NEURIMMINFL-Aimsandscope 4 4 1..1

An Official Journal of the American Academy of Neurology Neurology.org/nn○Online ISSN: 2332-7812 Volume 4, Number 4, July 2017 Neuroimmunology & Neuroinﬂ ammation

Autoimmune episodic Elsberg syndrome: Dynamic regulation of ataxia in patients A rarely recognized serum aryl hydrocarbon with anti-CASPR2 cause of cauda equina receptor agonists in antibody-associated syndrome and lower multiple sclerosis encephalitis thoracic myelitis Neurology.org/nn Online ISSN: 2332-7812 Volume 4, Number 4, July 2017

fi VISION: Neurology® Neuroimmunology & Neuroinflammation will be Academy Of cers the premier peer-reviewed journal for experts in the fields of Ralph L. Sacco, MD, MS, FAAN, President fl James C. Stevens, MD, FAAN, President Elect Neuroimmunology and Neuroin ammation. Ann H. Tilton, MD, FAAN, Vice President Carlayne E. Jackson, MD, FAAN, Secretary MISSION: Neurology: Neuroimmunology & Neuroinflammation will Janis M. Miyasaki, MD, MEd, FRCPC, FAAN, Treasurer provide neurologists with peer-reviewed articles, editorials, and Terrence L. Cascino, MD, FAAN, Past President reviews to enhance patient care, education, and clinical & Executive Office, American Academy of Neurology translational research. Catherine M. Rydell, CAE, Executive Director/CEO 201 Chicago Ave Editor Minneapolis, MN 55415 Neurology: Neuroimmunology & Neuroinflammation Tel: 612-928-6100 Josep O. Dalmau, MD, PhD fi ICREA-University of Barcelona Editorial Of ce Barcelona, Spain Patricia K. Baskin, MS, Executive Editor University of Pennsylvania Kathleen M. Pieper, Senior Managing Editor, Neurology Philadelphia, PA Sharon L. Quimby, Managing Editor, Neurology® Clinical Practice Tel: 612-928-6400 Morgan S. Sorenson, Managing Editor, Neurology: Neuroimmunology fl Toll-free: 800-957-3182 (US) & Neuroin ammation Fax: 612-454-2748 Lee Ann Kleffman, Managing Editor, Neurology® Genetics [email protected] Cynthia S. Abair, MA, Senior Graphics Editor Research focus: Autoimmune and paraneoplastic disorders, described Andrea R. Rahkola, Production Editor, Neurology a new category of disorders mediated by antibodies against Robert J. Witherow, Senior Editorial Associate neuronal cell surface and synaptic proteins Karen Skaja, Senior Editorial Associate Kaitlyn Aman Ramm, Editorial Assistant Kristen Swendsrud, Editorial Assistant Editor-in-Chief Andrea Willgohs, Editorial Assistant Neurology® Robert A. Gross, MD, PhD, FAAN Publisher Professor of Neurology and of Pharmacology and Physiology Wolters Kluwer Strong Epilepsy Center Baltimore, MD University of Rochester Medical Center Rochester, NY Publishing Staff Tel: 612-928-6400 Kim Jansen, Executive Publisher Toll-free: 800-957-3182 (US) Alexandra Lazerow, Production Team Leader, AAN Journals Fax: 612-454-2748 Mark Armacost, Production Editor [email protected] Stacy Drossner, Production Associate

Editorial Inquiries Tel: 612-928-6400 Toll-free: 800-957-3182 (US) Fax: 612-454-2748 [email protected] Neurology.org/nn Online ISSN: 2332-7812 Volume 4, Number 4, July 2017

Editor Christopher A. Beck, PhD Neurology® Neuroimmunology & Neuroinflammation University of Rochester Josep O. Dalmau, MD, PhD Rochester, NY ICREA-University of Barcelona Sue Leurgans, PhD Barcelona, Spain Rush University Medical Center University of Pennsylvania Chicago, IL Philadelphia, PA Level of Evidence Evaluations Editor-in-Chief Gary S. Gronseth, MD, FAAN Neurology® University of Kansas Robert A. Gross, MD, PhD, FAAN Kansas City, KS Professor of Neurology and of Pharmacology and Physiology Strong Epilepsy Center Podcasts University of Rochester Medical Center Ted M. Burns, MD Rochester, NY University of Virginia Charlottesville, VA Deputy Editors Andrew M. Southerland, MD, MSc, Deputy Podcast Editor Neurology University of Virginia Bradford B. Worrall, MD, MSc, FAAN Charlottesville, VA Professor of Neurology and Public Health Sciences University of Virginia Ombudsman Charlottesville, VA David S. Knopman, MD, FAAN Mayo Clinic Neurology: Neuroimmunology & Neuroinflammation Rochester, MN Scott S. Zamvil, MD, PhD, FAAN Scientific Integrity Advisor Department of Neurology Robert B. Daroff, MD, FAAN University of California, San Francisco Case Western Reserve University San Francisco, CA Cleveland, OH Research focus: Human and animal models of CNS autoimmune diseases, identified mechanisms used by B cells Editorial Board Associate Editors Rohit Aggarwal, MD, MS, University of Pittsburgh, Pittsburgh, PA Neurology: Neuroimmunology & Neuroinflammation Sergio E. Baranzini, PhD, University of San Francisco, San Francisco, CA Susanne Benseler, MD, Calgary, Alberta, Canada Michael Heneka, MD Jeffrey L. Bennett, MD, PhD, FAAN, University of Colorado, Aurora, CO University of Bonn Séverine Boillée, PhD, Hôpital de la Salpêtriere, Paris, France Bonn, Germany Matthijs C. Brouwer, MD, PhD, Academic Medical Center, Amsterdam, Specialties: Neurodegeneration, stroke The Netherlands Research focus: Interplay between neurodegeneration and the immune system David B. Clifford, MD, FAAN, Washington University School of Medicine, Dennis Kolson, MD, PhD St.Louis,MO University of Pennsylvania Russell C. Dale, MBChB, MRCP, PhD, University of Sydney, Sydney, Philadelphia, PA Australia Specialty: HIV, Infectious disease Oscar H. Del Brutto, MD, FAAN, School of Medicine, Research focus: Mechanisms and determinants of HIV-induced neuronal Universidad Espiritu Santo, Samborondon, Ecuador injury Thomas G. Forsthuber, MD, University of Texas, San Antonio, TX Friedmann Paul, MD Joseph El Khoury, MD, Massachusetts General Hospital, Boston, MA Charite University Hospital Markus P. Kummer, PhD, University of Bonn, Bonn, Germany Berlin, Germany Jun Li, MD, PhD, Vanderbilt University, Nashville, TN Specialties: Multiple sclerosis, neuromyelitis optica, neuroimaging Lin Mei, MD, PhD, Georgia Regents University, Augusta, Georgia (magnetic resonance imaging, optical coherence tomography) Christopher Power, MD, University of Alberta, Alberta, Canada Research focus: CNS autoimmune disease diagnosis using novel imaging Richard M. Ransohoff, MD, Biogen Idec, Boston, MA techniques Christine Stadelmann, MD, University of Göttingen, Göttingen, Germany Israel Steiner, MD, Rabin Medical Center, Petach Tikva, Israel Lan Zhou, MD, PhD Lawrence Steinman, MD, FAAN, Stanford University, Stanford, CA Mount Sinai School of Medicine JesperTegnér,PhD,MSc,KarolinskaUniversity Hospital, Stockholm, Sweden New York, NY SilviaN.Tenembaum,MD,NationalPediatricHospital,JuanP.Garrahan, Specialty: Neuromuscular Buenos Aires, Argentina Research focus: Macrophage subsets in acute skeletal muscle injury repair Maarten J. Titulaer, MD, PhD, Erasmus Medical Center, Rotterdam, Biostatistics The Netherlands Richard J. Kryscio, PhD Ari Waisman, PhD, University of Mainz, Mainz, Germany University of Kentucky Hugh J. Willison, MBBS, PhD, University of Glasgow, Glasgow, UK Lexington, KY Gregory F. Wu, MD, PhD, Washington University in St. Louis, St. Louis, MO Neurology.org/nn Online ISSN: 2332-7812 Volume 4, Number 4, July 2017

Level of Evidence Review Team David S. Gloss II, MD, MPH&TM, Charleston, WV Melissa J. Armstrong, MD, Gainesville, FL John J. Halperin, MD, FAAN, Summit, NJ Richard L. Barbano, MD, PhD, FAAN, Rochester, NY Jason Lazarou, MSc, MD, Toronto, Ontario, Canada Richard M. Dubinsky, MD, MPH, FAAN, Kansas City, KS Steven R. Messé, MD, FAAN, Philadelphia, PA Jeffrey J. Fletcher, MD, MSc, Ann Arbor, MI Pushpa Narayanaswami, MBBS, DM, FAAN, Boston, MA Gary M. Franklin, MD, MPH, FAAN, Seattle, WA Alex Rae-Grant, MD, Cleveland, OH

Publication Information Neurology® is a registered trademark of the American Academy of Neurology (registration valid in the United States). Neurology® Neuroimmunology & Neuroinflammation (eISSN 2332-7812) is an open access journal published online for the American Academy of Neurology, 201 Chicago Ave, Minneapolis, MN 55415, by Wolters Kluwer Health, Inc. at 14700 Citicorp Drive, Bldg 3, Hagerstown, MD 21742. Business offices are located at Two Commerce Square, 2001 Market St., Philadelphia, PA 19103. Production offices are located at 351 West Camden Street, Baltimore, MD 21201-2436. © 2017 American Academy of Neurology. Neurology® Neuroimmunology & Neuroinflammation is an official journal of the American Academy of Neurology. Journal website: Neurology.org/nn, AAN website: AAN.com Copyright and Permission Information: Please go to the journal website (www.neurology.org/nn) and click the “©Request Permissions” icon for the relevant article. Alternatively, send an email to [email protected]. General information about permissions can be found here: https://www.lww.com/journal-permission. Disclaimer: Opinions expressed by the authors and advertisers are not necessarily those of the American Academy of Neurology, its affiliates, or of the Publisher. The American Academy of Neurology, its affiliates, and the Publisher disclaim any liability to any party for the accuracy, completeness, efficacy, or availability of the material contained in this publication (including drug dosages) or for any damages arising out of the use or non-use of any of the material contained in this publication. Advertising Sales Representatives: Wolters Kluwer, 333 Seventh Avenue, New York, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215-2485, [email protected] and Elizabeth S. Hall, tel: 267-804-8123, [email protected]. In Europe: Avia Potashnik, Wolters Kluwer, tel: 144 207 981 0722; 144 7919 397 933 or e-mail: [email protected]. Careers & Events: Monique McLaughlin, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: (215) 521- 8468, fax: 215-521-8801; [email protected]. Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-555-1212 (office), 215-356-2721 (mobile); [email protected]; [email protected]. Special projects: US & Canada: Alan Moore, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8638, [email protected]. International: Andrew Wible, Senior Manager, Rights, Licensing, and Partnerships, Wolters Kluwer, [email protected]. Table of Contents Neurology.org/nn Online ISSN: 2332-7812 Volume 4, Number 4, July 2017

EDITOR’S CORNER e355 Elsberg syndrome: A rarely recognized cause of cauda e373 The case for autoimmune neurology equina syndrome and lower thoracic myelitis J. Dalmau F. Savoldi, T.J. Kaufmann, E.P. Flanagan, M. Toledano, and B.G. Weinshenker

EDITORIAL e366 Aryl hydrocarbon receptor activity may serve as e358 Comparative utility of disability progression measures a surrogate marker for MS disease activity in PPMS: Analysis of the PROMiSe data set J.J. Sabatino, Jr. and S.S. Zamvil M.W. Koch, G.R. Cutter, G. Giovannoni, B.M.J. Uitdehaag, J.S. Wolinsky, M.D. Davis, J.R. Steinerman, and V. Knappertz ARTICLES e359 Dynamic regulation of serum aryl hydrocarbon receptor agonists in MS e356 Optical coherence tomography and visual evoked V. Rothhammer, D.M. Borucki, M.I. Garcia Sanchez, potentials in pediatric MS M.A. Mazzola, C.C. Hemond, K. Regev, A. Paul, A.T. Waldman, G.T. Liu, A.M. Lavery, G. Liu, W. Gaetz, P. Kivisäkk, R. Bakshi, G. Izquierdo, H.L. Weiner, and T.S. Aleman, and B.L. Banwell F.J. Quintana

e360 Both cladribine and alemtuzumab may effect MS e343 Disruption of the leptomeningeal blood barrier in via B-cell depletion neuromyelitis optica spectrum disorder D. Baker, S.S. Herrod, C. Alvarez-Gonzalez, N. Asgari, E.P. Flanagan, K. Fujihara, H.J. Kim, L. Zalewski, C. Albor, and K. Schmierer H.P. Skejoe, J. Wuerfel, H. Kuroda, S.H. Kim, E. Maillart, R. Marignier, S.J. Pittock, F. Paul, and IgG-specific cell-based assay detects potentially B.G. Weinshenker e357 pathogenic MuSK-Abs in seronegative MG S. Huda, P. Waters, M. Woodhall, M.I. Leite, e350 Gene variants of adhesion molecules act as modifiers L. Jacobson, A. De Rosa, M. Maestri, R. Ricciardi, of disease severity in MS J.M. Heckmann, A. Maniaol, A. Evoli, J. Cossins, E. Dardiotis, E. Panayiotou, A. Provatas, D. Hilton-Jones, and A. Vincent K. Christodoulou, A. Hadjisavvas, A. Antoniades, A. Lourbopoulos, M. Pantzaris, N. Grigoriadis, Mucosal biopsy shows immunologic changes of the G.M. Hadjigeorgiou, and T. Kyriakides e362 colon in patients with early MS A.M. Moser, W. Spindelboeck, H. Strohmaier, e352 Abnormal brain metabolism on FDG-PET/CT is C. Enzinger, T. Gattringer, S. Fuchs, F. Fazekas, a common early finding in autoimmune encephalitis G. Gorkiewicz, P. Wurm, C. Högenauer, and M. Khalil J.C. Probasco, L. Solnes, A. Nalluri, J. Cohen, K.M. Jones, E. Zan, M.S. Javadi, and A. Venkatesan e363 Aquaporin-4 antibodies in patients treated with natalizumab for suspected MS e353 Antiepileptic drug therapy in patients with A. Gahlen, A.-K. Trampe, S. Haupeltshofer, autoimmune epilepsy M. Ringelstein, O. Aktas, A. Berthele, B. Wildemann, A.M. Feyissa, A.S. López Chiriboga, and J.W. Britton R. Gold, S. Jarius, and I. Kleiter In Primary Periodic Paralysis Helping to turn limitation into activation

KEVEYIS can help patients gain greater control over Primary Periodic Paralysis (PPP) by decreasing the number, severity, and duration of attacks1,2 KEVEYIS® (dichlorphenamide) is the first and only FDA-approved treatment indicated for: hyperkalemic, hypokalemic, and related variants of PPP. • The efficacy of KEVEYIS has been proven in 2 clinical studies1 • KEVEYIS has a demonstrated safety profile1

Call 844-KEVEYIS for Patient Support

Learn more about PPP and KEVEYIS at KEVEYIS.com IMPORTANT SAFETY INFORMATION Concomitant Use of Aspirin Indication Anorexia, tachypnea, lethargy, and coma have been reported with concomitant use of Falls KEVEYIS® is indicated for the treatment of primary hyperkalemic periodic paralysis, dichlorphenamide and high-dose aspirin. The concomitant use of KEVEYIS and high- KEVEYIS increases the risk of falls; risk is greater in the elderly and with higher doses. primary hypokalemic periodic paralysis, and related variants. dose aspirin is contraindicated. Use with caution in patients receiving low-dose aspirin. Consider dose reduction or discontinuation of KEVEYIS in patients who experience falls while treated with KEVEYIS. Contraindications Hypokalemia KEVEYIS increases potassium excretion and can cause hypokalemia. The risk of Pregnancy and Lactation • Hypersensitivity to dichlorphenamide or other sulfonamides hypokalemia is greater when KEVEYIS is used in patients with conditions associated with Use during pregnancy only if the potential benefit justifies the potential risk to the • Concomitant use of KEVEYIS and high-dose aspirin hypokalemia (eg, adrenocortical insufficiency, hyperchloremic metabolic acidosis, or fetus. It is not known in humans whether dichlorphenamide is excreted in human milk; • Severe pulmonary disease, limiting compensation to metabolic acidosis caused respiratory acidosis), and in patients receiving other drugs that may cause hypokalemia exercise caution when administered to a nursing woman. by KEVEYIS (eg, loop diuretics, thiazide diuretics, laxatives, antifungals, penicillin, and theophylline). Adverse Reactions • Hepatic insufficiency: KEVEYIS may aggravate hepatic encephalopathy Baseline and periodic measurements of serum potassium are recommended. The most common adverse reactions seen in clinical trials (incidence ≥ 10% and greater Warnings and Precautions If hypokalemia develops or persists, consider reducing the dose or discontinuing KEVEYIS. than placebo) include paresthesias, cognitive disorder, dysgeusia, and confusional state. Hypersensitivity/Anaphylaxis/Idiosyncratic Reactions Metabolic Acidosis Please see brief summary of Prescribing Information below. Fatalities associated with the administration of sulfonamides have occurred due to KEVEYIS can cause hyperchloremic non–anion gap metabolic acidosis.Concomitant use adverse reactions including Stevens-Johnson syndrome, toxic epidermal necrolysis, of KEVEYIS with other drugs that cause metabolic acidosis may increase the severity of fulminant hepatic necrosis, agranulocytosis, aplastic anemia and other blood dyscrasias. metabolic acidosis. Pulmonary involvement can occur in isolation or as part of a systemic reaction. Baseline and periodic measurements of serum bicarbonate are recommended. Discontinue KEVEYIS at the first appearance of skin rash or any sign of immune- If metabolic acidosis develops or persists, consider reducing the dose or mediated or idiosyncratic adverse reaction. discontinuing KEVEYIS.

References: 1. KEVEYIS Prescribing Information. Feasterville-Trevose, PA: Strongbridge Biopharma; 2017. 2. Sansone VA, Burge J, McDermott MP, et al; for the Muscle Study Group. Randomized, placebo-controlled trials of dichlorphenamide in periodic paralysis. Neurology. 2016;86:1408-1416. © 2017 Strongbridge Biopharma plc STRONGBRIDGE BIOPHARMA™ is a trademark of Strongbridge Biopharma plc. KEVEYIS® is a registered trademark licensed exclusively in the US to Strongbridge Biopharma plc. KEV069-02 04/2017

Metabolic Acidosis The following are adverse reactions which have been reported for dichlorphenamide KEVEYIS can cause hyperchloremic non-anion gap metabolic acidosis. Concomitant that were serious adverse events or are not reported in the previous section of use of KEVEYIS with other drugs that cause metabolic acidosis may increase the labeling: amnesia, cardiac failure, condition aggravated, convulsion, fetal death, severity of metabolic acidosis. hallucination, nephrolithiasis, pancytopenia, psychotic disorder, renal tubular necrosis, stupor, syncope, and tremor. BRIEF SUMMARY OF PRESCRIBING INFORMATION Baseline and periodic measurement of serum bicarbonate during KEVEYIS treatment are recommended. DRUG INTERACTIONS These highlights do not include all the information needed to use Aspirin and Salicylates KEVEYIS® safely and effectively. See Full Prescribing Information If metabolic acidosis develops or persists, consideration should be given to reducing for KEVEYIS®. the dose or discontinuing KEVEYIS. KEVEYIS may cause an elevation in salicylate levels in patients receiving aspirin. Falls Anorexia, tachypnea, lethargy, and coma have been reported with concomitant use of INDICATIONS AND USAGE dichlorphenamide and high-dose aspirin. KEVEYIS is indicated for the treatment of primary hyperkalemic periodic paralysis, KEVEYIS increases the risk of falls. The risk of falls is greater in the elderly and with primary hypokalemic periodic paralysis, and related variants. higher doses of KEVEYIS. Consider dose reduction or discontinuation of KEVEYIS in Concomitant use of KEVEYIS and high dose aspirin is contraindicated. KEVEYIS should be used with caution in patients receiving low dose aspirin. DOSAGE AND ADMINISTRATION patients who experience falls while treated with KEVEYIS. ADVERSE REACTIONS USE IN SPECIFIC POPULATIONS Initiate dosing at 50 mg twice daily. The initial dose may be increased or decreased Pregnancy based on individual response, at weekly intervals (or sooner in case of adverse Clinical Trials Experience reaction). The maximum recommended total daily dose is 200 mg. Because clinical trials are conducted under widely varying conditions, adverse reaction Pregnancy Category C. Primary hyperkalemic periodic paralysis, primary hypokalemic periodic paralysis, rates observed in the clinical trials of a drug cannot be directly compared to rates in There are no adequate and well-controlled studies in pregnant women. Teratogenic and related variants are a heterogeneous group of conditions, for which the response the clinical trials of another drug and may not reflect the rates observed in practice. effects (fetal limb reduction defects) were reported following oral administration of to KEVEYIS may vary. Therefore, prescribers should evaluate the patient’s response In a 9-week randomized controlled trial in adults with hyperkalemic or hypokalemic dichlorphenamide to pregnant rats during organogenesis at 350 mg/kg, or 17 times to KEVEYIS after 2 months of treatment to decide whether KEVEYIS should be periodic paralysis (Study 1), the most common adverse reactions in patients the maximum recommended human dose (200 mg/day) on a body surface area continued. treated with KEVEYIS, with rates greater than placebo, were paresthesia, cognitive (mg/m2) basis. A no-effect dose has not been established. KEVEYIS should be used CONTRAINDICATIONS disorder, dysgeusia, and confusional state. The mean dose of KEVEYIS was 94 mg/ during pregnancy only if the potential benefit justifies the potential risk to the fetus. Nursing Mothers KEVEYIS is contraindicated in the following circumstances: day in patients with hypokalemic periodic paralysis and 82 mg/day in patients with hyperkalemic periodic paralysis. It is not known whether dichlorphenamide is excreted in human milk. Because • Hypersensitivity to dichlorphenamide or other sulfonamides many drugs are excreted in human milk, caution should be exercised when • Concomitant use of KEVEYIS and high dose aspirin Adverse Reactions in Patients Treated with KEVEYIS with Incidence ≥ 5% dichlorphenamide is administered to a nursing woman. • Severe pulmonary disease, limiting compensation to metabolic acidosis caused and more common than in Patients Treated with Placebo in Study 1 Pediatric Use by KEVEYIS Adverse Reaction KEVEYIS Placebo Safety and effectiveness in pediatric patients have not been established. N = 36 N = 29 • Hepatic insufficiency: KEVEYIS may aggravate hepatic encephalopathy. (%) (%) Geriatric Use WARNINGS AND PRECAUTIONS Paresthesia 44 14 The risk of falls and of metabolic acidosis are greater in elderly patients. Hypersensitivity / Anaphylaxis / Idiosyncratic Reactions Cognitive disorder* 14 7 OVERDOSAGE Fatalities associated with the administration of sulfonamides have occurred due Dysgeusia 14 0 Symptoms of overdosage or toxicity may include drowsiness, anorexia, nausea, to adverse reactions including Stevens-Johnson syndrome, toxic epidermal Confusional state 11 0 vomiting, dizziness, paresthesias, ataxia, tremor, and tinnitus. necrolysis, fulminant hepatic necrosis, agranulocytosis, aplastic anemia and other Nervous system blood dyscrasias. Pulmonary involvement can occur in isolation or as part of a disorders Headache 8 7 In the event of overdosage, induce emesis or perform gastric lavage. The systemic reaction. Hypoesthesia 8 0 electrolyte disturbance most likely to be encountered from overdosage is hyperchloremic acidosis. KEVEYIS should be discontinued at the first appearance of skin rash or any sign of Lethargy 8 0 immune-mediated or idiosyncratic adverse reaction. Dizziness 6 0 Mfd. by: Taro Pharmaceutical Industries Ltd., Haifa Bay, Israel 2624761 Concomitant Use of Aspirin Gastrointestinal Diarrhea 6 3 Dist. by: Strongbridge US Inc., Trevose, PA 19053 Anorexia, tachypnea, lethargy, and coma have been reported with concomitant use disorders Nausea 6 0 KEVEYIS® is a registered trademark licensed exclusively in the US to of dichlorphenamide and high-dose aspirin. The concomitant use of KEVEYIS and General disorders and Fatigue 8 0 Strongbridge Biopharma plc. high dose aspirin is contraindicated. KEVEYIS should be used with caution in patients administration site Malaise 6 0 STRONGBRIDGE BIOPHARMA™ is a trademark of Strongbridge Biopharma plc. receiving low dose aspirin. conditions Strongbridge Biopharma plc is the parent company of Strongbridge US Inc. Hypokalemia Investigations Weight decreased 6 0 www.KEVEYIS.com Muscle spasms 8 0 KEVEYIS increases potassium excretion and can cause hypokalemia. Musculoskeletal and Revised: January 2017 The risk of hypokalemia is greater when KEVEYIS is used in patients with conditions connective tissue Arthralgia 6 3 associated with hypokalemia (e.g., adrenocortical insufficiency, hyperchloremic disorders Muscle twitching 6 0 metabolic acidosis, or respiratory acidosis), and in patients receiving other drugs that Dyspnea 6 0 may cause hypokalemia (e.g., loop diuretics, thiazide diuretics, laxatives, antifungals, Respiratory Pharyngolaryngeal pain 6 0 penicillin, and theophylline). Rash 8 0 Baseline and periodic measurement of serum potassium during KEVEYIS treatment Skin are recommended. Pruritus 6 0 * If hypokalemia develops or persists, consideration should be given to reducing the Cognitive disorder combined cases with the preferred terms of cognitive disorder, disturbance in attention, and mental impairment. dose or discontinuing KEVEYIS. In Primary Periodic Paralysis Helping to turn limitation into activation

Call 844-KEVEYIS for Patient Support

Metabolic Acidosis The following are adverse reactions which have been reported for dichlorphenamide KEVEYIS can cause hyperchloremic non-anion gap metabolic acidosis. Concomitant that were serious adverse events or are not reported in the previous section of use of KEVEYIS with other drugs that cause metabolic acidosis may increase the labeling: amnesia, cardiac failure, condition aggravated, convulsion, fetal death, severity of metabolic acidosis. hallucination, nephrolithiasis, pancytopenia, psychotic disorder, renal tubular necrosis, stupor, syncope, and tremor. BRIEF SUMMARY OF PRESCRIBING INFORMATION Baseline and periodic measurement of serum bicarbonate during KEVEYIS treatment are recommended. DRUG INTERACTIONS These highlights do not include all the information needed to use Aspirin and Salicylates KEVEYIS® safely and effectively. See Full Prescribing Information If metabolic acidosis develops or persists, consideration should be given to reducing for KEVEYIS®. the dose or discontinuing KEVEYIS. KEVEYIS may cause an elevation in salicylate levels in patients receiving aspirin. Falls Anorexia, tachypnea, lethargy, and coma have been reported with concomitant use of INDICATIONS AND USAGE dichlorphenamide and high-dose aspirin. KEVEYIS is indicated for the treatment of primary hyperkalemic periodic paralysis, KEVEYIS increases the risk of falls. The risk of falls is greater in the elderly and with primary hypokalemic periodic paralysis, and related variants. higher doses of KEVEYIS. Consider dose reduction or discontinuation of KEVEYIS in Concomitant use of KEVEYIS and high dose aspirin is contraindicated. KEVEYIS should be used with caution in patients receiving low dose aspirin. DOSAGE AND ADMINISTRATION patients who experience falls while treated with KEVEYIS. ADVERSE REACTIONS USE IN SPECIFIC POPULATIONS Initiate dosing at 50 mg twice daily. The initial dose may be increased or decreased Pregnancy based on individual response, at weekly intervals (or sooner in case of adverse Clinical Trials Experience reaction). The maximum recommended total daily dose is 200 mg. Because clinical trials are conducted under widely varying conditions, adverse reaction Pregnancy Category C. Primary hyperkalemic periodic paralysis, primary hypokalemic periodic paralysis, rates observed in the clinical trials of a drug cannot be directly compared to rates in There are no adequate and well-controlled studies in pregnant women. Teratogenic and related variants are a heterogeneous group of conditions, for which the response the clinical trials of another drug and may not reflect the rates observed in practice. effects (fetal limb reduction defects) were reported following oral administration of to KEVEYIS may vary. Therefore, prescribers should evaluate the patient’s response In a 9-week randomized controlled trial in adults with hyperkalemic or hypokalemic dichlorphenamide to pregnant rats during organogenesis at 350 mg/kg, or 17 times to KEVEYIS after 2 months of treatment to decide whether KEVEYIS should be periodic paralysis (Study 1), the most common adverse reactions in patients the maximum recommended human dose (200 mg/day) on a body surface area continued. treated with KEVEYIS, with rates greater than placebo, were paresthesia, cognitive (mg/m2) basis. A no-effect dose has not been established. KEVEYIS should be used CONTRAINDICATIONS disorder, dysgeusia, and confusional state. The mean dose of KEVEYIS was 94 mg/ during pregnancy only if the potential benefit justifies the potential risk to the fetus. Nursing Mothers KEVEYIS is contraindicated in the following circumstances: day in patients with hypokalemic periodic paralysis and 82 mg/day in patients with hyperkalemic periodic paralysis. It is not known whether dichlorphenamide is excreted in human milk. Because • Hypersensitivity to dichlorphenamide or other sulfonamides many drugs are excreted in human milk, caution should be exercised when • Concomitant use of KEVEYIS and high dose aspirin Adverse Reactions in Patients Treated with KEVEYIS with Incidence ≥ 5% dichlorphenamide is administered to a nursing woman. • Severe pulmonary disease, limiting compensation to metabolic acidosis caused and more common than in Patients Treated with Placebo in Study 1 Pediatric Use by KEVEYIS Adverse Reaction KEVEYIS Placebo Safety and effectiveness in pediatric patients have not been established. N = 36 N = 29 • Hepatic insufficiency: KEVEYIS may aggravate hepatic encephalopathy. (%) (%) Geriatric Use WARNINGS AND PRECAUTIONS Paresthesia 44 14 The risk of falls and of metabolic acidosis are greater in elderly patients. Hypersensitivity / Anaphylaxis / Idiosyncratic Reactions Cognitive disorder* 14 7 OVERDOSAGE Fatalities associated with the administration of sulfonamides have occurred due Dysgeusia 14 0 Symptoms of overdosage or toxicity may include drowsiness, anorexia, nausea, to adverse reactions including Stevens-Johnson syndrome, toxic epidermal Confusional state 11 0 vomiting, dizziness, paresthesias, ataxia, tremor, and tinnitus. necrolysis, fulminant hepatic necrosis, agranulocytosis, aplastic anemia and other Nervous system blood dyscrasias. Pulmonary involvement can occur in isolation or as part of a disorders Headache 8 7 In the event of overdosage, induce emesis or perform gastric lavage. The systemic reaction. Hypoesthesia 8 0 electrolyte disturbance most likely to be encountered from overdosage is hyperchloremic acidosis. KEVEYIS should be discontinued at the first appearance of skin rash or any sign of Lethargy 8 0 immune-mediated or idiosyncratic adverse reaction. Dizziness 6 0 Mfd. by: Taro Pharmaceutical Industries Ltd., Haifa Bay, Israel 2624761 Concomitant Use of Aspirin Gastrointestinal Diarrhea 6 3 Dist. by: Strongbridge US Inc., Trevose, PA 19053 Anorexia, tachypnea, lethargy, and coma have been reported with concomitant use disorders Nausea 6 0 KEVEYIS® is a registered trademark licensed exclusively in the US to of dichlorphenamide and high-dose aspirin. The concomitant use of KEVEYIS and General disorders and Fatigue 8 0 Strongbridge Biopharma plc. high dose aspirin is contraindicated. KEVEYIS should be used with caution in patients administration site Malaise 6 0 STRONGBRIDGE BIOPHARMA™ is a trademark of Strongbridge Biopharma plc. receiving low dose aspirin. conditions Strongbridge Biopharma plc is the parent company of Strongbridge US Inc. Hypokalemia Investigations Weight decreased 6 0 www.KEVEYIS.com Muscle spasms 8 0 KEVEYIS increases potassium excretion and can cause hypokalemia. Musculoskeletal and Revised: January 2017 The risk of hypokalemia is greater when KEVEYIS is used in patients with conditions connective tissue Arthralgia 6 3 associated with hypokalemia (e.g., adrenocortical insufficiency, hyperchloremic disorders Muscle twitching 6 0 metabolic acidosis, or respiratory acidosis), and in patients receiving other drugs that Dyspnea 6 0 may cause hypokalemia (e.g., loop diuretics, thiazide diuretics, laxatives, antifungals, Respiratory Pharyngolaryngeal pain 6 0 penicillin, and theophylline). Rash 8 0 Baseline and periodic measurement of serum potassium during KEVEYIS treatment Skin are recommended. Pruritus 6 0 * If hypokalemia develops or persists, consideration should be given to reducing the Cognitive disorder combined cases with the preferred terms of cognitive disorder, disturbance in attention, and mental impairment. dose or discontinuing KEVEYIS. Table of Contents continued e367 Safety/tolerability of the anti-semaphorin 4D antibody e354 Alemtuzumab and intrathecal methotrexate failed in the VX15/2503 in a randomized phase 1 trial therapy of Rasmussen encephalitis C. LaGanke, L. Samkoff, K. Edwards, L. Jung Henson, Z. Liba, P. Sedlacek, V. Sebronova, A. Maulisova, P. Repovic, S. Lynch, L. Stone, D. Mattson, A. Galluzzi, B. Rydenhag, J. Zamecnik, M. Kyncl, and P. Krsek T.L. Fisher, C. Reilly, L.A. Winter, J.E. Leonard, and M. Zauderer e361 Anti-DPPX encephalitis: Prominent nystagmus reﬂected by extraocular muscle FDG-PET avidity e369 Olfactory dysfunction in patients with primary L. Doherty, D. Gold, L. Solnes, J. Probasco, and progressive MS A. Venkatesan F.A. Schmidt, M.B. Maas, R. Geran, C. Schmidt, H. Kunte, K. Ruprecht, F. Paul, Ö. Göktas, and L. Harms e365 ANCA-associated vasculitis predominantly presenting with severe myalgias e371 Autoimmune episodic ataxia in patients with E. Bahou and L. Zhou anti-CASPR2 antibody-associated encephalitis B. Joubert, F. Gobert, L. Thomas, M. Saint-Martin, Vanishing spinal cord after varicella-zoster virus V. Desestret, P. Convers, V. Rogemond, G. Picard, e364 myelitis F. Ducray, D. Psimaras, J.-C. Antoine, J.-Y. Delattre, and J. Honnorat M. Sepúlveda, J. Almeida, J. Berenguer, N. Solá-Valls, J. Saura, Y. Blanco, S. Llufriu, F. Graus, and A. Saiz

CLINICAL/SCIENTIFIC NOTES e348 IVIG treatment for repeated hypothermic attacks e368 Acute liver injury in a Glatopa-treated patient with MS associated with LGI1 antibody encephalitis J.J. Sabatino, Jr., N.J. Mehta, S. Kakar, S.S. Zamvil, and Y. Hayashi, M. Yamada, A. Kimura, and T. Inuzuka B.A.C. Cree

e349 Encephalomyelitis following deﬁnitive Zika virus e370 An unusual case of miliary PML-IRIS in an HIV1 patient infection D. Cuendet, R. Sarro, L. Merz, M.C. Jiménez, P. Maeder, W. Roth, C. Tyshkov, K. Thakur, and W. Vargas R. Du Pasquier, and P. Tsouni

e351 Spontaneous remission lasting more than a decade e372 Dacrystic seizures: A cry for help in untreated AQP4 antibody-positive NMOSD A. Gadoth, J. Singh, J.W. Britton, E.P. Flanagan, and L. Pandit and S. Mustafa S.J. Pittock

Video LOE classiﬁcation LOE recommendation Podcast

tinyurl.com/NeurologyNN twitter.com/GreenJournal

Cover image: F18-FDG PET/CT scan demonstrates normal size of extraocular muscles but markedly asymmetric increased FDG activity localizing to the left medial rectus, right lateral rectus, and inferior rectus bilaterally. Stylized by Andrea Rahkola, Neurology Production Editor. See “Anti-DPPX encephalitis: Prominent nystagmus reﬂected by extraocular muscle FDG-PET avidity.” EDITOR’SCORNER The case for autoimmune neurology

Josep Dalmau, MD, PhD This year, at the Annual Meeting of the American diseases such as irritable bowel disease and MS. Ago- Academy of Neurology, the inaugural session of the nists of AHR come from a variety of sources and can be “Autoimmune Neurology” section was held. Atten- endogenous metabolic products (e.g., L-Kynurenine), Correspondence to dance and enthusiasm were high. The term “Autoim- synthetic (e.g., laquinimod) or derive from the diet, Dr. Dalmau: ” [email protected] mune Neurology is currently used to describe the environmental pollutants, or commensal bacteria. In emerging field that encompasses “immune-inflamma- several disease models, activation of AHR by a variety Neurol Neuroimmunol tory” diseases of the nervous system other than MS. of agonists produces nuclear translocation and control Neuroinflamm Autoimmune Neurology has emerged over the past of specific transcriptional programs and has been dem- 2017;4:e373; doi: 10.1212/ NXI.0000000000000373 15 years with an expanding list of disorders and has onstrated to result in anti-inflammatory and neuropro- attracted the attention of colleagues from other spe- tective effects. These authors have previously shown cialties. The identification of these diseases has facil- that AHR agonists generated by the gut microbiome itated the diagnosis and often successful treatment for can be detected in the CNS where they dampen CNS patients who previously were considered to have idi- inflammation by activating AHR in resident cells.2 In opathic or poorly defined diseases. In addition, stud- the current study, the authors detected a global ies of the underlying pathogenic mechanisms have decrease of circulating AHR agonists in patients with improved our understanding on how autoantibodies relapsing-remitting MS compared with healthy con- directly alter the function of neurons, astrocytes, and trols. By contrast, during acute stages of CNS inflam- synaptic networks, resulting in an extraordinary vari- mation in active MS or clinically isolated syndrome ety of neuropsychiatric symptoms. All AAN members (CIS), the authors found that AHR agonist activity interested in participating and receiving news about was increased. Moreover, AHR ligand levels in patients the Autoimmune Neurology section should consider with benign MS or mild clinical impairment were joining “Synapse” (Synapse AAN Online Communi- unchanged compared with healthy controls. These ties), the official communication platform of the data demonstrate that the serum AHR agonists are AAN Sections. dynamically modulated during the course of MS and A remarkable parallelism can be established could lead to the development of biomarkers to mon- between the growing interest in Autoimmune Neu- itor disease activity as discussed in an Editorial Com- rology and this journal, Neurology® Neuroimmunology ment by Sabatino and Zamvil.3 & Neuroinflammation, which was inaugurated in The study by Moser et al.4 in this issue also im- 2014 partly in response to the increasing number of plicates dysregulation of the gut microbiome in auto- excellent manuscripts focused on these fascinating immunity. In their proof-of-concept study, the disorders. This July issue of Neurology: Neuroimmu- authors examined mucosal specimens obtained dur- nology & Neuroinflammation reflects the forward ing ileocolonoscopy from 15 treatment-naive patients momentum of the field, with an abundance of inter- with CIS or early relapsing MS. They also evaluated esting articles that make it difficult to choose which to fecal samples from the patients for short-chain fatty highlight in this Editors Corner. acids (SCFAs), which have been implicated in medi- Among the articles is a study by Rothhammer ating regulatory intestinal immune responses, includ- et al.,1 which reports that agonists of aryl hydrocarbon ing Treg homeostasis. The results demonstrated that receptor (AHR) in serum are dynamically modulated compared with healthy controls, patients with CIS or during the course of MS. The AHR is a ligand- MS had significantly fewer dendritic cells (DCs), activated transcription factor that has critical regula- CD1031 tolerogenic DCs, and Tregs in their distal tory functions of the innate and adaptive immune colon as well as significantly lower total fecal content responses relevant to the pathogenesis of autoimmune of SCFAs and, in particular, of butyrate and acetate.

From the ICREA-IDIBAPS, Hospital Clínic, University of Barcelona, Spain; and Department of Neurology, University of Pennsylvania. Funding information and disclosures are provided at the end of the editorial. Go to Neurology.org/nn for full disclosure forms. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 These results indicate that homeostasis of colon DCs complex encephalitis more often had hypermetabolism and Tregs is disturbed in patients with CIS and MS than those with other types of AE. Longitudinal studies and may be associated with the SCFA depletion. on the pattern of the FDG-PET/CT abnormalities Although the authors state that these results cannot along the course of the disease were not obtained; this serve to establish a causal relationship with the demy- has been suggested to be important in patients with elinating disease, they provide a very nice discussion anti-NMDAR encephalitis and might as well be of how these changes in the immune constitution of important in other AE. Nevertheless, the sensitivity the gut could be clinically relevant. of FDG-PET/CT in AE was higher than any of the Joubert et al.5 report a novel clinical manifestation other paraclinical tests and seemed to be unaltered by of Caspr2 antibody–associated encephalitis consisting corticosteroids and sedatives, suggesting a potential, yet of paroxysmal or episodic ataxia. Their article de- undefined, role of FDG-PET/CT in the diagnosis, scribes a 61-year-old patient with limbic encephalitis follow-up, or prognosis of AE. who developed short-lasting (less than 1 minute) epi- In addition to these studies, this July issue of Neu- sodes of slurred speech, gait ataxia, and limb dysme- rology: Neuroimmunology & Neuroinflammation con- tria. These episodes repeated 3–4 times every day and tains 21 articles and an Editorial Comment on AHR showed a rapid response to immunotherapy. Based as a potential surrogate marker of MS disease activity, on this observation, the authors retrospectively iden- which I hope will catch your interest. tified episodic ataxia in 5 of 37 patients with Caspr2 antibody–associated encephalitis, all with limbic STUDY FUNDING involvement (and none with neuromyotonia or No targeted funding reported. Morvan syndrome). In these additional cases, the DISCLOSURE acute symptoms of cerebellar dysfunction lasted J. Dalmau is the editor of Neurology: Neuroimmunology & Neuroinflam- between a few minutes and 2 hours and were often mation; is on the editorial board for Neurology®, UpToDate; holds pat- triggered by orthostatism. In one of the patients, ents for and receives royalties from Ma2 autoantibody test, NMDA receptor autoantibody test, GABA(B) receptor autoantibody test, GA- the episodes preceded by 7 months the develop- BA(A) receptor autoantibody test, DPPX autoantibody test, and IgLON5 ment of full-blown autoimmune encephalitis (AE). autoantibody test; and receives research support from Euroimmun, NIH, In 4 of 5 cases, the ataxic episodes responded to Fundació CELLEX, and Instituto Carlos III (Fondo de Investigaciones immunotherapy and in 1 improved spontaneously. Sanitarias). Go to Neurology.org/nn for full disclosure forms. The underlying mechanism is unknown, but the REFERENCES authors emphasize the remarkable similarity 1. Rothhammer V, Borucki DM, Garcia Sanchez MI, et al. between these episodes and those of episodic Dynamic regulation of serum aryl hydrocarbon receptor ataxias related to hereditary channelopathies agonists in MS. Neurol Neuroimmunol Neuroinflamm involving KCNA1 (a gene coding for the voltage- 2017;4:e359. doi: 10.1212/NXI.0000000000000359. gated potassium channel Kv1.1) and CACNA1A 2. Rothhammer V, Mascanfroni ID, Bunse L, et al. Type I (Cav2.1 subunit of the P/Q type voltage-gated interferons and microbial metabolites of tryptophan modulate astrocyte activity and central nervous system calcium channel). inflammation via aryl hydrocarbon receptor. Nat Med 6 Probasco et al. report the fludeoxyglucose (FDG)- 2016;22:586–597. PET/CT findings in 61 patients with AE, 32 of them 3. Sabatino JJ, Zamvil SS. Aryl hydrocarbon receptor activity seropositive for antibodies. Fifty-two (85%) had may serve as a surrogate marker for MS disease activity. abnormal FDG-PET/CT, showing isolated hypome- Neurol Neuroimmunol Neuroinflamm 2017;4:e366. doi: tabolism in 42 (68%), isolated hypermetabolism in 2 10.1212/NXI.0000000000000366. 4. Moser AM, Spindelboeck W, Strohmaier H, et al. Mucosal (3%), and mixed areas of hypermetabolism and hypo- biopsy reveals immunologic changes of the colon in patients metabolism in 8 (13%). Overall, abnormal metabolic with early MS. Neurol Neuroimmunol Neuroinflamm findings occurred more frequently than CSF inflam- 2017;4:e362. doi: 10.1212/NXI.0000000000000362. mation (62%), abnormal MRI (40%), and abnormal 5. Joubert B, Gobert F, Thomas L, et al. Autoimmune epi- initial EEG (30%). The authors did not find differ- sodic ataxia in patients with anti-CASPR2 antibody-associ- ences across age groups, antibody status, antibody class, ated encephalitis. Neurol Neuroimmunol Neuroinflamm 2017;4:e371. doi: 10.1212/NXI.0000000000000371. or type of AE. This article has several limitations dis- 6. Probasco JC, Solnes L, Nalluri A, et al. Abnormal brain cussed by the authors, related to the small number of metabolism on FDG-PET/CT is a common early finding in patients with some subtypes of AE; for example, the autoimmune encephalitis. Neurol Neuroimmunol Neuroin- subsets of patients with anti-NMDAR or anti-VGKC flamm 2017;4:e352. doi: 10.1212/NXI.0000000000000352.

2 Neurology: Neuroimmunology & Neuroinflammation EDITORIAL Aryl hydrocarbon receptor activity may serve as a surrogate marker for MS disease activity

Joseph J. Sabatino Jr, MS diagnosis and prognosis is based on a combina- assay driven by an AHR-responsive promoter was MD, PhD tion of clinical criteria, MRI, and CSF findings. measured. They first demonstrated that AHR agonistic Scott S. Zamvil, MD, However, we still lack the ability to predict disease activity was reduced in a cohort of patients with PhD course, including relapse frequency and disability relapsing-remitting MS (RR-MS) compared with progression. Thus, reliable biomarkers of MS risk healthy controls. The authors then measured AHR and prognosis are sorely needed. In this issue of activity in MS patients with different stages of disease Correspondence to Neurology® Neuroimmunology & Neuroinflamma- activity. Surprisingly, they found that AHR agonistic Dr. Sabatino: 1 [email protected]. tion,Rothhammeretal. suggestthatchangesin activity was lower in patients with MS in remission org or Dr. Zamvil: serum agonistic activity of the aryl hydrocarbon compared with those with active inflammation (as con- [email protected] receptor (AHR) may be a useful measure of MS firmed by contrast-enhanced MRI), although both activity. groups were still lower overall compared with controls. Neurol Neuroimmunol Neuroinflamm The AHR is a ligand-activated transcription factor By contrast, AHR activity levels were higher in patients 2017;4:e366; doi: 10.1212/ that was initially described in response to toxins, but with clinically isolated syndrome at the time of their NXI.0000000000000366 was later found to be activated by a multitude of first clinical attack compared with healthy controls, diverse stimuli, including environmental toxins, whereas no difference was found between patients with microbial toxins, dietary compounds, and endogenous benign MS (long-standing RR-MS with minimal neu- metabolites.2 The AHR is found on a host of immune rologic impairment) and controls. cells, including T cells and B cells, and is found at The findings by Rothhammer et al. are the first particularly high levels in barrier tissues such as the indication that AHR activity is dynamically modu- gut, skin, and lungs.2 A number of endogenous AHR lated at various stages of MS. In addition to represent- ligands have been reported, including kynurenine ing a potential biomarker of MS activity, the AHR is (a tryptophan metabolite produced by indoleamine also a novel therapeutic target for MS, as in the case of 2,3-dioxygenase), indoles (produced by bacterial laquinimod.5,6 The reason why AHR agonistic metabolism of tryptophan and dietary intake), and activity is increased during a first clinical attack, but 6-formylindolo[3,2-b]carbazole.2 Depending on the decreased in subsequent relapses remains unclear. ligand, AHR stimulation can strongly influence the The authors did not identify any association between development of proinflammatory Th17 or anti- AHR agonistic activity and patient age, disease inflammatory regulatory T cells, leading to profoundly duration, or status on approved MS therapies. different outcomes in the MS animal model, experi- However, one must recognize that the patient num- mental autoimmune encephalomyelitis (EAE).3,4 The bers in their cohorts were relatively small. It is possi- AHR is therefore a putative MS therapeutic target in ble that AHR agonist levels may be influenced by MS. In this regard, laquinimod can alter the pheno- factors other than MS activity. The reporter assay type of antigen-presenting cells and autoreactive T and used by the authors captures the sum of all AHR B cells,5–8 and its therapeutic efficacy in EAE is depen- agonistic activity in the sera but does not measure dent on AHR activation.9 AHR stimulation in the individual AHR ligands, which can vary markedly CNS also seems to have a neuroprotective effect in in their effects on the immune system.3,4 Thus, these EAE.10 It therefore stands to reason that AHR activity findings may represent changes in endogenous AHR could provide a link between environmental expo- ligand levels in response to inflammation, differences sures, genetics, and MS risk and disease activity. in patient microbiota, or other extrinsic environmen- In a study by Rothhammer et al.,1 serum AHR tal sources. As the study by Rothhammer et al. was not agonistic activity in patients with MS using a reporter longitudinal, it is not clear to what extent AHR levels See article From the Multiple Sclerosis Center, Department of Neurology, University of California San Francisco. Funding information and disclosures are provided at the end of the editorial. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the Editorial Office. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 fluctuate throughout the course of individual patients 2. Stockinger B, Di Meglio P, Gialitakis M, Duarte JH. The with MS. It is also unknown whether changes in the aryl hydrocarbon receptor: multitasking in the immune – AHR agonistic activity drive changes in MS activity or system. Annu Rev Immunol 2014;32:403 432. 3. Quintana FJ, Basso AS, Iglesias AH, et al. Control of T represent a compensatory response. Further study into (reg) and T(H)17 cell differentiation by the aryl hydrocar- the dynamic changes of specific AHR ligands and bon receptor. Nature 2008;453:65–71. fluctuations during different stages of disease activity 4. Veldhoen M, Hirota K, Westendorf AM, et al. The aryl within individual patients and larger populations as hydrocarbon receptor links TH17-cell-mediated autoim- a whole is clearly needed. It is hoped that such studies munity to environmental toxins. Nature 2008;453:106– will shed light on the utility of AHR as a possible 109. 5. Schulze-Topphoff U, Shetty A, Varrin-Doyer M, et al. biomarker and therapeutic target in MS. The findings Laquinimod, a quinoline-3-carboxamide, induces type ii by Rothhammer et al. represent an exciting step myeloid cells that modulate central nervous system auto- forward toward realizing this goal. immunity. PLoS One 2012;7:1–10. 6. Jolivel V, Luessi F, Masri J, et al. Modulation of dendritic STUDY FUNDING cell properties by laquinimod as a mechanism for modu- No targeted funding reported. lating multiple sclerosis. Brain 2013;136:1048–1066. 7. Varrin-Doyer M, Pekarek KL, Spencer CM, et al. Treat- DISCLOSURE ment of spontaneous EAE by laquinimod reduces Tfh, B Dr. Sabatino received research support from NMSS, R25 NS070680. cell aggregates, and disease progression. Neurol Neuroim- Dr. Zamvil served on the data safety monitoring board for BioMS, Teva munol Neuroinflammation 2016;3:e272. doi: 10.1212/ Pharmaceuticals, Eli Lilly and Com; is a member of the clinical advisory NXI.0000000000000272. board for Myelin Repair Foundation; is deputy editor for Neurology: 8. Toubi E, Nussbaum S, Staun-Ram E, et al. Laquinimod Neuroimmunology & Neuroinflammation; has a patent pending for modulates B cells and their regulatory effects on T cells in Aquaporin-4 peptides and methods forusingthesame;hasconsulted – for Biogen, Teva, EMD Serono, Novartis, and Roche; served on the speak- Multiple Sclerosis. J Neuroimmunol 2012;251:45 54. er’s bureau for Biogen and Advanced Health Media; and received research 9. Kaye J, Piryatinsky V, Birnberg T, et al. Laquinimod ar- support from Biogen, Teva, NIH, NMSS, and Alexander M. and June rests experimental autoimmune encephalomyelitis by acti- L. Maisin Foundation. Go to Neurology.org/nn for full disclosure forms. vating the aryl hydrocarbon receptor. Proc Natl Acad Sci USA 2016;113:E6145–E6152. REFERENCES 10. Rothhammer V, Mascanfroni ID, Bunse L, et al. Type I 1. Rothhammer V, Borucki DM, Sanchez MIG, et al. interferons and micorbial metabolites of tryptophan mod- Dynamic regulation of serum aryl hydrocarbon receptor ulate astrocyte activity and central nervous system inflam- agonists in MS. Neurol Neuroimmunol Neuroinflamma- mation via the aryl hydrocarbon receptor. Nat Med 2016; tion 2017;4:e359. doi: 10.1212/NXI.0000000000000359. 22:586–597.

