Cell‐based therapies for amyotrophic lateral sclerosis/motor neuron disease

Abstract

Background

Amyotrophic lateral sclerosis (ALS), which is also known as motor neuron disease (MND), is a fatal disease associated with rapidly progressive disability, for which no definitive treatment exists. Current treatment approaches largely focus on relieving symptoms to improve the quality of life of those affected. The therapeutic potential of cell‐based therapies in ALS/MND has not been fully evaluated, given the paucity of high‐quality clinical trials. Based on data from preclinical studies, cell‐based therapy is a promising treatment for ALS/MND. This review was first published in 2015 when the first clinical trials of cell‐based therapies were still in progress. We undertook this update to incorporate evidence now available from randomised controlled trials (RCTs).

Objectives

To assess the effects of cell‐based therapy for people with ALS/MND, compared with placebo or no treatment.

Search methods

On 31 July 2019, we searched the Cochrane Neuromuscular Specialised Register, CENTRAL, MEDLINE, and Embase. We also searched two clinical trials registries for ongoing or unpublished studies.

Selection criteria

We included RCTs that assigned people with ALS/MND to receive cell‐based therapy versus a placebo or no additional treatment. Co‐interventions were allowed, provided that they were given to each group equally.

Data collection and analysis

We followed standard Cochrane methodology.

Main results

Two RCTs involving 112 participants were eligible for inclusion in this review. One study compared autologous bone marrow‐mesenchymal stem cells (BM‐MSC) plus riluzole versus control (riluzole only), while the other study compared combined intramuscular and intrathecal administration of autologous mesenchymal stem cells secreting neurotrophic factors (MSC‐NTF) to placebo. The latter study was reported as an abstract and provided no numerical data. Both studies were funded by biotechnology companies.

The only study that contributed to the outcome data in the review involved 64 participants, comparing BM‐MSC plus riluzole versus control (riluzole only). It reported outcomes after four to six months. It had a low risk of selection bias, detection bias and reporting bias, but a high risk of performance bias and attrition bias. The certainty of evidence was low for all major efficacy outcomes, with imprecision as the main downgrading factor, because the range of plausible estimates, as shown by the 95% confidence intervals (CIs), encompassed a range that would likely result in different clinical decisions.

Functional impairment, expressed as the mean change in the Amyotrophic Lateral Sclerosis Functional Rating Scale‐Revised (ALSFRS‐R) score from baseline to six months after cell injection was slightly reduced (better) in the BM‐MSC group compared to the control group (mean difference (MD) 3.38, 95% CI 1.22 to 5.54; 1 RCT, 56 participants; low‐certainty evidence). ALSFRS‐R has a range from 48 (normal) to 0 (maximally impaired); a change of 4 or more points is considered clinically important. The trial did not report outcomes at 12 months. There was no clear difference between the BM‐MSC and the no treatment group in change in respiratory function (per cent predicted forced vital capacity; FVC%; MD –0.53, 95% CI –5.37 to 4.31; 1 RCT, 56 participants; low‐certainty evidence); overall survival at six months (risk ratio (RR) 1.07, 95% CI 0.94 to 1.22; 1 RCT, 64 participants; low‐certainty evidence); risk of total adverse events (RR 0.86, 95% CI 0.62 to 1.19; 1 RCT, 64 participants; low‐certainty evidence) or serious adverse events (RR 0.47, 95% CI 0.13 to 1.72; 1 RCT, 64 participants; low‐certainty evidence). The study did not measure muscle strength.

Authors' conclusions

Currently, there is a lack of high‐certainty evidence to guide practice on the use of cell‐based therapy to treat ALS/MND. Uncertainties remain as to whether this mode of therapy is capable of restoring muscle function, slowing disease progression, and improving survival in people with ALS/MND. Although one RCT provided low‐certainty evidence that BM‐MSC may slightly reduce functional impairment measured on the ALSFRS‐R after four to six months, this was a small phase II trial that cannot be used to establish efficacy.

We need large, prospective RCTs with long‐term follow‐up to establish the efficacy and safety of cellular therapy and to determine patient‐, disease‐ and cell treatment‐related factors that may influence the outcome of cell‐based therapy. The major goals of future research are to determine the appropriate cell source, phenotype, dose and method of delivery, as these will be key elements in designing an optimal cell‐based therapy programme for people with ALS/MND. Future research should also explore novel treatment strategies, including combinations of cellular therapy and standard or novel neuroprotective agents, to find the best possible approach to prevent or reverse the neurological deficit in ALS/MND, and to prolong survival in this debilitating and fatal condition.

Plain language summary

Cell‐based therapies for amyotrophic lateral sclerosis/motor neuron disease (ALS/MND)

Review question

How effective and safe is cell‐based therapy in people with ALS/MND, when we compare it with an inactive treatment or no treatment?

Background

Amyotrophic lateral sclerosis (ALS; also known as motor neuron disease or MND) is a condition in which nerves in the brain and spinal cord that control movement (motor neurons) stop working. A person with ALS/MND has difficulty moving, swallowing, chewing and speaking, which become worse over time. Half of people with ALS/MND die within three years of their first symptoms. Weakness of muscles used in breathing often leads to death. The condition currently has no cure. Current treatment approaches largely focus on relieving symptoms to improve the quality of life of those affected.

Cell‐based therapy can be defined as injection of cellular material into a person to treat disease. Various types of cell‐based therapies have been tried in ALS/MND, including stem cell therapy. Stem cell therapy aims to provide new motor neurons, which may help stop or slow down disease progression in people with ALS/MND. Previous reviews supported the use of cell‐based therapy as a potential means of delaying the disease course in ALS/MND, but these were mainly based on preclinical animal models. Randomised controlled trials (RCTs) provide the most reliable evidence. In RCTs, one group receives the test treatment, and the other, 'control' group has an alternative treatment, a dummy treatment (placebo) or no treatment. Well‐performed RCTs provide the best evidence. Studies with no untreated group for comparison and small clinical trials have found no clinical benefits. Limited data from non‐RCTs involving a small number of people with ALS/MND and a short follow‐up period suggested that cell‐based therapy may slow disease progression. There is currently no approved cell‐based therapy for ALS/MND. We undertook this review to assess the RCT evidence now becoming available.

Study characteristics

Cochrane review authors searched medical databases for clinical trials. They found two completed RCTs that assessed the effects of cell‐based therapy over a six‐month follow‐up period. One study was not fully published and did not provide numerical data. Both studies were funded by stem cell companies. One study, which included 64 people with ALS/MND, provided data. The people taking part in the trial had an average time since symptom onset of about two years. They had mild to moderate problems with motor function (ability to perform physical tasks) at the start of the trial (with an average of 35 on the ALS Functional Rating Scale‐revised, on which a score of 0 indicates greatest impairment and 48 is normal function).

Key results and quality of the evidence

The study provided low‐quality evidence that stem cells obtained from people's own bone marrow (the cells in the centre of bone) did not result in significant side effects. The cell implantation procedure was well tolerated. Based on evidence from this trial, stem cell treatment may slightly reduce decline in motor function at six months, but may not improve breathing or quality of life at four months, or overall survival at six months. Based on the very limited evidence available, any benefit is uncertain due to there being only one poorly conducted study and results within the study varies. We urgently need large, well‐designed clinical trials to establish whether or not cell‐based therapies have a clear clinical benefit in ALS/MND. Major goals of future research are to identify the right type and amount of cells to use, and how best to administer them.

The evidence is up to date as of July 2019.

Summary of findings

Summary of findings for the main comparison. BM‐MSC compared to no treatment for amyotrophic lateral sclerosis/motor neuron disease.

BM‐MSC compared to no treatment for amyotrophic lateral sclerosis/motor neuron disease
Participants or population: people with probable or definite amyotrophic lateral sclerosis/motor neuron disease (treated with riluzole) Setting: hospital Intervention: BM‐MSC Comparison: no treatment
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with no treatment	Risk with BM‐MSC
Functional impairment, assessed using a functional rating scale (change from baseline to 6 months): mean change in ALSFRS‐R score Range: 48 (normal) to 0 (maximally impaired) Follow‐up: 6 months	The mean ALSFRS‐R score change in the no treatment group was –6.48 points	The mean ALSFRS‐R score in the BM‐MSC group was 3.38 points less than in the no treatment group (1.22 less to 5.54 less)	—	56 (1 RCT)	⊕⊕⊝⊝ Lowa,b	BM‐MSCs may reduce the decline in ALSFRS‐R score from baseline to 6 months.
Functional impairment, assessed using a functional rating scale (change from baseline to 12 months)	Not measured
Muscle strength: manual muscle testing (mean change from baseline to 12 months)	Not measured
Respiratory function: change in FVC% (mean change from baseline to 12 months) Follow‐up: 4 months	The mean change in FVC% in the no treatment group was –10.75%	The mean change in FVC% in the BM‐MSC group was –0.53% more (–5.37% more to 4.31% less)	—	56 (1 RCT)	⊕⊕⊝⊝ Lowa,c	BM‐MSCs may have little or no effect on FVC% change from baseline to 4 months.
Overall survival Follow‐up: 6 months	Study population		RR 1.07 (0.94 to 1.22)	64 (1 RCT)	⊕⊕⊝⊝ Lowa,d	BM‐MSCs may have little or no effect on overall survival at 6 months.
	903 per 1000	966 per 1000 (849 to 1000)
Adverse events, total Follow‐up: 4 months	Study population		RR 0.86 (0.62 to 1.19)	64 (1 RCT)	⊕⊕⊝⊝ Lowa,d	There may be little or no difference in total adverse events with BM‐MSCs compared to no treatment group.
	742 per 1000	638 per 1000 (460 to 883)
Serious adverse events Follow‐up: 4 months	Study population		RR 0.47 (0.13 to 1.72)	64 (1 RCT)	⊕⊕⊝⊝ Lowa,e	There may be little or no difference in serious adverse events with BM‐MSC compared to no treatment group.
	194 per 1000	91 per 1000 (25 to 333)
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). ALSFRS‐R: Amyotrophic Lateral Sclerosis Functional Rating Scale – Revised; BM‐MSC: bone marrow mesenchymal stem cells; CI: confidence interval; FVC: forced vital capacity; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: 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 certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aThe study that provided data was at high risk of bias for blinding of participants and personnel, which collectively posed a serious concern for the outcomes assessed (which require a degree of subjective judgement) and also at high risk of attrition bias.
 ^bDowngraded one level for imprecision. The 95% CI ranged from a relative trivial change in ALSFRS to a degree of change that was considered clinically important (changes in score of 4 or more) (Castrillo‐Viguera 2010).
 ^cDowngraded one level for imprecision. The 95% CI ranged from a substantial decrease to a substantial increase.
 ^dDowngraded one level for imprecision. The 95% CI encompassed a moderate decrease to a moderate increase in risk.
 ^eDowngraded one level for imprecision. The 95% CI encompassed a substantial decrease to a substantial increase in harms. Serious adverse events in the MSC group were not considered to be treatment‐related.

