Table 2). Relevant data from the studies were extracted by the first author, including outcome measures (Table 3).

Table 1.

TBI Model, Study Design, and Quality Assessment of Included Studies

Study	TBi model	Animal	Age	Sex	Weight	Design	Treated (n)	Control (n)	QA score
Ates 2006⁶⁰	CCi	Rat^a	Adult	M	200–250g	RCT	12	12	17
Bayir 2008⁶¹	CHi	Rabbit^b	Adult	MF (7,5)	2000–2500g	CT	6	6	16
Beni 2004⁶²	CHi mod/severe TBi	Mouse^c	Adult	M	30–42g	CT	10 (7 path)	23 (7 path)	16
Campolo 2013⁴⁰	CCi	Mouse^d	Adult	M	25–30g	RCT	10	10	15
Dehghan 2013⁶³	Marmarou severe TBi	Rat^e	Adult	M	250–300g	RCT	7	7	14
Ding 2014⁶⁴	Weight drop	Mouse^f	Adult	M	28–32g	CT	10, (6 path)	10, (6 path)	15
Ding 2015⁶⁵	Marmarou	Mouse^f	Adult	M	28–32g	CT	5	5	17
Kabadi 2010⁶⁶	Fluid percussion	Rat^g	Adult	M	300–350g	CT	7–8	9	16
Kelso 2011⁶⁷	CCi	Rat^g	Adult	M	225–275g	RCT	4	4	16
Lin 2016⁶⁸	CCi	Rat^g	Adult	M	220–250g	RCT	10 (5 path)	10 (5 path)	17
Mesenge 1998⁶⁹	CHi Weight drop	Mouse^h	Adult	M	25–30g	RCT	14–16	16	13
Ozdemir 2005⁷⁰	CHi Weight drop	Rat^a	Juvenile	MF	Age: P7	CT	7	7	17
Sarrafzadeh 2000⁷¹	CCi Moderate TBi	Rat^g	Adult	M	Mean: 300g	RCT	7	7–8	13
Wu 2016 ³⁰	CCi	Rat^g	Adult	M	300–330g	RCT	12	12	13
Yamakawa 2017⁷²	mTBi accel/decal	Rat^g	Juvenile	MF	Age: P30	RCT	5	5	15

Human studies		Setting	Population		Mean age (years)
Kemp 2004⁷³	TBI severity: mild (2), moderate (3), severe (2)	Community rehabilitation program	Adults with sleep problems	M	39.4 (range: 17–55)	Double-blind crossover RCT	7	7	19/29ⁱ
Kuczynski 2013⁴⁶	Mild TBI	Tertiary concussion clinic	Children with post-traumatic headaches	66% F	14.1 (SD 3.1)	Retrospective cohort study	12	32	11/29ⁱ

Animal species: ^aAlbino Wistar rat; ^bNew Zealand rabbit; ^cSabra mouse; ^dCD1 mouse; ^eAlbino N-Mary rat; ^fICR mouse; ^gSprague-Dawley rat; ^hSwiss mouse.

Traumatic brain injury (TBI) models: Controlled cortical impact (CCI); closed head injury (CHI); Marmarou method (diffuse injury); fluid percussion; acceleration/deceleration model (accel/decl).

Study designs: Randomized controlled trial (RCT); controlled trial (CT).

Quality assessment (QA) SYstematic Review Centre for Laboratory animal Experimentation (SYRCLE) risk of bias tool; ⁱDowns and Black criteria

Table 2.

Details of Intervention

Study	Treatment	Dose (mg/kg)	Total dose (mg/kg)	Control	Time post-injury	Route	Treated (n)	Control (n)
Ates 2006⁶⁰	MEL	100	100	TBI alone	Immediately	IP	12	12
Bayir 2008⁶¹	MEL	100	100	Mannitol 20% 2g/kg	4h	IP	6	6
Beni 2004⁶²	MEL	1	1	Vehicle (5% ethanol)	1h	IP	10	23
		5	5				23
		10	10				8
Campolo 2013⁴⁰	MEL	10	20	Vehicle (1% ethanol)	1h + 6h	IP	10	10
Dehghan 2013⁶³	MEL	5	20	Vehicle (ethanol)	1 + 24 + 48 + 72h	IP	7	7
Dehghan 2013⁶³	MEL	20	80	Vehicle (ethanol)	1 + 24 + 48 + 72h	IP	7	7
Ding 2014⁶⁴	MEL	10	50	Vehicle (5% ethanol)	Immed +1 + 2 + 3 + 4h	IP	10	10
Ding 2015⁶⁵	MEL	10	50	Vehicle (5% ethanol)	Immed +1 + 2 + 3 + 4h	IP	6	6
Kabadi 2010⁶⁶	MEL	100	100	Vehicle (PEG)	immediately	IP	7–8	9
Kabadi 2010⁶⁶	MEL	200	200	Vehicle (PEG)	immediately	IP	7–8	9
Kelso 2011⁶⁷	MEL	5	10	Vehicle (2% ethanol)	5min +1h	IP	4	4
Lin 2016⁶⁸	MEL	5mL/kg	15mL/kg	Vehicle (2% ethanol)	Immed +2 + 4h	IP	10	10
Mesenge 1998⁶⁹	MEL (Pilot I)	1.25	5	Vehicle (HCl)	Ia 20min pre-injury +5min +1 + 2h	IP	14	16
	MEL (Pilot I)	1.25	5	Vehicle (HCl)	Ib 5min +1 + 2+ 3h	IP	15	16
	MEL (Therapeutic window II)	1.25	5	Vehicle (HCl)	IIa 5min +1 + 2+ 3h	IP	15	13
		1.25	5		IIb 30min +1 + 2+ 3h	IP	15
		1.25	5		IIc. 1 + 2+ 3 + 4h	IP	15
	MEL (Dose finding III)	0.625, 1.25, or 2.5	2.5, 5, or 10	Vehicle (HCl)	5min, 1 + 2+ 3h	IP	15 each dose	13
Ozdemir 2005⁷⁰	MEL	5	5	Vehicle (5% ethanol)	immediately	IP	7	7
Sarrafzadeh 2000⁷¹	MEL	100	300	Vehicle (0.1% ethanol)	Day: Immed +1 + 2h	IP	7	7–8
Sarrafzadeh 2000⁷¹	MEL	100		Vehicle (0.1% ethanol)	Night: Immed +1 + 2h	IP	7	7–8
Wu 2016 ³⁰	MEL	10	50	Vehicle (1% ethanol)	Immed +1, 2, 3, and 4h	IP	12	12
Yamakawa 2017⁷²	MEL	4.5	4.5	Vehicle (ethanol)	Day 1–14	Oral	5	5

Human studies
Kemp 2004 ⁷³	MEL	5mg orally at night	-	AMIT 25mg	Cross over trial: 1st arm, 4 weeks; washout period, 2 weeks 2nd arm, 4 weeks	Oral	7	7
Kuczynski 2013⁴⁶	MEL	3–10 mg	-	AMT 10–50mg; other treatments	MEL at night for 4 weeks or AMT at night or other treatment for 8 weeks	Oral	12	AMT 18; other 14

MEL, melatonin; HCl, hydrogen chloride; PEG, polyethylene glycol; AMT, amitriptyline; IP, intraperitoneal.

Table 3.