2 Neurology: Neuroimmunology & Neuroinflammation Dynamic regulation of serum aryl hydrocarbon receptor agonists in MS

Veit Rothhammer, MD ABSTRACT Davis M. Borucki, BSc, Objective: Several factors influence the clinical course of autoimmune inflammatory diseases BA such as MS and inflammatory bowel disease. Only recently, the complex interaction between Maria Isabel Garcia the gut microbiome, dietary factors, and metabolism has started to be appreciated with regard Sanchez, MD to its potential to modulate acute and chronic inflammation. One of the molecular sensors that Maria Antonietta mediates the effects of these environmental signals on the immune response is the aryl hydrocar- Mazzola, MD bon receptor (AHR), a ligand-activated transcription factor with key functions in immune cells. Christopher C. Hemond, Methods: In this study, we analyzed the levels of AHR agonists in serum samples from patients MD with MS and healthy controls in a case-control study. Keren Regev, MD Anu Paul, PhD Results: We detected a global decrease of circulating AHR agonists in relapsing-remitting MS pa- Pia Kivisäkk, MD, PhD tients as compared to controls. However, during acute CNS inflammation in clinically isolated syn- Rohit Bakshi, MD drome or active MS, we measured increased AHR agonistic activity. Moreover, AHR ligand levels Guillermo Izquierdo, MD in patients with benign MS with relatively mild clinical impairment despite longstanding disease Howard L. Weiner, MD were unaltered as compared to healthy controls. Francisco J. Quintana, Conclusions: Collectively, these data suggest that AHR agonists in serum are dynamically modu- PhD lated during the course of MS. These findings may guide the development of biomarkers to monitor disease activity as well as the design of novel therapeutic interventions for MS. Neurol Neuroimmunol Neuroinflamm 2017;4:e359; doi: 10.1212/NXI.0000000000000359 Correspondence to Dr. Quintana: [email protected] GLOSSARY AHR 5 aryl hydrocarbon receptor; CIS 5 clinically isolated syndrome; DMT 5 disease-modifying therapy; HEK 5 human embryonic kidney; IBD 5 inflammatory bowel disease; Kyn 5 Kynurenine; RRMS 5 relapsing-remitting MS.

The Aryl hydrocarbon receptor (AHR) is a key regulator of innate and adaptive immune responses relevant to the pathogenesis of autoimmune diseases such as inflammatory bowel disease (IBD) and MS.1–4 AHR is a ligand-activated transcription factor, whose function is regulated by small agonists that promote AHR activation, nuclear translocation, and the control of specific transcriptional programs.5–14 These agonists are provided by diverse sources, including environmental pollutants, dietary components, microbial products, as well as endogenous metabolites.3,6–11,13–17 The relevance of endogenous AHR ligands during inflammation has been investigated in different experimental paradigms. L-Kynurenine (Kyn), for example, is an AHR agonist generated by endogenous metabolism. Of interest, Kyn is increased in the context of inflammation and damp- ens proinflammatory T-cell responses, limiting immune-mediated pathology.18,19 Similarly, synthetic agonists can also activate AHR to therapeutically modulate the immune response. Laquinimod is an AHR agonist that shows anti-inflammatory and neuroprotective effects in the MS model experimental autoimmune encephalomyelitis probably as a result of the inhibition of NF-kB activation in mouse and human dendritic cells.20–25 Indeed, beneficial effects of laqui- Supplemental data nimod were also documented in the Benefit-Risk Assessment of Avonex and Laquinimod at Neurology.org/nn From the Ann Romney Center for Neurologic Diseases (V.R., D.M.B., M.A.M., C.C.H., K.R., A.P., P.K., R.B., H.L.W., F.J.Q.), Brigham and Women’s Hospital, Harvard Medical School, Boston, MA; Molecular Biology Service and MS Unit (M.I.G.S., G.I.), University of Seville, Spain; and Broad Institute of MIT and Harvard (F.J.Q.), Cambridge, MA. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the NIH. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Table Characteristics of patients with MS and controls tested in luciferase assays

Figure Cohorts Females Age, y Disease duration EDSS Treatment

1 Controls (26) 16 (61.5) 29.0 (26.0, 30.3) None None None

RRMS (91) 71 (78) 44.2 (40.0, 48.9) 13.6 (7.8, 17.6) 1.8 (1.0, 2.5) 62 (68.1)

2 Controls (26) 16 (61.5) 31.9 (26.0, 38.0) None None None

RRMS remission (32) 26 (81.2) 44.7 (40.7, 51.6) 14.8 (12.0, 18.8) 1.8 (1.5, 2.0) 11 (34.4)

RRMS active (20) 16 (80) 38.3 (31.2, 49.7) 6.6 (3.0, 11.5) 1.6 (1.0, 2.5) 10 (50)

3 Controls (33) 20 (60.1) 36.2 (26.0, 45.5) None None None

CIS (15) 9 (60) 32.8 (28.1, 34.6) Onset 1.6 (1.5, 2.0) 0

4 Controls (7) 6 (85.7) 48.9 (42.0, 52.0) None None None

Benign (11) 10 (90.9) 54.1 (51.0, 56.0) 21.0 (17.0, 28.0) 0.6 (0, 1.5) 2 (18.2)

Abbreviations: CIS 5 clinically isolated syndrome; RRMS 5 relapsing-remitting MS. In column “Females,” numbers represent absolute numbers, and in parentheses, percentages of females. In columns Age, Disease duration, and Expanded Disability Status Scale (EDSS), numbers represent mean, numbers in parentheses represent 25% and 75% percentiles, respectively. In column “Treatment,” numbers represent absolute numbers of treated patients, and in parentheses, percentages of patients under immunomodulatory treatment.

(BRAVO) study, in which laquinimod-treated CNS inflammation in clinically isolated syn- patients with MS showed a reduction in the drome (CIS) or patients with RRMS, we rate of cerebral atrophy vs placebo that sug- detected increased AHR agonist levels as com- gested a neuroprotective role of AHR activation pared to healthy controls or clinically stable during CNS inflammation.26 patients with RRMS. Serum AHR agonists in Anti-inflammatory AHR ligands are also patients with benign MS with relatively mild provided by the diet and commensal bacte- clinical impairment despite longstanding dis- ria.2,9 These ligands have the capability to ease, however, exhibited unaltered AHR ligand dampen ongoing inflammation in the colonic levels as compared to healthy controls. Collec- mucosa and improve the outcome of experi- tively, these findings suggest that serum AHR mental colitis.6 Moreover, alterations in the agonists are dynamically modulated during the composition of the commensal flora as well course of MS. Low basal levels of circulating as genetic polymorphisms detected in IBD AHR agonists are detected in patients with patients have been shown to impair the gen- MS, probably reflecting deficits associated not eration of these protective AHR ligands, only with the diet and commensal flora but also ultimately contributing to immune dysregula- in the pathways that control the production tion and disease pathology.6 and degradation of AHR agonists. Inflamma- AHR agonists provided by the diet and tion increases AHR agonists in serum, probably commensal bacteria also contribute to the con- by promoting the production of endogenous trol of CNS inflammation. We have recently anti-inflammatory metabolites such as Kyn. shown that AHR agonists generated by the Finally, a fraction of patients with MS main- interaction of the gut microbiome and host tains control levels of circulating AHR agonists metabolism cross the blood-brain barrier and concomitant with a more benign disease dampen CNS inflammation by activating course, suggesting a protective role of AHR AHR in resident cells.5 Accordingly, we de- ligands in later stages of MS in the absence of tected decreased CNS AHR activation in acute inflammation. These observations might a small set of MS samples, as well as decreased guide the development of novel therapeutics circulating AHR agonists.5,6 for MS and biomarkers for risk stratification In this study, we analyzed AHR agonists in and treatment selection in patients with MS. serum samples from patients with MS and healthy controls. We detected a decrease in METHODS Determination of AHR agonistic activity. Fifteen thousand human embryonic kidney (HEK)-293 cells per serum AHR agonists in relapsing-remitting well were plated in 96-well plates (flat bottom). Twenty-four MS (RRMS) patients.5 However, during acute hours after plating, cells were transfected with equal amounts of

2 Neurology: Neuroimmunology & Neuroinflammation Figure 1 Detection of aryl hydrocarbon receptor ligands from different sources Figure 2 Aryl hydrocarbon receptor ligand levels are decreased in patients with relapsing-remitting MS

Aryl hydrocarbon receptor (AHR) agonistic activity in serum samples of healthy controls (controls, n 5 26) and relapsing- remitting MS (RRMS) patients (RRMS, n 5 91) was assessed in duplicates using an AHR ligand–sensitive luciferase assay. Relative activity was calculated by dividing firefly luciferase activity (pGud-Luc) by Renilla luciferase activity (pTK-Renilla). Values are means of duplicate measurements. Lines represent mean and error bars standard error of the mean (SEM). Significance levels were derived using the Student t test. **p , 0.01.

(Promega) as suggested by the manufacturer. After 24 hours, transfected cells were incubated with Dulbecco’s modified eagle medium (DMEM) supplemented with 10% of patient serum in duplicates. Luciferase activity was analyzed 24 hours later using the Dual-Luciferase Reporter System (Promega). Firefly luciferase activity was divided by Renilla luciferase activity and normalized to their respective control levels, which were set as 100%. The study was approved by the Institutional Review Board of Brigham and Women’s Hospital, and all participants provided written informed consent.

Statistical analysis. Statistical analyses were performed with Prism software (GraphPad, San Diego, CA), using the statistical Aryl hydrocarbon receptor (AHR) agonistic activity was measured for a collection of AHR li- tests indicated in the individual figure legends. No samples were gands from exogenous and exogenous sources, including the pollutant 2,3,7,8-Tetrachloro- excluded. The investigators were blinded as to sample cohorts dibenzo-p-dioxin (TCDD) (A), the diet-derived ligand Indole-3-carbinol (I3C) (B), ligands when performing AHR ligand level measurement and samples derived from microbial and host tryptophan metabolism Indole (C), Indoxyl-3-sulfate were run in duplicates. p Values of ,0.05 were considered sig- 9 9 9 (I3S) (D), Indirubin (E), and 2 Z-Indirubin (F), the mucosal ligand 2-(1 H-indole-3 -carbonyl)- nificant. All error bars represent SEM. thiazole-4-carboxylic acid methyl ester (ITE) (G), and the endogenous metabolite Kynurenine (H). Data are normalized to 100% (maximum activity per ligand) and are representative of 2 RESULTS independent experiments. AHR agonistic activity in serum is decreased in stable RRMS. To study circulating AHR agonistic activity in MS samples, we first analyzed sera from pGud-Luc (Firefly luciferase under control of AHR-responsive promoter element27) and pTK-Renilla (Renilla luciferase under a cohort of patients with RRMS and compared control of constitutively active thymidine kinase promoter; these to sera from healthy controls (table). In these Promega, Madison, WI) using Fugene Transfection Reagent studies, we used a reporter assay based on HEK-

Neurology: Neuroimmunology & Neuroinflammation 3 293 cells cotransfected with a plasmid containing Figure 3 Aryl hydrocarbon receptor ligand an AHR-responsive promoter element (xenobiotic levels are modulated by disease response element) driving firefly luciferase expres- activity sion (pGud-Luc27), and a thymidine kinase promoter-driven Renilla luciferase construct (pTK- Renilla) to control for transfection efficiency.5 Following transfection, the reporter cells were incubated with patient serum, and relative luciferase activities (pGud-Luc/pTK-Renilla) were determined after 24 hours using a commercial dual-luciferase assay. This assay detected AHR activation in response to a broad range of AHR agonists, including the pollutant 2,3,7,8-Tetrachlorodibenzo- p-dioxin (TCDD), the dietary ligand Indole-3- carbinol (I3C), ligands derived from microbial and host tryptophan metabolism such as Indole, Indoxyl-3-sulfate (I3S), Indirubin, and 29Z-In- dirubin, the mucosal ligand 2-(19H-indole-39-carbonyl)-thiazole-4-carboxylic acid methyl ester (ITE) and the endogenous metabolite Kyn (figure 1). Using this approach, we detected a global decrease in AHR agonistic activity in RRMS patient sera as compared to healthy controls (figure 2). Note that some patients displayed higher serum AHR agonistic activity than healthy controls, suggesting that additional disease-linked mechanisms may increase AHR agonits in patients with MS. However, patient age, Aryl hydrocarbon receptor (AHR) agonistic activity in serum disease duration, or the prevalence of disease- samples of healthy controls (controls, n 5 26), relapsing- modifying therapy (DMT) were not associated remitting MS (RRMS) patients during remission (RRMS remission, n 5 32), and patients with RRMS with active dis- with the detected AHR agonistic activity (figure e-1 ease (RRMS active, n 5 20) was assessed in duplicates using at Neurology.org/nn). an AHR ligand–sensitive luciferase assay. Values are means of duplicate measurements. Lines represent mean and error bars standard error of the mean (SEM). Significance levels Circulating AHR agonists are increased during acute were derived using 1-way analysis of variance followed by CNS inflammation. AHR ligands are generated during the Tukey multiple comparisons test. ****p , 0.0001. acute inflammation by different mechanisms including the enzymatic activity of indoleamine 2,3- MS, the risk of which can be assessed by evaluating dioxygenase (IDO) which produces anti-inflammatory – additional biomarkers, such as MRI, CSF composition, Kyn.1 3 Thus, we speculated that acute CNS inflam- or electrophysiologic studies, among others.28,29 mation such as that linked to MS relapses might Sera from CIS patients displayed increased AHR modulate AHR agonists in serum. To test this agonistic activity as compared to healthy controls hypothesis, we analyzed an additional cohort of patients (figure 4). Together with our findings on patients with MS with active CNS inflammation as determined with RRMS during a disease relapse, these findings by the presence of cerebral gadolinium-enhancing le- suggest that acute CNS inflammation results in sions in MRI at the time of sample acquisition and increased serum AHR agonist levels. compared them to a group of patients with RRMS with nonactive disease (table). While we still detected Unaffected AHR agonist levels in patients with benign a global decrease in AHR ligand levels in comparison to MS. Patients with benign MS present a relatively healthy controls, RRMS active patients displayed mild disease course, despite long disease duration increased AHR serum levels as compared to the samples and limited use of DMTs.30 Based on the anti- from the RRMS remission cohort (figure 3). inflammatory effects of AHR in several experimen- To further validate these findings, we used an inde- tal models of autoimmunity2,3,13 and potentially pendent cohort of patients who had been recently diag- MS,21,22 we analyzed circulating AHR agonist levels nosed with CIS, the first clinical manifestation of in a cohort of patients with benign MS characterized autoimmune CNS inflammation (table). CIS does not by mild clinical impairment despite longstanding fulfillanamnesticorMRItomographiccriteriaforMS RRMS (“Benign MS,” table). We found that serum and does not always convert into clinically definitive samples from patients with benign MS showed AHR

4 Neurology: Neuroimmunology & Neuroinflammation Figure 4 Aryl hydrocarbon receptor ligand Figure 5 Aryl hydrocarbon receptor ligand levels are increased in clinically levels in patients with benign isolated syndrome as compared to relapsing-remitting MS are unchanged healthy controls as compared to healthy controls

Aryl hydrocarbon receptor agonistic activity in serum sam- Aryl hydrocarbon receptor agonistic activity in serum sam- 5 ples of healthy controls (controls, n 33) and patients with ples of healthy controls (controls, n 5 7) and patients with 5 clinically isolated syndrome (CIS, n 15) was assessed in benign relapsing-remitting MS as defined by low EDSS duplicates. Values are means of duplicate measurements. scores despite longstanding disease (benign MS, n 5 11) Lines represent mean and error bars standard error of the was assessed in duplicates. Values are means of duplicate t mean (SEM). Significance level was derived by the Student measurements. Lines represent mean and error bars stan- **p , test. 0.01. dard error of the mean (SEM). Significance level was derived by the Student t test. n.s. 5 not significant. agonist levels comparable to those detected in controls (figure 5). the generation of AHR ligands from dietary tryptophan.35–37 Moreover, genetically defined factors have DISCUSSION In this work, we analyzed serum levels the potential to influence the composition of the gut of AHR agonists in patients with MS. Our data microbiome, for example, via the production of mi- suggest that AHR agonist levels are dynamically mod- croRNAs or altered cytokine signaling.6,38,39 Finally, ulated during the course of MS: in acute inflamma- the genetic background of patients with MS may tion, such as the first relapse in CIS or during impair the uptake of microbiota-produced AHR ag- relapses in RRMS, AHR agonistic activity is increased onists and their precursors, as well as their activation as compared to controls or patients with RRMS with into potent AHR agonists. Collectively, these factors stable disease, respectively. By contrast, during stable may influence AHR-dependent immunoregulation disease, AHR ligand levels negatively correlate with in MS. disease severity, since patients with benign MS exhibit Inflammation seems to increase circulating AHR higher levels of AHR agonistic activity than patients agonists in MS. Inflammation has profound effects with MS suffering from more severe disease (figure 6). on metabolism. Indeed, it has been reported that Several factors might contribute to the decrease in the AHR agonist Kyn is produced by the metabolism circulating AHR agonists detected in patients with during inflammation.18,19 Thus, together with addi- MS. It has become clear in recent years that genetic tional AHR agonists that may be generated during polymorphisms correlate with an increased risk of inflammation, Kyn may participate in a negative feed- developing MS. While most of these polymorphisms back loop aimed at limiting immunopathology. This have been linked to the immune system,31–34 meta- anti-inflammatory mechanism may cross-talk with bolic pathways relevant to the uptake, activation, or additional immunoregulatory pathways40 and/or degradation of AHR ligands are also affected, as indi- DMTs. Type I interferons, for example, modulate cated by reported alterations in enzymes that catalyze Kyn levels41 in patients with MS.

Neurology: Neuroimmunology & Neuroinflammation 5 potential effects of specific therapies, might constitute Figure 6 Dynamic modulation of aryl hydrocarbon receptor ligand levels in patients with relapsing-remitting MS additional confounding factors. Indeed, some patients showed an increased activity of serum AHR ligands, the reasons for which are not clear as of now. Future longitudinal studies may be helpful in determining the clinical relevance of this observation. Moreover, our assay determines the net agonistic activity of AHR ligands in biological samples. Thus, relative changes in specific agonistic or inhibitory AHR ligand levels could be masked or missed by our approach. Finally, technical aspects need to be taken into consideration, since AHR ligand binding and activation has been shown to be species and cell line specific.42–45 Thus, the use of different cell lines or transfection techniques (e.g., stable vs transient transfection) may lead to varying results in individual assay systems. Based on our observations, it is tempting to speculate that different sources of AHR agonists drive chronic and acute AHR activation in MS. Chronic AHR activation may be controlled by the genetic background, diet, and/or the commensal flora, with potential confounding effects provided by environmental factors such as sun exposure and daylight that may differentially influence specific cohorts of patients with MS and controls.31,46 Acute AHR activation may be controlled by AHR-activating metabolites, such as Kyn, produced in the context of inflammation to limit immunopathology. The integration of these multiple sources of AHR agonists determines the contribution of AHR signaling to immune modulation. Longitudinal studies based on metabolomic approaches are therefore needed to analyze the correlation between specific AHR agonists, their sources, and disease activity in MS and, potentially, other conditions such as IBD. More importantly, given the potential of AHR agonists to cross (A) Serum aryl hydrocarbon receptor (AHR) agonistic activity is increased during the first the blood-brain barrier and modulate CNS inflam- relapse of the disease but decreases below control levels in remission. Further relapses 5 increase AHR ligand levels during the relapse. (B) Benign courses of MS have control mation, AHR activation could represent a novel levels of AHR agonistic activity and might be increased during relapses. RRMS 5 relapsing- therapeutic avenue for MS. remitting MS. AUTHOR CONTRIBUTIONS Veit Rothhammer: AHR ligand measurement, data analysis, data inter- Several limitations and potential confounding fac- pretation, and manuscript writing and revision. Davis M. Borucki: tors have to be taken into consideration when assess- AHR ligand measurement, data analysis, and manuscript revision. Maria ing AHR agonist levels in human samples in our Antonietta Mazzola, Christopher C. Hemond, Anu Paul, Maria Isabel study. First, some of our cohorts were limited in Garcia Sanchez, Guillermo Izquierdo, Keren Regev, Pia Kivisäkk, Rohit Bakshi, and Howard L. Weiner: providing of patient samples and clinical patient numbers and exhibited imperfect matching data. Francisco J. Quintana: design and supervision of the study, data of age, disease duration, or prevalence of DMT. analysis and interpretation, and manuscript writing, revision, and editing. Although we did not detect systematic changes when analyzing the correlation of these factors with agonis- ACKNOWLEDGMENT The authors acknowledge the participants and the healthy controls in tic activity (figure e-1), additional potentially this study for their participation and to the Nodo Biobanco Hospital- unknown variables, such as preanalytical sample pro- ario Virgen Macarena (Biobanco Sistema Sanitario Público de cessing, storage conditions, or selective AHR ligand Andalucía) for its help and support in the gifts of clinical samples used degradation or enrichment during sample preparation in this work. The Biobank is integrated in the Spanish Biobanks Net- work (RetBioH; redbiobancos.es), and supported by Instituto de Salud cannot be excluded. Also, cohort-specific differences, Carlos III, integrated in the national I1D1i 2013–2016 and cofunded including dietary factors, changes in the gut flora, and by the European Union (ERDF/ESF, “Investing in your future”)

6 Neurology: Neuroimmunology & Neuroinflammation (grant no. PT13/0010/0041). They also thank Shahamat Tauhid and 9. Zelante T, Iannitti RG, Cunha C, et al. Tryptophan ca- Sheena L. Dupuy for technical assistance. tabolites from microbiota engage aryl hydrocarbon receptor and balance mucosal reactivity via interleukin-22. STUDY FUNDING Immunity 2013;39:372–385. This work was supported by grants ES025530, AI093903, and AI126880 10. Opitz CA, Litzenburger UM, Sahm F, et al. An endoge- from the NIH and RG4111A1 and JF2161-A-5 from the National Multi- nous tumour-promoting ligand of the human aryl hydro- ple Sclerosis Society to F.J.Q. V.R. received support from an educational carbon receptor. Nature 2011;478:197–203. grant from Mallinckrodt Pharmaceuticals (A219074) and by a fellowship 11. Monteleone I, Rizzo A, Sarra M, et al. Aryl hydrocarbon from the German Research Foundation (DFG RO4866 1/1). receptor-induced signals up-regulate IL-22 production and inhibit inflammation in the gastrointestinal tract. Gastro- DISCLOSURE enterology 2011;141:237–248, 248.e1. V. Rothhammer, D.M. Borucki, M.I. Garcia Sanchez, and M.A. Mazzola 12. Benson JM, Shepherd DM. Dietary ligands of the aryl report no disclosures. C.C. Hemond received research support from the hydrocarbon receptor induce anti-inflammatory and American Academy of Neurology. K. Regev and A. Paul report no disclo- immunoregulatory effects on murine dendritic cells. Tox- sures. P. Kivisakk received research support from EMD Serono, Sanofi icol Sci 2011;124:327–338. Genzyme, and Verily Life Sciences. R. Bakshi is editor-in-chief of the 13. Apetoh L, Quintana FJ, Pot C, et al. The aryl hydrocarbon Journal of Neuroimaging, consulted for EMD Serono, Genentech, Sanofi Genzyme, and Novartis, and received research support from Biogen, receptor interacts with c-Maf to promote the differentia- EMD Serono, Novartis, and Sanofi Genzyme; Dr. Bakshi’s spouse holds tion of type 1 regulatory T cells induced by IL-27. Nat stock/stock options in Biogen Idec. G. Izquierdo served on the scientific Immunol 2010;11:854–861. advisory board for Biogen Idec, Bayer, Sanofi, Roche, Novartis, Gen- 14. Quintana FJ, Murugaiyan G, Farez MF, et al. An endog- zyme, Almirall, Merck Serono, and Teva and consulted for Sanofi, enous aryl hydrocarbon receptor ligand acts on dendritic Merck, and Novartis. H.L. Weiner served on the advisory board for cells and T cells to suppress experimental autoimmune The Guthy Jackson Charitable Foundation, Teva, Biogen Idec, Novartis, encephalomyelitis. Proc Natl Acad Sci USA 2010;107: and Sanofi-Aventis, consulted for Therapix, Biogen, Novartis, Serono, 20768–20773. Teva, and Sanofi, and received research support from the National Mul- 15. Gagliani N, Amezcua Vesely MC, Iseppon A, et al. Th17 tiple Sclerosis Society. F.J. Quintana served on the scientific advisory cells transdifferentiate into regulatory T cells during reso- board for Reva, served on the editorial board for Systems Biomedicine, – Inmunologia (Spain), American Journal of Clinical Experimental Immunol- lution of inflammation. Nature 2015;523:221 225. ogy, Seminars in immunopathology, and associate editor for Immunology 16. Gramatzki D, Pantazis G, Schittenhelm J, et al. Aryl (UK), consulted for Teva, Blueprint Medicines, and AnTolRx, and hydrocarbon receptor inhibition downregulates the TGF- received research support from EMD_Serono, Sanofi, Novartis, Teva, beta/Smad pathway in human glioblastoma cells. Onco- ONO, NIH, Harvard Medical School, BADERC, and NMSS. Go to gene 2009;28:2593–2605. Neurology.org/nn for full disclosure forms. 17. Song J, Clagett-Dame M, Peterson RE, et al. A ligand for the aryl hydrocarbon receptor isolated from lung. Proc Received February 1, 2017. Accepted in final form March 23, 2017. Natl Acad Sci USA 2002;99:14694–14699. 18. Platten M, Ho PP, Youssef S, et al. Treatment of autoim- REFERENCES mune neuroinflammation with a synthetic tryptophan 1. Stockinger B, Di Meglio P, Gialitakis M, Duarte JH. The metabolite. Science 2005;310:850–855. aryl hydrocarbon receptor: multitasking in the immune 19. Bessede A, Gargaro M, Pallotta MT, et al. Aryl hydrocar- system. Annu Rev Immunol 2014;32:403–432. bon receptor control of a disease tolerance defence path- 2. Mascanfroni ID, Yeste A, Vieira SM, et al. IL-27 acts on way. Nature 2014;511:184–190. DCs to suppress the T cell response and autoimmunity by 20. Varrin-Doyer M, Pekarek KL, Spencer CM, et al. Treat- inducing expression of the immunoregulatory molecule ment of spontaneous EAE by laquinimod reduces Tfh, B CD39. Nat Immunol 2013;14:1054–1063. cell aggregates, and disease progression. Neurol Neuroim- 3. Quintana FJ. The aryl hydrocarbon receptor: a molecular munol Neuroinflamm 2016;3:e272. doi: 10.1212/NXI. pathway for the environmental control of the immune 0000000000000272. response. Immunology 2013;138:183–189. 21. Kaye J, Piryatinsky V, Birnberg T, et al. Laquinimod ar- 4. Cella M, Colonna M. Aryl hydrocarbon receptor: linking rests experimental autoimmune encephalomyelitis by acti- environment to immunity. Semin Immunol 2015;27:310– vating the aryl hydrocarbon receptor. Proc Natl Acad Sci 314. USA 2016;113:E6145–E6152. 5. RothhammerV,MascanfroniID,BunseL,etal.TypeI 22. Berg J, Mahmoudjanlou Y, Duscha A, et al. The immu- interferons and microbial metabolites of tryptophan modulate nomodulatory effect of laquinimod in CNS autoimmunity astrocyte activity and central nervous system inflammation via is mediated by the aryl hydrocarbon receptor. the aryl hydrocarbon receptor. Nat Med 2016;22:586–597. J Neuroimmunol 2016;298:9–15. 6. Lamas B, Richard ML, Leducq V, et al. CARD9 impacts 23. Mishra MK, Wang J, Keough MB, et al. Laquinimod colitis by altering gut microbiota metabolism of trypto- reduces neuroaxonal injury through inhibiting microglial phan into aryl hydrocarbon receptor ligands. Nat Med activation. Ann Clin Transl Neurol 2014;1:409–422. 2016;22:598–605. 24. Jolivel V, Luessi F, Masri J, et al. Modulation of den- 7. Mascanfroni ID, Takenaka MC, Yeste A, et al. Metabolic dritic cell properties by laquinimod as a mechanism for control of type 1 regulatory T cell differentiation by AHR modulating multiple sclerosis. Brain 2013;136:1048– and HIF1-alpha. Nat Med 2015;21:638–646. 1066. 8. Fukumoto S, Toshimitsu T, Matsuoka S, et al. Identifica- 25. Schulze-Topphoff U, Shetty A, Varrin-Doyer M, et al. tion of a probiotic bacteria-derived activator of the aryl Laquinimod, a quinoline-3-carboxamide, induces type II hydrocarbon receptor that inhibits colitis. Immunol Cell myeloid cells that modulate central nervous system auto- Biol 2014;92:460–465. immunity. PLoS One 2012;7:e33797.

Neurology: Neuroimmunology & Neuroinflammation 7 26. Vollmer TL, Sorensen PS, Selmaj K, et al. A randomized remitting multiple sclerosis. PLoS One 2015;10: placebo-controlled phase III trial of oral laquinimod for e0130715. multiple sclerosis. J Neurol 2014;261:773–783. 37. Zhu WH, Lu CZ, Huang YM, Link H, Xiao BG. A 27. Garrison PM, Tullis K, Aarts JM, Brouwer A, Giesy JP, putative mechanism on remission of multiple sclerosis dur- Denison MS. Species-specific recombinant cell lines as ing pregnancy: estrogen-induced indoleamine 2,3- bioassay systems for the detection of 2,3,7,8- dioxygenase by dendritic cells. Mult Scler 2007;13:33–40. tetrachlorodibenzo-p-dioxin-like chemicals. Fundam Appl 38. Jangi S, Gandhi R, Cox LM, et al. Alterations of the Toxicol 1996;30:194–203. human gut microbiome in multiple sclerosis. Nat Com- 28. Klotz L, Berthele A, Brück W, et al. Monitoring of blood mun 2016;7:12015. parameters under course-modified MS therapy: substance- 39. Liu S, da Cunha AP, Rezende RM, et al. The host shapes specific relevance and current recommendations for action the gut microbiota via fecal microRNA. Cell Host [in German]. Nervenarzt 2016;87:645–659. Microbe 2016;19:32–43. 29. Marcus JF, Waubant EL. Updates on clinically isolated 40. Mimran A, Mor F, Carmi P, et al. DNA vaccination with syndrome and diagnostic criteria for multiple sclerosis. CD25 protects rats from adjuvant arthritis and induces an Neurohospitalist 2013;3:65–80. antiergotypic respone. J Clin Invest 2004;113:924–932. 30. Pittock SJ, McClelland RL, Mayr WT, et al. Clinical 41. Amirkhani A, Rajda C, Arvidsson B, et al. Interferon-beta implications of benign multiple sclerosis: a 20-year affects the tryptophan metabolism in multiple sclerosis population-based follow-up study. Ann Neurol 2004;56: patients. Eur J Neurol 2005;12:625–631. 303–306. 42. Flaveny CA, Murray IA, Chiaro CR, Perdew GH. Ligand 31. Rothhammer V, Quintana FJ. Environmental control of selectivity and gene regulation by the human aryl hydro- autoimmune inflammation in the central nervous system. carbon receptor in transgenic mice. Mol Pharmacol 2009; Curr Opin Immunol 2016;43:46–53. 75:1412–1420. 32. Hemmer B, Kerschensteiner M, Korn T. Role of the 43. Shiizaki K, Ohsako S, Kawanishi M, Yagi T. Identification innate and adaptive immune responses in the course of of amino acid residues in the ligand-binding domain of the multiple sclerosis. Lancet Neurol 2015;14:406–419. aryl hydrocarbon receptor causing the species-specific 33. International Multiple Sclerosis Genetics Consortium response to omeprazole: possible determinants for binding (IMSGC), Beecham AH, Patsopoulos NA, et al. Analysis putative endogenous ligands. Mol Pharmacol 2014;85: of immune-related loci identifies 48 new susceptibility var- 279–289. iants for multiple sclerosis. Nat Genet 2013;45:1353–1360. 44. Bazzi R, Bradshaw TD, Rowlands JC, Stevens MF, Bell 34. International Multiple Sclerosis Genetics Consortium; DR. 2-(4-Amino-3-methylphenyl)-5-fluorobenzothiazole Wellcome Trust Case Control Consortium 2, Sawcer S, is a ligand and shows species-specific partial agonism of et al. Genetic risk and a primary role for cell-mediated the aryl hydrocarbon receptor. Toxicol Appl Pharmacol immune mechanisms in multiple sclerosis. Nature 2011; 2009;237:102–110. 476:214–219. 45. Henry EC, Gasiewicz TA. Molecular determinants of 35. Lanz TV, Williams SK, Stojic A, et al. Tryptophan-2,3- species-specific agonist and antagonist activity of Dioxygenase (TDO) deficiency is associated with subclinical a substituted flavone towards the aryl hydrocarbon recep- neuroprotection in a mouse model of multiple sclerosis. Sci tor. Arch Biochem Biophys 2008;472:77–88. Rep 2017;7:41271. 46. Farez MF, Mascanfroni ID, Méndez-Huergo SP, et al. 36. Mancuso R, Hernis A, Agostini S, et al. Indoleamine 2,3 Melatonin contributes to the seasonality of multiple scle- dioxygenase (IDO) expression and activity in relapsing- rosis relapses. Cell 2015;162:1338–1352.

8 Neurology: Neuroimmunology & Neuroinflammation Disruption of the leptomeningeal blood barrier in neuromyelitis optica spectrum disorder

Nasrin Asgari, MD, PhD ABSTRACT Eoin P. Flanagan, MD Objective: To describe leptomeningeal blood-barrier impairment reflected by MRI gadolinium- Kazuo Fujihara, MD enhanced lesions in patients with aquaporin-4 immunoglobulin G (AQP4-IgG)–positive neuro- Ho Jin Kim, MD myelitis optica spectrum disorder (NMOSD). Hanne P. Skejoe, MD Methods: A retrospective case series of 11 AQP4-IgG–positive NMOSD patients with lep- Jens Wuerfel, MD tomeningeal enhancement (LME) were collected from 5 centers. External neuroradiologists, Hiroshi Kuroda, MD blinded to the clinical details, evaluated MRIs. Su Hyun Kim, MD Elisabeth Maillart, MD Results: LME was demonstrated on postcontrast T1-weighted and fluid-attenuated inversion recovery Romain Marignier, MD images as a sign of leptomeningeal blood-barrier disruption and transient leakage of contrast agent into Sean J. Pittock, MD the subarachnoid space in 11 patients, 6 in the brain and 6 in the spinal cord. The patterns of LME were Friedemann Paul, MD* linear or extensive and were accompanied by periependymal enhancement in 5 cases and intraparen- Brian G. Weinshenker, chymal enhancement in all cases. The location of LME in the spinal cord was adjacent to intraparenchy- MD* mal contrast enhancement with involvement of a median number of 12 (range 5–17) vertebral segments. At the time of LME on MRI, all patients had a clinical attack such as encephalopathy (36%) and/or myelopathy (70%) with median interval between symptom onset and LME of 12 days

Correspondence to (range 2–30). LME occurred in association with an initial area postrema attack (44%), signs of systemic Dr. Asgari: infection (33%), or AQP4-IgG in CSF (22%) followed by clinical progression. LME was found at initial [email protected] clinical presentation in 5 cases and at clinical relapses leading to a diagnosis of NMOSD in 6 cases. Conclusion: This study suggests that altered leptomeningeal blood barrier may be accompanied by intraparenchymal blood-brain barrier breakdown in patients with AQP4-IgG–positive NMOSD during relapses. Neurol Neuroimmunol Neuroinflamm 2017;4:e343; doi: 10.1212/NXI.0000000000000343

GLOSSARY AP 5 area postrema syndrome/lesion; AQP4 5 aquaporin-4; BBB 5 blood-brain barrier; CS 5 cerebral syndromes; DWI 5 diffusion-weighted imaging; FLAIR 5 fluid-attenuated inversion recovery; Gd 5 gadolinium; IVMP 5 IV methylprednisolone; LETM 5 longitudinally extensive transverse myelitis; LME 5 leptomeningeal enhancement; MNL 5 mononuclear leukocyte; NMOSD 5 neuromyelitis optica spectrum disorder; ON 5 optic neuritis; PLEX 5 plasma exchange; PMNL 5 polymorphonuclear leukocyte; STIR 5 short tau inversion recovery; T1W 5 T1 weighted.

Neuromyelitis optica spectrum disorder (NMOSD) is an inflammatory, autoimmune disease of the CNS that follows a relapsing course in the majority of patients. NMOSD-related neurologic disability frequently occurs as a consequence of cumulative attacks.1 The inflammation of NMOSD is often associated with blood-brain barrier (BBB) leakage, reflected by MRI Supplemental data at Neurology.org/nn *These authors contributed equally to this work as co-last authors. From the Department of Neurobiology (N.A.), Institute of Molecular Medicine, University of Southern Denmark; Department of Neurology (E.P.F., S.J.P., B.G.W.), Mayo Clinic, Rochester, MN; Department of Multiple Sclerosis Therapeutics (K.F.), Fukushima Medical University School of Medicine; Multiple Sclerosis and Neuromyelitis Optica Center (K.F.), Southern TOHOKU Research Institute for Neuroscience, Koriyama, Japan; Department of Neurology (H.J.K., S.H.K.), Research Institute and Hospital of National Cancer Center, Goyang, Korea; Department of Radiology (H.P.S.), Aleris-Hamlet Hospital, Copenhagen, Denmark; Medical Image Analysis Center Basel (J.W.); Department of Biomedical Engineering (J.W.), University Basel, Switzerland; NeuroCure Clinical Research Center and Clinical and Experimental Multiple Sclerosis Research Center (J.W., F.P.), Department of Neurology, Charité Universitätsmedizin Berlin; Experimental and Clinical Research Center (J.W., F.P.), Max Delbrueck Center for Molecular Medicine and Charité Uni- versitätsmedizin Berlin, Germany; Department of Neurology (H.K.), Tohoku University Graduate School of Medicine, Sendai, Japan; Department of Neurology (E.M.), Hôpital Pitié-Salpêtrière, APHP, Paris, France; Service de Neurologie A and Eugène Devic EDMUS Foundation against Multiple Sclerosis (R.M.), Observatoire Français de la Sclérose en Plaques (OFSEP), Hôpital Neurologique Pierre Wertheimer, Hospices Civils de Lyon, Bron; and Lyon Neurosciences Research Center (R.M.), FLUID team, Inserm U 1028/CNRS 5292, France. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 gadolinium (Gd)-enhanced lesions.2–4 Serum- T2W and T1W with or without Gd and short tau inversion derived aquaporin-4 immunoglobulin G recovery (STIR) sequences were available for spinal cord images. Precontrast and postcontrast images were evaluated to determine (AQP4-IgG) antibody is believed to play an the presence of abnormal contrast enhancement on T1W and/or important and potentially pivotal role in FLAIR sequences according to the criteria previously reported.12–14 If NMOSD pathogenesis1; however, the mecha- performed, postcontrast FLAIR sequences obtained at least nism whereby AQP4-IgG contributes to BBB 10 minutes after a single dose of contrast were also used for evaluation. leakage remains poorly characterized. The Standard protocol approvals, registrations, and patient autoantigen AQP4 is densely localized on consents. All patients provided oral and written informed con- ependymal cells and astrocyte end-feet, form- sent for publication. ing the glia limitans of the BBB.5 The lack of a basement membrane and expression of tight RESULTS Retrospective MRI and clinical data from 11 AQP4-IgG–positive (measured either in a cell- junctions in ependymal cells may allow based assay or ELISA15) NMOSD patients with lep- AQP4-IgG in the CSF to get access to its tomeningeal enhancement (LME) were obtained for 5–7 target. AQP4-IgG may exert pathologic ef- review (table 1). The female: male ratio was 9:2, and fects by binding selectively to AQP4 on CNS the median age at onset of disease was 40.8 years (range astrocytes, inducing complement-dependent 15–71 years). Seven patients were East Asian, 1 African astrocyte damage and inflammation.8 American, 1 of mixed ancestry, and 2 whites. MRI-detected leptomeningeal enhance- Radiologic characteristics. LME localization. Contrast ment (LME), which reflects the leptomenin- enhancement in the leptomeningeal space was evalu- geal blood-barrier disruption, has been ated by precontrast and postcontrast T1W and/or reported in patients with NMOSD.2 LME re- FLAIR images and observed in all 11 patients, 6 in fers to an enhancement pattern that follows the brain and 6 in the spinal cord; 1 patient had both the pial surface of the CNS and fills the sub- brain and spinal cord LME. arachnoid spaces of sulci and cisterns. The In the brain, T1W contrast-enhanced images re- LME pattern depends on the extent and inten- vealed a linear or thick and extensive leptomeningeal and pial enhancement with localization in both cere- sity of the involved inflammatory area and bral hemispheres (2 patients), bilateral medial surfaces may provide useful clues to the diagnosis of, of the cerebral hemispheres (1), bilateral frontal, 9 e.g., neurosarcoidosis. medial parietal lobe (1), right parietal lobe (1), and In this retrospective case study, we investi- bilateral parieto-occipital lobes (1) (figure 1, A.a and b, gated whether compromised integrity of the C.a and b). FLAIR (figure 1, A.c, B.c, C.c) and intraparenchymal BBB occurs concurrently FLAIR postcontrast (figure 1B.a) sequences revealed with LME and whether such an alteration of similar leptomeningeal signal intensity changes. BBB is associated with disease activity in pa- In the spinal cord, T1W contrast-enhanced im- tients with AQP4-IgG–positive NMOSD. ages showed LME in 6 patients, conus to cauda equina (2), multifocal spinal cord (1), and thoracic METHODS Study design and population. The study was segments (3) (figure 2, A.a and c, B.a and C, figure designed as a retrospective case series with clinical and radiologic 3B, figure 4, A and D, figure e-1, H1 and H3 at follow-up and collected by a multicenter study group in 5 centers Neurology.org/nn). (Japan, South Korea, United States, France, and Denmark). Pa- Follow-up MRI demonstrated resolution of LME tients were included if they met the following criteria: (1) AQP4- within 1–4 weeks in 9 patients. In one other case, the IgG–positive NMOSD, qualifying for the 2006 Wingerchuk criteria10 or/and the International Panel for NMO Diagnosis next MRI was performed after 72 days when the (IPND) 201511 and (2) LME as visualized by MRI following the LME had disappeared. One patient did not have a standard protocols.12,13 Patients were excluded if they had (1) follow-up MRI. systemic inflammatory disease, e.g., sarcoidosis or (2) neoplastic LME in association with parenchymal enhancement. disease. The information was obtained by review of medical re- LME in all cases spatially corresponded to intraparen- cords and reevaluation of previous MRIs of the CNS by local chymal T1W contrast enhancement or/and FLAIR neuroradiologists. In addition, 2 expert neuroradiologists not involved in patient management evaluated images and were image abnormalities in the brain (figure 1, A.c, B.c, blinded to the clinical details (H.P.S. and J.W.). C.b and c). Periependymal linear T1W contrast enhancement was localized at the lateral ventricles (3 Neuroimaging. Because the study was retrospective, several patients) (figure 1B.b, figure 3D) and adjacent to the types of 1.5T MRI scanners were used. T2-weighted (T2W) and T1-weighted (T1W) images with or without Gd, diffusion- third and fourth ventricles (2) (figure 3F). Intracerebral weighted imaging, and fluid-attenuated inversion recovery cloud-like enhancement (3 patients) (figure 1C.c), (FLAIR) sequences were available for analysis for all brain scans. enhancement of hypothalamus (2) (figure 3G), and optic

2 Neurology: Neuroimmunology & Neuroinflammation Spinal cord MRI demonstrated intraparenchymal Table 1 Clinical characteristics of the NMOSD patients with LME T1W contrast enhancement in the underlying paren-

NMOSD 1 AQP4-IgG 1 LME chyma adjacent to the LME (figure 2, A.a and B.a, Patient characteristics (n 5 11) figure 3B, figure 4B, figure e-1H1), involving cervi- Sex cal, thoracic, and lumbar cords. Spinal cord lesions F 9 expanded over a median of 12 segments (range 5–17)

M 2 as demonstrated on contrast-enhanced T1W (figure 2, A.a and B.a, figure 3B, figure 4B, figure e-1H1) Ethnicity and T2W and STIR images (figure 2, A.b and d, B. East Asian 7 b and d, figure 3, A and C, figure 4, C and E, figure African American 1 e-1, H2 and H4). Mixed ancestry 1 CSF abnormalities. Lumbar puncture was performed White 2 after MRI investigation in all patients, and complete Median age, y (range) 40.8 (15–71) CSF analysis was available at the time of the LME for Median interval between symptom onset and 12 (2–30) 6 of the 11 patients. CSF examination showed pleo- detection of LME on MRI, d (range) cytosis .50 mL (/mm3) in all cases; 2 cases with Acute attack at the time of LME 11 increased polymorphonuclear leukocytes (PMNLs) Attack with LME as the initial clinical 5 had 80% (PMNL 166 and mononuclear leukocyte presentation of NMOSD [MNL] 41/mm3) and 98% (PMNL 212 and MNL Established NMOSD diagnosis at the time of 6 3 attack w/LME 4/mm ) PMNL, respectively. The combination of

LETM 4 LME and pleocytosis in all cases, especially in the 2 cases with increased CSF-PMNL and fever, led to CS 1 previous TM 1 extensive virologic and bacteriologic workup, which 1 LETM previous ON 1 proved negative both in blood and CSF. The median Clinical features at the time of LME on MRI CSF protein concentration was 169 (range 54–322) LETM 7 (mg/dL), and CSF glucose level was normal in all CS 4 cases. CSF-restricted oligoclonal bands or elevated

AP 1 LETM 3 IgG CSF/serum ratio were absent, except in 1 patient. AQP4-IgG in CSF was only measured in 2 patients, AP 1 CS 1 and it was positive in both patients. TM 1 CS 1

Median follow-up time after LME episode, mo 79 (2–186) LME and clinical disease activity. When LME was vis- (range) ible on MRI, all patients had experienced attacks with Disease course subacute onset and progressive neurologic deficits

Relapsing (F/M) 8 (88%/12%) over a few weeks (7 patients) or with acute onset

Monophasic (F/M) 3 (67%/33%) and rapid progression within a few days (4 patients). Median interval between symptom onset and MRI Acute treatment IVMP (11) (additional PLEX in 7 patients with LETM) scans displaying LME was 12 days (range 2–30 days).

Preventive therapy after attack with LME Seven patients had longitudinally extensive transverse myelitis (LETM) (4 subacute and 3 acute onset) and Oral corticosteroids 2 4 had cerebral syndromes (CS) (3 with subacute and Azathioprine 2 1 with acute onset) as the clinical presentation asso- Azathioprine 1 oral corticosteroids 1 ciated with LME (table e-1). 1 Cyclosporine oral corticosteroids 1 Spinal cord involvement. In the LETM group, 3 of Mycophenolate mofetil 1 the 4 patients with subacute onset initially had expe- Mitoxantrone 2 rienced an area postrema syndrome/lesion (episodes – Rituximab 2 of nausea, vomiting, and hiccups) (figure 3, A C) followed a few days later by extremity weakness, Abbreviations: AP 5 area postrema syndrome/lesion; AQP4-IgG 5 aquaporin-4 immunoglobulin G; CS 5 cerebral syndrome; IVMP 5 IV methylprednisolone; LETM 5 longitudinally exten- which progressed to tetraplegia (1 patient) or para- sive transverse myelitis; LME 5 leptomeningeal enhancement; NMOSD 5 neuromyelitis optica plegia (2) over 7, 12, and 30 days. MRI of the spinal spectrum disorder; ON 5 optic neuritis; PLEX 5 plasma exchange; TM 5 transverse myelitis. cord at this time revealed LETM with LME adjacent to LETM. Another patient with a subacute onset of chiasm (1) were observed on T1W contrast-enhanced LETM had atypical chest pain as the initial symptom, or/and FLAIR sequences. In addition, parenchymal le- which after 7 days progressed to paraplegia and MRI sions were observed on T2W or/and FLAIR sequences of the spinal cord revealed LME and LETM (figure 4, (figure 1, A.c, B.c, C.c, figure 3, A and E). A–E).

Neurology: Neuroimmunology & Neuroinflammation 3 Figure 1 Leptomeningeal enhancement accompanied by intraparenchymal enhancement during attacks in patients with NMOSD

Contrast-enhanced cerebral T1W (A.a, A.b, B.b, C.a, and C.b), FLAIR sequences (A.c, B.a, B.c, and C.c) and FLAIR post– contrast-enhanced image (B.a) MRI from 3 patients with AQP4-IgG–positive NMOSD (A, B, and C) during attacks, showing leptomeningeal enhancement (A.a, A.b, B.a, C.a, and C.b), marked with arrows. Parenchymal lesions are seen including cloud- like enhancement (B.c, A.c, and C.c). Notably, enhancement occurred at the ependymal surface of the lateral ventricles (B.a). AQP4-IgG 5 aquaporin-4 immunoglobulin G; FLAIR 5 fluid-attenuated inversion recovery; NMOSD 5 neuromyelitis optica spectrum disorder; T1W 5 T1 weighted.

In the remaining 3 patients with acute onset of 17 days after onset of symptoms revealed LME (figure LETM, 2 developed symptom/signs after an upper 1, A.a–c). Another patient experienced dizziness and respiratory tract infection, by 2 and 4 days respec- was treated with corticosteroids. A third patient had tively (without the identification of an infectious nonspecific symptoms such as drowsy mental state, agent) and 1 occurred 2 days after an induced abor- dysarthria, and recurrent episode of myelitis and optic tion. One presented with tetraplegia and 2 with para- neuritis (ON). LME were in those 2 cases who dem- plegia. MRI of spinal cord revealed LETM with LME onstrated on MRI 24 and 21 days after symptom adjacent to the intramedullary lesion (H1-H4). onset, respectively. The one patient with acute symp- Brain involvement. In the CS group, 1 patient ini- tom onset had a generalized seizure followed by tet- tially experienced an area postrema syndrome fol- raparesis from TM. LME was in this case lowed 10 days later by altered consciousness and demonstrated on MRI 5 days after symptom onset recurrent generalized seizures. Brain MRI performed (table e-1).

4 Neurology: Neuroimmunology & Neuroinflammation Figure 2 LETM in association with leptomeningeal enhancement

Sagittal (A.a and B.a) and axial (A.c and B.c) T1W contrast-enhanced images and sagittal T2W (A.b and B.b) and axial images (A.d and B.d) of spinal cord MRI from 2 (A and B) AQP4-IgG–positive patients with NMOSD. Images were obtained during acute LETM attacks with leptomeningeal and intraparenchymal (A.a and A.c at th10; B.a and B.c at th12) T1W contrast enhancement. The spinal cord lesions were also demonstrated on T2W and STIR images (A.b, A.d, B.b, and B.d). AQP4-IgG 5 aquaporin-4 immunoglobulin G; LETM 5 longitudinally extensive transverse myelitis; NMOSD 5 neuromyelitis optica spectrum disorder; STIR 5 short tau inversion recovery; T1W 5 T1 weighted; T2W 5 T2 weighted.

