Yilan Liu
Research Medical Student

Research Medical Student:  Yilan Liu
Research Site:  University of Alabama at Birmingham
Mentor:  Victor W. Mark, MD

Title of Project: A New Look at the Symbol Digit Modalities Test in Multiple Sclerosis and Disabilities

Background: Reduced information processing speed (IPS) is the most common cognitive impairment in MS, related to reduced employment and physical ability. The Symbol Digit Modalities Test (SDMT) is the gold standard measure of IPS in MS (Benedict et al 2017). Although many investigators have suggested that performance on the SDMT involves multiple cognitive processes, there has been little attempt to tease apart what specific cognitive processes may affect the IPS score on the SDMT. Improved understanding of the cognitive processes that contribute to IPS on the SDMT could support future cognitive rehabilitation trials to treat impaired IPS in MS.

Objectives: Assess specific eye movement measures during the SDMT to suggest what specific cognitive processes may underlie IPS on the SDMT in MS.

Methods: We recruited a convenience sample of 38 adults with MS, without clinical oculomotor impairment, who performed the SDMT while an infrared eye tracker recorded their eye movements. Eye positions were sampled at 10Hz. Data were exported to a database for subsequent specific eye movement measures: (1) Search organization as calculated by the “best r” metric (index of primarily orthogonal search progress; Mark et al 2004); (2) total of upward saccades from the lower area on the page of test symbols to the answer key at the top, inferred as a measure of symbol working memory. Spearman’s rho was performed to assess correlations between the variables and the SDMT IPS score (correct responses in 90 seconds).

Results: The mean SDMT IPS score was 38 (SD = 14). Both best r and total of upward saccades significantly positively correlated with the SDMT IPS score (best r, rho = 0.45, p = 0.004; upward saccades, rho = 0.62, p < 0.0001).

Conclusions: Both search organization and total of upward saccades were positively correlated with the IPS score. The latter result was unexpected: the faster on the SDMT, the more often subjects checked the answer key. The results suggest that lower-scoring MS subjects, who less often checked the answer key during the test, may have become lost during their visual search, as reflected by their poorer search organization. Upward saccades may not so much represent working memory but rather as a strategy to assure successful test completion that is not effectively used by lower-scoring persons with MS. Further research will be needed to assess the criterion validity of specific SDMT eye movement measures relative to standard cognitive test assessments.

Kaleigh Olmsted
Research Medical Student

Research Medical Student:  Kaleigh Olmsted
Research Site:  University of Houston Health Science Center at Houston
Mentor:  J. William LIndsey, MD

Title of Project: Effect of Ocrelizumab on Antibody Responses to EBV

Background:   Multiple Sclerosis (MS) is an autoimmune disease that affects the central nervous system (CNS). While the exact mechanism of pathogenesis is not fully understood, it is known that MS involves an immune-mediated process wherein the body’s immune system has an abnormal response that is directed against the CNS. This immune-mediated process causes inflammation that results in damage and eventual fibrosis of nerve fibers, the surrounding, insulating myelin of nerve fibers, as well as damage to the cells that synthesize myelin causing a disruption in the neuron’s ability to depolarize and respond to stimuli.

The etiology of MS is currently unknown, however, there are several studies that associate MS with Epstein-Barr virus (EBV); a human herpes virus infecting the majority of the population worldwide. People with MS consistently exhibit increased antibody titers to EBV antigens in comparison to healthy controls.1,2 It has also been reported that a more severe initial infection with EBV (infectious mononucleosis) is correlated with an increase in an individual’s risk of developing MS, and high levels of anti-EBV IgG in asymptomatic young adults also have increased risk for developing MS later in life.3-6 EBV contains multiple distinct antigens, and while not all antigens have been studied in detail, it is suggested that in people with MS, there is an  increased antibody response to the following EBV antigens: nuclear antigen EBNA-1, small capsid protein BFRF3, and tegument protein BRRF2.7 Despite the multiple lines of evidence supporting an association between EBV and MS, the role that the virus plays in the pathogenesis of MS is poorly understood.

Multiple mechanisms by which EBV might contribute to CNS damage have been proposed. Two suggested mechanisms include direct damage to the CNS by either an active EBV infection in the CNS or a latent EBV infection which activates the innate immunity in the CNS.8-11  However, the evidence that supports EBV’s presence in the CNS is controversial.12 A less direct mechanism involving molecular mimicry as a result of cross-reactivity between CNS proteins and virus antigens has also been suggested, whereby reactivation of an EBV infection outside the CNS stimulates an autoimmune process inside the CNS.13,14 Although its overall role in MS remains to be demonstrated, it has been previously shown that antibodies to EBNA-1 cross-react with self antigen heterogeneous nuclear ribonucleoprotein L (HNRNPL).15 It has also been found that antibodies to the EBV protein BFRF3 cross-react with the cytoplasmic protein septin-9 and antibodies to BRRF2 were noted in some subjects to exhibit a cross-reactivity to DLST, however the response against this antigen was weaker.16 These findings support the suggested molecular mimicry mechanism that EBV antigens cross-react with brain antigens and the possibility that reactivation of persistent EBV infection could provoke an immune response that both controls the viral infection while simultaneously causing recurrent autoimmune damage to the CNS.

