Multiple Sclerosis in Children: Current and Emerging Concepts

Abstract

Multiple sclerosis is being increasingly recognized and diagnosed in children. In the past several years, advances have been made in diagnosing multiple sclerosis in children, identifying new genetic and environmental risk factors, delineating underlying immunobiology, characterizing imaging findings, and implementing new treatment strategies. In this review, we discuss these advances. Future research into the determinants of multiple sclerosis in children and into new treatment options will be aided by continued international collaboration.

Multiple sclerosis (MS) is a chronic, immune-mediated, neurodegenerative disorder of the central nervous system (CNS). While the clinical symptoms of MS most commonly manifest between 20 and 40 years of age, approximately 3 to 10% of all MS patients report that their first clinical symptoms occurred in childhood or adolescence.^1,2 Most children with MS have a relapsing-remitting course that is highly active. As compared with adults with MS, rates of relapse are two to three times higher in children, and children have greater lesion volumes on MRI, especially in the posterior fossa.^3–5

The diagnosis of MS is based on the clinical history, neurologic examination, supportive MRI findings, and ancillary laboratory data, especially cerebrospinal fluid (CSF) testing. An early, yet accurate, diagnosis of MS in children is important for timely initiation of treatment. With the first Food and Drug Administration (FDA) approval for a disease-modifying therapy (DMT) for MS in children (and several other MS therapies currently in the clinical trials phase), the treatment landscape for children with MS has never been more promising. This review will present an update on the risk factors and clinical features of MS in children, the most recent diagnostic criteria, advances in neuroimaging, and current approaches to the treatment of children with MS.

Epidemiology and Risk Factors

Studies have reported an incidence of MS in children between 0.26 and 2.1 per 100,000 children, depending on the geographic location.^6–11 Some of these studies have reported a rising incidence of MS in children; however, improved awareness of the diagnosis and changes in diagnostic criteria likely impact these estimates.

Both genetic and environmental risk factors for MS in children have been identified, many of which are shared with adults.¹² Children with MS are in close temporal proximity to the biological onset of the disease, which aids the study of contributing environmental exposures as children are often living in a stable environmental milieu.

Vitamin D

The epidemiologic association between increased MS incidence/prevalence and increased distance from the equator first raised the hypothesis that the risk of MS may be related to sunlight exposure. A surrogate marker of sunlight exposure is vitamin D, and low serum 25-hydroxyvitamin D (the circulating form of vitamin D) levels were associated with an increased risk for the subsequent diagnosis of MS in a large study of Canadian children with incident demyelination (hazard ratio: 1.1 for every 10 nmol/L decrease).¹³

Viral Exposures

Early exposure to common viruses likely contributes to the risk of MS. Serological evidence of remote exposure to Epstein-Barr virus (EBV) was more commonly detected in children with MS than in controls in one international multi-center study, and was also associated with an increased risk of a diagnosis of MS in a separate study of Canadian children with incident demyelination (hazard ratio: 2.04).^13,14 In addition, previously EBV-exposed children with MS demonstrated a much higher rate of EBV reactivation compared with EBV-exposed children without MS in a longitudinal study that included monthly mouth swabs to detect EBV viral DNA.¹⁵ In contrast to EBV, evidence of remote exposure to cytomegalovirus (CMV)was associated with decreased risk of a diagnosis of MS in a study of Canadian children with incident demyelination (adjusted hazard ratio: 0.42). Similarly, serological evidence of remote CMV infection was less commonly detected in children with MS compared with controls in a U.S. study (odds ratio: 0.27).^16,17 Together, these studies raise the idea that CMV exposure may be protective. Finally, two U.S. studies reported that seropositivity for herpes simplex virus 1 increased the risk for MS in children who were white and did not have the HLA-DRB1*15:01 risk allele.^18,19

Smoking, Air Quality, and Household Chemicals

Second-hand smoke exposure is a risk factor for MS in children. In one population-based case–control study in France, children who had at least one parent who smoked at home had double the adjusted risk for a diagnosis of MS. Risk was higher for older children (adjusted risk ratio: 2.49), suggesting that a longer duration of exposure increases risk. In a Canadian study of children with incident demyelination, second-hand smoke exposure was reported in 37% of children with MS, compared with 30% of those with monophasic demyelination.^20,21