Background

Description of the condition

Motor neuron disease (MND) is a rare neurodegenerative disorder with an annual incidence of approximately 2 per 100,000 population. MND affects men and women of all ages, with a peak incidence at 50 to 70 years of age (Logroscino 2005; Logroscino 2008). The cause of MND is unknown, but up to 10% of cases are familial (Murray 2004). The clinical features of MND are attributable to the degeneration of neurons and corticospinal tracts from the primary motor cortex in the brain to the anterior horn cells in the spinal cord and brainstem nuclei (Rabin 1999). Four major categories of MND are recognised, namely amyotrophic lateral sclerosis (ALS), primary lateral sclerosis (PLS), progressive muscular atrophy (PMA) and progressive bulbar palsy (PBP). When the person presents with both upper and lower motor neuron signs, the disease is known as ALS, which is the most common form of MND. The terms PLS and PMA are applied when the initial presentation reflects only upper motor neuron involvement or only lower motor neuron involvement, respectively. PBP presents with weakness of bulbar muscles. The terminology can be confusing. In the UK and Australia, for example, MND is both the umbrella term for these conditions and is also used to refer to the ALS subtype. In the United States, however, ALS is the more usual umbrella term, but it also denotes the ALS subtype. This review uses ALS/MND.

Common clinical features of ALS/MND include wasting and weakness of the muscles for mastication, speech articulation and swallowing, intrinsic muscles of the hands and muscles of lower limbs. Respiratory failure due to respiratory muscle weakness is a late feature, leading to death (Caroscio 1987). Rarely, ALS/MND presents with acute respiratory failure (Chen 1996). The disease is virtually always fatal. Approximately half of people with ALS/MND die within three years from onset of symptoms, although 10% of people with ALS/MND live longer than 10 years (del Aguila 2003; Turner 2003).

The exact mechanism leading to selective cell death of motor neurons is not well understood and is likely to be multifactorial, involving genetic and environmental factors. Several genes have been identified as the cause of familial ALS/MND, including mutations in Cu²⁺/Zn²⁺ superoxide dismutase 1 (SOD1), TAR DNA‐binding protein 43 (TARDBP), fused in sarcoma (FUS), c9orf72, UBQLN2 and TANK‐binding kinase 1 (TBK1) (Deng 2011; Oakes 2017; Renton 2014). The genetic contribution to the majority of ALS/MND remains unknown. The neurodegenerative process of ALS/MND may involve a complex interplay between genetic factors, oxidative stress, glutamatergic excitotoxicity, protein aggregation, mitochondrial dysfunction, altered immune‐inflammation and impairment of axonal transportation. The surrounding glial cells have also been implicated in the pathogenesis via the release of inflammatory mediators, impaired neuronal metabolic support and dysfunctional signalling pathways (Lunn 2014; Shaw 2005). All these processes eventually lead to apoptosis of motor neurons.

To date, there is no curative treatment for ALS/MND. Current treatment approaches focus on relieving symptoms to improve the quality of life of those affected. Riluzole, an antiglutamate agent, is the most commonly used pharmacological treatment for ALS/MND. It has a small beneficial effect on bulbar function, limb function and survival, but no effect on muscle strength (Bensimon 1994; Goodall 2006; Miller 2012). In addition, edavarone (an antioxidant) was found to slow functional deterioration in some people with ALS/MND and is approved for use in Japan, Korea and the US (Abe 2014; Cruz 2018; Writing Group 2017). Many other pharmacological agents have been tried, but without clear benefit. In addition non‐pharmacological treatment, such as non‐invasive ventilation, prolongs median survival and improves quality of life in people with ALS/MND (Bourke 2006; Radunovic 2013).

Description of the intervention

Multipotential stem cells may provide an attractive therapeutic option because of their ability to migrate into damaged neural tissues and promote regeneration of neurons (neurogenesis). These multipotential stem cells produce neurotrophic (growth‐stimulating) factors, thus provoking the transdifferentiation of stem cells into neurons (Karussis 2010).

Cell‐based therapy can be defined as injection of cellular material into a person for therapeutic purposes. Various type of cells can be used including stem cells that are used to treat degenerative diseases (regenerative medicine), blood cancers and bone marrow diseases (bone marrow transplantation). To date, there have been numerous clinical trials of the treatment of ALS/MND with cell‐based therapy utilising cells isolated mostly from autologous (the person's own) bone marrow and peripheral blood, thus minimising the risk of rejection. The types of cells used for implantation have been bone marrow mononuclear cells (BM‐MNCs; Blanquer 2012; Deda 2009; Prabhakar 2012), bone marrow‐mesenchymal stem cells (BM‐MSCs; Baek 2012; Blanquer 2012; Karussis 2010; Martinez 2012; Mazzini 2003; Mazzini 2006; Mazzini 2012), granulocyte‐colony stimulating factor (G‐CSF)‐mobilised peripheral blood mononuclear cells (M‐PBMNCs; Cashman 2008; Chio 2011; Nefussy 2010), olfactory ensheathing stem cells (OESC; Chen 2007; Chew 2007; Giordana 2010; Huang 2008; Piepers 2010), and neural stem cells (NSCs; Feldman 2014).

BM‐MNCs are usually separated by a density gradient method from bone marrow aspirate obtained from the individual's hip bone. Mesenchymal stem cells (MSCs) can be easily isolated from bone marrow, placenta, muscle and fat. The cells are subsequently cultured for three to five weeks to provide large numbers for therapeutic application. These cells can be expanded in vitro with no risk of malignant transformation (Bernardo 2007). The process of obtaining M‐PBMNCs involves administration of G‐CSF to increase the number of M‐PBMNCs in the circulation, followed by their removal using a blood cell separation machine (apheresis). OESCs are extracted from human foetal olfactory bulb tissue and cultured for two to three weeks. NSCs used in clinical studies are cultured human NSCs derived from a single source human foetal spinal cord tissue of approximately eight gestational weeks and expanded serially by epigenetic means only (Feldman 2014).

Implantation of cells has been performed via several methods. The common methods include intrathecal (into the subarachnoid space via the spinal canal), intracortical (into the cerebral cortex) and direct transplantation of autologous MSCs into surgically exposed spinal cord under general anaesthesia. Studies have shown that direct transplantation of autologous cells into the spinal cord is well tolerated and feasible in people with ALS/MND (Feldman 2014; Mazzini 2012).

Several clinical trials have provided important insights into the safety and feasibility of stem cell mobilisation and transplantation in people with ALS/MND. Previous reviews supported the use of cell‐based therapy as a means of delaying the disease course in ALS/MND, mainly based on preclinical animal models (Goutman 2015; Thonhoff 2009). Single‐arm and small clinical trials observed no clinical benefits. Limited data from non‐RCTs involving a small number of people with ALS/MND and a short‐term follow‐up period suggested that cell‐based therapy slowed the rate of disease progression (Lunn 2014). Uncertainties remain, however, regarding its ability to achieve functional improvement and its long‐term safety profile; in particular, whether this mode of therapy is associated with acceleration of disease progression (Lunn 2014).

How the intervention might work

There are two possible mechanisms by which stem cell therapy may help in the treatment of ALS/MND. First, by using progenitor cells that have been generated ex vivo to regenerate dying neuronal cells. Experimental observations showed that transplanted stem cells and mononuclear cells have the capacity to stimulate the regenerative processes of motor neurons (Mazzini 2003). In animal models of ALS/MND, stem cell transplantation can significantly slow the progression of the disease and prolong survival (Mazzini 2003). Increasing numbers of preclinical studies have shown that transplanted stem cells are capable of migrating to regions of experimentally induced nerve injury, where they are able to proliferate and differentiate into neurons and glial cells (Jiang 2002; Liu 2000; McDonald 1999; Terada 2002; Woodbury 2000). The types of stem cell that have been tested in preclinical models include BM‐MSCs, MSCs, cord blood stem cells, embryonic stem cells, neural stem and progenitor cells, human glial restricted progenitors, and induced pluripotent stem cells (IPCs).

Second, stem cells promote the survival of existing neurons. MSCs are very attractive candidates for cell therapy in ALS/MND because of their great plasticity (Chen 2008), and immunomodulatory properties (Mazzini 2012). MSCs can induce a neuroprotective microenvironment via anti‐inflammatory and immunosuppressive effects on astrocytes and microglial cells (Uccelli 2008). MSCs release soluble molecules such as cytokines and chemokines, and express immune‐relevant receptors such as chemokine receptors and cell adhesion molecules that ameliorate inflammation and stimulate the survival of neuronal cells (Uccelli 2008). Preclinical data have shown that MSCs are capable of transdifferentiation into neurons and glial cells both in vitro and in vivo (Black 2001; Kim 2002; Sanchez‐Ramos 2000). In addition, NSCs have the ability to generate immunomodulatory cells, growth‐factor‐releasing cells and functional support cells to modify motor neuron survival and activity (Gowing 2011).

Most studies on the pathogenesis of ALS/MND thus far have been in animals. There are many limitations when extrapolating the findings observed in animal models into humans. First, there are interspecies differences in neuronal physiology and specific gene‐splicing patterns (Hardingham 2010). Second, there is an overemphasis on models based on rat SOD1, when most cases of human sporadic ALS/MND may not have a SOD1 defect. In this respect, stem cells could be used to model disease, allowing us to further explore the pathophysiological process of ALS/MND.

Why it is important to do this review

The lack of effective pharmacological treatment for ALS/MND and compelling preclinical data have provided a rationale for the therapeutic application of stem cells for this devastating incurable disease. Early clinical trials have suggested that stem cells could have the potential to replace and repair damaged motor neurons in people with ALS/MND (Martinez 2012; Mazzini 2003; Mazzini 2015). Moreover, the procedures of expansion and transplantation of these cells into people with ALS/MND are well tolerated and feasible. However, most of the clinical trials involved small numbers of participants, which can produce false‐positive results, or overestimate the magnitude of an association; consequently, the results have been inconsistent. Additionally, small trials may fail to detect rare adverse events. Combining available data in a systematic review may increase the likelihood of detecting a true effect of the intervention, thus allowing meaningful conclusions to be drawn. It is also important to know whether cells derived from different sources have different impacts on clinical outcomes among people with ALS/MND. For example, stem cells obtained from different sources may possess different biological properties (plasticity, self‐renewal, differentiation, homing, migration and secretion of trophic factors) and different immunological properties (modulating immune response). These differences may be attributed to the inherent biological properties of the stem cells or changes that occur during enrichment and processing. Moreover, questions regarding the optimal treatment approaches, including the cell dose, phenotype, preparation and delivery system, remain to be answered (Abdul 2018).

This systematic review set out to determine the efficacy, feasibility and safety of cell‐based therapy in people with ALS/MND. The findings of this review may facilitate design of the optimal cell‐based therapy programme for people with ALS/MND as well as identify critical areas for improvement and recommendations for future clinical trials. The review was first published in 2015 when the first clinical trials of cell‐based therapies were still in progress (Abdul‐Wahid 2016). We updated it in 2019 to incorporate evidence now available from randomised controlled studies.

Objectives

To assess the effects of cell‐based therapy in people with ALS/MND compared with a placebo or no treatment.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs), quasi‐RCTs and cluster RCTs. Quasi‐random methods of assignment to interventions are systematic methods that are not truly random, such as allocation using alternation, date of birth, day of visit or medical record number.

Types of participants

We included people of any age with a diagnosis of definite or probable ALS/MND according to accepted criteria, such as the revised El Escorial World Federation of Neurology criteria (Brooks 2000).