Ocrelizumab, a humanized anti-CD20 monoclonal antibody was recently approved for treatment of adults with relapsing MS or primary progressive MS. Studies have shown that ocrelizumab is a more effective therapy than therapy with interferon B-1a and this new treatment has also been effective in delaying progression in early progressive multiple sclerosis.17,18 The use of Ocrelizumab causes B cell depletion by tagging the B cells for phagocytosis and degradation. This, in concert with the increased efficacy of the drug, suggests that there is a role for B cells in the pathogenesis of MS. After initial infection, EBV enters a latent state and persists for the life of the host, and B lymphocytes are the main reservoir for latent infection.  Treatment with ocrelizumab should greatly reduce the EBV viral load. In the absence of stimulation with the EBV protein, B cells responsive to cross-reactive proteins are expected to regain self tolerance. Here in we investigate that ocrelizumab works through depletion of B lymphocytes, thereby eliminating the majority of EBV, resulting in a subsequent reduction in the anti-EBV immune response. If this is true, then patients on ocrelizumab should have decreased antibody responses to EBV antigen and the known cross-reactive human proteins. 

Methods: Samples of plasma from 36 patients diagnosed with MS were taken on three different occasions: before treatment with ocrelizumab, six months post treatment, and twelve months post treatment. 26 of those patients had received treatment with ocrelizumab, while the remaining 10 patients were under standard treatment without ocrelizumab, the control group. These samples were previously collected and were stored at -80oC. Antigens for ELISA and Western blots included recombinant EBV and human proteins. The vectors for these proteins and stocks of transfected bacteria were used for protein production for the immunoassays. Selected EBV protein, BFRF3, was produced in recombinant form in E. coli with an N-terminal hexahistidine tag using pET-45b(+) vector as previously described (Lindsey et al., 2016)15 and selected human protein, septin-9, was produced as described (Dooley et al., 2016)7. For each subject, the antibody responses for the three time points were measured in a single assay which included a standard curve for quantitative results using our published method for quantitative ELISA (Lindsey et al., 2016)15. We first performed a pilot assay to determine the best plasma concentration to keep the measured values in the linear range of the standard curve and all ELISA assays were done in triplicates. We used a standard curve of known amounts of human IgG for clinical use (Gammagard, Baxter) that were serial diluted and placed in wells coated with anti-IgG. For BFRF3, all wells were coated with 2.7 μl/well of recombinant BFRF3, and for total IgG concentrations all wells were coated with 1 μl/well of Goat anti-human IgG Fc fragment. Results for each individual specimen are expressed as the μg of reference IgG required to produce an OD equivalent to 1 μl of sample plasma. Intra-assay coefficients of variation were <10%. Lastly, the antibody response to septin-9, the cross-reactive antigen to EBV protein BFRF3, was measured using quantitative Western blots. 100 μl of the BFRF3 cross reactive antigen, Septin-9 was dissolved in 100 μl of Laemmli sample buffer, then 4 μl was run in each lane of a 4–12% gradient bis/tris polyacrylamide gel (Invitrogen, Carlsbad, CA) in MES buffer with colored molecular weight markers (Bio-Rad, Hercules, CA) in every 6th lane, and then blotted onto nitrocellulose membrane (Bio-Rad).Membranes were blocked with TBS containing 2.5% non-fat dried milk and 0.05% Tween 20. After blocking, membranes were cut into strips and incubated with 3 μl serum in 3 ml of blocking buffer. On each blot, known concentrations of human IgG for clinical use (Gammagard, Baxter) was used as the standards of reference. The secondary antibody was mouse anti-human IgG Fc conjugated to alkaline phosphatase (Southern Biotechnology) diluted 1:2000 in blocking buffer. After the secondary antibody, the blot was washed extensively in water, and then bound antibody was visualized with NBT/BCIP at 20 minutes. After color development, the blots were reassembled and band densities were measured by densitometry with a Hewlett Packard LaserJet 3030 scanner and Kodak 1D software. The amount of Septin-9 antibody was calculated as the band density relative to the same band in the reference serum on that blot of human IgG. The magnitude of changes in antibody response between the categories of Septin-9, recombinant BFRF3 proteins, and the total IgG were compared. Data was analyzed and comparisons of these categories were made using a One-Sample Signed Rank Test in SigmaPlot version 11.0, as none of the antibody responses were normally distributed.

Results: Data for the total IgG concentration, BFRF3 concentration and antibody response to septin-9 were not normally distributed. The mean percent change in total IgG concentration from baseline to 12 months showed an 8% decrease, with a mean percent interquartile interval of -18.4% and a mean percent third quartile of 9.2%. The 95 percent confidence interval for the population median was -13.3% to 7.84%. The difference between the median of the baseline IgG concentration and the median of the IgG concentration at 12 months is not great enough to reject the possibility that the difference is due to random sampling variability. There is not a significant difference between the two medians (P = 0.537).

