Multiple Sclerosis Journal

Multiplehttp://msj.sagepub.com/ Sclerosis Journal

Antibodies to Epstein-Barr virus and MRI disease activity in multiple sclerosis Silje Kvistad, Kjell-Morten Myhr, Trygve Holmøy, Søren Bakke, Antonie G Beiske, Kristian S Bjerve, Harald Hovdal, Kristin I Løken-Amsrud, Finn Lilleås, Rune Midgard, Gro Njølstad, Tom Pedersen, Jurate Saltyte Benth, Stig Wergeland and Øivind Torkildsen Mult Scler published online 19 May 2014 DOI: 10.1177/1352458514533843 The online version of this article can be found at: http://msj.sagepub.com/content/early/2014/05/16/1352458514533843

Published by: http://www.sagepublications.com

On behalf of: European Committee for Treatment and Research in Multiple Sclerosis

Americas Committee for Treatment and Research in Multiple Sclerosis

Pan-Asian Committee for Treatment and Research in Multiple Sclerosis

Latin American Committee on Treatment and Research of Multiple Sclerosis

Additional services and information for Multiple Sclerosis Journal can be found at: Email Alerts: http://msj.sagepub.com/cgi/alerts Subscriptions: http://msj.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Downloaded from msj.sagepub.com at Universitetsbiblioteket i Bergen on May 21, 2014

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - May 19, 2014 What is This?

Downloaded from msj.sagepub.com at Universitetsbiblioteket i Bergen on May 21, 2014

533843

research-article2014

MSJ0010.1177/1352458514533843Multiple Sclerosis JournalS Kvistad, KM Myhr

MULTIPLE SCLEROSIS MSJ JOURNAL

Research Paper

Antibodies to Epstein-Barr virus and MRI disease activity in multiple sclerosis Silje Kvistad, Kjell-Morten Myhr, Trygve Holmøy, Søren Bakke, Antonie G Beiske, Kristian S Bjerve, Harald Hovdal, Kristin I Løken-Amsrud, Finn Lilleås, Rune Midgard, Gro Njølstad, Tom Pedersen, Jūratė Šaltytė Benth, Stig Wergeland and Øivind Torkildsen

Abstract Background: Previous reports indicate an association between Epstein-Barr virus (EBV) antibody levels and multiple sclerosis (MS) disease activity, but the results have been conflicting. Objectives: The objective of this paper is to study if EBV antibody levels reflect MRI disease activity in MS and examine the potential for EBV antibody levels as biomarkers for treatment response. Methods: A total of 87 MS patients were followed for two years prior to and during interferon beta (IFNB) treatment, with MRI examinations and serum measurement of IgM and IgG antibodies to viral capsid antigen (VCA), EBV nuclear antigen 1 (EBNA-1) and early antigen (EA). Associations between EBV antibody levels and MRI activity were assessed by a logistic regression model. Results: Higher anti-EBNA-1 IgG levels were associated with increased MRI activity, OR = 2.95 (95% CI 1.07–8.10; p = 0.036) for combined unique activity (CUA; the sum of T1Gd+ lesions and new or enlarging T2 lesions). Although most patients were anti-VCA IgM negative, there was an inverse association, OR = 0.32 (95% CI 0.12–0.84; p = 0.021) with CUA during IFNB treatment. Conclusions: This study supports an association between anti-EBNA-1 IgG levels and MS disease activity. We also found an inverse association with anti-VCA IgM levels during IFNB treatment not previously described, indicating anti-VCA IgM as a possible biomarker for IFNB treatment response.

Multiple Sclerosis Journal 1–8 DOI: 10.1177/ 1352458514533843 © The Author(s), 2014. Reprints and permissions: http://www.sagepub.co.uk/ journalsPermissions.nav

Correspondence to: Silje Kvistad Norwegian MS Competence Centre, Haukeland University Hospital, Jonas Lies vei 65, 5053 Bergen, Norway. [email protected] Kjell-Morten Myhr Norwegian MS Competence Centre, Haukeland University Hospital, Norway/KG Jebsen Center for MS Research, University of Bergen, Norway/Norwegian MS Registry and Biobank, Haukeland University Hospital, Norway Trygve Holmøy Institute of Clinical Medicine, University of Oslo, Norway/Akershus University Hospital, Norway Søren Bakke Oslo University Hospital, Norway

Keywords: Multiple sclerosis, relapsing–remitting, MRI, interferon beta Date received: 21 August 2013; revised: 5 March 2014; accepted: 7 April 2014

