Sage Journals: Discover world-class research

Abstract

Background

Serum neurofilament light chain (sNfL) is a marker of neuroaxonal injury, and serum glial fibrillary acidic protein (sGFAP) reflects reactive astrogliosis. In adult multiple sclerosis (MS), sNfL correlates with relapsing disease activity while sGFAP correlates with progressive disease.

Objectives

We evaluate sNfL and sGFAP as biomarkers in pediatric-onset MS (POMS) compared to pediatric healthy controls (PHC), and correlations with the disease course.

Methods

In this single-center observational cross-sectional study, we extracted data from a longitudinal database and measured NfL and GFAP from bio-banked serum using single-molecule array technology.

Results

The analysis included 61 POMS patients and 45 PHC. Controlling for age and BMI, sNfL was 414% higher and sGFAP was 42.3% higher in POMS. Disability (EDSS) is associated with higher sNfL (β = 0.32, p = 0.002) and higher sGFAP (β = 0.11, p = 0.03). sNfL is associated with MRI lesion burden, recent disease activity (β =0.95, p < 0.001), and untreated status (β = 0.5, p = 0.006).

Conclusion

sNfL and sGFAP are elevated in POMS compared to PHC. Both biomarkers are associated with clinical disability. Elevated sGFAP may reflect early neurodegeneration in POMS, while sNfL reflects disease activity and DMT response. Elevated sNfL among some clinically and radiographically stable POMS patients suggests ongoing neuroaxonal injury with a potential role for sNfL monitoring disease stability.

Keywords

Multiple sclerosis pediatrics biomarker relapse neurofilaments glial fibrillary acid protein

Introduction

Multiple sclerosis (MS) is a leading cause of non-traumatic disability among young adults.¹ Pediatric-onset MS (POMS) accounts for approximately 3% to 5% of MS cases.² POMS is a highly inflammatory form of the disease in which children experience two to three times as many relapses^3,4 and develop a higher disability and cognitive impairment at younger ages than adult-onset MS (AOMS) patients.⁵ Disease modifying therapies (DMTs) control neuroinflammatory activity in relapsing-remitting MS (RRMS), thus reducing the risk of relapse, new lesion formation, and disability accrual. With multiple treatment options now available, including some highly effective immunosuppressive DMTs, it is particularly important to monitor the disease activity and DMT response in order to adjust treatment strategies accordingly, optimizing disease control while also avoiding excessive immunosuppression.

Fluid-based biomarkers are an exciting and emerging topic in the field of MS. The value of these biomarkers lies in their potential to monitor and predict disease activity. Two serum biomarkers of particular interest are serum neurofilament light chain (sNfL) and serum glial fibrillary acidic protein (sGFAP). In general, sNfL is a marker of neuroaxonal injury,⁶ and sGFAP is a marker of astrocyte activation.⁷

The neurofilament light chain is a component of the neuronal cytoskeleton that is released in the setting of neuroaxonal injury.⁶ A growing body of the literature has shown sNfL to both reflect and predict neuroinflammatory disease activity, burden, and severity in MS.^8–24 However, most relevant studies are limited to adult MS patients. We were able to identify only four studies that investigate sNfL in POMS.^14–17 As such, questions remain regarding the value of sNfL as a biomarker of disease activity, DMT response, and both clinical and radiographic disease burden in the pediatric MS population.

Glial fibrillary acidic protein is an intermediate filament of astrocytes.⁷ sGFAP is considered a marker of astrocyte activation and reactive astrogliosis, and is elevated in multiple neurodegenerative neurological disorders.⁷ While sNfL reflects inflammatory and relapsing disease activity in MS, GFAP has shown promise as a marker of neurodegeneration, disability accrual, and progressive disease.^{7,10,11,25,26} Whether this holds true for pediatric MS patients remains an important and unaddressed question. There have been no published studies on sGFAP in pediatric MS.

In this study, we aimed to evaluate the value of sNfL and sGFAP as biomarkers in pediatric multiple sclerosis. We planned to investigate sNfL and sGFAP as markers of disease activity, accumulated disease burden, and treatment response. In addition, as a novel contribution to the literature, we sought to evaluate the added utility of sNfL and sGFAP in detecting evidence of subclinical disease activity in a subset of POMS patients who were otherwise clinically and radiographically stable as per standard of care clinical assessments and MRI surveillance.

Patients and methods

General study design

This is an observational, cross-sectional, single-institution study.

Study population

Participants included POMS patients enrolled prospectively in a demyelinating disease database study and unmatched healthy pediatric controls recruited for this biomarker study. We used a consecutive sample of all enrolled patients who met this study's inclusion criteria and had provided serum between 2005 and 2021.

