Stem cell‐based interventions for the prevention of morbidity and mortality following hypoxic‐ischaemic encephalopathy in newborn infants

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To determine the efficacy and safety of stem cell‐based interventions for the treatment of hypoxic‐ischaemic encephalopathy (HIE) in newborn infants.

Background

Description of the condition

Hypoxic‐ischaemic encephalopathy (HIE) is a leading cause of mortality and long‐term neurological sequelae, affecting thousands of children in low‐, middle‐, and high‐income countries. Severity of HIE is scored in three stages: Stage I (Sarnat 1) – mild, Stage II (Sarnat 2) – moderate, Stage III (Sarnat 3) – severe (Sarnat 1976). Standard management in neonatal intensive care units consists of providing supportive care to maintain cerebral perfusion and metabolic balance. The only effective intervention is therapeutic hypothermia, which reduces mortality and improves neurocognitive outcome of newborns at 18 month of age (Edwards 2010; Jacobs 2013). The lower brain temperature induced by therapeutic hypothermia may decrease brain baseline metabolism and energy demand (Owji 2017; Pfister 2010; Polderman 2008). The mechanisms include reduction of apoptosis, mitochondrial dysfunction, and free radical production; decreased permeability of blood‐brain barrier; and mitigation of reperfusion injury and neuro‐inflammation (Goss 1995; Gunn 1997; Gunn 1998; Haaland 1997; Iwata 2007; Rothwell 1995; Silverstein 1997; Thoresen 1995). However, treatment should to be started during the latent phase of brain injury, which extends up to six hours from the primary brain insult (Gluckman 1992), to avoid reperfusion injury (Gunn 1997; Zhao 1996). Furthermore, the duration of treatment should be long enough to mitigate a secondary phase of neural injury with cytotoxic cell oedema, apoptosis, accumulation of extracellular excitotoxins, and delayed seizures due to reperfusion (Sirimanne 1996; Tan 1996; Williams 1991). Thus, hypothermia should cover the whole reperfusion phase of brain injury and last 72 hours (Wyllie 2015). Additional therapies for HIE are under investigation and include erythropoietin, allopurinol, xenon, topiramate, and magnesium sulphate. These strategies will be described briefly. Erythropoietin has anti‐inflammatory, antioxidative, antiapoptotic, and antiexcitotoxic properties (Villa 2003). In addition, it may promote neurogenesis and angiogenesis (Wang 2004). In animal models of HIE, erythropoietin attenuates brain damage, and improves learning memory (Gonzalez 2009; Kumral 2004; Wu 2012). These data seem to be confirmed in small clinical trials (Elmahdy 2010; Rogers 2014; Zhu 2009). Allopurinol may be neuroprotective through direct scavenging of hydroxyl radicals and neutralizing non‐protein ‐bound iron (Peeters‐Scholte 2003). Animal studies showed that allopurinol might be neuroprotective (Palmer 1993). Furthermore, its administration to term infants affected by HIE might improve neurocognitive outcome at four to eight years of age (Kaandorp 2012). Xenon has been reported to decrease brain injury in animal studies (David 2008; Ma 2005), with additional efficacy when combined with hypothermia (Hobbs 2008; Thoresen 2009). In one phase II clinical trial, xenon seemed to be safe (Dingley 2014). Topiramate may reduce brain damage, decrease the rate of apoptosis, diminish the infarcted area, and improved neurological outcome in animal models of HIE (Noh 2006; Ozyener 2012; Sfaello 2005). One ongoing trial aims to investigate the effects of topiramate on full‐term neonates with HIE undergoing hypothermia (NCT01765218). There are controversies regarding the neuroprotective role of magnesium sulphate in brain damage in term and late preterm infants (Galinsky 2014; Tagin 2013). The data from animal studies show lack of beneficial neuroprotective effect following HIE (de Haan 1997; Greenwood 2000; Penrice 1997). There is an ongoing clinical trial on magnesium sulphate during therapeutic hypothermia in term neonates with HIE (NCT01646619).

Description of the intervention

Mesenchymal stem/stromal cells (MSCs) have emerged as exciting and new therapeutic agents that could potentially ameliorate HIE (Nabetani 2018). MSCs are the most commonly used regenerative cells in clinical trials due to their relatively safe profile, ease of isolation/propagation, ability to reduce inflammation and oxidative stress, restore energy failure, and decrease cell apoptosis (Trounson 2015). The International Society for Cellular Therapy defines cells as MSCs by the following criteria: adhere to plastic in standard culture conditions, express specific surface antigen markers, and have the capacity for multipotent differentiation (Dominici 2006). Although MSCs are ubiquitously used in regenerative studies, additional stem cell‐based therapies (collectively referred to as 'regenerative cells') are also under consideration. For instance, mononuclear cells, oligodendrocyte progenitor cells, neural stem cells, hematopoietic stem cells, endothelial cells, and inducible pluripotent stem cells have demonstrated efficacy in animal models of brain injury (Pimentel‐Coelho 2012). Furthermore, one ongoing clinical trial is evaluating the efficacy of autologous umbilical cord blood (which contains a combination of MSCs, hematopoietic stem cells, and other mononuclear cells) as a treatment for HIE (NCT02612155). While inducible pluripotent stem cells are typically obtained from skin cells, they can also be retrieved from blood or MSCs to become a specialized cell, such as a neural stem cell or oligodendrocyte (Cai 2010).