Introduction Multiple sclerosis (MS) is an inflammatory, demyelinating disease of the central nervous system (CNS). The cause of the disease is unknown, but it develops as a result of a complex interplay between genetic and environmental risk factors.1 One of the most attractive and studied environmental risk factor candidates is the Epstein-Barr virus (EBV).2–5 The risk of MS in seronegative individuals is very low.6 Recent studies have questioned whether it is possible to develop the disease without previous infection with the virus.7,8 It has also been suggested that EBV antibodies are associated with MS disease activity,9,10 but this remains controversial.11–13 If changes in EBV antibody levels reflect changes in disease activity, these antibodies could possibly act as surrogate markers for

MS disease activity. We have previously performed a prospective study evaluating clinical and magnetic resonance imaging (MRI) disease activity during omega-3 fatty acids supplementation and subcutaneous interferon beta-1a (IFNB) treatment of human leukocyte antigen (HLA)-DRB1-typed relapsing– remitting MS (RRMS) patients (The OFAMSstudy).14,15All patients were followed for two years with repeated MRI scans and serum sampling. In this sub-study, prospectively collected serum samples from this cohort were analyzed for EBV antibodies to assess the association between serum EBV antibody levels and MRI disease activity and to determine whether EBV antibody levels have the potential to be used as a surrogate marker for IFNB treatment response in RRMS patients.

Antonie G Beiske MS Centre Hakadal, Norway Kristian S Bjerve St. Olav’s Hospital, Trondheim University Hospital, Norway/Children’s and Women’s Health, Norwegian University of Science and Technology, Norway Harald Hovdal St Olav’s Hospital, Trondheim University Hospital, Norway Kristin I Løken-Amsrud Institute of Clinical Medicine, University of Oslo, Norway/Innlandet Hospital Trust, Norway Finn Lilleås Curato Oslo, Norway Rune Midgard Molde Hospital, Norway/ Unit for Applied Clinical Research, Norwegian University of Science and Technology, Norway Gro Njølstad Haukeland University Hospital, Norway

http://msj.sagepub.com 1


Multiple Sclerosis Journal Tom Pedersen Unilabs Drammen, Norway Jūratė Šaltytė Benth Institute of Clinical Medicine, University of Oslo, Norway/Helse Sør-Øst Health Services Research Centre, Akershus University Hospital, Norway Stig Wergeland Øivind Torkildsen Norwegian MS Competence Centre, Haukeland University Hospital, Norway/KG Jebsen Center for MS Research, University of Bergen, Norway

Material and methods Study population and design The original OFAMS study was a randomized, double-blind, placebo-controlled trial that included 92 Norwegian patients with RRMS according to the McDonald criteria.14 The inclusion criteria were age 18–55 years and Expanded Disability Status Scale (EDSS) score ≤5. All patients had active disease with ≥1 relapse or new T1-weighted gadolinium-enhancing (T1Gd+) or T2-weighted (T2) lesions on MRI in the year prior to inclusion. None of the patients used immunomodulatory drugs at inclusion. At baseline the patients were randomized to receive either Triomar™ capsules (Pronova Biocare, Sandefjord, Norway) containing 1350 mg of eicosapentaenoic acid (EPA) and 850 mg of docosahexaenoic acid (DHA) or placebo (corn oil). From month 6 all patients received subcutaneous injections of IFNB three times weekly. Serum was collected at baseline and at months 1, 3, 6, 7, 9, 12, 18 and 24, and MRI was performed at baseline and monthly for nine months and then after12 and 24 months. For this substudy we analyzed paired MRI and serum samples collected at baseline and at months 6, 12 and 24, while EDSS evaluations were paired with serum analyses at baseline and at months 6, 12, 18 and 24. Clinical examination EDSS scorings and relapse recordings were conducted by experienced neurologists. A relapse was defined as the appearance of new or worsening of old neurologic symptoms or signs, in the absence of fever, persisting for more than 48 hours and causing objective changes on neurologic examination. EBV serology Immunoglobulin (Ig)G antibodies to viral capsid antigen (VCA), Epstein-Barr virus nuclear antigen 1 (EBNA-1) and early antigen (EA) and IgM antibodies to VCA were measured with the Liaison ® quantitative chemiluminescent assay according to instructions of the manufacturer, with serum dilution 1:100, and expressed as relative light units (RLU). This assay has been proven to have high sensitivity and specificity for detection of anti-EBNA-1 IgG.16Anti-VCA IgM and anti-VCA IgG were considered negative if 43 U/ml. Anti-EBNA-1 IgG was considered negative if 38 U/ml. Anti-EA IgG was considered negative if 40 U/ml. The upper reference range was 750 U/ml for anti-VCA