POMS patients met the 2017 revised McDonald criteria and had their first demyelinating attack prior to the age of 18 years.²⁷ All patients were evaluated at least once by a neurologist in the pediatric MS clinic at Massachusetts General Hospital. As many of the patients were last seen prior to the recognition of myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) as a separate diagnostic entity, we retrospectively reviewed charts and excluded cases with subsequent positive MOG antibody testing (titer 1:40 or higher) and/or presentations more consistent with MOG antibody-associated disease.²⁸

Definitions

We defined the DMT status as follows: Untreated patients had no DMT exposure or had been on DMT for <3 months. Treated patients had been on DMT for >3 months. Low-efficacy DMTs included beta-interferons and glatiramer acetate, moderate-efficacy DMTs included dimethyl fumarate, and high-efficacy DMTs included fingolimod, natalizumab, rituximab, cyclophosphamide, and ocrelizumab. We defined active disease as an adjudicated clinical relapse and/or a new gadolinium-enhancing (gad+) lesion on MRI in the 60 days prior to biomarker measurement. Stable disease or “remission” was defined as documented evidence of clinical stability and radiographic stability within at least 60 days before and after the time point in question. If disease activity or stability could not be determined due to incomplete records or unclear timing of new lesions, the disease status was considered “unknown.” T1 burden was considered “low” with fewer than five and “high” with greater than five hypointense T1 lesions; T2 burden was considered “low” with fewer than ten and “high” with greater than ten hyperintense T2 lesions.

Data collection

Most demographic and clinical variables of interest were obtained through database query. For variables that can change over time, such as EDSS, we used the measurement closest in time within 90 days to the biomarker measurement. A chart review was performed to validate all clinical relapses. Our laboratory has established workflows and the literature to guide this data validation.^9,22 A review of the electronic medical record and neuroimaging was also required to obtain additional variables including the date of MS disease onset and MRI brain lesion burden. sNfL and sGFAP levels from banked serum samples were quantified using the Quanterix single molecule array (SIMOA) technology.^9,22

Statistical analysis

As biomarker levels were right-skewed, data analysis focused on log-transformed values or geometric means. First, we compared average biomarker levels between healthy controls, MS patients, and subgroups of interest using a random intercept model to account for the repeated observations of subjects. The random intercept model was used to account for the within-subject correlation since several subjects contributed multiple measurements to the analysis. The estimated difference in the log-transformed mean as well as the percent change in the geometric mean are reported for the group comparisons. We generated unadjusted models and models adjusted for BMI and age. We adjusted the BMI for given multiple studies that have shown a higher BMI to be associated with lower sNFL and sGFAP.²⁹ Second, we used the same random intercept models to estimate the association between biomarker levels and clinical and demographic variables of interest. All analyses were completed in the statistical package R (www.r-project.org) or Stata version 17.

Standard protocol approvals, registrations, and patient consents

This study was approved through the institutional review board at Mass General Brigham, approval number 2005P001041. Parents or guardians signed informed consent for participation in this study.

Results

Study cohort characteristics

We identified 70 patients with an initial diagnosis of POMS, but then excluded seven due to a positive MOG titer and two more due to presentations more consistent with MOGAD than MS (recurrent bilateral optic neuritis and multiphasic disseminated encephalomyelitis). Our study cohort therefore included 61 POMS patients, of whom 38 provided serum at one timepoint and 23 who provided serum at multiple timepoints. We selected 45 healthy pediatric controls, of whom 37 provided serum at one time-point and 8 provided serum at two time-points. As such, this study analyzed a total of 110 serum samples from POMS patients and 53 from healthy controls. Among the POMS group, 64 serum samples (58.7%) were obtained between the years 2005 and 2009, 31 (28.4%) were obtained between 2010 and 2015, and only 13 (11.9%) were obtained after 2015.

Table 1 reports demographic and baseline clinical data. Most notably, POMS patients were significantly older (mean 15.9 years vs 11.7 years, p<0.005) and had a significantly higher BMI (27.6 vs 20.9, p<0.005) than the healthy controls. In general, our POMS cohort was relatively early in their disease course with a median duration of illness of 20.5 months. They also had fairly low disability, with a median EDSS score of 1.5. Thirty serum samples (27.5%) were provided by patients within 6 months of the first demyelinating event and prior to starting treatment. Just over half (54.1%) were untreated at the time of serum collection; of those who were on treatment, only five (11.1%) were on medium or high-efficacy DMTs. Approximately half (50.5%) of the samples were collected in the setting of recent disease activity, clinical or radiographic, in the last 60 days. Overall, a clinical relapse had occurred in 39.4% samples within the last 60 days and in 26.6% samples within the last 30 days. Of those who had a recent MRI, a new gadolinium-enhancing lesion was detected in 51.8% within the last 60 days and, in 26.6% within the last 30 days.