A recent U.S. case–control study reported an increased risk of MS in children exposed to pesticides (odds ratio: 2.18).²² In another U.S. study, exposure to air pollutants (particulate fine matter, sulfur dioxide, carbon monoxide, and lead) was associated with an increased risk of MS, but only in children living within 20 miles of an MS center.²³

Birth Factors

Maternal health and perinatal/postnatal factors may influence the risk for MS in children. A large U.S. case–control study reported that maternal illness during pregnancy was associated with an increased risk of a diagnosis of MS in childhood (adjusted odds ratio: 2.25). Cesarean delivery was associated with a reduced risk of a subsequent diagnosis of MS (adjusted odds ratio: 0.40).²² While this study did not demonstrate an association of infant breastfeeding and risk of MS, a separate study showed that a lack of infant breastfeeding was associated with an increased risk for a future diagnosis of MS (odds ratio: 4.43).²⁴

Sex Hormones

When the clinical onset of MS occurs prior to puberty, the occurrence of MS is equal between sexes; however, MS is two to three times more common in women after puberty, suggesting a role for sex hormones.²⁵ One case–control and one longitudinal study showed that a later age of menarche was associated with a lower risk of a diagnosis of MS in childhood.^26,27 In children with established MS, a 2.4 increased rate of relapses was reported for boys with onset of MS symptoms in the peripubertal period compared with postpuberty.²⁸ Similarly, relapse rates were higher in the perimenarche period compared with postmenarche (incidence rate ratio: 8.5) in a study of girls with MS.²⁹

Obesity and Diet

Obesity may be related to risk of MS in children. Two case–control studies reported increased body mass index in children with MS compared with controls, and this relationship was especially strong in girls.^27,30 In addition, one case–control study reported that the body mass index trajectories of children with MS were elevated not only at diagnosis but also in early childhood, years before the clinical onset of the disease.³¹ Adipokines secreted by adipocytes are involved in immune regulation.³² Obesity also serves as a risk factor for low vitamin D status³³ and earlier age of menarche.²⁷ While there is significant interest in dietary factors (e.g., salt intake) and risk of MS in children, no definite associations have been reported.^34,35

Genetic Factors

Alterations in the major histocompatibility complex (MHC) contribute to risk of MS. Specifically, the HLA-DRB1*15:01 allele conferred increased risk of a diagnosis of MS in children presenting with incident demyelination (odds ratio: 2.7).^13,36 The risk conferred by HLA alleles may be influenced by ancestry (e.g., European, African, Hispanic).³⁷

Several non-HLA single-nucleotide polymorphisms (SNPs) have been associated with an increased risk of MS in children, many of which are identical to SNPs detected in adults.³⁸ Many of these SNPs are implicated in the adaptive or innate immune pathways.

Gene–Environment Interactions

The interplay between genetics and environmental risk factors is important to consider. As an example, the HLA alleles may serve to modify the risk for pediatric MS when considered in the context of passive smoke exposure,²¹ vitamin D status,³⁹ obesity,⁴⁰ and certain viral exposures.¹⁸

Diagnosis of Multiple Sclerosis in Children

As in adults, the diagnosis of MS in children requires evidence of CNS demyelination that is separated in both space and in time. A child with two or more episodes of neurological symptoms characteristic of MS localizing to different areas of the CNS meets the dissemination in space (DIS) and time (DIT) criteria clinically. Diagnostic criteria for MS in adults outline MRI and laboratory factors that may be used to demonstrate DIS or DIT in an individual with only one clinical episode (Table 1).⁴¹ These criteria had a sensitivity of 71% and a specificity of 95% when applied in a longitudinal cohort of children with incident demyelination, supporting their application in children.⁴² Caution needs to be employed when using these diagnostic criteria in young children (<11 years) as many such children meet the clinical definition of acute disseminated encephalomyelitis (ADEM), which is typically monophasic. Therefore, a diagnosis of MS in a child with ADEM requires one or more additional non-ADEM clinical episode and fulfillment of DIS and DIT criteria.⁴¹

Table 1.