Acute and long-term immunotherapy. In all patients, LME and acute attack, 1 with LETM, and 1 with acute attacks were initially treated with IV methylpred- a generalized seizure followed by TM (table 1). nisolone (IVMP), which were administered between Three patients with previous attacks received the the first and the follow-up MRI. The 7 patients with NMOSD diagnosis at median 20 months (range LETM did not clinically respond to IVMP therapy 12–24) after attack with LME on MRI, with disease and were subsequently treated by plasma exchange. duration of median 116 months (range 108–132). Various immunosuppressive drugs as listed in They were suspected for MS and treated with table 1 were used as preventive therapy after the interferon-b previously, relapses occurred and treat- follow-up MRI, which showed resolution of LME; ment was discontinued. One patient was diagnosed one received azathioprine before the attack and with NMOSD 2 years before an attack with LME follow-up MRI examination. on MRI. She was treated with azathioprine. For all LME and NMOSD diagnosis. LME was found at the patients, median follow-up after onset of attack and initial attack of NMOSD in 5 (3: LETM, 2: area LME on MRI was 79.6 months (range 2–186). postrema syndrome) of the 11 patients (table 1). In the remaining 6 patients, symptoms at first attack DISCUSSION In this retrospective case series com- were ON (2 patients), TM (1), AP (1), and CS (2). prising 11 patients with AQP4-IgG–positive A total of 6 patients, including all patients who NMOSD, we observed contrast enhancement of had LME at the initial attack, qualified for the diag- the leptomeningeal structures during clinical attacks nosis of NMOSD except one, with area postrema accompanied by intraparenchymal contrast syndrome, who received the NMOSD diagnosis 2 enhancement spatially corresponding to the LME. years later at testing for AQP4-IgG. Of the 6 patients, LME was observed in association with either an initial 2 had a history of TM and ON, respectively, and area postrema attack, with precedent signs of systemic received the diagnosis of NMOSD in association with infection or with detectable AQP4-IgG in CSF,

Neurology: Neuroimmunology & Neuroinflammation 5 dysfunction (figure e-2). It could be speculated that Figure 3 An area postrema lesion in association with leptomeningeal enhancement and intraparenchymal BBB breakdown AQP4-IgG may be deposited at the astrocytic foot processes at pial vessels and cause astrocytic damage. The resultant dysfunction of AQP4 water channels in the pial and subpial surfaces could lead to impairment of leptomeningeal vessel barrier, which may be involved in the development of NMOSD lesions. In line with this study, pronounced meningeal inflammation has been reported in autopsy findings of patients with NMOSD.16 Our recent experimental study of intrathecal injection of AQP4-IgG may cor- roborate these findings, as pronounced deposition of AQP4-IgG was observed in the leptomeningeal area, including where pial vessels penetrate the brain parenchyma and form perivascular spaces into the brain parenchyma,17 the so-called paravascular pathway.18 Furthermore, AQP4-IgG initiated perivascular astrocyte-destructive lesions and consequently breakdown of the BBB.17 A significant proportion of subarachnoid CSF circulates through the parenchyma within these paravascular spaces19 which surround penetrating and cerebral surface vessels20–22 and is completely ensheathed by astrocyte end-feet.22 Of interest, experimental data have suggested that the flow of fluid along this paravascular route is mediated by astrocytic AQP4.19 Although direct inferences from an animal study are not possible, our observations suggest that intraparenchymal contrast enhancement lesions may form in anatomical proximity to regions where LME is vis- ible. LME was spatially associated with periependymal enhancement and with intraparenchymal enhancement in the brain, brainstem, hypothalamus, optic chiasm, and spinal cord. These CNS MRI abnormalities in patients with AQP4-IgG–positive NMOSD are believed to be common and typical of the disease with regards to localization and configu- ration.2,23 Conversely, LME has been reported in few Spinal cord MRI (A–C) and brain MRI (D–G) from an AQP4-IgG–positive patient with NMOSD who presented with recurrent myelopathy of 1 month duration and area postrema lesion, patients only and is presumably more common dur- which progressed into LETM. Images were obtained during acute attacks. An area postrema ing NMOSD relapses and in AQP4-IgG–positive lesion (A) in association with subacute progression to LETM with leptomeningeal and intra- patients.23,24 However, the frequency of LME is parenchymal T1W contrast enhancement (arrows, B). The spinal cord lesions were also likely to be underestimated given that most MRIs are demonstrated on T2W and STIR images (C). Cerebral MRI during another attack with periependymal enhancement on T1W images at the lateral (D) and third ventricles (F). FLAIR performed subsequently at onset of clinical symptoms images showing lesions along lateral ventricle (E), thalamus, and hypothalamus (G). (B) Used and thus at a time point where lesion formation is with permission of the publisher from Flanagan EP, Weinshenker BG. Neuromyelitis optica probably completed and LME may have vanished. spectrum disorders. Curr Neurol Neurosci Rep 2014;14:483. Copyright © 2014, Springer Timing of MRI may be important for the demon- Science1Business Media New York. AQP4-IgG 5 aquaporin-4 immunoglobulin G; BBB 5 blood-brain barrier; FLAIR 5 fluid-attenuated inversion recovery; LETM 5 longitudinally stration of LME, and evaluation of MRI should extensive transverse myelitis; NMOSD 5 neuromyelitis optica spectrum disorder; T1W 5 T1 consider the interval between symptom onset and weighted; T2W 5 T2 weighted. MRI scan. Last, LME may be suppressed by steroid treatment as immunosuppressive therapy probably followed by secondary worsening of symptoms/signs decreases BBB permeability.25 It has previously been and lesions on MRI (figure e-2). These findings reported that corticosteroid treatment before MRI suggest that, in some cases, entry of serum-derived may lead to the improvement or resolution of LME9 AQP4-IgG to the subarachnoid space may be facili- as part of the treatment response. In this study, pa- tated by primary inflammation at area postrema tients received treatment with IV corticosteroid by systemic infection or by blood-CSF barrier before follow-up MRI, and this may have led to the

6 Neurology: Neuroimmunology & Neuroinflammation between the leptomeningeal blood-barrier disruption Figure 4 Leptomeningeal enhancement in association with the presence of AQP4-IgG in CSF and LETM and BBB breakdown sites and severity of attacks require larger and prospective studies. In this study, all cases were associated with disease activity at the time of LME on MRI, and LME was accompanied by other characteristic findings such as LETM, area postrema lesion etc.11 In 7 of the 11 patients, the attacks occurred subacutely with a dynamic progression of symptoms/signs. Four of the seven patients had an initial area postrema syndrome. The dynamic formation of an LETM was previously reported with initial involvement of the area postrema during 12 days from the onset of symptoms as visualized on sequential brainstem/spinal cord MRI.26 These findings support the concept that area postrema, a circumventricular area which has incomplete BBB, may serve as entry route to the CNS for circulating AQP4-IgG.20,26,27 In NMOSD, the presence of AQP4-IgG in CSF has been reported to be associated with high titers of AQP4-IgG in serum,28–30 and to be strongly correlated with astrocyte damage, as reflected in elevated levels of glial fibrillary acidic protein in the CSF and BBB disruption during disease attacks.30,31 In this small retrospective study, titers of AQP4-IgG in serum were not determined and AQP4-IgG in CSF was only measured in 2 patients, both of whom had positive results. Infections may result in events that alter the BBB and thus permit access of AQP4-IgG to the CNS.32,33 Three cases with LETM were preceded acutely by symptoms suggestive of systemic infection. Cases of infectious or parainfectious events before the onset of neurologic symptoms and NMOSD have been reported, adding to the etiological diversity in NMOSD.32–36 It is possible that infection may act as a precipitating factor in the development of Sagittal contrast-enhanced T1W (A and B), sagittal (C), axial (D), and axial (E) T2W MRIs of the spinal cord of an AQP4-IgG–positive NMOSD patient during attacks, showing lep- NMOSD with the presence of AQP4-IgG. tomeningeal and intraparenchymal T1W contrast enhancement (A, B, and D). The spinal cord We observed that leptomeningeal blood-barrier lesions were also demonstrated on T2W and STIR images (C and E). The patient had a sub- integrity was compromised during relapses and was acute onset with atypical chest pain as the initial symptom, which after 7 days progressed to paraplegia and LETM on MRI. This patient was positive for AQP4-IgG in CSF. AQP4-IgG 5 accompanied by intraparenchymal BBB breakdown aquaporin-4 immunoglobulin G; LETM 5 longitudinally extensive transverse myelitis; in patients with AQP4-IgG–positive NMOSD. NMOSD 5 neuromyelitis optica spectrum disorder; STIR 5 short tau inversion recovery; Although our data are too preliminary to allow for 5 5 T1W T1 weighted; T2W T2 weighted. a firm statement on the putative mechanism of lesion formation in human NMOSD, they suggest in con- resolution of LME. The temporal evolution of alter- junction with previous animal experiments that the ations to the BBB permeability is difficult to deter- question of how pathogenic AQP4-IgG gains access mine and will require a kinetic study with sequential to its CNS target deserves further attention. More- and frequent MRI images that analyze the dynamic over, future imaging studies may inform us about the changes of enhancement in newly appearing active relevance of LME for attack severity and clinical lesions. The localization of the LME at the spinal cord course in NMOSD. in 70% of cases was associated with extremely long acute LETM lesions, and the patients were resistant AUTHOR CONTRIBUTIONS to steroid therapy. This may on MRI reflect the sever- N.A.: study concept and design, acquisition of data, interpretation of results, drafting, and revising of the manuscript. E.P.F., K.F., and H.J.K.: ity of attacks in the spinal cord at the time of LME. acquisition of data, interpretation of results, revising of the manuscript, However, such determination of an association and approving the final version. H.P.S. and J.W.: MRI reevaluation,

Neurology: Neuroimmunology & Neuroinflammation 7 interpretation of results, revising of the manuscript, and approving the B.G. Weinshenker served on the scientific advisory board for Novartis, final version. H.K., S.H.K., and E.M.: acquisition of data, revising of Biogen Idec, and Mitsubishi; served on the editorial board for Canadian the manuscript, and approving the final version. R.M. and S.J.P.: acqui- Journal of Neurological Sciences, Turkish Journal of Neurology,andNeurology; sition of data, interpretation of results, revising of the manuscript, and holds a patent for and receives royalties from NMO-IgG for diagnosis of approving the final version. F.P.: study design, interpretation of results, neuromyelitis optica; is an adjudication committee member for MedImmune revising of the manuscript, and approving the final version. B.G.W.: and Alexion; and received research support from Guthy-Jackson Charitable study design, acquisition of data, interpretation of results, revising of Foundation. Go to Neurology.org/nn for full disclosure forms. the manuscript, and approving the final version.

Received January 2, 2017. Accepted in final form March 2, 2017. STUDY FUNDING No targeted funding reported. REFERENCES 1. Wingerchuk DM, Lennon VA, Lucchinetti CF, Pittock SJ, DISCLOSURE Weinshenker BG. The spectrum of neuromyelitis optica. N. Asgari received research support from The Vejle Hospital Research Lancet Neurol 2007;6:805–815. Fund of the Region of Southern Denmark, Lundbeck Research Founda- 2. Kim HJ, Paul F, Lana-Peixoto MA, et al. MRI character- tion, and The University of Southern Denmark. E.O. Flanagan reports istics of neuromyelitis optica spectrum disorder: an inter- no disclosures. K. Fujihara served on the scientific advisory board for national update. Neurology 2015;84:1165–1173. Bayer Schering Pharma, Biogen Idec, Mitsubishi Tanabe Pharma Corpo- 3. Kim W, Park MS, Lee SH, et al. Characteristic brain magnetic ration, Novartis Pharma, Chugai Pharmaceutical, Ono Pharmaceutical, resonance imaging abnormalities in central nervous system Nihon Pharmaceutical, Alexion Pharmaceuticals, and MedImmune and aquaporin-4 autoimmunity. Mult Scler 2010;16:1229–1236. Medical Review; received travel funding and/or speaker honoraria from 4. Asgari N, Skejoe HP, Lillevang ST, Steenstrup T, Stenager E, Bayer Schering Pharma, Biogen Idec, Eisai Inc, Mitsubishi Tanable Pharma Corporation, Novartis Pharma, Astellas Pharma Inc, Takeda Pharmaceutical Kyvik KO. Modifications of longitudinally extensive Company Limited, Asahi Kasei Medical Co, Daiichi Sankyo, and transverse myelitis and brainstem lesions in the course Nihon Pharmaceutical; served on the editorial board for Clinical and of neuromyelitis optica (NMO): a population-based, Experimental Neuroimmunology; an advisory board member for Sri Lanka descriptive study. BMC Neurol 2013;13:33. Journal of Neurology; received research support from Bayer Schering Phar- 5. Nielsen S, Nagelhus EA, Amiry-Moghaddam M, Bourque C, ma, Biogen Idec Japan, Asahi Kasei Medical, The Chemo-Sero- Agre P, Ottersen OP. Specialized membrane domains for Therapeutic Research Institute, Teva Pharmaceutical, Mitsubishi Tanabe water transport in glial cells: high-resolution immunogold Pharma, Teijin Pharma, Chugai Pharmaceutical, Ono Pharmaceutical, cytochemistry of aquaporin-4 in rat brain. J Neurosci 1997; Nihon Pharmaceutical, Genzyme Japan, Grants-in-Aid for Scientific 17:171–180. Research from the Ministry of Education, Science and Technology of 6. Rash JE, Yasumura T, Hudson CS, Agre P, Nielsen S. Japan, and Ministry of Health, Welfare and Labor of Japan. H.J. Kim received travel funding and/or speaker honoraria from Bayer Schering Direct immunogold labeling of aquaporin-4 in square ar- Pharma, Biogen Idec, Genzyme, MedImmune, Merck, Novartis, rays of astrocyte and ependymocyte plasma membranes in Teva-Handok, and UCB; served as an associate editor for Journal of rat brain and spinal cord. Proc Natl Acad Sci USA 1998; Clinical Neurology, co-editor of Multiple Sclerosis Journal—Experimental, 95:11981–11986. Translational and Clinical; consulted for Bayer Schering Pharma, Biogen 7. Jarius S, Paul F, Franciotta D, et al. Mechanisms of dis- Idec, Genzyme, HanAll Biopharma, MedImmune, Merck, Novartis, ease: aquaporin-4 antibodies in neuromyelitis optica. Nat Teva-Handock, and UCB; received research support from Genzyme, Clin Pract Neurol 2008;4:202–214. Kael-GemVax, Merck Serono, Teva-Handock, UCB, and The Ministry of 8. Lucchinetti CF, Mandler RN, McGavern D, et al. A role Science, ICT and Future Planning. H.P. Skejoe reports no disclosures. J. for humoral mechanisms in the pathogenesis of Devic’s Wuerfel served on the scientific advisory board for Biogen, Novartis, Gen- neuromyelitis optica. Brain 2002;125:1450–1461. zyme, Teva, and Roche; received travel funding and/or speaker honoraria from Novartis, Bayer, Biogen, and Novartis; was CEO at MIAC 9. Moravan M, Segal BM. Treatment of CNS sarcoidosis AG; received support from German Ministry of Education and Research, with infliximab and mycophenolate mofetil. Neurology German Ministry of Economy, EXIST program, and ZIM program. 2009;72:337–340. H. Kuroda received research support from Ministry of Education, Culture, 10. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Sports, Science and Technology. S.H. Kim reports no disclosures. Weinshenker BG. Revised diagnostic criteria for neuromyelitis E. Maillart served on the scientific advisory board for Genzyme, Roche, optica. Neurology 2006;66:1485–1489. Novartis, and Merck; received travel funding and/or speaker honoraria 11. Wingerchuk DM, Banwell B, Bennett JL, et al. Interna- from Biogen, Teva Pharmaceuticals, Genzyme, Roche, and Novartis. tional consensus diagnostic criteria for neuromyelitis opti- R. Marignier served on the scientific advisory board for MedImmune. ca spectrum disorders. Neurology 2015;85:177–189. S.J. Pittock holds patents that relate to functional AQP4/NMO-IgG 12. Fukui MB, Meltzer CC, Kanal E, Smirniotopoulos JG. assays and NMO-IgG as a cancer marker; has consulted for Alexion Pharma, MedImmune, and Chugai Pharma (all consulting compensation MR imaging of the meninges: part II: neoplastic disease. – is paid to Mayo Clinic); received research support from Alexion, RO1, Radiology 1996;201:605 612. and Guthy-Jackson Charitable Foundation Grant. F. Paul served on the 13. Meltzer CC, Fukui MB, Kanal E, Smirniotopoulos JG. MR scientific advisory board for Novartis OCTIMS study and MedImmune; imaging of the meninges: part I: normal anatomic features received travel funding and/or speaker honoraria from Bayer, Novartis, and nonneoplastic disease. Radiology 1996;201:297–308. Biogen Idec, Teva, Sanofi-Aventis/Genzyme, Merck Serono, Alexion- 14. Eisele P, Griebe M, Szabo K, et al. Investigation of lep- ® Chugai, MedImmune, and Shite; is an associate editor for Neurology tomeningeal enhancement in MS: a postcontrast FLAIR Neuroimmunology & Neuroinflammation, an academic editor for PLoS MRI study. Neurology 2015;84:770–775. One; consulted for Sanofi Genzyme, Biogen Idec, MedImmune, Shite, 15. Zekeridou A, Lennon VA. Aquaporin-4 autoimmunity. and Alexion; received research support from Bayer, Novartis, Biogen Idec, Neurol Neuroimmunol Neuroinflamm 2015;2:e110. Teva, Sanofi-Aventis/Genzyme, Alexion and Merck Serono, German Research Council, Werth Stiftung of the City of Cologne, German doi: 10.1212/NXI.0000000000000110. Ministry of Education and Research, Arthur Arnstein Stiftung Berlin, 16. Saji E, Arakawa M, Yanagawa K, et al. Cognitive impair- Arthur Arnstein Foundation Berlin, Guthy-Jackson Charitable Founda- ment and cortical degeneration in neuromyelitis optica. tion, and National Multiple Sclerosis Society of the United States. Ann Neurol 2013;73:65–76.

8 Neurology: Neuroimmunology & Neuroinflammation 17. Asgari N, Berg CT, Morch MT, Khorooshi R, Owens T. 27. Apiwattanakul M, Popescu BF, Matiello M, et al. Intrac- Cerebrospinal fluid aquaporin-4-immunoglobulin G dis- table vomiting as the initial presentation of neuromyelitis rupts blood brain barrier. Ann Clin Trans Neurol 2015;2: optica. Ann Neurol 2010;68:757–761. 857–863. 28. TakahashiT,FujiharaK,NakashimaI,etal.Anti-aquaporin-4 18. Iliff JJ, Lee H, Yu M, et al. Brain-wide pathway for waste antibody is involved in the pathogenesis of NMO: a study on clearance captured by contrast-enhanced MRI. J Clin antibody titre. Brain 2007;130:1235–1243. Invest 2013;123:1299–1309. 29. Jarius S, Franciotta D, Paul F, et al. Cerebrospinal fluid 19. Iliff JJ, Wang M, Liao Y, et al. A paravascular pathway antibodies to aquaporin-4 in neuromyelitis optica and facilitates CSF flow through the brain parenchyma and the related disorders: frequency, origin, and diagnostic rele- clearance of interstitial solutes, including amyloid beta. Sci vance. J Neuroinflammation 2010;7:52. Trans Med 2012;4:147ra111. 30. Sato DK, Callegaro D, de Haidar Jorge FM, et al. Cere- 20. Broadwell RD, Sofroniew MV. Serum proteins bypass the brospinal fluid aquaporin-4 antibody levels in neuromye- blood-brain fluid barriers for extracellular entry to the litis optica attacks. Ann Neurol 2014;76:305–309. central nervous system. Exp Neurol 1993;120:245–263. 31. Takano R, Misu T, Takahashi T, Sato S, Fujihara K, 21. Attwell D, Buchan AM, Charpak S, Lauritzen M, Itoyama Y. Astrocytic damage is far more severe than Macvicar BA, Newman EA. Glial and neuronal control demyelination in NMO: a clinical CSF biomarker study. of brain blood flow. Nature 2010;468:232–243. Neurology 2010;75:208–216. 22. Rangroo Thrane V, Thrane AS, Plog BA, et al. Paravascular 32. Graber DJ, Levy M, Kerr D, Wade WF. Neuromyelitis optica microcirculation facilitates rapid lipid transport and astrocyte pathogenesis and aquaporin 4. J Neuroinflammation 2008;5: signaling in the brain. Scientific Rep 2013;3:2582. 22. 23. Pekcevik Y, Orman G, Lee IH, Mealy MA, Levy M, 33. Sellner J, Hemmer B, Muhlau M. The clinical spectrum Izbudak I. What do we know about brain contrast and immunobiology of parainfectious neuromyelitis optica enhancement patterns in neuromyelitis optica? Clin Imag- (Devic) syndromes. J Autoimmun 2010;34:371–379. ing 2016;40:573–580. 34. Jarius S, Ruprecht K, Wildemann B, et al. Contrasting 24. Tahara M, Ito R, Tanaka K, Tanaka M. Cortical and disease patterns in seropositive and seronegative neuro- leptomeningeal involvement in three cases of neuromyeli- myelitis optica: a multicentre study of 175 patients. J tis optica. Eur J Neurol 2012;19:e47–e48. Neuroinflammation 2012;9:14. 25. Barkhof F, Tas MW, Frequin ST, et al. Limited duration 35. Wingerchuk DM, Hogancamp WF, O’Brien PC, of the effect of methylprednisolone on changes on MRI in Weinshenker BG. The clinical course of neuromyelitis multiple sclerosis. Neuroradiology 1994;36:382–387. optica (Devic’s syndrome). Neurology 1999;53:1107–1114. 26. Asgari N, Skejoe HP, Lennon VA. Evolution of longitu- 36. Jarius S, Wildemann B, Paul F. Neuromyelitis optica: dinally extensive transverse myelitis in an aquaporin-4 clinical features, immunopathogenesis and treatment. Clin IgG-positive patient. Neurology 2013;81:95–96. Exp Immunol 2014;176:149–164.

Neurology: Neuroimmunology & Neuroinflammation 9 Gene variants of adhesion molecules act as modifiers of disease severity in MS

Efthimios Dardiotis, MD ABSTRACT Elena Panayiotou, PhD Objective: To assess the potential effect of variants in genes encoding molecules that are impli- Antonios Provatas, MD cated in leukocyte trafficking into the CNS on the clinical phenotype of multiple sclerosis (MS). Kyproula Christodoulou, Methods: A total of 389 Greek MS cases and 336 controls were recruited in 3 MS centers from PhD Cyprus and Greece. We genotyped 147 tagging single nucleotide polymorphisms (SNPs) in 9 Andreas Hadjisavvas, PhD genes encoding for P-selectin (SELP), integrins (ITGA4, ITGB1,andITGB7), adhesion molecules Athos Antoniades, PhD (ICAM1, VCAM1,andMADCAM1), fibronectin 1 (FN1), and osteopontin (SPP1) involved in lym- Athanasios Lourbopoulos, phocyte adhesion and trafficking into the CNS. Clinical end points of the study were age at MS MD onset and MS severity as measured by the Multiple Sclerosis Severity Score. Permutation testing Marios Pantzaris, MD was applied to all analyses. Nikolaos Grigoriadis, MD Georgios M. Results: SNPs rs6721763 of the ITGA4 and rs6532040 of the SPP1 were found to significantly Hadjigeorgiou, MD* influence disease severity (permutation p values: 3.00e-06 and 0.009884, respectively). SNP Theodoros Kyriakides, rs1250249 of the FN1 had a dose-dependent effect on age at disease onset (permutation p MD* value: 0.0002). Conclusions: This study provides evidence implicating variants encoding adhesion molecules, responsible for lymphocyte adhesion and trafficking within the CNS, as modifiers of MS disease Correspondence to severity. These genetic biomarkers, which can be available at the time of diagnosis, may be used Dr. Kyriakides: to assess the biological aggressiveness of the disease and thus guide decisions on treatment. [email protected] Neurol Neuroimmunol Neuroinflamm 2017;4:e350; doi: 10.1212/NXI.0000000000000350

GLOSSARY DMT 5 disease-modifying therapy; GWAS 5 genome-wide association study; MAF 5 mean allele frequency; PI 5 progression index; SNP 5 single nucleotide polymorphism.

Multiple sclerosis (MS) is considered to result from an interplay between environmental expo- sures and genetic background.1–3 The majority of genetic studies have focused on susceptibility variants, and this effort has yielded more than 110 genetic risk factors of MS.4 In addition, there is evidence that genetic factors may influence disease phenotype such as age at onset, disease severity, and the clinical course.5–7 However, 3 genome-wide association studies (GWASs) investigating MS severity have failed to reach a genome-wide significance threshold.8–10 A recent meta-analysis of 10 cohorts in 7,125 patients with MS also failed to detect any significant signal.11 Disability in MS, either resulting from relapses or accumulating during the progressive phase of the disease, is partly a result of immune attack from adaptive and innate immunity. A number of molecules are important in leukocyte adhesion and diapedesis through the blood-brain barrier, as well as their migration through the extracellular matrix to the target tissue. The use of natalizumab, a monoclonal antibody against a4 integrin, was found to markedly reduce relapses, Supplemental data at Neurology.org/nn *These authors contributed to this manuscript as senior authors. From The Cyprus Institute of Neurology and Genetics (E.D., E.P., K.C., A.H., M.P., T.K.), Nicosia, Cyprus; Department of Neurology (E.D., A.P., G.M.H.), Laboratory of Neurogenetics, University of Thessaly, University Hospital of Larissa; Head of Research and Developments (A.A.), Stremble Ventures LTD, Limassol, Cyprus; and 2nd Department of Neurology (A.L., N.G.), AHEPA University Hospital, Aristotle University of Thessaloniki, Greece. Funding information and disclosures are provided at the end of the editorial. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 new lesion formation, and disability progres- FN1, 2q34), and (9) osteopontin (OMIM #166490, SPP1, sion12,13 by blocking lymphocyte adhesion and 4q21-q25). To cover most of the polymorphic variability along with the migration through the vascular endothelial above-mentioned genes, we used the linkage disequilibrium tag- cells and by blocking interaction with extracel- ger algorithm of the HapMap project for selection of tag-SNPs. lular matrix proteins. The impressive effective- Selection of tagging single nucleotide polymorphisms (SNPs) 1 ness of natalizumab argues strongly in favor of was based on the HapMap Release 27 database (Phase II III, Feb09, on NCBI B36 assembly, dbSNP b126) for CEU popu- the significance of this pathogenic pathway. lation, using the criteria of pairwise r2 values $0.8 and minor This study was designed to test the effect of allele frequency .0.05. This approach retrieved 32 tagging SNPs genetic variants of molecules, that are impli- in SELP gene, 36 in ITGA4 gene, 16 in ITGB1,3inITGB7,5in ICAM1,15inVCAM1,3inMADCAM1,38inFN1, and 8 in cated in leukocyte trafficking through blood SPP1 which sum to 156 SNPs. Chromosome position of each vessel walls into the CNS, on disease severity SNP, the minor allele frequency for CEU population and TSI as measured by the Multiple Sclerosis Severity population (Tuscan in Italy), which is considered closer to the Score (MSSS). Greek population, and the minor allele frequency (MAF) of our control group are presented in table e-1 at Neurology.org/nn. TaqMan 59-nuclease assays were used to genotype SNPs by METHODS Study population. Greek patients with MS and acquiring specific assays for each SNP and then performing allelic controls were recruited from 3 MS centers: The Cyprus Institute discrimination assays on the ABI 7900HT Real-Time System of Neurology and Genetics in Cyprus, the University Hospital of (Applied Biosystems, Foster City, CA). Genotype success rates Larissa, Greece, and the AHEPA Hospital of Aristotle University were above 96%. However, 9 SNPs as shown in table e-1 failed to in Thessaloniki, Greece.14,15 In total, 389 MS cases and 336 be genotyped and therefore were not included in the analysis. controls were recruited for the study. Patients were included in the study if they had clinically definite MS according to the 2005 Statistical analysis. Quality control was performed by evaluat- revised McDonald criteria, were aged $18 years, and had disease ing the genotype distributions of all SNPs in the control samples duration of $5 years. Cases with primary progressive MS were for the Hardy-Weinberg equilibrium using an exact test17 with excluded. a threshold of p # 0.05. MAFs were estimated and used for Clinical and demographic data were collected using the same quality control (threshold of MAF ,0.05%). Linear regression criteria in each center. Neurologists specialized in MS collected analysis for the age at disease onset and MSSS were performed for data such as sex, current age, and age at disease onset, as age at association testing. The cutoff threshold for significance was the first neurologic symptom suggestive of MS. Use of disease- a priori set to a permutation-derived p value of at most 0.01. modifying therapies (DMTs) and treatment duration were also Analyses for age at onset were adjusted for sex and site. Analyses recorded. Patients were receiving one of the following DMTs: with the MSSS, as a dependent variable, were additionally interferon-b, glatiramer acetate, or natalizumab. Disability status adjusted for age at onset and treatment status. For treatment was assessed in the absence of relapse for at least 6 months, using status, the duration of treatment with DMT in months was used the Kurtzke Expanded Disability Status Scale (EDSS). Disease as a covariate to ensure that any detected effect was not driven by duration was defined as the time (in years) from the first symptom treatment duration. Permutation testing was applied to adjust for to the last EDSS assessment. The rate of disease progression was multiple testing for known or unknown factors that may affect expressed by the progression index, which is the ratio of EDSS to the balance of the distribution of the data. A total of 106 per- MS duration. Finally, disease severity in each individual was as- mutations were performed by using a high-performance com- sessed by means of MSSS which is a useful algorithm for disability puting platform. distribution in MS groups evaluated cross-sectionally.16 The con- Label-swapping permutations were used with 2 sets of empir- trol group consisted of healthy volunteers from the same ethic ical significance p values to estimate individual SNPs significance. regions and was matched with cases for age and sex. The minimum possible p value depends on the number of tests performed; in our case, it was 1026. Because the permutation Standard protocol approvals, registrations, and patient scheme followed preserves the correlational structure between consents. The study protocol was approved by the local ethics SNPs, there is no assumption of independence unlike Bonferroni, committees, and written informed consent was obtained from which assumes all tests are independent. As this is a candidate all patients. gene study with relatively dense coverage, there is a likelihood of dependence between neighboring SNPs due to linkage disequi- Isolation of DNA, SNP selection, and genotyping. DNA librium; therefore, permutation offers a better alternative for ad- was isolated from peripheral blood samples using the Puregene justing for multiple testing. DNA purification kit (Qiagen, Valencia, CA). Based on the study rationale, we chose to genotype genes that encode for the following adhesion and extracellular matrix molecules: (1) P-selectin RESULTS Clinical and demographic data of controls (OMIM #173610) encoded by the SELP gene located on chro- and patients as a group and in each MS center are pre- mosome 1q23-q25, (2) integrin a-4 (OMIM #1929750, ITGA4, sented in table 1. More than two-thirds of the pa- 2q31-q32), (3) integrin-b–1 (OMIM #135630; ITGB1, tients were women. The mean disease duration was b– 10p11.2), (4) integrin- 7 (OMIM #147559, ITGB7, 12.9 years, and most patients (70%) were under 12q13.13), (5) intercellular adhesion molecule 1 (OMIM DMT. The mean EDSS was 3.4, whereas half of the #147840, ICAM1, 19p13.3-p13.2), (6) vascular cell adhesion molecule 1 (OMIM #192225, VCAM1, 1p32-p31), (7) mucosal patients had EDSS less than 3. Severe accumulation vascular addressin cell adhesion molecule 1 (OMIM #102670, of disability (MSSS: 7–10) was observed in approxi- MADCAM1, 19p13.3), (8) fibronectin 1 (OMIM #135600, mately 20% of the patients. Patients recruited from

2 Neurology: Neuroimmunology & Neuroinflammation Table 1 Demographic and clinical characteristics of study participants

Total patients Controls Cyprus patients Cyprus controls Larissa patients Larissa controls Thessaloniki patients

No. 389 336 104 127 206 209 79

Female, n (%) 251 (64.5) 234 (69.6) 74 (71.2) 88 (69.3) 126 (61.2) 146 (69.9) 51 (64.6)

Male, n (%) 138 (35.5) 102 (30.4) 30 (28.8) 39 (30.7) 80 (38.8) 63 (30.1) 28 (35.4)

Female:male ratio 1.8:1 2.3:1 2.5:1 2.3:1 1.6:1 2.3:1 1.8:1

Age at disease onset, y, 30.4 (18–63) 35.3 (18–58) 29.0 (18–63) 27.9 (18–50) mean (range)

Age at the time of analysis, y, 43.3 (23–74) 36.3 (23–78) 43.8 (25–66) 47.4 (20–78) 43.0 (23–74) 29.6 (23–58) 41.9 (23–63) mean (range)

Disease duration, y, mean 12.9 12.44 12.77 13.90

Median (range) 11 (5–52) 11 (5–35) 10 (5–52) 12 (5–38)

DMT, n (%) 271 (69.7) 71 (68.3) 131 (63.6) 69 (87.3)

DMT time, mo, mean (SD) 72.1 (44.2) 73.6 (49.6) 70.4 (43.3) 73.9 (40.2)

EDSS, n (%)

<3 188 (47.8) 69 (66.3) 84 (40.8) 33 (41.8)

3to<6 114 (29.3) 27 (26.0) 54 (26.2) 33 (41.8)

6–6.5 40 (10.3) 6 (5.8) 25 (21.1) 9 (11.4)

‡7 35 (9.0) 2 (1.9) 31 (15.0) 2 (2.5)

Mean (SD) 3.40 (2.16) 2.44 (1.54) 3.83 (2.46) 3.61 (1.59)

Progression index, mean (SD) 0.31 (0.24) 0.22 (0.15) 0.34 (0.24) 0.35 (0.31)

MSSS

Median 4.00 3.00 5.00 5.00

Mean (SD) 4.39 (2.61) 3.18 (1.98) 4.92 (2.89) 4.71 (2.00)

Benign MS (MSSS 5 [1–2]), n (%) 109 (28.0) 49 (47.1) 51 (24.8) 9 (11.4)

Severe MS (MSSS 5 [7–10]), n (%) 86 (22.1) 9 (8.7) 61 (29.6) 16 (20.3)

Abbreviations: DMT 5 disease-modifying therapy; EDSS 5 Kurtzke Expanded Disability Status Scale; MSSS 5 Multiple Sclerosis Severity Score.

the Larissa and Thessaloniki sites presented a more increase from a mean MSSS of 3.538 to 4.034 and severe phenotype regarding EDSS, progression index, 4.176 in AA, Ab, and bb genotypes, respectively. and MSSS scores compared with the patients re- Permutation analyses of the effects of SNP alleles cruited from Cypriot (p , 0.05). on age at onset are shown in table e-3. The permuta- The distribution of SNP alleles in patients, their tion p value significance threshold of 0.01 was ex- respective scores in MSSS, and the results of linear ceeded by rs1250249 of FN1 (p 5 0.0002). A regression analysis are presented in table e-2. Signifi- dose-dependent effect of this polymorphism on age cant associations with p , 0.05 were found with 4 at onset was observed: age at onset dropped from 32.3 ITGA4 polymorphisms (rs10930969, rs2124440, years in AA homozygotes to 28.1 years in Ab hetero- rs3770136, and rs6721763), in rs3780873 ITGB1, zygotes and to 26.3 years in bb homozygotes. in 2 FN1 polymorphisms (rs1250258 and rs13432972), and in rs9138 of SPP1. Permutation DISCUSSION Despite the considerable progress in analysis was performed to adjust for multiple testing identifying genetic susceptibility variants,4 there is (table 2). Significant associations were found with limited evidence in identifying genetic factors that rs6721763 of ITGA4 (permutation p value: 3.00e- may contribute to MS disease severity. Genetic asso- 06) and rs6532040 of SPP1 (permutation p value: ciation studies have focused on candidate genes such 0.009884). A dose-dependent effect of minor allele as APOE, genes related to immune function, autoim- on MSSS was noted with worse scores in minor allele munity, neurobiology, myelin structure, and the homozygotes for both SNPs. In particular, for human leukocyte antigen (HLA), which resulted in rs6721763 of ITGA4, the mean MSSS increased from conflicting data.18,19 In addition, 3 GWAS investigat- 3.191 in AA homozygotes to 3.94 in Ab heterozy- ing MS severity failed to surpass a genome-wide sig- gotes and finally 5.012 in bb homozygotes. The same nificance threshold.8–10 However, analysis of SNPs in trend was observed for rs6532040 of SPP1 with an genes involved in glycosylation and calcium and

Neurology: Neuroimmunology & Neuroinflammation 3 4

Table 2 Permutation analysis results in MSSS

Permutation Permutation Permutation Permutation Permutation Permutation SNPs p value SNPs p value SNPs p value SNPs p value SNPs p value SNPs p value

SELP 27. rs6128 0.1919 23. rs4667310 0.1279 1. rs3825084 0.6863 2. rs6510810 0.2457 25. rs13652 0.6824

1. rs732314 0.9147 28. rs3917843 0.3343 24. rs7607758 0.1746 2. rs2272300 0.2438 3. rs3745925 0.2051 26. rs13432972 0.06204

2. rs3917862 0.6987 29. rs3917853 0.1037 25. rs11689738 0.6311 3. rs11170465 0.5258 FN1 27. rs1250203 0.7171

3. rs3917683 0.9941 30. rs3917854 0.1719 26. rs12469449 0.5248 ICAM1 1. rs1250258 0.0634 28. rs7608342 0.6461 erlg:Nuomuooy&Neuroinflammation & Neuroimmunology Neurology: 4. rs3917686 0.5239 ITGA4 27. rs6721763 3.00e-06a 1. rs5030390 0.6959 2. rs2577290 0.7659 29. rs6707530 0.348

5. rs2244529 0.4432 1. rs12988934 0.3794 28. rs6728886 0.4004 2. rs1799969 0.6981 3. rs1250246 0.209 30. rs1250211 0.9535

6. rs2244526 0.9039 2. rs3770138 0.7438 29. rs6707704 0.2942 3. rs5498 0.5103 4. rs1898536 0.9657 31. rs2289200 0.8463

7. rs3917709 0.6725 3. rs10930969 0.01169 30. rs1449259 0.3535 4. rs3093032 0.2641 5. rs1250250 0.885 32. rs7568287 0.6808

8. rs2236868 0.6458 4. rs2124440 0.01847 31. rs1047307 0.2376 5. rs281437 0.1168 6. rs11693652 0.9635 33. rs33996776 0.7985

9. rs3917714 0.7683 5. rs3770136 0.01521 ITGB1 VCAM1 7. rs16854041 0.5775 34. rs10498037 0.7494

10. rs6125 0.7683 6. rs155149 0.3515 1. rs11009157 0.2657 1. rs3176860 0.3093 8. rs1250249 0.5565 35. rs7579867 0.8239

11. rs3917727 0.2448 7. rs3770132 0.3451 2. rs1187070 0.5029 2. rs3176861 0.3179 9. rs1250248 0.1671 36. rs11651 0.8754

12. rs6131 0.3165 8. rs711801 0.8058 3. rs1009002 0.696 3. rs3917009 0.7551 10. rs17517509 0.9489 SPP1

13. rs3917731 0.4606 9. rs1449264 0.2445 4. rs2488336 0.3562 4. rs3917010 0.2939 11. rs1437799 0.9584 1. rs2853749 0.7142

14. rs3917739 0.8051 10. rs1449265 0.8729 5. rs10827164 0.7011 5. rs3765685 0.2097 12. rs724617 0.9754 2. rs11728697 0.1396

15. rs3917740 0.8371 11. rs3770126 0.06531 6. rs16933819 0.2905 6. rs3176867 0.2108 13. rs13021679 0.6185 3. rs6840362 0.4792

16. rs2076074 0.9684 12. rs11694175 0.9604 7. rs9417094 0.2937 7. rs3917012 0.1209 14. rs2372545 0.4479 4. rs6532040 0.009884a

17. rs3917744 0.6447 13. rs17224277 0.3591 8. rs2504001 0.6565 8. rs3176869 0.539 15. rs7575234 0.6593 5. rs6811536 0.4773

18. rs2142760 0.4416 14. rs155087 0.04326 9. rs2488331 0.4218 9. rs3181088 0.1947 16. rs7572169 0.6593 6. rs4754 0.6128

19. rs3917768 0.6865 15. rs2305588 0.1206 10. rs10763902 0.7189 10. rs2209627 0.288 17. rs10172425 0.6431 7. rs1126772 0.6849

20. rs2205895 0.6823 16. rs155099 0.07616 11. rs3780873 0.1267 11. rs3176876 0.5796 18. rs34255697 0.5046 8. rs9138 0.03449

21. rs3917779 0.9829 17. rs3755021 0.1005 12. rs4587680 0.3719 12. rs3917016 0.533 19. rs1525351 0.7776

22. rs2420378 0.9013 18. rs155103 0.3347 13. rs2488320 0.8601 13. rs3917018 0.3098 20. rs2692228 0.917

23. rs3917786 0.7186 19. rs155106 0.9674 14. rs1316757 0.853 14. rs3917019 0.4335 21. rs11883812 0.5302

24. rs3917788 0.5852 20. rs3770120 0.1899 15. rs2488318 0.8561 15. rs3176878 0.857 22. rs10804242 0.9766

25. rs3917792 0.6765 21. rs17225354 0.6191 16. rs2153875 0.8525 MADCAM1 23. rs6709607 0.9355

26. rs1569471 0.743 22. rs2305586 0.1351 INTEGRIN, b-7 1. rs12982646 0.5785 24. rs7588661 0.7678

Abbreviations: MSSS 5 Multiple Sclerosis Severity Score; SNP 5 single nucleotide polymorphism. a Significant. glutamate signaling that may represent disease a modest trend for association. In particular, patients modifiers reached levels close to genome-wide sig- with the wild-type genotype were less likely to have nificance threshold. The effect of HLA on disease a mild course (p 5 0.08) and had increased risk of onset was relatively consistent among studies.20,21 having a secondary-progressive course of the disease Carriers of the DRB1*15:01 allele had an earlier (p 5 0.05).33 In addition, a haplotype study of 4 onset of MS, with each copy accounting for a decrease SNPs (rs4754, rs1126616, rs1126772, and rs9138; in age at onset by about 10.6 months.22 0.49 kb, 1.65 kb, 1.98 kb, and 2.14 kb away from This study detected a statistically significant asso- our polymorphism, respectively) showed that homo- ciation between rs6721763 of ITGA4 gene and zygotes combining the wild-type alleles of all 4 SNPs disease severity as measured with the MSSS in displayed a milder disease course, slower accumula- a dose-dependent manner. ITGA4 encodes for the tion of disability, and delayed conversion from a4 subunit of a4b1 integrin, which is involved in relapsing-remitting to secondary-progressive disease leukocyte adhesion and migration through the blood- course.31 The same research group replicated these brain barrier.23 The monoclonal antibody, natalizu- findings in an independent group and also detected mab, which is approved for the treatment of similar effects on a polymorphism located in the pro- relapsing-remitting MS, exerts its favorable effects moter region of the osteopontin gene (rs7687316, on relapses and disability progression by targeting 5.5 kb distant from our polymorphism).34 However, a4 integrin.24 Polymorphisms in ITGA4 have been a study of 8 SNPs located in exons 6 and 7 failed to linked so far only to MS susceptibility.25–27 In partic- detect any significant association with MS severity.35 ular rs1143676, a nonsynonymous polymorphism Finally, our study found that rs1250249 of FN1 (R878Q) located in exon 24 of ITGA4 has been re- exerts a dose-dependent effect on disease onset. The ported to be associated with disease susceptibility in presence of the minor allele predisposes to an earlier Slovaks25 but not in other populations.26,27 This poly- age at onset. Fibronectin 1 is an extracellular matrix morphism is located 9.3 kb distant from our poly- glycoprotein that interacts with a number of integrins morphism. The second ITGA4 polymorphism that mediating a variety of functions including cytoskele- was linked so far to disease susceptibility in a Basque ton organization, cell adhesion, and migration.36 group is the intronic SNP rs1449263,26 which how- Polymorphisms of FN1 have been associated so far ever is 66.5 kb away from our polymorphism. Finally, among others with schizophrenia37 and the develop- rs155100 (36 kb far from rs6721763) has been re- ment of intracranial aneurysms.38 ported to influence the risk of autism.28 SNP The difficulty in identifying genetic modifiers of rs6721763 is an intronic polymorphism (intron 17– MS course and severity may reflect either the com- 18) located only just 929 bp from the splicing site of plexity or diversity of inflammatory and neurodegen- exon 18 of ITGA4. erative mechanisms that contribute to the clinical The second genetic modifier of disease course expression or some inherent flaws in measuring rele- identified in our study is rs6532040 of SPP1. The vant clinical endpoints. In addition, the advent of presence of the minor allele worsens the disease dis- immunomodulatory treatments has proved to alter ability course. Osteopontin is a secreted matrix pro- the course of the disease.39,40 Early initiation and timely tein with pleiotropic, but mainly, proinflammatory escalation of disease-modifying drugs are known to functions involved in T-cell activation, proliferation, prevent the accumulation of disability. Therefore, the and migration. A part of its actions is mediated by inclusion in a genetic association study of patients with binding to integrins. In addition, intracellular osteo- MS without any treatment along with those who may pontin activates intracellular signaling pathways.29 or may not receive the best medical treatment is an Osteopontin transcripts are upregulated in MS brain important confounding factor that may obscure the lesions and in normal-appearing white matter. In the detection of significant genetic effects. mouse model of experimental autoimmune encepha- Our study included a clinically well-characterized lomyelitis, osteopontin was implicated in disease and ethnically homogenous group. Disease severity severity and activity.30 Genetic polymorphisms in each individual was assessed by means of MSSS, altered osteopontin levels.31 A genetic association a robust and validated measure, that shows stability study showed that a G/A polymorphism located in over time. In addition, patients were only included the 39 untranslated region of exon 7 (position if they had at least 5 years duration of the disease, 9583) had a significant effect on the age at onset.32 as it is known that the first 5 years of the disease rep- SNP rs6532040, which emerged as significant in our resent a significant outcome milestone of MS. Also, study, is located in intron 5–6, which is only 1.5 kb a number of meaningful clinical predictors that may distant from the start of exon 7. In another study, have an effect on long-term severity were collected rs9138 of the osteopontin gene, located at exon 7 and and included as covariates in the analysis. Moreover, 2.14 kb distant from our polymorphism, showed genetic variability across the tested genes was covered

Neurology: Neuroimmunology & Neuroinflammation 5 by a large number of tagging SNPs. Finally, permuta- 5. DeLuca GC, Ramagopalan SV, Herrera BM, et al. An ex- tion analysis, a powerful unbiased test, was imple- tremes of outcome strategy provides evidence that multiple mented to adjust for multiple testing of known and sclerosis severity is determined by alleles at the HLA-DRB1 locus. Proc Natl Acad Sci USA 2007;104:20896–20901. unknown factors. 6. Gray OM, Abdeen H, McDonnell GV, Patterson CC, In brief, our study provides evidence implicating Graham CA, Hawkins SA. An investigation of susceptibil- variants in genes that encode adhesion molecules, ity loci in benign, aggressive and primary progressive mul- responsible for lymphocyte trafficking and activation, tiple sclerosis in Northern Irish population. Mult Scler in the severity of MS. Replication of these associa- 2009;15:299–303. tions in other independent populations will render 7. Kroner A, Mehling M, Hemmer B, et al. A PD-1 polymorphism is associated with disease progression in multi- our findings more robust. Such biomarkers, if vali- ple sclerosis. Ann Neurol 2005;58:50–57. dated, may become valuable prognostic tools at an 8. Baranzini SE, Wang J, Gibson RA, et al. Genome-wide individual level at the time of diagnosis. Furthermore, association analysis of susceptibility and clinical phenotype therapies targeting adhesion molecules may be worth in multiple sclerosis. Hum Mol Genet 2009;18:767–778. exploring in the future. 9. Brynedal B, Wojcik J, Esposito F, et al. MGAT5 alters the severity of multiple sclerosis. J Neuroimmunol 2010;220: AUTHOR CONTRIBUTIONS 120–124. Efthimios Dardiotis, Georgios M. Hadjigeorgiou, and Theodoros Kyria- 10. International Multiple Sclerosis Genetics Consortium. kides: original idea, study design, clinical management and cohort recruit- Genome-wide association study of severity in multiple ment, interpretation of results, and drafting of the manuscript. Elena sclerosis. Genes Immun 2011;12:615–625. Panayiotou, Kyproula Christodoulou, and Andreas Hadjisavvas: 11. George MF, Briggs FB, Shao X, et al. Multiple sclerosis experimental work. Antonios Provatas, Athanasios Lourbopoulos, Marios risk loci and disease severity in 7,125 individuals from 10 Pantzaris, and Nikolaos Grigoriadis: clinical management and cohort studies. Neurol Genet 2016;2:e87. doi: 10.1212/NXG. recruitment. Athos Antoniades: statistical analysis. 0000000000000087. 12. Polman CH, O’Connor PW, Havrdova E, et al. A ran- STUDY FUNDING domized, placebo-controlled trial of natalizumab for The study was supported by the Project YGEIA/BIOS/0609 (BIE/01), relapsing multiple sclerosis. N Engl J Med 2006;354: which is co-financed by the European Regional Development Fund 899–910. and the Republic of Cyprus through the Research Promotion 13. Tsivgoulis G, Katsanos AH, Mavridis D, et al. The efficacy Foundation. of natalizumab versus fingolimod for patients with relapsing-remitting multiple sclerosis: a systematic review, DISCLOSURE indirect evidence from randomized placebo-controlled tri- E. Dardiotis, E. Panayiotou, and A. Provatas report no disclosures. als and meta-analysis of observational head-to-head trials. K. Christodoulou received research support from Research Promotion Foundation (Cyprus), Telethon Grants (Cyprus), and Norwegian Finan- PLoS One 2016;11:e0163296. cial Mechanism, EU. A. Hadjisavvas reports no disclosures. A. Anto- 14. Aggelakis K, Zacharaki F, Dardiotis E, et al. Interleukin- niades is the CEO and Head of Research and Development of Stemble 1B and interleukin-1 receptor antagonist gene polymor- Ventures Ltd. A. Lourbopoulos reports no disclosures. M. Pantzaris phisms in Greek multiple sclerosis (MS) patients with received travel funding and/or speaker honoraria from Sanofi-Genzyme bout-onset MS. Neurol Sci 2010;31:253–257. Greece, Genesis Pharma Greece. N. Grigoriadis served on the scientific 15. Hadjigeorgiou G, Dardiotis E, Tsivgoulis G, et al. Correc- advisory board for Biogen, Nayer, Novartis, Celgene, Genesis Pharma, tion: observational study assessing demographic, economic Merck Serono, Sanofi-Aventis, Paradig MS, Navigating MS, and Helani and clinical factors associated with access and utilization of and has consulted for Biogen, Bayer, Novartis, Celgene, Genesis health care services of patients with multiple sclerosis Pharma, Merck Serono, and Sanofi-Aventis. G.M. Hadjigeorgiou served under treatment with interferon beta-1b (EXTAVIA). on the scientific advisory board for Genesis Pharma, Genzyme, Merck Serono, Bayer Hellas, and TEVA and received speaker honoraria from PLoS One 2015;10:e0140308. Genesis Pharma, Genzyme, Merck Serono, Bayer Hellas, and TEVA. 16. Roxburgh RH, Seaman SR, Masterman T, et al. Multiple T. Kyriakides reports no disclosures. Go to Neurology.org/nn for full Sclerosis Severity Score: using disability and disease duration disclosure forms. to rate disease severity. Neurology 2005;64:1144–1151. 17. Wigginton JE, Cutler DJ, Abecasis GR. A note on exact Received December 13, 2016. Accepted in final form March 1, 2017. tests of Hardy-Weinberg equilibrium. Am J Hum Genet 2005;76:887–893. REFERENCES 18. Ramagopalan SV, Deluca GC, Degenhardt A, Ebers GC. 1. Compston A, Sadovnick AD. Epidemiology and genetics The genetics of clinical outcome in multiple sclerosis. of multiple sclerosis. Curr Opin Neurol Neurosurg 1992; J Neuroimmunol 2008;201–202:183–199. 5:175–181. 19. Ramagopalan SV, Deluca GC, Morrison KM, et al. Anal- 2. Sokratous M, Dardiotis E, Tsouris Z, et al. Deciphering ysis of 45 candidate genes for disease modifying activity in the role of DNA methylation in multiple sclerosis: emerg- multiple sclerosis. J Neurol 2008;255:1215–1219. ing issues. Auto Immun Highlights 2016;7:12. 20. Hensiek AE, Sawcer SJ, Feakes R, et al. HLA-DR 15 is 3. Efthymiou G, Dardiotis E, Liaskos C, et al. Anti-hsp60 associated with female sex and younger age at diagnosis in antibody responses based on Helicobacter pylori in multiple sclerosis. J Neurol Neurosurg Psychiatry 2002; patients with multiple sclerosis: (ir)relevance to disease 72:184–187. pathogenesis. J Neuroimmunol 2016;298:19–23. 21. MastermanT,LigersA,OlssonT,AnderssonM,OlerupO, 4. Sawcer S, Franklin RJ, Ban M. Multiple sclerosis genetics. Hillert J. HLA-DR15 is associated with lower age at onset in Lancet Neurol 2014;13:700–709. multiple sclerosis. Ann Neurol 2000;48:211–219.