Table 1.

Demographic and clinical data for POMS patients and healthy controls.

		Pediatric MS patients	Pediatric healthy controls	Significance (p-value)
Number of participants		61	45	–
Number of serum samples		109	53	–
Sex	Female	46 (75.4%)	26 (57.8%)	p = 0.087
Race	African American	9 (14.8%)	3 (6.7%)	p = 0.189
	Asian	1 (1.6%)	1 (2.2%)
	White	39 (63.9%)	36 (80.0%)
	Other/multiple	12 (19.7%)	5 (11.1%)
Ethnicity	Hispanic	20 (32.8%)	5 (11.1%)	p = 0.001
	Non-Hispanic	41 (67.2%)	35 (77.8%)
	Unknown	0	5 (11.1%)
Average age (years)		15.9 (range 7.9–17.9)	11.7 (range 4.4–17.9)	p < 0.001
Average BMI		27.6 (range 15.6–49.6)	20.9 (range 13.6–45.9)	p <0.001
Average disease duration (range)		20.5 months (2 days-7 years)
Average EDSS (range)		1.5 (0–6)
Presence of oligoclonal bands		33/55 (60%)
Treatment Status	Untreated	59/109 (54.1%)
	Low efficacy DMT	45/109 (41.3%)
	Medium efficacy DMT	1/109 (<1%)
	High efficacy DMT	4/109 (3.7%)
Disease Activity*	Active disease	55/109 (50.5%)
	Remission	45/109 (41.3%)
	Unknown status	9/109 (8.3%)
Clinical relapse within the last 30 days		29/109 (26.6%)
Clinical relapse within the last 60 days		43/109 (39.4%)
New gad+ MRI lesion within the last 30 days		30/71 (42.3%)
New gad+ MRI lesion within the last 60 days		43/83 (51.8%)
New T2 MRI lesion within the last 6 months		85/97 (87.6%)
New T2 MRI lesion within the last 12 months		99/104 (95.2%)

This table shows demographic and clinical data for our POMS cohort and healthy pediatric controls at baseline.

* Active disease is defined as new gad+ lesion(s) or onset of a validated clinical relapse within the prior 60 days, and remission is defined as > 60 days of clinical and radiographic stability.

Biomarker levels in POMS versus controls

Summary statistics for sNfL and sGFAP in our healthy controls and POMS cohort are reported in Figure 1(a). Our POMS cohort had a geometric mean sNfL of 27.8 pg/L (SD 39.7, range 1.3–238) and sGFAP of 118.7 pg/L (SD 78.8, range 26–580), while healthy controls had a geometric mean sNFL of 5.2 pg/L (SD 2.4, range 1.8–14) and sGFAP of 115.9 pg/L (SD 48.2, range 33–237). In an unadjusted random effects model, the geometric means of sNFL and sGFAP were 276% and 2.4% higher, respectively, among POMS patients compared to healthy controls. After controlling for repeated observations and adjusting for age and BMI, levels of both biomarkers were significantly higher in the POMS group. As determined by adjusted percent change in geometric means, POMS patients had sNfL levels 414% higher (95% CI [236.9, 684.3]) and sGFAP levels 42.3% higher (95% CI [11.8, 81.1]) than healthy controls (Figure 1).

Figure 1.

sNFL and sGFAP in healthy controls, POMS patients, and subgroups of interest. 1a – sNfL in healthy controls, POMS patients, and subgroups of interest. 1b – sGFAP in healthy controls, POMS patients, and subgroups of interest. The box plots in Figure 1a compare sNfL levels and in Figure 1b compare sGFAP levels among healthy controls vs POMS patients (panel A), POMS patients with active disease vs remission* (panel B), and POMS patients on any DMT vs untreated (panel C). Note that these are raw sNfL and sGFAP levels, not yet adjusted for BMI and age. HC = healthy controls; DMT = disease-modifying treatment; Active disease defined as new gad+ lesion(s) or onset of a validated clinical relapse within the prior 60 days, and remission defined as > 60 days of clinical and radiographic stability.

Biomarker association with demographic features

In Tables 2 and 3 we report associations between baseline variables and biomarker levels. In univariate analyses, three demographic variables—older age, higher BMI, and Black race—were associated with lower sNfL and lower sGFAP. After adjusting the BMI, the relationship between age and biomarker level was no longer statistically significant.

Table 2.

Variables associated with log-transformed sNfL in POMS.