Several factors must be taken into account when optimizing regenerative cells. For example, the tissue source, laboratory processing, passage number, dose, frequency, timing, and route of administration are all variables that may affect efficacy (Möbius 2015). The perinatal period offers an opportune time to collect umbilical cord tissue/blood, amniotic fluid or placental tissue (Garcia 2014; Parolini 2014; Sanberg 2014; Taghizadeh 2014). These sources, once considered medical waste, offer a vast supply of regenerative cells that are obtained non‐invasively and recognized for their high proliferative rates/differentiation potential, paracrine release of biological factors and their low likelihood of mounting an immune response after transplantation (Batsali 2013; Möbius 2015; Moreira 2017; Parolini 2008). The therapeutic potential of regenerative cells may also be influenced by different laboratory cell processing techniques (e.g. oxygen tension, cell expansion media, passage number, fresh versus cryopreserved cells) (Frey 2006; Kaplan 2017; Parody 2013). Fewer passages is preferred, as multiple passages may impair cell function (Bellayr 2014; Wagner 2008). Interestingly, larger amounts of immature hematopoietic progenitors and MSCs with high proliferative potential have been detected in the cord blood of preterm as compared with term infants (Podesta 2015). In one systematic review of animal models of HIE, neuroprotective dosages of MSCs ranged between 200,000 to 3,500,000 cells, with most of the studies administering a one‐time intracerebral/intranasal dose of cells 72 hours or less after injury (Archambault 2017).

Though regenerative cells are characterized by low immunogenicity (Gebler 2012), autologous transplantation is likely to be associated with lower risks for infections and immune rejection. However, allogeneic transplantation may offer significant practical advantages such as rapid availability of disease‐free products that are generated in a cost‐effective manner (Hare 2017).

How the intervention might work

It is currently unknown which regenerative cell(s) promotes the therapeutic effects in HIE; however, we will present a brief overview of the potential role(s) each cell has after neurological injury (see also Table 1).

Table 1.

Types of regenerative cells

Cell type	Source	Rationale	Mechanism of action	Preclinical/clinical results	References
MSC	Human umbilical cord tissue/blood; rodent/human bone marrow	Safe and feasible in phase I RCT for bronchopulmonary dysplasia Low immunogenicity (low MHC II), easily obtainable, rapid expansion Autologous/allogeneic administration Paracrine release of trophic factors	Paracrine release of IGF‐1, EGF, VEGF, BDNF Immunomodulatory: regulate T‐cell, B‐cell function, and production of inflammatory cytokines Mitochondrial transfer	Nerve fibre remyelination and axonal regeneration Improve behavioural/motor tests Enhance neural cell proliferation, survival, function Decrease infarct size	Ahn 2016; Boshuizen 2018; Chopp 2002; Hsu 2016; Islam 2012; Liu 2010; Murphy 2013; Park 2016
MNC	Human umbilical cord blood	Readily collected and large supply in cord blood with high plasticity Safe and feasible in phase I RCT for hypoxic‐ischaemic encephalopathy Low immunogenicity (minimal HLA matching) Paracrine release of trophic factors Autologous/allogeneic administration	Increase expression of BDNF, NGF, VEGF, GDNF Activation of pro‐survival Akt pathway Decrease TNF‐α and increase IL‐10 gene expression Reduce CD4+ T cell infiltration Regulate hedgehog signalling	Decrease neuronal apoptosis, astrogliosis, inflammation Improve oligodendrocyte survival Induce axonal growth Improve neurobehavioral outcome	Aridas 2016; Cotten 2014; Fan 2005; McDonald 2018; Pimentel‐Coelho 2012; Rowe 2010; Wang 2013
OPC	Rodent/human embryonic stem cell; human NSC derivation	Differentiate into oligodendrocytes (cells highly susceptible to hypoxic‐ischaemic injury) Remyelinate injured axons	Diffuse biodistribution along white matter tracts and differentiate into myelin sheath‐producing oligodendrocytes	Promote myelin sheath formation, NSC proliferation, and inhibit apoptosis Motor recovery following CNS injury	Chen 2015; Gopagondanahalli 2016; Kim 2018; Manley 2017; Niimi 2018; Xu 2015
NSC	Human fetal striatum; human ESC; human iPSC	Differentiate into cells necessary for brain repair, including: neurons, astrocytes, and oligodendrocytes Paracrine release of trophic factors Low immunogenicity and tumorigenicity	Immunomodulation Paracrine secretion of BDNF, VEGF, and EGF Attenuate NF‐κB signalling Upregulate glutamate transport	Stimulate survival and migration of endogenous NSCs and neurons Reduce inflammation and reactive oxygen species production Improve axonal growth, motor function, decreased infarct size	Daadi 2016; Huang 2018; Ji 2015; Mine 2013
HSC	Umbilical cord blood	Paracrine release of neurotrophic factors Multipotent capacity and ability to transdifferentiate into neuronal cells Autologous/allogeneic administration	Reduce microglial cells and T lymphocytes Secrete VEGF, HGF, IGF‐1	Decrease infarct size and maintains cerebral blood flow Enhance axonal growth Ameliorate neuronal apoptosis and postischaemic inflammation	Schwarting 2008; Tsuji 2014; Verina 2013
EPC	Human umbilical vein; human umbilical cord blood; human adipose stem cell; human iPSC	Umbilical cord‐derived EPCs have higher regenerative potential than adult bone marrow‐derived EPCs Endothelial cell protection, repair, angiogenesis Low immunogenicity Paracrine release of regenerative factors Autologous administration	Anti‐inflammatory effects: reduce CD4+ infiltration to the brain Activation of PI3/Akt pathway Axonal growth: BDNF secretion Angiogenesis: VEGF, IGF‐1 secretion	Improve cognitive and motor function Inhibit neuronal apoptosis Stimulate blood vessel formation and reduce infarct size	Grandvuillemin 2017; Kidani 2016; McDonald 2018; Nabetani 2018; Wang 2016; Wu 2013
iPSC	Skin fibroblasts, umbilical cord tissue, amniotic tissue	Autologous administration Differentiation into multiple neural lineage cells Low immunogenicity	Differentiate into functional neural cells (electrophysiological properties) Decrease infiltration of MPO+ neutrophils and CD11b+ microglia VEGF expression and organelle transfer	Improve survival and sensorimotor function Establish axonal connections Inhibit inflammation, neural apoptosis, and glial scar formation	Cai 2010; Hsu 2016; Oki 2012; Pluchino 2013; Qin 2015; Tornero 2013