IgG and 600 U/ml for anti-EBNA-1 IgG. The antibody levels in this cohort were high for EBNA-1 IgG, and two-thirds of the patients had test results above the upper reference range. To detect differences accounting for the higher levels, RLU values were therefore used in the primary analyses. As the correlation between RLU and U/ml is less clear above the upper reference range, we also included an additional analysis using U/ml, giving all measurements above the upper reference limit the same value (600 U/ml). HLA-DRB1 typing The HLA-DRB1 status was determined by DNA sequencing using the SeCoreLoc DRB1 SEQ kit (Invitrogen, Carlsbad, CA, USA) at the Department of Immunology, Oslo University Hospital, Rikshospitalet. Patients carrying at least one DRB1*15 allele were considered HLA-DRB1*15 positive. MRI The MRIs were performed according to a standardized protocol comprising T2-weighted and T1-weighted Gd+ scans using a standard head coil with a 1.5 Tesla MRI unit. Blinded assessments of T1Gd+ lesions, T2 lesions, and combined unique activity (CUA; the sum of T1Gd+ lesions and new or enlarging T2 lesions) were conducted by two experienced neuroradiologists. Missing values A total of nine MRI scans (one during study months 1–6 and eight during study months 7–24) were missing. Seven blood samples for EBV analyses were missing for months 0–6 and 40 for months 7–24. Ethylenediaminetetraacetic acid (EDTA) blood for HLA-DRB1 typing was missing from four patients. EDSS scores were missing from two patients at month 24. Missing values were not replaced. Statistical analyses Demographic and clinical characteristics at baseline are presented as mean values with standard deviations (SD). The MRI outcomes were skewed toward none or one lesion and therefore dichotomized as present or absent. EBV antibody levels were skewed and exhibited large variance. These were therefore log10transformed for further analyses. There was a certain degree of intra-individual variation observed in EBV antibody levels, as measured by intra-class correlation coefficient (ICC). The association between EBV

2 http://msj.sagepub.com


S Kvistad, KM Myhr et al. Table 1. Demographic and clinical characteristics at baseline. Patients, n (%) Female Male Caucasian, n (%) HLA-DRB1*15-positivea, n (%) Proportion of T1Gd+ lesions, n, (%) Age, mean years (SD) Years since diagnosis, mean (SD) Years since first symptom, mean (SD) EDSS score, mean (SD)

87 (100) 57 (65.5) 30 (34.5) 87 (100.0) 58 (66.7) 45 (51.7) 38.7 (8.36) 1.89 (3.1) 5.68 (5.62) 1.9 (0.84)

an

= 84; T1Gd+: gadolinium-enhancing T1 lesions; EDSS: Expanded Disability Status Scale.

antibody levels and MRI activity was therefore assessed by a logistic regression model for repeated measurements to estimate the odds ratio (OR) for simultaneous MRI activity associated with 10-fold increase in each antibody level (RLU) and antibody activity (U/ml). The SAS GLIMMIX procedure was used to fit a model with random intercepts for patients and fixed effects for antibody levels. Only paired measurements of EBV antibody levels and MRI outcomes were included in the regression model. The effect of IFNB treatment was examined by performing the same analyses on two sub-sets of data, before and after the treatment. Because of log10-transformation of serological markers, ORs for MRI activity for 10-fold increase in EBV antibody levels were estimated. HLA-DRB1*15 status, body mass index (BMI), age, gender, vitamin D, neutralizing antibodies against IFNB (NAB) and omega-3 levels were entered in the regression models to adjust for possible confounding. The VCA IgM analyses were also adjusted for EBNA-1 IgG. The statistical analyses were conducted using SPSS version 20.0 and SAS version 9.2. P values less than 0.05 were considered significant. Results Study population Combined serum analyses and MRI scans from 87 patients were available for analysis. The demographic and baseline characteristics are shown in Table 1. EBV seropositivity A total of 403 serum samples were analyzed in 87 patients, all of whom (100%) showed serologic