		Unadjusted		Adjusted (for BMI and age)
		β, p-value	% change (95% CI)	β, p-value	% change (95% CI)
Demographic factors
Sex	Female Male	Reference 0.02, p = 0.937	2.5 (−45.0, 90.8)	−0.05, p = 0.877	−4.6 (−47.8, 74.4)
Race	White	Reference −0.83, p = 0.032 −0.27, p = 0.468	−56.4 (−79.6, −7.0) −24.0 (−64.2, 61.6)	−0.78, p = 0.028 0.14, p = 0.698	−54.1 (−77.0, −8.2) 15.5 (−45.0, 142.3)
	Black
	Other
Ethnicity	Non−Hispanic Hispanic	Reference 0.26, p = 0.355	30.2 (−26.0, 129.1)	0.33, p = 0.214	39.6 (−18.0, 137.5)
BMI		−0.07, p = 0.001	−6.4 (−10.0, −2.7)	−0.06, p = 0.004	−5.8 (−9.5, −1.9)
Age (years)		−0.12, p = 0.039	−11.6 (−21.3, −0.6)	−0.09, p = 0.129	−8.5 (−18.4, 2.7)
Measures of disease burden
Disease duration (months)		−0.01, p = 0.066	−1.0 (−2.0, 0.1)	−0.01, p = 0.110	−0.9 (−2.1, 0.2)
EDSS score		0.34, p = 0.001	40.4 (14.8, 71.9)	0.32, p = 0.002	37.1 (12.8, 66.6)
High T2 lesion burden		1.13, p < 0.001	209.7 (89.2, 407.0)	1.03, p < 0.001	180.5 (72.3, 357)
High T1 lesion burden		1.38, p < 0.001	297.0 (148.3, 534.9)	1.25, p < 0.001	250.0 (119.9, 457.0)
Measures of disease activity and control
Active disease*		1.06, p < 0.001	189.4 (107.6, 303.5)	0.95, p < 0.001	159.0 (87.6, 257.5)
Clinical relapse within 30 days		0.83, p < 0.001	130.0 (57.8, 235.4)	0.74, p < 0.001	109.7 (44.9, 203.4)
New gad+ lesion within 30 days		1.05, p < 0.001	185.4 (77.0, 360.0)	1.04, p < 0.001	181.6 (76.1, 350.3)
Treatment	Untreated	Reference −0.60, p = 0.001	−45.2 (−61.8, −21.3)	−0.50, p = 0.006	−39.4 (−57.5, −13.7)
Treatment	Treated	Reference −0.60, p = 0.001	−45.2 (−61.8, −21.3)	−0.50, p = 0.006	−39.4 (−57.5, −13.7)

This table shows results from a linear mixed model with log-transformed sNfL as the outcome and the listed variables as predictors. β – Difference in log-transformed means as estimated from linear mixed models with the log-transformed sNfL as the outcome and the listed variable as the predictor; % change – The percent change in the geometric mean as estimated by back transforming the coefficient from the linear mixed model; CI=confidence interval; *Active disease defined as new gad+ lesion(s) or onset of a validated clinical relapse within the prior 60 days, and remission defined as > 60 days of clinical and radiographic stability.

Table 3.

Variables associated with log-transformed sGFAP in POMS.

		Unadjusted		Adjusted (for BMI and age)
		β, p-value	% change (95% CI)	β, p-value	% change (95% CI)
Demographic factors
Sex	Female Male	Reference −0.12, p = 0.459	−11.7 (−36.8, 23.3)	−0.19, p = 0.257	−17.7 (−41.4, 15.7)
Race	White	Reference −0.41, p = 0.045−0.40, p = 0.047	−33.5 (−55.4, -0.9) −33.2 (−55.1, -0.5)	−0.40, p = 0.048 −0.26, p = 0.223	−33.1 (−55.1, -0.4) −22.9 (−49.5, 17.7)
	Black
	Other
Ethnicity	Non-Hispanic Hispanic	Reference0.07, p = 0.659	7.1 (−21.3, 45.5)	0.08, p = 0.584	8.8 (−20.0, 47.9)
BMI		−0.03, p = 0.002	-3.3 (−5.4, -1.3)	−0.03, p = 0.007	−3.0 (−5.1, -0.8)
Age (years)		−0.06, p = 0.056	-5.4 (−10.6, 0.1)	−0.03, p = 0.237	−3.4 (−8.8, 2.3)
Measures of disease burden
Disease duration (months)		−0.004, p = 0.077	-0.45 (−0.94, 0.05)	−0.002, p = 0.506	−0.21 (−0.85, 0.43)
EDSS score		0.13, p = 0.011	14.1 (3.2, 26.2)	0.11, p = 0.030	11.7 (1.1, 23.4)
High T2 lesion burden		0.34, p = 0.020	40.5 (5.6, 86.8)	0.29, p = 0.054	33.7 (−0.5, 79.7)
High T1 lesion burden		0.35, p = 0.017	42.2 (6.8, 89.4)	0.27, p = 0.069	30.5 (−2.0, 73.9)
Measures of disease activity and control
Active disease*		0.08, p = 0.346	8.8 (−8.9, 30.0)	0.03, p = 0.703	3.3 (−12.8, 22.5)
Clinical relapse w/in 30 days		0.08, p = 0.411	8.3 (−10.6, 31.2)	0.01, p = 0.937	0.8 (−16.6, 21.8)
Gad+ lesion w/in 30 days		0.21, p = 0.076	22.8 (−2.2, 54.1)	0.15, p = 0.182	16.5 (−7.1, 46.2)
Treatment	Untreated	Reference −0.029, p = 0.752	−2.9 (−19.0, 16.5)	0.016, p = 0.863	1.6 (−15.1, 21.6)
Treatment	Treated	Reference −0.029, p = 0.752	−2.9 (−19.0, 16.5)	0.016, p = 0.863	1.6 (−15.1, 21.6)