BDNF: brain‐derived neurotrophic factor; CNS: central nervous system; ESC: embryonic stem cell; EGF: epidermal growth factor; GDNF: glial cell‐line derived neurotrophic factor; HGF: hepatocyte growth factor; HLA: human leukocyte antigen; HSC: hematopoietic stem cells; IGF‐1: insulin‐like growth factor 1; iPSC: inducible pluripotent stem cells; IL: interleukin; MHC: major histocompatibility complex; MSC: mesenchymal stem cell; MNC: mononuclear cells; NF‐κB: nuclear factor kappa beta; NGF: nerve growth factor; NSC: neural stem cells; OPC: oligodendrocyte progenitor cells; RCT: randomized controlled trial; TNF: tumour necrosis factor; VEGF: vascular endothelial growth factor.

Mesenchymal stem cells: the beneficial effects of MSCs occurs at the cellular and functional level. The proposed mechanism of action is through the paracrine release of factors known to improve neurogenesis, such as basic fibroblast growth factor, insulin‐like growth factor‐1, and anti‐inflammatory cytokines (Li 2002; Murphy 2013; Qu 2007; van Velthoven 2009; van Velthoven 2011; van Velthoven 2012). MSCs can also modulate the local immune response by regulating the function of immune cells, such as T‐cells and B‐cells, macrophages, and dendritic cells (Iyer 2008). Furthermore, in injured brain tissue, MSCs upregulate the expression of genes associated with cell proliferation (e.g. Spp1 and interleukin (IL) 17 (IL‐17)), neurogenesis (neural cell adhesion molecule and nerve growth factor), migration (CXCR4), neuronal survival (glial‐derived neurotrophic factor), and downregulation of genes involved in inflammation (i.e. IL‐1β) (van Velthoven 2011). In addition, MSCs may inhibit apoptosis by transporting mitochondria through tunnelling nanotubules in efforts to rescue aerobic respiration (Liu 2014).

At the functional level, MSC treatment, given within the first 10 days after injury, prevents neuroinflammation and apoptosis (Bonestroo 2015; Donega 2013). The administration of umbilical cord‐derived MSCs resulted in nerve fibre remyelination and axonal regeneration, diminished loss of white and grey matter and improved sensorimotor function, and better long‐term neurological recovery (Donega 2013; Donega 2015; Liu 2010; Morán 2017).

Mononuclear cells: in a newborn lamb model of HIE, treatment with umbilical cord blood mononuclear cells showed a decrease in neuronal apoptosis and inflammation, along with a trend towards a reduction in seizures (Aridas 2016). Mononuclear cells also reduce motor deficits and cortical brain loss by decreasing CD4+ T‐cell and microglial infiltration to the injury site (McDonald 2018).

Oligodendrocyte progenitor cells: administration of oligodendrocyte progenitor cells to a small animal model of cervical spinal injury increased myelination and enhanced motor performance (Manley 2017). Differentiation into myelin‐producing cells and expression of brain‐derived neurotrophic factor and bcl‐2 appear to be mechanisms by which oligodendrocyte progenitor cells encourage learning and memory in neonatal asphyxia (Chen 2015).

Neural stem cells: Mine and colleagues showed that intrastriatal transplantation of neural stem cells slows inflammation (microglia/macrophage activation) and promotes axonal connections (Mine 2013). Moreover, it is posited that neural stem cell therapy inhibits IL‐1β expression and improves neural plasticity by upregulating nuclear factor kappa β signalling (Ji 2015).