6 Neurology: Neuroimmunology & Neuroinflammation 22. International Multiple Sclerosis Genetics Consortium, multiple sclerosis in Japanese patients. J Neuroimmunol Wellcome Trust Case Control Consortium 2, Sawcer S, 2003;136:125–129. et al. Genetic risk and a primary role for cell-mediated 33. Caillier S, Barcellos LF, Baranzini SE, et al. Osteopontin immune mechanisms in multiple sclerosis. Nature 2011; polymorphisms and disease course in multiple sclerosis. 476:214–219. Genes Immun 2003;4:312–315. 23. Ley K, Laudanna C, Cybulsky MI, Nourshargh S. Getting 34. Comi C, Cappellano G, Chiocchetti A, et al. The impact to the site of inflammation: the leukocyte adhesion cascade of osteopontin gene variations on multiple sclerosis devel- updated. Nat Rev Immunol 2007;7:678–689. opment and progression. Clin Dev Immunol 2012;2012: 24. Rudick RA, Sandrock A. Natalizumab: alpha 4-integrin 212893. antagonist selective adhesion molecule inhibitors for MS. 35. Hensiek AE, Roxburgh R, Meranian M, et al. Osteopon- Expert Rev Neurother 2004;4:571–580. tin gene and clinical severity of multiple sclerosis. J Neurol 25. Durmanova V, Shawkatova I, Javor J, et al. VLA4 gene 2003;250:943–947. polymorphism and susceptibility to multiple sclerosis in 36. Goossens K, Van Soom A, Van Zeveren A, Favoreel H, Slovaks. Folia Biol (Praha) 2015;61:8–13. Peelman LJ. Quantification of fibronectin 1 (FN1) splice 26. O’Doherty C, Roos IM, Antiguedad A, Aransay AM, variants, including two novel ones, and analysis of integ- Hillert J, Vandenbroeck K. ITGA4 polymorphisms and rins as candidate FN1 receptors in bovine preimplantation susceptibility to multiple sclerosis. J Neuroimmunol embryos. BMC Dev Biol 2009;9:1. 2007;189:151–157. 37. Nakata K, Ujike H, Sakai A, et al. Association study 27. Andreoli V, Cittadella R, Valentino P, et al. The role of between the fibronectin gene and schizophrenia. Am J VLA4 polymorphisms in multiple sclerosis: an association Med Genet B Neuropsychiatr Genet 2003;116B:41–44. study. J Neuroimmunol 2007;189:125–128. 38. Ruigrok YM, Rinkel GJ, van’t Slot R, Wolfs M, Tang S, 28. Correia C, Coutinho AM, Almeida J, et al. Association of Wijmenga C. Evidence in favor of the contribution of the alpha4 integrin subunit gene (ITGA4) with autism. genes involved in the maintenance of the extracellular Am J Med Genet B Neuropsychiatr Genet 2009;150B: matrix of the arterial wall to the development of intracra- 1147–1151. nial aneurysms. Hum Mol Genet 2006;15:3361–3368. 29. Rittling SR, Singh R. Osteopontin in immune-mediated 39. Tsivgoulis G, Katsanos AH, Grigoriadis N, et al. The diseases. J Dent Res 2015;94:1638–1645. effect of disease modifying therapies on disease progression 30. Chabas D, Baranzini SE, Mitchell D, et al. The influence of in patients with relapsing-remitting multiple sclerosis: a sys- the proinflammatory cytokine, osteopontin, on autoim- tematic review and meta-analysis. PLoS One 2015;10: mune demyelinating disease. Science 2001;294:1731–1735. e0144538. 31. Chiocchetti A, Comi C, Indelicato M, et al. Osteopontin gene 40. Tsivgoulis G, Katsanos AH, Grigoriadis N, et al. The haplotypes correlate with multiple sclerosis development and effect of disease-modifying therapies on brain atrophy in progression. J Neuroimmunol 2005;163:172–178. patients with clinically isolated syndrome: a systematic 32. Niino M, Kikuchi S, Fukazawa T, Yabe I, Tashiro K. review and meta-analysis. Ther Adv Neurol Disord Genetic polymorphisms of osteopontin in association with 2015;8:193–202.

Neurology: Neuroimmunology & Neuroinflammation 7 Abnormal brain metabolism on FDG-PET/CT is a common early finding in autoimmune encephalitis

John C. Probasco, MD ABSTRACT Lilja Solnes, MD Objective: To compare the rate of abnormal brain metabolism by FDG-PET/CT to other paraclin- Abhinav Nalluri, BS ical findings and to describe brain metabolism patterns in autoimmune encephalitis (AE). Jesse Cohen, BA Methods: A retrospective review of clinical data and initial dedicated brain FDG-PET/CT studies Krystyna M. Jones, MD for neurology inpatients with AE, per consensus criteria, treated at a single tertiary center over Elcin Zan, MD 123 months. Z-score maps of FDG-PET/CT were made using 3-dimensional stereotactic surface Mehrbod S. Javadi, MD projections with comparison to age group–matched controls. Brain region mean Z-scores with Arun Venkatesan, MD, magnitudes $2.00 were interpreted as significant. Comparisons were made to rates of abnormal PhD initial brain MRI, abnormal initial EEG, and presence of intrathecal inflammation. Results: Sixty-one patients with AE (32 seropositive) underwent brain FDG-PET/CT at median 4 Correspondence to weeks of symptoms (interquartile range [IQR] 9 weeks) and median 4 days from MRI (IQR 8.5 Dr. Probasco: days). FDG-PET/CT was abnormal in 52 (85%) patients, with 42 (69%) demonstrating only hypo- [email protected] metabolism. Isolated hypermetabolism was demonstrated in 2 (3%) patients. Both hypermetabolic and hypometabolic brain regions were noted in 8 (13%) patients. Nine (15%) patients had normal FDG-PET/CT studies. CSF inflammation was evident in 34/55 (62%) patients, whereas initial EEG (17/56, 30%) and MRI (23/57, 40%) were abnormal in fewer. Detection of 2 or more of these paraclinical findings was in weak agreement with abnormal brain FDG- PET/CT (k 5 0.16, p 5 0.02). Conclusions: FDG-PET/CT was more often abnormal than initial EEG, MRI, and CSF studies in neurology inpatients with AE, with brain region hypometabolism the most frequently observed. Neurol Neuroimmunol Neuroinflamm 2017;4:e352; doi: 10.1212/NXI.0000000000000352

GLOSSARY AE 5 autoimmune encephalitis; FDG-PET 5 18F-fluorodeoxy-glucose PET; FLAIR 5 fluid-attenuated inversion recovery; IQR 5 interquartile range; NMDAR 5 NMDA receptor; VGKCc 5 voltage-gated potassium channel-complex.

As early immunotherapy seems to contribute to improved outcomes in autoimmune encephalitis (AE), recent criteria have been proposed to facilitate early diagnosis.1 18F-fluorodeoxyglucose PET (FDG-PET) is only included in criteria for definite autoimmune limbic encephalitis.1 However, FDG-PET has been recognized as a potentially useful biomarker in suspected AE.2–7 In autoimmune limbic encephalitis, hypermetabolism on FDG-PET in otherwise normal mesial temporal lobe structures by MRI suggests that FDG-PET may be more sensitive than MRI.2–4 Also, particular patterns of metabolism noted by FDG-PET have been identified in certain AE syndromes.8–11 The majority of prior studies of FDG-PET in AE have been limited to qualitative description of FDG-PET findings,2,3,5,6,10,12–16 used nondedicated brain FDG-PET studies,8 have been restricted to specific syndromes,9,11,17,18 or have made limited comparisons to other diagnostic results incorporated in the current clinical criteria.3–6 Supplemental data at Neurology.org/nn From the Department of Neurology (J.C.P., A.N., J.C., A.V.), Johns Hopkins Encephalitis Center, Department of Neurology (J.C.P.), Johns Hopkins Center for Refractory Status Epilepticus and Neuroinflammation, and Russell H. Morgan Department of Radiology and Radiological Sciences (L.S., K.M.J., E.Z., M.S.J.), Johns Hopkins University School of Medicine, Baltimore, MD. Funding information and disclosures are provided at the end of the editorial. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 We sought to semiquantitatively describe with regions demonstrating an average Z-score magnitude greater dedicated brain FDG-PET/CT findings for than 2.00 (i.e., greater than 2 SDs abnormal relative to the CortexID database) was recorded as quantitatively abnormal. neurology inpatients who met recent AE con- Using these methods, by chance a healthy control would have sensus criteria relative to a database of healthy a 26% probability of an abnormal study (i.e., hyper- or hypo- controls, with comparisons between seronega- metabolism in at least 1 of the 6 brain regions evaluated). Brain tive and seropositive patients with AE. We also FDG-PET/CT figures were generated using CortexID or the GE Advanced Workstation software package (GE Healthcare). sought to describe the rate of abnormal patterns of brain region metabolism relative to Review of brain MRI. Blinded review of brain MRI was performed by 2 fellowship trained neuroradiologists (L.S. and E.Z.). other paraclinical findings in these AE cases Clinical MRIs were performed as per the institutional protocol at as well as prior case series. either 1.5- or 3-T on a Philips (Best, Netherlands), GE Healthcare or Siemens (Erlangen, Germany) scanner. For pur- METHODS Standard protocol approvals, registrations, poses of this study, T2/fluid-attenuated inversion recovery and patient consents. This study was approved by the Institu- (FLAIR) signal, diffusion-weighted imaging and apparent tional Review Board of Johns Hopkins University. diffusion coefficient, and T1 pre- and post-administration of gadolinium sequences were reviewed and rated by each reviewer Patients. We identified admitted patients with AE who under- as consistent or inconsistent with AE, with differences in rating went FDG-PET/CT at Johns Hopkins Hospital through the reconciled by discussion between the reviewers. course of admission using the diagnostic terms encephalitis and positron emission tomography (PET) to search the administrative Statistical methods. The Mann-Whitney U test was used for database (December 1, 2005, to March 15, 2016). Patients were comparisons of continuous variables. Categorical variables were cross-referenced with the Johns Hopkins Hospital PET/CT compared using the x2 test or Fisher exact test, as appropriate. Center database. Kappa measurement of agreement was performed to assess in- Included patients underwent a brain FDG-PET/CT study termodality agreement of MRI, EEG, and CSF inflammatory and had possible or definite AE, including definite limbic enceph- markers with brain FDG-PET/CT metabolic patterns. Kappa alitis, per consensus criteria.1 Diagnostic findings consistent with measurement of agreement was performed for brain FDG-PET/ AE included MRI and EEG abnormalities and the presence of CT metabolic patterns with detection of only 1 or at least 2 intrathecal inflammation on routine testing.1 Seropositive pa- diagnostic findings consistent with AE. p , 0.05 was considered tients had a paraneoplastic or cell surface antibody detected in significant. either the serum or the CSF using commercially available assays A Friedman test was performed to compare median Z-scores (Athena Diagnostics, Worcester, MA; Mayo Clinic Laboratories, across brain regions for all patients. Serial Wilcoxon rank-sum Rochester, MN). tests were performed to compare Z-scores between brain regions The electronic medical record was reviewed; data collected for all patients with p , 0.008 considered significant after Bon- were demographic information, clinical history and presentation, ferroni correction. Split-plot analyses of variance with significance diagnostic results, and whether corticosteroids or sedatives were of p , 0.008 after Bonferroni correction were performed to administered within 24 hours preceding FDG-PET/CT compare patterns of FDG-avidity across 6 FDG-PET/CT brain study.19,20 regions within patients and between the seropositive and seronegative AE groups, definite and possible AE groups, as well as for Brain FDG-PET/CT review. Blinded review of brain FDG- those treated with corticosteroids and those treated with sedatives PET/CT was performed by 2 board-certified nuclear medicine within 24 hours of brain FDG-PET/CT and those not treated. radiologists (L.S. and M.S.J.). Dedicated 10-minute 3D brain Comparisons included Z-scores for both left and right hemi- FDG-PET/CT acquisitions were performed as per the institu- sphere regions for all patients. tional clinical protocol following whole-body acquisition and did not require additional radiopharmaceutical dose administration. Review of the literature. A PubMed search was performed All brain FDG-PET/CT images were acquired using a Discovery using positron emission tomography and encephalitis as search DRX or LX (GE Healthcare, Waukesha, WI) or Biograph mCT terms, updated up to October 27, 2016. Included studies and (Siemens, Knoxville, TN) in 3D mode for 10 minutes with in- case series reported brain FDG-PET findings (hypermetabolism line CT for attenuation correction. Filtered back projection and and/or hypometabolism) of at least 5 patients with AE or para- ordered subset expectation maximization methods were used to neoplastic encephalitis. When provided, reports of abnormalities reconstruct images, with respective reconstructions used in the on brain MRI, EEG, and CSF assays were reviewed. blinded review. The reconstructed data sets were fused and pro- jected to predefined surface pixels (3-dimensional stereotactic RESULTS Clinical characteristics of patients with AE surface projections) after anatomic standardization.21 Qualitative undergoing brain FDG-PET/CT. Of the 296 inpatients and quantitative PET image analysis was performed using a commercially available database of over 250 age-stratified with the diagnosis of encephalitis, 61 patients met the healthy controls, CortexID (GE Healthcare).21,22 Z-scores were consensus criteria for AE and underwent brain calculated for standard brain regions, and these regions were also FDG-PET/CT with studies available for review scored as normal, hypometabolic, or hypermetabolic by the 2 (table 1). Thirty-two of the 61 (52%) patients had board-certified nuclear medicine radiologists. Patients younger antibodies identified in the serum or CSF, 28 of than 30 years were compared with the lowest age group of con- whom with antibodies with known AE/paraneo- trols for Z-score calculations. The following standard CortexID brain regions were used as they could be reliably validated by plastic encephalitis significance, 24/61 (39%) with radiologists’ visual inspection: caudate, cerebellum, frontal lobe, definite AE antibodies (figure 1). Of the other sero- occipital lobe, parietal lobe, and temporal lobe. FDG-PET/CT positive patients, 4 were anti–voltage-gated potassium

2 Neurology: Neuroimmunology & Neuroinflammation Table 1 Clinical characteristics of patients with AE

All Seropositive Seronegative (N 5 61) (N 5 32) (N 5 29) p Value

Age, y, median (IQR) 54 (37) 39 (44) 57 (23) ,0.05a

Sex, female, n (%) 33 (54) 17 (53) 16 (55) 1.00

Race, n (%) 0.39

White 37 (61) 17 (53) 20 (69)

Black 9 (15) 5 (16) 4 (14)

Other 15 (24) 10 (31) 5 (17)

History of cancer, n (%) 11 (18) 3 (9) 8 (28) 0.09

Lymphoma 50 5

Breast 32 1

Testicular 11 0

Meningioma 10 1

Prostate and renal cell 10 1

Diagnosed with cancer during admission, n (%) 6 (10) 4 (13) 2 (7) 0.67

Small cell lung 33 0

Breast 10 1

Ovarian teratoma 10 1

Seminoma 11 0

Duration of neurologic symptoms before admission, 4 (7.5) 6 (10) 4 (8) 0.85 wk, median (IQR)

Neurologic signs and symptoms on admission, n (%)

Lethargy 47 (77) 22 (69) 25 (86) 0.13

Short-term memory impairment 46 (75) 25 (78) 21 (72) 0.77

Hallucinations 5 (8) 4 (13) 1 (3) 0.36

Cerebellar signs 47 (77) 21 (66) 26 (90) ,0.05a

Focal weakness 37 (61) 17 (53) 20 (69) 0.29

Focal numbness 35 (57) 15 (47) 20 (69) 0.12

Movement disorder 39 (64) 19 (59) 20 (69) 0.59

Seizures 25 (41) 16 (50) 9 (31) 0.19

Status epilepticus 10 (16) 4 (13) 6 (21) 0.50

Cranial neuropathy 14 (23) 8 (25) 6 (21) 0.77

Aphasia 25 (41) 12 (38) 13 (45) 0.61

Psychiatric symptoms 22 (36) 15 (47) 7 (24) 0.11

Focal neurologic findings on admission 58 (95) 30 (94) 28 (97) 1.00

Multiple focal neurologic findings on admission 50 (82) 22 (69) 28 (97) ,0.01a

Abbreviations: AE 5 autoimmune encephalitis; IQR 5 interquartile range. Clinical characteristics of patients with AE who underwent brain FDG-PET/CT through the course of inpatient evaluation. a Significant. channel-complex (VGKCc) seropositive for anti- Seropositive patients were younger than the sero- bodies different from anti-LGI1 and anti-CASPR2 (3 negative patients (median 39 vs 57 years, p 5 0.01). with supportive CSF, EEG, and/or MRI); 2 were Durations of symptoms before admission were similar anti-a3 AChR seropositive (1 with supportive EEG); for both groups (median 6 vs 4 weeks, p 5 0.85). and 2 were anti–striational antibody seropositive (1 Twelve of the 13 patients were younger than 30 years with supportive CSF and MRI). Two of the 4 anti– (5 anti-NMDA receptors [NMDARs]). Fourteen GAD65-seropositive patients had reviewable anti- (23%) patients’ CSF were tested for antibodies, with body levels (9,500 and 53,650 U/mL), whereas all the anti–NMDAR antibody detected in 3 patients, all had supportive CSF, EEG, and/or MRI. of whom were negative in the serum. No other

Neurology: Neuroimmunology & Neuroinflammation 3 Figure 1 Antibody status of patients with AE

Antibody status of patients with AE who underwent dedicated brain FDG-PET/CT (N 5 61). AE 5 autoimmune encephalitis; ANNA-1 5 anti–neuronal nuclear antibody 1; CRMP5 5 collapsin response mediator protein 5; GAD65 5 65 kDa glutamic acid decarboxylase enzyme; VGKCc 5 voltage-gated potassium channel-complex antibodies different from leucine-rich inactivated 1 protein (LGI1) and contactin-associated protein-2 (CASPR2); AChR 5 acetylcholine receptor antibody.

antibodies were detected in tested CSF samples. Of the onset (interquartile range [IQR] 9 weeks) and 17 patients with a history or subsequent diagnosis of a median of 4 days (IQR 8.5 days) from brain cancer, 7 were found to be seropositive: small cell lung MRI. Brain FDG-PET/CT was abnormal in 52/ cancer (1 anti-GAD65, 1 anti-Hu, and 1 anti-CV2), 61 patients (85%, figure e-1) when compared with breast (1 anti-NMDAR and 1 anti-LGI1), seminoma the healthy control database. FDG-PET/CT (1 anti-Ma2), and testicular cancer (1 anti-Ma2). demonstrated brain region hypometabolism alone in 42/61 (69%), hypermetabolism alone in 2/61 Other paraclinical findings. Routine CSF studies were (3%) patients, and regions of abnormal hypo- consistent with intrathecal inflammation in 34/55 pa- metabolism and abnormal hypermetabolism in 8/ tients (62%, figure e-1 at Neurology.org/nn). Initial 61 (13%) of patients (figure e-1, table e-3). No EEG for 17/56 (30%) patients demonstrated tem- differences were observed across age group, anti- poral area slowing, epileptiform discharges, or seizures body status, antibody class, or AE classification consistent with the diagnosis of AE. This was more (table e-1). No difference in proportion with frequently observed among patients younger than 30 abnormal metabolism was noted between those years (8/13, 62%) than others (9/43, 21%, p 5 0.01, evaluated by FDG-PET/CT within 4 weeks of table e-1). Brain MRI studies for 57 patients were symptoms (27/31 [87%]) and those evaluated later available for blinded review, and 23 (40%) were (25/30 [83%], p 5 0.73). consistent with the diagnosis of AE. No differences Across brain regions in patients with AE, metabo- were observed across antibody status, antibody class, lism was greater for the caudate (21.28, IQR 2.43) or AE classification (table e-1). relative to the frontal (22.24, IQR 2.69, p , 0.005), Fifty-one of the 61 patients (84%) had at least 1 temporal (21.80, IQR 1.77, p 5 0.002), parietal paraclinical finding consistent with AE on routine (22.49, IQR 1.61, p , 0.005), and occipital CSF analysis, brain MRI, or EEG. Thirty-one pa- (22.09, IQR 2, p , 0.005) brain regions (p , tients (51%) had only 1 paraclinical finding; 17 pa- 0.005, figure 2A). Brain region metabolism patterns tients (28%) had 2 findings; and 3 patients (5%) did not vary between seropositive and seronegative had 3 findings consistent with AE. Ten patients were AE patient groups (F(1,120) 5 3.18, p 5 0.08, h 2 included based on clinical criteria, 4 of whom with p 0.03, figure 2B) nor definite AE and possible AE (F definite AE based on detected antibodies, and 3 sero- (1,120) 5 2.69, p 5 0.10, h 2 0.02). Similarly, brain positive for other antibodies (table e-2). p region metabolism patterns did not vary between Brain FDG-PET/CT findings. Brain FDG-PET/CT those treated with corticosteroids (F(1,120) 5 5 h 2 was performed a median of 4 weeks after symptom 0.200, p 0.656, p 0.002) nor those treated with

4 Neurology: Neuroimmunology & Neuroinflammation Figure 2 Metabolism across brain regions in AE

Boxplots of Z-scores for FDG-avidity for brain areas on dedicated FDG-PET/CT for (A) patients meeting consensus criteria for AE, (B) seronegative and seropositive patients meeting the consensus criteria for AE. Z-scores varied across brain regions for patients with AE (p , 0.005), with values for the caudate being greater than those for frontal (p , 0.005), temporal (p 5 0.002), parietal (p , 0.005), and occipital (p , 0.005) lobes. No difference was noted between seronegative and seropositive patient groups (p 5 0.08). AE 5 autoimmune encephalitis.

5 5 h 2 k 5 5 sedatives (F(1,120) 1.95, p 0.165, p 0.016) metabolism by FDG-PET/CT ( 0.16, p within 24 hours of brain FDG-PET/CT and those 0.02; table e-4). not treated. Literature review. Fourteen studies were identified Concordance between FDG-PET/CT and other para- which met the inclusion criteria (table 2). Of the clinical findings. The finding of an abnormal meta- 139 FDG-PET studies reported, 120 (86%) were bolic pattern on brain FDG-PET/CT was not in abnormal, with 55 (40%) demonstrating both hypo- agreement with the presence of CSF inflammation on and hypermetabolism, 30 (22%) demonstrating only routine assessment (table e-4). By contrast, brain hypometabolism, and 35 (25%) demonstrating only MRI findings consistent with AE were in weak hypermetabolism. This is compared with the sum agreement with the finding of abnormal metabolism report of 38/75 (51%) EEGs, 68/114 (60%) brain on brain FDG-PET/CT (k 5 0.17, p , 0.05), most MRI, and 45/86 (52%) routine CSF studies consis- notably with hypometabolism (k 5 0.25, p , 0.05; tent with the diagnosis of possible AE. table e-4, figure 3). In general, the presence of any FDG-PET/CT abnormality was not in agreement DISCUSSION Here, we describe dedicated semi- with the presence of EEG findings consistent with AE quantitative brain FDG-PET/CT findings among (table e-4). However, detection of EEG findings patients meeting the consensus AE criteria. Dedicated consistent with the diagnosis of AE was in weak brain FDG-PET/CT was abnormal in 85% of pa- agreement with detection of brain region hyperme- tients with AE, and FDG-PET abnormalities were tabolism (k 5 0.16, p , 0.05) and in fair agreement more sensitive for AE compared with EEG, MRI, or with having regions of both hypermetabolism and routine CSF findings. Although brain region hypo- hypometabolism in the same FDG-PET/CT study metabolism was most commonly noted, some studies (k 5 0.26, p , 0.05, table e-4). demonstrated areas of both hyper- and hypo- Detection of at least 1 paraclinical finding consis- metabolism and a minority demonstrated hyperme- tent with AE was not in agreement with detection of tabolism alone. The combination of abnormalities in abnormal metabolism by FDG-PET/CT (table e-4). at least 2 of the 3 other paraclinical tests (routine Detection of 2 or more consistent findings was in CSF studies, brain MRI, and EEG) was in fair weak agreement with detection of abnormal brain agreement with abnormal findings on dedicated brain

Neurology: Neuroimmunology & Neuroinflammation 5 Figure 3 Brain MRI, brain FDG-PET/CT and hypometabolic 3D-SSP maps for 3 patients with AE

Brain MRI, brain FDG-PET/CT, and hypometabolic 3D-SSP maps, respectively, for patients with anti-NMDAR encephalitis (A–C), anti-LGI1 encephalitis (D–F), and seronegative AE (G–I). For the anti-NMDAR encephalitis patient, note normal T2/ FLAIR MRI (A) with right basal ganglia, right frontotemporoparietal, left frontal, and bilateral posterior cortical hypometabolism centered on the middle occipital lobe on FDG-PET/CT and 3D-SSP maps (B and C). For the anti-LGI1 patient, note normal T2/FLAIR MRI (D) with relatively normal basal ganglia metabolism (E) in setting of diffuse frontotemporoparietal hypometabolism on FDG-PET/CT and 3D-SSP maps (E and F). For the seronegative AE patient, again note normal T2/FLAIR MRI (G) with diffuse frontotemporoparietal hypometabolism on FDG-PET/CT and 3D-SSP maps (H and I). A 5 anterior; AE 5 autoimmune encephalitis; L 5 left; LGI1 5 leucine-rich inactivated 1 protein; NMDAR 5 NMDA receptor; P 5 posterior; R 5 right; and 3D-SSP, 3-dimensional stereotactic surface projection.

FDG-PET/CT. Our results suggest that brain FDG- AE is reversible. Moreover, although widespread PET/CT may be helpful in supporting evidence of regional hypometabolism was observed across vari- brain dysfunction in suspected patients with AE. ous AE syndromes, there are likely syndrome- Brain region hypometabolism in multiple regions specific patterns of brain region metabolism.8,9,15 likely reflects widespread impairment of neuronal Previous series primarily report hypermetabolism activity in AE.23 Whether such hypometabolism re- in AE. These series contain larger proportions of sults from functional changes, structural changes, or patients with anti-NMDAR (36/130 reported pa- a combination of both is not yet clear. Many of the tients) or patients with anti-LGI1, anti-CASPR2, areas of regional hypometabolism did not have cor- or anti-VGKCc antibodies (39/130) than our relates on MRI, suggesting the possibility of neuro- cohort, potentially limiting their generalizability to nal dysfunction in the absence of structural other seropositive and seronegative AE.4–6 We disturbance. Longitudinal studies will be needed to observed brain region hypermetabolism in a subset clarify whether the observed hypometabolism in of patients, many of whom had anti-NMDAR or

6 Neurology: Neuroimmunology & Neuroinflammation Table 2 Literature review of FDG-PET findings in AE

No. of PET Comparison PET CSF inflammation studies/ to control Abnormal demonstrated PET demonstrated PET demonstrated EEG consistent MRI consistent demonstrated/ Reference patients in population Serum and/or CSF PET/PET hypometabolism, hypermetabolism, both hyper-/hypo- with AE/EEG with AE/MRI lumbar punctures no. series performed? antibody status (N) performed N N metabolism, N performed performed performed

2 7/7 No Seronegative (7) 6/7 1 4 1 2/7 6/7 6/7

3 9/9 No NMDAR (3), 9/9 3 6 7/9 seronegative (6)

4 18/18 Yes Hu (2), Ri (1), GAD65 (1), 13/18 8 5 10/18 10/18 LGI1 (1), CASPR2 (2) VGKCa (3), NMDAR (2), NMDAR/VGKCa (1), seronegative (5)

5 13/10 No Hu (2), VGKCa (2), 12/13 1 6 5 15/17 NMDAR (1), “Neuronal Cell Memberane” (2), “Atypical” (1), seronegative (2)

6 12/16 No LGI1 (1), VGKCa (1), 11/12 9 1 1 10/16 10/16 NMDAR (2), GAD65 (2), Neuropil (1), Ma2 (1), Ma2/Hu (1), nontype (1), seronegative (6) erlg:Nuomuooy&Neuroinflammation & Neuroimmunology Neurology: 8 6/6 Yes NMDAR (6) 6/6 3 3 2/6

9 10/10 Yes NMDAR (6), LGI1 (4) 10/10 10 0/10 2/10 6/10

10 13/6 No NMDAR (6) 12/13 4 2 6 6/6 2/6 6/6

11 18/8 Yes NMDAR (8) 14/18 1 1 12

12 6/8 No LGI1 (5), CASPR2 (2), 3/6 1 2 1/8 6/8 2/8 CASPR2/LGI1 (1)

13 6/6 No Seronegative (6) 6/6 1 5 3/5

14 5/5 No GABA(B) (5) 3/5 1 2 3/5 2/5 2/5

15 10/14 No LGI1 (10) 9/10 8 1 10/14 10/14 1/13

16 6/7 No VGKCa (6), NMDAR (1) 6/6 6 6/7 3/7 2/7

Total 120/139 30 35 55 38/75 68/114 45/86

Abbreviations: AE 5 autoimmune encephalitis; CASPR2 5 contactin-associated protein-2; GAD65 5 65 KDa glutamic acide decarboxylase; LGI1 5 leucine-rich inactivated 1 protein; NMDAR 5 NMDA receptor; VGKC 5 voltage-gated potassium channel. – – Systematic review of case series reporting FDG-PET findings in AE along with the available EEG, MRI, and CSF study reports.2 6,8 16 Only anti-VGKCa seropositivity reported and potentially includes those seropositive for anti-LGI1, anti-CASPR2, or other VGKC-complex antibodies. 7 anti-VGKCc encephalitis, compatible with previous Future prospective studies involving serial FDG-PET literature.5,8–12,15 studies may help clarify the specificity and evolution The current AE consensus criteria only include of patterns of metabolism through the phases of FDG-PET findings in the criteria for definite autoim- encephalitis, as has been suggested in cases series of mune limbic encephalitis. Bilateral FLAIR/T2 abnor- specific encephalitides such as anti-NMDAR enceph- malities of the medial temporal lobes are required, alitis.11 Our study included the initial FDG-PET and in the absence of such findings, FDG-PET studies for patients regardless of antibody status, hypermetabolism in the medial temporal lobes may and future larger prospective studies of specific anti- meet this requirement. Observations provided here body syndromes may further clarify patterns of suggest that AE may lead to broader metabolic abnor- abnormality, pattern associations with clinical status, malities detectable by FDG-PET outside the confines and pattern changes in the setting of immune therapy of the medial temporal lobes and these may inform as has been observed in cases of autoimmune demen- future FDG-PET AE criteria. tia.27 Not all patients underwent CSF antibody test- Concerns raised regarding the incorporation of ing, which may be more sensitive, and thus we may FDG-PET in the evaluation of patients with AE underestimate the number of seropositive patients. include availability of FDG-PET imaging modalities However, there was no difference noted in brain in urgent clinical situations. Moreover, as a newer region metabolism between seropositive and seroneg- modality, further work is needed to validate it as ative groups. Also, 4 patients were anti-VGKCc sero- a method in the diagnosis of AE.24 Worldwide, positive without further specification, and although 3 FDG-PET/CT represents one of the medical imaging had other findings supportive of AE, the clinical value modalities with the largest growth in terms of number of such antibodies is unknown and cautious interpre- of scanners.25 In addition, FDG-PET/CT has been tation is advised. One-third of patients studied here found to be diagnostically superior to other conven- were treated with either corticosteroids or sedatives tional imaging modalities in other clinical settings, within 24 hours of FDG-PET/CT. Although both and it has demonstrated cost-effectiveness in settings corticosteroids and sedatives have been reported to such as non–small lung cancer staging.25 FDG-PET decrease cortical metabolism,19,20 no differences in also plays an important role in screening for occult brain region metabolism were noted between patients malignancy in paraneoplastic syndromes, including exposed and unexposed to these medications before encephalitis.26 Thus, FDG-PET is likely to become FDG-PET/CT. In addition, FDG-PET/CT metabo- an increasingly used modality in the evaluation of lism patterns for patients with AE were not compared patients with suspected AE beyond occult malignancy with other patients with neurologic diseases (such as screening. Many institutions use a “vertex to toe” infectious encephalitis); psychiatric diseases; intoxica- field of view for their whole-body protocols. The tions; or other syndromes which may also have abnor- addition of a 10-minute dedicated 3D PET acquisi- mal FDG-PET findings.21,22,28–31 It will be important tion of the brain requires no extra radiopharmaceuti- for future prospective studies to incorporate patients cal administration, is easily incorporated in with other neurologic, psychiatric, and medical dis- conventional clinical workflows, provides increased eases to assess the specificity of metabolic findings by statistical quality in comparison with “vertex to toe” FDG-PET described here. Finally, the CortexID con- imaging, and allows for higher-resolution images with trol population used for comparison ranges from 30 more robust quantitation. As the utility of FDG-PET to 85 years. The 13 patients younger than 30 years is evaluated further, collaborative evaluation by neu- studied had a similar rate of abnormal brain region rologists and radiologists will be necessary for careful metabolism compared with those older. Ideally, a con- characterization and correlation of syndromes with current age- and sex-matched control population associated imaging findings, with comparisons to could be used for direct comparison, although such healthy and other neurologic patient populations. data collection is limited by the radiation exposure to A major limitation of this study is that it is retro- otherwise normal patients. spective, involving all patients meeting the criteria Here, brain FDG-PET/CT was commonly abnor- for AE who underwent FDG-PET/CT at a single ter- mal in AE, most often demonstrating brain region tiary medical center with associated selection bias. hypometabolism. The frequency of metabolic abnor- Although performed at a single center, it benefits malities was greater than that of diagnostic studies from the consensus inclusion criteria for AE and uni- currently included in consensus criteria for the diag- formity of PET equipment, protocols, and analyses. nosis of AE. Overall, FDG-PET/CT may represent Also, the observations reported here were limited to a sensitive and early biomarker for AE and could play those patients admitted to the hospital for onset of a complementary role to currently proposed tests in symptoms of 3 months or less and do not include the diagnosis of AE. Future prospective studies may findings for those with longer duration of symptoms. further clarify the role FDG-PET may play in the

8 Neurology: Neuroimmunology & Neuroinflammation diagnosis and monitoring of AE in general and spe- 10. Lagarde S, Lepine A, Caietta E, et al. Cerebral (18)Fluo- cific antibody syndromes in particular. roDeoxy-Glucose Positron Emission Tomography in pae- diatric anti N-methyl-d-aspartate receptor encephalitis: – AUTHOR CONTRIBUTIONS a case series. Brain Dev 2016;38:461 470. 11. Yuan J, Guan H, Zhou X, et al. Changing brain Dr. Probasco: design and conceptualization of the study, analysis and interpretation of the data, and drafting and revising of the manuscript. metabolism patterns in patients with ANMDARE: Dr. Solnes: design and conceptualization of the study, analysis and inter- serial 18F-FDG PET/CT findings. Clin Nucl Med pretation of the data, and revising of the manuscript. Mr. Nalluri: anal- 2016;41:366–370. ysis and interpretation of data. Mr. Cohen, Dr. Jones, and Dr. Zan: 12. Chen Y, Xing XW, Zhang JT, et al. Autoimmune enceph- analysis and interpretation of the data and revising of the manuscript. alitis mimicking sporadic Creutzfeldt-Jakob disease: a ret- Dr. Javadi and Dr. Venkatesan: design and conceptualization of the rospective study. J Neuroimmunol 2016;295-296:1–8. study, analysis and interpretation of the data, and revising of the 13. Lee BY, Newberg AB, Liebeskind DS, Kung J, Alavi A. manuscript. FDG-PET findings in patients with suspected encephalitis. Clin Nucl Med 2004;29:620–625. STUDY FUNDING 14. Kim TJ, Lee ST, Shin JW, et al. Clinical manifestations No targeted funding reported. and outcomes of the treatment of patients with GABAB encephalitis. J Neuroimmunol 2014;270:45–50. DISCLOSURE 15. Shin YW, Lee ST, Shin JW, et al. VGKC-complex/LGI1- J.C. Probasco serves on the editorial board for The Neurohospitalist,isan antibody encephalitis: clinical manifestations and response associate editor for The Neurohospitalist, and is editor-in-chief for NEJM to immunotherapy. J Neuroimmunol 2013;265:75–81. Journal Watch Neurology. L. Solnes, A. Nalluri, J. Cohen, K.M. Jones, 16. Newey CR, Sarwal A, Hantus S. [(18)F]-Fluoro-deoxy- E. Zan, and M.S. Javadi report no disclosures. A. Venkatesan received speaker honoraria from Almirall, served as a medical expert for U.S. glucose positron emission tomography scan should be Government Vaccine Injury Compensation Program, received research obtained early in cases of autoimmune encephalitis. Auto- support from NIH, and served as medical expert for Carnival Cruise immune Dis 2016;2016:9450452. Lines. Go to Neurology.org/nn for full disclosure forms. 17. Lee EM, Kang JK, Oh JS, Kim JS, Shin YW, Kim CY. 18F-Fluorodeoxyglucose positron-emission tomography Received January 23, 2017. Accepted in final form March 27, 2017. findings with anti-N-methyl-d-aspartate receptor encephalitis that showed variable degrees of catatonia: three cases REFERENCES report. J Epilepsy Res 2014;4:69–73. 1. Graus F, Titulaer MJ, Balu R, et al. A clinical approach to 18. Novy J, Allenbach G, Bien CG, Guedj E, Prior JO, Rossetti diagnosis of autoimmune encephalitis. Lancet Neurol AO. FDG-PET hyperactivity pattern in anti-NMDAr 2016;15:391–404. encephalitis. J Neuroimmunol 2016;297:156–158. 2. Ances BM, Vitaliani R, Taylor RA, et al. Treatment- 19. Fulham MJ, Brunetti A, Aloj L, Raman R, Dwyer AJ, Di responsive limbic encephalitis identified by neuropil Chiro G. Decreased cerebral glucose metabolism in pa- antibodies: MRI and PET correlates. Brain 2005;128: tients with brain tumors: an effect of corticosteroids. 1764–1777. J Neurosurg 1995;83:657–664. 3. Fisher RE, Patel NR, Lai EC, Schulz PE. Two different 20. Matheja P, Weckesser M, Debus O, et al. Drug-induced 18F-FDG brain PET metabolic patterns in autoimmune changes in cerebral glucose consumption in bifrontal epi- limbic encephalitis. Clin Nucl Med 2012;37:e213–e218. lepsy. Epilepsia 2000;41:588–593. 4. Baumgartner A, Rauer S, Mader I, Meyer PT. Cerebral 21. Minoshima S, Frey KA, Koeppe RA, Foster NL, Kuhl DE. FDG-PET and MRI findings in autoimmune limbic A diagnostic approach in Alzheimer’s disease using three- encephalitis: correlation with autoantibody types. dimensional stereotactic surface projections of fluorine-18- J Neurol 2013;260:2744–2753. FDG PET. J Nucl Med 1995;36:1238–1248. 5. Masangkay N, Basu S, Moghbel M, Kwee T, Alavi A. 22. Josephs KA, Duffy JR, Strand EA, et al. Characterizing Brain 18F-FDG-PET characteristics in patients with para- a neurodegenerative syndrome: primary progressive apraxia neoplastic neurological syndrome and its correlation with of speech. Brain 2012;135:1522–1536. clinical and MRI findings. Nucl Med Commun 2014;35: 23. Dalmau J, Lancaster E, Martinez-Hernandez E, Rosenfeld MR, 1038–1046. Balice-Gordon R. Clinical experience and laboratory investiga- 6. Aupy J, Collongues N, Blanc F, Tranchant C, Hirsch E, tions in patients with anti-NMDAR encephalitis. Lancet De Seze J. Autoimmune encephalitis, clinical, radiological Neurol 2011;10:63–74. and immunological data [in French]. Rev Neurol [Paris] 24. Graus F, Dalmau J. Role of (18)F-FDG-PET imaging in 2013;169:142–153. the diagnosis of autoimmune encephalitis - Authors’ reply. 7. Morbelli S, Djekidel M, Hesse S, Pagani M, Barthel H. Lancet Neurol 2016;15:1010. Role of (18)F-FDG-PET imaging in the diagnosis of auto- 25. Buck AK, Herrmann K, Stargardt T, Dechow T, Krause BJ, immune encephalitis. Lancet Neurol 2016;15:1009–1010. Schreyogg J. Economic evaluation of PET and PET/CT in 8. Leypoldt F, Buchert R, Kleiter I, et al. Fluorodeoxyglucose oncology: evidence and methodologic approaches. J Nucl positron emission tomography in anti-N-methyl-D- Med Technol 2010;38:6–17. aspartate receptor encephalitis: distinct pattern of disease. 26. McKeon A, Apiwattanakul M, Lachance DH, et al. Pos- J Neurol Neurosurg Psychiatry 2012;83:681–686. itron emission tomography-computed tomography in par- 9. Wegner F, Wilke F, Raab P, et al. Anti-leucine rich glioma aneoplastic neurologic disorders: systematic analysis and inactivated 1 protein and anti-N-methyl-D-aspartate review. Arch Neurol 2010;67:322–329. receptor encephalitis show distinct patterns of brain glu- 27. Flanagan EP, McKeon A, Lennon VA, et al. Autoimmune cose metabolism in 18F-fluoro-2-deoxy-d-glucose positron dementia: clinical course and predictors of immunother- emission tomography. BMC Neurol 2014;14:136. apy response. Mayo Clinic Proc 2010;85:881–897.

Neurology: Neuroimmunology & Neuroinflammation 9 28. Foster NL, Heidebrink JL, Clark CM, et al. FDG-PET 30. Hubele F, Bilger K, Kremer S, Imperiale A, Lioure B, improves accuracy in distinguishing frontotemporal Namer IJ. Sequential FDG PET and MRI findings in dementia and Alzheimer’s disease. Brain 2007;130: a case of human herpes virus 6 limbic encephalitis. Clin 2616–2635. Nucl Med 2012;37:716–717. 29. Wong KK, Tolia B, Bohnen N. Chronic sequelae of her- 31. Vollenweider FX, Kometer M. The neurobiology of psy- pes simplex encephalitis demonstrated on interictal F-18 chedelic drugs: implications for the treatment of mood FDG PET/CT. Clin Nucl Med 2008;33:443–444. disorders. Nat Rev Neurosci 2010;11:642–651.

10 Neurology: Neuroimmunology & Neuroinflammation Antiepileptic drug therapy in patients with autoimmune epilepsy

Anteneh M. Feyissa, MD, ABSTRACT MSc Objective: We aimed to report the pattern of usage and efficacy of antiepileptic drugs (AEDs) in A. Sebastian López patients with autoimmune epilepsy (AE). Chiriboga, MD Methods: We retrospectively studied the Mayo Clinic’s electronic medical record of patients with Jeffrey W. Britton, MD AE in which seizures were the main presenting feature. Clinical data, including demographics, seizure characteristics, type of AED and immunotherapy used, presence of neural antibody, and treatment outcomes, were reviewed. Correspondence to Dr. Feyissa: Results: The medical records of 252 adult patients diagnosed with autoimmune encephalitis and [email protected] paraneoplastic disorders were reviewed. Seizure was the initial presentation in 50 patients (20%). Serum and/or CSF autoantibodies were detected in 41 (82%) patients, and 38 (76%) patients had neural autoantibodies. The majority (n 5 43, 86%) received at least 1 form of immunotherapy in combination with AEDs, while the remainder received AEDs alone. Twenty-seven patients (54%) became seizure free: 18 (36%) with immunotherapy, 5 (10%) with AEDs alone, and 4 (8%) with AEDs after immunotherapy failure. Levetiracetam was the most commonly used (42/50); however, it was associated with 0% seizure-free response. AED seizure-free responses occurred with carbamazepine (n 5 3) [3/16, 18.8%], lacosamide (n 5 3) [3/18, 16.6%] with phenytoin (n 5 1) [1/8, 12.5%], or oxcarbazepine (n 5 2) [2/11, 18.1%]. Regardless of the type of therapy, voltage-gated potassium channel-complex antibody–positive patients were more likely to become seizure free compared with glutamic acid decarboxylase 65 antibody–positive cases (12/17 vs 2/10, p 5 0.0183). Conclusions: In select patients, AEDs alone were effective in controlling seizures. AEDs with sodium channel blocking properties resulted in seizure freedom in a few cases. Prospective studies are needed to clarify AED selection and to elucidate their immunomodulatory properties in AE. Neurol Neuroimmunol Neuroinflamm 2017;4:e353; doi: 10.1212/NXI.0000000000000353

GLOSSARY AE 5 autoimmune epilepsy; AED 5 antiepileptic drug; EMR 5 electronic medical record; G-AChR 5 ganglionic acetylcholine receptor; NMDAR 5 NMDA receptor; TNFa 5 tumor necrosis factor–alpha; TPO 5 thyroid peroxidase; VGKC-complex 5 voltage-gated potassium channel–complex.

Autoimmune epilepsy (AE) has been linked to both antineural antibodies targeting neural intracellular proteins (GAD65, ANNA-1, Ma, etc) and cell surface antigens (voltage-gated potassium channel–complex [VGKC-complex], NMDA receptor, AMPA, GABA-B, mGluR5, etc). AE can occasionally occur in the absence of detected neural antibodies as well.1 Valuable clinical clues for AE are subacute onset, an unusually high seizure frequency, intraindividual seizure variability or multifocality, antiepileptic drug (AEDs) resistance, personal or family history of autoimmunity, or history of recent or past neoplasia.2 Seizures occur as an early and prominent feature in AE, and these are characteristically refractory to conventional AED therapy.1

From the Department of Neurology (A.M.F., A.S.L.C.), Mayo Clinic, Jacksonville, FL; and Department of Neurology (J.W.B.), Mayo Clinic, Rochester, MN. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the Mayo Clinic. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology 1 Although intractability is a common feature usage and relative efficacy of AEDs in patients in AE, some respond to AEDs, and they with AE. remain an important aspect of therapy. The role of AEDs in these patients is also relevant METHODS To identify all relevant adult cases of AE, we as even after controlling the inflammatory searched the medical record index system of the Mayo Clinic, Ro- chester, for the terms “autoimmune encephalitis,”“autoimmune response, some patients remain at risk of epilepsy,”“autoimmune seizure,” and “limbic encephalitis” from recurrent seizures, particularly if cerebral dam- January 1, 2013, through December 31, 2015. Our search resulted age occurred during the acute encephalitic in 252 cases. Among these, 50 patients were identified who fulfilled the following criteria: (1) seizures as the exclusive or predominant phase of the illness. Some AEDs affect cellular presenting complaint and (2) an autoimmune etiology suspected on 3–6 and humoral immune responses. For exam- the basis of clinical presentation, inflammatory CSF, MRI charac- ple, carbamazepine and valproate have been teristics suggesting inflammation, or detection of serum and/or CSF 7 shown to increase the serum levels of IL-1b, neural autoantibody (figure 1). Demographic, clinical (seizure semiology, course, and associated symptoms), autoimmune serol- IL-2, IL-4, IL-6, IL-17, and tumor necrosis ogy, type of AED and immunotherapy used, EEG and radiologic factor–alpha (TNFa) production.4,5 At this characteristics, and treatment outcomes were reviewed. time, no single AED stands out as superior to Baseline seizure frequency was determined by reviewing the others in AE, and the medically refractory seizure frequency stated in the clinical record prior to initiation of treatment and categorized as daily (.1 seizure per day), weekly nature of seizures in these patients is typically (.1 seizure per week but not daily), or monthly (.1 seizure per emphasized. However, occasional patients do month but not weekly). Response to AED with and without respond to AED treatment, and it is not cur- immunotherapy was determined by review of the record docu- menting course following treatment initiation. rently known if this occurs with medications featuring particular mechanisms of action. In Statistical analysis. Data were expressed as mean, range, and this study, we aimed to explore the pattern of SD for continuous variables, and counts (percentages) for

Figure 1 Flowchart showing methodology of study design

AANA-1 5 antineuronal nuclear antibody–1; AED 5 antiepileptic drug; EMR 5 electronic medical record; G-ACR 5 ganglionic acetylcholine receptor; GAD65 5 glutamic acid decarboxylase 65; NMDAR 5 NMDA receptor; SSA 5 Sjögren’s- syndrome-related antigen A; TPO 5 thyroid peroxidase; VGCC 5 voltage-gated calcium channel; VGKC-complex 5 voltage- gated potassium channel–complex.

2 Neurology: Neuroimmunology & Neuroinflammation categorical variables. We compared treatment response by antibody type (VGKC-complex antibodies vs GAD65 vs antibody Table 1 Summary of clinical characteristics of 50 patients with suspected AE negative) using the Fisher exact test. All analyses were performed using SAS software (version 9.3). All statistical tests were 2 sided, Characteristics and p , 0.05 was considered statistically significant. Sex, n (%) Standard protocol approvals, registrations, and patient consents. The study was approved by the Mayo Clinic Institu- Male 26 (52) tional Review Board, and all patients consented to the use of their Female 24 (48) medical records for research purposes. Mean age at onset, y 41

Seizure type, n (%) RESULTS Clinical characteristics. Patients age ranged Focal seizures with/without loss of 33 (66) from 10 to 87 years (mean age 41.3 years); 26 (52%) awareness were men. The mean age at onset was 41 years. The Faciobrachial dystonic seizures 4(8) median follow-up period was 18 months (range 3–44 Epilepsia partialis continua 4(8) months). Seizure types were focal seizures with and GTCs with/without status 14 (28) without loss of awareness, faciobrachial dystonic seiz- Others (absence, myoclonic, hypermotor, 5 (10) ures, epilepsy partialis continua, and generalized tonic- and atonic) clonic seizures with/without status. EEG was abnormal No. of AEDs, n (%) with interictal discharges or documented seizure activity £ in the majority of cases. MRI data were abnormal in 23 1 19 (38) patients. Brain PET CT was performed in 4 patients, ‡2 31 (62) and it was abnormal in each. Summary of clinical Seizure frequency, n (%) characteristics of the 50 patients is provided in table 1. ‡1/day 26 (52)

‡1/week 9 (18) Autoantibody evaluation. Neural autoantibodies were identified in 38 patients (76%), of which 26 had anti- ‡1/month 5 (10) bodies targeting plasma membrane proteins: VGKC- <1/month 10 (20) complex antibody (n 5 17) (LGI1 [11], CASPR2 Seizure freedom, n (%) 30 (60)

[1], and double negative [5]); ganglionic acetylcholine With immunotherapy (IVMP and/or IVIG) 19 (38) receptor (G-AChR) (n 5 5); NMDA receptor (n 5 3); AEDs alone 5 (10) P/Q-type calcium channel (n 5 1), and 12 had anti- AEDs after failing immunotherapy 4(8) bodies targeting cytoplasmic or nuclear proteins: P GAD65 (n 5 10), ANNA-1 (n 5 1), and ANNA-2 (n Others (MAD, VNS, etc) 2(4) 5 1). Three patients (6%) had non-neuronal antibodies Neuronal autoantibody status, n (%) (thyroid peroxidase [TPO], Ant-Ro, and anti–dsDNA), Neuronal autoantibody 38 (76) and nine patients (18%) were autoantibody negative. VGKC-complex 17 (34)

Treatments and response. The majority (n 5 43, 86%) GAD65 10 (20) received at least 1 form of immunotherapy in combi- G-AChR 5 (10) nation with AEDs, while the reminder received AEDs NMDAR 3(6) alone. The median number of AEDs used was 2 (range P/Q-type calcium channel 1(2) 1–6). Levetiracetam was the most common medication ANNA-1 1(2) used (n 5 42), followed by lacosamide (n 5 18), ANNA-2 1(2) carbamazepine (n 5 16), oxcarbazepine (n 5 11), – lamotrigine (n 5 11), and phenytoin (n 5 8). Others (TPO, Ant-Ro, anti dsDNA) 3(6) Summary of AED use for the cohort is provided in Negative 9 (18) figure 2. In addition, 4 patients received vagus nerve MRI with probable inflammatory 23 (46) changes, n (%) stimulation therapy, and 3 patients had unsuccessful EEG with IEDs or documented seizure 38 (76) epilepsy surgeries. The rationale for selecting one AED activity, n (%) over the other was not clarified in the electronic medical record (EMR). Twenty-seven patients (54%) Abbreviations: AANA-1 5 antineuronal nuclear antibody–1; AANA-2 5 antineuronal nuclear antibody–2; G-AChR 5 gan- became seizure free: 18 (36%) with immunotherapy, 5 glionic acetylcholine receptor; AE 5 autoimmune epilepsy; AED (10%) with AEDs alone, and 4 (8%) with AEDs after 5 antiepileptic dug; GAD65 5 glutamic acid decarboxylase 65; failing immunotherapy. Initiation of carbamazepine (n GTC 5 generalized tonic-clonic seizure; IED 5 interictal dis- 5 5 charge; IVIG 5 IV immunoglobulin; IVMP 5 IV methylprednis- 3) (3/16, 18.8%), lacosamide (n 3) (3/18, 5 5 5 olone; MAD modified Atkins diet; NMDAR NMDA receptor; 16.6%), phenytoin (n 1) (1/8, 12.5%), or oxcar- TPO 5 thyroid peroxidase; VGKC-complex 5 voltage-gated bazepine (n 5 2) (2/11, 18.1%) resulted in seizure-free potassium channel–complex; VNS 5 vagaus nerve stimulation.