evidence of a previous EBV infection. The antibody levels were high for most parameters, with a mean RLU value of 181,097 (744–985,837, SD: 249,171) for anti-EA IgG, 1,440,677 (2436–17,191,742, SD: 1,102,240) for anti-EBNA-1 IgG, 7 112 (940–52,207, SD: 7637) for anti-VCA IgM and 935,201 (27,693– 10,811,595, SD: 699,053) for anti-VCA IgG. The ICC varied between 89.6 (VCA IgG) and 96.3 (EA), implying that between 3.7% and 10.4% of the variability was due to intra-individual variations. EBNA-1 IgG antibodies A total of 393 (97.5%) of the serum samples were positive for anti-EBNA-1 IgG. One patient had negative anti-EBNA-1 IgG tests on all occasions, whereas 75 had values above the upper reference range. In 20 patients, we were able to repeat the measurement at additional 1:20 dilution. Sixteen of these had RLU values above the upper reference range in the initial analysis. There was a strong correlation (Pearson correlation coefficient 0.995) between RLU in the original undiluted sample and antibody activity (expressed as U/ml) after additional 1:20 dilution (data not shown). Because of the large number of measurements of anti EBNA-1 IgG above the upper reference limit, and because RLU values above the upper reference limit correlated strongly with U/ml, we first analyzed the relationship with MRI using antibody activity measured as RLU. Analyzing the whole study period, a 10-fold increase in anti-EBNA-1 IgG RLU levels was associated with an increased occurrence of new T1Gd+ lesions, OR = 2.20 (95% confidence interval (CI) 1.04–4.67; p = 0.040), new T2 lesions, OR = 2.90 (95% CI 1.03–8.21; p = 0.045), and the combined unique MRI activity (CUA), OR = 2.95 (95% CI 1.07–8.10; p = 0.036) (Table 2). After adjusting for gender, age, HLA-DRB1*15 status and BMI, the associations remained significant and tended to be stronger—for new T1Gd+ lesions; OR = 3.05 (95% CI 1.27–7.35; p = 0.013), new T2 lesions; OR = 4.93 (95% CI 1.25–19.40; p = 0.023) and CUA; OR = 4.96 (95% CI 1.35–18.22; p = 0.016). There appeared to be an association prior to IFNB treatment, OR = 8.78 (95% CI 1.05–73.26; p = 0.045) for CUA after adjusting for gender, age, HLA status and BMI. However, no significant association was found during treatment with IFNB after adjustments, OR = 3.54 (95% CI 0.66–19.07; p = 0.138) for CUA. There were no differences in antibody levels before and after treatment with IFNB, indicating that IFNB treatment did not suppress antibody levels (eFigure



Multiple Sclerosis Journal Table 2. Odds ratios for MRI activity associated with 10-fold increase in anti-EBNA-1 IgG RLU levels. MRI measure

Total study period, n = 87

Before IFNB treatment (months 1–6)

During IFNB treatment|(months 7–24)

Odds ratio(95% CI)

p value

Odds ratio(95% CI)

p value

Odds ratio (95% CI)

p value

New T1Gd+ lesions New T2 lesions CUA

2.20 (1.04–4.67)

0.040

2.53 (1.00–6.38)

0.050

1.67 (0.41–6.82)

0.467

2.90 (1.03–8.21) 2.95 (1.07–8.10)

0.045 0.036

6.86 (0.91–51.61)a 6.54 (1.02–42.03)a

0.061 0.048

1.81 (0.57–5.75) 1.82 (0.56–5.90)

0.306 0.316

Adjusted for gender, age, BMI and HLA status Total study period, n = 82



During IFNB treatment (months 7–24)

Odds ratio(95% CI)

p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

3.05 (1.27–7.35)

0.013

3.13 (1.10–8.95)

0.034

5.15 (0.43–61.99)

0.193

4.93 (1.25–19.40) 4.96 (1.35–18.22)

0.023 0.016

9.93 (0.91–108.53)b 8.78 (1.05–73.26)b

0.060 0.045

3.34 (0.66–16.99) 3.54 (0.66–19.07)

0.144 0.138

MRI: magnetic resonance imaging; EBNA: Epstein-Barr virus nuclear antigen; RLU: relative light units; Gd+ lesions: gadolinium-enhancing lesions; CUA: combined unique activity; BMI: body mass index; HLA: human leukocyte antigen; IFNB: interferon beta; CI: confidence interval. an = 82. bn = 77.

1). Anti-EBNA-1 IgG levels at baseline and at month 6 were not associated with MRI activity the subsequent month (eTables 1(a) and (b)), thus indicating that the increase in anti-EBNA-1 IgG levels is more likely to be a downstream effect than an early phenomenon in MS. A large proportion of the anti EBNA-1 IgG measurements were above the upper reference range. Although there was a strong correlation between RLU and U/ml also above the upper reference range in 20 samples tested at additional dilution, this relationship is less clear than for values within the reference range. We therefore repeated the analysis, giving all 284 measurements above the upper reference limit this value (600 U/ml). Adjusted for gender, age, the presence of HLA-DRB1*15 and BMI, each 10-fold increase in U/ ml of anti-EBNA-1 IgG was associated with an increased occurrence of new T1Gd+ lesions; OR = 3.83 (95% CI 1.02–14.35: p = 0.047) and higher CUA; OR = 8.96 (95% CI 1.05–61.48; p = 0.045) (Table 3). We have previously shown that fluctuations in vitamin D levels are associated with radiologic disease activity, and others have shown a negative correlation between levels of 25-hydroxyvitamin D (25(OH)D) and antibody reactivity against EBNA-1.15,17 Adjusting for 25(OH)D levels in each serum sample did not, however, meaningfully alter our result (data not shown). Since our patient cohort included

subjects recruited from a previous trial of omega 3 fatty acids vs. placebo, showing no effect of omega 3 acids on MRI activity or inflammation markers in serum,14,18 we also analyzed our data adjusting for omega 3 and placebo treatment, but this did not influence our results (data not shown).