Types of interventions

Mononuclear cells or stem cells compared with a placebo or no additional treatment. We would have permitted the use of co‐interventions including standard treatment such as riluzole and symptomatic treatment, provided that they were administered to each group equally.

Types of outcome measures

Primary outcomes

1. Functional impairment, assessed using a functional rating scale (change from baseline to six months): change in functional rating scale, such as the Amyotrophic Lateral Sclerosis Functional Rating Scale‐Revised (ALSFRS‐R) (Cedarbaum 1999) at 6 or 12 months.

A change in ALSFRS‐R score of 4 or higher is considered clinically important (Castrillo‐Viguera 2010).

Secondary outcomes

1. Functional impairment, assessed using a functional rating scale (change from baseline to 12 months): change in functional rating scale, such as the ALSFRS‐R (Cedarbaum 1999).

2. Muscle strength: change in manual muscle testing of the upper and lower limbs (Medical Research Council (MRC) grade) at 6 and 12 months.

3. Respiratory function: change in upright forced vital capacity (FVC) at 6 and 12 months.

4. Nerve conduction: change in compound muscle action potential (CMAP), neurophysiological index (NI), combined motor index (CMI), motor unit number estimation (MUNE) and motor unit number index (MUNIX) at 6 and 12 months (Escorcio‐Bezerra 2016; Gawel 2016; Stein 2016).

5. Mood state and quality of life: change in mood state and quality of life using the Profile of Mood State (POMS) and quality of life scale questionnaires (such as ALS Assessment Questionnaires, ALSAQ‐40 or ALSQ5, Short‐Form 36 (SF‐36) Health Survey and EQ‐5D) at 6 and 12 months (Jenkinson 2000; Jenkinson 2007; Rabin 2001; Ware 1992).

6. Structural changes in serial magnetic resonance imaging (MRI): such as T2‐weighted and Fluid‐Attenuated Inversion Recovery (FLAIR) hyperintense signals in corticospinal tracts, precentral and frontal cortex at 6 and 12 months.

7. Overall survival: at 6 and 12 months.

8. Adverse events: including an inflammatory reaction at the cell injection site, cardiovascular and thromboembolic complications, adverse events, serious adverse events and the rate of withdrawal from the study.

Search methods for identification of studies

Electronic searches

We searched the following databases:

1. the Cochrane Neuromuscular Specialised Register via the Cochrane Register of Studies (CRS‐Web; 31 July 2019; Appendix 1);

2. the Cochrane Central Register of Controlled Trials (CENTRAL) via the CRS‐Web (31 July 2019; Appendix 2);

3. MEDLINE (1946 to 30 July 2019; Appendix 3);

4. Embase (1974 to 29 July 2019; Appendix 4).

We applied no language restrictions. We included studies reported as full‐text publications as well as those published as abstracts and proceedings.

We also conducted a search of the following trials registries US National Institutes of Health Clinical Trials Registry (www.clinicaltrials.gov), and the World Health Organization International Clinical Trials Registry Platform (ICTRP; apps.who.int/trialsearch/) to identify other ongoing and unpublished studies.

For the first version of the review (Abdul‐Wahid 2016), we searched the National Institute for Health Research Database of Abstracts and Reviews of Effects (DARE) and Health Technology Assessments (HTA) database to identify reviews and assessments for inclusion in the 'Discussion' section. We searched National Health Service Economic Evaluation Database (NHS EED) for any available cost information for the 'Discussion' section. We did not search these databases for this update as they are no longer included in the Cochrane Library.

Searching other resources

We searched for additional references in reference lists of all primary studies and review articles. We contacted the authors of RCTs and other experts in the field to obtain any additional published or unpublished studies. We searched relevant manufacturers' websites for trial information to identify further relevant studies. In addition, we handsearched journals for relevant articles: Cytotherapy (January 1999 to June 2018), Cell Transplantation (Issue 1, 2001 to Issue 6, 2018), Cell Stem Cell (Issue 1, 2007 to Issue 6, 2018) and Stem Cells (Issue 1, 1993 to Issue 6, 2018).

Data collection and analysis

Selection of studies

Two review authors (SFAW and ZKL) independently screened titles and abstracts of all studies identified from the first round of searching. We coded potentially relevant studies or studies that required further assessments as 'retrieve' based on the relevance of the population, intervention and outcomes to our review question. We coded studies clearly not relevant as 'do not retrieve'. Two review authors (SFAW and ZKL) inspected the full‐text versions of the studies coded 'retrieve' to further identify trials to be included in our meta‐analysis, based on the relevance of the population, intervention, comparison and study design. Among studies retrieved but excluded, we recorded reasons for exclusion. We resolved any disagreement through discussion and did not require consultation with a third person. We identified and excluded duplicates, and collated multiple reports of the same study, making each study the unit of interest in the review rather than each report. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Moher 2009) and Characteristics of excluded studies table.

We did not identify any included studies for quantitative analysis. We provide a qualitative narration of the excluded studies in the Discussion.

Data extraction and management

Two review authors (SFAW and ZKL) independently extracted the following study characteristics from included studies according to the methods described in our protocol (Abdul Wahid 2015).

1. Methods: study design, date and duration, details of any 'run‐in' period (time in a study before participants receive treatment), number of study centres and location, setting, withdrawals.

2. Participants: number, age (mean or median age, range), gender, disease severity, diagnostic criteria, inclusion and exclusion criteria.

3. Interventions: intervention and co‐intervention and comparison.

4. Outcomes: primary and secondary outcomes specified and collected, and time points reported.

5. Notes: date trial conducted, funding for trial, notable conflicts of interest of trial authors.

Two review authors (SFAW and ZKL) independently extracted the outcome data and noted in the Characteristics of included studies table if outcome data were not reported in a usable way, with disagreements resolved by consensus or by involving another review author (NAI). One review author (NAI) transferred data into Review Manager 5 (RevMan 5) software (Review Manager 2014), one review author (SFAW) checked the outcome data entries, and another review author (NML) spot‐checked study characteristics for accuracy against the trial report.

If reports had required translation, the authors would have extracted data from the translation provided, with data cross‐checked against the original report if possible.

Assessment of risk of bias in included studies

Two review authors (NML and SFAW) independently assessed the risk of bias for each study according to the domains listed below, as outlined in the Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a), with resolution of disagreement by discussion or by involving another author (NAI).

1. Random sequence generation.

2. Allocation concealment.

3. Blinding of participants and personnel.

4. Blinding of outcome assessment.

5. Incomplete outcome data.

6. Selective outcome reporting.

7. Other bias, such as premature termination and extreme baseline imbalance.

We made a judgement on each of the criteria above as to whether the study was at high, low or unclear risk of bias. We assessed blinding for each category of outcomes (objective and subjective) separately when possible. We completed a 'Risk of bias' table for each eligible study and resolved disagreement among review authors through discussion leading to consensus. We presented an overall assessment of the risk of bias using the 'Risk of bias' graph and the 'Risk of bias' summary.

Measures of treatment effect

For dichotomous data (adverse event, serious adverse event and overall survival), we used risk ratio (RR) with 95% confidence intervals (CI) to measure outcome estimates on the same scale. For continuous data (ALSFRS‐R score change, FVC % change, SF‐36 change), we used mean differences (MD) with 95% CIs for conceptually similar outcomes measured on the same scale, or standardised mean differences (SMD) with 95% CIs for conceptually similar outcomes measured on different scales. In this case, we would have adjusted all the scales to achieve a consistent direction of effect.

We planned to undertake meta‐analyses only where the participants, intervention, comparison and outcomes were similar enough for pooling to be meaningful, and to only narratively describe skewed data reported as medians and interquartile ranges.

Unit of analysis issues

For cluster RCTs (in other words, trials in which the assignment to intervention or control group was made at the level of the unit/ward rather than the individual participant), we would have assessed whether the study authors had made appropriate adjustments for the effects of clustering, using appropriate analysis models such as the Generalized Estimating Equation model. We would have inspected the width of the standard error (SE) or 95% CI of the estimated treatment effects to double‐check the possible unit of analysis in the study. If we found an inappropriately small SE or a narrow 95% CI, we would have asked the authors of the study to confirm the unit of analysis.

If no adjustment had been made for the effects of clustering, we would have performed adjustments by multiplying the SEs of the final effect estimates by the square root of the 'design effect', represented by the formula, 1 + (M – 1) × ICC, where M is the mean cluster size (number of participants per cluster) and ICC is the intracluster correlation. The mean cluster size (M) from each trial would be determined by dividing the total number of participants by the total number of clusters. We planned to use an assumed ICC of 0.10, as we consider this to be a realistic general estimate that is derived from previous studies on implementation research (Campbell 2001). We would have combined the adjusted final effect estimates from each trial with their SEs in meta‐analyses using generic inverse variance methods, as stated in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).

If the determination of the unit of analysis was not possible, we planned to include the studies concerned in a meta‐analysis using the effect estimates reported by the authors. We would also have performed sensitivity analyses to assess how the overall results were affected by the removal of the studies in which adjustment of unit of analysis was appropriate but not possible and the unit of analysis was unknown.

Dealing with missing data

If key information were missing, such as study characteristics, methods or outcome data, we planned to contact investigators to obtain the relevant information. Where this was not possible, we would have conducted a deterministic sensitivity analysis at the study level by adopting an approach as recommended in chapter 16.2.3 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a) for handling sensitivity analysis for continuous outcomes, for instance, our primary outcome of change in functional rating scales, such as the ALSFRS or Expanded Disability Status Scale (EDSS) measured at six months. In this approach, we would have assumed a fixed difference of 10% between the mean of the missing data and the observed mean of the available data in the intervention arm as well as the control arm, with the direction of the difference in opposition to the observed direction of the effect size in the analysis. Following is a possible scenario: in our analysis of the available data, the intervention arm has a higher mean, indicating that the effect favours the intervention arm (higher value in the scale indicating a better outcome). In this instance, we would have performed the sensitivity analysis by assuming that the mean of the missing data in the intervention arm was 10% lower than the mean of the available data in the intervention arm, and accordingly, the mean of the missing data in the control arm was 10% higher than the mean of the available data in the control arm. We would then have pooled the imputed data with the observed data for each arm and re‐assessed the imputed effect estimate having compared the imputed pooled estimate of the two arms. If the effect estimate of the study changed substantially following our sensitivity analysis using this approach, we would have considered the study at high risk of attrition bias. At the review level, we would have conducted a sensitivity analysis to explore the impact of including such studies with high risk of attrition bias in the overall pooled estimates of the major outcomes.

Assessment of heterogeneity

We would have used the I² statistic to measure heterogeneity among the trials in each analysis. If we had identified substantial unexplained heterogeneity (as shown by an I² statistic greater than 50%), we would have explored possible causes by prespecified subgroup analyses (see Subgroup analysis and investigation of heterogeneity).

Assessment of reporting biases

If we had been able to pool more than 10 trials, we would have created a funnel plot to explore possible publication biases. If we had found significant asymmetry in the funnel plot, which might indicate possible publication bias, we would have reported this with a note of caution in the discussion, taking into account the area of the void in the funnel plot. We did not plan to further explore publication bias using statistical methods in view of the limitations of these methods in the presence of the relatively small number of studies in a typical systematic review (Higgins 2011a).

