Serum Neurofilament Light Chain in Relapsing Multiple Sclerosis Patients on a Ketogenic Diet

Abstract

BACKGROUND:

Ketogenic diets have anti-inflammatory and neuroprotective properties which make these diets an attractive complimentary treatment approach for patients living with multiple sclerosis (MS). The objective of this study was to assess the impact of ketogenic diets on neurofilament light chain (NfL), a biomarker of neuroaxonal injury.

METHODS:

Thirty-nine subjects with relapsing MS completed a 6-month ketogenic diet intervention. NfL levels were assayed at both baseline (pre-diet) and 6-months on-diet. In addition, ketogenic diet study participants were compared to a cohort (n=31) of historical, untreated MS controls.

RESULTS:

Baseline (pre-diet) mean NfL was 5.45 pg/ml (95% CI 4.59 – 6.31). After 6 months on ketogenic diet, mean NfL was not significantly changed (5.49 pg/ml; 95% CI 4.82 – 6.19). Compared to untreated MS controls (mean 15.17 pg/ml), NfL levels for the ketogenic diet cohort were relatively low. MS subjects with higher levels of ketosis (as measured by serum beta-hydroxybutyrate) exhibited greater reductions in NfL between baseline and 6-months on ketogenic diet.

CONCLUSIONS:

Ketogenic diets do not worsen biomarkers of neurodegeneration in relapsing MS patients, with stable, low levels of NfL observed throughout the diet intervention. Subjects with greater biomarkers of ketosis experienced a higher degree of improvement in serum NfL.

Clinical Trial Identifier:

NCT03718247 – “Utilization of the Ketogenic Diet in Patients with Relapsing-Remitting MS” https://clinicaltrials.gov/ct2/show/NCT03718247

INTRODUCTION

Ketogenic diets rely on high consumption of healthy fats, minimal carbohydrate intake, and adequate protein provision to mimic a fasting state. On these diets, the body utilizes fatty acids as the main source of fuel, and energy production is shifted from glycolytic to oxidative phosphorylation energetics. Ketogenic diets have neuroprotective, anti-inflammatory properties that may prove useful for neurodegenerative disorders, such as multiple sclerosis (MS). These properties include reduction of reactive oxygen species (ROS) generation, upregulation of antioxidant pathways, suppression of the NLRP3 inflammasome, activation of neuroprotective macrophages, and suppression of pro-inflammatory cytokine production (1–5).

In the murine model of MS, ketogenic diets reduce demyelinating lesion volumes and hippocampal atrophy, correlating with attenuation of pro-inflammatory cytokines and ROS production (6). We previously demonstrated the safety and feasibility of ketogenic diets in a relapsing MS population, in addition to several potential clinical benefits (7, 8). Limited available work suggests that ketogenic diets exhibit a neuroprotective effect in MS, with significant reductions in serum neurofilament light chain (NfL), a biomarker of neuroaxonal degeneration (9).

As a metric of safety and potential efficacy, we assessed change in NfL levels from a cohort of clinically-stable relapsing MS patients who underwent ketogenic diet intervention for 6-months.

METHODS

Sixty-five subjects with relapsing-remitting MS enrolled into the ketogenic diet study. Inclusion and exclusion criteria have been previously outlined (8). Importantly, patients had to exhibit clinical- and neuroimaging-stability on their current disease modifying therapy (DMT), if any, for at least 6 months prior to enrolment.

Subjects were evaluated at each study visit with anthropometric measures (body mass index, waist circumference), clinical outcome metrics (MS Functional Composite, symbol digit modality test, 6-minute walk), patient-reported measures (self-reported fatigue, depression, and quality of life), and laboratory testing. Laboratory testing was checked at baseline, 3-, and 6-months on-diet and included fasting lipid panels, 25-hydroxyvitamin D, leptin/adiponectin, insulin, glucose, hemoglobin a1C, and bicarbonate. Serum beta-hydroxybutyrate (BHB), a ketone body that is produced secondary to fatty acid metabolism, was checked at 3- and 6-months on diet. For the last 45 subjects enrolled, serum was frozen and stored at baseline (pre-diet) and at 6-months on-diet. Of the eligible 45 subjects, 39 attended both the baseline and 6-month visit and were thus included for analysis.