MRI and laboratory criteria for a diagnosis of multiple sclerosis^a

Dissemination in space	Dissemination in time
Lesions in two or more of the following locations: Periventricular Juxtacortical Infratentorial Spinal cord	Any of the following: ≥ 1 enhancing lesion New lesions on subsequent MRI ≥ 2 oligoclonal bands present in cerebrospinal fluid (but not serum)

Abbreviation: MRI, magnetic resonance imaging.

^aThese criteria do not apply to children with acute disseminated encephalomyelitis (ADEM), in whom a second non-ADEM clinical neurological event is needed to confirm a diagnosis of multiple sclerosis.

Imaging Characteristics of Children with Multiple Sclerosis

Magnetic Resonance Imaging of Inflammatory Activity

Magnetic resonance imaging (MRI) is important both to support a diagnosis of MS in children and to provide insights into underlying biology. While one study found similar T2 lesion volumes in children and adults with MS matched for disease duration,⁴³ two other studies reported either higher T2 lesion numbers or volumes in children with MS compared with adults, especially in the posterior fossa.^5,44 In one study, children also had a greater number of enhancing lesions.⁴⁴ In the same study, follow-up MRIs in children with MS were more likely to demonstrate new lesions and enhancing lesions, suggesting a more aggressive radiological course in children, which parallels known differences in clinical disease activity. A representative MR image from a child with MS is given in Fig. 1.

Fig. 1.

Representative fluid-attenuated inversion recovery magnetic resonance images from a child with multiple sclerosis.

Structural and Functional MRI Analyses

A worrisome finding for clinicians is that children with MS demonstrate less-than-expected brain growth. One study of Canadian children and one from France reported that children with MS have reduced whole brain volumes when assessed at a single time point and lower-than-expected brain growth over multiple time points compared with matched controls.^45,46 In the Canadian study, reduced regional brain growth in the thalamus was especially marked. Inconsistent associations have been reported between gray matter involvement and white matter lesion volume.^46,47

Ultra-high field MRI at 7T (Tesla) and advanced MRI sequences performed at 3T provide additional information about cortical gray matter involvement in pediatric MS. Cortical lesions can be seen on MRI in children with MS, despite their young age, particularly in intracortical and leukocortical locations.^48,49 No strong associations have been established between cognitive function and the presence of cortical lesions on MRI in children with MS.^47,50 The clinical correlates of cortical lesions in children are an area for future study.

Beyond traditional MRI, advanced imaging modalities provide more information about structural white matter tract integrity and brain function. These techniques include diffusion tensor imaging (DTI), magnetization transfer ratio (MTR), and functional MRI (fMRI).

DTI can detect decreased microstructural integrity of both lesional and normal appearing white matter in children with MS, based on impaired water diffusivity. In children with MS, these microstructural alterations have been demonstrated in nonlesional white matter, especially in the corpus callosum, a heavily myelinated structure.^51–54 Studies have consistently shown both decreased fractional anisotropy and more variable changes in radial and axial diffusivity, which indicate axonal loss and demyelination, respectively.⁵²

MTR is an indicator of remyelination and repair. Studies have reported that younger children with MS show greater MTR recovery within lesions than adults⁵⁵ and even older adolescents.⁵⁶ This finding may correlate with the modest physical disability often observed in pediatric MS patients despite high lesion number and volumes on MRI.

fMRI can be used to learn about functional reorganization in pediatric MS patients. In one study, children with MS had increased (rather than decreased) resting-state functional connectivity in the default and frontoparietal networks, similar to cognitively intact adults with early MS.⁵³ The underlying adaptive properties of this finding is unclear, but may be a protective mechanism in pediatric MS patients.⁵⁷ However, the adult MS literature indicates that this apparent plasticity is lost over time and that later functional reorganization may not be beneficial.^53,57,58