Hematopoietic stem cells: human umbilical cord 34+ cells given 48‐hours after middle cerebral artery occlusion in mice transiently increased cerebral blood flow and blood vessel diameter in the peri‐infarct area (Tsuji 2014). These findings can be attributed to the release of growth factors (i.e. vascular endothelial growth factor and glial‐derived neurotrophic factor) known to stimulate neurogenesis and angiogenesis (Verina 2013).

Endothelial cells: the neuroprotective effects of endothelial cells are ascribed to their ability in decreasing neuroinflammation and cell apoptosis (Grandvuillemin 2017). Intraperitoneal injection of human umbilical vein endothelial cells preserved microvessels and lessened apoptosis in the cortex of treated animals, via regulation of stromal cell‐derived factor 1 and CXC chemokine receptor 4 (Wu 2013).

Inducible pluripotent stem cells: using human skin‐derived inducible pluripotent stem cells, Tornero and colleagues produced cortical progenitor cells that survived and differentiated into functional neurons that improved performance of impaired limb movement in a rat stroke model (Tornero 2013). Furthermore, inducible pluripotent stem cell treatment reduced the number of inflammatory cells and glial scar formation in a preclinical study of hemorrhagic stroke (Qin 2015).

Taken together, regenerative cells may exert their therapeutic benefit through multiple routes to establish a favourable environment for tissue regeneration, which ultimately leads to better functional outcomes following hypoxic‐ischaemic damage.

Why it is important to do this review

It is important to conduct this review as the use of regenerative cells in neonatology continues to grow. To date, US clinical trials studying the safety, feasibility, efficacy (or a combination of these) of regenerative cells have been registered for bronchopulmonary dysplasia, HIE, hypoplastic left heart syndrome, and intraventricular haemorrhage. The most widespread tissue used to derive regenerative cells is the umbilical cord (Mitsialis 2016; Yoon 2016). While MSCs dominate as cell type, a few trials are evaluating neural progenitor cells, mononuclear cells, and placenta/cord blood cells (NCT02434965; NCT02854579; NCT02999373).

The efficacy and safety of use of regenerative cell administration has been assessed in several systematic reviews and meta‐analyses. For instance, the safety of MSCs has been evaluated in a meta‐analysis across different disciplines (eight studies including 321 adults): there was no association between acute infusional toxicity, organ system complications, infection, death or malignancy (Lalu 2012). However, the risk of potential tumourigenicity related to MSC‐based interventions needs to be further elucidated (Barkholt 2013). The Cochrane review "Stem cell transplantation for ischaemic stroke" included three small trials in adults (Boncoraglio 2010). In newborns, one Cochrane Review has been conducted on MSC for the prevention and treatment of bronchopulmonary dysplasia in preterm infants (Pierro 2017). Early phase trials have been conducted (or are underway) on the use of MSCs or cord blood cells (or both) for bronchopulmonary dysplasia, severe IVH (NCT02274428), and HIE (Chang 2014; Cotten 2014).

Regenerative cells might be an effective intervention for the prevention of morbidity and mortality following HIE, which is one of the most severe morbidities in newborn infants.

Objectives

To determine the efficacy and safety of stem cell‐based interventions for the treatment of hypoxic‐ischaemic encephalopathy (HIE) in newborn infants.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomized controlled trials (RCTs), quasi‐RCTs, and cluster trials.

Types of participants

• Term infants (37 weeks or greater) and late preterm infants (34+0 to 36+6 weeks' gestation) 10 days of age or less.

• Evidence of peripartum asphyxia, characterized by evidence of neonatal or foetal distress with each enrolled infant satisfying at least one of the following criteria:

  • cord gas or postnatal blood gas (within the first hour of life) with pH 7.0 or less or base deficit 12 mEq/L or greater;

  • Apgar score 5 or less at five minutes;

  • need for mechanical ventilation or resuscitation at 10 minutes of life;

  • with or without evidence of encephalopathy (moderate or severe) according to Sarnat staging (Sarnat 1976):

    • Stage 1 (mild): hyperalertness, hyper‐reflexia, dilated pupils, tachycardia, absence of seizures;

    • Stage 2 (moderate): lethargy, hyper‐reflexia, miosis, bradycardia, seizures, hypotonia with weak suck and Moro;

    • Stage 3 (severe): stupor, flaccidity, small‐to‐mid position pupils that react poorly to light, decreased stretch reflexes, hypothermia, and absent Moro.

• No major congenital abnormalities recognizable at birth.

Types of interventions

Comparison 1: stem cell‐based interventions (any type) compared to control (placebo or no treatment).

Comparison 2: use of MSCs of type (e.g. number of doses or passages) or source (e.g. autologous versus allogeneic, or bone marrow versus cord) versus MSCs of other type or source.

Comparison 3: use of stem cell‐based interventions other than MSCs of type (e.g. mononuclear cells, oligodendrocyte progenitor cells, neural stem cells, hematopoietic stem cells, and inducible pluripotent stem cells) or source (e.g. autologous versus allogeneic, or bone marrow versus cord) versus stem cell‐based interventions other than MSCs of other type or source.

Comparison 4: MSCs versus stem cell‐based interventions other than MSCs.