Data synthesis

We would have performed meta‐analyses in Review Manager 5 using a fixed‐effect model (Review Manager 2014). If suitable data had been available, we planned to perform a sensitivity analysis to assess the change in the overall results with a random‐effects model.

'Summary of findings' table

We created a 'Summary of findings' table comparing cell‐based therapy versus placebo or no additional treatment using the following outcomes.

1. Functional impairment, assessed using a functional rating scale (change from baseline to six months): ALSFRS.

2. Functional impairment, assessed using a functional rating scale (change from baseline to 12 months): ALSFRS.

3. Muscle strength: manual muscle testing (at six months).

4. Respiratory function: change in upright FVC (at six months).

5. Overall survival (at six and 12 months).

6. Adverse events (at any given time point).

Our judgement on the overall certainty of the body of evidence was guided by the five GRADE considerations, namely limitations in study design, consistency of effect, imprecision, indirectness and publication bias. We used methods and recommendations described in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b) using the GRADE profiler (GRADEpro) software (GRADEpro 2018). We downgraded the certainty of evidence once if a GRADE consideration was present to a serious degree and twice if very serious. We justified decisions to downgrade or upgrade the certainty of evidence using footnotes in the 'Summary of findings' table.

Subgroup analysis and investigation of heterogeneity

We planned to carry out a subgroup analysis based on the type of cell‐based therapy received (i.e. BM‐MNCs, BM‐MSCs, M‐PBMNCs, OESCs or NSCs). We would also have conducted a subgroup analysis based on delivery method (i.e. intrathecal, intracranial, intraspinal and intravenous) and for the primary endpoint that was measured at two separate time points. We would have used functional scales, such as the EDSS and ALSFRS, FVC, quality of life scores, MRI changes, survival rate, neurophysiological index and adverse events as outcome measures.

We planned to use the formal test for subgroup interactions in Review Manager 5 (Deeks 2011).

Sensitivity analysis

We planned to carry out the following sensitivity analyses if the review included sufficient studies.

1. Repeat the analysis excluding studies at high risk of selection and attrition biases.

2. Repeat the analysis with a random‐effects model.

3. Repeat the analysis excluding unpublished studies.

If the overall results were affected substantially by the sensitivity analysis, we would have placed a note of caution in our discussion and conclusions regarding the certainty of our estimates, and proposed a need for further research where appropriate to explore the possible sources of variation in the outcome estimates.

Results

Description of studies

Results of the search

The previous version of this review identified no studies for inclusion. In 2019, we identified 151 new papers from database searches as potentially relevant and after we reviewed these, two RCTs (112 participants) met the inclusion criteria for the review. Of the two studies eligible for inclusion, one study contributed outcome data (Oh 2018), as the outcome data of the other study were not presented in an extractable form for meta‐analysis (Gothelf 2017). The results of Gothelf 2017 were published as an abstract only and conclusions were stated without any numbers being given. We also identified four ongoing trials (NCT01254539; NCT02286011; NCT02290886; NCT03280056). Figure 1 shows a summary of the results of the search.

1.

Study flow diagram.

Included studies

See Characteristics of included studies table for full details of the included studies. The following are major characteristics of the included studies.

Country

One study was conducted in the US (Gothelf 2017), the other in the Republic of Korea (Oh 2018).

Methods

Both studies were phase 2, parallel‐group RCTs.

Population

Both studies enrolled adults (aged 18 to 75 years) with probable or definite diagnosis of ALS/MND, following the revised El Escorial World Federation of Neurology criteria (Brooks 2000).

Interventions and comparators

The two included studies made two different comparisons. Gothelf 2017 compared autologous mesenchymal stem cells secreting neurotrophic factors (MSC‐NTF) cells by combined intramuscular and intrathecal administration (dosage not stated) with placebo (follow‐up period: 24 weeks). Oh 2018 compared BM‐MSCs (two intrathecal injection with 1 × 10⁶ cells per kilogram bodyweight at an interval of 26 days) with riluzole treatment with riluzole alone (follow‐up period: 6 months).

Outcomes

As both studies were phase 2 trials, safety outcomes were the primary outcomes of interest, and these included the assessment of overall adverse effects (Gothelf 2017; Oh 2018), and serious adverse effects (Oh 2018). Additionally, Gothelf 2017 evaluated changes in function as measured by ALSFRS and changes in vital capacity, both measured at 24 weeks as secondary outcomes. Oh 2018 evaluated changes in the degree of disability as measured by the ALSFRS‐R as the primary outcome, with changes in Appel scale, FVC, and change in health‐related quality of life as measured by the change in SF‐36, as secondary outcomes.

Excluded studies

We assessed and excluded five studies, reported in 12 full‐text articles, as follows.

1. Four non‐randomised trials (11 records). These trials investigated the safety and efficacy of foetal human NSCs in six participants (Mazzini 2015), MSC‐NTF in 38 participants (Karussis 2013; Petrou 2015), and BM‐MSC in 15 participants (Gabr 2017).

2. One study assessed the safety and feasibility of intravenous and intrathecal injections of BM‐MSCs in people with ALS/MND (Nabavi 2019).

See Characteristics of excluded studies table.

Ongoing studies

Four trials are ongoing (NCT01254539; NCT02286011; NCT02290886; NCT03280056). NCT01254539 is a randomised, double‐blind trial to assess the feasibility and safety of intraspinal and intrathecal infusion of autologous bone marrow stem cells. NCT02286011 is a phase 1 randomised, double‐blind, controlled clinical trial on intramuscular infusion of autologous BM‐MNC. NCT02290886 is a phase 1/2 randomised, placebo‐controlled, triple‐blind trial to evaluate the safety and efficacy of intravenous autologous adipose tissue‐derived MSCs at three different doses (one million, two million, and four million autologous MSCs). Finally, NCT03280056 is a phase 3, randomised, double‐blind, multicentre trial to evaluate the safety and efficacy of three intrathecal administrations of MSC‐NTF cells at a bi‐monthly intervals, compared to placebo. See Characteristics of ongoing studies table.

Risk of bias in included studies

Both included studies were parallel‐group, randomised controlled, phase 2 trials, with one labelled as "double‐blind" (Gothelf 2017), and another open‐label and assessor blinded (Oh 2018). For a summary of 'Risk of bias' assessments, see Figure 2.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation (selection bias)

We judged one study at low risk of selection bias, as the participants were randomised using an interactive web response system (Oh 2018). Authors of the other study did not mention the methods of sequence generation and allocation, hence we accorded it an unclear risk of selection bias (Gothelf 2017).

Blinding (performance bias and detection bias)

We judged one study to have a high risk of performance bias, as it was stated to be open label, but at low risk of detection bias, as neurologists and evaluators who were blinded to treatment assignments performed the outcome assessment (Oh 2018). Another study was labelled "double‐blind" (Gothelf 2017); however, there was no detailed description of the blinding mechanism and so we accorded it an unclear risk of performance bias. The risk of detection bias was unclear, as the authors did not clearly mention the blinding status of the outcome assessors.

Incomplete outcome data (attrition bias)

We judged one study to have high risk of attrition bias (Oh 2018). The overall dropout rate was high and the number of dropouts was unequal between the two groups. Four participants in the control group (withdrawal: 2; death: 1; serious adverse event: 1) and one participant in the MSC group (withdrawal: 1) were excluded from full analysis because these events occurred before the baseline visit. The number of participants analysed was not stated in the records obtained for the other included study (Gothelf 2017), hence we judged the risk of bias to be unclear.

Selective reporting (reporting bias)

We judged Oh 2018 to have low risk of reporting bias, as the published article published all the endpoints and outcomes specified in the protocol, such as ALSFRS‐R score change, FVC% change, and SF‐36 change; the trial authors also reported adverse events, serious adverse events, and overall survival (the last as a post hoc analysis). We judged Gothelf 2017 to have a high risk of bias for selective reporting, because the outcome data were not presented in an extractable form for meta‐analysis and attempts to contact study authors were unsuccessful.

Other potential sources of bias

We screened for other potential sources of biases including extreme baseline imbalance and unit of analysis issues. One included study was at low risk of other potential sources of bias (Oh 2018), while the information for another study, presented in abstract form as well as trial registry, were insufficient for an assessment of other potential biases (Gothelf 2017).

Effects of interventions

See: Table 1

Autologous bone marrow‐mesenchymal stem cells versus no treatment

One study, which had 64 participants, compared BM‐MSC injection to no treatment (Oh 2018). Participants in both groups received riluzole. This study did not report clinical outcomes at 12 months after cell injection and structural outcomes measured by MRI, which we had proposed in our protocol (Abdul Wahid 2015). Outcomes were measured at four months, other than the ALSFRS‐R score, which was measured at four and six months. See Table 1.

Primary outcomes

Functional impairment, assessed using a functional rating scale (change from baseline to six months)

ALSFRS‐R

The mean decline in ALSFRS‐R score (range: 48 (normal) to 0 (maximally impaired) from baseline to six months after cell injection was slightly less in the BM‐MSC group than in the control group (MD 3.38, 95% CI 1.22 to 5.54; 56 participants; low‐certainty evidence; Analysis 1.1). We downgraded the certainty of evidence one level each for risk of bias and imprecision. A change in ALSFRS‐R score of 4 or more points is considered clinically important (Castrillo‐Viguera 2010).

1.1. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 1 Functional impairment: change in Amyotrophic Lateral Sclerosis Functional Rating Scale – Revised score.

Secondary outcomes

Functional impairment, assessed using a functional rating scale (change from baseline to 12 months)

The included study did not measure outcomes beyond six months.

Muscle strength

The study did not report data for muscle strength.

Respiratory function

The study did not measure respiratory function at six or 12 months.

There was no clear difference between the two groups in the changes in FVC% from baseline to four months (MD –0.53, 95% CI –5.37 to 4.31; 56 participants; low‐certainty evidence; Analysis 1.2). We downgraded the certainty of the evidence one level each for risk of bias and imprecision.

1.2. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 2 Respiratory function: change in % forced vital capacity.

Nerve conduction

The study did not measure nerve conduction.

Mood state and quality of life

The study did not measure mood and did not report quality of life at six or 12 months.

The study reported mean changes in SF‐36 scores at four months. There was no clear difference in the changes in SF‐36 between the two groups (MD 2.77, 95% CI –3.50 to 9.04; 57 participants; Analysis 1.3).

1.3. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 3 Quality of life: change in 36‐item Short Form score.

Structural changes in serial magnetic resonance imaging

The study did not measure structural changes in serial MRI.

Overall survival

In the BM‐MSC group, 32/33 participants were alive at six months, compared to 28/31 participants in the control group. One death in the BM‐MSC group and two deaths in the control group were caused by respiratory failure as a result of disease progression. One control group participant died from a sudden cardiac arrest before the treatment phase of the study. BM‐MSC had little or no effect on overall survival at six months (RR 1.07, 95% CI 0.94 to 1.22; 64 participants; low‐certainty evidence; Analysis 1.4; Table 1). The study did not report overall survival at 12 months.

1.4. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 4 Survival at 6 months.