This table shows results from a linear mixed model with log-transformed sGFAP as the outcome and the listed variables as predictors. β – Difference in log-transformed means as estimated from linear mixed models with log-transformed sGFAP as the outcome and the listed variable as the predictor; % change – The percent change in the geometric mean as estimated by back transforming the coefficient from the linear mixed model; CI=confidence interval; *Active disease defined as new gad+ lesion(s) or onset of a validated clinical relapse within the prior 60 days, and remission defined as > 60 days of clinical and radiographic stability.

Biomarker associations with measures of disease burden

Greater disease burden, as measured clinically by EDSS and radiographically by MRI lesion burden, was associated with higher sNfL and higher sGFAP by univariate regression (Figure 2, Tables 2 and 3). For EDSS, the association with both sNfL and sGFAP remained statistically significant after adjusting for age and BMI. For each single point increase in the EDSS, the geometric mean of sNfL increased by 37.1% [95% CI [12.8, 66.6]) and that of sGFAP increased by 11.7% (95% CI [1.1, 23.4]). Biomarker levels also varied by radiographic lesion burden. Greater T1 and T2 lesion counts were associated with higher sNfL and higher sGFAP among the POMS patients by univariate regression. However, while the relationships between sNfL and radiographic burden remained robust after adjusting for age and BMI, the relationships between sGFAP and radiographic lesion burden dropped just below the level of statistical significance in these multivariate analyses (Tables 2 and 3). After adjustment, the geometric mean of sNfL was 181% greater (95% CI [72.3, 357]) in the high T2 lesion group and 250% higher (95% CI [119.9, 457.0]) in the high T1 lesion group.

Figure 2.

sNfL and sGFAP by EDSS in POMS. This figure shows the distribution of sNfL and sGFAP levels by EDSS. Each dot represents a single measurement from one POMS patient. Some POMS patients contributed more than one measurement at different time points. The raw data with all measurements are included the plots above. However, the random intercept model was used to account for repeated observations in regression analyses.

Biomarker association with measures of current disease activity and control:

Active disease, as defined by recent clinical relapse and/or new gad+ lesions within 60 days, was associated with higher sNfL, and DMT treatment was associated with lower sNfL (Figure 1(a), Table 2). After adjusting for age and BMI, sNfL was 173% higher (95% CI [105.0, 263.9]) in active disease compared to remission. Clinical relapse within the last 30 days and new gad+ lesion(s) within the last 30 days were associated with a 110% (95% CI [44.9, 203.4]) and 182% (95% CI [76.1, 350.3]), respectively, increase in sNfL in adjusted models. Of note, while POMS patients in remission had lower sNfL than those with active disease, both of these POMS subgroups had higher sNfL levels compared to healthy controls (Figure 1(a)). Relative to healthy controls, the geometric mean of sNfL was 148% higher (95% CI [63.0, 276.7]) among POMS patients in remission and 577% higher (95% CI [362.2, 890.8]) among POMS patients with active disease. There was no association between sGFAP and active disease, recent clinical relapse, or recent gad+ lesion (Figure 1(b), Table 3).

Regarding treatment effects, the geometric mean sNFL among treated POMS patients was 39.4% lower (95% CI [(−57.5, −13.7]) than that of untreated POMS patients after adjusting for age, BMI, and repeated measurements. We observed lower mean sNFL levels among the five patients on moderate/high efficacy DMTs compared to those on low efficacy treatments, but this difference did not reach statistical significance. There was no association between sGFAP and treatment status (Table 3).