Neurology: Neuroimmunology & Neuroinflammation 3 Figure 2 Pattern of AED use in 50 patients with suspected AE

AE 5 autoimmune epilepsy; AED 5 antiepileptic drug.

outcomes, while none of the patients became seizure immunotherapy is often the focus of management,7 free with levetiracetam (n 5 0) (0/42, 0%). The this study suggests that in select cases AEDs alone timing of therapeutic response to AEDs (i.e., cessation may confer seizure freedom. of clinical seizures) was immediate: within 24 hours in Of interest in our study, 9 patients became sei- 1 patient, within 1 week in 2, within 2 weeks in 3, and zure free after the initiation of AEDs with sodium within 4 weeks in the remainder. channel blocking properties7 (carbamazepine, n 5 Among those who responded to AED therapy, 4 3; oxcarbazepine, n 5 2; lacosamide alone, n 5 3, were VGKC-complex antibody positive (1 LGI1, 1 or with phenytoin, n 5 1). The response to AEDs CASPR2, and 2 double negative), 1 was GAD65 wasseenwithin2–4 weeks of the initiation of ther- antibody positive, 1 was AChR antibody positive, 1 apy in all patients. In a few patients, AEDs alone was TPO antibody positive, and 2 were autoantibody were effective in controlling seizures. However, negative. Regardless of the type of therapy, VGKC- improvement of cognitive symptoms and other complex antibody–positive patients were more likely manifestations of autoimmune encephalitis in most to become seizure free compared with GAD65- patients requires immunotherapy.7 It is conceivable positive and autoimmune antibody–negative cases that these therapeutic responses could be secondary (12/17 vs 2/10 vs 2/9 respectively, p , 0.05). Spe- to immunomodulatory properties of these medica- cifically, of the VGKC-complex antibody patients tions.4–6 For example, carbamazepine has been who became seizure free, 4 responded to AEDs shown to reduce levels of proinflammatory cyto- (25%), while the remainder became seizure free with kines IL-1b and TNFa in the hippocampus of rats4 immunotherapy (75%). One patient with GAD65 and inhibits the development of different types of ab-associated AE responded to carbamazepine, while inflammation through dose-dependent reduction the other became seizure free with immunotherapy. of prostaglandin E2 and substance P.5 Carbamaze- Two of the autoantibody-negative cases responded to pine and oxcarbazepine have also been shown to AEDs (carbamazepine), and 2 became seizure free decrease serum levels of IL-1 and IL-2 in healthy with immunotherapy. Details of patients who re- subjects.5 Although disruption of blood-brain sponded to AEDs alone or after failure of immuno- barrier permeability has been suggested as a possible therapy are summarized in table 2 and figure 3. pathway for cytokines influencing seizures and epilepsy,8 the exact role of cytokines in human epi- DISCUSSION Although the seizures in AE are lepsy is complex and not yet completely elucidated. often considered to be medication resistant, and Moreover, whether there are unique aspects of the

4 Neurology: Neuroimmunology & Neuroinflammation pathophysiology of seizure onset in these patients which make them more susceptible to sodium channel blockage is not known. The observed

voltage-gated response could also be due to the effectiveness of 5

generalized tonic- sodium channel blockers in focal onset epilepsy 5 Duration of seizure freedom, mo 16 11 which is seen in 66% of patients in our cohort. Of note, while AED responders responded to sodium channel blockers, no patient responded to lamotrigine, which also has sodium channel properties.9 Therefore, if there is a unique responsivity of seizures in these patients to this class of medications, the response may not be uniform throughout None 48 IVMP (later for cognitive impairment) NoneNone 68 7 cognitive impairment) the entire class. Studies are needed to elucidate

thyroid peroxidase; VGKC-complex whether any differential efficacy is based on their ve 5 1 effect on seizure mechanisms or immunomodula-

glutamic acid decarboxylase 65; GTC tory properties. 5 Regardless of the type of therapy, VGKC- ve (0.15 nm/L) ve, (0.19 nm/L) male; TPO 2 2 –

5 complex antibody positive patients were more likely to become seizure free compared with ve (0.10 nm/L) (0.33 nm/L) Autoantibody profile (titer) Immunotherapy tried Negative IVMP 24 CASPR2 1 VGKC-complex, LGI1 and CASPR2 G-AChR (0.64 nm/L) IVIG then IVMP 30 TPO, 21 IU/mL IVMP (later for Negative NoneGAD65-positive 20 and autoimmune antibody– negative cases (figure 2). This is consistent with previous observations of superior immunotherapy response in patients with neural surface–related autoantibody-mediated encephalitides as com- enhancement

1 pared to those associated with intracellular antigen 1,7,10 L hippocampal autoantibodies. Of note, among the patients swollen amygdala . 1 atrophy atrophy T2 signal change, amygdala change who responded to AEDs in our cohort, the major- ganglionic acetylcholine receptor; GAD65 leucine-rich inactivated 1 protein; M

5 ity were VGKC-complex antibody positive (1 5 LGI1 positive; 1 CASPR2 positive, and 2 double negative). Two of these responded to oxcarbazepine, one to lacosamide, and the other to carbama- EEG abnormalities MRI abnormalities Temporal IEDs Mild diffuse cerebral slowing zepine (figure 1). Indeed, sodium-channel female; G-AChR

5 blocking agents such as carbamazepine and phenytoin have been successfully used to treat

1 autoimmune VGK channelopathies such as neuro- 11 12 IV methylprednisolone; LGI1 myotonia and episodic ataxia type 1. These 5 agents act by reducing neuronal repetitive firing AEDs responded to Carbamazepine Temporal IEDs T2 signal hyperintensity Phenytoin lacosamide Lacosamide Temporal Lacosamide Temporal IEDs Lt temporal T2 signal through interaction with voltage-gated sodium channels.11,12 Perhaps, the observed favorable 1 1 response of these agents in VGKC-complex ab– associated AE is attributable to this mechanism. – contactin-associated protein-2; F In our series, one patient with GAD65 ab 5

IV immunoglobulin; IVMP associated AE became seizure free with carbama- AEDs previously tried lamotrigine (rash) Levetiracetam Carbamazepine Multifocal IEDs R Topiramate Oxcarbazepine Temporal IEDs Swollen left amygdala VGKC-complex, LGI1 and Levetiracetam Oxcarbazepine Temporal IEDs Unremarkable VGKC-complex, LGI1 Levetiracetam Lacosamide Temporal IEDs Unremarkable VGKC-complex, CASPR2 Phenytoin (low dose) Levetiracetam Carbamazepine Temporal IEDs Unremarkable GAD65, 404 nm/L IVIG 37 Carbamazepine (rash) Levetiracetam phenytoin (rash)

5 zepine after failing immunotherapy. This is in

1 contrast to a similar report that found lack of AED –

1 response in patients with GAD65 antibody positive AE following which it was postulated cognitive

1 that GAD65 ab–associated AEs are intractable to AEDs.13 Regardless of the type of therapy, 4/9

antiepileptic drug; CASPR2 (44.4%) autoantibody-negative patients become interictal discharge; IVIG 5 cognitive impairment cognitive challenges subacute cognitive memory impairment cognitive impairment personality change complex. 5 – seizure free, 2 with AEDs (one with lacosamide Focal dyscognitive seizures 1 Focal dyscognitive seizures 1 Focal seizures with GTCs cognitive impairment Focal dyscognitive seizures 1 decline Focal dyscognitive seizures 1 Focal dyscognitive seizures 1 Focal seizures impairment Focal dyscognitive seizures 1 dysphasia Refractory focal seizures Levetiracetam

Clinical characteristics of patients who become seizure free with AED therapy and the other with carbamazepine after failing immunotherapy), while the other two responded to immunotherapy. Others have noted the potential for response in patients with seronegative AE F/61 M/58 Sex/age, y Presenting symptoms M/70 M/68 M/68 F/67 M/63 F/63 M/59 Table 2 14 clonic seizure; IED potassium channel Abbreviations: AED to antiepileptic medication and immunotherapy.

Neurology: Neuroimmunology & Neuroinflammation 5 based on the extraction of relevant patient informa- Figure 3 Response to therapy by antibody status tion from EMRs. The lack of control group and the possible confounder of concomitant immunotherapy should also be acknowledged. In addition, selection bias could be present due to the tertiary nature of these patients. Finally, seizure quantification is only based on verbal report, thus recall bias could result in the event of amnestic seizures, and related to concurrent cognitive dysfunction that may have been present in a significant proportion of our cohort. Prospective studies are needed to clarify whether there is an optimal AED selection for AE. As the awareness of AE increases, finding effective strategies for seizure control is crucial.

AUTHOR CONTRIBUTIONS Study concept and design: Dr. Feyissa and Dr. Britton. Acquisition of data: Dr. Feyissa, Dr. Lopez, and Dr. Britton. Drafting of the manuscript: Dr. Feyissa and Dr. Lopez. Critical revision of the manuscript for important intellectual content: Dr. Feyissa and Dr. Britton.

STUDY FUNDING Dr. Feyissa is supported by the Mayo Clinic Neuroscience Focused Research Team Program. The number of patients who responded per the total number of patients receiving individual therapy is provided in each bar. Ab negative 5 negative to antibody testing (n 5 9); AED 5 DISCLOSURE antiepileptic drug; GAD65 5 glutamic acid decarboxylase 65 (n 5 10); VGKC-complex 5 Dr. Feyissa and Dr. Lopez report no disclosures. Dr. Briton received voltage-gated potassium channel–complex (n 5 17). research support from Grifols and Dr. and Mrs. David Hawk. Go to Neurology.org/nn for full disclosure forms. The rationale for selecting one AED over the Received February 7, 2017. Accepted in final form March 28, 2017. other was not clarified in the EMR of our cohort. Levetiracetam was the most common used drug REFERENCES (42/50, figure 2); however, none of the patients 1. Britton JW. Autoimmune epilepsy. Handb Clin Neurol in our series were rendered seizure free by this 2016;133:219–245. medication (0%). This finding is in contrast to 2. Bakpa OD, Reuber M, Irani SR. Antibody-associated epi- lepsies: clinical features, evidence for immunotherapies and previous reports suggesting that levetiracetam future research questions. Seizure 2016;41:26–41. may have, at least in part, anti-inflammatory prop- 3. Beghi E, Shorvon S. Antiepileptic drugs and the immune 3 erties. The reason for the prevalent use of levetir- system. Epilepsia 2011;52(suppl 3):40–44. acetam in our series is likely due to its ease of 4. Bianchi M, Rossoni G, Sacerdote P, Panerai AE, Berti F. dosing and lack of drug-drug interactions.9 Cer- Carbamazepine exerts anti-inflammatory effects in the rat. – tainly, AEDs with enzyme induction properties Eur J Pharmacol 1995;294:71 74. such as carbamazepine and phenytoin could alter 5. Himmerich H, Bartsch S, Hamer H, et al. Modulation of cytokine production by drugs with antiepileptic or mood the pharmacokinetics of immunosuppressive ther- stabilizer properties in anti-CD3- and anti-Cd40- 9 apiesandotheragentsusedinthesepatients. Per- stimulated blood in vitro. Oxid Med Cell Longev 2014; haps, newer AEDs with sodium channel blocking 2014:806162. properties and more favorable pharmacokinetic 6. Gomez CD, Buijs RM, Sitges M. The anti-seizure drugs profiles (such as oxcarbazepine and lacosamide) vinpocetine and carbamazepine, but not valproic acid, could be considered in this patient population. It reduce inflammatory IL-1beta and TNF-alpha expression in rat hippocampus. J Neurochem 2014;130:770–779. is also imperative to monitor for hyponatremia 7. Toledano M, Britton JW, McKeon A, et al. Utility of an (e.g., oxcarbazepine and carbamazepine) and aller- immunotherapy trial in evaluating patients with presumed gic cutaneous reactions (e.g., carbamazepine and autoimmune epilepsy. Neurology 2014;82:1578–1586. phenytoin), as these have been shown to occur at 8. Li G, Bauer S, Nowak M, et al. Cytokines and epilepsy. high rate in patients with VGKC-complex anti- Seizure 2011;20:249–256. body.15 Prospective studies are needed to clarify 9. French JA, Gazzola DM. Antiepileptic drug treatment: new drugs and new strategies. Continuum (Minneap AED selection and to elucidate their immunomod- Minn) 2013;19:643–655. ulatory properties in AE. 10. Tan KM, Lennon VA, Klein CJ, et al. Clinical spectrum There are several limitations to our study. First, of voltage-gated potassium channel autoimmunity. data collection for this retrospective study was Neurology 2008;70:1883–1890.

6 Neurology: Neuroimmunology & Neuroinflammation 11. Ahmed A, Simmons Z. Isaacs syndrome: a review. Muscle 14. von Rhein B, Wagner J, Widman G, Malter MP, Elger CE, Nerve 2015;52:5–12. Helmstaedter C. Suspected antibody negative autoimmune 12. Rajakulendran S, Schorge S, Kullmann DM, Hanna MG. limbic encephalitis: outcome of immunotherapy. Acta Episodic ataxia type 1: a neuronal potassium channelop- Neurol Scand 2017;135:134–141. athy. Neurotherapeutics 2007;4:258–266. 15. Irani SR, Stagg CJ, Schott JM, et al. Faciobrachial dystonic 13. Malter MP, Helmstaedter C, Urbach H, Vincent A, Bien CG. seizures: the influence of immunotherapy on seizure con- Antibodies to glutamic acid decarboxylase define a form of trol and prevention of cognitive impairment in a broaden- limbic encephalitis. Ann Neurol 2010;67:470–478. ing phenotype. Brain 2013;136(pt 10):3151–3162.

Neurology: Neuroimmunology & Neuroinflammation 7 Elsberg syndrome A rarely recognized cause of cauda equina syndrome and lower thoracic myelitis

Filippo Savoldi, MD ABSTRACT Timothy J. Kaufmann, Objective: Elsberg syndrome (ES) is an established but often unrecognized cause of acute lumbo- MD sacral radiculitis with myelitis related to recent herpes virus infection. We defined ES, determined Eoin P. Flanagan, its frequency in patients with cauda equina syndrome (CES) with myelitis, and evaluated its clin- MBBCh ical, radiologic, and microbiologic features and outcomes. Michel Toledano, MD Methods: We searched the Mayo Clinic medical records for ES and subsequently for combinations Brian G. Weinshenker, of index terms to identify patients with suspected CES and myelitis. MD Results: Our search yielded 30 patients, 2 diagnosed with ES and an additional 28 with clinical or radiologic evidence of CES retrospectively suspected of having ES. We classified patients in 5

Correspondence to groups according to diagnostic certainty. MRI and EMG confirmed that 2 had only myelitis, 5 only Dr. Weinshenker: radiculitis, and 16 both. Two had preceding sacral herpes infection and 1 oral herpes simplex. [email protected] Spinal cord lesions were commonly multiple, discontinuous, not expansile, and centrally or ventrally positioned. Lesions generally spared the distal conus. Nerve root enhancement was occasionally prominent and was smooth rather than nodular. Lymphocytic CSF pleocytosis was common. Thirteen patients (43%) had viral isolation studies, which were commonly delayed; the delay may have accounted for the low rate of viral detection. Acyclovir was administered to 6 patients. Most patients recovered with sequelae; 1 patient experienced encephalomyelitis and died. Conclusion: ES is a definable condition likely responsible for 10% of patients with combined CES and myelitis. Radiologic findings are not entirely specific but may help in differentiating ES from some competing diagnostic considerations. We propose criteria to facilitate diagnosis. Neurol Neuroimmunol Neuroinflamm 2017;4:e355; doi: 10.1212/NXI.0000000000000355

GLOSSARY CES 5 cauda equina syndrome; ES 5 Elsberg syndrome; HSV2 5 herpes simplex virus 2; ICD 5 International Classification of Disease; NMO 5 neuromyelitis optica; VZV 5 varicella zoster virus.

Elsberg syndrome (ES) is a presumed infectious syndrome consisting of acute or subacute bilateral lumbosacral radiculitis, often accompanied by myelitis confined to the lower spinal cord, and is frequently a manifestation of reactivation, or occasionally, primary herpes simplex virus 2 (HSV2) infection.1 Clinicians often do not consider ES in the differential diagnosis of acute cauda equina syndrome (CES) and do not perform appropriate testing for HSV infection in a timely way to facilitate definitive diagnosis. However, especially when accompanied by clinical or radiologic findings of myelitis, ES is not a rare cause of CES. ES has been the subject of case reports and small series,2–7 the largest of which included 17 patients.8 A formal case definition that defines the minimum criteria to diagnose this syndrome and identifies exclusionary findings that suggest alternative diagnoses does not exist. Sarcoidosis, lymphoma, other neoplasms, and vascular disorders (dural arteriovenous fistula) may mimic ES and are in the differential diagnosis.

From the Departments of Neurology (F.S., E.P.F., M.T., B.G.W.), and Neuroradiology (T.J.K.), Mayo Clinic, Rochester, MN; and Neuroimaging Research Unit (F.S.), Department of Neurology, San Raffaele Scientific Institute and University, Milan, Italy. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 We conducted a retrospective review of only 4 patients. To find unrecognized cases, we searched again for lumbosacral myeloradiculitis evaluated at a combination of key terms common to those of the 4 ES cases in patients evaluated in the neurology, urology, gynecology, or theMayoClinictoassessthefrequencyof infectious disease departments who had both myelitis and radi- ES, determined retrospectively as well as culitis (figure 1). After testing several possible combinations of prospectively. Among cases who did not terms, we selected patients identified by at least 1 of 4 searches of have other established diagnoses and who the medical records, components of which are highlighted in flow chart in figure 1, and thereby identified 1,035 patients. We met certain minimum criteria, we reviewed excluded patients with an International Classification of Disease Mayo clinical records to classify the level (ICD) diagnosis of MS or neuromyelitis optica (NMO), reducing of diagnostic certainty using a scheme we the cohort to 837 subjects. Exclusion of cases without terms referring to urinary retention or sacral sensory symptoms further derived based on a review of the medical liter- reduced the cohort to 337 subjects. A review of individual clinical ature (table 1). We identified clinical, demo- records identified 213 patients with other definitive alternative graphic, and radiologic findings of ES, diagnoses (table 2). Subsequently, we conducted a detailed review concentrating on distinction from other causes of clinical records to confirm the presence of urinary retention or other sacral symptoms (saddle anesthesia, paresthesias, neuralgic of lower thoracic/conus myelitis. pain, constipation, impotence, urinary incontinence, and fecal incontinence) yielding 49 patients who were strongly suspected METHODS We performed a retrospective review of Mayo to have ES (figure 1). We abstracted their demographic character- electronic medical records from 2000 to March 2016 interrogat- istics, neurologic examination findings, CSF, viral, EMG, and “ ” ing for the term Elsberg in the history, examination, review of MRI results from clinical records. The Mayo Clinic Institutional systems, impression, and final diagnosis and thereby identified Review Board granted approval to review retrospectively obtained patient clinical information. We defined diagnostic criteria for ES with 5 levels of diagnos- Table 1 Elsberg syndrome according to diagnostic certainty tic certainty based on the review of the medical literature: laboratory-supported definite, clinically definite, clinically proba- No. of ble, clinically possible, and excluded (table 1). Radiculitis, defined Categories Criteria patients either clinically (table 1, A1) or by MRI and/or electrophysiologic

1. Laboratory-supported (A1 OR A2) AND B5 3 evidence (table 1, A2), was required to establish a diagnosis of ES. definite A neurologist (B.G.W.) categorized each case according to these

2. Clinically definite A1 OR A2; B1 AND two of B2–B4; B1 and B2 (if 9 criteria and with special attention to potential alternative diagno- concomitant) ses. A neuroradiologist (T.K.) reviewed abnormal radiology to

3. Clinically probable A1 OR A2; B1 AND one of B2–B4 10 confirm and further characterize the abnormalities in each case and to consider whether there were unique or characteristic radio- 4. Clinically possible A1 OR A2; one of B1–B4 8 logic features of ES after the review of the entire cohort. 5. Excluded Neither of A1 nor of A2; any of D1–D3 19 We report the frequencies of key clinical and radiologic fea-

A. Required tures and summarize distinctive radiologic features.

A1. Clinical symptoms and signs of cauda equina involvement: urinary hesitancy or retention; bowel incontinence, or severe constipation (erectile dysfunction insufficient on its own) (n 5 RESULTS Forty-nine patients satisfied the inclusion 39, 80%) and exclusion criteria. Clinical and MRI findings are A2. MRI or electrophysiologic evidence of cauda equina involvement: enhancement of cauda equina; EMG evidence of radiculopathy (n 5 20, 41%) summarized in table 3 and the clinical course in

B. Supportive but not required table 4. Three patients met criteria for laboratory- supported definite, 9 for clinically definite, 10 for B1. Time course: acute/subacute onset; no relapse; progression over less than 3 mo (n 5 42, 86%) clinically probable, and 8 for clinically possible ES.

B2. Coexisting or recently preceding symptoms of genital herpes infection OR other clinical Nineteen patients were excluded, 16 (84%) because symptoms of herpes virus infection (n 5 5, 10%) of exclusionary criteria (table 1) and 3 (16%) because B3. Clinical (e.g., exaggerated reflexes and Babinski signs) or MRI evidence of myelitis in conus they failed to satisfy any of the established levels of (n 5 28, 57%) diagnostic certainty (table 1), including 2 of the 4 B4. CSF pleocytosis (n 5 23, 47%) diagnosed with ES at the time of their clinical eval- B5. Documented herpes virus infection from CSF by PCR, culture, or detection of IgM serology uation but with insufficient information to confirm (n 5 5, 10%) the diagnosis. Based on MRI and EMG, available in C. Red flags 29 and 21 cases respectively, 2 patients had only C1. Relapses beyond 1 y from onset (n 5 4, 8%) myelitis, 5 only radiculitis, and 16 both. Only 2 D. Exclusionary included patients had been diagnosed with ES at the

D1. Myelitis extending rostral to T9 (n 5 12, 25%) time of their clinical evaluation.

D2. Other neurologic symptoms suggestive of alternative etiology: optic neuritis, brain/ Of the 30 patients with a diagnosis of lab-sup- brainstem syndrome (n 5 3; 6%) ported, definite, probable, or possible ES, 24 (80%) D3. Other etiology proven/more likely for syndrome: NMOSD, dural arteriovenous fistula, viral were men and 6 women (20%) and their median 5 transverse myelitis, other causes of myelopathy (n 11, 22%) age was 53 years. Five patients (17%) experienced Abbreviations: IgM 5 immunoglobulin M; NMOSD 5 neuromyelitis optica spectrum disorder. prodromal symptoms, including fever, headache,

2 Neurology: Neuroimmunology & Neuroinflammation Figure 1 Flow chart of patient selection

Interrogation of the electronic database for specific key terms highlighted in the first box returned a list of 1,035 patients. We excluded patients with an ICD code identifying NMO and MS, yielding 837 patients. Further interrogation of the database for patients with the following key terms returned 337 patients: “urinary retention” OR “paresthesia” OR “neuralgic pain” OR “constipation” OR “impotence” OR “perineal” OR “retention” OR “anus” OR “anal” OR “saddle”.After the review of individual clinical records of the remaining patients, we excluded 213 subjects who had likely or established alternative diagnoses, as listed in the figure, and only patients who were confirmed to have urinary retention and other sacral sensory symptoms were retained. Of the remaining 49 patients, 19 did not satisfy any of the levels of diagnostic certainty proposed in table 1. The remaining 30 were assigned a specific diagnostic category according to the level of suspicion for ES. ES 5 Elsberg syndrome; ICD 5 International Classification of Disease;NMO5 neuromyelitis optica. photophobia, diarrhea, myalgia, and general malaise. required indwelling catheterization in 21 (70%) pa- Sacral herpes infection (zoster, n 5 1; simplex, n 5 1) tients. Saddle anesthesia (n 5 15, 50%) and consti- immediately preceded clinical presentation in 2 pa- pation (n 5 13, 43%) were common. Fifteen (50%) tients and oral herpes simplex in 1 patient. Urinary patients had weakness of the lower limbs worse than retention was present in 23 (77%) patients and 3/5 on the MRC scale in 7 patients (23%). Lower

Neurology: Neuroimmunology & Neuroinflammation 3 (60%) and cord gadolinium enhancement in 13 pa- Table 2 Alternative diagnoses established or likely (213 patients) tients (43%). Sixteen patients (53%) had MRI evidence of concomitant myelitis and radiculitis (figure Condition No. 2). Six patients (20%) had relatively minor MRI Spinal stenosis (intraspinal or extraspinal tumors, 50 abnormalities within the spinal cord that extended disk herniation, bone compression, trauma, – ankylosing spondylitis, tuberculosis, osteomyelitis, rostral to T9 T10; in patients whose dominant cord or spinal abscess) lesion was caudal to T9, minor exceptions were al- Vascular myelopathy (dural arteriovenous fistula, 41 lowed to the rule of excluding those with spinal cord infarction) abnormalities extending rostral to T9. Spinal cord Nonneurological 21 lesions were multiple and discontinuous in 9 cases Peripheral nerve disorders: Guillain-Barré 19 (30%) and were centrally or ventrally positioned syndrome, chronic inflammatory demyelinating polyneuropathy, or others within the cord. In 2 patients, the lesions were mildly

MS or clinically isolated demyelinating syndrome 16 expansile. In 3 cases (10%), the lesion was only present in the lower thoracic cord, not in the conus. Myelitis secondary to systemic autoimmune 14 diseases (sarcoidosis, SLE, Sjögren, Behçet, and Nerve root enhancement was always smooth rather systemic sclerosis) than nodular, occasionally diffuse, and sometimes NMO 11 limited to ventral or dorsal roots. Thirteen patients Infectious myelitis: viral (not herpes!), bacterial, 10 (43%) had MRI evidence of nerve root thickening. fungal, or parasitic. Nerve root enhancement often occurred contempo- Infectious brain infection (encephalitis or 4 meningitis) or autoimmune CNS condition (neither raneously with spinal cord involvement; evaluation of MS nor NMO) MRI at multiple time points, when available, showed Neurodegenerative brain condition or movement 4 that in 3 cases (10%) root enhancement occurred at disorder a later point of the illness as cord lesions were Myofascial pain syndrome or fibromyalgia 4 resolving. Paraneoplastic myelitis 3 The median CSF white blood cell count in the Primary bladder pathology 3 Elsberg group was 9; 15 (50%) patients had more m Motor neuron disease 3 than 5 leukocytes/ L; the average percentage of lym-

Brain tumor or cord tumor (symptoms not due to 3 phocytes was 69%; 2 patients had an average 65% compression) mononuclear predominance. CSF proteins were con- Radiation myelopathy 2 sistently elevated (mean 143 6 131 mg/dL), whereas

Brain vascular event 2 oligoclonal bands were rarely detected (table 3). ES was considered in the differential diagnosis of 3 Metabolic myelopathy (vitamins B12, D, E, and 1 copper deficiencies) (10%) patients. Acyclovir was administered in 6

Delirium 1 (20%) patients and corticosteroids in 15 (50%) pa-

Restless leg syndrome 1 tients, typically methylprednisolone 1 g per day over 3–5 days. Follow-up data were available for 13 (43%) 5 5 Abbreviations: NMO neuromyelitis optica; SLE sys- patients (table 4). One patient (3%) died with temic lupus erythematosis. encephalomyelitis and cardiac arrest; 3 patients (10%) had no neurologic recovery; 8 moderate recov- limb tendon reflexes were pathologically exaggerated ery (27%); and 1 complete recovery (3%). Only 2 in 10 patients (33%) and absent in 10 (33%); 4 patients (7%) relapsed, one at 33 and the other at 82 (13%) had unequivocal and 6 (20%) patients had months from onset. equivocal Babinski signs. Eighteen patients (60%) were tested for viral DISCUSSION ES is poorly defined and rarely reported, infection, in particular for HSV (n 5 16, 53%), var- but it is probably frequently unrecognized. Based on the icella zoster virus (VZV) (n 5 15, 50%), cytomega- historical literature and our clinical experience, we for- lovirus (n 5 15, 50%), Epstein-Barr virus (n 5 17, mulated objective criteria for ES with 4 different levels 57%), and HIV (n 5 17, 57%). PCR confirmed viral of certainty (table 1). Radiculitis, manifest either clini- infection in 3 patients (10%); in 2 patients, VZV was cally (table 1, A1), radiologically or electrophysiologi- demonstrated in the CSF and, in 1 patient, HSV2 cally (table 1, A2) is required. A secure diagnosis is from a genital swab. Three patients were tested for possible when evidence for infection by a causative virus Lyme disease, all with negative results. Radiculitis was is demonstrated and alternative causes excluded, but confirmed in 21 patients (70%) by MRI with gado- given the sensitivity of the serologic and virologic stud- linium (n 5 17, 57%) or EMG (n 5 12, 40%). MRI ies9 and frequent delay in obtaining appropriate samples, showed lumbar or lower thoracic cord T2- clinical and radiologic evidence may suggest ES regard- hyperintense signal abnormality in 18 patients less of the microbiologic information. For this reason,

4 Neurology: Neuroimmunology & Neuroinflammation The onset of ES was consistently acute in our Table 3 Key clinical and radiologic features cohort; insidious onset and evolution suggest alterna- Men/women 24/6 tive explanations such as sarcoidosis or neoplasm.

Median age at onset, y (range) 52.9 (10.7–82.7) Suspected cauda equina involvement, which leads to

Herpes zoster immediately before episode, n (%) 0 (0.0) urinary retention, is confirmed with MRI and/or EMG. Concomitant myelitis detected radiologically Sacral herpes infection immediately before episode, n (%) 2 (6.7) or clinically further supports the diagnosis. Myelitis, Herpes labialis immediately before episode, n (%) 1 (3.3) associated with many inflammatory CNS conditions Presence of prodromal symptoms, n (%)a 5 (16.7) such as MS, NMO, anti-MOG antibody myelitis, Urinary retention, n (%) 23 (76.7) sarcoidosis, and others, is not usually associated with

Catheter required for retention, n (%) 21 (70.0) radiculitis, whereas radiculitis is a required feature for

Urinary incontinence, n (%) 7 (23.3) the diagnosis of ES. We did not identify pathognomonic or specific Bowel incontinence (%) 3 (10.0) neuroradiologic findings. Nevertheless, some features Saddle anesthesia, n (%) 15 (50.0) were typical and could help distinguish ES from other Constipation, n (%) 13 (43.3) disorders. Spinal cord lesions were commonly multi- Exaggerated/pathologic lower limb tendon reflexes, n (%) 10 (30.0) ple and discontinuous, usually not expansile, and cen- Absent lower limb tendon reflexes, n (%) 10 (33.3) trally or ventrally positioned in the cord. Unlike the

Loss of limb sensation, n (%) 24 (80.0) imaging usual pattern of dural arteriovenous fistula, which can also cause both upper and lower motor Leg weakness, n (%) 15 (50.0) neuron problems, the cord lesions in ES are not Leg weakness <3/5 (MRC scale), n (%) 7 (23.3) consistently located in the most caudal portion of Extensor plantar response, n (%) 4 (13.3) the conus. Dural arteriovenous fistula lesions, in con- Immunosuppression, n (%) 1 (3.3) tradistinction to ES, are usually continuous, extend- Patients tested for virus (either PCR or serology from CSF, 18 (60.0) ing rostral from the lowermost portion of the serum, or genital swab), n (%) conus, involve the most of the thickness of the cord, PCR performed, n (%) 17 (56.7) and are commonly associated with swelling. Nerve Patients with positive PCR, n 3 (10.0) root enhancement was sometimes diffuse and some- Median delay in CSF sample collection, d (range) 32 (1–443) times limited to ventral or dorsal roots. It was always

CSF obtained, n (%) 25 (83.3) smooth and continuous and never nodular or plaque like as can be seen with neoplastic nerve root lesions.10 Median cell count in CSF/mL (range) 9(0–1,057) Nerve root enhancement often occurred contempo- Patients with >5 nucleated cells/mL in CSF, n (%) 15 (50.0) raneously with spinal cord involvement, although > m Mean percentage of lymphocytes in those with 5 WBC/ L (SD) 69 (24.0) either nerve root enhancement or spinal cord paren- Patients with CSF pleocytosis and >5% neutrophils, n (%) 3 (20.0) chymal signal abnormality could predominate; in Median number of oligoclonal bands (range) 0(0–7) some cases, root enhancement occurred at a later

Mean CSF protein mg/dL (SD) 143 (131) point of the illness as cord lesions were resolving.

EMG evidence of radiculopathy, n (%) 12 (40.0) Finally, plaque or subpial cord enhancement, typical of granulomatous conditions,11 did not occur in pa- MRI T2 cord hyperintensity in conus, n (%) 15 (50.0) tients with ES. MRI T2 cord hyperintensity in lower thoracic cord (below T10), n (%) 18 (60.0) CSF examination demonstrated lymphocytic MRI gadolinium enhancement (from T10), n (%) 13 (43.3) pleocytosis in 15 patients (50%), and in 3 of those MRI nerve root enhancement, n (%) 17 (56.7) (20%), neutrophils exceeded 5%. Root thickening in cauda, n (%) 13 (43.3) Although HSV infection is widely believed to be

Pial/subpial enhancement, n (%) 3 (10.0) responsible for a majority of cases, almost half of the cases (13/30) were never investigated for the pres- Thoracic or cervical MRI cord abnormal, n (%) 6 (20.0) ence of a viral infection even though lumbar puncture Abbreviation: WBC 5 white blood cell. was performed, indicative of the low level of diagnos- a Prodromal symptoms: fever; headache; myalgia; malaise; diarrhea; photophobia; tic suspicion for an infectious etiology for this condi- lymphadenopathy. tion. Furthermore, when PCR was performed, only 2 of the 30 (7%) were positive in the CSF and 1 from we established criteria for clinically definite, probable, a genital swab. Despite CSF viral detection being con- and possible ES (see supportive feature in table 1, B1– sidered the diagnostic gold standard, its diagnostic B5). In addition, we identified red flags that are cau- accuracy is imperfect in real-life settings,9,12 with tionary but not exclusionary as well as exclusionary cri- a negative predictive value of 82% and a positive pre- teria (table 1, C and D). dictive value of 54% in one study.9 Therefore,

Neurology: Neuroimmunology & Neuroinflammation 5 Table 4 Summary of treatments and clinical course

Treatment with Treatment with acyclovir corticosteroids Both Neither

Patients, n (%) 4 (13.3) 13 (43.3) 2 (6.7) 11 (36.7)

Patients with follow-up, n (%) 3 (10.0) 7 (23.3) 2 (6.7) 1 (3.3)

Patients with >6 mo follow-up, n (%) 2 (6.7) 7 (23.3) 2 (6.7) 1 (3.3)

Median follow-up (range), mo 23 (1–81) 92 (21–385) 101 (17–185) 261

Fatal outcome, n (%) 0 1 (3.3) 0 0

No recovery, n (%) 0 1 (3.3) 1 (3.3) 1 (3.3)

Partial recovery, n (%) 3 (10.0) 3 (10.0) 1 (3.3) 1 (3.3)

Complete recovery at follow-up, n (%) 0 1 (3.3) 0 0

Patients with relapses at follow-up, n (%) 0 2 (6.7) 0 0

a negative result cannot absolutely exclude the pres- first patient, who was immunocompetent, had a few ence of infection. Furthermore, rapid viral clearance discrete lesions higher in the cord and also had from the CSF has been shown for HSV and VZV encephalitic manifestations; the second patient, who infections.13 The highest yield of viral detection by was HIV positive and immunocompromised, had PCR occurred in samples obtained 3–14 days after a longitudinally extensive lesion spanning up to 9 symptom onset.9 In our cohort, samples were vertebral segments of the thoracic spine and had evi- acquired outside this interval in 14 of the 18 cases dence of meningeal irritation and necrotic cord dam- (78%); the only 2 positive CSF PCR samples were age. Ascending necrotic myelitis has been described in obtained at days 11 and 37. Of the 3 patients who the setting of HSV2.16 Although the patients with experienced an HSV infection immediately preceding VZV meet the criteria for ES, encephalitic symptoms ES, only 1 (33%) was tested within 14 days and 1 at as well as long thoracic and cervical cord MRI lesions day 15; none were positive for HSV infection in the and meningeal irritation may be considered red flags, CSF and 1 tested positive only from a genital swab. and caution in the diagnostic process should be main- Reported experience with CSF PCR in patients tained before providing a final diagnosis. with ES is mostly limited to case reports and some- HSV and VZV are treatable when acyclovir is ini- times other viral detection methods, such as serology tiated early, but the potential benefits of treatments or culture, have been used.5 Antibody studies may be with antiviral drugs have not been documented. In particularly useful when CSF is obtained late (e.g., our cohort, none of the 5 patients with follow-up .14 days) in the course of the disease. Demonstra- who received acyclovir experienced a fatal outcome. tion of intrathecal production of HSV- or VZV- Yet, one patient treated with both acyclovir and cor- specific immunoglobulin G (IgG) antibodies can be ticosteroids had no clinical recovery as did one patient diagnostic.14,15 Reliance on an antibody index com- who received corticosteroid treatment alone and paring virus-specific IgG levels in CSF and serum another who received neither (table 4). Therefore, with correction for blood-brain barrier leak using superiority of acyclovir treatment cannot be proven. either the CSF/serum albumin ratio or the ratio of Nonetheless, acyclovir treatment was started only in 6 CSF/serum antibody titers of a nonpathogenic virus patients, whereas high-dose corticosteroids were ensures that serum HSV or VZV antibodies have not administered as primary treatment in 15 patients, reached the CSF by means of a sink mechanism. illustrating the low index of suspicion for this condi- Although rarely demonstrated, the presence of tion and the tendency to treat as an idiopathic inflam- virus-specific immunoglobulin M (IgM) in the CSF matory condition. is strongly indicative of acute or active chronic CNS The study is limited by its retrospective nature. infection as these antibodies do not tend to cross the Because of the referral pattern of patients to our insti- blood-brain barrier. tution, patients were evaluated initially at outside in- The 2 patients with a positive CSF PCR were pos- stitutions and as this study was retrospective, itive for VZV. VZV may be considered an etiologic investigations were not standardized. Potentially, we agent of ES,1 but is also recognized to be a cause of did not capture all cases by our electronic search strat- myelitis and/or myeloradiculitis occurring elsewhere egy. The proportion of laboratory-supported definite in the spinal cord. The 2 patients with herpes zoster cases and of patients with proven HSV infection was detection experienced more widespread involvement low. This might reflect low rates of awareness of this of the spinal cord than did other Elsberg cases. The syndrome and consequent low proportion of cases

6 Neurology: Neuroimmunology & Neuroinflammation with acute and convalescent serology results. Limita- Figure 2 MRI evidence of myelitis and radiculitis in 6 patients tions of sensitivity of viral isolation methods particularly when samples are obtained late in the course may be an additional contributor. Perhaps, next- generation sequencing may enhance diagnostic sensitivity as already shown with other infections.17 We recognize that the scientific literature is mov- ing away from eponymous definitions. Our intent was not to entrench an eponymous definition, but to formalize diagnostic criteria that are flexible to aid in the identification of patients with combinations of radiculitis and caudal myelitis who have potentially treatable viral infections after considering clinical and radiologic clues that identify other potential etiologies for this syndrome. Improved testing and virus identification methodologies may establish the role of antiviral treatment. ES is a definable condition that may be responsible for approximately 5%–15% of patients with CES and myelitis based on our review of cases seen at the Mayo Clinic. It is usually self-limited but commonly leaves some degree of permanent neurologic deficit and in one instance led to a fatal outcome. Although radiologic findings are not specific, they can help to rule out competing diagnoses, such as dural arteriovenous fistula, sarcoidosis, and other conditions. When both enhancement of cauda equina and signal abnormality of the distal spinal cord occur acutely, Elsberg may be a common cause of the syndrome. The diagnostic criteria that we formulated may bring attention to this condition and permit a more precise diagnosis when integrated with tests to exclude other mimics. Detection of viral infection is useful, but sensitivity appears to be low. Treatment with acyclovir is appropriate even in the absence of demonstration of viral infection given the favorable risk-benefit profile of this drug.

AUTHOR CONTRIBUTIONS Brian G. Weinshenker conceived the idea for the manuscript, reviewed and interpreted source data, and oversaw all aspects of the project. Filippo Savoldi collected and interpreted data and drafted the manuscript. Timothy J. Kaufmann reviewed MRI scans of all patients suspected to have Elsberg syndrome. All authors revised the manuscript critically for important intellectual content and provided final approval of the version to be published.

STUDY FUNDING No targeted funding reported.

DISCLOSURE Patient A presents with concomitant presence of multiple and discontinuous T2 hyperin- F. Savoldi reports no disclosures. T.J. Kaufmann served on the editorial tense lesions (A.a) that enhance on T1-weighted images after gadolinium injection (A.b) con- board for American Journal of Neuroradiology and Neuro-Oncology; con- comitantly with nerve roots of the cauda equina (A.b and A.c). Nerve root enhancement is sulted for SpinThera; and received research support from the National prominent in 2 other patients (B.a, B.b, C.a, and C.b), with greater nerve root thickening in Cancer Institute. E.P. Flanagan and M. Toledano report no disclosures. the latter (C.c). Two other patients had multifocal, discontinuous, T2-hyperintense lesions B.G. Weinshenker served on the data safety monitoring board for No- (D.a and E.a) as well as enhancement (D.b and E.b) in both the conus and lower thoracic cord. vartis, Biogen, and Mitsubishi; served on the editorial board for Canadian Finally, cord T2-hyperintense abnormality (F.a) may precede the onset of nerve root Journal of Neurological Sciences, Turkish Journal of Neurology, and enhancement (F.b) by 26 days. Neurology®; holds a patent for and receives royalties from NMO-IgG for the diagnosis of neuromyelitis optica; consulted for Caladrius; and

Neurology: Neuroimmunology & Neuroinflammation 7 served as an adjudication committee member for MedImmune and 9. Davies NW, Brown LJ, Gonde J, et al. Factors influencing Alexion. Go to Neurology.org/nn for full disclosure forms. PCR detection of viruses in cerebrospinal fluid of patients with suspected CNS infections. J Neurol Neurosurg Psy- Received March 5, 2017. Accepted in final form April 5, 2017. chiatry 2005;76:82–87. 10. Smith JK, Lury K, Castillo M. Imaging of spinal and REFERENCES spinal cord tumors. Semin Roentgenol 2006;41:274–293. 1. Eberhardt O, Kuker W, Dichgans J, Weller M. HSV-2 11. Smith JK, Matheus MG, Castillo M. Imaging manifesta- sacral radiculitis (Elsberg syndrome). Neurology 2004;63: tions of neurosarcoidosis. AJR Am J Roentgenol 2004; 758–759. 182:289–295. 2. Caplan LR, Kleeman FJ, Berg S. Urinary retention prob- 12. Jeffery KJM, Read SJ, Peto TEA, Mayon-White RT, ably secondary to herpes genitalis. N Engl J Med 1977; Bangham CRM. Diagnosis of viral infections of the central 297:920–921. nervous system: clinical interpretation of PCR results. Lan- 3. Hemrika DJ, Schutte MF, Bleker OP. Elsberg syndrome: cet 1997;349:313–317. a neurologic basis for acute urinary retention in patients 13. Kleines M, Scheithauer S, Schiefer J, Hausler M. Clinical with genital herpes. Obstet Gynecol 1986;68:37S–39S. application of viral cerebrospinal fluid PCR testing for 4. Nakajima H, Furutama D, Kimura F, et al. Herpes sim- diagnosis of central nervous system disorders: a retrospec- plex virus myelitis: clinical manifestations and diagnosis by tive 11-year experience. Diagn Microbiol Infect Dis 2014; the polymerase chain reaction method. Eur Neurol 1998; 80:207–215. 39:163–167. 14. Gobbi C, Tosi C, Städler C, Merenda C, Bernasconi E. 5. Suarez-Calvet M, Rojas-Garcia R, Querol L, Sarmiento Recurrent myelitis associated with herpes simplex virus LM, Domingo P. Polyradiculoneuropathy associated to type 2. Eur Neurol 2001;46:215–218. human herpesvirus 2 in an HIV-1-infected patient 15. Takahashi T, Tamura M, Miki K, et al. Varicella zoster (Elsberg syndrome): case report and literature review. virus myelitis in two elderly patients: diagnostic value of Sex Transm Dis 2010;37:123–125. nested polymerase chain reaction assay and antibody index 6. Aurelius E, Forsgren M, Gille E, Skoldenberg B. Neuro- for cerebrospinal fluid specimens. Case Rep Neurol 2013; logic morbidity after herpes simplex virus type 2 menin- 5:81–90. gitis: a retrospective study of 40 patients. Scand J Infect 16. Wiley CA, VanPatten PD, Carpenter PM, Powell HC, Dis 2002;34:278–283. Thal LJ. Acute ascending necrotizing myelopathy caused 7. Yoritaka A, Ohta K, Kishida S. Herpetic lumbosacral rad- by herpes simplex virus type 2. Neurology 1987;37:1791– iculoneuropathy in patients with human immunodefi- 1794. ciency virus infection. Eur Neurol 2005;53:179–181. 17. Wilson MR, Naccache SN, Samayoa E, et al. Actionable 8. Oates JK, Greenhouse PR. Retention of urine in anogen- diagnosis of neuroleptospirosis by next-generation ital herpetic infection. Lancet 1978;1:691–692. sequencing. N Engl J Med 2014;370:2408–2417.

8 Neurology: Neuroimmunology & Neuroinflammation Comparative utility of disability progression measures in PPMS Analysis of the PROMiSe data set

Marcus W. Koch, MD, ABSTRACT PhD Objective: To assess the comparative utility of disability progression measures in primary pro- Gary R. Cutter, PhD gressive MS (PPMS) using the PROMiSe trial data set. Gavin Giovannoni, MD, Methods: Data for patients randomized to placebo (n 5 316) in the PROMiSe trial were included in PhD this analysis. Disability was assessed using change in single (Expanded Disability Status Scale Bernard M.J. Uitdehaag, [EDSS], timed 25-foot walk [T25FW], and 9-hole peg test [9HPT]) and composite disability meas- MD, PhD ures (EDSS/T25FW, EDSS/9HPT, and EDSS/T25FW/9HPT). Cumulative and cross-sectional Jerry S. Wolinsky, MD unconfirmed disability progression (UDP) and confirmed disability progression (CDP; sustained Mat D. Davis, PhD for 3 months) rates were assessed at 12 and 24 months. Joshua R. Steinerman, MD Results: CDP rates defined by a $20% increase in T25FW were higher than those defined by Volker Knappertz, MD EDSS score at 12 and 24 months. CDP rates defined by T25FW or EDSS score were higher than those defined by 9HPT score. The 3-part composite measure was associated with more CDP events (41.4% and 63.9% of patients at 12 and 24 months, respectively) than the 2-part

Correspondence to measure (EDSS/T25FW [38.5% and 59.5%, respectively]) and any single measure. Cumulative Dr. Koch: UDP and CDP rates were higher than cross-sectional rates. [email protected] Conclusions: The T25FW or composite measures of disability may be more sensitive to disability progression in patients with PPMS and should be considered as the primary endpoint for future studies of new therapies. CDP may be the preferred measure in classic randomized controlled trials in which cumulative disability progression rates are evaluated; UDP may be feasible for cross-sectional studies. Neurol Neuroimmunol Neuroinflamm 2017;4:e358; doi: 10.1212/ NXI.0000000000000358

GLOSSARY 9HPT 5 9-hole peg test; CDP 5 confirmed disability progression; DMT 5 disease-modifying therapy; EDSS 5 Expanded Disability Status Scale; GA 5 glatiramer acetate; PPMS 5 primary progressive MS; T25FW 5 timed 25-foot walk; UDP 5 unconfirmed disability progression.

Patients with primary progressive MS (PPMS) represent 10%–15% of patients with MS and suffer the highest neurodegeneration-related disability.1–4 There are currently no approved treatments to ameliorate disease progression in PPMS, and there are several barriers to developing therapeutics in progressive MS.2,5 Specifically, individual disability measures may lack the sensitivity required to reveal all cases of disease progression within the defined time frame of a clinical trial.6 The Expanded Disability Status Scale (EDSS) remains the typical outcome measure in PPMS trials. However, it has been widely recognized as inadequate7 and is associated with several important weaknesses.3 Other widely used measures of disability in MS include the timed 25- foot walk (T25FW) and the 9-hole peg test (9HPT).8,9 Recent studies have suggested that Supplemental data at Neurology.org/nn From the Departments of Clinical Neurosciences and Community Health Sciences (M.W.K.), University of Calgary, Alberta, Canada; University of Alabama at Birmingham (G.R.C.); Barts and The London School of Medicine and Dentistry (G.G.), London, UK; Vrije Universiteit University Medical Center (B.M.J.U.), Amsterdam, The Netherlands; McGovern Medical School (J.S.W.), Department of Neurology, University of Texas Health Science Center at Houston; Teva Pharmaceutical Industries (M.D.D., J.R.S., V.K.), Frazer, PA; and Heinrich-Heine Universität Düsseldorf (V.K.), Germany. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology 1 the results of an analysis of the PROMiSe data Table 1 Baseline demographic and clinical characteristics of patients who took set undertaken to better understand the utility part in the PROMiSe trial and were of the different measures of disability progres- randomized to receive placebo sion—EDSS, T25FW, and 9HPT alone and Characteristic Placebo cohort (n 5 314) in combinations—and to inform the design of

Male, n (%) 163 (51.9) future trials in PPMS.

White, n (%) 284 (90.5) METHODS Standard protocol approvals, registrations, Mean age, y (range) 50.2 6 8.1 (23–66) and patient consents. Participating study sites in each individ- Mean time from first 10.7 (7.7) ual country locally ensured all necessary regulatory approvals symptom, y (SD) (e.g., institutional review boards/institutional ethics committees) Mean time from first 5.1 (5.5) in accordance with local regulations, including local data protec- diagnosis, y (SD) tion regulations used in the original collection of the data for this Mean EDSS score (SD) 4.9 (1.2) secondary data analysis. Mean T25FW, s (SD) 11.2 (10.5) Study design. The data set for this analysis was derived from the a Mean 9HPT, s (SD) 30.9 (19.6) PROMiSe study, a randomized, double-blind, placebo-controlled clinical trial with a planned duration of 3 years.11 The design of Abbreviations: 9HPT 5 9-hole peg test; EDSS 5 Expanded 11 Disability Status Scale; T25FW 5 timed 25-foot walk. the PROMiSe study has been reported previously. In summary, 5 a Average across 4 trials. patients who took part in the PROMiSe study (N 943) were randomized in a 2:1 ratio to receive glatiramer acetate (GA) 20 mg once daily (n 5 627) or identical-appearing placebo (n 5 a combined disability measure that incorpo- 316) by daily subcutaneous injection.11 The primary endpoint rates scoring from multiple instruments—for was time to confirmed disability progression (CDP), with CDP example, using T25FW, 9HPT, and EDSS— being defined as a change of $1 point on the EDSS sustained for 3 months in patients with a baseline EDSS score of 3.0–5.0, or may be associated with more progression a change of $0.5 points for 3 months in those with a baseline events among patients with PPMS, and thus score of 5.5–6.5.11 Patients also completed the MS Functional higher reported event rates than achieved with Composite (MSFC) evaluation, a 3-part test including the any single instrument alone.9,10 T25FW and 9HPT and the Paced Auditory Serial Addition Test (PASAT).11 The PROMiSe study represents one of the best and largest (N 5 943) data sources on Patients. Patients eligible to participate in the PROMiSe study were aged between 30 and 65 years, with an EDSS score of patients with PPMS and offers a unique 3.0–6.5, investigator-confirmed PPMS, and progressive symp- opportunity to assess the value of different toms including myelopathy for $6 months before screening. functional measures of disability progression Additional eligibility criteria included evidence of pyramidal in a clinical trial setting.11 We report here damage on neurologic examination, including a pyramidal functional system score $2, and evidence of multilevel CNS disease.11 Patients with any history of MS relapse, those with , Table 2 Cumulative and cross-sectional PROMiSe disease progression rates at lymphopenia 3,000 cells/mL, and those who had used an b 12 and 24 months interferon- drug, immunosuppressant, immunomodulating agent, corticosteroid, or investigational drug within 3 months of Progression rates Progression rates study initiation were not eligible to take part.11 at 12 months, % at 24 months, % Only data for patients randomized to placebo in the PROM- iSe trial were included in this analysis in order to compare clinical Cross Cross Cumulative sectional Cumulative sectional measures in patients with PPMS for whom the natural course of the disease was not potentially affected by exposure to GA UDP 3M CDP UDP 3M CDP UDP 3M CDP UDP 3M CDP treatment.