EA IgG antibodies Anti-EA IgG was positive in 100 and borderline in 57 serum samples from 40 patients. We found no significant associations between anti-EA IgG RLU levels and any of the MRI measures (eTable2). This was not influenced by IFNB treatment or by adjusting for gender, age, HLA-DRB1*15 status and BMI. VCA IgG antibodies All patients had positive VCA IgG tests on all occasions with a total of 403 positive analyses. We found no significant association between anti-VCA IgG RLU levels and any of the MRI measures (eTable3). This was independent of IFNB treatment, and was still non-significant after adjusting for gender, age, HLA-DRB1*15 status and BMI. VCA IgM antibodies Anti-VCA IgM was positive in nine and borderline in 24 serum samples from 12 patients. Including also RLU values below the lower reference range from the



S Kvistad, KM Myhr et al. Table 3. Odds ratio for new MRI disease activity assessed with 10-fold increase in anti-EBNA-1 IgG activity (U/ml). Total period n = 87

Odds ratio (95% CI)

p value

New T1 Gd+ lesions New T2 lesions CUA

2.35 (0.80–6.87) 3.96 (0.84–18.56) 3.67 (0.87–15.56)

0.119 0.081 0.078

Total period n = 82. Adjusted for gender, age, BMI and HLA status

Odds ratio (95% CI)

p value

New T1 Gd+ lesions New T2 lesions CUA

3.83 (1.02–14.35) 9.62(0.98–94.74) 8.05(1.05–61.48)

0.047 0.053 0.045

MRI: magnetic resonance imaging; EBNA: Epstein-Barr virus nuclear antigen; Gd+ lesions: gadolinium-enhancing lesions; Ig: immunoglobulin; CUA: combined unique activity; BMI: body mass index; HLA: human leukocyte antigen; CI: confidence interval.

whole study period, 10-fold increases in anti-VCA IgM RLU levels were associated with significantly less MRI disease activity as measured by CUA; OR = 0.32 (95% CI 0.12–0.84; p = 0.021), but not for separate analyses of new T1Gd+ lesions or for new T2 lesions (Table 4). After adjusting for gender, age, HLA-DRB1*15 status and BMI, the OR for CUA were 0.31 (95% CI 0.11–0.87; p = 0.026). No significant associations between anti-VCA IgM RLU levels and MRI activity prior to IFNB treatment were detected, but increasing anti-VCA IgM RLU levels were associated with significantly less MRI disease activity during IFNB treatment; OR = 0.12 (95% CI 0.02–0.79; p = 0.028) for new T1Gd+ lesions, 0.23 (95% CI 0.06–0.95; p = 0.043) for new T2 lesions and OR = 0.19 (95% CI 0.05–0.83; p = 0.028) for CUA. These associations persisted after adjustments; OR = 0.09 (95% CI 0.01–0.73; p = 0.026) for new T1Gd+ lesions, OR = 0.21 (95% CI 0.05–0.95; p = 0.042) for new T2 lesions and OR = 0.18 (95% CI 0.04–0.84; p = 0.030) for CUA. Adjusting for EBNA-1 IgG did not change these findings (data not shown). The difference in antibody levels before and after treatment with IFNB was small, indicating that IFNB treatment did not influence antibody levels (eFigure 1).

MRI and clinical disease activity The MRI disease activity, given by mean number of new lesions during the study period, is outlined in eTable4. A total of 42 relapses were recorded in 23 patients, of which 28 occurred after the IFNB treatment commenced. The median (range) EDSS score

was 2.0 (0.0–4.0) at baseline and 2.0 (0.0–6.5) at month 24. Twenty-six patients progressed ≥1 EDSS point during the whole study period, of whom 22 had confirmed progression sustained > 6 months. All patients still had a relapsing–remitting disease course, thus no patients had converted to secondary progressive MS. Neither anti-EBNA-1 IgG nor anti-VCA IgM were associated with the occurrence of relapses or the progression in EDSS. The mean anti-VCA IgM RLU level during the total study period showed a trend toward higher levels in patients who progressed ≥1 EDSS point compared to the stable patients (8436.8 compared to 7079.0, p = 0.059, independent t-test).