Discussion

We found sNfL to be dramatically higher in our POMS cohort compared to healthy pediatric controls. This finding is consistent with the literature from the adult MS population^{11,18,23,25,30} and four existing studies from the POMS population.^14–17 It is likely due to the greater degree of neuroaxonal injury in pediatric MS compared to healthy controls. This association between sNfL and neuroaxonal injury has been described across many different central nervous system disorders.⁶ It is also important to note that, on average, clinically and radiographically stable POMS patients had a higher sNfL than the healthy controls. This finding indicates ongoing neuroaxonal injury in a subset of cases with otherwise apparently good disease control. It also supports a role for sNfL to more comprehensively assess disease stability when combined with the current practice of routine clinical exams and surveillance MRIs.

After adjusting for important covariates age and BMI, we also found that a subset of POMS patients drives an overall higher sGFAP among POMS patients compared to controls. This is a novel finding in POMS but consistent with the adult MS literature.^11,25,31 Elevated sGFAP indicates astrocyte activation and glial scarring, the phenomena observed in the setting of numerous inflammatory and neurodegenerative conditions.⁷ In adult MS, sGFAP is associated with progression and disability accrual, with higher sGFAP in progressive MS compared to relapsing-remitting MS.^{7,11,13,25,26,31} This leads to a particularly salient inference: Elevated sGFAP in a subset of our young, recently diagnosed relapsing-remitting POMS cohort suggests that neurodegenerative processes underlying progressive MS likely begin very early in the course of the disease. This may support a role for sGFAP in monitoring underlying and poorly understood mechanisms of progression that are subclinical in the early stages of MS, undetectable by standard MRI protocols, and not well suppressed by current DMTs.

Our study finds that sNfL and sGFAP both correlate with greater accumulated disease burden, as measured by higher disability scores and/or MRI lesion counts. These sNfL findings are consistent with the adult MS literature and three recent POMS studies,^{10,14,15,17,18,20,23,24} and the sGFAP findings align with the adult MS literature.^{7,11,13,31,32} Also consistent with prior studies,,^{9,10,11,15,17–19,22,24,25,30} we find that sNfL, but not sGFAP, reflects active, relapsing neuroinflammatory activity, with sNfL higher in active disease – characterized by recent clinical relapse and/or gad+ lesion(s) – compared to remission.

In short, while sGFAP may serve as a marker of accumulated disease burden and progressive MS, sNfL is a better marker of relapsing MS and active neuroinflammation. Thus, sNfL could be a useful indicator of treatment response. Indeed, we find that sNfL, but not sGFAP, is significantly lower among treated POMS patients compared to untreated patients. Our data also demonstrates a trend toward lower sNfL with higher efficacy DMTs. However, as most of our cohort was enrolled prior to the now widespread use of higher efficacy DMTs for pediatric MS, the number of patients on moderate and high efficacy DMTs in our cohort was too small to adequately power this sub-analysis. Nevertheless, the finding is consistent with adequately powered studies in adult MS that reveal greater declines in sNfL following initiation of higher efficacy DMTs.^{12,13,15,17–20,23,24,30} Altogether, these data suggest that lower sNfL levels may reflect more favorable treatment response and better disease control. As such, sNfL could potentially inform clinical decisions around medication management.

A common goal in the management of MS is NEDA-3—the absence of clinical relapses, new MRI lesions, and disability progression.³³ Unfortunately, it has become clear that some relapsing MS patients who achieve NEDA-3 nevertheless do eventually develop secondary progressive MS.^33,34 Recently, the concept of NEDA-4—which includes the additional criterion of annual brain volume loss <0.4%—is gaining traction as a way to more comprehensively assess disease stability, specifically as it relates to the risk of progression independent of relapse.^34,35 Similarly, the field may consider sNfL as an additional means of monitoring disease activity and ultimately defining NEDA-5. Future research should identify absolute or relative sNfL levels that correlate with long-term disease stability, including the absence of relapses and progression.

While this study presents some novel and noteworthy findings, there are limitations. As POMS is a rare disease, the size of our POMS cohort by convenience sampling was not large enough to power all of our analyses. And, as our POMS patients enrolled from a single academic referral center, some were only seen in the clinic for one-second opinion consultation; this reduces the number of patients with follow-up data, effectively decreasing the sample size for some analyses even further. Another limitation was the convenience sampling that led the pediatric control group to be poorly matched for age and BMI, two factors known to correlate with both MS diagnosis and serum biomarker measurements. In addition, we acknowledge limitations inherent to database research, including missing data and data entry errors, though we did mitigate this risk with validation through chart review. Finally, our cohort enrolled a median of about fifteen years ago, when only low-efficacy DMTs were available to most POMS patients. Therefore, our cohort has more relapsing disease activity and less highly effective treatment exposure than most patients today. This, in addition to the single-center nature of our study, limits generalizability. It will be important to re-examine the value of sNfL and sGFAP in a diverse cohort of more contemporary POMS patients, particularly those who have better-controlled MS activity on higher efficacy treatment.