Details of Outcome Measures

Study	Outcome measure	Function	Scale and direction	Time assessed
Contusion/Traumatic lesion size
Ates 2006⁶⁰	Contusion volume, %	Quantitative digital analysis of the lesion size (5mm section, % ipsilateral hemisphere)	0–100%Lower score is more favorable outcome	14 days
Beni 2004⁶²	Contusion size, %	Quantitative digital analysis of lesion, (1mm sections, % of opposite hemisphere × 2)	0–100%Lower is better	24h
Campolo 2013⁴⁰	Lesion volume, mm³	Lesion volume calculated from 2mm sections	Lower is better	24h
Kelso 2011⁶⁷	Cortical tissue spared, mm³	Percentage of total cortical tissue volume	0–100Higher is better	12 days
Lin 2016⁶⁸	Lesion volume, mm³	Quantitative digital analysis of lesion (0.5mm sections)	Lower is better	8h
Sarrafzadeh 2000⁷¹	Contusion volume, mm³	Quantitative digital analysis of lesion (0.4mm sections)	Lower is better	24h
Cerebral edema
Dehghan 2013⁶³	Brain water content, %	Brain water content increases as cerebral edema increases: % water content of whole brain calculated by wet weight/dry weight method (WW/DW)	0–100Lower is better	72h
Lin 2016⁶⁸				8h
Wu 2016 ³⁰				24h
Ding 2014⁶⁴		% ipsilateral hemisphere, WW/DW		24h
Ding 2015⁶⁵				24h
Kabadi 2010⁶⁶				48h
Sarrafzadeh 2000⁷¹				24h
Sarrafzadeh 2000⁷¹	Hemispheric swelling, %	% of contralateral hemisphere;[(WWL - WWR)/WWR] × 100		24h
Neurological status
Beni 2004⁶²	Neurological Severity Score (NSS) change	Change in NSS: 10 behavioral and motor tasks, 1 point awarded for each failed task	0–10Lower = less deterioration	Change from baseline at 1, 4, and 7 days;
Ding 2014⁶⁴	NSS	NSS total score	0–10Lower is better	1, 24,^a and 72h
Ding 2015⁶⁵	NSS	Change in NSS	0–10Higher is better	1,^a 3, and 7 days
Wu 2016³⁰	Modified NSS	Modified NSSwith motor, sensory, reflex, and balance tests	0–18Lower is better	24h
Dehghan 2013⁶³	Vetnary Coma Score (VCS)	Neurological status based on motor (1–8), visual (1–4), and respiratory responses (1–3)	3–15Higher is better	Baseline, 1, 24, 48 and 72h
Cognitive performance
Kelso 2011⁶⁷	Morris Water Maze (MWM) (sec)	MWM Escape latency: time to reach platform (max 60 sec) × 5 trials (1) Data acquisition test over 5 days and (2) retention 3 days later (distance – partially reported)	Maximum 60 secLower is better	(1) Day 5–9(2) Day 12
Lin 2016⁶⁸	MWM (sec)	Escape latency(sec): (1) Hidden platform, (2) visible platform	Maximum 100 secLower is better	(1) Day 11–15(2) Day 16
Ozdemir 2005⁷⁰	MWM (sec)	(1) Escape latency (sec) and(2) Time in correct quadrant (sec)	Maximum 60 sec(1) Lower is better(2) Higher is better	(1) Days 1 to 4(2) Day 5
Yamakawa 2017⁷²	Behavioral test battery:Beam walking (BW), Open field (OF), Elevated Plus Maze (EPM), Novel Context Mismatch (NCM), and Force Swim (FS)	BW: The beam walking task measures acute balance, coordination, and motor disturbances (sec);OF: a measure of locomotion and exploratory behavior;EPM: a behavioral test for anxiety;NCM: a measure of short-term working memory;FS: a measure of depressive-like behavior	BW: lower is betterOF: higher is betterEPM: higher is betterNCM: higher is betterFS: lower is better	BW day 1, OF day 2, EPM day 3, NCM day 7–10, FS day 13
Motor function and coordination
Campolo 2013⁴⁰	Rotarod test (sec)	Time on Rotarod (average of 3 trials)	0–60Higher is better	1 and 6h;
Lin 2016⁶⁸	Balance beam (sec)	Time on beam	Higher is better	Day 1–5
Mesenge 1998⁶⁹	Grip score (sec)	Length of time remains on a string (mainly motor function)	Maximum 0–30Higher is better	24h
Other outcomes
Dehghan 2013⁶³	Intracranial pressure (ICP) (mmHg)	ICP measured over 3 days (baseline, 1, 24, 48, and 72h)	Lower is better	72h^a
Bayir 2008⁶¹	(1) MRI pathology change	Pathologies categorized: parenchymal or extra cerebral hemorrhage, contusion, ventricular effacement; 36 hMRI at 3 h post-injury, intervention given, MRI repeated at 36 h	Lower amounts of change are more favorable; i.e., less cerebral edema	36h
Bayir 2008⁶¹	(2) Brain protrusion through craniotomy	Maximal extrusion point of brain tissue outside craniotomy site	Lower is better	36h
Kemp 2004 ⁷³	Sleep diary;Speed and Capacity of Language-Processing Test (SCOLP), Adult Memory and Processing Battery (AMIPB), and Hospital Anxiety and Depression Scale (HAD)	Tests for change in memory and information processing speeds: (1) Neuropsychological SCOLP test; (2) information processing test from AMIPB; mood assessment:HAD	0–100(1) and (2) higher is better(3) lower is better	End of each treatment arm
Kuczynski 2013⁴⁶	> 50% headache frequency	Reduction in headache frequency	Dichotomous	2 months post- treatment

Reported.

Data synthesis

Quantitative analysis was conducted using Review Manager (RevMan) version 5.3 software (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). Treatment effects were first calculated separately for each study outcome. Treatment effects for pooled data were then calculated across trials, where possible, for specific outcomes and domains. For all analyses, a random effect, inverse variance model was used to calculate standardized mean differences (SMD) and 95% confidence intervals (CI). The effect of heterogeneity (I²) was used to measure the degree of inconsistency across pooled studies caused by variability rather than chance, with larger values indicative of high heterogeneity.

Data quality

Two independent reviewers (K.M.B. and M.V.) assessed the risk of bias and the methodological quality using the SYstematic Review Centre for Laboratory animal Experimentation (SYRCLE) Risk of Bias tool⁵⁸ for animal studies and Downs and Black Criteria⁵⁹ for human studies (Tables 2 and 4). The SYRCLE scale has demonstrated reliability and consists of 10 items each with a score of “high,” “unclear,” or “low” risk of bias. Where no mention of treatment randomization was made, the study was awarded a “high” risk of bias, but randomization without mention of the process was given a score of “unclear.” A “low” risk of bias was awarded when baseline characteristics of weight, sex, age, and strain of the animals were provided. When animals were euthanized immediately without being returned to housing, this category of study was ranked as “low” risk. If outcome was objectively assessed using computerized methods only, this category of study was awarded “low” risk of bias. Otherwise, studies with blinding of outcome assessors were categorized as “high” risk when no specific mention of blinding was made. Selective outcome reporting was not scored, as most animal protocols are not yet published/registered. The quality of human studies was evaluated using the Downs and Black quality assessment tool, which assigns an individual score calculated out of 29 total points for each study. Discrepancies were resolved by discussion. The protocol for this systematic review and meta-analysis was registered with the International Prospective Register of Systematic Reviews (PROSPERO; registration number PROSPERO 2017 CRD4201707302).

Table 4.