EDSS alone 29.5 14.9 21.3 17.7 47.6 32.4 29.4 26.6 Outcome evaluations. Disability was assessed using change T25FW 20% alone 51.7 33.5 34.9 26.5 71.7 52.0 48.6 44.4 in single (EDSS, T25FW, and 9HPT) and composite disability

9HPT 20% alone 26.6 8.4 15.4 9.6 38.0 18.0 19.6 15.4 measures. Combination endpoints were defined as progression by EDSS or T25FW, 9HPT or T25FW, or by EDSS or EDSS or T25FW 20% 58.1 38.5 42.3 32.7 77.9 59.5 53.7 49.5 T25FW or 9HPT. The PASAT was not included in this anal- T25FW 20% or 59.1 36.3 41.2 32.0 77.6 56.7 52.8 47.7 ysis of the PROMiSe database because of the potential for 9HPT 20% practice effects occurring from repeated administration. More- EDSS or T25FW 20% 63.8 41.4 47.8 37.5 82.1 63.9 57.5 52.8 over, the test may have limited measurement reliability, is or 9HPT 20% sometimes difficult to administer, has been used less frequently

Abbreviations: 3M CDP 5 3-month confirmed disability progression; 9HPT 5 9-hole peg in recent MS trials, and is not anticipated to be featured in test; EDSS 5 Expanded Disability Status Scale; T25FW 5 timed 25-foot walk; UDP 5 future trials. unconfirmed disability progression. As per the PROMiSe study protocol, worsening of the EDSS For confirmed progression rates, patients who did not have the confirmation visit by the was defined as a change from baseline of $1 point (for baseline specified time point were censored for analysis purposes. EDSS 3.0–5.0) or $0.5 points (for baseline EDSS 5.5–6.5).11

2 Neurology: Neuroimmunology & Neuroinflammation after baseline due to worsening disability.12 For each measure, Figure 1 Comparison of cumulative and cross-sectional UDP and CDP rates for rates of unconfirmed disability progression (UDP) and of 3- EDSS, T25FW 20%, and 9HPT 20% month CDP were determined. Two trial methodologies were evaluated: cumulative and cross-sectional. The cumulative approach describes the design of the PROMiSe trial and similar large-scale randomized clinical trials, with outcome measurement every 3 months and disability progression at any time point during follow-up contributing to an overall CDP proportion. The cross-sectional approach describes the design for simple, straightforward investigator-initiated trials such as futility trials13 and observational studies on PPMS disability progression.10 In this case, the measurement at one follow- up time point (e.g., 12 months) is compared with the baseline measurement. This analysis is restricted to patients in the PROMiSe trial receiving placebo in order to compare clinical measures in patients with PPMS for whom the natural course of the disease was not affected by exposure to GA treatment.

Statistical analyses. Cumulative progression rates were estimated using the Kaplan-Meier methodology. For summaries by month, 1 month was defined as 28 days. For confirmed disability events, patients who experienced a worsening event in the measure of interest that could not be confirmed due to the patient’s discontinuation from the trial prior to the confirmation visit were censored at the time of the worsening. Other patients who did not experience the progression event were censored at their last day in the study. Cross-sectional progression rates were determined by comparing baseline and follow-up disability measures. Patients missing any of the 3 disability measures at baseline, the time point of interest, or the corresponding confirmation assessment were excluded from the analysis. The relative percentage contribution of each disability measure to disease progression rates observed for the combined EDSS or T25FW or 9HPT measure was derived by determining the number of individual patients progressing on 1, 2, or all 3 outcomes measures.

RESULTS A total of 316 patients were assigned to receive placebo in the PROMiSe study; 314 of these patients had a postbaseline EDSS assessment, and these patients constituted the population for the cumulative analyses presented here. Of these, 272 patients were included in the 12-month cross-sectional analysis, and 214 patients constitute the population for the 24-month cross-sectional analysis. Baseline demographic and clinical characteristics for the 314 patients who received placebo are shown in table 1.11 The baseline characteristics of the subgroup of patients assigned to receive placebo were consistent with those of the entire PROMiSe patient population.11 Table 2 shows cumulative and cross-sectional UDP and CDP rates at 12 and 24 months. Among (A) EDSS rates. (B) T25FW 20% rates. (C) 9HPT 20% rates. 3M CDP 5 3-month confirmed $ disability progression; 9HPT 5 9-hole peg test; EDSS 5 Expanded Disability Status Scale; single measures, a 20% worsening in T25FW T25FW 5 timed 25-foot walk; UDP 5 unconfirmed disability progression. (T25FW 20%) was associated with the highest cumulative progression rates, followed by EDSS and then a $20% worsening in 9HPT (9HPT 20%) at both Worsening of the T25FW was defined as an increase of $20% the 12- and 24-month time points. Among combined compared with baseline or as the inability to complete the measures, the composite of all 3 individual measures T25FW after baseline due to worsening disability (increase of (i.e., EDSS or T25FW 20% or 9HPT 20%) had the EDSS to 7.0 or higher).12 Worsening of the 9HPT was defined as an increase of $20% compared with baseline (calculated as the highest cumulative 3-month CDP rate at both 12 and average across 4 trials: 2 in the dominant hand and 2 in the 24 months (41.4% and 63.9%, respectively). Cross- nondominant hand) or as the inability to complete the 9HPT sectional 3-month CDP rates were also highest for the

Neurology: Neuroimmunology & Neuroinflammation 3 composite of all 3 scores than for any single or dual been identified. In other words, a greater proportion combination of scores at both 12 and 24 months of patients identified as progressors by T25FW were (37.5% and 57.5%, respectively). UDP proportions not identified as progressors by EDSS or 9HPT, as were consistently higher than CDP proportions for all compared with the proportion of patients identified measures, with a considerably greater difference as progressors by EDSS who were not identified as between UDP and CDP in the cumulative incidence progressors by T25FW or 9HPT (figure e-2). of the composite of all 3 individual measures (1 year: 63.8% and 41.4%, respectively; 2 years: 82.1% and DISCUSSION The data presented here show that 63.9%, respectively) compared with the cross- among patients with PPMS randomized to receive sectional proportions (1 year: 47.8% and 37.5%, placebo in the PROMiSe study, T25FW 20% at 12 respectively; 2 years: 57.5% and 52.8%, respectively). and 24 months was associated with higher rates of dis- The proportion of patients with UDP and 3- ability progression than EDSS, and both T25FW month CDP for each of the disability metrics over 20% and EDSS were associated with higher disability the course of the study is shown in figure 1 for each rates than 9HPT 20%. The 3-part combination study paradigm (cumulative and cross sectional). The measure (EDSS or T25FW 20% or 9HPT 20%) was T25FW 20%-alone values were closely aligned with associated with more 3-month CDP events (41.4% those from the combined EDSS-or-T25FW 20% val- and 63.9% of patients as 12 and 24 months, ues at all time points over the course of the study respectively) than the 2-part measure (EDSS or (table 3). T25FW 20% [38.5% and 59.5%, respectively]), and The relative contribution of each of the individual more progression events than any single measure component measures—together and in different (EDSS, 14.9% and 32.4%, respectively; T25FW, combinations—of a 3-part composite measure 33.5% and 52.0%, respectively; 9HPT, 8.4% and (EDSS or T25FW 20% or 9HPT 20%) is shown 18.0%, respectively). in figure 2 and figure e-1 at Neurology.org/nn. Minimal changes in the 9HPT 20% outcome T25FW 20% had the greatest contribution to the were noted over the course of the PROMiSe trial. overall measure, followed by EDSS, and then As such, this measure may be less suitable than EDSS 9HPT 20%. These figures support the notion that and T25FW 20% when used alone; however, 9HPT CDP as defined by T25FW identifies individuals as may still have utility as part of a composite measure. progressors that the other measures would not have Moreover, future studies should explore thresholds other than 20% worsening on performance measures such as 9HPT. Table 3 Disease progression rates over the course of the PROMiSe study A marked difference was noted between cumulative and cross-sectional progression rates, which Month may inform the design of PPMS trials. Cumulative

Measure, % patients 36 9 1215182124 progression rates were generally higher than cross- sectional rates, but more strikingly, there was a greater Unconfirmed disease progression difference between unconfirmed and CDP using the

EDSS alone 2.2 7.7 16.0 24.2 31.8 37.2 42.5 44.9 cumulative approach. Based on our results, the nor-

T25FW 20% alone 2.6 20.4 40.3 48.7 53.4 60.6 66.6 70.3 mative approach to use CDP in classic randomized controlled trials is supported. Failure to confirm dis- 9HPT 20% alone 1.3 10.6 17.0 21.9 27.6 32.6 34.9 37.0 ability progression may have a neurobiological basis EDSS or T25FW 20% 3.5 24.9 44.8 53.8 60.1 66.9 72.2 75.9 or may be due to measurement unreliability; it would T25FW 20% or 9HPT 20% 3.5 26.8 46.4 54.2 61.1 67.7 73.3 77.0 be preferable to exclude these from pivotal trial anal- EDSS or T25FW 20% 4.1 30.9 50.5 58.9 66.1 72.6 77.8 81.2 yses in both cases. For trials using the cross-sectional or 9HPT 20% approach, measures of UDP may still be feasible. 3-month confirmed disease progression Such studies would be resource-sparing and less bur- densome for participants, but would have acknowl- EDSS alone 1.3 4.8 10.3 15.5 22.6 26.2 31.0 32.4 edged data sparsity and methodological limitations. T25FW 20% alone 1.9 13.1 25.6 33.5 36.5 43.2 48.2 53.1 The observations presented here add to a growing 9HPT 20% alone 0.6 4.2 5.5 8.8 11.8 14.8 16.2 18.0 body of evidence that suggests that combining spe- EDSS or T25FW 20% 2.6 15.9 30.1 39.2 43.8 50.4 56.5 60.5 cific disability measurements may be more useful in T25FW 20% or 9HPT 20% 2.2 15.0 27.5 36.6 41.6 48.6 53.0 57.8 the clinical trial setting for assessing progression in pa- 9,10,14–16 EDSS or T25FW 20% 2.9 17.9 32.0 42.4 48.6 55.5 60.9 64.9 tients with PPMS than single measures. In or 9HPT 20% a retrospective database study of 181 patients with

Abbreviations: 9HPT 5 9-hole peg test; EDSS 5 Expanded Disability Status Scale; T25FW progressive MS (primary 47% of patients and second- 5 timed 25-foot walk. ary 53% of patients), combining change in T25FW

4 Neurology: Neuroimmunology & Neuroinflammation Figure 2 Contribution of individual clinical measures evaluating patient worsening at 24 months (cross- sectional 3-month CDP) within a 3-part combined measure (EDSS or T25FW 20% or 9HPT 20%)

(A) Cumulative UDP. (B) Cumulative 3-month CDP. (C) Cross-sectional UDP. (D) Cross-sectional 3-month CDP. 3M CDP 5 3-month confirmed disability progression; 9HPT 5 9-hole peg test; EDSS 5 Expanded Disability Status Scale; T25FW 5 timed 25-foot walk; UDP 5 unconfirmed disability progression. with EDSS was significantly more predictive of patients with an initial confirmed progression event patient prognosis than EDSS alone.15 An earlier ret- at 12 and 24 weeks, respectively.9 Confirmed pro- rospective database study of 161 patients with PPMS gression rates for EDSS alone were 38.5% and found that when comparing single measures or 2- 30.4%; for T25FW alone the rates were 51.0% instrument combinations involving EDSS, T25FW, and 44.5%; and for the 9HPT alone the rates were and 9HPT, T25FW/9HPT predicted the greatest 21.1% and 17.4%, respectively.9 Most recently, the number of progression events after 1 year (46% of large, prospective INFORMS (fingolimod) study, patients progressed compared with 17%, 34%, and comprising patients with PPMS treated for at least 20% for the individual measures, respectively), and 3 years, reported that a composite of EDSS, T25FW, the combination of T25FW/EDSS predicted the or 9HPT predicted more progression events (80.3% greatest number of progression events after the second cumulative probability of CDP) than any of the indi- year (57% of patients progressed compared with vidual components alone (EDSS 58.7%, T25FW 32%, 46%, and 24% for the individual measures, 70.0%, and 9HPT 41.3% of patients) among those respectively).11 An evaluation of data from the 96- randomized to placebo.14 week Olympus (rituximab) trial17 in the 147 patients Although those previous published studies offer with PPMS randomized to placebo found that a com- similar conclusions to this study regarding the relative bined measure consisting of EDSS/T25FW/9HPT utility of outcome measures, they differ in several was associated with more CDP events than did any important ways that distinguish them from our anal- single or dual combination measure.11 Progression ysis of the PROMiSe data set. The retrospective data- rates at 96 weeks were 61.9% and 54.8% for those base study included both primary and secondary

Neurology: Neuroimmunology & Neuroinflammation 5 progressive patients with MS, was primarily focused AUTHOR CONTRIBUTIONS on outcomes of early vs late changes on clinical scales, Marcus W. Koch: conception of the study, data analysis, writing of the first included patients who were exposed to disease- draft, and revision of the manuscript. Gary R. Cutter: study/analysis review, manuscript review, and approval. Gavin Giovannoni: study/analysis design, modifying therapies (DMTs) during the course of manuscript review, and approval. Bernard M.J. Uitdehaag: study/analysis the study, did not set a minimum disability criterion review, manuscript review, and approval. Jerry S. Wolinsky: overall conduct as per EDSS for study entry, and included less- and image analysis of the original clinical trial, warehoused components 15 of the trial data, early investigation of the database for clinical-biomarker frequent patient assessment. Similarly, the earlier correlations, evaluation plan and data review, and preparation of this study likely included patients exposed to DMTs, manuscript. Mat D. Davis: study/analysis design, data analysis, manuscript while its entry criteria included an EDSS threshold preparation, review, and approval. Joshua R. Steinerman: study/analysis score of 2.0–6.5, representing a less-disabled patient design, manuscript preparation, review, and approval. Volker Knappertz: study/analysis design, data analysis, manuscript preparation, review, and 10 population than that of the PROMiSe study. The approval. same EDSS criteria applied to the Olympus study, whereas none of these 3 studies, nor the INFORMS ACKNOWLEDGMENT study, compared confirmed with unconfirmed pro- The authors thank the patients, investigators, and site personnel involved gression or offered a cross-sectional analysis of disease with the PROMiSe study. Dr. Justin Potuzak provided editorial support at the express direction of the authors. He had no role in the design or con- 9,14 progression data. ceptualization of the study, in analysis or interpretation of the data, or in Two other recently published studies address revising the manuscript for intellectual content. similar themes and arrive at compatible conclusions to those of this study.18,19 In one of these studies, STUDY FUNDING however, the patient population consisted entirely This study was funded by Teva Pharmaceutical Industries. of those with secondary progressive MS,18 whereas DISCLOSURE the other had a lower EDSS inclusion threshold M.W. Koch served on the scientific advisory board for Roche and Bio- (1–7), allowed inclusion of patients who had been gen; received travel funding and/or speaker honoraria from Genzyme treated with DMTs as recently as 3 months prior to Sanofi and Biogen; and received research support from Teva and Alberta study entry, and was intended as a validation study Innovated Health Solutions. G.R. Cutter served on the data and safety monitoring committees for AMO Pharmaceuticals, Apotek, Gilead, of a novel disability score (CombiWISE); that scor- Horizon, Modigenetech/Prolor, Merck, Merck/Pfizer, OPKO Biologics, ing measure, in addition to EDSS, T25FW, and Sanofi-Aventis, Reata, Receptos/Celgene, Teva, NHLBI, and NICHD; 9HPT, also includes the Scripps Neurological Rat- received speaker honoraria from Consortium of MS Centers and Teva; ’ 19 served on the editorial board for Multiple Sclerosis, JASN, Alzheimer s& ing Scale. And, as with the previously noted stud- Dementia: Translational Research & Clinical Interventions; is president of ies, neither unconfirmed disease nor a cross- Pythagoras; has consulted for Atara Biotherapeutics, Biogen GmbH, Cer- sectional analysis was included in the design of eSpir Inc., Consortium of MS Centers, Genzyme, Genentech, Innate these studies.18,19 Thus, this study offers several Therapeutics, Janssen, Klein-Buendel Incorporated, MedImmune, Med- Day, Nivalis, Novartis, Opexa Therapeutics, Roche, Savara Inc., Somah- unique features that make it a potentially valuable lution, Teva, Transparency Life Sciences, and TG Therapeutics; contribution to the literature, such as reflecting the participated in NARCOMS MS Patient Registry; and received research untreated natural history of PPMS by including support from NIH/NAID, NIH/National Institute of Neurological Dis- only patients receiving placebo, including patients orders and Stroke, NIH/NHLBI, NIH/NICHD, NIH/NIA, Consor- tium of MS Centers, US Department of Defense, NIH/NIAMS, NIH/ with more severe disability, comparing confirmed NIDDK, UAB/UCSD, Children’s Hospital (Boston), the State of Ala- and unconfirmed disease progression, and includ- bama, and the Myasthenia Gravis Foundation. G. Giovannoni served on ing a cross-sectional analysis. the scientific advisory board for Biogen Idec, Five Prime, Genzyme, GQ Pharma, Ironwood, Merck Serono, Novartis, Roche, Sanofi-Aventis, Syn- There is an unmet need for controlling disease thon BV, Teva, Vertex, AbbVie, and Canbex; received speaker honoraria progression in patients with PPMS. For these pa- from Biogen Idec, Genzyme, GW Pharma, Merck Serono, Novartis, tients, individual measures of disease progression Roche, and Teva; is editor for Multiple Sclerosis and Related Disorders; may limit the potential to assess the benefit of new consulted for Biogen Idec, Five Prime, Genzyme, GW Pharma, Iron- wood, Merck Serono, Novartis, Roche, Sanofi-Aventis, Synthon BV, agents to the extent that the composite measures have Teva, Vertex, AbbVie, and Canbex; served on the speakers’ bureau for higher event rates. Based on the observations pre- Novartis and Teva; and received research support from Genzyme and sented here and those from previous studies, future Merck. B.M.J. Uitdehaag consulted for Novartis, Merck Serono, Biogen studies of agents for the treatment of PPMS may Idec, Teva, Sanofi Genzyme, and Roche. JS Wolinsky served on the scientific advisory board for Clene Nanomedicine, Forward Pharma, benefit from either using T25FW 20% as a single MedDay, Novartis, Roche/Genentech, and Sanofi/Genzyme; received outcome measure or using T25FW 20% in combina- travel funding and/or speaker honoraria from ACTRIMS, Academic tion with the EDSS as a primary efficacy endpoint CME, CMSC, PRIME, WebMD, University of Kansas, University of Maryland, ECTRIMS, and ACTRIMS; served on the editorial board for rather than EDSS alone, particularly if higher event Multiple Sclerosis Journal; was editor for Multiple Sclerosis and Related rates over shorter exposure periods are sought. It is Disorders; has consulted for AbbVie, Acetylon, Alkermes, Bayer Health- largely unknown whether specific disability outcomes Care, Celgene, Forward Pharma, EMD Serono, Genzyme, Novartis, would differentially detect treatment effects or Roche, Takeda, and Teva; received research support from NIH; and receives royalties for monoclonal antibodies out-licensed through the whether this would differ according to the mecha- University of Texas Health Science Center at Houston to Millipore nism of the treatment under study. (Chemicon International) Corporation. M.D. Davis is employed by

6 Neurology: Neuroimmunology & Neuroinflammation and holds stock or stock options in Teva Pharmaceutical Industries. J.R. 9. Zhang J, Waubant E, Cutter G, Wolinsky J, Leppert D. Steinerman served on the data safety monitoring board for NIH; has Composite end points to assess delay of disability progres- been employed by Teva Pharmaceutical Industries; consulted for Click sion by MS treatments. Mult Scler 2014;20:1494–1501. Therapeutics; and holds stock or stock options in Teva and Click Ther- 10. Bosma LV, Kragt JJ, Brieva L, et al. The search for respon- apeutics. V. Knappertz is employed by Teva Pharmaceutical Industries sive clinical endpoints in primary progressive multiple scle- and holds stock or stock options in Teva and Knopp Biosciences. Go to rosis. Mult Scler 2009;15:715–720. Neurology.org/nn for full disclosure forms. 11. Wolinsky JS, Narayana PA, O’Connor P, et al. Glatiramer Received January 3, 2017. Accepted in final form March 23, 2017. acetate in primary progressive multiple sclerosis: results of a multinational, multicenter, double-blind, placebo-controlled REFERENCES trial. Ann Neurol 2007;61:14–24. 1. Stellmann JP, Neuhaus A, Lederer C, Daumer M, Heesen 12. Schwid SR, Goodman AD, Apatoff BR, et al. Are quan- C. Validating predictors of disease progression in a large titative functional measures more sensitive to worsening cohort of primary-progressive multiple sclerosis based on MS than traditional measures? Neurology 2000;55: a systematic literature review. PLoS One 2014;9:e92761. 1901–1903. 2. Koch MW, Cutter G, Stys PK, Yong VW, Metz LM. 13. Koch MW, Korngut L, Patry DG, et al. The promise of Treatment trials in progressive MS—current challenges futility trials in neurological diseases. Nat Rev Neurol and future directions. Nat Rev Neurol 2013;9:496–503. 2015;11:300–305. 3. Fox RJ, Thompson A, Baker D, et al. Setting a research 14. Lublin F, Miller DH, Freedman MS, et al. Oral fingolimod agenda for progressive multiple sclerosis: the International in primary progressive multiple sclerosis (INFORMS): a phase Collaborative on Progressive MS. Mult Scler 2012;18: 3, randomised, double-blind, placebo-controlled trial. Lancet 1534–1540. 2016;387:1075–1084. 4. Holland NJ, Schneider DM, Rapp R, Kalb RC. Meeting 15. Bosma LV, Kragt JJ, Knol DL, Polman CH, Uitdehaag the needs of people with primary progressive multiple scle- BM. Clinical scales in progressive MS: predicting long- rosis, their families, and the health-care community. Int J term disability. Mult Scler 2012;18:345–350. MS Care 2011;13:65–74. 16. van Winsen LM, Kragt JJ, Hoogervorst EL, Polman CH, 5. Coetzee T, Zaratin P, Gleason TL. Overcoming barriers in Uitdehaag BM. Outcome measurement in multiple scle- progressive multiple sclerosis research. Lancet Neurol rosis: detection of clinically relevant improvement. Mult 2015;14:132–133. Scler 2010;16:604–610. 6. Ziemssen T, Rauer S, Stadelmann C, et al. Evaluation of 17. Hawker K, O’Connor P, Freedman MS, et al. Rituximab study and patient characteristics of clinical studies in pri- in patients with primary progressive multiple sclerosis: re- mary progressive multiple sclerosis: a systematic review. sults of a randomized double-blind placebo-controlled PLoS One 2015;10:e0138243. multicenter trial. Ann Neurol 2009;66:460–471. 7. Ontaneda D, LaRocca N, Coetzee T, Rudick R; NMSS 18. Cadavid D, Cohen JA, Freedman MS, et al. The EDSS- MSFC Task Force. Revisiting the multiple sclerosis func- Plus, an improved endpoint for disability progression in tional composite: proceedings from the National Multiple secondary progressive multiple sclerosis. Mult Scler 2017; Sclerosis Society (NMSS) Task Force on Clinical Disabil- 23:94–105. ity Measures. Mult Scler 2012;18:1074–1080. 19. Kosa P, Ghazali D, Tanigawa M, et al. Development of 8. Bosma LV, Sonder JM, Kragt JJ, Polman CH, Uitdehaag a sensitive outcome for economical drug screening for BM. Detecting clinically-relevant changes in progressive progressive multiple sclerosis treatment. Front Neurol multiple sclerosis. Mult Scler 2015;21:171–179. 2016;7:131.

Neurology: Neuroimmunology & Neuroinflammation 7 Optical coherence tomography and visual evoked potentials in pediatric MS

Amy T. Waldman, MD, ABSTRACT MSCE Objective: To determine the relative ability of optical coherence tomography (OCT) and Grant T. Liu, MD pattern-reversal visual evoked potentials (pVEPs) to detect visual pathway involvement in Amy M. Lavery, PhD pediatric-onset MS. Geraldine Liu, MA Methods: Pediatric-onset MS participants (onset ,18 years) and healthy controls (HCs) under- William Gaetz, PhD went OCT (Cirrus HD-OCT) and pVEPs. Retinal nerve fiber layer (RNFL), ganglion cell layer to inner Tomas S. Aleman, MD plexiform layer (GCL-IPL), and P100 pVEP latency were measured. Generalized estimating equa- Brenda L. Banwell, MD tion models were used to compare the groups, adjusting for age and intereye correlations. Results: Twenty-four pediatric MS participants, 14 with a history of remote (.6 months) optic Correspondence to neuritis (ON) in one eye (8 participants) or both the eyes (6 participants), and 24 HCs were Dr. Waldman: enrolled. RNFL thinning (,83 mm, 2 SDs below HC eyes) occurred in 50% of ON eyes vs [email protected] 5% of non-ON eyes. Prolonged VEP latency (.109 msec) occurred in 58% of ON eyes and 55% of non-ON eyes. A clinical history of ON predicted RNFL (p , 0.001) and GCL-IPL thinning (p 5 0.011), whereas prolonged pVEP latency in children with MS occurred independent of ON history. Conclusions: OCT and pVEPs provide complementary but distinct insights. OCT is sensitive to retinal changes in the context of clinical ON, whereas pVEPs are useful to detect disseminated lesions of the visual pathway in children with MS. Neurol Neuroimmunol Neuroinflamm 2017;4: e356; doi: 10.1212/NXI.0000000000000356

GLOSSARY ETDRS 5 Early Treatment Diabetic Retinopathy Study; GCL-IPL 5 ganglion cell layer to inner plexiform layer; GEE 5 generalized estimating equation; HC 5 healthy control; OCT 5 optical coherence tomography; ON 5 optic neuritis; pVEP 5 pattern-reversal visual evoked potential; RNFL 5 retinal nerve fiber layer; VA 5 visual acuity.

Optical coherence tomography (OCT) and pattern-reversal visual evoked potentials (pVEPs) assess the structural and functional integrity of the visual pathway. Although OCT and pVEPs are well-substantiated tools to evaluate MS in adults, less is known about the comparative utility of these tools in pediatric-onset MS. Importantly, because OCT interrogates retinal injury and pVEPs assess visual pathway pathology, these tools may inform differently. Prior studies evaluating VEP alone in children with optic neuritis (ON) have demonstrated prolonged latencies and/or reduced amplitudes in 83% of children with acute ON.1 PVEPs are abnormal in 56% of children with confirmed MS at the time of their first attack, even if this attack did not include clinical ON.2 In OCT studies, thinning of the retinal nerve fiber layer (RNFL) and the ganglion cell layer (GCL) occurs after ON, with maximal thinning in the ON eye recorded at 6 months post-ON.3 In adults with MS, thinning of the retina has also been documented in the absence of ON, although thinning in non-ON eyes in pediatric MS has been variably detected.4–6 In pediatric-onset MS, the likelihood of an OCT or VEP abnormality in MS eyes, Supplemental data at Neurology.org/nn From the Division of Neurology (A.T.W., A.M.L., G.L., B.L.B.), Neuro-ophthalmology Service (G.T.L.), Division of Ophthalmology, and Division of Radiology (W.G.), Children’s Hospital of Philadelphia, PA; and Departments of Neurology (A.T.W., B.L.B., G.T.L.), Pediatrics (A.T. W., B.L.B.), and Ophthalmology (G.T.L., T.S.A.), Perelman School of Medicine, University of Pennsylvania, Philadelphia. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 with or without ON, has not been calculated. Full-field (20°) pVEPs were performed in CHOP’s Neuro- We sought to compare OCT and pVEPs in physiology Laboratory by trained technicians on a Natus System (software version Nicolet EDX21.1; Natus Neurology, Pleasan- a cohort of pediatric-onset MS participants, ton, CA) using a Queens’ Square montage and following the with and without a history of ON, in order recommendations of the International Society of Clinical Elec- to inform on the relative utility of these tests in trophysiology of Vision (ISCEV).11 Participants were positioned the pediatric context. 1 m from the computer screen and asked to maintain visual fixation on a red cross in the center of an alternating checkerboard pattern of black and white squares (check size 329) at 100% METHODS Pediatric-onset MS participants (whose first attack contrast. Monocular recordings were performed while the partic- occurred prior to age 18 years) were recruited for this cross- ipants wore their corrective lenses (if required) and with a patch sectional study from The Children’s Hospital of Philadelphia over the opposite eye. Analyses focused on the midoccipital elec- (CHOP) between November, 2013, and July, 2015. Healthy trode channel. The P100 latency was recorded, and the tracings controls (HCs) were recruited through local advertisement. were reviewed and manually corrected if needed by 2 of the Healthy participants without neurologic, ocular, or systemic investigators (A.T.W. and T.S.A.). Technician and VEP machine disease were eligible provided they had a normal corrected visual availability were dependent on clinical workflow and were not acuity (VA) (defined as 47 or more letters correct using the Early available for every participant. Treatment Diabetic Retinopathy Study (ETDRS) chart, which is RNFL and GCL-IPL thicknesses (in micrometers) were approximately 20/25 using Snellen acuity). defined as abnormal if they were beyond 2 SDs of the mean values The diagnosis of MS was confirmed using the 2010 obtained from HC eyes. Normality of OCT data was assessed McDonald criteria.7 Prior ON was confirmed by history and through histograms, measures of kurtosis and skewness, and the documentation of reduced high-contrast VA, red desatura- Shapiro-Wilk W test. P100 latency was considered prolonged if tion, pain with ocular movement, and/or visual field defect. the latency was .2 SDs above the control mean. Proportions A chart review was performed to abstract the maximum VA of eyes with normal and abnormal tests were determined. The deficit, presence of optic disc edema, color vision loss, pain sensitivity, specificity, positive and negative predictive values of with eye movements, and findings on confrontation visual OCT and VEP measures to detect MS-ON, MS fellow, and MS field examination as recorded at the time of ON. MRI of non-ON eyes were calculated. Generalized estimating equation the orbits (if available), brain, and spine were reviewed for (GEE) models, adjusting for age and intereye correlations, were all participants to confirm ON and MS diagnoses. The sub- used to compare the groups. sequent development of optic nerve pallor, as adjudicated by Intereye differences were determined by subtracting values a neurologist or neuro-ophthalmologist, was documented. between the right and left eyes for RNFL, GCL-IPL, and VEP None of the participants had active acute ON at the time of latency. Intereye differences for each participant exceeding 2 evaluation, and all were at least 6 months from an acute attack SDs above the mean differences for HC eyes were assessed and of ON. Eyes were categorized as ON eyes, fellow eyes (the compared using a t test. unaffected eye in a participant with unilateral ON), or non- ON eyes (no history of ON in either eye). Standard protocol approvals, registrations, and patient All study procedures occurred on the same day, beginning consents. The study was approved by the Institutional ’ with the measurement of high-contrast VA using the ETDRS Review Board at The Children s Hospital of Philadelphia. VA chart with the participant’s own corrective lenses or glasses Written informed consent and child assent were obtained if needed. Although we did not perform cycloplegic refraction, all from all participants. participants had their refractive error for each eye recorded using an autorefractor (VisuRef; Carl Zeiss Meditec, Inc., Dublin, CA) RESULTS Twenty-four pediatric-onset MS partic- as a refractive error can affect OCT results.8,9 Refractive errors did ipants and 24 HCs were enrolled (table 1). Eleven not exceed 4 diopters. additional pediatric-onset MS participants declined Spectral domain OCT was obtained in a dark room without enrollment because of time constraints or lack of windows by a single trained technician using Cirrus HD-OCT interest in research. Among the MS participants (Model 5000, software version 6.5; Carl Zeiss Meditec, Inc.) at 27,000 A-scans per second. Participants underwent the Optic enrolled, 14 (58%) had a history of ON in one eye 5 5 Disc Cube (200 3 200 scans) and Macular Cube (512 3 128) (N 8) or both the eyes (N 6). In total, 20 MS- protocols for each undilated eye while fixating on landmarks ON eyes, 8 MS-fellow eyes, 20 MS non-ON eyes, positioned nasally or centrally to the Optic Disc Cube and Mac- and 48 HC eyes were included. Children with MS ular Cube, respectively. The technician verified that the images and ON had excellent recovery of high-contrast VA: were focused and centered with uniform illumination and as- of the 20 ON eyes, 17 (85%) achieved a VA of 20/32 sessed for artifacts as recommended by the OSCAR-IB Consensus or better at the time of the last follow-up examina- Criteria for Retinal OCT Quality Assessment.10 Only scans meeting these criteria and having a signal strength $7 were used for tion. Optic nerve pallor was seen in 14/20 eyes (70%) the analysis. Segmentation was performed using the built-in auto- with a history of ON. None of the fellow eye optic matic segmentation algorithm of the OCT system. Inspection of nerves were pale. Only one of the eyes in a child the resulting Macular Cube cross sections revealed no evidence without ON had temporal nerve pallor. for eccentric fixation as the fovea was always centered at fixation. OCT scans were excluded for failing quality con- RNFL thickness measured from an interpolated 1.7-mm-radius trol (due to movement and artifacts) including RNFL circle centered on the optic nerve head obtained from the Optic Disc Cube scan. GCL to inner plexiform layer thickness (GCL- images for 1 HC eye and GCL-IPL measurements for IPL) was obtained from analysis of the parafoveal and perifoveal 1 HC and 3 MS eyes. Eight MS participants and 6 retina of the Macular Cube scan. controls did not undergo VEP because of scheduling

2 Neurology: Neuroimmunology & Neuroinflammation GCL-IPL thinning, and delayed VEP latencies), and Table 1 Demographics 3 of these 4 eyes had a remote history of ON.

Healthy controls In MS participants, the presence of ON pallor MS (N 5 24) (N 5 24) p Value (N 5 15 eyes) was associated with a greater likelihood Mean age, y (range) 17.5 (11–24) 15.4 (8–22) 0.046 of RNFL thinning (mean 80 mm, SD 13) compared Sex (F:M) 14:10 16:8 0.105 with RNFL thickness in nonpale eyes (N 5 33 eyes; m , Race, n (%) 0.152 mean RNFL 94 m, SD 13; p 0.001). GCL-IPL was also thinner in pale optic nerves (mean 70 mm, Caucasian 15 (62.5) 13 (56.5) SD 10) compared with nonpale optic nerves (mean African American 9 (37.5) 5 (21.7) 79, SD 8; p 5 0.002). VEP latency, however, did not Asian 0 (0) 2 (8.7) differ as a function of optic nerve pallor. Unknown/mixed 0 (0) 3 (13.0) Table 4 compares the sensitivity, specificity, pos- Ethnicity, n (%) 0.259 itive predictive value, and negative predictive value for

Hispanic 3 (12.5) 1 (4.2) OCT and pVEPs in MS-ON and non-ON eyes com-

Non-Hispanic 21 (87.5) 20 (83.3) pared with HC eyes. RNFL thinning has moderate sensitivity (50%) for the detection of MS-ON eyes Unknown/not reported 0 3 (8.3) but not for the detection of MS in the absence of ON – Mean disease duration, y (range) 2.4 (0.3 7.7) NA NA (5%). GCL-IPL thinning was not sensitive for detect- Mean duration from last episode 2.9 (0.5–6.0) NA NA of optic neuritis, y (range) ing MS in any of the groups. Although the specificity and positive predictive values of GCL-IPL thickness EDSS, median (range) 1(0–3.5) NA NA were high (100%), this was driven by the absence of Abbreviations: EDSS 5 Expanded Disability Status Scale; NA 5 not applicable. GCL-IPL thinning in the HCs. The sensitivity, specificity, and positive and negative predictive values for constraints. Three additional VEP tracings (2 MS VEPs were similar in MS eyes regardless of ON his- eyes and 1 HC eye) were excluded because of poor tory, indicating that pVEPs are supportive of an MS fixation or movement with irregular waveforms. diagnosis but not helpful in determining prior ON. RNFL values were normally distributed in MS By defining a cut-off for abnormal retinal thick- participants and controls, whereas GCL-IPL thickness and VEP latencies, which was based on 2 SDs ness demonstrated a departure from normality based below the normative values, it is possible to miss sig- on the Shapiro-Wilk W test (table e-1 and figure e-1 nificant loss or slowing in an individual whose results at Neurology.org/nn). Mean and SD values for are asymmetric but fall within 1–2 SD of the mean. RNFL and GCL-IPL thicknesses and VEP latencies We explored intereye differences for OCT and VEP for MS participants and HCs are listed in table 2. In results to determine whether our definition of abnor- the 48 MS eyes, 27% met our criteria for RNFL mal (2 SDs below the mean) failed to detect poten- abnormality (thickness ,83 mm). The proportions tially more subtle differences in pediatric-onset MS of participants with abnormal metrics stratified by eyes. Intereye differences for RNFL and GCL-IPL whether the eye was affected by ON are presented thicknesses and VEP latency for HC and MS partic- in table 3. Among those participants who completed ipants are presented in table 5. We defined abnormal OCT and VEP, only 4 of 30 (13%) MS eyes were intereye differences as the mean intereye difference deemed abnormal by all 3 measures (RNFL thinning, plus 2 SDs from HC (RNFL 6 mm, GCL-IPL

Table 2 Mean OCT and VEP values

MS eyes

p Valuea p Valuea p Valuea p Valuea Healthy vs control Non-ON eyes vs control Fellow eyes vs control vs control eyes (N 5 48) All eyes (N 5 48) eyes (N 5 20) eyes (N 5 8) eyes ON eyes (N 5 20) eyes

RNFL (SD) 100 (8), N 5 47 90 (14), N 5 48 0.002 100 (10), N 5 20 0.193 87 (12), N 5 8 ,0.001 81 (12), N 5 20 ,0.001

GCL-IPL (SD) 83 (6), N 5 45 76 (9), N 5 46 0.001 79 (9), N 5 20 0.103 77 (10), N 5 7 0.032 72 (9), N 5 18 ,0.001

VEP latency (SD)b 101 (4), N 5 35 115 (17), N 5 30 ,0.001 119 (17), N 5 11 ,0.001 108 (6), N 5 7 ,0.001 115 (20), N 5 12 0.001

Abbreviations: GCL-IPL 5 ganglion cell layer to inner plexiform layer; OCT 5 optical coherence tomography; ON 5 optic neuritis; pVEP 5 pattern-reversal visual evoked potential; RNFL 5 retinal nerve fiber layer. a Generalized estimating equation models were used to compare each group with healthy eyes, adjusting for age and intereye correlations. b The clinical features of the MS participants who underwent pVEPs (N 5 30 eyes) did not differ from the patients who did not complete this test (N 5 18 eyes). There was no difference between the groups with respect to the number of ON eyes (p 5 0.580), the age at testing (p 5 0.811), disease duration (p 5 0.937), or mean high-contrast letter acuity scores (p 5 0.1435).

Neurology: Neuroimmunology & Neuroinflammation 3 Table 3 Proportions with abnormal findings

MS (N 5 48)

Proportions of patients with HC eyes (N 5 48) All eyes Non-ON eyes Fellow eyes ON eyes

Visual acuity deficit 0 7/48 5 15% 3/20 5 15% 1/8 5 13% 3/20 5 15%

Pallor present 0 15/48 5 31% 1/20 5 5% 0 14/20 5 70%

RNFL abnormal (<83 mm) 2/47 5 4% 13/48 5 27% 1/20 5 5% 1/7 5 13% 10/20 5 50%

GCL-IPL abnormal (<71 mm) 0 10/45 5 22% 2/20 5 10% 2/8 5 25% 6/18 5 33%

VEP latency prolonged 2/35 5 6% 16/30 5 53% 6/11 5 55% 3/7 5 43% 7/12 5 58% (>109 msec)

Abbreviations: GCL-IPL 5 ganglion cell layer to inner plexiform layer; HC 5 healthy control; ON 5 optic neuritis; RNFL 5 retinal nerve fiber layer; VEP 5 visual evoked potential. One HC had normal high-contrast visual acuity (47 letters correct in each eye) but had an RNFL of 82 and 81 mminthe right and left eyes. The GCL-IPL and VEP latencies for this individual were within the normal range. One HC had prolonged VEP latencies (112 and 116 msec in the right and left eyes), despite normal high-contrast visual acuity (59 and 60 letters), RNFL, and GCL-IPL thicknesses.

2 mm, and pVEP latency 4 msec). Intereye differ- HCs for RNFL thickness (p 5 0.565) or GCL-IPL ences greater than 6 mm for RNFL thickness occurred thickness (p 5 0.051). Prolonged VEP latency in 10 MS participants, 8 of whom had a history of occurred independent of ON history (p , 0.001). ON (5 with unilateral ON and 3 with bilateral ON); For MS eyes, the presence of pallor in an eye pre- an intereye difference greater than 2 mm in GCL-IPL dicted RNFL thinning (p , 0.001) and GCL-IPL occurred in 11 MS participants, 8 of whom had ON thinning (p 5 0.001) but not VEP latency (p 5 (5 with unilateral ON and 3 with bilateral ON), and 0.159). When ON and pallor were both included intereye difference in VEP latency greater than 4 msec in the model which adjusted for age and intereye occurred in 9 MS participants, 6 of whom had ON (4 correlations, there was no difference in RNFL thick- with unilateral, 2 with bilateral ON). Of note, if ness between MS patient and HC eyes (p 5 0.607) or intereye difference criteria were used to define optic GCL-IPL thickness (0.061). VEP latency differed nerve involvement by OCT, 3 MS patients with clin- between MS and control eyes, even after accounting ical histories of ON would have been identified whose for ON and pallor (p 5 0.001). RNFL or GCL-IPL measures failed to meet the absolute cut-off criteria for abnormality (,2 SDs below DISCUSSION OCT and VEP provide valuable, but the control mean) (see figure e-2 for a representative different information in pediatric-onset patients participant). Furthermore, in 3 MS additional pa- with MS. OCT confirms the history of ON, but tients, intereye OCT measures suggested relative unlike the findings in adult MS, OCT infrequently (but not absolute) thinning in one eye in the absence identifies retinal abnormalities in pediatric-onset pa- of any history of ON. tients with MS who have not experienced clinical A clinical history of ON predicted RNFL thinning evidence of optic nerve involvement. Only 5% of our (p , 0.001) using regression analysis. After account- patients with MS demonstrated RNFL thinning and ing for ON and intereye correlations in GEE models, 10% had GCL-IPL thinning in the absence of prior OCT parameters did not differ between MS eyes and ON. By contrast, prolonged pVEP latency occurred

Table 4 Capacity of OCT and VEP to correctly identify MS eyes with and without ON compared with healthy eyes

MS-ON eyes compared with HC eyes MS fellow eyes (non-ON eye) vs HCs MS non-ON eyes vs healthy controls

RNFL GCL-IPL Prolonged RNFL GCL-IPL Prolonged RNFL GCL-IPL Prolonged thinning, % thinning, % VEP latency, % thinning, % thinning, % VEP latency, % thinning, % thinning, % VEP latency, %

Sensitivity 50 33 58 25 13 43 5 10 55

Specificity 96 100 94 96 100 94 96 100 94

Positive predictive 83 100 78 50 100 60 33 100 75 value

Negative predictive 82 79 87 88 88 89 70 71 87 value

Abbreviations: GCL-IPL 5 ganglion cell layer to inner plexiform layer; HC 5 healthy control; OCT 5 optical coherence tomography; ON 5 optic neuritis; RNFL 5 retinal nerve fiber layer; VEP 5 visual evoked potential.

4 Neurology: Neuroimmunology & Neuroinflammation Table 5 Intereye differences for OCT and VEP metrics

HC MS p Value

RNFL, mean difference, mm 2.3 (SD 1.7) [range 0–6] 7.6 (SD 7.1) [range 0–30] 0.0011

GCL-IPL, mean difference, mm 0.7 (SD 0.8) [range 0–2] 5.1 (SD 6.0) [range 0–24] 0.0012

VEP latency, mean difference, msec 1.5 (SD 1.3) [range 0.2–5] 13.3 (SD 13.6) [range 0.8–45.4] 0.0012

Abbreviations: GCL-IPL 5 ganglion cell layer to inner plexiform layer; HC 5 healthy control; OCT 5 optical coherence tomography; RNFL 5 retinal nerve fiber layer; VEP 5 visual evoked potential. in 53% of patients with MS, a finding that was not OCT measures, we did identify 3 patients with MS influenced by a history of ON. without a history of ON whose intereye differences ex- Our data are aligned with adult studies in which ceeded the intereye variability seen in our control cohort. VEP abnormalities are more frequently seen than However, 2 of these 3 patients had RNFL and GCL-IPL RNFL thinning in non-ON eyes.12,13 Our VEP re- values above the normative mean (despite the intereye sults are also similar to a prior study in which in 48 of differences), rendering it difficult to propose evidence of the 85 (56%) pediatric patients with MS demon- non-ON global MS-related injury in our cohort. strated VEP prolongation more than 2.5 SDs beyond There are several important caveats to our work. control mean latencies and/or reduced amplitudes.2 The rarity of pediatric MS limits sample size. We Also in line with our findings, another study of 14 did not perform formal cycloplegic refractions, an pediatric patients with MS reported no difference in important issue given that severe myopia is associated VEP latencies between MS-ON eyes and MS non- with a thinner retina, and VA testing and VEPs ide- ON eyes (p 5 0.524) or between MS-ON eyes and ally should be performed with the patient’s best MS fellow eyes (p 5 0.654).5 refractive correction. To mitigate this concern at least OCT has been reported in 54 pediatric patients in part, we used an autorefractor to estimate the with MS collectively between 3 prior studies, all of refractive error and none of the patients had severe which reported a significant decrease in mean myopia (.24.00). Furthermore, nearly all of our RNFL thickness among MS-ON eyes compared patients with MS had visual acuities near or equal with HC eyes.4–6 In this study, we demonstrated to 20/40, and all HCs by inclusion criteria had 20/ a mean reduction of 19 mminRNFLthickness 20 VA in each eye, making even mild uncorrected among pediatric MS-ON eyes vs HCs, which is myopia unlikely to have been a factor in this study. comparable with a meta-analysis of adult OCT data These results can inform clinical decision making. which revealed a mean RNFL loss of 20.38 mm For a patient with suspected or confirmed MS who (95% CI 222.86 to 217.91) among 956 MS- describes a remote history of vision loss, an abnormal ON eyes vs 1,107 HC eyes.14 OCT (defined either by absolute or intereye measures Although the comparison of mean scores enables that deviate substantially from normative values) may the identification of population differences, such data be used to support a prior episode of ON. However, do not necessarily inform on the utility of testing for a normal OCT does not exclude the possibility of the individual patient. In our cohort, only 50% of prior ON. VEP abnormalities detect dissemination children with MS and a history of ON will have of disease in patients being investigated for MS and reduced RNFL that is 2 SDs below the mean for are informative even in the absence of ON. HCs; this proportion of children with abnormal RNFL results in MS-ON eyes is substantiated by 2 AUTHOR CONTRIBUTIONS previous pediatric MS OCT studies based on their Amy T. Waldman: study concept/design, acquisition of data, statistical published mean data.4,6 Although OCT parameters analysis, interpretation of data, drafting/revising the manuscript for content, including medical writing for content and obtaining funding, and in half of the MS-ON eyes may fall within a normal study supervision and coordination. Grant T. Liu: interpretation of data, reference range, intereye differences in RNFL and drafting/revising the manuscript for content, including medical writing GCL-IPL thicknesses may help the clinician in iden- for content. Amy M. Lavery: acquisition of data, analysis and interpreta- tifying a remote history of ON. tion of data, statistical analysis, and drafting/revising the manuscript for content, including medical writing for content. Geraldine Liu: acquisition In this study, the mean RNFL thickness in MS of data and drafting/revising the manuscript for content, including med- non-ON eyes was equal to the mean of HCs. We ical writing for content. William Gaetz, Tomas S. Aleman, and Brenda L. did not replicate findings of 2 other OCT studies of Banwell: interpretation of data and drafting/revising the manuscript for content, including medical writing for content. pediatric MS participants in which mean RNFL thickness was reduced in non-ON eyes, findings that STUDY FUNDING were interpreted as supportive of a more degenerative This study was supported by the NIH (NINDS) (K23NS069806, PI: global MS pathology.4,5 Using intereye differences in A. Waldman).

Neurology: Neuroimmunology & Neuroinflammation 5 DISCLOSURE 5. Yilmaz U, Gucuyener K, Erin DM, et al. Reduced retinal A.T. Waldman received travel funding from Novartis; receives publishing nerve fiber layer thickness and macular volume in pediatric royalties from UpToDate; has consulted for Optum; received research sup- multiple sclerosis. J Child Neurol 2012;27:1517–1523. port from Biogen and NIH/National Institute of Neurological Disorders and 6. Waldman AT, Hiremath G, Avery RA, et al. Monocular Stroke; and holds stock or stock options in Pfizer and Spark Therapeutics. and binocular low-contrast visual acuity and optical coher- G.T. Liu serves on the editorial board for Journal of Neuro-Ophthalmology; ence tomography in pediatric multiple sclerosis. Mult Scler receives royalties from Elsevier. A.M. Lavery reports no disclosures. G. Liu’s Relat Disord 2014;3:326–334. spouse receives royalties from Elsevier. W. Gaetz reports no disclosures. T.S. 7. Polman CH, Reingold SC, Banwell B, et al. Diagnostic Aleman received research support from NIH. B.L. Banwell served on the scientific advisory board for Biogen, Sanofi, Eli Lilly, and Novartis; received criteria for multiple sclerosis: 2010 revisions to the McDonald – travel funding and/or speaker honoraria from Biogen, Merck Serono, Teva, criteria. Ann Neurol 2011;69:292 302. and Bayer; serves on the editorial board for Neurology®; consulted for Biogen 8. Samarawickrama C, Wang XY, Huynh SC, Burlutsky G, Idec, Eli Lilly, Sanofi, and Novartis; has spoken at an event supported by the Stapleton F, Mitchell P. Effects of refraction and axial Consortium of MS Centers; and received research support from Multiple length on childhood optic disk parameters measured by Sclerosis Society of Canada, Multiple Sclerosis Scientific Research Founda- optical coherence tomography. Am J Ophthalmol 2007; tion, and National Multiple Sclerosis Society. Go to Neurology.org/nn for 144:459–461. full disclosure forms. 9. El-Dairi MA, Asrani SG, Enyedi LB, Freedman SF. Opti- cal coherence tomography in the eyes of normal children. Received January 27, 2017. Accepted in final form March 24, 2017. Arch Ophthalmol 2009;127:50–58. 10. Tewarie P, Balk L, Costello F, et al. The OSCAR-IB REFERENCES consensus criteria for retinal OCT quality assessment. 1. Tekavcic-Pompe M, Stirn-Kranjc B, Brecelj J. Optic neu- PLoS One 2012;7:1–7. ritis in children—clinical and electrophysiological follow- 11. Odom JV, Bach M, Brigell M, et al. ISCEV standard up. Doc Ophthalmol 2003;107:261–270. for clinical visual evoked potentials: (2016 update). Doc 2. Pohl D, Rostasy K, Treiber-Held S, et al. Pediatric Ophthalmol 2016;133:1–9. multiple sclerosis: detection of clinically silent lesions 12. Naismith RT, Tutlam NT, Xu J, et al. Optical coherence by multimodal evoked potentials. J Pediatr 2006;149: tomography is less sensitive than visual evoked potentials 125–127. in optic neuritis. Neurology 2009;73:46–52. 3. Costello F, Hodge W, Pan YI, Eggenberger E, Coupland 13. Di Maggio G, Santangelo R, Guerrieri S, et al. Optical S, Kardon RH. Tracking retinal nerve fiber layer loss after coherence tomography and visual evoked potentials: which optic neuritis: a prospective study using optical coherence is more sensitive in multiple sclerosis? Mult Scler 2014;20: tomography. Mult Scler 2008;14:893–905. 1342–1347. 4. Yeh EA, Weinstock-Guttman B, Lincoff N, et al. Retinal 14. Petzold A, de Boer JF, Schippling S, et al. Optical coher- nerve fiber thickness in inflammatory demyelinating dis- ence tomography in multiple sclerosis: a systematic review eases of childhood onset. Mult Scler 2009;15:802–810. and meta-analysis. Lancet Neurol 2010;9:921–932.