Neutralizing antibodies against IFNB During the study period NAB occurred in 27 patients. Adjusting for NAB status did not influence our results (eTables 5(a) and (b)). Discussion These results indicate that increasing levels of antiEBNA-1 IgG are associated with increased MRI disease activity in MS, assessed by both occurrence of new T1Gd+ and T2 lesions. The finding was consistent for analyses using antibody activity measured directly as RLU and as U/ml. Further, we detected an even stronger inverse association between anti-VCA IgM RLU levels and new MRI lesions, particularly during IFNB treatment, which has not previously been described. However, as most anti-VCA IgM RLU levels were below the reference range, this must be interpreted with care.



Multiple Sclerosis Journal Table 4. Odds ratios for MRI activity associated with 10-fold increase in anti-VCA IgM RLU levels. MRI measure

Total study period n = 87



Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value


0.51 (0.22–1.22) 0.39 (0.14–1.04) 0.32 (0.12–0.84)

0.129 0.059 0.021

0.73 (0.24–2.21) 0.63 (0.17–2.38)a 0.42 (0.11–1.61)a

0.577 0.488 0.203

0.12 (0.02–0.79) 0.23 (0.06–0.95) 0.19 (0.05–0.83)

0.028 0.043 0.028

Adjusted for gender, age, BMI and HLA status

Total study period n = 82

Odds ratio (95% CI)


0.49 (0.20–1.23) 0.38(0.14–1.09) 0.31(0.11–0.87)



p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

0.129 0.071 0.026

0.76(0.23–2.48) 0.63(0.16–2.56)b 0.41(0.10–1.71)b

0.641 0.513 0.219

0.09(0.01–0.73) 0.21(0.05–0.95) 0.18(0.04–0.84)

0.026 0.042 0.030

MRI: magnetic resonance imaging; VCA: viral capsid antigen; RLU: relative light units; Gd+ lesions: gadolinium-enhancing lesions; CUA: combined unique activity; BMI: body mass index; HLA: human leukocyte antigen; Ig: immunoglobulin; IFNB: interferon beta; CI: confidence interval. an = 82. bn = 77.

The main strength of our study is the prospective design with repeated and paired MRI scans and measurement of EBV antibodies before and during IFNB treatment in a well-characterized cohort of HLA-DR-typed RRMS patients. All MRI, biochemical and clinical assessments were performed with strict and standardized procedures. There are, however, some limitations to our study. MS patients have previously been described as having high levels of EBV viral antibodies,19,20 which were consistent with the findings in our study. Measuring antibody activity as recommended by the manufacturer yielded a large number of measurements above the upper reference limit. When using RLU levels above the upper detection levels, there is an increasing risk of less correlation with antibody levels. Re-analysis of 20 samples at additional dilution did, however, reveal a very strong correlation between RLU and U/ ml also above the upper reference limit, supporting that RLU values reflect antibody activity faithfully above the upper reference limit as well. Moreover, analyzing the same material using antibody levels expressed as U/ml, giving all measurements above the upper reference limit the same value corresponding to this limit, confirmed the results obtained using RLU levels. Unfortunately, the amount of available serum did not allow repeated measurements with higher dilution of all samples, which would have allowed a more accurate estimation of the association between MRI and antibody activity expressed as U/ml. Previous studies have found that individuals with MS have particularly high antibody reactivity against certain EBNA-1 epitopes.21,22 There is a

possibility that antibody levels against specific EBNA-1 epitopes could be better biomarkers for disease activity. Several factors may influence antiEBV levels, including smoking, which is associated with higher anti-VCA IgG and anti-EBNA-1 IgG levels.23 Although adjusted for a number of confounders, our results were not adjusted for smoking status. All patients in this study showed evidence of previous EBV infection, which confirms several other reports suggesting an EBV seroprevalence in MS patients of 100% and that EBV might play a role in MS pathogenesis.2,7,9,24–27 Several studies indicate that high levels of anti-EBNA-1 IgG increase the risk of developing MS.19,28–31 Other studies have found an association between anti-EBNA-1 IgG levels and conversion from clinical isolated syndrome (CIS) to RRMS.10,19 Whether EBNA-1 IgG levels vary with MS disease activity is a controversial topic, and only a few studies have addressed this issue. Higher titers of anti-EBNA-1 IgG were associated with the development of T1Gd+ lesions and were predictive of change in T2 lesion volume and EDSS progression in a large MS/CIS cohort followed for five years.10 This is consistent with our findings, although a recent follow-up study did not find any significant association between antibody levels and clinical disease activity.13 Unlike our study, prospective MRI examinations were not performed and subclinical disease may have been missed. This was shown in our study with a significant increase in MRI activity with increasing levels of anti-EBNA-1 IgG, but not in clinical activity measured by MS relapses or EDSS progression. In studies with a short follow-up period and few