Conclusions

Both sGFAP and sNfL are elevated in POMS patients compared to healthy pediatric controls. While both are higher in the setting of greater clinical and radiographic disease burden, only sNfL seems to reflect active relapsing disease activity and treatment response. However, in some apparently clinically and radiographically stable patients, sNfL remains elevated, suggesting ongoing neuroaxonal injury. This supports a potential role for sNfL in monitoring disease stability, which should be the subject of future research endeavors. The sGFAP elevation among POMS patients suggests that astrogliosis and mechanisms underlying MS progression may begin in the very early stages of multiple sclerosis, even among children.

Supplemental Material

sj-docx-1-mso-10.1177_20552173241274567 - Supplemental material for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis

Supplemental material, sj-docx-1-mso-10.1177_20552173241274567 for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis by Laura Saucier, Brian C Healy, Shrishti Saxena, Eunnindy Sanon and Tanuja Chitnis in Multiple Sclerosis Journal – Experimental, Translational and Clinical

Supplemental Material

sj-docx-2-mso-10.1177_20552173241274567 - Supplemental material for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis

Supplemental material, sj-docx-2-mso-10.1177_20552173241274567 for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis by Laura Saucier, Brian C Healy, Shrishti Saxena, Eunnindy Sanon and Tanuja Chitnis in Multiple Sclerosis Journal – Experimental, Translational and Clinical

Supplemental Material

sj-docx-3-mso-10.1177_20552173241274567 - Supplemental material for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis

Supplemental material, sj-docx-3-mso-10.1177_20552173241274567 for Glial fibrillary acidic protein and neurofilament light chain as biomarkers in pediatric multiple sclerosis by Laura Saucier, Brian C Healy, Shrishti Saxena, Eunnindy Sanon and Tanuja Chitnis in Multiple Sclerosis Journal – Experimental, Translational and Clinical

Footnotes

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr. Chitnis has served as a consultant for Siemens, Octave Biosciences and Roche Diagnostics.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Multiple Sclerosis Society (grant number Clinical Care Fellowship Award CF-2007-36764).

ORCID iD

Tanuja Chitnis

Supplemental material

Supplemental material for this article is available online.

References

Filippi

Bar-Or

Piehl

, et al. Multiple sclerosis. Nat Rev Dis Prim 2018; 4: 43.

Chitnis

Glanz

Jaffin

, et al. Demographics of pediatric-onset multiple sclerosis in an MS center population from the Northeastern United States. Mult Scler 2009; 15: 627–631.

Benson

Healy

Gorman

, et al. Elevated relapse rates in pediatric compared to adult MS persist for at least 6 years. Mult Scler Relat Disord 2014; 3: 186–193.

Gorman

Healy

Polgar-Turcsanyi

, et al. Increased relapse rate in pediatric-onset compared with adult-onset multiple sclerosis. Arch Neurol 2009; 66: 54–59.

Renoux

Vukusic

Mikaeloff

, et al. Natural history of multiple sclerosis with childhood onset. N Engl J Med 2007; 356: 2603–2613.

Khalil

Teunissen

Otto

, et al. Neurofilaments as biomarkers in neurological disorders. Nat Rev Neurol 2018; 14: 577–589.

Abdelhak

Foschi

Abu-Rumeileh

, et al. Blood GFAP as an emerging biomarker in brain and spinal cord disorders. Nat Rev Neurol 2022; 18: 158–172.

Bittner

Havrdová

, et al. The potential of serum neurofilament as biomarker for multiple sclerosis. Brain 2021; 144: 2954–2963.

Rosso

Gonzalez

Healy

, et al. Temporal association of sNfL and gad-enhancing lesions in multiple sclerosis. Ann Clin Transl Neurol 2020; 7: 945–955.

10.

Momtazmanesh

Shobeiri

Saghazadeh

, et al. Neuronal and glial CSF biomarkers in multiple sclerosis: a systematic review and meta-analysis. Rev Neurosci 2021; 32: 573–595.

11.

Malmestrom

Haghighi

Rosengren

, et al. Neurofilament light protein and glial fibrillary acidic protein as biological markers in MS. Neurology 2003; 61: 1720–1725.

12.

Axelsson

Malmeström

Gunnarsson

, et al. Immunosuppressive therapy reduces axonal damage in progressive multiple sclerosis. Mult Scler J 2014; 20: 43–50.

13.

Gunnarsson

Malmeström

Axelsson

, et al. Axonal damage in relapsing multiple sclerosis is markedly reduced by natalizumab. Ann Neurol 2011; 69: 83–89.