The mean percent change in BFRF3 concentration from baseline to 12 months showed an 8.9% decrease, with a mean percent interquartile interval of -49.8% and a mean percent third quartile of 24.7%. The 95 percent confidence interval for the population median was -17.9% to -3.6%. The difference between the median of the baseline BFRF3 concentration and the median of the BFRF3 concentration at 12 months is not great enough to reject the possibility that the difference is due to random sampling variability. There is not a significant difference between the two medians. (P = 0.247).

The mean percent change in septin-9 concentration from baseline to 12 months showed an 3% decrease, with a mean percent interquartile interval of -30.4% and a mean percent third quartile of 39.8%. The 95 percent confidence interval for the population median was -23.8% to 32.1%. The difference between the median of the baseline septin-9 concentration and the median of the septin-9 concentration at 12 months is not great enough to reject the possibility that the difference is due to random sampling variability. There is not a significant difference between the two medians (P = 0.808).

Conclusions: In this investigation, our data showed that there was no significant change in antibody response to EBV protein BFRF3 and is cross-reactive cytoplasmic protein septin-9. This suggests that the efficacy of  treatment with Ocrelizumab does not work though the mechanism of significantly decreasing the EBV viral load and its subsequent self-reactive proteins as we hypothesized. However, in this experiment, data was based on samples taken at 12 months compared to baseline and its that possible that a time period greater than 12 months is necessary to show a significant change in antibody levels compared to baseline. Perhaps a larger sample size would have also benefited this study by increasing its statistical power, potentially providing the ability to reveal a significant difference between 12 months and baseline. In the future, it could be valuable to consider testing antibody levels to more antigens such as EBNA-1 and BRRF2 and their corresponding self-reactive proteins, heterogenous nuclear ribonucleoprotein L (HNRNPL) and mitochondrial protein DLST. It may also be of value to pursue other mechanisms of Ocrelizumab’s therapeutic effects, such as its ability to treat MS though a cell-mediated response.

Babita Bisht, PT, PhD
Whitaker Platform Presenter, 2020 CMSC Virtual Meeting

WHITAKER: Dietary Patterns and Health-Related Quality of Life of Individuals with Multiple Sclerosis

Category: Complementary and alternative therapies

Background: Individuals with multiple sclerosis (MS) look for dietary changes to improve their disease outcome. Information regarding what specific dietary changes are being implemented by individuals with MS and if these changes affect quality of life (QoL) can be useful in shaping future research.

Objectives: 1) To assess prevalence of MS-specific diets (e.g. Wahls diet, Swank diet, vegan diet) and dietary patterns of individuals with MS. 2) To investigate effects of intake of certain food groups on relapses and self-reported QoL measure.

Methods: Individuals with MS participated in an online survey and completed questions regarding intake of specific diets, frequency of intake of specific foods, relapses and Patient-Reported Outcome Measurement Information System (PROMIS) based on past 6 months. PROMIS Global scores (range 10-50) were used to measured QoL where higher scores reflect higher QoL. We estimated daily servings of foods from food frequency questions. In this cross-sectional analysis, we included 977 participants who provided complete data. 

Results: Individuals with MS with mean age 47.7 (SD 11) years and average 10 (SD 8.9) years since diagnosis participated in this study. Specific diets for MS were followed by 72% participants suggesting that most individuals with MS are implementing dietary changes. Most prevalent diets were Wahls (26%), paleolithic (16%) and anti-inflammatory (13%) diets. Some participants (11%) reported following multiple diets as well. Only 12% reported experiencing a relapse in past 6 months. Mann-Whitney U tests showed that individuals who did not experience any relapse had higher median daily intake of alcohol (0.08 vs 0.05 ounce equivalent, p=0.016) than individuals who experienced relapses in prior 6 months. PROMIS Global average scores were 33.4 (SD 6.1). Median (interquartile range) daily servings intake of different foods were as follows: total fruits and vegetables 2.2 (1.2-3.4), total dairy 0.1 (0-0.5), total grains 0.4 (0.1-1.4), total meat and fish 3.7 (1.9-5.1), total alcohol 0.1 (0-0.4) and total eggs 0.1 (0-0.4). These results show low intake of dairy, grains, eggs and alcohol among individuals with MS. Spearman’s correlation did not show any significant relationship between PROMIS Global scores and dosage of different food intake. However, Mann-Whitney U test showed that individuals who were not taking dairy had higher PROMIS scores than who were taking dairy (median 36 vs 33, p<0.001). Additionally, individuals taking alcohol had higher PROMIS scores than who were not taking any alcohol (median 34 vs 32, p=0.001). These results suggest potential beneficial effects of avoiding dairy and consuming alcohol on QoL of individuals with MS.

Conclusions: Most individuals with MS report making dietary changes to improve disease outcome. Dairy and alcohol may affect relapse rate and QoL of these individuals. Future studies should assess role of dietary changes as complimentary treatment for MS.