Immunology of Multiple Sclerosis in Children

MS has traditionally been thought of as a T-cell–mediated disease, precipitated by an abnormal balance between regulatory T-cells (T_reg) and CNS-reactive effector T cells. In a German study that compared circulating T-cell profiles from children with MS to age-matched controls, children with MS had a lower proportion of circulating naive T-cells, a higher proportion of memory T-cells, and reduced suppressive capacity of T_reg subsets.⁵⁹ Of note, the ratio of naive to memory T-cells in children with MS was similar to those of controls 20 to 30 years older in age, suggesting that there may be a premature impairment in the ability of the thymus to produce new T-cells in children with MS. A U.S. study reported that T-cell proliferation in response to exposure to myelin peptides was increased in 10 pediatric MS patients, relative to 10 age-matched controls and 10 adults with MS, suggesting an increased peripheral immune response to myelin-derived antigens.⁶⁰ Finally, one study showed that the serum T-cell cytokine signatures from pediatric MS patients differed from healthy controls, with certain serum cytokines (IL-10, IL-21, IL-23, and IL-27) predicting MS relapse.⁶¹

B-cell immunology plays an important role in mediating inflammatory CNS disease. Circulating serum antibodies against myelin basic protein were not elevated in children with MS compared with controls in one study.⁶² In a Canadian study of children with incident demyelination, CSF antibodies against nodal/paranodal assembling proteins were more commonly present in children who later were diagnosed with MS. Antibodies toward myelin oligodendrocyte (MOG) protein are more commonly found in children with ADEM compared with those diagnosed with MS.⁶³ Antibodies to KIR4.1, an inward rectifying potassium channel, were detected in more than half of children with incident-acquired demyelination of the CNS in a German study.⁶⁴

Treatment of Children with Multiple Sclerosis

After a diagnosis of MS has been made, initiation of a DMT is recommended to decrease relapses and to reduce the accumulation of disability from the disease. Initiation of DMT in children requires the coordinated efforts of a multidisciplinary team to address the practical issues of administration, emphasize the importance of adherence to therapy, address potential barriers, and assess the psychological burden of a chronic disease.

Disease-Modifying Therapies

DMTs come in the form of injections, infusions, and oral medications. While there are several DMTs approved by the FDA for treatment of adults with MS, there are few randomized controlled trials of these medications for children (Table 2). In addition, there is only one DMT approved by the FDA for use in children with MS. The PARADIGMS study was a phase 3 study of fingolimod versus weekly interferon-β1a.⁶⁵ The annualized relapse rate was 82% lower with fingolimod than interferon-β1a, and there were fewer new and enlarging T2 lesions on MRI. Clinical trials are underway for dimethyl fumarate and teriflunomide. Future studies are anticipated for ocrelizumab and cladribine. Until these studies in children are completed, adult data inform their off-label use in children. DMTs with evidence only in adults include only siponimod, cladribine, and ocrelizumab.

Table 2.