S Kvistad, KM Myhr et al. relapses, the sensitivity for clinical activity will be low and MRI is therefore the best assessment of disease activity. Corroborating other studies, we found that most patients were positive for anti-EBNA-1 IgG (97.5%) and anti-VCA IgG (100%) while few patients were positive for anti-VCA IgM (2.2%).10,11 Using RLU values measuring the low and fluctuating levels of anti-VCA IgM we found an inverse association between anti-VCA IgM levels and MRI activity during IFNB therapy in patients with anti-VCA IgM below the lower set concentration limit, thus considered anti-VCA IgM negative. The association persisted after adjusting for anti-EBNA-1 IgG RLU levels, gender, age, HLA-DRB1*15 status, NAB and BMI, and was even stronger than the association between anti-EBNA-1 IgG levels and MRI activity. To our knowledge, no study so far has quantified the levels of anti-VCA IgM and examined the associations with disease activity, but several studies have analyzed the presence of anti-VCA IgM in serum of MS patients.10,11,24,31 There are generally few seropositive patients in each study. Two of the studies found no association between VCA IgM seropositivity and disease activity.10,11 The significance and mechanism underlying the observed association between MRI activity and the low anti-VCA IgM levels recorded in this study are currently unclear. Increasing anti-VCA IgM has low specificity and may be observed during immunosuppression or other viral infections.32 As the majority of measurements in this study were below the lower reference range, the result is less robust and may represent unspecific mechanisms rather than alterations in the specific immune response against EBV. However, if confirmed, and independent of the underlying mechanism, our finding may suggest that anti-VCA IgM is a marker of disease activity during IFNB treatment. Anti-VCA IgM may be a marker for reactivation of EBV and can indicate active viral replication.11,33 Previous studies have explored the association between EBV reactivation and disease activity. In a prospective study of 108 MS patients and 163 healthy controls, EBV reactivation was associated with disease activity,9 but other studies have not confirmed these findings.10–12 In this study we did not find any association between anti-EA IgG RLU levels and MRI activity, suggesting that EBV reactivation was not associated with disease activity, and thus not likely to explain the observed association between MRI activity and anti-VCA IgM.

There is a need for biomarkers of disease activity and treatment response in MS. Our study demonstrates that increasing anti-EBNA-1 IgG levels are associated with increased MRI disease activity. However, the association was not significant after initiation of IFNB treatment, which indicates that anti-EBNA-1 IgG is not an ideal biomarker for disease activity during IFNB treatment. Our findings of an association between anti-VCA IgM and MRI activity during IFNB treatment may implicate this as a possible biomarker for treatment response, but it needs to be addressed in further studies. Conflict of interest None declared. Funding This research received no specific grant from any funding agency in the public, commercial, or not-forprofit sectors. References 1. Compston A and Coles A. Multiple sclerosis. Lancet 2008; 372: 1502–1517. 2. Ascherio A, Munger KL and Lünemann JD. The initiation and prevention of multiple sclerosis. Nat Rev Neurol 2012; 8: 602–612. 3. Munger KL and Ascherio A. Risk factors in the development of multiple sclerosis. Expert Rev Clin Immunol 2007; 3: 739–748. 4. Tselis A. Epstein-Barr virus cause of multiple sclerosis. Curr Opin Rheumatol 2012; 24: 424–428. 5. Thacker EL, Mirzaei F and Ascherio A. Infectious mononucleosis and risk for multiple sclerosis: A meta-analysis. Ann Neurol 2006; 59: 499–503. 6. Ascherio A and Munger KL. Environmental risk factors for multiple sclerosis. Part II: Noninfectious factors. Ann Neurol 2007; 61: 504–513. 7. Pakpoor J, Disanto G, Gerber JE, et al. The risk of developing multiple sclerosis in individuals seronegative for Epstein-Barr virus: A meta-analysis. Mult Scler 2013; 19: 162–166. 8. Deuschle K, Hofmann J, Otto C, et al. Are there Epstein-Barr virus seronegative patients with multiple sclerosis? Mult Scler 2013; 19: 1242–1243. 9. Wandinger K, Jabs W, Siekhaus A, et al. Association between clinical disease activity and Epstein-Barr virus reactivation in MS. Neurology 2000; 55: 178–184. 10. Farrell RA, Antony D, Wall GR, et al. Humoral immune response to EBV in multiple sclerosis is