Adverse events

Total adverse events

The common adverse events in the MSC group at six months were influenza‐like illness (seven participants), back pain (five participants) and musculoskeletal pain (five participants). The proportion of participants with any adverse event was 9% (3/33, four events) in the BM‐MSC group, and included headache (two events), pyrexia (one event), and pain at administration site (one event); these adverse drug reactions were mild and transient, occurred within 48 hours postinjection, and were self‐limited or subsided with simple analgesics within 48 hours of treatment. There was no clear difference in the proportion of participants with any adverse event between the two groups (RR 0.86, 95% CI 0.62 to 1.19; 64 participants; low‐certainty evidence, downgraded one level each for risk of bias and imprecision; Analysis 1.5; Table 1).

1.5. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 5 Adverse events, total.

Serious adverse events

The incidence of serious adverse events during the entire follow‐up period was 9% (3/33, three events) in the MSC group versus 19% (6/31, six events) in the control group; serious adverse events in the MSC group were not considered to be treatment‐related. There was no clear difference in serious adverse events between the two groups (RR 0.47, 95% CI 0.13 to 1.72; 64 participants; low‐certainty evidence, downgraded one level each for risk of bias and imprecision; Analysis 1.6).

1.6. Analysis.

Comparison 1 Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment, Outcome 6 Serious adverse events.

Four deaths occurred during the six‐month follow‐up period: 1/33 participants in the MSC group (respiratory failure at five months postinjection, related to disease progression) and 3/31 participants in the control group (two of respiratory failure and one of sudden cardiac arrest before visit 3).

Autologous mesenchymal stem cells secreting neurotrophic factors versus placebo

Gothelf 2017 investigated the effects of MSC‐NTF versus placebo in 48 participants. The trial reported an improvement in ALSFRS‐R slope of decline in the treatment group with no improvement in the placebo group, but provided no numerical data. Gothelf 2017 also reported that there were no deaths or treatment‐related serious adverse events. Treatment‐related adverse events were mostly expected and transient, but the report did not elaborate further. We emailed the contact author twice in 2018 for additional information, but did not obtain a response.

Discussion

Summary of main results

We identified two RCTs of cell‐based therapy in ALS/MND. The total number of participants was 112; however, one trial was published as an abstract which did not provide numerical outcome data. Our findings are from one trial of intrathecal BM‐MSCs (Oh 2018), which involved 64 participants and reported outcomes after four or six months.

People with ALS/MND who received intrathecal BM‐MSCs appeared to have a small but statistically significant reduction in functional impairment as measured by changes in ALSFRS‐R scores over four or six months, with no clear differences compared to the control group in change in per cent predicted FVC, change in quality of life measured by the SF‐36, survival at six months, or total and serious adverse events (Table 1). The study did not measure muscle strength. It is uncertain if the difference in ALSFRS‐R is clinically meaningful, since the MD was 3.38 (95% CI 1.22 to 5.54) at six months, which was less than the 4‐point change previously considered by people with ALS/MND to be the minimal clinically important change (Castrillo‐Viguera 2010). Furthermore, there is imprecision in the effect size estimates as the 95% CI ranges from a relative trivial change in ALSFRS to a degree of change that is considered clinically important.

A second RCT with 48 participants reported that combined intrathecal and intramuscular autologous MSC‐NTF led to an improvement in the ALSFRS‐R slope of decline in the treatment group, with no improvement in the placebo group (Gothelf 2017). There were no deaths, treatment‐related serious adverse events or transient treatment‐related adverse events. However, the result was only published as an abstract at the time of this review and we were thus not able to include this study in the quantitative analysis.

Overall completeness and applicability of evidence

This systematic review showed that there is a lack of high‐certainty evidence to support the use of cell‐based therapy for people with ALS/MND. Uncertainties remain as to whether this mode of therapy is capable of restoring muscle function, slowing disease progression and improving survival in people with ALS/MND. Although low‐certainty evidence from one RCT showed that intrathecal BM‐MSC probably improves disability in people with ALS/MND compared to a control group, this was a small phase 2 trial, which cannot be used to establish efficacy.

Currently, the data regarding the neuroprotective properties of different cell‐based products are limited. Furthermore, the key elements, including cell source, phenotype, dose and method of implantation, which will be critical in designing optimal cell‐based therapy for people with ALS/MND, remain unclear.

Certainty of the evidence

The certainty of evidence was low for all major outcomes, with risk of bias and imprecision as the main downgrading factors, due to high risk of bias in blinding of participants and personnel as well as incomplete outcome data, and also because the range of plausible estimates, as shown by the 95% CIs, encompassed a range that would likely result in different clinical judgements and decisions. This, in turn, reflected the fact that there was only a single included study of limited sample size with some risk of bias concerns.

Potential biases in the review process

We performed comprehensive searches in the Cochrane Neuromuscular Specialised Register, CENTRAL, MEDLINE and Embase. Research conducted to answer the review question is still in its early stages, with several phase 1 and phase 2 studies identified, although only two phase 2 RCTs were published to‐date, with four ongoing studies. Despite our comprehensive searches, it is possible that we missed some relevant articles and conference presentations not listed in the databases above or not captured in the handsearching process.

Agreements and disagreements with other studies or reviews

As most human studies to date focus on the safety of various types of cell‐based products and the feasibility of the surgical implantation technique; these issues have been the focus of previous reviews. Despite the lack of reports of harms of cell‐based therapy, data on efficacy in humans are still very limited. Preclinical animal studies of stem cell therapy show promising efficacy of stem cell in ALS/MND but, in contrast, there is a lack of high‐certainty evidence on the efficacy of stem cell therapy for people with ALS/MND, due to the paucity of high‐quality RCTs.

One meta‐analysis of nine preclinical in vivo studies and 12 retrospective descriptive clinical studies that were published between March 2009 and March 2015 confirmed the efficacy of stem cell therapy in improving survival in animal models (transgenic mice expressing human mutated superoxide dismutase 1), where an MD of 9.79 days (95% CI 4.45 to 15.14) in lifespan favoured stem cell therapy (Moura 2016). In contrast, the data in the Moura 2016 review from clinical studies were insufficient to assess effectiveness of stem cell therapy, and could only demonstrate itsfeasibility and "the absence of serious adverse events". However, even this conclusion should be interpreted with caution because all clinical studies were heterogeneous and of an unsatisfactory quality.

Jeong 2015 performed a meta‐analysis of 11 single‐arm studies of stem cell‐based therapy in people with ALS/MND. This single‐arm meta‐analysis showed that the pooled MD in ALSFRS from baseline was decreased by 3.3 points (95% CI –5.38 to –1.22) in which ALSFRS declined 0.4 points per month, while pooled MD in FVC from baseline was decreased by 14% (95% CI –19% to –6%), mean decline in FVC was 1.2% per month. According to natural history data, the ALSFRS score declines 1.01 points and FVC declines by 2.7% per month in people with ALS/MND. The authors of Jeong 2015 concluded that stem cell therapy in people with ALS/MND can slow disease progression compared to natural history, based on single‐arm clinical studies.

Taken together, based on evidence from non‐RCTs, the previous systematic reviews concluded that the effectiveness of stem cell therapy for people with ALS/MND has not been established and needs re‐evaluation.

The findings of the present systematic review appear in line with two preliminary studies that evaluated safety and feasibility of cell transplantation procedures into the brain, spinal cord and thecal sac of people with ALS/MND (Goutman 2015; Lunn 2014). These single‐arm, small clinical phase 1/2 trials showed that, in general, direct cell implantation into the cerebral cortex, spinal cord and thecal sac of people with ALS/MND appeared feasible and was not associated with an acceleration in disease progression, although there are great uncertainties in the findings, due to the non‐randomised and preliminary nature of the trials.

Studies have evaluated a variety of cell‐based products designed to either replace the lost motor neurons or improve the metabolic supply of the affected neurons, thus delaying neuronal death. MSCs, obtained mainly from autologous bone marrow, are the most frequently used cell type in clinical trials because they are readily available in large numbers, release growth factors and are non‐immunogenic. In people with ALS/MND, no serious adverse effects and no detrimental effects on neurological function occurred following intraspinal transplantation of variable doses of MSCs (7 × 10⁶ to 152 × 10⁶ cells). Some trials reported improvement in motor function in people with ALS/MND (Karussis 2010; Oh 2015; Petrou 2015).

Conflicting findings exist regarding the effect of OESC transplantation in people with ALS/MND. Two non‐randomised clinical studies reported favourable outcomes: Huang 2008 reported a significantly greater reduction in functional deterioration three to four months post OESC transplantation (15 participants) than in an untreated control group (20 participants) and a single‐arm trial (42 participants) reported improvement in neurological and lung function after repeated cell administration (Chen 2007). However, other reports did not support these clinical observations (Chew 2007; Giordana 2010; Piepers 2010). Giordana 2010 reported postmortem findings from two people with ALS/MND who received intracranial injection of foetal OESCs. OESC transplantation did not modify the neuropathology of ALS/MND and there was an absence of axonal regeneration, neuronal differentiation and myelination.

Taken together, the findings in this review and previous studies suggest that the potential use of autologous BM‐MSC to may slightly reduce the decline in motor function in people with ALS/MND (Karussis 2010; Oh 2015; Petrou 2015).

Apart from riluzole and non‐invasive ventilation, current management strategies have shown very limited effects on survival in ALS/MND (Miller 2012; Radunovic 2013). The only RCT presented in our review found no survival benefit of BM‐MSC in people with ALS/MND at six months (Oh 2018). This may be because of the relative short follow‐up period in Oh 2018. However, a long‐term post hoc analysis of survival in Oh 2018 found no statistically significant difference in mean survival time in the MSC group (55 (SE 4) months) compared to the control group (48 (SE 6) months) (log‐rank test for survival P = 0.487). In contrast, two previous non‐RCTs showed that median survival in the cell‐based therapy group was significantly longer than in the control group (66 months in the CD133+ cell group versus 19 months in the control group; P = 0.0111, 10 participants; 87.76 (standard deviation (SD) 10.45) months in the BM‐MNC group versus 57.38 (SD 5.31) months in the control group) (Martinez 2009; Sharma 2015). A subgroup analysis of the intervention arm revealed that younger age at disease onset (less than 50 years), limb onset (compared to bulbar onset), and concurrent lithium therapy were associated with longer survival. The impact of cell‐based therapy on survival of people with ALS/MND, in particular the effect of the different types of cell‐based products and transplantation regimens needs to be examined in future RCTs.

Four ongoing RCTs are addressing some of these key issues related to cell treatment protocols in people with ALS/MND. These trials are expected to be completed between 2019 and 2021. NCT01254539 is a randomised, double‐blind trial to assess the feasibility and the security of the intraspinal and intrathecal infusion of autologous bone marrow stem cells. NCT02286011 is a phase 1 randomised, double‐blind, controlled clinical trial on intramuscular infusion of autologous BM‐MNC. NCT02290886 is a phase 1/2 randomised, placebo‐controlled, triple‐blind trial to evaluate the safety and efficacy of intravenous autologous adipose tissue‐derived MSCs in three different doses (one million, two million, and four million autologous MSCs). Finally, NCT03280056 is a phase 3, randomized, double‐blind, multicentre trial to evaluate the safety and efficacy of three intrathecal administrations of NurOwn (MSC‐NTF cells) at a bi‐monthly interval, compared to placebo.

Authors' conclusions

Implications for practice.

Limited evidence of very low‐to‐low certainty from a single randomised controlled trial shows that autologous bone marrow‐mesenchymal stem cells treatment may reduce the decline in disability at six months in people with amyotrophic lateral sclerosis/motor neuron disease (ALS/MND), but the available evidence is not sufficient to enable a clear conclusion regarding the therapeutic efficacy and safety of cell‐based therapy in ALS/MND.