6 Neurology: Neuroimmunology & Neuroinflammation Both cladribine and alemtuzumab may effect MS via B-cell depletion

David Baker, PhD ABSTRACT Samuel S. Herrod, BSc Objective: To understand the efficacy of cladribine (CLAD) treatment in MS through analysis of Cesar Alvarez-Gonzalez, lymphocyte subsets collected, but not reported, in the pivotal phase III trials of cladribine and PhD alemtuzumab induction therapies. Lukasz Zalewski, PhD Methods: The regulatory submissions of the CLAD Tablets Treating Multiple Sclerosis Orally Christo Albor, PhD (CLARITY) (NCT00213135) cladribine and Comparison of Alemtuzumab and Rebif Efficacy in Klaus Schmierer, PhD Multiple Sclerosis, study one (CARE-MS I) (NCT00530348) alemtuzumab trials were obtained from the European Medicine Agency through Freedom of Information requests. Data were extracted and statistically analyzed. Correspondence to Dr. Schmierer: Results: Either dose of cladribine (3.5 mg/kg; 5.25 mg/kg) tested in CLARITY reduced the annu- [email protected] alized relapse rate to 0.16–0.18 over 96 weeks, and both doses were similarly effective in reducing the risk of MRI lesions and disability. Surprisingly, however, T-cell depletion was rather modest. Cladribine 3.5 mg/kg depleted CD41 cells by 40%–45% and CD81 cells by 15%–30%, whereas alemtuzumab suppressed CD41 cells by 70%–95% and CD81 cells by 47%–55%. However, either dose of cladribine induced 70%–90% CD191 B-cell depletion, similar to alemtuzumab (90%). CD191 cells slowly repopulated to 15%–25% of baseline before cladribine redosing. However, alemtuzumab induced hyperrepopulation of CD191 B cells 6–12 months after infusion, which probably forms the substrate for B-cell autoimmunities associated with alemtuzumab. Conclusions: Cladribine induced only modest depletion of T cells, which may not be consistent with a marked influence on MS, based on previous CD41 T-cell depletion studies. The therapeutic drug-response relationship with cladribine is more consistent with lasting B-cell depletion and, coupled with the success seen with monoclonal CD201 depletion, suggests that B-cell suppression could be the major direct mechanism of action. Neurol Neuroimmunol Neuroinflamm 2017;4: e360; doi: 10.1212/NXI.0000000000000360

GLOSSARY ALEM 5 alemtuzumab; ANOVA 5 analysis of variance; CARE-MS I 5 Comparison of Alemtuzumab and Rebif Efficacy in Multiple Sclerosis, study one; CLAD 5 cladribine; CLARITY 5 CLAD Tablets Treating Multiple Sclerosis Orally; DMT 5 disease-modifying therapy; FOI 5 Freedom of Information; NK 5 natural killer; PDF 5 portable document format; pwRMS 5 people with relapsing MS; SAI 5 secondary B-cell autoimmunity.

MS is a CNS demyelinating disease responding to immunosuppression.1 Pulsed induction therapies with cladribine (CLAD)2,3 and alemtuzumab (ALEM)4–6 can induce long-term remission,3,6 while reducing risks of a permanent immunosuppressive state through continuous drug use.7 Pivotal trials of an oral CLAD, 2-chlorodeoxyadenosine triphosphate, prodrug8 and ALEM, CD52-depleting antibody,9 suggest that both drugs have comparable clinical efficacy in controlling relapses,2–5 but markedly different side-effect profiles.2,4,5 Although both drugs induce lymphocyte depletion,8–11 only ALEM causes significant secondary B-cell autoimmunity 6,9 2 Supplemental data (SAI) in people with MS. It was suggested that CLAD may create a cancer risk, which in the at Neurology.org/nn From the BartsMS (D.B., S.S.H., C.A.G., C.A., K.S.), Blizard Institute, ITS Research (L.Z.), Queen Mary University of London; and Barts Health NHS Trust (K.S.), Emergency Care & Acute Medicine Neuroscience Clinical Academic Group, London, UK. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the Queen Mary University of London. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 absence of additional trial data caused regula- FOI requests. After termination of the commercial develop- tors to refuse licensing and halted CLAD ment of oral CLAD in 2011, and subsequent conversations with 12 the UK Medicines and Healthcare products Regulatory Agency development in 2011. However, a subse- about approaches to develop generic CLAD, the full regulatory 13 14 quent CLAD trial and meta-analysis indi- submission of the CLARITY trial2 was obtained through a FOI cated that the CLAD-associated cancer request (Submitted May 2013, obtained November 2013). The frequency was no different to natural aging data set was provided in portable document format (PDF). Files containing relevant data were identified and converted into Mi- 14 or other pivotal MS-drug trials. This sug- crosoft Excel spreadsheets using a PDF parser developed on gested that even in the absence of oral CLAD, a Python 2.7 platform at the MidPlus computational facilities injectable generic CLAD may still have value at Queen Mary University of London (code available on request). in treating active MS.15,16 The converted data were validated by comparing sample records between PDF and spreadsheet versions of the files. In addition, Although the mechanism of action of CLAD we obtained redacted copies of the regulatory submission of the in MS is unclear,17 efficacy of ALEM has been CARE-MS I trial.4 The data were provided in PDF batches attributed to CD41 T-cell deletion and relative during the third and fourth quarter of 2016. However, primary 9,18 (raw) white cell counts were not included in this package. The sparing of T-regulatory cells and SAI to data presented here were therefore extracted from the tabulated homeostatic T-cell proliferation and lack of documents. 19 thymic repopulation. Although immune- Lymphocyte phenotyping data. In both CLARITY and reconstitution kinetics after ALEM have CARE-MS I, lymphocyte subsets were analyzed using flow cy- been reported,10,18,20 the lymphocyte subset tometry. Both data sets included the following: CD31,CD41 1 1 1 1 of pivotal CLAD/ALEM trials was only par- and CD8 T cells; CD19 B cells; CD16 /CD56 (natural killer [NK]) cells; CD41/CD45RA1 (naive T-helper) cells; and 2,4,5 tially disclosed, yet meeting abstracts CD41/CD45RO1 (memory T) cells. The data were presented as indicated that lymphocyte subset data were absolute numbers/unit volume in both studies. 21,22 collected and analyzed years ago. We Statistical analysis. Statistical analysis comparing 2 or more hypothesized that differences in the CLAD/ unpaired independent nominal variables was performed using ALEM lymphocyte repopulation kinetics x2 test for heterogeneity. If statistical significance was detected x2 may offer insights into the efficacy of CLAD for comparison of more than 2 variables, post hoc test for heterogeneity with Bonferroni correction for multiple compari- and adverse-effect profile of ALEM. sons was applied. Data are represented as mean 6 SD unless described otherwise. For comparison of unpaired independent METHODS Standard protocol approvals, registrations, continuous variables, a 2-tailed Student t test for unpaired sam- and patient consents. Freedom of Information (FOI) re- ples assuming unequal variances was used. For comparison of quests to the European Medicines Agency (EMA) for the full more than 2 unpaired independent continuous variables, 1-way regulatory submissions of phase III “CLAD Tablets Treating analysis of variance (ANOVA) with post hoc Tukey was used. It Multiple Sclerosis Orally” (CLARITY; NCT00213135)2 and was evident that the whole population was not analyzed at every “Comparison of Alemtuzumab and Rebif Efficacy in Multiple time point as indicated in the text; no adjustments for such Sclerosis, study one” (CARE-MS I; NCT00530348)4 trials were missing data were made. made. Although these trials were recruited following ethical approval of the trials and informed consent, as previously re- RESULTS In total, 309/1,326 pwRMS from the 2,4 ported, no specific ethical approval was obtained or required CLARITY trial2 had lymphocyte subsets analyzed to view and use these “public domain” documents. The details (figure 1), with n 5 101 in the placebo arm, n 5 of participants were anonymous. Information relevant to study 5 design, setting, participants, eligibility, variables, randomization, 103 in the CLAD 3.5 mg/kg arm, and n 105 in the blinding, study size, bias reduction, flow diagrams of participants, CLAD 5.25 mg/kg arm (tables e-1 and e-2 at and the elements relating to the CONSORT and STROBE re- Neurology.org/nn). There was no significant differ- porting guidelines can be obtained from the original CLARITY2 ence for any demographics or clinical characteristics 4 and CARE-MS I publications. among the study arms (table e-2). It is important that, Trial designs. The full details of the trials have been reported compared with placebo, both doses of CLAD caused previously.2,4 Briefly, in the 96-week CLARITY trial, people significant and comparable (p 5 0.953, ANOVA post with relapsing MS (pwRMS) were randomized 1:1:1 to receive hoc Tukey) reduction in the annualized relapse rate either placebo or one of 2 doses of oral CLAD. Patients were (relative reduction compared with placebo 55% and given tablets containing either 10 mg/d (60–69.9 kg body 61% for CLAD 3.5 mg/kg and 5.25 mg/kg, respec- weight) or 20 mg/d (70–79 kg body weight) CLAD prodrug administered for 4–5 days in weeks 0 and 5 (year 1) and weeks 48 tively, over 96 weeks) and MS-related MRI parame- and 52 (year 2) to result in a total cumulative dose of 3.5 mg/kg. ters (table e-2). The sample was representative of the Those randomized to the 5.25 mg/kg arm were given additional overall population in the CLARITY study.2 doses in weeks 9 and 13.2 In the CARE-MS I, pwRMS were randomly allocated 1:1 to receive either interferon b-1a (Rebif 44 Full blood count. Following administration of CLAD, mg tiw) or ALEM 12 mg/d on days 1–5 in year 1, followed by 12 only minor and nonsignificant depletion of platelets mg/d on days 1–3 one year later.4 and red blood cells occurred. However, there was

2 Neurology: Neuroimmunology & Neuroinflammation Figure 1 Cladribine targets mainly lymphocytes

Mean number of red blood cells and leucocytes following treatment with either placebo (n 5 42–101. Typically, the lower limit of sample size was n 5 63, except week on 55) or a total doses of either 3.5 mg/kg (n 5 47–103. Typically, the lower limit of sample size was n 5 67, except on week 55) or 5.25 mg/kg (n 5 38–104. Typically, the lower limit of sample size was n 5 62, except on week 55). Placebo (circle) or cladribine (CLAD) that was administered in monthly courses (inverse triangle) at 0, 5 and 48 and 52 weeks (diamond; 3.5 mg per dose) and additionally at 9 and 13 weeks (hexagon; 5.25 mg per dose). The results show the mean 6 SEM of (A) red blood cells (B) platelets (C) white blood cells, (D) lymphocytes, (E) monocytes, (F) polymorphonuclear neutrophils, (G) eosinophils, or (H) basophils.

Neurology: Neuroimmunology & Neuroinflammation 3 significant (p , 0.01) depletion of the total leukocyte depletion of CD191 B cells was very similar with population within a month of the second cycle of both dosing schedules (figure 2B). B-cell numbers both doses of CLAD (figure 1A). As expected, lym- dropped markedly following the first course of phocytes (figure 1B)2 were markedly depleted over CLAD with a nadir (85%–90% depletion) at the the 96-week observation period (p , 0.01) com- time of the second dose of cycle 1. B cells did not pared with more subtle influences on monocytes drop further following administration of additional (figure 1C), polymorphonuclear neutrophils (figure CLAD doses as part of the 5.25 mg/kg schedule and 1D),2 eosinophils (figure 1E), and basophils (figure recovered to 30% of baseline prior to the second 1F). Following lymphocyte depletion with 2 doses of treatment cycle after 48 weeks, which again led to CLAD, there was further depletion following the use significant depletion (80% of baseline; figure 2B). In of 2 additional doses (figure 1B). contrast to the B-cell population, a dose-response effect with CD41 and CD81 T cells was detected Lymphocyte phenotyping. Although reduction of CD31 (figure 2, C and D). In the 3.5 mg/kg group, CD41 T cells showed some dose dependency (figure 2A), T-cell depletion by 20% occurred following the first

Figure 2 Cladribine preferentially depletes B lymphocytes compared with a modest depletion of T cells

The results represent the mean percentage 6 SEM of blood lymphocytes compared with baseline following treatment with either placebo (circle; n 5 56–79) or total doses of either 3.5 mg/kg (diamond; n 5 62–82) or 5.25/kg (hexagon; n 5 66–81) cladribine (CLAD) administered in monthly courses (inverse triangle) at 0, 5 and 48 and 52 weeks (3.5 mg per dose) and additionally at 9 and 13 weeks (5.25 mg per dose). Results show the numbers of (A) CD31 T cells, (B) CD191 B cells, (C) CD41 T cells (D) CD81 T cells, and (E) CD41-naive and (F) CD41 memory T cells.

4 Neurology: Neuroimmunology & Neuroinflammation dose and by 45% following the second dose of cycle cells were depleted by 45% and 60% of baseline 1. This level of depletion was maintained until the following treatment cycles 1 and 2, respectively second cycle of treatment was given, which led to (figure 2E) with reductions in memory T cells being a maximum depletion by 60% of baseline, over the slightly less pronounced (figure 2F). Maximum duration of the study (96 weeks). In the 5.25 mg/kg reduction of naive T cells in the 5.25 mg/kg group group, depletion was more pronounced, by 70% of was nearly 80% depletion from the baseline within baseline after the first cycle, remaining at this level 12 weeks after treatment cycle 1, subsequently re- until the end of the observation period (figure 2C). maining at 70% throughout the remainder of the CD81 T cells followed a rather similar kinetic as study, whereas memory T cells described a similar CD41 T cells, although depletion was overall less curve of depletion and (minor) recovery at 60%–65%. pronounced. In the 3.5 mg/kg group, CD81 T cells Three months after treatment initiation, naive CD81 were reduced by 30% and 40% of baseline following T cells in the 3.5 mg/kg group were reduced by 26.5% treatment cycles 1 and 2, respectively. In the 5.25 and memory CD81 T cells by less than 10% of mg/kg group, CD81 T cells were reduced by 50% of baseline levels. No difference in CD3, CD4, CD8, or baseline after the first treatment cycle with no further CD19 lymphocyte counts was detected between the depletion following the second cycle. Again, similar patient cohorts remaining free of relapses and patients to CD41 T cells, CD81 T cells did not show any who developed at least 1 relapse (figure 3). significant recovery during the observation period (figure 2D). The two CD41 T-cell subsets analyzed Comparison of lymphocyte kinetics after CLAD and (naive, i.e., CD45RA1 and memory, i.e., ALEM administration. Although the demographics of CD45RA2) T cells were both affected and revealed pwRMS were different between CLARITY2 and rather similar kinetics of depletion and recovery CARE-MS I4 studies (table e-1), both CLAD and (figure 2, E and F). In the 3.5 mg/kg group, naive T ALEM exhibited high efficacy.2–5 However, analysis

Figure 3 The incidence of relapse does not relate to peripheral blood CD3, CD4, CD8, CD19, T- and B-cell levels

Cladribine (CLAD) was administered as weekly courses at 0, 5 and 48 and 52 weeks. The results represent the mean 6 SEM absolute number (per cubic millimeter) of peripheral blood: (A) CD3, (B) CD19, (C) CD4, and (D) CD8 lymphocytes following treatment (inverse triangles) with oral CLAD in the CLAD Tablets Treating Multiple Sclerosis Orally (CLARITY) trial administered with 3.5 mg/kg CLAD and those divided into groups who remained healthy (circle; n 5 121–136 per group) or those who had at least 1 relapse (diamond; n 5 29–34). pwRMS 5 people with relapsing MS.

Neurology: Neuroimmunology & Neuroinflammation 5 of lymphocyte subsets following CLAD and ALEM 2, although neither dose ever caused a nadir similar treatment revealed several differences (figure 4). (1) to ALEM (reduction by .90% of baseline). (3) The One month after the first treatment cycle with kinetics of CD81 T-cell depletion were comparable ALEM, CD41 T cells were almost completely (by for both ALEM and CLAD (high and low doses) over 95% of baseline) depleted (figure 4A). Even 1 with their respective CD41 T-cell behavior (figure year after treatment, just before the second treat- 4B). (4) Although ALEM, as well as both doses of ment cycle, CD41 T cells remained depleted by over CLAD, induced depletion of NK cells, this cell type 70% of baseline. The 5.25 mg/kg dose of CLAD led rapidly recovered following ALEM treatment and to a reduction of CD41 T cells by 70% of baseline was even slightly above baseline at 6 and 12 months. within 3 months, with very little recovery until the By contrast, both doses of CLAD caused pro- second treatment cycle nearly 9 months later. (2) nounced and virtually identical NK cell depletion, Withthe3.5mg/kgdoseofCLAD,whichwas by 47% of baseline, within 9 weeks of treatment clinically as effective as the 5.25 mg/kg dose3,4 (table cycle 1 (n 5 71–72) (figure 4C). (5) The most e-2), depletion of CD41 T cells during year 1 was striking difference became evident, comparing the much less pronounced than that with ALEM or effect of ALEM and CLAD on the CD191 B-cell CLAD 5.25 mg/kg. The gap between the CD41 T- population (figure 4D). Although the level of cell depletion curves induced by the 2 different depletion was initially similar between ALEM CLAD dosing schedules only narrowed during year and both doses of CLAD, the repopulation

Figure 4 Depletion of lymphocyte subsets following alemtuzumab and cladribine treatment

Cladribine (CLAD) was administered as weekly courses of CLAD at 0, 5 and 48 and 52 weeks (time of initiation of cycle; inverse triangle) or weekly coursesof alemtuzumab (ALEM) at 0 and 52 weeks. The results represent the mean absolute number of peripheral blood lymphocytes (per cubic millimeter) during the CLAD Tablets Treating Multiple Sclerosis Orally (CLARITY) trial in people with relapsing MS (pwRMS) treated with either placebo (circle; n 5 68–80), CLAD 3.5 mg/kg (diamond; n 5 77–86), CLAD 5.25 mg/kg (hexagon; n 5 79–84), and pwRMS in the Comparison of Alemtuzumab and Rebif Efficacy in Multiple Sclerosis, study one (CARE-MS I) trial treated with 12 mg/d ALEM (square; n 5 171–184).

6 Neurology: Neuroimmunology & Neuroinflammation characteristics were very different. After CLAD, B CLAD that are reported in approximately 90% of cells slowly repopulated, remaining significantly pwRMS treated with ALEM.4,5 below their baseline level until the second treatment Although MS has commonly been considered to cycle reduced this subset yet again. By contrast, B- be a CD41 T cell–mediated (Th1/Th17) disease,23 cell numbers after ALEM repopulated back to a concept supported by the marked (95%) CD41 T- baseline within less than 6 months and then hy- cell depletion induced by ALEM, CD41 T-cell perrepopulated well above baseline by 9 and 12 depletion in isolation has arguably failed to control months (figure 4D). Surprisingly, neither of the MS.24,27 Depletion of CD41 T cells using a CD4- peer-reviewed phase III trial reports of ALEM in MS specific antibody, dosed to maintain CD41 T-cell provided any indication of the latter but only re- numbers above 250 cells/mm3 (equivalent to a deple- ported that B cells reach normal levels 6 months tion of approximately 70%) to limit immunodefi- after drug administration.4,5 ciency thresholds,27 did not effectively suppress the development of new lesions on MRI and was consid- DISCUSSION We analyzed the lymphocyte repo- ered to have failed in MS.27 Based on our analysis, the pulation kinetics following depletion with oral 45%–50% depletion of CD41 T cells achieved by CLAD prodrug and IV ALEM using data sets from CLAD 3.5 mg/kg leaves these cells above the thresh- their regulatory submissions. Using data mining of old required for optimal disease inhibition in both these publicly available documents, it was possible CD4 T cell–mediated experimental autoimmune to obtain additional information not previously encephalomyelitis in animals28 and MS.27 Therefore, published in peer-reviewed papers. This may pro- CD41 T-cell depletion may not account for the vide value for understanding the mechanisms of efficacy of this drug, in terms of relapse reduction and action and side-effect profiles of these disease- its effect on new MRI lesions, given a strong treat- modifying therapies (DMTs), which have focused ment effect of CLAD was already detectable within onthepathogenicroleofTcells.23 However, the the first year of treatment.2,3,13 therapeutic drug–response relationship with CLAD The perceived failure of CD4-depleting monoclo- is perhaps more consistent with lasting B-cell nal antibody led some groups to speculate that CD81 depletion of both doses and helps create a focus on T cells may be the pathogenic drivers of MS as sug- the role of the B cells in control of MS, which may gested by the predominance of this cell type in MS or may not act via T cells.24 The data obtained with lesions.29 Again, given the depletion threshold estab- ALEM are consistent with previous smaller scale lished for CD4 depletion,27,28 we hypothesize that studies.10,25 this is unlikely to be a key mechanism of action of There is unequivocal evidence from the marked CLAD, given CLAD 3.5 mg/kg had only a minor reduction of disease following hematopoietic stem- (15%–28%) CD81 T cell–depleting effect in year cell treatment that relapsing MS is due to aberrant 1, when a strong disease-modifying effect was already function of the immune system.26 CLAD induced evident.3,13 However, again, this comes with the pro- a modest depletion in T and NK cells, but a more viso that an effective threshold of CD81 T-cell deple- marked depletion of B cells. This was consistent with tion required for effective immunotherapy in MS has the known selective lymphocyte-depleting effect yet to be defined. By contrast, ALEM markedly de- because they express high levels of deoxycytidine pletes CD81 T cells. This, coupled with depletion of kinase that phosphorylates CLAD, a deoxyadenosine other immune subsets, may potentially contribute to analog, to cytotoxic 2-chlorodeoxyadenosine triphos- the high number of reported infections that occurred phate.8 Although causality cannot be ascertained, the in 77% of pwRMS participating in CARE-MS I, phase III trial data suggest that lymphocyte depletion compared with only 48% in CLARITY.2,4 In addi- may contribute to the reduction in disease activity in tion, ALEM is more likely to effect CD81 regulatory/ pwRMS.2,3 CLAD also effectively inhibited the evo- suppressor-cell responses that can control tolerance lution from a clinically isolated syndrome of demye- and potential SAI induction.28 lination to definite MS.13 Most of the ongoing disease Although it has been shown that the cancer risk in activity in that trial occurred within the first 3 months pwRMS on CLAD was no different compared with of the study.13 Based on the data presented here, all contemporaneously licensed DMT,14 NK cells maximum depletion using the dosing schemes used represent an important part of cancer immunosurveil- takes between 1 and 3 months. Rebaselining of effi- lance. As CLAD induced a modest depletion of NK cacy data at 3 months after first treatment course cells, vigilance is, therefore, advised over the interme- would have made the suppression of disease activity diate and longer term to truly establish safety, even more impressive.13 Furthermore, the slower whether or not oral CLAD becomes a licensed depletion kinetics of CLAD probably contribute to DMT for pwRMS, or a generic preparation of CLAD the lack of administration-related reactions with is being used off-label.16,30

Neurology: Neuroimmunology & Neuroinflammation 7 CLAD induced a marked and long-lasting CD191 interest that sequestration of immune cells in lym- B-cell depletion that did not reach baseline levels phoid tissue, such as in the spleen and notably bone within the 12-month treatment cycle, a repopulation marrow, following sphingosine-1-phosphate receptor kinetic evidently contrasting with ALEM. Although modulation35 may limit the capacity of ALEM to the capacity of ALEM for hyperrepopulation of B control MS,36 as ALEM may deplete less effectively cells has been reported previously,10,20 it was subse- in lymphoid tissues compared with the blood.36,37 quently largely ignored, notably in the phase III trial Alternatively, it is possible that the important path- reports.4,5 We have recently reported that CD81 T- ogenic cells may not yet have been adequately ana- cell depletion by CD52-specific antibodies can block lyzed, as it is likely that they form only a small immunologic tolerance induction,28 suggesting that it component within the pool of immune cells. Fur- could contribute to the rapid hyperrepopulation of B thermore, it appears that in addition to any T-cell cells that may be the prelude of SAI. Such hyperre- inhibitory activity,23 most agents that inhibit MS population of CD191 B cells is not a feature of treat- deplete memory B cells.24 Relating levels of memory ment with CLAD or CD20-specific B cell–depleting B cells to therapeutic activity have yet to be per- therapies31 neither of which are associated with B formed in MS, but this has been used to personalize cell–driven autoimmunities2,32 or significant T-cell retreatment to limit relapse in other autoimmune depletion.31 Following ALEM, the CD191 B-cell diseases that are sensitive to CD20-depleting anti- hyperrepopulation into the blood is due to the bodies.38,39 Although it is clear that ALEM markedly accumulating immature B cells, probably from the depletes memory B cells,20 the memory B cell– bone marrow, where they differentiate into mature/ depleting capacity of CLAD is unknown. It is naive B cells.20 When this occurs in the absence of T- tempting to speculate that suppression of MS disease cell regulation, previous studies have shown that B- activity by CLAD may relate to long-term depletion cell autoreactivity can develop, and this may be of memory B cells. Further work, including func- a cause of SAI following the use of ALEM in tional studies and cytokine analysis, additional to pwRMS.33,34 Previously, it has been shown that the that already known9,17,18 will be required to fully apparent increase in the number of CD191 B cells, understand the definitive mechanism of action of generated by the overproduction of immature/mature these treatments, as both qualitative and quantitative B cells following ALEM infusion, masks a long- changes in lymphocyte subsets will be important in lasting depletion of CD191/CD271 memory B defining their therapeutic activity.9,18 cells,20 which also occurs at varying levels with other agents that control MS,24,31 suggesting that these cells AUTHOR CONTRIBUTIONS may be important in disease control by DMT.24 Freedom of Information requests: Klaus Schmierer. Concepts: David Baker and Klaus Schmierer. Data extraction: Lukasz Zalewski. Data Anal- However, it must be recognized that there are ysis: David Baker, Samuel S. Herrod, and Cesar Alvarez Gonzalez. Draft- limitations of solely examining peripheral blood ing: David Baker, Samuel S. Herrod, Cesar Alvarez Gonzalez, and Klaus immune subset levels, when it is likely that the path- Schmierer. ogenic cells form only a minor population within any ACKNOWLEDGMENT single subtype. It is therefore perhaps not be surpris- 1 1 1 This research used MidPlus computational facilities at Queen Mary, sup- ing that peripheral blood CD4 ,CD8 ,andCD19 ported by QMUL Research-IT and funded by EPSRC grant EP/ cell levels have not demonstrated biomarker activity K000128/1. The authors thank the European Medicines Agency for sup- to predict disease activation following immune plying the trial documents. reconstitution after CLAD and ALEM treatment,25 suggesting that further functional analysis may be STUDY FUNDING This research received no specific grant from any funding agency in the required to define the mechanisms of action. Fur- public, commercial, or not-for-profit sectors. thermore, without access to all of the primary data, we were unable to directly compare all the subset DISCLOSURE analysis performed to clinical efficacy and safety, so None considered relevant; however, D.B. received a grant from Sanofi their true relationship remains to be established. The Genzyme with the past 3 years. K.S. has been involved in trials spon- sored by, and received meeting support from, Sanofi Genzyme and lack of peripheral blood biomarker activity may also received honoraria from Merck KGaA. Go to Neurology.org/nn for be due to compartmentalization of autoimmunity full disclosure forms. outside the peripheral blood. This could be due to the actions of immune cells sequestered within the Received January 16, 2017. Accepted in final form April 5, 2017. CNS during MS, which trigger disease activity.24 REFERENCES Alternatively, compartmentalization of the immune 1. Scolding N, Barnes D, Cader S, et al. Association of Brit- response within lymphoid tissues may also be a reason ish Neurologists: revised (2015) guidelines for prescribing for the inability to detect a biomarker of autoim- disease-modifying treatments in multiple sclerosis. Pract mune activity within the blood. As such, it is of Neurol 2015;1:273–279.

8 Neurology: Neuroimmunology & Neuroinflammation 2. Giovannoni G, Comi G, Cook S, et al. A placebo- 19. Jones JL, Thompson SA, Loh P, et al. Human autoimmu- controlled trial of oral cladribine for relapsing multiple nity after lymphocyte depletion is caused by homeostatic sclerosis. N Engl J Med 2010;362:416–426. T-cell proliferation. Proc Natl Acad Sci USA 2013;110: 3. Giovannoni G, Cook S, Rammohan K, et al. Sustained 20200–20205. disease-activity-free status in patients with relapsing- 20. Thompson SA, Jones JL, Cox AL, et al. B-cell reconstitu- remitting multiple sclerosis treated with cladribine tablets tion and BAFF after alemtuzumab (CAMPATH-1H) in the CLARITY study: a post-hoc and subgroup analysis. treatment of multiple sclerosis. J Clin Immunol 2010; Lancet Neurol 2011;10:329–337. 30:99–105. 4. Cohen JA, Coles AJ, Arnold DL, et al. Alemtuzumab 21. Rieckmann P, Comi G, Cook S, et al. Effects of cladribine versus interferon beta 1a as first-line treatment for pa- tablets on peripheral lymphocyte subtypes implicated in tients with relapsing-remitting multiple sclerosis: multiple sclerosis immunopathogenesis: surface marker a randomised controlled phase 3 trial. Lancet 2012; analysis for a subset of patients from the 96-week, phase 380:1819–1828. III, double-blind, placebo controlled CLARITY study. 5. Coles AJ, Twyman CL, Arnold DL, et al. Alemtuzumab Mult Scler J 2009;15(suppl 9):S248–S249. for patients with relapsing multiple sclerosis after disease- 22. Hartung HP, Arnold DL, Cohen J, et al; CARE-MS I modifying therapy: a randomised controlled phase 3 trial. Investigators. Lymphocyte subset dynamics following Lancet 2012;380:1829–1839. alemtuzumab treatment in the CARE-MS I study. Mult 6. Tuohy O, Costelloe L, Hill-Cawthorne G, et al. Alemtu- Scler J 2012;18(suppl 4):427–428. zumab treatment of multiple sclerosis: long-term safety 23. Martin R, Sospedra M, Rosito M, et al. Current multiple and efficacy. J Neurol Neurosurg Psychiatry 2015;86: sclerosis treatments have improved our understanding of 208–215. MS autoimmune pathogenesis. Eur J Immunol 2016;46: 7. Levin SN, Kaplan TB. Infectious complications of 2078–2090. novel multiple sclerosis therapies. Curr Infect Dis Rep 24. Baker D, Marta M, Pryce G, et al. Memory B cells are 2017;19:7. major targets for effective immunotherapy in relapsing 8. Warnke C, Leussink VI, Goebels N, et al. Cladribine as multiple sclerosis. EBioMedicine 2017;16:41–50. a therapeutic option in multiple sclerosis. Clin Immunol 25. Kousin-Ezewu O, Azzopardi L, Parker RA, et al. Accel- 2012;142:68–75. erated lymphocyte recovery after alemtuzumab does not 9. Ruck T, Bittner S, Wiendl H, Meuth SG. Alemtuzumab predict multiple sclerosis activity. Neurology 2014;82: in multiple sclerosis: mechanism of action and beyond. Int 2158–2164. J Mol Sci 2015;16:16414–16439. 26. Atkins HL, Bowman M, Allan D, et al. Immunoablation 10. Coles AJ, Wing M, Smith S, et al. Pulsed monoclonal and autologous haemopoietic stem-cell transplantation for antibody treatment and autoimmune thyroid disease in aggressive multiple sclerosis: a multicentre single-group multiple sclerosis. Lancet 1999;354:1691–1695. phase 2 trial. Lancet 2016;388:576–585. 11. Mitosek-Szewczyk K, Tabarkiewicz J, Wilczynska B, et al. 27. van Oosten BW, Lai M, Hodgkinson S, et al. Treatment Impact of cladribine therapy on changes in circulating of multiple sclerosis with the monoclonal anti-CD4 anti- dendritic cell subsets, T cells and B cells in patients with body cM-T412: results of a randomized, double-blind, multiple sclerosis. J Neurol Sci 2013;332:35–40. placebo-controlled, MR-monitored phase II trial. Neurol- 12. Available at: ema.europa.eu/docs/en_GB/document_ ogy 1997;49:351–357. library/Medicine_QA/2011/02/WC500102304.pdf. Ac- 28. von Kutzleben S, Pryce G, Giovannoni G, et al. Depletion cessed March 30, 2017. of CD52-positive cells inhibits the development of CNS 13. Leist TP, Comi G, Cree BA, et al; Oral cladribine for early autoimmune disease, but deletes an immune-tolerance MS (ORACLE MS) Study Group. Effect of oral cladribine promoting CD8 T cell population: implications for sec- on time to conversion to clinically definite multiple scle- ondary autoimmunity of alemtuzumab in multiple sclero- rosis in patients with a first demyelinating event (ORA- sis. Immunology 2017;150:444–445. CLE MS): a phase 3 randomised trial. Lancet Neurol 29. Salou M, Nicol B, Garcia A, Laplaud DA. Involvement of 2014;13:257–267. CD8(1) T cells in multiple sclerosis. Front Immunol 14. Pakpoor J, Disanto G, Altmann DR, et al. No evidence for 2015;6:604. higher risk of cancer in patients with multiple sclerosis 30. Alvarez-Gonzalez C, Allen-Philbey K, Mathews J, et al. taking cladribine. Neurol Neuroimmunol Neuroinflamm Treating multiple sclerosis with generic cladribine. Mult 2015;2:e158. doi: 10.1212/NXI.0000000000000158. Scler J 2016;22(suppl 3):S604–S605. 15. Lublin FD, Reingold SC, Cohen JA, et al. Defining the 31. Palanichamy A, Jahn S, Nickles D, et al. Rituximab effi- clinical course of multiple sclerosis: the 2013 revisions. ciently depletes increased CD20-expressing T cells in mul- Neurology 2014;83:278–286. tiple sclerosis patients. J Immunol 2014;193:580–586. 16. Beutler B, Sipe JC, Romine JS, et al. The treatment of 32. Kappos L, Li D, Calabresi PA, et al. Ocrelizumab in chronic progressive multiple sclerosis with cladribine. Proc relapsing-remitting multiple sclerosis: a phase 2, rand- Natl Acad Sci USA 1996;93:1716–1720. omised, placebo-controlled, multicentre trial. Lancet 17. Korsen M, Bragado Alonso S, Peix L, Bröker BM, Dressel 2011;378:1779–1787. A. Cladribine exposure results in a sustained modulation 33. Meffre E, Wardemann H. B-cell tolerance checkpoints in of the cytokine response in human peripheral blood mono- health and autoimmunity. Curr Opin Immunol 2008;20: nuclear cells. PLoS One 2015;10:e0129182. 632–638. 18. Freedman MS, Kaplan JM, Markovic-Plese S. Insights 34. Kinnunen T, Chamberlain N, Morbach H, et al. Accu- into the mechanisms of the therapeutic efficacy of alemtu- mulation of peripheral autoreactive B cells in the absence zumab in multiple sclerosis. J Clin Cell Immunol 2013;4: of functional human regulatory T cells. Blood 2013;121: 1000152. 1595–1603.

Neurology: Neuroimmunology & Neuroinflammation 9 35. Cyster JG, Schwab SR. Sphingosine-1-phosphate and lym- 38. Kim SH, Kim W, Li XF, et al. Repeated treatment with phocyte egress from lymphoid organs. Annu Rev Immunol rituximab based on the assessment of peripheral circu- 2012;30:69–94. lating memory B cells in patients with relapsing neuro- 36. Willis M, Pearson O, Illes Z, et al. An observational study myelitis optica over 2 years. Arch Neurol 2011;68: of alemtuzumab following fingolimod for multiple sclero- 1412–1420. sis. Neurol Neuroimmunol Neuroinflamm 2017;4:e320. 39. Trouvin AP, Jacquot S, Grigioni S, et al. Usefulness of doi: 10.1212/NXI.0000000000000320. monitoring of B cell depletion in rituximab-treated rheu- 37. Hu Y, Turner MJ, Shields J, et al. Investigation of the matoid arthritis patients in order to predict clinical relapse: mechanism of action of alemtuzumab in a human CD52 a prospective observational study. Clin Exp Immunol transgenic mouse model. Immunology 2009;128:260–270. 2015;180:11–18.

10 Neurology: Neuroimmunology & Neuroinflammation IgG-specific cell-based assay detects potentially pathogenic MuSK-Abs in seronegative MG

Saif Huda, MD ABSTRACT Patrick Waters, PhD Objective: To increase the detection of MuSK-Abs using a CBA and test their pathogenicity. Mark Woodhall, PhD Methods: Sera from 69 MuSK-RIA–positive patients with myasthenia gravis (MG) (Definite Maria Isabel Leite, DPhil MuSK-MG), 169 patients negative for MuSK-RIA and AChR-RIA (seronegative MG, SNMG), 35 Leslie Jacobson, DPhil healthy individuals (healthy controls, HCs), and 16 NMDA receptor-Ab–positive (NMDAR-Ab) Anna De Rosa, MD disease controls were tested for binding to MuSK on a CBA using different secondary Michelangelo Maestri, MD antibodies. Roberta Ricciardi, MD Jeannine M. Heckmann, Results: Initially, in addition to 18% of SNMG sera, 11% of HC and 19% of NMDAR-Ab sera PhD showed positive binding to MuSK-transfected cells; this low specificity was due to anti-IgG Angelina Maniaol, PhD (H1L) detection of IgM bound nonspecifically to MuSK. Using an IgG Fc gamma-specific second- Amelia Evoli, MD ary antibody, MuSK-Abs were detected by CBA in 68/69 (99%) of Definite MuSK-MG, 0/35 HCs, Judy Cossins, DPhil 0/16 NMDAR-Ab, and 14/169 (8%) of SNMG sera, providing increased sensitivity with high David Hilton-Jones, MD specificity. The RIA-negative, CBA-positive MuSK-IgG sera, but not IgM-MuSK–binding sera, Angela Vincent, FRCPath reduced agrin-induced AChR clustering in C2C12 myotubes, qualitatively similar to RIA-positive MuSK-Abs. Conclusions: An IgG-specific MuSK-CBA can reliably detect IgG MuSK-Abs and increase sensitiv- Correspondence to ity. In the MuSK-CBA, IgG specificity is essential. The positive sera demonstrated pathogenic Prof. Vincent: potential in the in vitro AChR-clustering assay, although less effective than Definite MuSK-MG [email protected] sera, and the patients had less severe clinical disease. Use of IgG-specific secondary antibodies may improve the results of other antibody tests. Classification of evidence: This study provides Class III evidence that an IgG-specific MuSK-CBA identifies patients with MG. Neurol Neuroimmunol Neuroinflamm 2017;4:e357; doi: 10.1212/ NXI.0000000000000357

GLOSSARY AChEI 5 acetylcholinesterase inhibitor; CBA 5 cell-based assay; CI 5 confidence interval; DMEM 5 Dulbecco Modified Eagle Medium; FCS 5 fetal calf serum; HC 5 healthy control; HEK 5 human embryonic kidney; MG 5 myasthenia gravis; MGFA 5 Myasthenia Gravis Foundation of America; PSA 5 Penicillin, Streptomycin, and Amphotericin; RT 5 room temperature; SNMG 5 seronegative MG.

Several methods are available for the detection of antigen-specific antibodies (Abs) in the serum or CSF of patients with autoantibody-mediated CNS and peripheral nervous system diseases.1 In the case of suspected autoimmune myasthenia gravis (MG), sera are routinely tested by radioimmunoprecipitation assays (RIAs) for Abs to AChR or MuSK. However, indirect immunofluorescence on live cells transiently transfected with AChRs and clustered by rapsyn, as they are at the neuromuscular junction, increases the detection of AChR-IgG2–5; CBAs have been found to be sensitive and specific for many antibodies, e.g., for aquaporin-4 (AQP-4) antibodies in patients with neuromyelitis optica.6 Here, to see whether a CBA might increase sensitivity for Supplemental data at Neurology.org/nn From the Nuffield Department of Clinical Neurosciences (S.H., P.W., M.W., M.I.L., L.J., J.C., D.H.-J., A.V.), University of Oxford, UK; Department of Clinical and Experimental Medicine (A.D.R., M.M., R.R.), Neurology Unit, Pisa, Italy; Division of Neurology (J.M.H.), University of Cape Town, South Africa; Oslo University Hospital (A.M.), Norway; and Department of Neuroscience (A.E.), Catholic University, Rome, Italy. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the Nuffield Department of Clinical Neurosciences, Oxford University. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 MuSK-Abs, we established a specific MuSK- Twenty-four hours later, coverslips were transferred to a 24- Ab assay, tested previously MuSK-Ab–negative well cell culture plate and incubated with human sera diluted 1:20 in assay buffer (DMEM, 20 mM HEPES, 1% bovine serum patients, and evaluated the pathogenic poten- albumin) for 1 hour at room temperature (RT). After washing tial of the antibodies detected. and fixing in 3% formaldehyde, coverslips were washed and incubated with one of the following secondary Abs in assay buffer METHODS Primary research question. Can a cell-based as appropriate: goat anti-human IgG Alexa Fluor 568 (Invitrogen, assay reliably enhance the detection of MuSK antibodies? binds both heavy and light chains), goat anti-human IgG Fc(g) (Thermo Fisher Scientific, Waltham, MA), goat anti-human IgM Classification of evidence. This study provides Class III evi- Alexa Fluor 594 (Invitrogen), or IgG subclass–specific anti- dence that an IgG-specific MuSK-CBA identifies patients with human IgG1, IgG2, IgG3, or IgG4 (Sigma-Aldrich). For the MG. The optimized MuSK-CBA had a sensitivity of 99% (95% IgG subclass and Fc(g) assays, the secondary Abs were not fluo- – confidence interval [CI] 92.2 100) and a specificity of 100% rescently labeled, and a third layer of goat anti-mouse Alexa Fluor – (95% CI 93.0 100). 568 was used. After final washing, the coverslips were mounted Standard protocol approvals, registrations, and patient on mounting media (DakoCytomation, Cambridge, UK) with 9 9 consents. The use of patient sera was approved by the Oxford- 1% DAPI (4 ,6 -diamidino-2-phenylindole dichloride). Slides – shire Research Ethics Committee A (07 Q160X/28). were read the following day with an Axion Zeiss inverted fluorescent microscope and all photographs taken under identical 2 Patient sera. All sera had been stored at 20°C sera. During conditions with a MacProbe v4.3 digital image system. use, aliquots were kept at 4°C to avoid repeated freeze/thaw cycles. Archived sera from MuSK-IgG–positive RIA-definite Cell-based assay scoring. Scoring of clustered AChR, LRP4, MG (Definite MuSK-MG; n 5 69), MuSK- and AChR-RIA– and MuSK-CBAs was performed by 2 masked observers as negative MG (seronegative MG [SNMG]; n 5 169), NMDA previously described.2 Binding of the red fluorescent–labeled receptor-Ab–positive (NMDAR-Ab; n 5 16), and aquaporin-4- secondary antibodies was scored based on the degree of cell Ab–positive (AQP-4-Ab; n 5 10) patients and healthy controls surface fluorescence and colocalization with EFGP-labeled (HCs; n 5 35) were studied. Definite MuSK-MG sera were AChRs or MuSK. Complementary DNAs used for the LRP4 from Italy (Pisa). The SNMG sera were from cohorts (Italy CBA did not contain EGFP, and only the cell surface staining was [Rome] 5 16, Norway 5 50, United Kingdom 5 32, and South scored. Nonspecific binding was excluded when sera positive for 1 Africa 5 33; Germany 5 5, Turkey 5 7, Japan 5 3, South antigen were negative for the 2 other antigens. If sera bound 2 Korea 5 12, and Philippines 5 11). All patients had been seen antigens, negativity on the third antigen was used to exclude by MG specialists who provided brief clinical features, and nonspecific binding to HEK cells. If sera bound 3 antigens, supportive features of (1) neurophysiology (evidence of decre- nonspecific binding was assessed using colocalization and ment on repetitive nerve stimulation [decrement of fourth preadsorption studies (see below). Mean end-point titers between CMAP amplitude greater than 10% of baseline value]) and/or anti-human IgG Fc(g) and anti-human IgG(H1L) were deter- neuromuscular jitter on single-fiber EMG, (2) treatment mined by identifying the highest dilution at which the serum response to cholinesterase inhibitors, and (3) treatment response resulted in a score of 1. to immunotherapy in 132/169 (78%) cases. Colocalization. To confirm specific binding to extracellular Tissue culture. Human embryonic kidney-293 (HEK, derived MuSK, following incubation with human sera, the coverslips from the European Collection of Authenticated Cell Cultures) were washed and incubated with commercial polyclonal goat cells were grown in Dulbecco Modified Eagle Medium (DMEM) anti-MuSK-Ab AF562 (R&D Systems, Minneapolis, MN) for supplemented with 10% fetal calf serum (FCS) (Sigma-Aldrich, 1 hour at RT. After washing and fixing in 3% formaldehyde, the St. Louis, MO) and 1% each of Penicillin, Streptomycin, and coverslips were washed and incubated with goat anti-human IgG Amphotericin B (PSA, Invitrogen, Carlsbad, CA) at 37°C in an Alexa Fluor 488 (green, Invitrogen) and rabbit anti-goat IgG atmosphere of 5% CO2. C2C12 myoblasts were grown in growth Alexa Fluor 568 (red, Invitrogen). Coverslips were then washed medium (DMEM supplemented with 15% FCS and 1% PSA) and prepared for analysis as previously described. and differentiated for 5–6 days in differentiation medium (DMEM supplemented with 2% FCS and 1% PSA) at 37°C in Preadsorption. For preadsorption studies, the highest positive

an atmosphere of 8% CO2. SNMG sera (1:10 dilution) were adsorbed sequentially thrice against 9 3 106 live untransfected HEK cells or MuSK- Cell-based assays. HEK-293 cells were detached using trypsin transfected HEK cells in solution for 1 hour at RT. The and centrifuged for 5 minutes at 1,100g. The cells were resus- adsorbed sample (equivalent to 1:20 preadsorption) was tested by 3 5 pended, counted, and plated at a density of 2 10 cells/well on MuSK-CBA to confirm adsorption. poly-L-lysine–coated 13-mm glass coverslips in 6-well cell culture plates. For the MuSK-CBA, 3 mg of MuSK-EGFP comple- Agrin production and AChR-clustering assays. MuSK-Abs mentary DNA was transfected into HEK cells. For the clustered inhibit agrin-induced AChR clustering, as previously described.7,8 AChR (clustered-AChR) assay, cells were transiently For the production of agrin, a T175 flask of HEK-293 cells was cotransfected, using polyethylenimine, with plasmids encoding transfected with neural agrin (originally donated by the late Dr. the 4 subunits of human adult AChR and rapsyn-EFGP (with Werner Hoch). After 24 hours, the culture medium was changed, a total of 3 mgofa:b:d:e:rapsyn at 2:1:1:1:1, respectively).2 For and after 48 hours, the conditioned medium was centrifuged at the LRP4 CBA, LRP4, covalently linked to the transmembrane 1,200g for 10 minutes at RT, aliquoted, and stored at 220°C. To and cytosolic domains of CASPR2 (to preserve transmembrane demonstrate inhibition of AChR clustering by the samples, the positioning) and low-density lipoprotein receptor–related sera were heat inactivated at 55°C for 30 minutes, dialyzed, and protein-associated protein 1 pcDNA3.1 were cotransfected with filter sterilized before use. C2C12 myotubes were incubated with a 5:1 ratio (6 mg). In each case, the medium was changed 16 patient sera (1:10) for 30 minutes followed by 1:1,000 agrin in hours posttransfection. differentiation medium for 16 hours. Samples were masked

2 Neurology: Neuroimmunology & Neuroinflammation before application. AChR clusters were labeled using Alexa Fluor 1 showed low IgG2 subclass binding. Moreover, the 594-conjugated a-bungarotoxin (Invitrogen) diluted 1:1,000 in CBA scores for IgM binding to MuSK correlated differentiation media for 60 minutes at 37°C, 8% CO2. My- broadly with the results obtained with anti-human otubes were then washed in differentiation media, fixed in 3% IgG(H1L) (R 5 0.5, p , 0.0001; data not formaldehyde, and washed and stored in phosphate-buffered saline at 4°C. Twenty fields containing myotubes were selected shown). These results suggested that much of the with bright field and red fluorescent images were taken using MuSK-CBA reactivity identified with anti-human SimplePCI (Digital Pixel) software. Images were analyzed blind IgG(H1L) was due to detection of IgM and not for AChR cluster number using a macro with ImageJ software. IgG antibodies binding to MuSK.

Statistics. Graphs and statistical analysis were performed with Optimization of anti-IgG Fc(g). The data above con- GraphPad Prism version 6.0A and MedCalc 16.4.3. End-point firm previous observations that anti-human IgG titers were compared using the Wilcoxon matched-pairs signed- 1 rank test. Correlation coefficients were assessed using the (H L) binds not only to IgG heavy chains but also Spearman correlation coefficient and an R value. Clinical data cross-reacts with different Igs via the light chains were compared using a 2-tailed Fisher exact test and the Mann- which are not Ig class specific.10 To improve the CBA Whitney U test. for the detection of IgG-MuSK antibodies, we replaced the anti-human IgG(H1L) secondary anti- 1 RESULTS CBA with anti-human IgG(H L). To assess body with unlabeled goat anti-human IgG Fc(g) and the sensitivity of the MuSK-CBA, we first used a goat detected its binding to human antibodies with a ter- 1 anti-human IgG(H L) Alexa Fluor 568 (Invitrogen) tiary rabbit anti-goat IgG(H1L) Alexa Fluor 568 as a secondary antibody to detect binding of IgG antibody. antibodies. Surprisingly, with this secondary anti- A representative CBA image of a Definite MuSK- body, only 54/69 Definite MuSK-MG (RIA-positive) MG serum is shown in figure 2B. Four representative patients bound detectably (sensitivity 78.3%). Definite MuSK-MG sera demonstrated higher visual Moreover, the specificity of the assays was poor with scores, with higher end-point dilution titers, with 4/35 (11%) HCs and 3/16 NMDAR-Ab sera (19%) anti-IgG Fc(g) compared with anti-IgG(H1L) demonstrating positive binding (specificity 86.3%). (p 5 0.0005 Wilcoxon matched-pairs signed-rank With this assay, MuSK-Abs were also detected in 19/ test) (figure 2C). With this improved IgG-specific 104 SNMG sera (18%), but because of the poor method, we found IgG MuSK-Abs in 68/69 (sensi- specificity, these initial results were unreliable (figure tivity 99% [95% CI 92.2–100]) of the Definite “ 1A). Of interest, even with the apparently false- MuSK-MG patients (the one negative sample ” positive results, a positive HC serum binding [MuSK-RIA titer 0.91 nM] bound nonspecifically colocalized with MuSK-GFP surface expression to HEK cells by CBA and because of limited volume (figure 1B), suggesting binding of the patient for further examination was designated negative). immunoglobulins to MuSK. Moreover, the 35 HC sera and 16 NMDAR- To determine if this was also the case with SNMG antibody–positive sera were negative for MuSK-IgG, – sera, a sample that scored 3 4 on the MuSK-CBA was demonstrating 100% specificity. We subsequently selected for further study. Binding of this serum tested 10 AQP4-antibody–positive sera that were immunoglobulin colocalized with surface labeling of also negative. Finally, 14/169 (8%) SNMG, includ- MuSK using a commercial MuSK-Ab, as expected for ing 6 of the 19 that were positive with anti-IgG an antibody to MuSK (figure 1C). Moreover, this (H1L), were positive for IgG-specific MuSK-Abs reactivity was adsorbed by incubation with MuSK- (figure 2D). transfected HEK cells (but not with untransfected In addition, 18/169 (11%) sera were positive for HEK cells) confirming that the binding was specific clustered AChR-Abs, 1/169 for LRP4-Abs, and 1/ for MuSK (data not shown). 169 for both clustered AChR-Abs and LRP4-Abs. A flowchart summarizing all results of CBA testing Subclasses and isotype. The high proportion and path- with anti-IgG(H1L) or anti-IgG Fc(g) is shown in ogenic potential of MuSK-IgG4 Abs are a recognized figure e-1 at Neurology.org/nn. feature in MuSK-MG.9 We used subclass and IgM- specific secondary antibodies2,8 to study 12 SNMG Functional relevance of IgG and IgM-MuSK antibodies. sera (of sufficient remaining volume) positive for MuSK-MG sera inhibit the agrin/MuSK/LRP4/ MuSK-Abs by anti-human IgG(H1L), comparing DOK7 pathway (figure 3A).7,8,11 Six SNMG sera with 2 Definite MuSK-MG sera. The latter were with IgG-specific MuSK-Abs (figure 3B), but not 8 mainly of IgG class with all subclasses variably rep- with only IgM nonspecific MuSK-Abs (figure 3C), resented at the low dilution (1:20) tested (figure 2A). reduced the number of agrin-induced AChR clusters By contrast, 11/12 SNMG sera positive with anti- on C2C12 myotubes, suggesting that only the IgG- IgG(H1L) showed strong evidence of IgM binding specific Abs had pathogenic potential. Nevertheless, to MuSK but no IgG antibodies (figure 2A), and only they were clearly less effective than the 2 Definite

Neurology: Neuroimmunology & Neuroinflammation 3 Figure 1 CBA with anti-IgG(H1L)

(A) Scatter plots of results from patients and controls. (B) Representative CBA images from SNMG and HC sera. Scale bar 5 50 mM. (C) Colocalization of commercial MuSK-AF562 antibody (red) with anti-IgG(H1L)(green)inanSNMGserum. Scale bar 5 20 mM. CBA 5 cell-based assay; HC 5 healthy control; MG 5 myasthenia gravis; SNMG 5 seronegative MG.