Methodology Quality Assessment of Included Studies Using SYRCLE Risk of Bias Tool⁵⁸

	Random sequence generation	Baseline characteristics	Allocation concealment	Random housing	Blinding to allocated treatment	Random outcome assessment	Outcome assessor blinding	Incomplete outcome data	Other sources of bias	Total SYRCLE score	Any randomization	Any blinding	Sample size
Ates 2006	2	1	2	2	3	2	3	1	1	17	Y	N	N
Bayir 2008	3	1	2	2	3	1	1	1	1	16	N	Y	N
Beni 2004	3	1	2	2	2	1	3	1	1	16	N	N	N
Campolo 2013	2	1	2	2	2	1	3	1	1	15	Y	N	N
Dehghan 2013 2013	2	1	2	2	2	1	2	1	1	14	Y	N	N
Ding 2014	3	1	2	2	1	2	1	1	2	15	N	Y	N
Ding 2015	3	1	2	2	1	2	1	3	2	17	N	Y	N
Kabadi 2010	3	1	2	2	2	1	3	1	1	16	N	N	N
Kelso 2011	2	1	2	2	2	1	3	1	2	16	Y	N	N
Lin 2016	2	1	2	2	2	2	3	1	2	17	Y	N	N
Mesenge 1998	2	1	2	2	2	1	1	1	1	13	Y	Y	N
Ozdemir 2005	3	1	2	1	2	3	3	1	1	17	N	N	N
Sarrafzadeh 2000	2	1	2	2	2	1	1	1	1	13	Y	N	N
Wu 2016	2	1	2	2	2	1	1	1	1	13	Y	Y	N
Yamakawa 2017	1	3	2	1	2	1	3	1	1	15	Y	N	N

Results

The selection process for included studies is shown in Figure 1. After removing duplicates, 213 studies were screened by title and abstract review. Twenty-two studies were identified for detailed review and a further three studies were identified from reference lists of review articles. Fifteen pre-clinical studies with unique data examining the neuroprotective effects of MEL and two clinical studies examining its use for symptom management after TBI were included (see Tables 1 and 2). Sample sizes were small in all studies: 4–16 in experimental pre-clinical groups, and 7 and 12 in clinical studies. Sixteen studies were controlled trials; one study was a retrospective controlled cohort study.

FIG. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram of included and excluded studies.

Pre-clinical studies

TBI model and severity of injury

A variety of animals were used: nine studies used rats, five used mice, and one used rabbits, (see Table 1).^{30,40,46,60

–73} Most were young adults, although two models were juvenile rat models: postnatal day 7 and 30.^70,72 In keeping with the influence of sex on outcome following animal and human TBI, adult animals were all male, except for the rabbit model.⁶¹ Several different mechanistic models of TBI were employed; more focal injuries were obtained using closed head injury (CHI), controlled cortical impact (CCI), or fluid percussion models (FPI).^{30,40,60,61,64,66

–71,74} and diffuse injuries were obtained either by the Marmarou method^63,65 or the acceleration/deceleration model⁷² The resultant injuries were moderate to severe in all cases except for one mild TBI model.⁷² Although Bayir and coworkers report their TBI model as mild, the injuries on MRI would suggest moderate/severe TBI.⁶¹

Interventions

The range of MEL doses is shown in Table 2 and ranged from 0.625 mg/kg to 200 mg/kg, although the most frequent dose was 5 mg/kg. The intervention was given once in six studies^{60,61,66,69,70,74} and in nine studies the dose was repeated up to 14 times.⁷² The total MEL dose was between 0.625 and 300 mg/kg. The control intervention was the agent or vehicle used to dissolve and administer the MEL (usually ethanol and saline). All interventions were given by intraperitoneal (IP) injection except in one study, in which it was administered orally.⁷² The interventions were commenced 20 min before the injury,⁶⁹ within 5 min of the injury,^{30,60,64,65,66, 67

–71} or at 1 h^63,74 or 4 h post-injury.⁶¹ No adverse events caused by the interventions were reported in the pre-clinical studies.

Qualitative analysis

None of the pre-clinical studies had published protocols nor were registered with Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies (CAMARADES). Therefore, the selective outcome reporting item on the SYRCLE tool was not scored. There was insufficient information reported for many (41%) of the remaining nine questions, which were scored as “unclear.” Overall, all studies had significant risks of bias (see Table 4). The average SYRCLE score was 15.3 (95% CI: 14.4, 16.1). Although four studies^60,65,68,69 scored above the 95% CI, they were not sufficiently remarkable as to be excluded from any analyses. Nine studies reported any randomization, although details were not given. Only 30% reported any blinding, either of investigators, animal handlers, or outcome assessors (see Fig. 2). None of the studies reported sample size calculations. Indeed, nine studies (60%) had sample sizes of six per group or less.

FIG. 2.

Assessment of bias in 15 animal studies evaluating the efficacy of melatonin to improve traumatic brain injury (TBI) outcome using the SYstematic Review Centre for Laboratory animal Experimentation (SYRCLE) risk of bias tool.

Outcome

A variety of outcome measures were used (see Table 3). The commonest were (1) brain water content as a marker for cerebral edema (eight studies) reported between 8 and 72 h post- injury, (2) lesion/contusion volume (seven studies) measured between 8 h and 14 days post-injury, (3) Neurological Symptom Scores (NSS) or variant (five studies), and (4) the Morris Water Maze test (MWM) (escape latency; used in four studies). The NSS, when used, was measured repeatedly between 1 h and 7 days post-injury (see Table 3). Similarly, the MWM test was measured repeatedly between day 1 and 4 post-injury except in one study.⁶⁸ These measures were used in meta-analyses of the pre-clinical studies. All motor outcomes differed in each of the four studies (see Table 3). Some motor outcomes required more balance and coordination (e.g., Rotarod test) than others (e.g., Grip test). Correct effect sizes and significance levels are reported in Table 5.

Table 5.

Mean, Standard Deviation, Effect Sizes, and 95% Confidence Intervals for Individual Pre-Clinical Study Results