Implications for research.

To date, there is no conclusive evidence that cell‐based therapy alters the natural course of ALS/MND and prolongs survival, with currently available evidence limited by imprecision as a result of small sample size. There were significant shortcomings related to the design of the available published studies. We found significant variability between trials with regards to selection criteria, outcome measures, and types of cells and methods of cell implantation. Moreover, these trials were generally underpowered to show any clinical benefit. Prospective RCTs with larger sample size and longer‐term follow‐up are urgently required to assess the clinical benefits of cell‐based therapy including improvement in disease progression and quality of life and prolongation of survival in people with ALS/MND. Importantly, data from well‐designed trials might determine patient‐, disease‐ and cell treatment‐related factors that could potentially influence the clinical outcomes of cell‐based therapy.

Questions remain as to the optimal cell source, phenotype and dose, as well as transplantation method and protocol that would be key elements in designing an optimal cell‐based therapy programme for people with ALS/MND; these should be the major goals of future research.

Combination of cellular therapy with standard therapy (riluzole) or novel neuroprotective agents should also be explored to strengthen the therapeutic efficacy and to find the best possible approach to prevent or reverse the neurological deficit and to prolong survival of this otherwise debilitating and fatal condition.

Studies to investigate the mechanisms of cellular neuroprotection induced by cell implantation would be vital in developing an effective cell‐based targeted therapy in ALS/MND. Future clinical trials should consider the following factors.

1. Improved trial designs and participant selection criteria, with relevant clinical outcomes.

2. Long‐term follow‐up to establish long‐term safety and durability of the clinical benefit of cell‐based therapy.

3. Standardisation of cell products used and cell implantation protocols.

4. Detailed characterisation of cells used for implantation, including viability and immunophenotype.

5. In vivo cell tracking using advanced imaging technologies to provide insight into the survival and migratory potential of the grafted cells as well as safety issues such as aberrant differentiation and migration.

6. Postmortem pathological analysis of brain or spinal cord specimens from people with ALS/MND, to detect evidence of neuroprotection or axonal regeneration of the diseased motor neurons and evidence of aberrant differentiation or tumour formation.

7. Use of novel cellular sources and treatment approaches (such as induced pluripotent stem cells, cell lines expressing neutrotrophic growth factors and combining mesenchymal stem cell and neural stem cell transplantation).

What's new

Date	Event	Description
27 November 2018	New citation required and conclusions have changed	No studies were eligible for inclusion in the previous review; the review now includes two randomised controlled trials (RCTs).
15 November 2018	New search has been performed	Two RCTs involving 112 participants were eligible for inclusion in this review. We reported the results of newly included studies, assessed risk of bias and graded the certainty of the evidence. We also: listed the sources that we handsearched; added another scale as an example of functional rating scale for amyotrophic lateral sclerosis and added the reported minimally important difference in Amyotrophic Lateral Sclerosis Functional Rating Scale.

Appendices

Appendix 1. Cochrane Neuromuscular Specialised Register via the Cochrane Register of Studies (CRS‐Web) search strategy

Search date 31 July 2019

#1 MeSH DESCRIPTOR Motor Neuron Disease Explode All AND INREGISTER
 #2 "motor neuron disease*" or "motor neurone disease*" AND INREGISTER
 #3 "motorneuron disease*" or "motorneurone disease*" AND INREGISTER
 #4 "motoneuron disease*" or "motoneurone disease*" AND INREGISTER
 #5 "charcot disease" AND INREGISTER
 #6 "amyotrophic lateral sclerosis" AND INREGISTER
 #7 als:ti or als:ab or nmd:ti or mnd:ab AND INREGISTER
 #8 #1 or #2 or #3 or #4 or #5 or #6 or #7 AND INREGISTER
 #9 mononuclear NEAR2 leukocyte* AND INREGISTER
 #10 MeSH DESCRIPTOR Stem Cells Explode All AND INREGISTER
 #11 "stem cell*" AND INREGISTER
 #12 MeSH DESCRIPTOR Stem Cell Transplantation Explode All AND INREGISTER
 #13 "bone marrow" AND INREGISTER
 #14 mesenchymal NEAR cell* AND INREGISTER
 #15 mononuclear NEAR cell* AND INREGISTER
 #16 angiogenesis NEAR therap* AND INREGISTER
 #17 #9 or #10 or #11 or #12 or #13 or #14 or #15 AND INREGISTER
 #18 #8 and #17 AND INREGISTER

Appendix 2. Cochrane Central Register of Controlled Trials (CENTRAL) via the Cochrane Register of Studies (CRS‐Web) search strategy

Search date 31 July 2019

#1 MeSH DESCRIPTOR Motor Neuron Disease Explode All AND CENTRAL:TARGET
 #2 "motor neuron disease*" or "motor neurone disease*" AND CENTRAL:TARGET
 #3 "motorneuron disease*" or "motorneurone disease*" AND CENTRAL:TARGET
 #4 "motoneuron disease*" or "motoneurone disease*" AND CENTRAL:TARGET
 #5 "charcot disease" AND CENTRAL:TARGET
 #6 "amyotrophic lateral sclerosis" AND CENTRAL:TARGET
 #7 als:ti or als:ab or nmd:ti or mnd:ab AND CENTRAL:TARGET
 #8 #1 or #2 or #3 or #4 or #5 or #6 or #7 AND CENTRAL:TARGET
 #9 mononuclear NEAR2 leukocyte* AND CENTRAL:TARGET
 #10 MeSH DESCRIPTOR Stem Cells Explode All AND CENTRAL:TARGET
 #11 "stem cell*" AND CENTRAL:TARGET
 #12 MeSH DESCRIPTOR Stem Cell Transplantation Explode All AND CENTRAL:TARGET
 #13 "bone marrow" AND CENTRAL:TARGET
 #14 mesenchymal NEAR cell* AND CENTRAL:TARGET
 #15 mononuclear NEAR cell* AND CENTRAL:TARGET
 #16 angiogenesis NEAR therap* AND CENTRAL:TARGET
 #17 #9 or #10 or #11 or #12 or #13 or #14 or #15 AND CENTRAL:TARGET
 #18 #8 and #17 AND CENTRAL:TARGET

Appendix 3. MEDLINE (OvidSP) search strategy

Database: Ovid MEDLINE(R) and Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations and Daily <1946 to July 30, 2019>
 Search strategy:
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
 1 randomized controlled trial.pt. (486353)
 2 controlled clinical trial.pt. (93185)
 3 randomi#ed.tw. (580003)
 4 placebo.ab. (199708)
 5 drug therapy.fs. (2127161)
 6 randomly.ab. (315646)
 7 trial.ab. (472150)
 8 groups.ab. (1939321)
 9 or/1‐8 (4518366)
 10 exp animals/ not humans.sh. (4603844)
 11 9 not 10 (3913504)
 12 exp Motor Neuron Disease/ (26171)
 13 (moto$1 neuron$1 disease$1 or moto?neuron$1 disease).mp. (8621)
 14 ((Lou Gehrig$1 adj5 syndrome$1) or (Lou Gehrig$1 adj5 disease)).mp. (200)
 15 charcot disease.tw. (24)
 16 Amyotrophic Lateral Sclerosis.mp. (25602)
 17 or/12‐16 (35296)
 18 Leukocytes, Mononuclear/ (35660)
 19 Mesenchymal Stromal Cells/ (33145)
 20 Bone Marrow Transplantation/ (44128)
 21 exp stem cells/ (201255)
 22 exp Stem Cell Transplantation/ (76955)
 23 (mononuclear adj5 cell$1).tw. (81721)
 24 mesenchymal stem cell$1.tw. (38264)
 25 (angiogenesis adj3 therap$).tw. (3187)
 26 bone marrow.tw. (204618)
 27 stem cells.tw. (155001)
 28 or/18‐27 (548288)
 29 11 and 17 and 28 (159)
 30 remove duplicates from 29 (159)

Appendix 4. Embase (OvidSP) search strategy

Database: Embase <1974 to 2019 July 29>
 Search strategy:
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
 1 crossover‐procedure.sh. (60061)
 2 double‐blind procedure.sh. (163402)
 3 single‐blind procedure.sh. (36004)
 4 randomized controlled trial.sh. (561727)
 5 (random$ or crossover$ or cross over$ or placebo$ or (doubl$ adj blind$) or allocat$).tw,ot. (1668280)
 6 trial.ti. (276639)
 7 controlled clinical trial/ (464227)
 8 or/1‐7 (1985978)
 9 exp animal/ or exp invertebrate/ or animal.hw. or non human/ or nonhuman/ (26239888)
 10 human/ or human cell/ or human tissue/ or normal human/ (20033631)
 11 9 not 10 (6268194)
 12 8 not 11 (1764798)
 13 limit 12 to (conference abstracts or embase) (1493112)
 14 motor neuron disease/ or amyotrophic lateral sclerosis/ (42033)
 15 (moto$1 neuron$1 disease$1 or moto?neuron$1 disease$1).mp. (13637)
 16 ((Lou Gehrig$1 adj5 syndrome$1) or (Lou Gehrig$1 adj5 disease)).mp. (227)
 17 charcot disease.tw. (28)
 18 amyotrophic lateral sclerosis.tw. (28745)
 19 or/14‐18 (46810)
 20 mononuclear cell/ (44111)
 21 exp mesenchyme cell/ (78826)
 22 exp bone marrow transplantation/ (63495)
 23 exp stem cell/ (347407)
 24 exp stem cell transplantation/ (141205)
 25 (mononuclear adj5 cell$1).tw. (108990)
 26 mesenchymal stem cell$1.tw. (56351)
 27 (angiogenesis adj3 therap$).tw. (4368)
 28 bone marrow.tw. (292794)
 29 stem cell$1.tw. (360632)
 30 or/20‐29 (838941)
 31 13 and 19 and 30 (93)
 32 remove duplicates from 31 (93)

Appendix 5. ClinicalTrials.gov search strategy

ClinicalTrials.gov basic search

(motor neuron disease OR amyotrophic lateral sclerosis) AND (cell based OR leukocytes OR mesenchymal OR mesenchym OR mononuclear OR bone marrow OR stem cell OR angiogenesis)

Appendix 6. WHO ICTRP search strategy

WHO ICTRP advanced search

motor neuron disease AND cell‐based OR motor neuron disease AND leukocytes OR motor neuron disease AND mesenchymal OR motor neuron disease AND mesenchym OR motor neuron disease AND mononuclear OR motor neuron disease AND bone marrow OR motor neuron disease AND stem cell OR motor neuron disease AND angiogenesis OR amyotrophic lateral sclerosis AND cell‐based OR amyotrophic lateral sclerosis AND leukocytes OR amyotrophic lateral sclerosis AND mesenchymal OR amyotrophic lateral sclerosis AND mesenchym OR amyotrophic lateral sclerosis AND mononuclear OR amyotrophic lateral sclerosis AND bone marrow OR amyotrophic lateral sclerosis AND stem cell OR amyotrophic lateral sclerosis AND angiogenesis.