Treatment options for children with multiple sclerosis

Disease-modifying therapy	Method of administration73	Mechanism of action73	Dosing	Side effects	Monitoring73	ARR	MRI response	Level of evidence in children
INF-β1a	Injectable	Polypeptide causing inhibition of T-lymphocyte proliferation, reduced migration of inflammation cells across the blood-brain barrier (produced in mammalian cell lines)	22–44 μg three times a week74	Injection site reactions (28%), Influenza-like symptoms (24%), hepatic disorders (14%), blood cell disorders (5%), allergic reactions (2%), seizure (2%), thyroid disorder (1%), autoimmune disorders (1%)74	Liver function, blood counts	0.4774 0.744–0.795 (US), 0.2313–0.426 (other countries)75 1.38 (high-dose IFN-β)76 0.4–0.977	New or enlarging T2 lesions (annualized rate): 9.2765	Retrospective studies,74–76 open label prospective77
INF-β1b	Injectable	Similar to β1a, but produced in bacterial expression systems (not glycosylated)	250 μg every other day	Injection-site reactions, influenza-like symptoms, elevated liver enzymes	Liver function, blood counts at initiation and every 6 mo	50% reduction in ARR; 0.67 for treatment < 1 y	None in children	Retrospective study78
Glatiramer acetate	Injectable	Synthetic peptides similar to myelin basic protein, promote Th2 deviation under development of Th2 glatiramer acetate reactive CD4+ T-cells	20 mg daily or 40 mg three times a week76	Dyspnea (7%), skin injection reaction (13%) Adults: lipoatrophy (permanent)73	None	0.277 0.5376	Adults: CELs per MRI scan at 1 y: 0.41 New or enlarging T2 lesions at 1 y:1.4979	Retrospective study,76 open label prospective study77
Fingolimod	Oral	Sphingosine1-phosphate receptor modulator, inhibits ability of autoreactive lymphocytes to leave lymph nodes	0.5 mg daily (0.25 mg if <40 kg)65	Headache (32%), URI (38%), leukopenia (14%), convulsions (6%)65; Adults: transient bradycardia/AV block with first dose, macular edema, elevated liver enzymes, one death from fulminant primary varicella-zoster infection73	Monitor with electro-cardiogram for 6 h after the first dose. Liver function, routine eye exams. Screening for varicella-zoster antibodies before initiation	0.1265 0.14–0.2780	New or enlarging T2 lesions (annualized rate): 4.4–7.565,80	Randomized controlled trials: age 10–17 y old,65 age < 20 y old80
Dimethyl fumarate	Oral	Activation of the nuclear factor (erythroid-derived 2)-like 2 transcriptional pathway	240 mg twice a day81	Flushing (60%), GI symptoms (54%), rash (23%), malaise (15%), treatment discontinuation from side effects (23%)81 Adults: reports of PML in JCV-positive patients with prolonged lymphopenia73	Liver function, blood counts. JCV antibodies if prolonged lymphopenia (<500 cells/mm3)	0.681 0.882	New T2 lesions at 1 y: 33% 3-fold reduction in new T2 lesions82	Retrospective study,81 open label phase II trial (FOCUS),82 Phase III trial, (NCT02283853), ongoing
Teriflunomide	Oral	Inhibits de novo pyrimidine synthesis, reducing circulating lymphocytes	14 mg daily73	Headache (16%), hepatic enzyme elevation (15%), diarrhea (14%), hair thinning (14%), serious infections (3%)83	Liver function, blood counts, blood pressure	Adults: 31–35% reduction in ARR83	Adults: decrease in CELs compared with placebo: 80.4%84	Phase III trial (NCT02201108, “TERIKIDS”), ongoing
Mitoxantrone	Infusion	Synthetic anthracenedione derivative, inhibits proliferating immune cells and induces apoptosis	20 mg (10 mg if <12 y old) monthly for 2–4 doses85	Cardiotoxicity (palpitations or cardiomyopathy—26%), nausea/vomiting (74%), fatigue (53%), alopecia (47%). Mild leukopenia (5%), transient amenorrhea (11%)85 Adults: therapy-related leukemia, teratogenicity73	Liver function, blood counts. Echocardiograms before, during, and after treatment	0.4785	New T2 lesions at the end of treatment duration: 28.5%85	Retrospective study85
Natalizumab	Infusion	Monoclonal antibody, prevents adherence of activated leucocytes to inflamed endothelium	300 mg every month86	Headache (6%), GI symptoms (4%)69 URI symptoms (9%)87 PML	JCV antibodies every 6 mo, liver function, blood counts	0.0686 0.187 0.3888 0.489	MRI activity: 12.5%86 New T2 or CELs at 1 y: 7%87	Open label prospective study86–87 Retrospective study88–89
Rituximab	Infusion	Monoclonal antibody against CD20+ B cells90	375 or 500 mg/m2 (up to 1,000 mg) Initiation; two doses, 2 wk apart Maintenance: every 6 mo90,91	Rash (5%), hypotension (4%)69 Lymphopenia (41%), hypogammaglobulinemia (22%), anaphylaxis (2%), severe infection or death from infection (2.8%)90	CD19 every 6 mo, liver function, blood counts, immunoglobulin levels	Adults: 0.0491	Adults: decrease in CELs per MRI: 0.8–0.0591	Observational cohort study69
Cyclophosphamide	Infusion	Alkylating agent: interferes with DNA transcription in rapidly dividing cells	Monthly based on lymphocyte counts +/− preceding induction	Nausea/vomiting (88%), lymphopenia (100%), anemia (63%), thrombocytopenia (19%), hematuria (6%), alopecia/hair thinning (59%), menstrual irregularities (29%), infection (18%), paresthesias (6%), central line complications (6%), fatigue (12%), urticarial (6%), myalgias (6%), amenorrhea (18%), sterility (6%), osteoporosis (12%), bladder cancer (6%)92	Liver function, blood counts, urinalysis	1.192	75% with new T2 lesions at 1 y92	Retrospective study of children with highly active MS92