Serum NfL was measured using the single molecular array (Simoa) NF-Light Advantage Kit on the SR-X instrument (Quanterix, USA) according to the manufacturer’s instructions. Each sample was plated in replicates at 1:4 dilution. Archived samples from thirty-one untreated individuals with MS, obtained at the time of diagnosis and available from a separate natural history study, were used as historical controls. Estimated Disability Status Scale (EDSS) score and demographics were obtained via chart review at the time of sample collection. The intra-assay coefficient of variation (CV) was 9.5% and inter-assay CV was 8.0% for the NfL assay. The study was approved by the University of Virginia Institutional Review Board. Subjects provided informed consent before conducting any study-related procedures.

Statistical Analysis

Statistical analyses were conducted using SAS 9.4 software. Descriptive statistics were calculated and reported using standard appropriate statistics (e.g., means, frequencies, and t-statistic). Paired t-tests and χ2 tests were used as appropriate for continuous and categorical variables to provide in-group comparisons from baseline to 6-months on diet. Spearman correlation coefficients were used to assess the relationship between change in NfL and change in outcome metrics. A 2-sided p value of <0.05 was defined as statistically significant.

Any data not published within the article are available, and the anonymized data will be shared by request from any qualified investigator.

RESULTS

The baseline mean NfL level for the ketogenic diet study cohort was 5.45 pg/ml (95% CI 4.59 – 6.31). Compared to historical, untreated MS controls (mean 15.17 pg/ml, 95% CI 9.83 – 20.52), the baseline mean NfL levels for the study cohort were relatively low, indicating low levels of axonal damage. Demographics of the two study cohorts are found in Table 1.

Table 1:

Demographics of historical, untreated control MS cohort and ketogenic diet study cohort.

	Historical Controls (n=31)	Ketogenic Diet Participants (n=39)
Age	37.3 ± 10.9	39.4 ± 7.9
Sex, n (% female)	22 (71%)	33 (85%)
Race
n (% White)	18 (58%)	33 (85%)
n (% Black)	13 (42%)	6 (15%)
Hispanic, n (% yes)	1 (2.5%)	3 (8%)
MS Duration	3.0 ± 4.4	8.0 ± 5.9
Last Clinical Relapse (years)	0.6 ± 0.6	3.8 ± 3.7
Baseline EDSS Score	2.4 ± 1.8	2.4 ± 0.9
Current DMT, n (%)
None	31 (100%)	1 (3%)
Interferon/Glatiramer acetate	-----	7 (18%)
Teriflunomide	-----	3 (8%)
Dimethyl Fumarate	-----	4 (10%)
Fingolimod	-----	6 (15%)
Anti-CD20 therapy	-----	4 (10%)
Natalizumab	-----	12 (31%)
Alemtuzumab	-----	2 (5%)

After 6 months of ketogenic diet, the mean NfL level for the cohort was 5.49 (95% CI 4.82 – 6.19). There was no statistical difference between baseline and treatment levels, indicating that serum NfL remained stable following 6 months of ketogenic diet (Figure 1).

Figure 1:

Neurofilament light chain levels from the ketogenic diet study cohort (N=39), shown at baseline and at 6-months on ketogenic diet. Boxplots demonstrate median with interquartile range. Whiskers represent the range.

The change in NfL did not correlate with change in clinical outcome metrics or patient-reported outcomes. Change in serum NfL had a moderate, direct correlation with serum bicarbonate levels (r=0.31; p=0.05), suggesting that a greater decrease in bicarbonate over time was associated with a greater decline in NfL levels.