14.

Wendel

E-M

Bertolini

Kousoulos

, et al. Serum neurofilament light-chain levels in children with monophasic myelin oligodendrocyte glycoprotein-associated disease, multiple sclerosis, and other acquired demyelinating syndrome. Mult Scler J 2022; 28: 1553–1561.

15.

Kuhle

Chitnis

Banwell

, et al. Plasma neurofilament light chain in children with relapsing MS receiving teriflunomide or placebo: a post hoc analysis of the randomized TERIKIDS trial. Mult Scler 2023; 29: 385–394.

16.

de Vries RM

vdV

Wong

YYM

Mescheriakova

, et al. High neurofilament levels are associated with clinically definite multiple sclerosis in children and adults with clinically isolated syndrome. Mult Scler J 2019; 25: 958–967.

17.

Reinert

Benkert

Wuerfel

, et al. Serum neurofilament light chain is a useful biomarker in pediatric multiple sclerosis. Neurol Neuroimmunol Neuroinflammation 2020; 7(4): e749.

18.

Disanto

Barro

Benkert

, et al. Serum Neurofilament light: A biomarker of neuronal damage in multiple sclerosis. Ann Neurol 2017; 81: 857–870.

19.

Varhaug

Barro

Bjørnevik

, et al. Neurofilament light chain predicts disease activity in relapsing-remitting MS. Neurol Neuroimmunol NeuroInflammation 2018; 5: 1–8.

20.

Cantó

Barro

Zhao

, et al. Association between serum neurofilament light chain levels and long-term disease course among patients with multiple sclerosis followed up for 12 years. JAMA Neurol 2019; 76: 1359–1366.

21.

Bhan

Jacobsen

Myhr

, et al. Neurofilaments and 10-year follow-up in multiple sclerosis. Mult Scler J 2018; 24: 1301–1307.

22.

Rosso

Healy

Saxena

, et al. MRI lesion state modulates the relationship between Serum neurofilament light and age in multiple sclerosis. J Neuroimaging 2021; 31: 388–393.

23.

Kuhle

Barro

Disanto

, et al. Serum neurofilament light chain in early relapsing remitting MS is increased and correlates with CSF levels and with MRI measures of disease severity. Mult Scler 2016; 22: 1550–1559.

24.

Kuhle

Kropshofer

Haering

, et al. Blood neurofilament light chain as a biomarker of MS disease activity and treatment response. Neurology 2019; 92: E1007–E1015.

25.

Norgren

Sundstrom

Svenningsson

, et al. Neurofilament and glial fibrillary acidic protein in multiple sclerosis. Neurology 2004; 63: 1586–1590.

26.

Högel

Rissanen

Barro

, et al. Serum glial fibrillary acidic protein correlates with multiple sclerosis disease severity. Mult Scler J 2020; 26: 210–219.

27.

Thompson

Banwell

Barkhof

, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol 2018; 17: 162–173.

28.

Banwell

Bennett

Marignier

, et al. Diagnosis of myelin oligodendrocyte glycoprotein antibody-associated disease: international MOGAD panel proposed criteria. Lancet Neurol 2023; 22: 268–282.

29.

Manouchehrinia

Piehl

Hillert

, et al. Confounding effect of blood volume and body mass index on blood neurofilament light chain levels. Ann Clin Transl Neurol 2020; 7: 139–143.

30.

Novakova

Zetterberg

Sundström

, et al. Monitoring disease activity in multiple sclerosis using serum neurofilament light protein. Neurology 2017; 89: 2230–2237.

31.

Axelsson

Malmeström

Nilsson

, et al. Glial fibrillary acidic protein: a potential biomarker for progression in multiple sclerosis. J Neurol 2011; 258: 882–888.

32.

Abdelhak

Huss

Kassubek

, et al. Serum GFAP as a biomarker for disease severity in multiple sclerosis. Sci Rep 2018; 8: 14798.

33.

Rotstein

Healy

Malik

, et al. Evaluation of no evidence of disease activity in a 7-year longitudinal multiple sclerosis cohort. JAMA Neurol 2015; 72: 152–158.

34.

Kappos

De Stefano

Freedman

, et al. Inclusion of brain volume loss in a revised measure of “no evidence of disease activity” (NEDA-4) in relapsing-remitting multiple sclerosis. Mult Scler 2016; 22: 1297–1305.

35.

Rotstein

Solomon

Sormani

, et al. Association of NEDA-4 with no long-term disability progression in multiple sclerosis and comparison with NEDA-3: a systematic review and meta-analysis. Neurol Neuroimmunol Neuroinflammation 2022; 9: e200032.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.02 MB

0.03 MB