Abbreviations: ARR, annualized relapse rate; CELs, contrast enhancing lesions; GI, gastrointestinal; INF, interferon; JCV, John Cunningham virus; MRI, magnetic resonance imaging; MS, multiple sclerosis; PML, progressive multifocal leukoencephalopathy; URI, upper respiratory infection.

Several factors need to be considered when choosing DMTs, including likelihood of compliance, exposure status to John Cunningham virus (given its association with progressive multifocal leukoencephalopathy), and the individual risks/benefits. For many years, injectable DMTs were considered first-line therapies. However, children face several logistical and social factors that decrease compliance with, or give a negative perception of, injectable DMTs.⁶⁶

There are two main approaches to DMT initiation and switching of DMTs: escalating therapy versus initiating early high-efficacy therapy. Escalating therapy consists of starting with less potent therapies with close follow-up, with a switch to higher-efficacy therapies with any subsequent relapses or radiographic disease progression. Early high-efficacy therapy consists of the early initiation of potent DMTs, with the goal of inducing and maintaining remission early.⁶⁷ Arguments for the use of early high-efficacy therapy in children include the higher relapse rate,³ higher volume of brain lesions on MRI,⁵ and lower prevalence of NEDA with treatment (no evidence of disease activity—clinical or radiographic).⁶⁸ However, many of the DMTs commonly used in adults have limited evidence in children. Therefore, careful consideration and discussion with families of potential risks and benefits are needed. While treatment patterns may vary, practice patterns in the United States have been evolving toward the earlier use of more potent therapies earlier in the disease course.⁶⁹

Emerging Concepts and Future Directions for Treatment

In addition to much-needed additional randomized controlled trials for children, a greater understanding of how DMTs alter the developing immune system and their long-term effects is imperative. We also need to better understand how lifestyle factors may influence the course of disease. Vitamin D, exercise, and dietary factors may have a favorable effect on relapse rate. For example, for vitamin D, one study reported that every 10 ng/mL increase in adjusted serum 25-hydroxyvitamin D level was associated with a 34% decrease in relapse rate.⁷⁰ Trials of dietary supplements have not been conducted independent of concurrent DMT use, and thus cannot be recommended as monotherapies. In one study, increased intake of fat, especially saturated fat, was associated with an increased risk of relapse. Conversely, increased consumption of vegetables was associated with a decreased risk of relapse.⁷¹ In another study, higher levels of strenuous physical activity were associated with both reduced T2 lesion volumes on MRI and lower relapse rates.⁷² Together, these findings suggest that certain diet and lifestyle interventions may have a benefit. This is an area for future study.

Conclusion

Multiple sclerosis in children is being increasingly diagnosed worldwide. This has been aided by improved recognition of the disease, consensus diagnostic criteria, and improved imaging. The past several years have led to an improved understanding of factors that contribute to the risk of MS in children and the pathobiology of the disease. Therapeutic options for MS in children are increasing; however, clinical trials for new and emerging treatments are still needed for children. Multicenter collaborations such as those fostered by the International Pediatric Multiple Sclerosis Study Group will aid in improving future clinical care and advancing research.