We will include all types of transplantation regardless of cell source (bone marrow, cord blood versus Wharton's jelly, placenta, adipose tissue, peripheral blood), type of graft (autologous or allogeneic), and dose. We will exclude stem cell‐derived cerebral organoids (Di Lullo 2017).

Characteristic of the interventions and cointerventions (e.g. cooling) are specified in Subgroup analysis and investigation of heterogeneity.

Types of outcome measures

Primary outcomes

• All‐cause neonatal mortality (mortality less than 28 days of age).

• Major neurodevelopmental disability: cerebral palsy, developmental delay (Bayley Mental Developmental Index (Bayley 1993; Bayley 2006) or Griffiths Mental Development Scale (Griffiths 1954) assessment greater than two standard deviations (SD) below the mean), intellectual impairment (intelligence quotient (IQ) greater than two SDs below the mean), blindness (vision less than 6/60 in both eyes), or sensorineural deafness requiring amplification (Jacobs 2013). We will separately assess data on children aged 18 to 24 months and aged three to five years.

• Death or major neurodevelopmental disability assessed at 18 to 24 months of age (defined as cerebral palsy, developmental delay (Bayley or Griffith assessment more than two SD below the mean) or intellectual impairment (IQ more than two SD below mean), blindness (vision less than 6/60 in both eyes), or sensorineural deafness requiring amplification) (Bayley 1993; Bayley 2006; Griffiths 1954; Jacobs 2013).

Secondary outcomes

• All‐cause mortality prior to first hospital discharge.

• Each component of major neurodevelopmental disability (these components of long‐term outcome will be reported for all studies that have evaluated children after 18 months' chronological age; separate analyses will be performed for children aged 18 to 24 months and three to five years):

  • cerebral palsy;

  • developmental delay:

    • Bayley or Griffith assessment more than two SD below the mean;

    • neuromotor development (Bayley Scales of Infant Development – Psychomotor Development Index (BSID PDI)) assessed in survivors;

    • cognitive development (Bayley Scales of Infant Development – Mental Development Index (BSID MDI)) assessed in survivors;

  • intellectual impairment (IQ more than two SD below mean);

  • blindness (vision less than 6/60 in both eyes);

  • sensorineural deafness requiring amplification.

• Seizures (suspected clinically or identified by electroencephalogram (EEG) or amplitude‐integrated electroencephalogram (aEEG)):

  • seizures during neonatal period (after MSC administration);

  • seizures or need for anticonvulsants at follow‐up 12 to 14 months of age.

• Cystic periventricular leukomalacia on brain ultrasound in the first month of life.

• Brain magnetic resonance imaging (MRI) abnormalities (yes/no), defined as white matter lesions (i.e. cavitations; Rutherford 2010) and punctate lesions (Cornette 2002); germinal matrix‐intraventricular haemorrhage (GM‐IVH) (Parodi 2015); or cerebellar haemorrhage (Limperopoulos 2007).

• Duration of hospital stay (days).

• Tumour formation, any type, any location, detected by MRI or computed tomography (to assess the risk of tumourigenicity of donor MSCs).

• Immune‐rejection or any serious adverse event (certain, probable or possible according to the World Health Organization probability scale). We will consider post‐hoc analyses for any unexpected adverse effects reported by the studies.

Search methods for identification of studies

We will use the criteria and standard methods of Cochrane and Cochrane Neonatal (see the Cochrane Neonatal search strategy for specialized register).We will search for errata or retractions from included studies published in full‐text on PubMed (www.ncbi.nlm.nih.gov/pubmed) and report the date this was done within the review.

Electronic searches

We will conduct a comprehensive search including: Cochrane Central Register of Controlled Trials (CENTRAL, current issue) in the Cochrane Library; MEDLINE via PubMed (1996 to current); Embase (1980 to current); and CINAHL (1982 to current) using search terms detailed in Appendix 1 for the full search strategies for each database). We will apply no language restrictions. We will search clinical trials registries for ongoing or recently completed trials (clinicaltrials.gov; the World Health Organization’s International Trials Registry and Platform, and the ISRCTN Registry).

Searching other resources

We will review the reference lists of all identified articles for relevant articles not identified in the primary search.

Data collection and analysis

We will use the standard methods of the Cochrane Neonatal Review Group, as described below.

Selection of studies

Two review authors (MB, OR) will independently search for and identify eligible trials that meet the inclusion criteria. We will screen the titles and abstracts to identify potentially relevant citations, and will retrieve the full texts of all potentially relevant articles; we will independently assess the eligibility of studies by filling out eligibility forms designed in accordance with the specified inclusion criteria. We will exclude studies published only in abstract form unless the final results of the trial are reported and all necessary information can be ascertained from the abstract or authors, or both. We will review studies for relevance by assessing study design, types of participants, interventions provided, and outcome measures reported. We will resolve disagreements by discussion and, if necessary, by consultation with a third review author (DL). We will provide details of excluded studies in the 'Characteristics of excluded studies' table, along with reasons for exclusion. We will contact trial authors if details of primary trials are unclear.

Data extraction and management

Two review authors (MB, OR) will independently extract data using a data extraction form integrated with a modified version of the Cochrane Effective Practice and Organisation of Care Group data collection checklist (Cochrane EPOC Group 2017).