Data and analyses

Comparison 1. Autologous bone marrow‐mesenchymal stem cells (BM‐MSC) versus no treatment.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Functional impairment: change in Amyotrophic Lateral Sclerosis Functional Rating Scale – Revised score	1	56	Mean Difference (IV, Fixed, 95% CI)	3.38 [1.22, 5.54]
2 Respiratory function: change in % forced vital capacity	1	56	Mean Difference (IV, Fixed, 95% CI)	‐0.53 [‐5.37, 4.31]
3 Quality of life: change in 36‐item Short Form score	1	57	Mean Difference (IV, Fixed, 95% CI)	2.77 [‐3.50, 9.04]
4 Survival at 6 months	1	64	Risk Ratio (M‐H, Fixed, 95% CI)	1.07 [0.94, 1.22]
5 Adverse events, total	1	64	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.62, 1.19]
6 Serious adverse events	1	64	Risk Ratio (M‐H, Fixed, 95% CI)	0.47 [0.13, 1.72]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Gothelf 2017.

Methods	Study design: phase 2, randomised, double‐blind, placebo‐controlled, multicentre study
Participants	Country: US Number of participants: 48 Mean age: 51.1 years Sex: 35 men, 13 women Inclusion criteria Men and women aged 18–75 years, inclusive ALS/MND diagnosed as possible, laboratory‐supported probable, probable or definite, as defined by revised El Escorial criteria Disease onset, as defined by first reported occurrence of symptomatic weakness, spasticity or bulbar symptoms, of ˃ 12 months and ≤ 24 months Current disease symptoms must have included limb weakness ALSFRS‐R ≥ 30 at the screening visit Upright slow vital capacity measure ≥ 65% of predicted for gender, height and age at the screening visit Participants must have been taking a stable dose of riluzole for at least 30 days prior to enrolment or not be taking riluzole, and not have been on it for ≥ 30 days prior to enrolment (riluzole‐naive participants permitted in the study) Capable of providing informed consent and willing and able to follow study procedures, including willingness to undergo lumbar puncture Geographic accessibility to the study site and willingness and ability to comply with follow‐up Women of child‐bearing potential must have agreed not to become pregnant for the duration of the study. Women must have been willing to consistently use 2 forms of contraceptive therapy throughout the course of the trial, and undergo a pregnancy test 1 week before BM aspiration. Men must have been willing to consistently use 2 forms of contraceptive if their partners were of child‐bearing age Citizen or permanent resident of the US Exclusion criteria Prior stem cell therapy of any type Inability to lie flat for the duration of intrathecal cell transplantation or BM biopsy, or inability to tolerate study procedures for any other reason History of autoimmune disease (excluding thyroid disease), myelodysplastic or myeloproliferative disorder, leukaemia or lymphoma, whole body irradiation, hip fracture or severe scoliosis Any unstable clinically significant medical condition other than ALS/MND (e.g. myocardial infarction, angina pectoris or congestive heart failure within 6 months of baseline), treatment with anticoagulants that, in the opinion of the investigator, would compromise the safety of participants Any history of malignancy including any malignancy affecting the central nervous system and melanoma, within the previous 5 years, with the exception of localised skin cancers (with no evidence of metastasis, significant invasion or re‐occurrence within 3 years of baseline) Serum AST or ALT values > 3 times the upper normal limit Serum creatinine value > 2 times the upper normal limit Positive test for HBV, HCV or HIV Current use of immunosuppressant medication or use of such medication within 4 weeks of screening visit History of acquired or inherited immune deficiency syndrome Exposure to any other experimental agent (off‐label use or investigational), or participation in a clinical trial within 30 days prior to screening visit Use of non‐invasive ventilation, diaphragm pacing system or invasive ventilation (tracheostomy) History of either substance abuse within the past year or unstable psychiatric disease according to investigators' judgement Placement or usage of feeding tube Pregnant or currently breastfeeding
Interventions	Autologous MSC‐NTF cells; cell dose not mentioned Placebo
Outcomes	Primary outcome Number of participants with adverse events Secondary outcomes Change in ALSFRS slopes from the pretransplantation period to the post‐transplantation period between the treatment and placebo groups for 24 weeks after transplantation Change in slow vital capacity slopes from the pretransplantation period to the post‐transplantation period between the treatment and placebo groups for 24 weeks after transplantation
Funding	Funded Brainstorm‐Cell Therapeutics
Conflicts of interest	Not clearly stated.
Notes	Information obtained from a combination of 2 oral presentation abstracts and a trial registry record. Repeated attempts to contact study authors for outcome data were unsuccessful.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "This Phase 2 multicenter double blind, placebo‐controlled trial enrolled 48 participants with ALS randomized 3:1 (active:placebo)." Methods of sequence generation not stated.
Allocation concealment (selection bias)	Unclear risk	No information in the abstracts and trial records to enable a meaningful assessment of the independence of sequence generation and allocation.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "double blind, placebo‐controlled." However, there was no detailed description of the blinding mechanism.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	It was unclear whether the outcome assessor was blinded for the major outcomes, such as adverse effects and improvement in ALSFRS‐R slope of decline in function.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	The authors stated that 48 participants were recruited, but it was unclear how many participants were included in the analysis, as the results were presented only in percentages.
Selective reporting (reporting bias)	High risk	It appeared that the major outcomes appropriate for a phase 2 trial, namely adverse effects and clinical improvements, were reported. However, the reports provided insufficient detail, as the authors only reported the results in terms of percentages and P values, and did not provide the number of participants analysed and the absolute number of participants with events.
Other bias	Unclear risk	The oral presentation abstracts and the trial records provided insufficient detail for assessment of other biases.

Oh 2018.

Methods	Study design: parallel‐group, randomised, controlled phase 2 trial
Participants	Country: Republic of Korea Number of participants: 64 Mean age: 53.3 years Sex: 33 men, 31 women Study dates: enrolled between December 2011 and November 2012 and followed up until July 2013 Inclusion criteria People aged 25–75 years diagnosed with clinically probable or definite ALS/MND according to the revised El Escorial criteria. ALSFRS‐R score 31–46 Stable riluzole treatment (50 mg, twice daily) for at least 3 months before screening Disease duration < 5 years after the onset of first symptoms Exclusion criteria Participation in other clinical trials within the past 12 months FVC < 40% of the predicted value Presence of any comorbidity that might interfere with the outcome Tracheostomy or non‐invasive ventilation Any haemorrhagic tendency Administration of any drug that could affect the BM Follow‐up period: 6 months
Interventions	Intrathecal BM‐MSCs plus riluzole (33 participants) Riluzole alone (control group; 31 participants)
Outcomes	Primary outcomes Adverse effects, any or serious from baseline (0 month) to month 6 Difference in changes of ALSFRS‐R from baseline (0 month) to months 4 and 6 Secondary outcomes Change in Appel ALS Rating Scale from baseline (0 month (visit 5) to month 4 (visit 9) Change in FVC (percent of predicted normal) from baseline (0 month (visit 5) to month 4 (visit 9) Change in SF‐36 from baseline (0 month (visit 5) to month 4 (visit 9)
Funding	Ministry for Health & Welfare Affairs, Republic of Korea (HI10C1673 and HI15C0876), and Corestem Biotechnology, Gyeonggi‐do, Republic of Korea
Conflicts of interest	SHK received funding for postmarketing survey of Autologous Bone Marrow‐Derived Mesenchymal Stem Cells (HYNR‐CS) from Corestem Biotechnology according to the safety guideline of Korean Ministry of Food and Drug Safety (KMFDS) after conditional approval of HYNR‐CS as an orphan drug from KMFDS. The remaining authors had nothing to report
Notes	In the control group, sham procedures related to stem cell therapy, including BM aspiration, CSF collection for BM‐MSCs suspension and lumbar puncture were not performed due to ethical considerations. ClinicalTrials.gov ID: NCT01363401
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	All eligible participants were randomised (1:1) into 2 groups using an interactive web response system.
Allocation concealment (selection bias)	Low risk	Interactive web response system was used.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants were not blinded to study group assignment. There was no sham procedure (BM aspiration, CSF collection for BM‐MSCs suspension or lumbar puncture) performed in the control group, due to ethical considerations.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Neurologists and evaluators were blinded to treatment assignments.
Incomplete outcome data (attrition bias) All outcomes	High risk	4 participants in the control group (withdrawal: 2; death: 1; serious adverse event: 1) and 1 participant in the MSC group (withdrawal) were excluded from the full analysis set because these events occurred before visit 5 (baseline). The disproportionate number of dropouts between the control and treatment groups could affect the balance of confounders between the study groups and, therefore, we judged this study at high risk for attrition bias.
Selective reporting (reporting bias)	Low risk	All major outcomes appropriate for the phase 2 trial were reported in sufficient detail.
Other bias	Low risk	We identified no other potential sources of bias.

ALS: amyotrophic lateral sclerosis; ALSFRS‐R: Amyotrophic Lateral Sclerosis Functional Rating Scale – Revised; AST: aspartate aminotransferase; ALT: alanine aminotransferase; BM: bone marrow; BM‐MSC: bone marrow‐derived mesenchymal stem cells; CSF: cerebrospinal fluid; FVC: forced vital capacity; HBV: hepatitis B virus; HCV: hepatitis C virus; MSC: mesenchymal stem cell; MSC‐NTF: mesenchymal stem cells secreting neurotrophic factors; MND: motor neuron disease; SF‐36: 36‐item Short Form.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Gabr 2017	Not an RCT. Evaluation of the safety and efficacy of bone marrow‐derived MSCs injected intrathecally in restoring damaged motor neurons.
Karussis 2013	Not an RCT. A phase 1/2 clinical trial to evaluate the safety and tolerability of intramuscular and intrathecal treatment with autologous MSCs differentiated to secrete neurotrophic factors in people with ALS/MND.
Mazzini 2015	Not an RCT. Phase 1 clinical trial using foetal human neural stem cells from natural in utero death administered into the anterior horns of the spinal cord to test for the safety of both cells and neurosurgical procedures in people with ALS/MND.
Nabavi 2019	Not an RCT. To assess the safety and feasibility of intravenous and intrathecal injections of bone marrow derived mesenchymal stromal cells in people with ALS/MND.
Petrou 2015	Not an RCT. To evaluate the safety and efficacy of transplantation of autologous bone marrow‐derived MSCs induced to secrete neurotrophic factors in people with ALS/MND.

ALS/MND: amyotrophic lateral sclerosis/motor neuron disease; MSC: mesenchymal stem cell; RCT: randomised controlled trial.

Characteristics of ongoing studies [ordered by study ID]

NCT01254539.