		Melatonin			Control
Study		n	Mean	SD	n	Mean	SD	p value for mean diff (2 tailed t test)	Bias corrected effect size (Hedges)		95% CI
Contusion/lesion size
Ates⁶⁰		%	12	11.7	1.2	12	16.9	1.8	<0.001	−3.274	−2.0	−4.5
Beni⁷⁴		%	7	6.2	1.0	7	11.1	1.0	<0.001	−4.454	−2.5	−6.4
Campolo⁴⁰		mm³	10	37.0	4.7	10	56.0	3.2	<0.001	−4.517	−2.9	−6.2
Kelso⁶⁷			4	84.0	8.0	4	86.0	10.0	0.765	−0.192	1.2	−1.6
Lin⁶⁸			5	8.0	1.0	5	11.2	0.8	0.001	−3.190	−1.3	−5.1
Sarrafzadeh⁷¹ (night)			8	30.5	12.8	7	41.8	13.9	0.126	−0.796	0.3	−1.8
Sarrafzadeh⁷¹ (day)			8	35.6	6.8	8	41.8	16.0	0.334	−0.473	0.5	−1.5
Cerebral Edema
Dehghan⁶³ MEL 5mg/kg		Brain water content, %	7	73.1	0.8	4	75.7	0.4	<0.001	−3.468	−1.6	−5.4
Dehghan⁶³ MEL 20mg/kg			7	72.9	0.7	3	75.7	0.5	<0.001	−3.879	−1.7	−6.1
Wu³⁰			12	79.8	0.3	12	80.9	0.4	<0.001	−3.004	−1.8	−4.2
Lin⁶⁸			5	78.4	1.3	5	81.8	1.6	0.006	−2.105	−0.6	−3.7
Ding 2014⁶⁴			6	80.2	1.7	6	81.9	1.6	0.104	−0.953	0.2	−2.1
Ding 2015⁶⁵			5	80.4	0.9	5	81.9	0.9	0.029	−1.521	−0.1	−2.9
Kabadi⁷⁵ MEL 100mg/kg			8	79.5	0.8	5	80.5	0.8	0.060	−1.110	0.1	−2.3
Kabadi⁷⁵ MEL 200mg/kg			8	79.1	0.3	4	80.5	0.6	<0.001	−3.087	−1.4	−4.8
Sarrafzadeh⁷¹			7	80.5	0.4	7	80.7	1.1	0.696	−0.199	0.9	−1.2
Sarrafzadeh⁷¹		Hemispheric swelling, %	7.9	7	3.4	9.2	7	4.5	0.55	−0.300	0.75	−1.36
Neurological status
Beni⁷⁴ MEL 1mg/kg		NSS	10	4.7	1.0	8	6.5	0.9	0.001	−1.889	−0.8	−3.0
Beni⁷⁴ MEL 5mg/kg			23	4.5	2.2	8	6.5	0.9	0.019	−0.989	−0.1	−1.8
Beni⁷⁴ MEL 10mg/kg			8	4.9	2.9	7	6.5	0.8	0.184	−0.684	0.4	−1.7
Ding 2014⁶⁴ NSS			10	2.9	2.5	10	4.9	1.9	0.060	−0.858	0.1	−1.8
Ding 2015⁶⁵ NSS change		ΔNSS^a	5	3.2	1.3	5	1.6	1.1	0.075	−1.170	0.2	−2.5
Wu³⁰ NSS		mNSS	5	7.5	0.4	5	9.5	0.6	<0.001	−3.541	−1.6	5.5
Dehghan⁶³ MEL 5mg/kg		VSS^a	7	13.1	0.8	3	11.5	0.9	0.021	1.787	0.2	3.4
Dehghan⁶³ MEL 20mg/kg dose		VSS^a	7	14.2	0.8	4	11.5	1.0	0.001	2.850	1.1	4.6
Cognitive function
Memory and new learning
Kelso⁶⁷		MWM	4	48.4	21.2	4	45.3	34.8	0.884	−0.093	1.3	−1.5
			10	58.4	16.4	10	80.5	20.6	0.016	−1.137	−0.2	−2.1
Ozdemir⁷⁰ 5mg/kg			7	38.8	1.4	4	44.9	3.4	0.002	−2.454	−0.9	−4.1
Ozdemir⁷⁰ 20mg/kg			7	40.2	1.6	3	44.9	3.4	0.015	−1.935	−0.3	−3.5
Yamakawa⁷²		NCM^a	5	33.7	8.7	5	25	8.7	0.153	0.903	2.2	1.0
Exploratory behavior
Yamakawa⁷²		OF^a	5	6132	724	5	4887	724	0.026	1.552	0.1	3.0
Anxiety
Yamakawa⁷²		EPM^a	5	38.9	9.2	5	31.2	6.5	0.163	0.876	−0.4	2.2
Depression
Yamakawa⁷²		FS	5	33.8	12.2	5	24.5	7.1	0.179	0.841	0.5	−2.1
Motor function
Yamakawa⁷²		BW	6	5	4.2	6.9	5	5.8	0.786	−0.160	−1.1	1.4
Campolo⁴⁰		MR^a	30.25	10	18.2	25	10	10.27	0.437	0.340	−0.5	1.2
Lin⁶⁸		BB^a	48	10	6.5	25.9	10	6.5	<0.001	3.256	1.9	4.6
Mesenge⁶⁹	Ia	Grip^a	18.9	14	10.9	7	8	7.9	0.014	1.150	0.2	2.1
Mesenge⁶⁹	Ib		17.9	15	11.6	7	8	0.9	0.016	1.108	0.2	2.0
	IIa		22.8	15	10.1	13.3	5	8.7	0.077	0.928	−0.1	2.0
	IIb		15.8	15	12.4	13.3	4	7.8	0.709	0.204	−0.9	1.3
	IIc		10.5	15	13.3	−13.3	4	0.8	0.685	0.221	−0.9	1.3
	IIIa		23.3	15	8.9	14.9	5	8.5	0.081	0.913	−0.1	2.0
	IIIb		19.5	15	10.8	14.9	5	8.5	0.400	0.426	−0.6	1.4
	IIIc		12.9	15	11.6	-14.9	5	8.5	0.729	0.174	−0.8	1.2
Intracranial pressure
Dehghan⁶³ MEL 5mg		ICP	12.1	7	1.1	21.9	3	0.7	<0.001	−8.717	4.7	12.8
Dehghan⁶³ MEL 20mg		ICP	10.9	7	1.1	21.9	4	1	<0.001	−9.415	5.3	13.5
Bayir⁶¹		Brain protrusion	2	6	0.8	3.4	6	1.7	0.098	−0.972	−0.2	2.2

Higher score = more favorable outcome.

Bold indicates statistical significance.

NSS, Neurological Severity Score; mNSS, modified mNSS; VCS, vetinary coma score; MWM, Morris Water Maze; NCM, Novel Context Mismatch, OF, Open Field; EPM, Elevated Plus Maze; FS, Forced Swim; BW, Beam Walking; MR, Motorod; BB, Balance Beam; ICP, intracranial pressure.

Stratified meta-analyses

There was a significant effect of treatment with MEL on both pathological outcome measures (Fig. 2), and behavioral measures (Fig 3).

FIG. 3.

A Forest plot demonstrating the effect of melatonin in comparison with control (vehicle) on pathological outcome measures in pre-clinical studies of traumatic brain injury.

Pathological outcomes

MEL decreased the size of the contusion by an SMD of 2.22 (95% CI: 0.84, 3.59; six studies, seven comparisons). The effect size was large, being >3.0 for four of the six studies, with the largest being 4.5.⁴⁰ MEL decreased cerebral edema by an SMD of 1.91 (95% CI: 1.08, 2.74; six studies, nine comparisons). Six of nine studies reported significantly large effect sizes ranging from 1.5 to 3.9. There was significant heterogeneity among the studies (I²: 84%). However, this finding was supported by two studies not included in the meta-analysis. One reported a decrease in intracranial pressure (ICP) in association with MEL treatment.⁶³ In another study, MEL had a similar effect on brain protrusion (a marker of increased ICP) as 20% mannitol,⁶¹ a first-tier therapy for increased ICP following TBI.⁷⁶

Behavioral outcomes

MEL treatment had a favorable effect on the neurological status (see Fig. 3), by an SMD of 1.35 (95% CI: 0.83, 1.88; five studies, eight comparisons). The outcome closest to 24 h post-injury was compared in the meta-analysis. Five of the eight experiments reported significantly large effect sizes associated with MEL treatment, ranging from 1.0 to 3.5. Meta-analysis demonstrated that overall, MEL improved the performance on a memory-based cognitive task by an SMD of 1.16 (95% CIs: 0.4, 1.92). There was a large effect size ranging from 1.1 to 2.4. Only one study examined the effect of MEL on mood and anxiety.⁷² This was done between days 7 and 13 post-injury, and showed no effect of treatment. Motor outcomes were also improved by MEL (SMD 0.93; 95% CI:0.69, 1.93).

Overall effect

When all studies and comparisons were combined, the overall outcome was improved by an SMD of 1.51 (95% CI, 1.06, 1.96; 15 studies, 20 comparisons, see Fig. 4). One outcome measure from each study was included in the following order of priority: neurological status, cognitive function, contusion size, and cerebral edema. As expected, the heterogeneity was moderately high (I² 58%). Nine comparisons had effect sizes >1.5.

FIG. 4.