We will extract the following characteristics from each included study.

• Administrative details: study author(s); published or unpublished; year of publication; year in which study was conducted; presence of vested interest; details of other relevant papers cited.

• Details of the study: study design; type, duration, and completeness of follow‐up (e.g. greater than 80%); country and location of study; informed consent; ethics approval.

• Details of participants: sex, birth weight, gestational age, number of participants.

• Details of interventions: initiation, dose, and duration of MSCs administration; cointervention such as cooling.

• Details of outcomes as mentioned above under Types of outcome measures.

We will resolve disagreements by discussion. We will describe ongoing studies identified by our search, when available, detailing the primary author, research question(s), methods, and outcome measures, together with an estimate of the reporting date.

Should any queries arise, or in cases for which additional data are required, we will contact study investigators/authors for clarification. Two review authors (MB, OR) will use Cochrane software for data entry (Review Manager 2014). We will replace any standard error of the mean (SEM) by the corresponding SD.

Assessment of risk of bias in included studies

Two review authors (OR, MB) will independently assess the risk of bias (low, high, or unclear) of all included trials using the Cochrane ‘Risk of bias’ tool for the following domains (Higgins 2017).

• Sequence generation (selection bias).

• Allocation concealment (selection bias).

• Blinding of participants and personnel (performance bias).

• Blinding of outcome assessment (detection bias).

• Incomplete outcome data (attrition bias).

• Selective reporting (reporting bias).

• Any other bias.

We will resolve any disagreements by discussion or by consultation with a third review author (AM). See Appendix 2 for a more detailed description of risk of bias for each domain.

Measures of treatment effect

We will use risk ratios (RRs), risk differences (RDs), numbers needed to treat for an additional beneficial outcome (NNTB) or numbers needed to treat for an additional harmful outcome (NNTH) for categorical variables, and mean differences (MDs) for continuous variables. We will replace any within‐group SEM reported in a trial by its corresponding SD using the formula SD = SEM × √n, where n is the number of participants. We will report 95% confidence intervals (CIs) for each statistic.

Unit of analysis issues

We will include all RCTs and quasi‐RCTs in which the unit of allocation was the individual infant. If we find any cluster‐RCTs, we will adjust analysis for the designed effect using the method stated in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017).

Dealing with missing data

We will obtain a dropout rate for each study. If we find a significant dropout rate (e.g. greater than 20%), we will contact study author(s) to request additional data. We will perform a sensitivity analysis to evaluate the overall results with and without inclusion of studies with a significant dropout rate. If a study reports outcomes only for participants completing the trial or only for participants who followed the protocol, we will contact study author(s) to ask them to provide additional information to facilitate an intention‐to‐treat analysis; in instances when this is not possible, we will perform a complete‐case analysis.

Assessment of heterogeneity

We will assess clinical heterogeneity by comparing the distribution of important participant factors between trials and trial factors (randomization concealment, blinding of outcome assessment, loss to follow‐up, treatment type, cointerventions). We will assess statistical heterogeneity by examining the I² statistic (Higgins 2017), a quantity that describes the proportion of variation in point estimates that is due to variability across studies rather than to sampling error.

We will interpret the I² statistic as follows, as described by Higgins 2003.

• Less than 25%: no (none) heterogeneity.

• 25% to 49%: low heterogeneity.

• 50% to 74%: moderate heterogeneity.

• 75% or greater: high heterogeneity.

We will consider statistical heterogeneity to be substantial when the I² statistic is 50% or greater. In addition, we will employ the Chi² test of homogeneity to determine the strength of evidence that heterogeneity is genuine. We will explore clinical variation across studies by comparing the distribution of important participant factors among trials and trial factors (randomization concealment, blinding of outcome assessment, loss to follow‐up, treatment types, and cointerventions). We will consider a threshold of P value less than 0.1 as an indicator of whether heterogeneity (genuine variation in effect sizes) is present.

Assessment of reporting biases

We will examine the possibility of within‐study selective outcome reporting for each study included in the review. We will search for trial protocols of included trials on electronic sources such as PubMed, ClinicalTrials.gov, and the WHO ICTRP to assess whether outcome reporting seems to be sufficiently complete and transparent. We will investigate publication by using funnel plots if we include 10 or more clinical trials in the systematic review (Egger 1997; Higgins 2017).

Data synthesis

We will perform statistical analyses according to the recommendations of Cochrane Neonatal (neonatal.cochrane.org/en/index.html) using Review Manager 5 (Review Manager 2014). We will analyse all infants randomized on an intention‐to‐treat basis. We will analyse treatment effects in the individual trials. We will use a fixed‐effect model to combine the data. For any meta‐analyses, we will synthesize data using RR, RD, NNTB, NNTH, MD, and 95% CI. We will analyse and interpret individual trials separately when we judge meta‐analysis to be inappropriate.