Multiple Sclerosis Journal associated with disease activity on MRI. Neurology 2009; 73: 32–38. 11. Buljevac D, van Doornum GJ, Flach HZ, et al. Epstein-Barr virus and disease activity in multiple sclerosis. J Neurol Neurosurg Psychiatry 2005; 76: 1377–1381. 12. Torkildsen Ø, Nyland H, Myrmel H, et al. EpsteinBarr virus reactivation and multiple sclerosis. Eur J Neurol 2008; 15: 106–108. 13. Ingram G, Bugert JJ, Loveless S, et al. Anti-EBNA-1 IgG is not a reliable marker of multiple sclerosis clinical disease activity. Eur J Neurol 2010; 17: 1386–1389. 14. Torkildsen O, Wergeland S, Bakke S, et al. Omega-3 fatty acid treatment in multiple sclerosis (OFAMS Study): A randomized, double-blind, placebocontrolled trial. Arch Neurol 2012; 69: 1044–1051. 15. Løken-Amsrud KI, Holmøy T, Bakke SJ, et al. Vitamin D and disease activity in multiple sclerosis before and during interferon-beta treatment. Neurology 2012; 79: 267–273. 16. Dobson R, Topping J and Giovannoni G. Comparison of two commercial ELISA systems for evaluating antiEBNA1 IgG titers. J Med Virol 2013; 85: 128–131. 17. Salzer J, Nyström M, Hallmans G, et al. EpsteinBarr virus antibodies and vitamin D in prospective multiple sclerosis biobank samples. Mult Scler 2013; 19: 1587–1591. 18. Holmoy T, Løken-Amsrud KI, Bakke SJ, et al. Inflammation markers in multiple sclerosis: CXCL16 reflects and may also predict disease activity. PloS One 2013; 8: e75021. 19. Lunemann JD, Tintore M, Messmer B, et al. Elevated Epstein-Barr virus-encoded nuclear antigen-1 immune responses predict conversion to multiple sclerosis. Ann Neurol 2010; 67: 159–169. 20. Munger KL, Levin LI, O’Reilly EJ, et al. Anti-EpsteinBarr virus antibodies as serological markers of multiple sclerosis: A prospective study among United States military personnel. Mult Scler 2011; 17: 1185–1193. 21. Sundström P, Nyström M, Ruuth K, et al. Antibodies to specific EBNA-1 domains and HLA DRB1*1501 interact as risk factors for multiple sclerosis. J Neuroimmunol 2009; 215: 102–107. Visit SAGE journals online http://msj.sagepub.com

SAGE journals

22. Sundqvist E, Sundström P, Lindén M, et al. EpsteinBarr virus and multiple sclerosis: Interaction with HLA. Genes Immun 2012; 13: 14–20.

23. Lucas RM, Hughes AM, Lay ML, et al. Epstein-Barr virus and multiple sclerosis. J Neurol Neurosurg Psychiatry 2011; 82: 1142–1148. 24. Wagner HJ, Hennig H, Jabs WJ, et al. Altered prevalence and reactivity of anti-Epstein-Barr virus antibodies in patients with multiple sclerosis. Viral Immunol 2000; 13: 497–502. 25. Myhr KM, Riise T, Barrett-Connor E, et al. Altered antibody pattern to Epstein-Barr virus but not to other herpesviruses in multiple sclerosis: A population based case-control study from western Norway. J Neurol Neurosurg Psychiatry 1998; 64: 539–542. 26. Bagert BA. Epstein-Barr virus in multiple sclerosis. Curr Neurol Neurosci Rep 2009; 9: 405–410. 27. Cepok S, Zhou D, Srivastava R, et al. Identification of Epstein-Barr virus proteins as putative targets of the immune response in multiple sclerosis. J Clin Invest 2005; 115: 1352–1360. 28. Csuka D, Simon D, Hobor R, et al. Serum concentration of immunoglobulin G-type antibodies against the whole Epstein-Barr nuclear antigen 1 and its aa35-58 or aa398-404 fragments in the sera of patients with systemic lupus erythematosus and multiple sclerosis. Clin Exp Immunol 2013; 171: 255–262. 29. Levin LI, Munger KL, Rubertone MV, et al. Temporal relationship between elevation of EpsteinBarr virus antibody titers and initial onset of neurological symptoms in multiple sclerosis. JAMA 2005; 293: 2496–2500. 30. Sundström P, Juto P, Wadell G, et al. An altered immune response to Epstein-Barr virus in multiple sclerosis: A prospective study. Neurology 2004; 62: 2277–2282. 31. DeLorenze GN, Munger KL, Lennette ET, et al. Epstein-Barr virus and multiple sclerosis: Evidence of association from a prospective study with long-term follow-up. Arch Neurol 2006; 63: 839–844. 32. Robertson P, Beynon S, Whybin R, et al. Measurement of EBV-IgG anti-VCA avidity aids the early and reliable diagnosis of primary EBV infection. J Med Virol 2003; 70: 617–623. 33. Hon GM, Hassan MS, van Rensburg SJ, et al. Assessment of Epstein-Barr virus in blood from patients with multiple sclerosis. Metab Brain Dis 2012; 27: 311–318.