Forest plot demonstrating the effect of melatonin in comparison with control (vehicle) on behavioral outcome in pre-clinical studies of traumatic brain injury.

Dose, timing of intervention, and frequency

Four studies examined the effect of differing doses on outcome.^{63,66, 69,74} The dose, timing, and frequency of interventions varied greatly, making it difficult to draw conclusions. The effects of frequency of dose regimes and timing of interventions were analyzed in ad hoc subgroup analyses of the overall outcome. Interventions were considered to be “early” if given before 30 min and late if given after 60 min (see Fig. 5). There was no significant effect of single versus multiple dosing (χ² 2.26, df 1; p = 0.13) or of “early” or “late” treatment subgroups (χ² 0.2, df 1; p = 0.66, see Fig. 5).

FIG. 5.

Subgroup meta-analysis: efficacy of intervention versus control (vehicle) in pre-clinical studies examining single versus multiple dose and early (within 30 min) versus late treatment (≥ 1 h). Random effects assumed; 95% confidence intervals; effect of heterogeneity, I²; 200 mg/kg (1, 12), 5mg/kg (2, 10, 11, 13, 19, 21), 20 mg /kg (3, 5, 14, 15), 100 mg/kg (4, 16), 1 mg/kg (6), nighttime administration (7,17), 10 mg/kg (8, 20), daytime administration (9,18).

Clinical studies

Two clinical studies were identified. There was no placebo-controlled study. Kemp and coworkers examined the effectiveness of MEL in seven male adults with sleep disturbance and the study was of good methodological quality except for being significantly underpowered.⁷⁶ The TBIs ranged from mild to severe and had occurred 36 months (range: 9 − 73) previously. A double-blind randomized crossover study was performed using 5 mg of oral MEL in comparison with 25 mg amitriptyline for 4 weeks.⁷⁶ There was no effect of MEL treatment on sleep or neuropsychological parameters compared with the active comparator, amitriptyline. No adverse events were reported.

Kuczynski and coworkers reported an open-label retrospective cohort study of children (mean age 14 years, SD 3.1) with post-concussion syndrome and post-traumatic headaches 8 months following mild TBI.⁴⁶ Although a beneficial effect of MEL was reported – 75% (95% CI: 49, 88) had >50% improvement in the number of headaches – the study was uncontrolled. The effect on sleep was not reported. No adverse events were reported. Caution should be observed when drawing any conclusion from this study, because of its low methodological quality.

Discussion

The goal of this review was to systematically evaluate the literature for the efficacy of MEL as a potential treatment to improve outcome after TBI. Pre-clinical intervention studies, although often focused on a biological construct, ideally have either a pathological or a behavioral outcome consistent with clinical realities in order to add validity and allow clinical correlation. By necessity, pre-clinical studies often have small sample sizes. A systematic review combined with meta-analysis allow the data to be methodologically and objectively assessed. This is the first systematic review of the evidence supporting the use of MEL in animal models and humans with TBI. Based on the results of our meta-analysis in pre-clinical studies, there is potential for treatment with MEL to improve functional outcome.

MEL significantly improved neurobehavioral outcome in neurological, cognitive, and motor domains, as well as in histopathological domains (contusion size and cerebral edema). There was an overall positive standardized mean difference of 1.51 (95% CI, 1.06, 1.96). This is similar to its effect in a systematic review of pre-clinical stroke models⁷⁷ and meta-analysis of MEL as a treatment for pain in adults.⁷⁸ Although the results of this meta-analysis are strengthened by the effect seen across multiple species (mouse, rat, and rabbit), it is significantly weakened by the significant methodological differences, especially dose and timing of treatments.

After injury, MEL was administered within minutes or up to several hours later, and was used as a single treatment in some and as repeated treatment in other studies. Timing theoretically could target different mechanisms; for example, early administration affecting glutamate toxicity and free radical formation; and later doses targeting neuroinflammation.^13,21 As secondary brain injury is a rapidly progressive process, especially in the first few days, timing of drug administration is likely to be important.⁴ In humans, MEL achieves maximum concentrations at ∼45 min when given orally (2 h if given intravenously), except when slow-release preparations are used (2 h), and its half-life is ∼2 h.⁵⁷ Here, although earlier treatment was found to be beneficial in one larger study,⁶⁹ we did not find an effect of single versus multiple dosing regimens or of early versus late initiation of treatment in subgroup meta-regression analysis. However, the variability of drug dosage (total dose: 2.5–300 mg) make these assessments less reliable. This is especially relevant as larger doses of MEL result in greater serum concentrations, at which supraphysiological non-receptor-mediated effects occur (e.g., direct free radical scavenging, enhancement of mitochondrial function).²⁷ For example, two studies found MEL 5 mg/kg was more effective than 1 mg/kg or 2.5 mg/kg.^69,74

Most pre-clinical studies had moderate or unclear risks of bias. A key factor in this was the lack of reporting detail to allow satisfactory experimental evaluation. Further, many investigators failed to either randomize or blind assessors and animal carers to the treatment condition. No studies provided sample size estimation or published protocols. This emphasizes the need for more rigor in applying reporting standards and for the publishing of experimental protocols.

In clinical studies, there was a lack of well-designed and adequately powered trials, and the two studies identified in the review evaluated the use of MEL in the chronic phase of TBI recovery for specific symptoms, headache,⁴⁶ and sleep.⁷⁶ These were of low quality. The randomized controlled crossover trial was significantly underpowered, and was unlikely to detect difference from an active comparator (amitriptyline).⁷⁶ The other clinical study was retrospective.⁴⁶ Overall, although MEL was well tolerated, there was a lack of data to support the use of MEL as a neuroprotective agent or for specific symptom management in clinical populations.

This current study has many strengths. First, it is a systematic review of the current evidence and followed a published protocol method to ensure a diligent and rigorous review process. By conducting a meta-analysis, studies were combined to increase overall sample size and statistical precision. Finally, included studies had functional neurological outcomes across the many domains of neurological impairment seen after TBI, and, therefore, clinical inferences can be easily made.

Although we attempted to be comprehensive, this review may have failed to identify studies and also may be subject to publication bias. We did not include unpublished data, and this may have skewed our results. All the studies in this review had small sample sizes and when combined with a relative paucity of studies, the meta-analysis is limited by a small data set. This systematic review has not investigated dosage or length of treatment beyond the first few days. It was not possible to examine the effects of MEL in specific TBI populations such as youth and the elderly, who may have different responses to the treatment. Nor were there sufficient studies to allow the evaluation of its use in different TBI pathologies; for example, diffuse versus focal injury.

Conclusions

There is evidence from pre-clinical studies that MEL treatment after TBI significantly improves both behavioral and pathological outcomes; however, significant research gaps exist. There are insufficient clinical data to support routine use following TBI, but further clinical research is warranted to evaluate whether MEL is a viable and safe adjunctive treatment to improve outcome after traumatic brain injury.

Footnotes

Acknowledgments

This work was supported by the Alberta Children's Hospital Foundation Neurotrauma Grant (RT34396). Dr. Barlow was supported in part by a tuition scholarship from the University of Queensland.

Author Disclosure Statement

No competing financial interests exist.

References

Faul

, Xu

, Wald

, and Coronado

V.G.

(2010). Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations and Deaths 2002–2006. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control: Atlanta, pps. 2–70.

Feigin

V.L.

, Theadom

, Barker-Collo

, Starkey

N.J.

, McPherson

, - Kahan

, Dowell

, Brown

, Parag

, Kydd

, Jones

, Ameratunga

, and Ameratunga

(2013). Incidence of traumatic brain injury in New Zealand: a population-based study. Lancet Neurol. 12, 53–64.