Subjects who exhibited a serum beta-hydroxybutyrate (BHB) level of ≥ 1.0 mmol/L at 3-months (n=10) and 6-months (n=5) on diet had a greater degree of improvement in serum NfL at 6-months compared to subjects with a BHB level of <1.0 mmol/L at 3-months (−1.2 ± 3.1 vs +0.5 ± 1.8; p<0.05) and 6-months (−2.3 ± 4.3 vs +0.4 ± 1.7; p=0.01).

DISCUSSION

Serum NfL, a biomarker of neuroaxonal injury, correlates with clinical disease activity and markers of inflammation in people living with MS (10). An association between a rise in NfL and worsening disability has been demonstrated (11) and DMT treatment for MS is associated with reduced NfL levels (12). In our study, serum NfL remained low and stable throughout the study’s dietary intervention, further supporting the conclusion that ketogenic diet does not worsen peripherally-derived biomarkers of neurodegeneration in individuals with MS.

At baseline, the mean NfL for the study cohort was low, particularly in comparison with an untreated MS control population. The low levels of NfL for the ketogenic diet cohort is secondary to our study’s recruited subject “phenotype” - a clinically- and radiologically-stable cohort with the majority of subjects (97%) on a disease modifying therapy. There was no significant difference between baseline and treatment NfL levels after 6 months on the ketogenic diet, indicating that a biomarker of neuronal injury in MS remained stable throughout the diet intervention.

BHB, a ketone body produced during fatty acid metabolism, provides a measurement of the degree of ketosis. Subjects who had notably elevated serum BHB levels on diet had improved levels of serum NfL compared to those with minimal BHB elevations. This could suggest that the degree of ketosis is directly related to attenuation of neuroaxonal injury in patients with MS. Likewise, lower serum bicarbonate levels can be secondary to greater levels of ketosis. To this end, reduction in serum bicarbonate on ketogenic diet was directly correlated with change in serum NfL. Taken together, these findings suggest that ketosis itself, as opposed to macronutrient restriction(s) and/or weight loss, is the important mediator between ketogenic diet and neuroprotection.

Bock and colleagues tested an “adapted” ketogenic diet in MS subjects (n=17) and showed a significant reduction in mean serum NfL from baseline (8.5 pg/mL) to 6 months (7.1 pg/mL) with comparison to a control diet (n=9) and caloric restriction (n=14) (9). Importantly, the majority of subjects in this study’s ketogenic diet group were untreated (29%) or on low-efficacy DMTs (48%). Consequently, baseline NfL levels were higher than those noted in our subjects. This study also employed more liberal carbohydrate restriction (<50 grams) compared to our study (net carbohydrates <20 grams) but did not provide longitudinal BHB measurements; thus degree of ketosis cannot be compared between studies. Nonetheless, this study shows the neuroprotective potential of ketogenic diets, particularly in patients that are untreated or on lower-efficacy DMTs.

Limitations of our study include the lack of a longitudinal control cohort. Further, our study provides an assessment of a clinically-stable MS population, thus our findings are not reflective of an actively-relapsing or progressive MS population. Serum BHB levels were drawn at the 3- and 6-month study visits and reflect degree of serologic ketosis at the time of blood draw. This test does not inform on the average ketosis across the study duration. More frequent sampling of serum BHB may yield important insight for next-step assessments of the ketogenic diet’s effect on biomarkers of neurodegeneration.

Our findings add to the current science of ketogenic diets in MS, providing further evidence of their short-term safety in terms of neuroaxonal injury. Biomarkers denoting greater levels of ketosis correlate with the degree of improvement in NfL. Future confirmatory work will be necessary, particularly longitudinal assessments of MS patients who are untreated or living with clinically-active MS.

Highlights.

• Ketogenic diets are safe in MS patients with sustained, low NfL levels while on diet

• Those with greater evidence of serologic ketosis exhibit greater improvement in NfL

• Reduction in serum bicarbonate on ketogenic diet directly correlates with NfL level