Quality of evidence

We will use the GRADE approach to assess the quality of evidence for the following (clinically relevant) outcomes (Schünemann 2013): 1) All‐cause neonatal mortality (mortality less than 28 days of age); 2) Major neurodevelopmental disability: cerebral palsy, developmental delay (Bayley Mental Developmental Index or Griffiths Mental Development Scale assessment greater than two standard deviations (SD) below the mean), intellectual impairment (intelligence quotient (IQ) greater than two SDs below the mean), blindness (vision less than 6/60 in both eyes), or sensorineural deafness requiring amplification; 3) Death or major neurodevelopmental disability assessed at 18 to 24 months of age (defined as cerebral palsy, developmental delay (Bayley or Griffith assessment more than two SD below the mean) or intellectual impairment (IQ more than two SD below mean), blindness (vision less than 6/60 in both eyes), or sensorineural deafness requiring amplification; 4) All‐cause mortality prior to first hospital discharge; 5) cognitive development (Bayley Scales of Infant Development – Mental Development Index (BSID MDI)) assessed in survivors; 6) seizures or need for anticonvulsants at follow‐up 12 to 14 months of age; 7) Immune‐rejection or any serious adverse event (certain, probable or possible according to the World Health Organization probability scale).

Two review authors will independently assess the quality of the evidence for each of the seven outcomes above. We will consider evidence from RCTs as high quality but downgrade the evidence one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates, and presence of publication bias. We will use the GRADEpro GDT Guideline Development Tool to create a 'Summary of findings' table to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence in one of four grades.

• High: we are very confident that the true effect lies close to that of the estimate of the effect.

• Moderate: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

• Low: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.

• Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Subgroup analysis and investigation of heterogeneity

We will perform the following subgroup analyses.

For MSCs trials:

• gestational age: term infants (37 weeks or greater), late preterm infants (34 to 36+6 weeks' gestation);

• HIE severity stage: mild, moderate, and severe (Sarnat);

• chronological age: less than three days, three days or greater;

• cointervention: with/without cooling;

• MSCs source: bone marrow, cord blood versus Wharton's jelly, placenta, adipose tissue, peripheral blood;

• type of graft: autologous or allogeneic;

• preconditioned (yes, no);

• fresh or frozen and thawed;

• MSCs dose: less than 2 × 10⁷/kg; 2 × 10⁷/kg or greater;

• number of doses: multiple or single administration;

• passage number (i.e. removing cells from a culture flask and plating them into more culture flasks, see Description of the intervention): less than three; three to six; greater than six.

For other cell‐based interventions:

• gestational age: term infants (37 weeks or greater), late preterm infants (34 to 36+6 weeks' gestation);

• HIE severity stage: mild, moderate and severe (Sarnat);

• chronological age: less than three days, three days or greater;

• cointervention: with/without cooling;

• cell source: bone marrow, cord blood, peripheral blood, placenta;

• type of graft: autologous or allogeneic;

• fresh or frozen and thawed;

• number of doses: multiple or single administration.

Sensitivity analysis

We will conduct sensitivity analyses to explore the effect of the methodological quality of trials, checking to ascertain whether studies with a high risk of bias will overestimate the effect of treatment. Differences in study design of included trials might affect the results of the systematic review. We will perform a sensitivity analysis to compare the effects of MSCs in truly randomized trials as opposed to quasi‐randomized trials.

Appendices

Appendix 1. Search strategy

PubMed:

("Stem Cells"[Mesh] OR "Stem Cell Transplantation"[Mesh] OR "Stromal Cells"[Mesh] OR “stem cell”[tiab] OR “stem cells”[tiab] OR “mesenchymal cell” OR “mesenchymal cells” OR “mononuclear cell” OR “mononuclear cells” OR “progenitor cell” OR “progenitor cells” OR “cord blood cell” OR “cord blood cells” OR “regenerative cell” OR “regenerative cells” OR "stromal cell" OR "stromal cells") AND (brain injury OR neuro‐protect* OR neuro‐restorative OR neuroprotect* OR Asphyxia Neonatorum[MeSH] OR Hypoxia‐Ischemia, Brain[MeSH] OR Asphyxia* OR Hypoxia OR Hypoxic OR Hypoxemia OR Hypoxaemia OR Ischemia OR Ischaemia OR Ischemic OR Ischaemic OR anoxia)

The PubMed search will be translated for the other databases detailed in the search methods, and combined with the following Cochrane Neonatal standard search strategies:

PubMed: ((infant, newborn[MeSH] OR newborn OR neonate OR neonatal OR premature OR low birth weight OR VLBW OR LBW or infan* or neonat*) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab] OR placebo [tiab] OR drug therapy [sh] OR randomly [tiab] OR trial [tiab] OR groups [tiab]) NOT (animals [mh] NOT humans [mh]))

Embase: ((exp infant) OR (infan* OR newborn or neonat* OR premature or very low birth weight or low birth weight or VLBW or LBW).mp AND (human not animal) AND (randomized controlled trial or controlled clinical trial or randomized or placebo or clinical trials as topic or randomly or trial or clinical trial).mp

CINAHL: (infan* OR newborn OR neonat* OR premature OR low birth weight OR VLBW OR LBW) AND (randomized controlled trial OR controlled clinical trial OR randomized OR placebo OR clinical trials as topic OR randomly OR trial OR PT clinical trial)

Cochrane Library: (infan* or newborn or neonat* or premature or preterm or very low birth weight or low birth weight or VLBW or LBW)

Appendix 2. Risk of bias tool

We will use the standard methods of Cochrane and Cochrane Neonatal to assess the methodological quality of the trials. For each trial, we will seek information regarding the method of randomization, blinding, and reporting of all outcomes of all the infants enrolled in the trial. We will assess each criterion as being at a low, high, or unclear risk of bias. Two review authors will separately assess each study. We will resolve any disagreements by discussion. We will add this information to the 'Characteristics of included studies' table. We will evaluate the following issues and enter the findings into the 'Risk of bias' table.