Carney

, Totten

A.M.

, O'Reilly

, Ullman

J.S.

, Hawryluk

G.W.

, Bell

M.J.

, Bratton

S.L.

, Chesnut

, Harris

O.A.

, Kissoon

, Rubiano

A.M.

, Shutter

, Tasker

R.C.

, Vavilala

M.S.

, Wilberger

, Wright

D.W.

, and Ghajar

(2017). Guidelines for the Management of Severe Traumatic Brain Injury, Fourth Edition. Neurosurgery, 80, 6–15.

Prins

, Greco

, Alexander

, and Giza

C.C.

(2013). The pathophysiology of traumatic brain injury at a glance. Dis. Model. Mech., 6, 1307–1315.

Stocchetti

, Carbonara

, Citerio

, Ercole

, Skrifvars

M.B.

, Smielewski

, Zoerle

, and Menon

D.K.

(2017). Severe traumatic brain injury: targeted management in the intensive care unit. Lancet Neurol. 16, 452–464.

Diaz-Arrastia

, Kochanek

P.M.

, Bergold

, Kenney

, Marx

C.E.

, Grimes

C.J.B.

, Loh

L.T.C

.Y., Adam

L.T.C

.G.E., Oskvig

, Curley

K.C.

, and Salzer

C.W.

(2014). Pharmacotherapy of traumatic brain injury: state of the science and the road forward: report of the Department of Defense Neurotrauma Pharmacology Workgroup. J. Neurotrauma, 31, 135–158.

Redman

, Armstrong

, and Ng

K.T.

(1983). Free-running activity rhythms in the rat: entrainment by melatonin. Science, 219, 1089.

Redman

J.R.

, and Armstrong

S.M.

(1988). Reentrainment of rat circadian activity rhythms: effects of melatonin. J. Pineal Res., 5, 203–215.

Herrera

E.A.

, Macchiavello

, Montt

, Ebensperger

, Diaz

, Ramirez

, Parer

J.T.

, Seron-Ferre

, Reyes

R.V.

, and Llanos

A.J.

(2014). Melatonin improves cerebrovascular function and decreases oxidative stress in chronically hypoxic lambs. J. Pineal Res., 57, 33–42.

10.

Aly

, Elmahdy

, El-Dib

, Rowisha

, Awny

, El-Gohary

, Elbatch

, Hamisa

, and El-Mashad

A.R.

(2015). Melatonin use for neuroprotection in perinatal asphyxia: a randomized controlled pilot study. J. Perinatol., 35, 186–191.

11.

Chumboatong

, Thummayot

, Govitrapong

, Tocharus

, Jittiwat

, and Tocharus

(2017). Neuroprotection of agomelatine against cerebral ischemia/reperfusion injury through an antiapoptotic pathway in rat. Neurochem. Int., 102, 114–122.

12.

Herrera

, Sainz

R.M.

, Mayo

J.C.

, Martín

, Antolín

, and Rodriguez

(2001). Glutamate induces oxidative stress not mediated by glutamate receptors or cystine transporters: protective effect of melatonin and other antioxidants. J. Pineal Res., 31, 356–362.

13.

Esposito

and Cuzzocrea

(2010). Antiinflammatory activity of melatonin in central nervous system. Curr. Neuropharmacol., 8, 228.

14.

Fernández-Gajardo

, Matamala

J.M.

, Carrasco

, Gutiérrez

, Melo

, and Rodrigo

(2014). Novel therapeutic strategies for traumatic brain injury: acute antioxidant reinforcement. CNS Drugs, 28, 229–248.

15.

Maldonado

M.D.

, Murillo-Cabezas

, Terron

M.P.

, Flores

L.J.

, Tan

D.X.

, Manchester

L.C.

, and Reiter

R.J.

(2007). The potential of melatonin in reducing morbidity-mortality after craniocerebral trauma. J. Pineal Res., 42, 1–11.

16.

Hardeland

, Cardinali

D.P.

, Srinivasan

, Spence

D.W.

, Brown

G.M.

, and Pandi-Perumal

S.R.

(2011). Melatonin—a pleiotropic, orchestrating regulator molecule. Prog. Neurobiol., 93, 350–384.

17.

Luchetti

, Canonico

, Betti

, Arcangeletti

, Pilolli

, Piroddi

, Canesi

, Papa

, and Galli

(2010). Melatonin signaling and cell protection function. FASEB J. 24, 3603–3624.

18.

Reiter

R.J.

, Tan

D.-X.

, and Fuentes-Broto

(2010). Melatonin: a multitasking molecule. Prog. Brain Res., 181, 127–151.

19.

Becker-André

, Wiesenberg

, Schaeren-Wiemers

, André

, Missbach

, Saurat

J.H.

, and Carlberg

(1994). Pineal gland hormone melatonin binds and activates an orphan of the nuclear receptor superfamily. J. Biol. Chem., 269, 28, 531–28,534.

20.

Leon

, Acuña-Castroviejo

, Sainz

R.M.

, Mayo

J.C.

, Tan

D.X.

, and Reiter

R.J.

(2004). Melatonin and mitochondrial function. Life Sci. 75, 765–790.

21.

Tan

D.-X.

, Manchester

L.C.

, Terron

M.P.

, Flores

L.J.

, and Reiter

R.J.

(2007). One molecule, many derivatives: a never-ending interaction of melatonin with reactive oxygen and nitrogen species?. J. Pineal Res., 42, 28–42.

22.

Allegra

, Reiter

R.J.

, Tan

D.-X.

, Gentile

, Tesoriere

, and Livrea

M.A.

(2003). The chemistry of melatonin's interaction with reactive species. J. Pineal Res., 34, 1–10.

23.

Rosen

, Than

N.N.

, Koch

, Poeggeler

, Laatsch

, and Hardeland

(2006). Interactions of melatonin and its metabolites with the ABTS cation radical: extension of the radical scavenger cascade and formation of a novel class of oxidation products, C2-substituted 3-indolinones. J. Pineal Res., 41, 374–381.

24.

Tan

D.X.

, Manchester

L.C.

, Reiter

R.J.

, Qi

W.B.

, Karbownik

, and Calvo

J.R.

(2000). Significance of melatonin in antioxidative defense system: reactions and products. Biol. Signals Recept., 9, 137–159.

25.

Rodriguez

, Mayo

J.C.

, Sainz

R.M.

, Antolín

, Herrera

, Martín

, and Reiter

R.J.

(2004). Regulation of antioxidant enzymes: a significant role for melatonin. J. Pineal Res., 36, 1–9.

26.

Acuña Castroviejo

, López

L.C.

, Escames

, López

, García

J.A.

, and Reiter

R.J.

(2011). Melatonin-mitochondria interplay in health and disease. Curr. Top. Med. Chem., 11, 221–240.

27.

Acuña-Castroviejo

, Escames

, León

, Carazo

, and Khaldy

(2003). Mitochondrial regulation by melatonin and its metabolites. Adv. Exp. Med. Biol., 527, 549–557.

28.

León

, Acuña-Castroviejo

, Escames

, Tan

D.-X.

, and Reiter

R.J.

(2005). Melatonin mitigates mitochondrial malfunction. J. Pineal Res., 38, 1–9.

29.

, Wang

, Liu

, Pan

, Zhang

, Fang

, Jiang

, Yin

, Zou

, Li

, Zhang

, Lin

, and Xiao

(2017). Melatonin reduces hypoxic-ischaemic (HI) induced autophagy and apoptosis: An in vivo and in vitro investigation in experimental models of neonatal HI brain injury. Neurosci. Lett., 653, 105–112.