Trial name or title	Clinical trial on the use of autologous bone marrow stem cells in amyotrophic lateral sclerosis (Extension CMN/ELA)
Methods	Randomised, double‐blind study
Participants	Inclusion criteria Diagnosis established using the World Federation of Neurology criteria > 6 and < 36 months of evolution of the disease Bulbar onset Aged > 18 and < 70 years FVC ≥ 50% Total time of oxygen saturation < 90% inferior to 5% of the sleeping time Signed informed consent Exclusion criteria Neurological or psychiatric disease Need of parenteral or enteral nutrition through percutaneous endoscopic gastrostomy or nasogastric tube Concomitant systemic disease Treatment with corticosteroids, immunoglobulins or immunosuppressors in the last 12 months Inclusion in other clinical trials Inability to understand informed consent
Interventions	Autologous bone marrow stem cells intraspinal transplantation at T3‐T4 Intrathecal infusion of autologous bone marrow stem cells Intrathecal infusion of placebo (saline solution) Cell dose not mentioned
Outcomes	Primary outcome FVC Secondary outcomes Neurological variables: ALSFRS, MRC and Norris scales Absence of adverse events Neurophysiological variables: electromyography, polysomnography, evoked potentials Neuroradiological variables: spinal magnetic resonance imaging Respiratory variables: maximal inspiratory pressure, maximal expiratory pressure, sniff nasal, oxymetry Psychological variables: Health Questionnaire (EuroQol‐5D), Profile of Mood States
Starting date	October 2010
Contact information	Jose María Moraleda Jiménez, MD, PhD; E‐mail: paqui iniesta martínez‐iniesmar@yahoo.es
Notes	Recruitment status completed. Contacted authors. Not yet finished analysis of the results.

NCT02286011.

Trial name or title	Intramuscular infusion of autologous bone marrow stem cells in people with amyotrophic lateral sclerosis (TCIM/ELA)
Methods	Prospective, randomised, double‐blind study
Participants	Inclusion criteria Diagnosis of definite or probable ALS/MND according to the criteria established by the World Federation of Neurology Reasonable assurance of adherence to protocol Neurophysiological data confirming lower motor neuron lesions in the lumbar region Assessment of motor deficits in dorsiflexion of both feet (4 or 5 points on the MRC scale; MRC scale grade muscle power as 0 = no contraction, 1 = flicker of contraction, 2 = active movement, with gravity eliminated, 3 = active movement against gravity, 4 = active movement against gravity and resistance and 5 = normal power) Participant must fulfil all inclusion criteria Exclusion criteria Diabetes mellitus Other diseases that may present with polyneuropathy Previous history of stroke Prior pathology of the peripheral nervous system affecting 1 or both lower limbs with or without clinically evident neurological sequelae Pregnant or breastfeeding women Women physiologically capable of becoming pregnant, unless they are using reliable contraception People with cardiac, renal, hepatic, systemic or immune conditions that could influence survival during the test Positive serology for hepatitis B, hepatitis C or HIV Clinical and anaesthesiological criteria contraindicating either sedation or extraction of bone marrow (an altered coagulation system or anticoagulated with inability to withdraw anticoagulation, haemodynamic instability, altered skin puncture site, etc.) Included in other clinical trials in the last 6 months
Interventions	Intramuscular infusion of autologous BM‐MNC (550 million cells (100–1200 million) diluted in 2 mL saline) in tibialis anterior muscle of 1 of the lower limbs Intramuscular infusion of placebo (2 mL saline) in the contralateral lower limb [sic] (control)
Outcomes	Primary outcome Rate of serious and non‐serious adverse events related to cellular therapy in participants with ALS/MND Secondary outcomes Estimated number of motor units Compound muscle action potential Fibre density Muscle strength: MRC score Maximum force developed in an isometric contraction of the tibialis anterior muscle Maximum transversal area of the tibialis anterior muscle
Starting date	November 2014 (duration: 38 months)
Contact information	Natalia García Iniesta +34968381221; E‐mail: nagarini@yahoo.es
Notes

NCT02290886.

Trial name or title	Clinical trial phase I/II, randomised, controlled with placebo, triple blind to evaluate safety, and indications of efficiency of the intravenous administration of the therapy with 3 doses of MSC in participants with moderate to severe ALS
Methods	Phase 1/2, multicentre, randomised, placebo‐controlled, triple‐blind study
Participants	Inclusion criteria Men and women aged > 18 years Good understanding of the protocol and able to grant informed consent Definite or probable diagnosis of sporadic ALS/MND in agreement with the criteria of El Escorial criteria, of the World Federation of Neurology FVC ≥ 50% of normal for sex, height and age Disease onset (beginning of symptoms) > 6 months and < 36 months previously Possibility of obtaining ≥ 50 g of adipose tissue Treatment with riluzole for ≥ 1 month before inclusion Exclusion criteria Any concomitant disease that under investigator's criteria could affect measures of the clinical variables (hepatic, renal or cardiac insufficiency, diabetes mellitus, etc.) Previous therapy with stem cells Participation in another clinical trial within 3 months prior to entry in this trial Any lymphoproliferative disease Tracheostomy, gastrostomy or both Haemophilia, haemorrhagic diathesis or current anticoagulation therapy Specific hypersensitivities "Medical precedents" of HIV infection or any serious immunocompromised condition Positive HBV or HCV serology Creatinine level (see www.ClinicalTrials.gov)
Interventions	Intravenous 1 million autologous MSC Intravenous 2 million autologous MSC Intravenous 4 million autologous MSC Intravenous placebo (control)
Outcomes	Primary outcomes Number of serious unexpected adverse reactions attributable to the treatment Infusion site complications Appearance of a new neurological sign or symptom not attributable to the natural progression of ALS/MND Secondary outcomes Change in disease progression Change in muscular force Change in FVC Change in muscular mass estimated by nuclear magnetic resonance imaging of the upper and low extremities Change in neurophysiological parameters Change in quality of life Need for and time to tracheotomy or permanent assisted ventilation
Starting date	July 2014 (duration: 62 months)
Contact information	Fernández O; E‐mail: oscar.fernandez.sspa@juntadeandalucia.es
Notes

NCT03280056.

Trial name or title	Safety and efficacy of repeated administrations of NurOwn® in ALS patients
Methods	Randomized, double‐blind, placebo‐controlled multicentre study
Participants	Inclusion criteria ALS/MND diagnosed as possible, laboratory‐supported probable, probable or definite as defined by revised El Escorial criteria. Having onset of ALS/MND disease symptoms, including limb weakness within 24 months at the screening visit. ALSFRS‐R ≥ 25 at the screening visit. Upright slow vital capacity measure ≥ 65% of predicted for gender, height and age at the screening visit Rapid progressors Participants taking a stable dose of riluzole are permitted Citizen or permanent resident of the US or Canadian citizen able to travel to a US site for all follow‐up study visits Exclusion criteria Prior stem cell therapy of any type History of autoimmune or other serious disease (including malignancy and immune deficiency) that may confound study results Current use of immunosuppressant medication or anticoagulants (per Investigator discretion) Exposure to any other experimental agent or participation in an ALS/MND clinical trial within 30 days prior to screening visit Use of RADICAVA (edaravone injection) within 30 days of screening or intent to use edaravone at any time during the course of the study including the follow‐up period Use of non‐invasive ventilation (bilevel positive airway pressure), diaphragm pacing system or invasive ventilation (tracheostomy) Feeding tube Pregnant women or women currently breastfeeding
Interventions	Active comparator: NurOwn (MSC‐NTF cells): 3 intrathecal administrations of NurOwn (MSC‐NTF cells) at bi‐monthly intervals
Outcomes	Primary outcomes ALSFRS‐R (28 weeks following the first treatment) Safety (not further specified) Secondary outcomes Biomarkers (such as cell‐secreted neurotrophic factors, inflammatory factors and cytokines in pg/mL) in the cerebrospinal fluid as well as in serum samples (time frame: through selected post‐treatment time points up to 20 weeks after transplant)
Starting date	August 2017
Contact information	Ralph Z Kern; E‐mail: rkern@brainstom‐call.com Yael D Gothelf; E‐mail: ygothelf@brainstom‐call.com
Notes

ALS/MND: amyotrophic lateral sclerosis/motor neuron disease; ALSFRS: Amyotrophic Lateral Sclerosis Functional Rating Scale; ALSFRS‐R: Amyotrophic Lateral Sclerosis Functional Rating Scale – Revised; ALT: alanine transaminase; AST: aspartate transaminase; BM‐MNC: bone marrow mononuclear cells; FVC: forced vital capacity; HBV: hepatitis B virus; HCV: hepatitis C virus; MRC: Medical Research Council; MSC: mesenchymal stem cells; MSC‐NTF: mesenchymal stem cells secreting neurotrophic factors.

Differences between protocol and review

1. In our protocol (Abdul Wahid 2015), we did not specifically state the other sources that we handsearched. In the current review, we handsearched publications from the following journals: Cytotherapy, Cell Transplantation, Cell Stem Cell and Stem Cells for relevant articles. We searched the Centre for Reviews and Dissemination (CRD) Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment (HTA) and National Health Service Economic Evaluation Database (NHSEED) for the original review in 2015 with no relevant papers found. We did not search them for this update as they are no longer included in the Cochrane Library.

2. We made the following changes to the outcomes for the first version of the review, which had no included studies.

   1. The primary outcome originally defined in the protocol was 'Change in Expanded Disability Status Scale (EDSS) or Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS) at 6 and 12 months.' In this update, we moved the 12‐month ALSFRS score to secondary outcomes and removed the change in EDSS outcome.

   2. We specified that the FVC to be reported was upright FVC.

   3. We changed the outcome 'Change in compound muscle action potential (CMAP) and neurophysiological index (NI) at 6 and 12 months' to 'Change in compound muscle action potential (CMAP), neurophysiological index (NI), combined motor index (CMI), motor unit number estimation (MUNE), and motor unit number index (MUNIX) at 6 and 12 months.

   4. We added examples of acceptable quality of life scales (such as ALS Assessment Questionnaires, ALSAQ‐40 or ALSQ5, Short‐Form 36 (SF‐36) Health Survey and EQ‐5D) at 6 and 12 months (Jenkinson 2000; Jenkinson 2007; Rabin 2001; Ware 1992).

   5. We added detail to the specification of our adverse events outcome: "Adverse events include an inflammatory reaction at the cell injection site, and cardiovascular and thromboembolic complications. We would have reported the rate of adverse events and the rate of withdrawal from the study." In the present review, we also reported serious adverse events.

3. We removed "Repeat the analysis excluding large studies to assess the effect of these studies on the overall results" from our planned sensitivity analyses on the advice of the Cochrane Neuromuscular statistician.

4. In the protocol, we specified that the outcomes for inclusion in the 'Summary of findings' table were: EDSS and ALSFRS scales, manual muscle testing, FVC, survival rate, and adverse events at 12 months. As 12‐month data were not available, we reported the longest time points from the included study, which was four or six months, depending on the outcome.

Contributions of authors

SFAW, ZKL, NML and NAI wrote the review and approved the review in its final form.

SFAW designed the project and review.

SFAW, ZKL and NAI drafted the search strategy.

SFAW, ZKL and NAI performed the study screening and selection.

NML and NAI reviewed the analysis of the search

Sources of support

Internal sources

• No sources of support supplied

External sources

• No sources of support supplied, Malaysia.

Declarations of interest

SFAW is a consultant haematologist and professor of internal medicine and performed haematopoietic stem cell transplantation for people with blood and bone marrow diseases. She has been actively involved in many clinical trials utilising cell and cell‐based products for blood cancers and various degenerative conditions. No known conflicts of interest.

ZKL is a neurologist and manages patients with a range of neurological diseases including ALS/MND. No known conflicts of interest.

NAI is a research officer and a clinical study co‐ordinator involved in production of clinical grade cell and cell‐based products and managing clinical trials. No known conflict of interest.

NML is a Paediatrician and Neonatologist by training, and a Cochrane Trainer. He has no known conflict of interest.

New search for studies and content updated (conclusions changed)