1. Sequence generation (checking for possible selection bias). Was the allocation sequence adequately generated?

For each included study, we will categorize the method used to generate the allocation sequence as:

• low risk (any truly random process, e.g. random number table; computer random number generator);

• high risk (any non‐random process, e.g. odd or even date of birth; hospital or clinic record number); or

• unclear risk.

2. Allocation concealment (checking for possible selection bias). Was allocation adequately concealed?

For each included study, we will categorize the method used to conceal the allocation sequence as:

• low risk (e.g. telephone or central randomization; consecutively numbered sealed opaque envelopes);

• high risk (open random allocation; unsealed or non‐opaque envelopes, alternation; date of birth); or

• unclear risk.

3. Blinding of participants and personnel (checking for possible performance bias). Was knowledge of the allocated intervention adequately prevented during the study?

For each included study, we will categorize the methods used to blind study participants and personnel from knowledge of which intervention a participant received. Blinding will be assessed separately for different outcomes or class of outcomes. We will categorize the methods as:

• low risk, high risk, or unclear risk for participants; and

• low risk, high risk, or unclear risk for personnel.

4. Blinding of outcome assessment (checking for possible detection bias). Was knowledge of the allocated intervention adequately prevented at the time of outcome assessment?

For each included study, we will categorize the methods used to blind outcome assessment. Blinding will be assessed separately for different outcomes or class of outcomes. We will categorize the methods as:

• low risk for outcome assessors;

• high risk for outcome assessors; or

• unclear risk for outcome assessors.

5. Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations). Were incomplete outcome data adequately addressed?

For each included study and for each outcome, we will describe the completeness of data including attrition and exclusions from the analysis. We will note whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomized participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information is reported or supplied by the trial authors, we will reinclude missing data in the analyses. We will categorize the methods used to deal with missing data as:

• low risk (less than 20% missing data);

• high risk (20% or greater missing data); or

• unclear risk.

6. Selective reporting bias. Are reports of the study free of suggestion of selective outcome reporting?

For each included study, we will describe how we investigated the possibility of selective outcome reporting bias and what we found. We will search study protocols of the included trials in ClinicalTrials.gov; the World Health Organization’s International Trials Registry and Platform, and the ISRCTN Registry. For studies in which study protocols were published in advance, we will compare prespecified outcomes versus outcomes eventually reported in the published results. If the study protocols were not published in advance, we will contact study authors to gain access to the study protocol. We will assess the likelihood of selective reporting bias as:

• low risk (where it is clear that all of the study's prespecified outcomes and all expected outcomes of interest to the review were reported);

• high risk (where not all the study's prespecified outcomes were reported; one or more reported primary outcomes were not prespecified outcomes of interest and were reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported); or

• unclear risk.

7. Other sources of bias. Was the study apparently free of other problems that could put it at a high risk of bias?

For each included study, we will describe any important concerns we had about other possible sources of bias (e.g. whether there was a potential source of bias related to the specific study design or whether the trial was stopped early due to some data‐dependent process). We will assess whether each study was free of other problems that could put it at risk of bias as:

• low risk;

• high risk; or

• unclear risk

If needed, we will explore the impact of the level of bias through undertaking sensitivity analyses.

Contributions of authors

MB, OR, and AM reviewed the literature and wrote the protocol.

DL and BT commented on and reviewed the protocol.

Sources of support

Internal sources

• Institute for Clinical Sciences, Lund University, Lund, Sweden.

  MB, OR and DL are employed by this organization

• The University of Texas Health San Antonio School of Medicine Clinical Investigator Kickstart Pilot Grant, USA.

  to AM

External sources

• Vermont Oxford Network, USA.

  Cochrane Neonatal Reviews are produced with support from Vermont Oxford Network, a worldwide collaboration of health professionals dedicated to providing evidence‐based care of the highest quality for newborn infants and their families.

Declarations of interest

OR: no known conflicts of interest.

MB: no known conflicts of interest.

AM: is supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant KL2 TR001118. AM also has submitted a provisional patent to the US Patent and Trademark Office that involves byproducts of stem cells ("cell‐free therapies"). This patent does not conflict with this planned systematic review, as the review is only considering different types of cell‐based interventions.

BT: work on stem cells is supported by the Canadian Institute for Health Research, the Canadian Stem Cell Network, the Canadian Thoracic Society, the Ottawa Hospital Research Institute and the Children's Hospital of Eastern Ontario Research Institute, and the Ontario Institute of Regenerative Medicine.

DL: no known conflicts of interest.

New