30.

, Shao

, Zhao

, Chen

, Yu

, Zhou

, Liang

, Shi

, Dixon

B.J.

, Wang

, Ling

, Hong

, and Zhang

(2016). Melatonin attenuates neuronal apoptosis through up-regulation of K(+) -Cl(-) cotransporter KCC2 expression following traumatic brain injury in rats. J. Pineal Res., 61, 241–250.

31.

Das

, McDowell

, Pava

M.J.

, Smith

J.A.

, Reiter

R.J.

, Woodward

J.J.

, Varma

A.K.

, Ray

S.K.

, and Banik

N.L.

(2010). The inhibition of apoptosis by melatonin in VSC4.1 motoneurons exposed to oxidative stress, glutamate excitotoxicity, or TNF-alpha toxicity involves membrane melatonin receptors. J. Pineal Res., 48, 157–169.

32.

Hovda

D.A.

, Lee

S.M.

, Smith

M.L.

, Von Stuck

, Bergsneider

, Kelly

, Shalmon

, Martin

, Caron

, Mazziotta

. and Phelps

, and Becker

D.P.

(1995). The neurochemical and metabolic cascade following brain injury: moving from animal models to man. J. Neurotrauma, 12, 903–906.

33.

Scott

F.F.

, Belle

M.D.C.

, Delagrange

, and Piggins

H.D.

(2010). Electrophysiological effects of melatonin on mouse Per1 and non-Per1 suprachiasmatic nuclei neurones in vitro. J. Neuroendocrinol., 22, 1148–1156.

34.

Paula-Lima

A.C.

, Louzada

P.R.

, De Mello

F.G.

, and Ferreira

S.T.

(2003). Neuroprotection against Abeta and glutamate toxicity by melatonin: are GABA receptors involved?. Neurotox. Res., 5, 323–327.

35.

Matsuta

, Yusup

, Tanase

, Ishida

, Akino

, and Yokoyama

(2010). Melatonin increases bladder capacity via GABAergic system and decreases urine volume in rats. J. Urol., 184, 386–391.

36.

Cheng

X.-P.

, Sun

, Ye

Z.-Y.

, and Zhou

J.-N.

(2012). Melatonin modulates the GABAergic response in cultured rat hippocampal neurons. J. Pharmacol. Sci., 119, 177–185.

37.

Andrews-Zwilling

, Bien-Ly

, Xu

, Li

, Bernardo

, Yoon

S.Y.

, Zwilling

, Yan

T.X.

, Chen

, and Huang

(2010). Apolipoprotein E4 causes age- and Tau-dependent impairment of GABAergic interneurons, leading to learning and memory deficits in mice. J. Neurosci., 30, 13, 707–13,717.

38.

Cherry

J.D.

, Tripodis

, Alvarez

V.E.

, Huber

, Kiernan

P.T.

, Daneshvar

D.H.

, Mez

, Montenigro

P.H.

, Solomon

T.M.

, Alosco

M.L.

, Stern

R.A.

, McKee

A.C.

, and Stein

T.D.

(2016). Microglial neuroinflammation contributes to tau accumulation in chronic traumatic encephalopathy. Acta Neuropathol. Commun., 4, 112.

39.

Villapol

, Fau

, Renolleau

, Biran

, Charriaut-Marlangue

, and Baud

(2011). Melatonin promotes myelination by decreasing white matter inflammation after neonatal stroke. Pediatr. Res., 69, 51–55.

40.

Campolo

, Ahmad

, Crupi

, Impellizzeri

, Morabito

, Esposito

, and Cuzzocrea

(2013). Combination therapy with melatonin and dexamethasone in a mouse model of traumatic brain injury. J. Endocrinol, 217, 291–301.

41.

Tsai

M.C.

, Chen

W.J.

, Tsai

M.S.

, Ching

C.H.

, and Chuang

J.I.

(2011). Melatonin attenuates brain contusion-induced oxidative insult, inactivation of signal transducers and activators of transcription 1, and upregulation of suppressor of cytokine signaling-3 in rats. J. Pineal Res., 51, 233–245.

42.

Bramley

, Henson

, Lewis

M.M.

, Kong

, Stetter

, and Silvis

(2017). Sleep disturbance following concussion is a risk factor for a prolonged recovery. Clin. Pediatr. (Phila) 56, 1280–1285.

43.

Dong

, Sheng

, Tong

, Li

, Xu

, and Hou

(2013). Risk factors associated with sleep disturbance following traumatic brain injury. Presented at the Sleep Medicine Conference, 5th International World Association of Sleep Medicine Congress and the 22nd Annual Congress of the Spanish Sleep Society, Valencia, Spain.

44.

Ponsford

J.L.

, Ziino

, Parcell

D.L.

, Shekleton

J.A.

, Roper

, Redman

J.R.

, Phipps-Nelson

, and Rajaratnam

S.M.

(2012). Fatigue and sleep disturbance following traumatic brain injury—their nature, causes, and potential treatments. J. Head Trauma Rehabil., 27, 224–233.

45.

Barlow

K.M.

(2016). Postconcussion syndrome: a review. J. Child Neurol., 31, 57–67.

46.

Kuczynski

, Crawford

, Bodell

, Dewey

, and Barlow

K.M.

(2013). Characteristics of post-traumatic headaches in children following mild traumatic brain injury and their response to treatment: a prospective cohort. Dev. Med. Child Neurol., 55, 636–641.

47.

Max

J.E.

, Lopez

, Wilde

E.A.

, Bigler

E.D.

, Schachar

R.J.

, Saunders

, Ewing-Cobbs

, Chapman

S.B.

, Yang

T.T.

, and Levin

H.S.

(2015). Anxiety disorders in children and adolescents in the second six months after traumatic brain injury. J. Pediatr. Rehabil. Med., 8, 345–355.

48.

Dodge

N.N.

, and Wilson

G.A.

(2001). Melatonin for treatment of sleep disorders in children with developmental disabilities. J. Child Neurol., 16, 581–584.

49.

Dai

, Cui

, Li

, Chen

, and Wang

(2007). Melatonin attenuates the development of antinociceptive tolerance to delta-, but not to mu-opioid receptor agonist in mice. Behav. Brain Res., 182, 21–27.

50.

Wilhelmsen

, Amirian

, Reiter

R.J.

, Rosenberg

, and Gögenur

(2011). Analgesic effects of melatonin: a review of current evidence from experimental and clinical studies. J. Pineal Res., 51, 270–277.

51.

Peres

M.F.P.

, Zukerman

, da Cunha Tanuri

, Moreira

F.R.

, and Cipolla-Neto

(2004). Melatonin, 3 mg, is effective for migraine prevention. Neurology, 63, 757.

52.

Miano

, Parisi

, Pelliccia

, Luchetti

, Paolino

M.C.

, and Villa

M.P.

(2008). Melatonin to prevent migraine or tension-type headache in children. Neurol. Sci., 29, 285–287.

53.

Citera

, Arias

M.A.

, Maldonado-Cocco

J.A.

, Lázaro

M.A.

, Rosemffet

M.G.

, Brusco

L.I.

, Scheines

E.J.

, and Cardinalli

D.P.

(2000). The effect of melatonin in patients with fibromyalgia: a pilot study. Clin. Rheumatol., 19, 9–13.

54.

Mozaffari

, Rahimi

. and Abdollahi

(2010). Implications of melatonin therapy in irritable bowel syndrome: a systematic review. Curr. Pharm. Des., 16, 3646–3655.

55.

Dominguez-Rodriguez

, Abreu-Gonzalez

, de la Torre-Hernandez

J.M.

, Gonzalez-Gonzalez

, Garcia-Camarero

, Consuegra-Sanchez

, Garcia-Saiz

M.D.

, Aldea-Perona

, Virgos-Aller

, Azpeitia

, and Reiter

R.J.

(2017). Effect of intravenous and intracoronary melatonin as an adjunct to primary percutaneous coronary intervention for acute ST-elevation myocardial infarction: results of the melatonin adjunct in the acute myocardial infarction treated with angioplasty trial. J. Pineal Res. 62.

56.

Gringras

, Nir

, Breddy

, Frydman-Marom

, and Findling

R.L.

(2017). Efficacy and safety of pediatric prolonged-release melatonin for insomnia in children with autism spectrum disorder. J. Am. Acad. Child Adolesc. Psychiatry, 56, 948–957.

57.

Harpsøe

N.G.

, Andersen

L.P.H.

, Gögenur

, and Rosenberg

(2015). Clinical pharmacokinetics of melatonin: a systematic review. Eur. J. Clin. Pharmacol., 71, 901–909.

58.

Zeng

, Zhang

, Kwong

J.S.

, Zhang

, Li

, Sun

, Niu

, and Du

(2015). The methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review. J. Evid. Based Med., 8, 2–10.

59.

Downs

S.H.

, and Black

(1998). The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J. Epidemiol. Community Health, 52, 377–384.

60.

Ates

, Cayli

, Gurses

, Yucel

, Iraz

, Altinoz

, Kocak

, and Yologlu

(2006). Effect of pinealectomy and melatonin replacement on morphological and biochemical recovery after traumatic brain injury. Int. J. Dev. Neurosci., 24, 357–363.

61.

Bayir

, Kiresi

D.A.

, Kara

, Cengiz

S.L.

, Kocak

, Ozdinc

, Ak

, and Bodur

(2008). The effects of mannitol and melatonin on MRI findings in an animal model of traumatic brain edema. Acta Neurol. Belg., 108, 149–154.

62.

Beni

S.M.

, Kohen

, Reiter

R.J.

, Tan

D.X.

, and Shohami

(2004). Melatonin-induced neuroprotection after closed head injury is associated with increased brain antioxidants and attenuated late-phase activation of NF-kappaB and AP-1. FASEB J. 18, 149–151.

63.

Dehghan

, Khaksari Hadad

, Asadikram

, Najafipour

, and Shahrokhi

(2013). Effect of melatonin on intracranial pressure and brain edema following traumatic brain injury: role of oxidative stresses. Arch. Med. Res., 44, 251–258.

64.

Ding

, Wang

, Xu

, Lu

, Zhang

, and Zhu

(2014). Melatonin reduced microglial activation and alleviated neuroinflammation induced neuron degeneration in experimental traumatic brain injury: Possible involvement of mTOR pathway. Neurochem. Int., 76, 23–31.

65.

Ding

, Xu

, Wang

, Zhang

, Wu

, and Li

(2015). Melatonin protects the brain from apoptosis by enhancement of autophagy after traumatic brain injury in mice. Neurochem. Int., 91, 46–54.

66.

Kabadi

S.V.

, and Maher

T.J.

(2010). Posttreatment with uridine and melatonin following traumatic brain injury reduces edema in various brain regions in rats. Ann. N. Y. Acad. Sci., 1199, 105–113.

67.

Kelso

M.L.

, Scheff

N.N.

, Scheff

S.W.

, and Pauly

J.R.

(2011). Melatonin and minocycline for combinatorial therapy to improve functional and histopathological deficits following traumatic brain injury. Neurosci. Lett., 488, 60–64.

68.

Lin

, Chao

, Li

, Xu

, Liu

, Hou

, Liu

. and Ji

(2016). Melatonin attenuates traumatic brain injury-induced inflammation: a possible role for mitophagy. J. Pineal Res., 61, 177–186.

69.

Mesenge

, Margaill

, Verrecchia

, Allix

, Boulu

R.G.

, and Plotkine

(1998). Protective effect of melatonin in a model of traumatic brain injury in mice. J. Pineal Res., 25, 41–46.

70.

Ozdemir

, Tugyan

, Uysal

, Sonmez

, Acikgoz

, Ozdemir

, Duman

, and Ozkan

(2005). Protective effect of melatonin against head trauma-induced hippocampal damage and spatial memory deficits in immature rats. Neurosci. Lett., 385, 234–239.

71.

Sarrafzadeh

A.S.

, Thomale

U.W.

, Kroppenstedt

S.N.

, and Unterberg

A.W.

(2000). Neuroprotective effect of melatonin on cortical impact injury in the rat. Acta Neurochir. 142, 1293–1299.

72.

Yamakawa

G.R.

, Salberg

, Barlow

K.M.

, Brooks

B.L.

, Esser

M.J.

, Yeates

K.O.

, and Mychasiuk

(2017). Manipulating cognitive reserve: pre-injury environmental conditions influence the severity of concussion symptomology, gene expression, and response to melatonin treatment in rats. Exp. Neurol., 295, 55–65.

73.

Kemp

, Biswas

, Neumann

, and Coughlan

(2004). The value of melatonin for sleep disorders occurring post-head injury: a pilot RCT. Brain Inj. 18, 911–919.

74.

Beni

S.M.

, Kohen

, Reiter

R.J.

, Tan

D.X.

, and Shohami

(2004). Melatonin-induced neuroprotection after closed head injury is associated with increased brain antioxidants and attenuated late-phase activation of NF-kappaB and AP-1. FASEB J. 18, 149–151.

75.

Kochanek

P.M

, Carney

, Adelson

P.D.

, Ashwal

, Bell

M.J.

, Bratton

, Carson

, Chesnut

R.M.

, Ghajar

, Goldstein

, Grant

G.A.

, Kissoon

, Peterson

, Selden

N.R.

, Tasker

R.C.

, Tong

K.A.

, Vavilala

M.S.

, Wainwright

M.S.

, Warden

C.R.

, American Academy of Pediatrics-Section on Neurological

Surgery

, American Association of Neurological Surgeons/ Congress of Neurological

Surgeons

, Child Neurology

Society

, European Society of Pediatric and Neonatal Intensive

Care

, Neurocritical Care

Society

, Pediatric Neurocritical Care Research

Group

, Society of Critical Care

Medicine

, Paediatric Intensive CareSociety

, Society for Neuroscience in Anesthesiology and Critical

Care

, and World Federation of Pediatric Intensive and Critical Care Societies (2012). Guidelines for the acute medical management of severe traumatic brain injury in infants, children, and adolescents—second edition. Pediatr. Crit. Care Med., 13, Suppl. 1., S1–82.

76.

Kemp

, Biswas

, Neumann

, and Coughlan

(2004). The value of melatonin for sleep disorders occurring post-head injury: a pilot RCT. Brain Inj. 18, 911–919.

77.

O'Collins

V.E.

, Macleod

M.R.

, Cox

S.F.

, Van Raay

, Aleksoska

, Donnan

G.A

,. and Howells

D.W.

(2011). Preclinical drug evaluation for combination therapy in acute stroke using systematic review, meta-analysis, and subsequent experimental testing. J. Cereb. Blood Flow Metab., 31, 962–975.

78.

Zhu

, Xu

, Duan

, Li

, Zhang

, Huang

, Zhao

, Wang

, Li

, Feng

, Li

, Hu

, and Yin

(2017). Exogenous melatonin in the treatment of pain: a systematic review and meta-analysis. Oncotarget, 8, 100, 582–100,592.