MuSK-MG samples tested in parallel that abolished cyclophosphamide, or rituximab. Postintervention AChR clusters (figure 3, A–C, figure e-1). status suggested a favorable response to immunotherapy in all patients, and remission was achieved in Comparison of MuSK-CBA and RIA phenotypes. The 3/11 (27%) patients. country of origin and individual clinical data of Table 2 compares the clinical features of MuSK- MuSK-Ab CBA–positive (RIA-negative) patients are Ab CBA positive with Definite MuSK-MG patients. summarized in table 1. Patients were predominantly A female preponderance was noted in both groups female (3.3:1) and presented at a median age of 25 (3.3:1 vs 9:1; p 5 0.18), but the median age at onset years (range 16–79 years). At presentation to the was later in Definite MuSK-MG patients (40 years neurologists, 8 patients had disease confined to the [range 12–82 years] compared with 25 years [range ocular muscles (median duration of follow-up 2 years 16–79 years]; p 5 0.01). Generalized MG was re- [range 1–4.5 years]), and Myasthenia Gravis Foun- ported in all 69 Definite MuSK-MG patients, com- dation of America (MGFA) grades were #2 in 11/13 pared with 8/13 of MuSK-CBA–positive patients (p (85%) patients. Patients were treated with therapies , 0.0001), with only 18/67 (27%) MGFA #2at including acetylcholinesterase inhibitors (AChEIs) onset (p 5 0.0001). Neurophysiology was only only, or combinations with prednisone, azathioprine, positive in 3/6 examined MuSK-CBA patients.

4 Neurology: Neuroimmunology & Neuroinflammation Figure 2 Improving the CBA with anti-IgG Fc(g)

(A) Antibodies in 2 Definite MuSK-MG sera were mainly IgG subclasses, particularly IgG1 and IgG4. By contrast, 12 SNMG sera detected with anti-IgG(H1L) were also detected with anti-IgM secondary antibody but not with anti-IgG subclass secondary antibodies. (B) Representative CBA images of a Definite MuSK-MG serum detected with anti-IgG Fc(g). Scale bar 5 50 mM. (C) End-point titrations in 4 Definite MuSK-MG sera were higher with IgG Fc(g)vsIgG(H1L). (D) Scatter plots of results from the SNMG patients and 51 disease and healthy controls. CBA 5 cell-based assay; MG 5 myasthenia gravis; SNMG 5 seronegative MG.

Reflecting the relatively mild phenotypes, ACHEI the diagnosis of MG. Antibodies binding to clustered was effective in 75% of MuSK-CBA patients, com- AChRs on CBAs have improved the diagnosis of pared with only 20% of Definite MuSK-MG a proportion of previously AChR-Ab–negative patients, but it is important that immunotherapies patients, and CBAs for LRP4 antibodies are being appeared equally effective in both groups, and final used by several research laboratories,2,4,12 many using disease status did not differ, although follow-up times anti-human IgG(H1L) to detect the antibodies were clearly different. Here, we found it necessary to optimize a CBA for MuSK-Abs to avoid detecting bound IgM which gave DISCUSSION Cell-based assays are now commer- poor disease specificity. Using an unlabeled secondary cially available for the detection of some CNS anti- antibody specific for the Fc(g) chain of IgG, and bodies but have not yet been developed widely for applying a tertiary fluorescent antibody for detection,

Neurology: Neuroimmunology & Neuroinflammation 5 Figure 3 Functional effects of MuSK-IgG and IgM on agrin-LRP4-MuSK–clustering pathway

(A) Representative images of myotubes, fluorescent labeled for AChR clusters in the presence of different sera. Definite MuSK-MG and CBA MuSK-IgG–positive sera reduced the number of AChR clusters, but CBA MuSK-IgM sera did not reduce clusters. Scale bar 5 50 mM. (B) Mean results from 2 experiments of CBA MuSK-IgG and control sera on AChR clusters. (C) Mean results of 2 experiments with CBA MuSK-IgM–positive and control sera. Values shown are mean 1 SEM. HC 5 healthy control; MG 5 myasthenia gravis.

we increased MuSK-Ab detection in Definite MuSK- The patients with MuSK-CBA antibodies were MG patients, while avoiding the detection of the mainly females but younger than the Definite nonspecific IgM bound to MuSK, which had initially MuSK-MG group (positive by RIA) and had milder occurred in 11% and 19% of healthy and disease disease. After a median follow-up of 2 years, within control samples, respectively. Overall, the IgG- which the risk of secondary generalization is highest,13 specific MuSK-CBA detected antibodies in 8% of ocular MG still compromised 39% of these patients. previously negative MG samples, and the clinical All patients treated responded to immunotherapy, relevance of these MuSK-IgG antibodies was but a beneficial response was also seen in 75% of demonstrated by inhibition of the agrin-LRP4- patients treated with AChEIs, either alone or in MuSK–clustering pathway, which did not occur in combination with other therapies. By contrast, 80% the presence of the samples that only contained IgM- of the Definite MuSK-MG cohort responded poorly MuSK reactivity. Clustered AChR antibodies were or with cholinergic side effects in keeping with detected in another 11% of patients, but only 1 previous reports.14 MuSK antibodies by RIA are often additional patient had LRP4 antibodies detected by associated with a particularly severe bulbar form of the same IgG-specific approach. MG,15 but it is well recognized that not all patients

6 Neurology: Neuroimmunology & Neuroinflammation Table 1 Clinical data from 13 patients only positive on IgG-specific MuSK-CBA

Age at MGFA MGFA last MuSK-CBA positive/country Sex onset, y onset MGFA Max visit Treatment AChEI (1)PIS

Patient 1/Germany F 42 IIa IIa IIa P — I

Patient 2/Germany F 19 I IIa 0 ACHEI, P, A — PR

Patient 3/Philippines M 21 I I I ChI, P Yes —

Patient 5/UK F 41 IIIb IIIb IIa Previous ACHEI, now P and MMF Yes MM

Patient 6/Italy F 16 I IIb IIb ACHEI, previous P No MM

Patient 7/Norway F 21 I I 0 ACHEI — PR

Patient 8/Norway M 79 I I I ACHEI, P ——

Patient 9/Norway F 20 I I I ACHEI Yes MM

Patient 10/Norway F 27 I I 0 ACHEI, previous P and A, then Ritux — CSR

Patient 11/South Africa F 37 IVb IVb IVb Previous Cyclph, A, IVIG now Cyclosp, P Yes I

Patient 12/South Africa F 31 IIa IIa IIa A, P Yes MM

Patient 13/South Africa M 25 IIa IIa I MTX, P No MM

Patient 14/South Africa F 25 I V IIIa A, previous PLEX and IVIG Yes MM

Abbreviations: — 5 data not available; A 5 azathioprine; ACHEI 5 acetylcholinesterase inhibitor; CBA 5 cell-based assay; CSR 5 complete stable remission; Cyclosp 5 cyclosporine; Cyclph 5 cyclophosphamide; I 5 improved; MGFA 5 Myasthenia Gravis Foundation of America; MM 5 minimal manifestation; MMF 5 mycophenolate mofetil; MTX 5 methotrexate; P 5 prednisone; PIS 5 postintervention status; PLEX 5 plasma exchange; PR 5 pharmacologic remission; Ritux 5 rituximab. Patient 4 had insufficient clinical data and has been omitted from both tables.

have severe disease, and it appears that the MuSK- function in vivo, the lack of disease specificity argues CBA–positive patients are within the mild spectrum. against it. IgM is normally in the form of pentamers The IgM bound to MuSK in SNMG sera was with up to ten binding sites. One possible explanation specific for MuSK by colocalization and preadsorp- for IgM detection in these assays could be the tion experiments. However, the MuSK-IgM sera nonantigen-specific binding of IgM pentamers to tested did not inhibit the formation of agrin- certain types of protein expressed at high levels on induced clusters. Although this does not exclude the cells. Since similar problems have been found with possibility that MuSK-IgM may act on MuSK myelin oligodendroglial antibodies,10 it may be due to the extracellular immunoglobulin-like domains and/or glycosylation state of these particular proteins

Table 2 Comparison between MuSK-Ab CBA–positive and Definite MuSK-MG and suggests that caution should be used regarding patients the use of anti-IgG(H1L) for the detection of IgG binding to similar antigens. Indeed, the Oxford MuSK-CBA positive MuSK-RIA positive (n 5 13) (n 5 69) p Value laboratory now screens all diagnostic sera with anti- IgG(H1L) and then confirms IgG with the IgG- Female:male 3.3:1 9:1 0.1816 specific secondary and fluorescence-labeled tertiary a Age at onset, y, median (range) 25 (16–79) 40 (12–82) 0.0105 antibody. This approach both improves specificity Generalized:ocular 8:5 69:0 ,0.0001 and the use of a tertiary layer increases a little the MGFA onset £2, n (%) 11/13 (84.6) 18/67 (26.9) 0.0001 fluorescent signal obtained.

MGFA follow-up £2, n (%) 11/13 (84.6) 58/69 (84.1) 1.000 A large multinational study reported the detection

Neurophysiology positive, n (%) 3/6 (50) ——of MuSK-Abs in 13% (83/633) of seronegative patients with MG using a CBA.12 Further testing of Treatment, n (%) 25 sera (positive by MuSK-CBA) showed that all 1 ACHEI response ( ) 6/8 (75) 14/69 (20) 0.0032 contained MuSK-IgM, but of these, only 2/25 Immunotherapy response 11/11 (100) 67/69 (97) 1.000 (8%) also contained MuSK-IgG Abs, very similar to Postintervention status our results; that study may also have benefitted using No. with remission (%) 3/11 (27.3) 26/69 (37.7) 0.7376 a more IgG-specific secondary antibody. MuSK-Abs

Median follow-up, y (range) 2(1–4.5) 8 (2–31) ,0.0001 have also been reported in SNMG cohorts from Sri Lanka (3/10; 10%) and China (3/8; 38%), but in 5 5 5 Abbreviations: AChEI acetylcholinesterase inhibitor; MG myasthenia gravis; MGFA both these previous studies, an anti-human IgG Myasthenia Gravis Foundation of America. 1 16,17 Unless otherwise specified all statistics were performed with the Fisher exact test. (H L) antibody was used. The results, therefore, a Mann-Whitney test. may need to be reassessed.

Neurology: Neuroimmunology & Neuroinflammation 7 This study has limitations. First, archived sera DISCLOSURE stored at 220°C may have degraded over time and S. Huda served on the scientific advisory board for Medefield. with prior freeze-thaw cycles, and many samples were P. Waters received honoraria from Biogen Idec Japan, Euroimmun AG, and Mereo Biopharma; holds a patent for and receives royalties from obtained after immunotherapies. Second, because of assays for the detection of antibodies to LGI1, CASPR2, Contactin2 limited volumes of sera, experiments to determine and GABAAR; his laboratory runs routine diagnostic assays, one of the functional relevance of MuSK-IgG antibodies by which is for GABAAR antibodies; and received research support from Euroimmun AG. M. Woodhall reports no disclosures. M.I. Leite CBA were prioritized over the determination of IgG received travel funding and speaker honoraria from Biogen Idec; subclass. It would be interesting in future studies to received a travel grant from Novartis; served on the editorial board determine if these antibodies are predominantly of the for Neuromuscular Disorders; and received support from NHS National IgG4 subclass and capable of inhibiting the coimmu- Specialised Commissioning Group for Neuromyelitis Optica UK and NIHR Oxford Biomedical Research Centre. L. Jacobson, A. De Rosa, noprecipitation of MuSK and LRP4 as shown for M. Maestri, and R. Ricciardi report no disclosures. J.M. Heckmann Definite MuSK-MG samples.8 Ultimately, only received research support from South Africa Medical Research Council auditing the results of routine diagnostic service testing and AFM Telethon. A. Maniaol received research support from The with the improved assay will demonstrate unequivo- Norwegian Association of Muscle Diseases. A. Evoli served as a scientific award jury member for Grifols. J. Cossins reports no disclosures. cally the clinical utility of this IgG-specific MuSK- D. Hilton-Jones served on the editorial board for Neuromuscular Dis- CBA. orders, Practical Neurology and receives royalties from Churchill Living- With characteristic clinical features, a diagnosis of stone and Cambridge University Press. A. Vincent received travel funding and speaker honoraria from Bial Commercial; served as an SNMG may be relatively straightforward despite associate editor for Brain, editorial board member for Neurology®; holds reduced neurophysiologic sensitivity, as for instance a patent with Oxford University for MuSK antibodies and for LGI1/ in ocular MG. By contrast, clinically “ambiguous” CASPR2 antibodies; received royalties from Blackwell Publishing, Mac cases with unsupportive neurophysiology may be Keith Press, and Elsevier; consulted for Athena Diagnostics; and has received research support from Athena Diagnostics, Euroimmun AG, due to MG or an entirely different pathology. In these and NIHR. Go to Neurology.org/nn for full disclosure forms. instances, the absence or detection of an antibody can have important implications for the diagnostic Received November 19, 2016. Accepted in final form April 6, 2017. process and management. We suggest that in addition to the established clustered AChR and LRP4 CBAs, REFERENCES which are beginning to be used more widely, RIA- 1. Waters P, Pettinghill P, Lang B. Detection methods for neural antibodies. In: Pittock S, Vincent A, editors. negative MG sera should be tested for IgG-specific Autoimmune Neurology. Amsterdam: Elsevier; 2016: antibodies by MuSK-CBA. Confirmation of positive 147–163. results with an IgG-specific test, as described here, 2. Leite MI, Jacob S, Viegas S, et al. IgG1 antibodies to may also improve the sensitivity and specificity of acetylcholine receptors in “seronegative” myasthenia other antibody assays. gravis. Brain 2008;13:1940–1952. 3. Jacob S, Viegas S, Leite MI, et al. Presence and pathogenic AUTHOR CONTRIBUTIONS relevance of antibodies to clustered acetylcholine receptor in ocular and generalised myasthenia gravis. Arch Neurol Saif Huda: drafting/revising the manuscript, study concept and design, – acquisition of data, statistical analysis, and interpretation of data. Patrick 2012;69:994 1001. Waters: drafting/revising manuscript, study concept and design, and 4. Devic P, Petiot P, Simonet T, et al. Antibodies to clustered interpretation of data. Mark Woodhall: drafting/revising the manuscript, acetylcholine receptor: expanding the phenotype. Eur J study concept and design, and acquisition of data. Maria Isabel Leite, Neurol 2014;21:130–134. Leslie Jacobson, Anna De Rosa, Michelangelo Maestri, Roberta Ricciardi, 5. Rodriguez Cruz PM, Al-Hajjar M, Huda S, et al. Clinical Jeannine M. Heckmann, Angelina Maniaol, Amelia Evoli, Judy Cossins, features and diagnostic usefulness of antibodies to clus- and David Hilton-Jones: drafting/revising the manuscript and acquisition tered acetylcholine receptors in the diagnosis of seronega- of data. Angela Vincent: drafting/revising the manuscript, study concept tive myasthenia gravis. JAMA Neurol 2015;72:642–649. and design, analysis and interpretation of data, and contribution of vital 6. Waters P, Riendl M, Saiz A, et al. Multicentre comparison of reagents/tools/patients and study supervision. All authors provided final a diagnostic assay: aquaporin-4 antibodies in neuromyelitis approval of the version to be published and took responsibility for the – conduct of the research. optica. J Neurol Neurosurg Psychiatry 2015;87:1005 1015. 7. Hoch W, McConville J, Helms S, Newsom-Davis J, Melms A, Vincent A. Auto-antibodies to the receptor tyro- ACKNOWLEDGMENT sine kinase MuSK in patients with myasthenia gravis with- The authors thank Dr. Jonathan Cheung for developing the ImageJ macro out acetylcholine receptor antibodies. Nat Med 2001;7: used for automated counting of AChR clusters. They also thank – Dr. Arthur Melms, Dr. Feza Deymeer, Dr. Motomura Masa, Dr. Sung 365 368. Jung-Joon, and Dr. Raymond Rosales for providing some of the SNMG 8. Koneczny I, Cossins J, Waters P, Beeson D, Vincent A. sera that were used in this study. MuSK myasthenia gravis IgG4 disrupts the interaction of LRP4 with MuSK but both IgG4 and IgG1-3 can disperse STUDY FUNDING preformed agrin-independent AChR clusters. PLoS One 2013;8:e80695. S.H. was supported by a Watney Trust/MGA/NIHR Oxford Bio- medical Research Centre Fellowship. Additional support was 9. Klooster R, Plomp JJ, Huijbers MG, et al. Muscle-specific provided by the NIHR Oxford Biomedical Research Centre (A.V.), the kinase myasthenia gravis IgG4 autoantibodies cause severe Medical Research Council (J.C.), and the Muscular Dystrophy neuromuscular junction dysfunction in mice. Brain 2012; Campaign (J.C.). 135:1081–1101.

8 Neurology: Neuroimmunology & Neuroinflammation 10. Waters P, Woodhall M, O’Connor KC, et al. MOG cell- 14. Guptill JT, Sanders DB, Evoli A. Anti-MuSK antibody based assay detects non-MS patients with inflammatory myasthenia gravis: clinical findings and response to treat- neurologic disease. Neurol Neuroimmunol Neuroinflamm ment in two large cohorts. Muscle Nerve 2011;44:36–40. 2015;2:e89. doi: 10.1212/NXI.0000000000000089. 15. Evoli A, Tonali PA, Monaco ML, et al. Clinical correlates 11. Huijbers MG, Zhang W, Klooster R, et al. MuSK IgG4 with anti-MuSK antibodies in generalized seronegative antibodies cause myasthenia gravis by inhibiting binding myasthenia gravis. Brain 2003;126:2304–2311. between MuSK and LRP4. Proc Natl Acad Sci USA 2013; 16. Chang T, Leite MI, Senanayake S, et al. Clinical and 110:20783–20788. serological study of myasthenia gravis using both radio- 12. Tsonis AI, Zisimopoulou P, Lazaridis K, et al. MuSK auto- immunoprecipitation and cell-based assays in a South antibodies in myasthenia gravis detected by cell based assay— Asian population. J Neurol Sci 2014;343:82–87. a multinational study. J Neuroimmunol 2015;284:10–17. 17. Yang L, Maxwell S, Leite MI, et al. Non-radioactive sero- 13. Bever CT, Aquino AV, Penn AS, Lovelace RE, Rowland logical diagnosis of myasthenia gravis and clinical features LP. Prognosis of ocular myasthenia. Ann Neurol 1983;14: of patients from Tianjin, China. J Neurol Sci 2011;301: 516–519. 71–76.

Neurology: Neuroimmunology & Neuroinflammation 9 Mucosal biopsy shows immunologic changes of the colon in patients with early MS

Adrian Mathias Moser, ABSTRACT * MD Objective: To investigate immune cells of the colonic mucosa and fecal short-chain fatty Walter Spindelboeck, acids (SCFAs) in treatment-naive patients with a clinically isolated syndrome (CIS) or early MD* relapsing MS. Heimo Strohmaier, PhD Methods: In this cross-sectional proof-of-concept study, we obtained mucosal specimens during Christian Enzinger, MD ileocolonoscopy from 15 untreated patients with CIS/MS and 10 controls. Mucosal immune cells Thomas Gattringer, MD were analyzed by FACS, and gas chromatography-mass spectrometry measurements of stool Siegrid Fuchs, MD samples served to determine SCFA. Franz Fazekas, MD Gregor Gorkiewicz, MD Results: The number of total dendritic cells (DCs), CD1031 tolerogenic DCs, and CD41251127– Philipp Wurm, MSc regulatory T cells (Tregs) was significantly reduced in the distal colon of patients with CIS/MS Christoph Högenauer, compared with controls, whereas we found no differences in the proximal colon. The patients’ fecal MD samples also showed a substantially lower content of SCFA and especially lower levels of butyrate Michael Khalil, MD, PhD and acetate. Conclusions: Our findings indicate a disturbed homeostasis of colonic DCs and Tregs in patients with MS which could be associated with colonic SCFA depletion. Although not implying causality, Correspondence to these findings confirm parallel abnormalities of the gut in MS and warrant further research if mod- Prof. Högenauer: ulation of the colonic SCFA profile or the colonic Treg pool can serve to modify the course of MS. christoph.hoegenauer@ medunigraz.at or Prof. Khalil: Neurol Neuroimmunol Neuroinflamm 2017;4:e362; doi: 10.1212/NXI.0000000000000362 [email protected]

GLOSSARY CIS 5 clinically isolated syndrome; CNS 5 central nervous system; DC 5 dendritic cell; EAE 5 experimental autoimmune encephalomyelitis; EDSS 5 Expanded Disability Status Scale; EI 5 electron impact; FCS 5 fetal calf serum; FMO 5 fluorescence minus one; GC-EI/MS 5 gas chromatography-electron impact-mass spectrometry; HBSS 5 Hanks balanced salt solution; HPLC 5 high-performance liquid chromatography; IQR 5 interquartile range; IRB 5 institutional review board; LPDC 5 lamina propria DC; PEG 5 polyethylene glycol; RPMI 5 Roswell Park Memorial Institute; SCFA 5 short-chain fatty acid; Tregs 5 regulatory T cell.

MS is a complex immune-mediated disease that is believed to arise in genetically susceptible individuals together with environmental factors such as viral infections, vitamin D deficiency, or smoking that promote or may even trigger autoimmunity.1–3 More recently, there have also been speculations on an important role of the gut in this process.4 Dysregulation of the gut microbiome may induce autoimmune responses at different sites, including the central nervous system (CNS).5–8 These insights are mainly derived from animal models of MS. Studies in experimental autoimmune encephalomyelitis (EAE) in mice showed that alterations in the gut microbiome induced by dietetic modifications to antibiotic ablation were associated with immunologic changes and influenced the severity of the disease.5,9 In this process, intestinal dendritic cells (DCs) appear to play a critical role as they act as antigen-presenting cells and are crucial in maintaining immune tolerance by interaction with numerous intestinal antigens.10 Lamina Supplemental data at Neurology.org/nn *These authors contributed equally to this work. From the Department of Internal Medicine (A.M.M., W.S., C.H.), Division of Gastroenterology and Hepatology, Theodor Escherich Laboratory for Microbiome Research (A.M.M., W.S., G.G., P.W., C.H.), Center for Medical Research (H.S.), Department of Neurology (C.E., T.G., S.F., F.F., M.K.), and Institute of Pathology (G.G., P.W.), Medical University of Graz, Austria. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 propria DCs (LPDCs) transport bacterial anti- Figure 1 FACS analysis of colonic mucosal DCs and Tregs gens and products to the mesenteric lymph nodes, where the balance between tolerance of the commensal microbiota and surveillance of invading pathogenic bacteria is tightly regulated.11,12 This critical function is further underlined by the ability of DCs to induce regulatory T cells (Tregs) and therefore maintain or establish immune tolerance.13 In particular, the CD1031 DC subtype is vital for the differentiation of Tregs, which is believed to exhibit tolerogenic properties in mice and humans.14,15 Tregs have been implicated in the protection against CNS demyelination in mice, and their expansion has been reported to occur in gut mucosa and draining lymph nodes of EAE-protected germ-free mice.16,17 All these processes are tightly regulated by a complex interplay of the gut flora, dietary intake, and mucosal immune cells.5,9,18 In this context, short-chain fatty acids (SCFAs) have been suggested to play a pivotal role, and murine CNS demyelination was inhibited by peroral treatment with propionate through expansion of intestinal Tregs.19 Other studies underline the role of SCFAs, especially acetate, butyrate, and propionate in mediating regulatory intestinal immune responses. Treg homeostasis has, thereby, been shown to be dependent on the presence and extent of SCFAs produced by commensal bacteria in mice. Moreover, SCFAs endow DCs with an enhanced ability to induce the differentiation of Tregs.18,20,21 Recently, oral propionate administration also resulted in expanded peripheral Treg counts in patients with MS.22 On the basis of these data, we hypothesized that corresponding changes in immune cells of the gut mucosa and in SCFAs should also be present in patients with MS. We speculated to find lower numbers of DCs, in particular CD1031 DCsinthecolonmucosaof patients with MS, which might be associated with fewer Tregs. To minimize confounding in-

Colonic specimens were obtained separately from the proximal and distal colon, and lamina fluences, we focused on patients without disease- propria mononuclear cells were isolated. (A) DCs were identified as live single cells that stain modifying treatment in the early stage of MS. negatively for lin (lineage) 1 markers (CD3, CD14, CD16, CD19, CD20, CD56, and CD34) and positively for HLA-DR. CD103-positive DCs were gated out of the DC population. (B) Tregs were identified as live single cells that stain positively for CD3, CD4, and CD25 METHODS Patients and controls. Between 2013 and markers and negatively for CD127. DC 5 dendritic cell; FACS 5 fluorescence-activated cell 2014, we consecutively recruited patients from the MS outpatient sorting; Treg 5 regulatory T cells. clinic of the Department of Neurology, Medical University of Graz for this cross-sectional study who were interested in our research question including the results of a colonic biopsy.

2 Neurology: Neuroimmunology & Neuroinflammation Inclusion criteria were as follows: (1) early MS as evidenced by was verified by visual inspection. After passing through a 70-mm a clinically isolated syndrome (CIS) suggestive of MS or relapsing- cell strainer, the released cells were washed twice with RPMI remitting MS within the first year after diagnosis,23 (2) age complete medium (containing 10% fetal calf serum [FCS] and 1% between 18 and 60 years, and (3) informed consent. Exclusion penicillin/streptomycin) and finally resuspended in RPMI criteria were as follows: (1) a known gastrointestinal pathology, complete medium. The cell suspension was kept on ice until further (2) previous or ongoing disease-modifying treatment of MS, (3) analysis. evidence for a concomitant other autoimmune disease, (4) gas- Flow cytometry. Mucosal immune cell subsets were character- trointestinal infection during the last 2 months before endoscopy, ized by FACS analysis as previously described (figure 1).25,26 The (5) a steroid pulse ,30 days before endoscopy, (6) antibiotic cell suspension was washed once with staining buffer (phosphate- therapy during the last month before endoscopy, (7) clinically buffered saline containing 3% FCS and 2 mM EDTA), and the relevant coagulopathy, (8) Expanded Disability Status Scale cells were stained with 100 mL staining buffer for 20 minutes at (EDSS) $7, (9) severe infectious disease, (10) alcohol or drug room temperature in the dark. For enumeration of LPDCs, abuse, and (11) history of a prior technically impossible colono- directly labeled monoclonal antibodies for the following markers scopy. As controls, we included healthy individuals (recruited were used: lin (lineage) 1-FITC (CD3, CD14, CD16, CD19, between 2013 and 2014 from the Department of Gastroenter- CD20, CD56, and CD34), HLA-DR-PerCP-Cy5.5, CD11c- ology and Hepatology, Medical University of Graz) meeting the APC, and CD103-PE. LPDCs were identified as lin1-/ following inclusion criteria: (1) age between 18 and 60 years, (2) HLADR1 cells. For determination of Tregs, anti-CD3-APC- screening colonoscopy for colorectal cancer due to age, positive Cy7, anti-CD4-V450, anti-CD8-FITC, anti-CD25-PE, and family history, or patient request, and (3) informed consent. The anti-CD127-Alexa Fluor 647 antibodies were used. With the exclusion criteria were identical to those of the patients with CIS/ exception of CD103-PE (eBioscience, San Diego, CA), all anti- MS. None of the controls had signs of an acute illness or received bodies were purchased from BD Bioscience (San Jose, CA). FMO any long-term medication. (fluorescence minus one) controls were used to set the boundaries for gating of positively stained cells. After the staining reaction, the Ileocolonoscopy and isolation of LP mononuclear cells. cells were washed once with staining buffer and resuspended in Ileocolonoscopy was performed with standard equipment (Olympus, 100 mL staining buffer. For the exclusion of dead cells, propidium Hamburg, Germany) in sedated subjects. Tissue specimens iodide was added to the samples immediately before acquisition on were obtained by forceps biopsy. Biopsies were separately taken an LSR II (BD Bioscience) flow cytometer. The data files were from the proximal (cecum and ascending colon) and distal analyzed using FlowJo (FlowJo, LLC, Ashland, OR) software. (descending and sigmoid colon) colon during the retraction of the colonoscope to investigate possible differences of colonic Gas chromatography-electron impact-mass spectrometry 24 immune cells related to colonic compartments. Colonic biopsy (GC-EI/MS) of SCFAs. SCFAs previously investigated in animal specimens were immediately preserved in chilled Roswell Park studies to study cellular immune reactions (acetic acid, propionic Memorial Institute (RPMI) medium (Sigma-Aldrich, St. Louis, acid, isobutyric acid, butyric acid, isovaleric acid, and valeric acid) MO; supplemented with penicillin, streptomycin, and amphoter- were extracted from stool frozen at 280°C.18–20 SCFA concentra- icin). Biopsies were washed once with calcium- and magnesium- tions were measured by GC-MS equipped with a polyethylene free Hanks balanced salt solution (HBSS) (Life Technologies, glycol (PEG) DB-WAXetr (30 m; 0.25 mm ID; 0.25 mmfilm) Vienna, Austria) and then incubated in calcium- and magnesium- column. SCFAs were extracted from feces by sequential addition of free HBSS containing 1 mM dithiothreitol and 5 mM EDTA at 1mLphosphoricacid(0.5%)and1mLmethyltert-butylether, 37°C for 20 minutes with gentle agitation to remove mucus and 10 minutes shaking, 10 minutes centrifugation, and removal of the epithelial cells. Following a brief wash with calcium- and upper organic layer. Before extraction, 100 nmol of D-acetic acid, magnesium-free HBSS, tissue was digested with 1 mg/mL Col- D-propionic acid, D-butyric acid, and D-valeric acid were added as lagenase A (Roche, Basel, Switzerland) and 5 U/mL DNase I internal standards. Calibration curves by stable isotope dilution were (Roche) in HBSS at 37°C for 60 minutes on a shaker and performed from 0.1 to 2,000 mM for acetic acid, propionic acid, mechanically disrupted by gentle pipetting. Complete dissociation isobutyric acid, butyric acid, isovaleric acid, and valeric acid. A 7890B/5977A MSD GC-MS (Agilent Technologies, Waldbronn, Germany) equipped with a PEG DB-WAXetr (30 m; 0.25 mm ID; Table Clinical and demographic data of 15 MS-p and 10 healthy controls (Ctrl) 0.25 mmfilm)columnwasused.Heliumwasusedascarriergasat included in the study 1.3 mL/min in splitless mode at 250°C injector temperature. The initial oven temperature of 60°C was held for 2 minutes, and then a p MS-p Ctrl Value the temperature first was ramped up to 150°C at a rate of 15°C/min. No. of patients (% women) 15 (53) 10 (40) NSb This was followed by a ramp of 5°C/min up to 170°C and 20°C/ min up to 250°C, where the temperature was held for another Age at colonoscopy, y, mean (SD) 35 (7) 44 (13) ,0.05c 2 minutes. The mass spectrometer was run in EI mode, where the Age at disease onset, y, mean (SD) 33 (7) NA NA fatty acids were detected in selected ion monitoring mode on m/z Disease duration at colonoscopy, mo, median (range) 7(4–29) NA NA 60, 63, 73, 74, 76, 79, and 80. The source temperature was set to 250°C, and the transfer line temperature was 280°C. Data analysis EDSS at time of colonoscopy, median (range) 1.5 (1.0–2.0) NA NA was performed by MassHunter (Agilent Technologies). No. of patients with disease-modifying treatment 0 (0.0) NA NA at colonoscopy (%) Statistics. Statistical analyses were performed using SPSS 22 (IBM Corp., Armonk, NY) and GraphPad Prism (GraphPad Abbreviations: Ctrl 5 control patients; EDSS 5 Expanded Disability Status Scale; MS-p 5 patients with MS; NA 5 not applicable; NS 5 not significant. Software, Inc., La Jolla, CA). Values are presented as number a The MS-p cohort consisted of 10 patients with clinically isolated syndromes and 5 (%), mean (SD), or median (interquartile range [IQR]) as appro- patients with relapsing-remitting MS. priate. For the comparison of categorical variables, we applied b Fisher exact test. Fisher exact test. Group differences of continuous variables were c Mann-Whitney U test. determined by Mann-Whitney U test or t test depending on

Neurology: Neuroimmunology & Neuroinflammation 3 and the interval between first symptoms and colonoscopy Figure 2 Reduction of mucosal DCs in the distal colon of MS-p was 0.5 years. Median EDSS27 was 2 (IQR 1.0–2.0) at the time of colonoscopy. As controls, we included 10 healthy individuals, with a mean age of 44 (SD 13) years, who were scheduled for screening colonoscopy during thesameperiod.Thisgroup consisted of 6 men and 4 women (table).

DCs and Tregs in the colon mucosa. Using FACS analysis, we explored the number of total DCs (defined as lin1-/HLADR1 cells), CD1031 DCs, and CD41251127–Tregs in the mucosa of the proximal and distal colon. No differences between patients with CIS/MS and controls were observed in the proximal colon (figure 2, A and B). Patients with CIS/ MS had significantly fewer overall DCs and CD1031 DCs in the distal colon compared with controls (total DCs: 3.0 [2.2–3.8] vs 5.0 [3.6–6.3], p 5 0.019 [& of living cells] and 255 [93–381] vs 404 [335– 592], p 5 0.026 [absolute numbers]; CD1031 DCs: 1.4 [0.9–1.8] vs 2.5 [1.4–3.1], p 5 0.036 [& of living cells] and 114 [38–174] vs 183 [124–303], p 5 0.048 [absolute numbers], figure 2, C and D). Furthermore, significant reductions of Tregs in patients with CIS/MS were likewise only observed in the distal colon (Tregs: 3.1 [1.3–3.7] vs 4.7 [3.4–6.1], p 5 0.032 [& of living cells] and 153 [83–205] vs 226 [136–510], p 5 0.023 [absolute numbers], figure 3, A–D). Levels of CD41, CD81 T cells (figure e-1, A–D at Neurology.org/nn), and natural killer cells (figure e-2, A–D) did not differ between patients with CIS/MS and controls. As we observed a significant age difference between patients with CIS/MS and controls, we calculated the cor- DC numbers from the proximal (A and B) and distal (C and D) colon, see central scheme. (A and relations between age and total DCs, CD1031 DCs, B) No significant differences in the number of total and CD1031 DCs were found in the and Tregs. Significant (negative) correlations were proximal colon of MS-p compared with controls (Ctrl). (C and D) In the distal colon, significant observed between age and Tregs in both compartments reductions of total and CD1031 DCs were observed in MS-p compared with Ctrl (total DCs: 3.0 [2.2–3.8] vs 5.0 [3.6–6.3], p 5 0.019 [‰ of living cells] and 255 [93–381] vs 404 [335– (proximal Tregs: p 5 0.048, r 520.461; distal Tregs: 592], p 5 0.026 [absolute numbers]; CD1031 DCs: 1.4 [0.9–1.8] vs 2.5 [1.4–3.1], p 5 0.036 p 5 0.029, r 520.489, figure e-3, A–C). Correla- [‰ of living cells] and 114 [38–174] vs 183 [124–303], p 5 0.048 [absolute numbers]) (median tions between age and total DCs (proximal: p 5 0.949, – U t [Q1 Q3]; Mann-Whitney test or test to compare non-Gaussian and Gaussian variables. r 52 5 r 5 Boxplots according to Tukey). DC 5 dendritic cell; MS-p 5 patients with MS. 0.013; distal: p 0.155, 0.299) and CD1031DCs (proximal: p 5 0.668, r 520.092; distal: p 5 0.093, r 5 0.351) remained not significant. non-Gaussian and Gaussian data distribution. Boxplots are depicted according to Tukey. Correlation analyses were per- SCFA levels in fecal samples. High-performance liquid formed using Spearman rank correlation coefficient (r). chromatography (HPLC) was used to assess levels of SCFAs in fecal samples obtained at the time of ileoco- Standard protocol approvals, registrations, and patient lonoscopy and showed that total SCFA content as consents. All individuals signed informed consent before study well as acetate and butyrate levels of patients with inclusion. The study was approved by the local institutional CIS/MS were significantly reduced by 50%–60% review board (IRB no. IRB00002556), vote 25-287 ex 12/13. compared with control stools. Propionate and valerate were also reduced in patients with CIS/MS, but this RESULTS Patients and controls. Overaperiodof10 difference did not reach statistical significance months, we screened consecutive patients for the respec- (figure 4). The order of absolute abundances of single tive inclusion and exclusion criteria, and 15 (10 with SCFAs was similar in patients with CIS/MS and con- a CIS) consented to the study. Their mean age was 35 trols (acetate . propionate . butyrate . valerate . (SD 7) years. Eight of the fifteen patients were women, isovalerate . isobutyrate).

4 Neurology: Neuroimmunology & Neuroinflammation and distal colon have been shown to differ in Figure 3 Reduction of mucosal Tregs in the distal colon of MS-p embryologic origin, gene-expression patterns, gene- methylation status, microbiota load, and microbiota metabolic activity (e.g., SCFAs).29–31 In addition, mucosal DC subsets differ between the proximal and distal colon of healthy individuals.24 Altogether, these findings suggest a concept of colonic immune compartmentalization which may well explain our different observations in the proximal and distal colon. Although the causes of the observed regional differences are not clear, altered amounts of bacterial metabolites along the colon may be of relevance.31 We also found a decrease of the total SCFA content as well as of acetate and butyrate levels accompanied by trends toward lower levels of propionate and valerate in stools of early MS patients compared with controls. These observations were hitherto not made in humans and could suggest a concept where changes of mucosal immune cells may be linked to reduced levels of certain bacterial metabolites in the colon of patients with MS. Propionate has been shown to be the most potent Treg inducer in vitro, and oral gavage ameliorated CNS demyelination in mice.19 Propionate levels were lower in our patients with MS although not to a significant extent. However, this may have been due to the small sample size. Otherwise we observed significantly lower levels of acetate and butyrate in patients with MS than controls, which have been shown to potently induce colonic Tregs in mice as well. It could, therefore, be speculated that their decrease may be of greater clinical relevance in patients with MS than that of propionate, although this interpretation again needs great caution because of the few individuals investi- Treg numbers from the proximal (A and B) and distal (C and D) colon, see central scheme. (A and B) No significant differences in the number of Tregs were found in the proximal colon of gated. To what extent this reduction is a consequence MS-p compared with controls (Ctrl). (C and D) In the distal colon, significant reductions of of specific microbial alterations or of dietary intake is Tregs were observed in MS-p compared with Ctrl (Tregs: 3.1 [1.3–3.7] vs 4.7 [3.4–6.1], yet unclear. Of interest, food patterns compatible p 5 0.032 [‰ of living cells] and 153 [83–205] vs 226 [136–510], p 5 0.023 [absolute numbers]) (median [Q1–Q3]; Mann-Whitney U test or t test to compare non-Gaussian and with western lifestyle have been proposed to increase Gaussian variables. Boxplots according to Tukey). MS-p 5 patients with MS; Treg 5 regula- the incidence of MS, and fecal microbiota disturban- tory T cells. ces have been reported to occur in patients with MS compared with healthy controls.32–34 In turn, adher- DISCUSSION This study demonstrates changes in ence to a Mediterranean diet elevated butyrate, ace- the immune constitution of the intestinal mucosa of tate, and propionate levels in healthy individuals.35 untreated patients with early MS when compared Thus, dietary factors may promote dysregulation of with control biopsies. Immuncytologic findings mucosal immune cells through an altered intestinal consisted of a reduced number of total DCs and of microenvironment. Diet, substrate availability, and the CD1031 DC subset in the mucosa of the distal synthesis capacities of bacteria may potentially colon and were paralleled by a simultaneous reduc- contribute to the amounts of intestinal SCFAs, and tion of the number of mucosal Tregs in the same therefore constitute potential explanations for the colonic region. These findings provide strong reduction of SCFAs in patients with MS.31 However, evidence for a regional immuneregulatory deficiency this certainly needs further investigation. of the colonic mucosa in patients with CIS/MS and Our findings in humans also support immuno- strengthen a role of the gut in MS. Inhomogeneous logic models hitherto only described in animals with pathologies along the colon are also seen clinically in EAE, which postulated that luminal SCFA content other immune-mediated diseases (e.g., ulcerative impacts on the regulatory capacity of mucosal DCs colitis and Crohn disease).28 Moreover, the proximal and the formation of the colonic Treg pool.18,19,36

Neurology: Neuroimmunology & Neuroinflammation 5 Figure 4 Decreased SCFA levels in fecal samples of MS-p

(A–C) Fecal samples were analyzed by HPLC. SCFA contents are shown in order of their absolute quantity in mmol/g (total . acetate . propionate . butyrate . valerate . isovalerate . isobutyrate). A significant reduction in MS-p (gray) vs controls (white) was found in total SCFA (252 [148–354] vs 121 [82–162]; p 5 0.017), acetate (180 [85–227] vs 70 [45–96]; p 5 0.025), and butyrate contents (33 [18–48] vs 14 [8–21]; p 5 0.018). Trends were observed in propionate (p 5 0.084) and valerate contents (p 5 0.075) (median [Q1–Q3]; Mann-Whitney U test or t test to compare non-Gaussian and Gaussian variables. Boxplots according to Tukey). HPLC 5 high-performance liquid chromatography; MS-p 5 patients with MS; SCFA 5 short-chain fatty acid.

Based on our data, one may hypothesize that lower changes in the colonic mucosa may further increase SCFA content (especially acetate and butyrate) with more advanced disease, which needs to be together with a reduced number of total and tolero- acknowledged as a limitation of the study. This study genic DCs may trigger a quantitative impairment of design also prohibited a more close matching of Tregs in the distal colon of patients with MS. These patients and controls. However, our analysis of the alterations are also likely to impact on a systemic level, effect of age on Tregs indicates that the significantly as changes in the lymphatic immune system in mice higher age of controls should have led to an underes- could be observed after manipulation of the intestinal timation rather than an overestimation of observed ecosystem. Thereby, activation of CD1031 DCs in differences. We also defined a large number of exclu- cervical lymph nodes played a major role in attenu- sion criteria to avoid any systematic bias. Neverthe- ating CNS autoimmunity.16 Of interest, a recent less, participating individuals may not have been report provides evidence for the CNS lymphatic fully representational of patients with CIS/MS and system communicating with deep cervical lymph healthy individuals in general, which needs to be nodes.37 Because the lymphatic system is known to acknowledged as another possible limitation. Our communicate throughout the whole body, informa- findings thus will need to be replicated by further tion from the intestinal lymphatic system may also studies. Future studies on a larger number of patients reach the CNS.37,38 In this context, T cells have should also allow us to probe potential associations of already been shown to gain autoreactive function in immunologic changes in the colonic mucosa with clinical extracerebral tissues such as the lungs and to migrate and radiologic findings. It is also important to stress that to the CNS transgressing the blood-brain barrier.39 ourfindingsinthecolonofpatients with CIS/MS cannot The concept of extracerebral immune-cell alterations serve to establish a causal relationship with their CNS at mucosal sites may, therefore, be supported by our disease. The immunologic changes observed could also data and warrants further mechanistic investigation just mirror a more general immune dysregulation of such regarding the gut in patients with MS. patients. Furthermore, a suggested interplay between the In this study, we solely focused on patients with changes in immune cells of the distal colon and content early MS and therefore we do not know if observed in SCFAs has to remain speculative as our observations

6 Neurology: Neuroimmunology & Neuroinflammation are based on cross-sectional data. These important points 3. Fitzgerald KC, Munger KL, Kochert K, et al. Association also need to be addressed in future studies. of vitamin D levels with multiple sclerosis activity and Altogether, our findings extend and substantiate progression in patients receiving interferon beta-1b. JAMA Neurol 2015;72:1458–1465. previous speculations on a disturbed immune milieu 4. Tremlett H, Fadrosh DW, Faruqi AA, et al. Gut micro- in the colon of patients with MS and support the biota composition and relapse risk in pediatric MS: a pilot possibility of a role of the colonic microenvironment study. J Neurol Sci 2016;363:153–157. in the immunopathogenesis of MS. Once confirmed, 5. Ochoa-Reparaz J, Mielcarz DW, Begum-Haque S, Kasper our observations could give rise to test if therapeutic in- LH. Gut, bugs, and brain: role of commensal bacteria in terventions that modulate the colonic SCFA profile or the control of central nervous system disease. Ann Neurol – the colonic Treg pool can modify the course of MS.40 2011;69:240 247. 6. Wekerle H, Berer K, Krishnamoorthy G. Remote control- triggering of brain autoimmune disease in the gut. Curr AUTHOR CONTRIBUTIONS Opin Immunol 2013;25:683–689. A.M.M. and W.S. planned the study and experiments and analyzed data. 7. Hohlfeld R, Dornmair K, Meinl E, Wekerle H. The A.M.M., W.S., M.K., and F.F. wrote the paper. C.H., M.K., F.F., S.F., search for the target antigens of multiple sclerosis, part and C.E supervised the study planning and implementation and gave sci- 1: autoreactive CD41 T lymphocytes as pathogenic effec- entific advices. H.S. performed immunologic analyses. T.G., M.K., C.E., – F.F., and S.F. recruited patients. All authors were involved in the prepa- tors and therapeutic targets. Lancet Neurol 2015;15:198 ration of the manuscript and data discussion. 209. 8. Hohlfeld R, Dornmair K, Meinl E, Wekerle H. The search for the target antigens of multiple sclerosis, part ACKNOWLEDGMENT 2: CD81 T cells, B cells, and antibodies in the focus of The authors thank Jennifer Ober for her efforts in performing FACS anal- reverse-translational research. Lancet Neurol 2016;15: yses and Harald Köfeler for his support regarding SCFA measurements. – They acknowledge the support in endoscopic examinations by Elisabeth 317 331. Krones, Patrizia K. Kump, and Andreas Eherer. 9. Joscelyn J, Kasper LH. Digesting the emerging role for the gut microbiome in central nervous system demyelination. Mult Scler 2014;20:1553–1559. STUDY FUNDING 10. Scott CL, Aumeunier AM, Mowat AM. Intestinal No targeted funding reported. CD1031 dendritic cells: master regulators of tolerance? Trends Immunol 2011;32:412–419. DISCLOSURE 11. Macpherson AJ, Uhr T. Induction of protective IgA by A.M. Moser reports no disclosures. W. Spindelboeck received travel intestinal dendritic cells carrying commensal bacteria. funding from AbbVie. H. Strohmaier reports no disclosures. C. Enzinger Science 2004;303:1662–1665. served on the scientific advisory board for Biogen-Idec, Bayer-Schering, 12. Voedisch S, Koenecke C, David S, et al. Mesenteric Merck Serono, Novartis, and Teva Ratiopharm; received travel funding lymph nodes confine dendritic cell-mediated dissemina- and speaker honoraria from Biogen-Idec, Teva-Aventis, Merck Serono, tion of Salmonella enterica serovar Typhimurium and Bayer-Schering, and Novartis; consulted for Biogen-Idec, Novartis, Genzyme, Merck Serono, and Bayer-Schering; and received research sup- limit systemic disease in mice. Infect Immun 2009;77: – port from Teva-Aventis, Biogen-Idec, and Merck Serono. T. Gattringer 3170 3180. served on the scientific advisory board for Boehringer Ingelheim; received 13. Worbs T, Bode U, Yan S, et al. Oral tolerance originates travel funding and speaker honoraria from Boehringer Ingelheim and in the intestinal immune system and relies on antigen Bayer Austria. S. Fuchs served on the scientific advisory board for carriage by dendritic cells. J Exp Med 2006;203:519–527. Biogen-Idec, Novartis, Merck Serono, Genzyme, Roche, and Teva Ratio- 14. Yuan J, Zhang G, Yang X, Liu K, Wang F. Transplanta- pharm and received speaker honoraria and/or travel funding from tion of allograft transforming growth factor-beta1 trans- Biogen-Idec, Merck Serono, Genzyme, Roche, and Teva Ratiopharm. fected CD103(1) lamina propria dendritic cells could F. Fazekas served on the scientific advisory board for Biogen-Idec, effectively induce antigen-specific regulatory T cells Genzyme, Merck, Novartis, Perceptive Informatics, Roche, Teva Ratio- – pharm, Actelion, and University of Zurich; received travel funding in vivo. Transpl Proc 2013;45:3408 3413. and/or speaker honoraria from Almirall, Merck, Novartis, Roche, and 15. Worthington JJ, Czajkowska BI, Melton AC, Travis MA. Teva; served on the editorial board for Cerebrovascular Diseases, Journal Intestinal dendritic cells specialize to activate transforming of Neurology, Polish Journal of Neurology and Neurosurgery, Stroke, Swiss growth factor-beta and induce Foxp31 regulatory T cells Archives of Neurology and Psychiatry, Multiple Sclerosis, European Stroke via integrin avb8. Gastroenterology 2011;141:1802–1812. ® ’ Journal,andNeurology ; served on the speaker s bureau for Merck 16. Ochoa-Reparaz J, Mielcarz DW, Wang Y, et al. A poly- Serono; and received an unrestricted research grant to his institution by saccharide from the human commensal Bacteroides fragilis Teva Ratiopharm. G. Gorkiewicz is an editorial board member for Mi- protects against CNS demyelinating disease. Mucosal crobiome. P. Wurm and C. Högenauer report no disclosures. M. Khalil – received travel funding and speaker honoraria from Bayer-Schering, Immunol 2010;3:487 495. Novartis, Genzyme, Merck Serono, Biogen, and Teva. Go to 17. Lee YK, Menezes JS, Umesaki Y, Mazmanian SK. Proin- Neurology.org/nn for full disclosure forms. flammatory T-cell responses to gut microbiota promote experimental autoimmune encephalomyelitis. Proc Natl Received January 3, 2017. Accepted in final form April 13, 2017. Acad Sci USA 2011;108(suppl 1):4615–4622. 18. Arpaia N, Campbell C, Fan X, et al. Metabolites produced REFERENCES by commensal bacteria promote peripheral regulatory T-cell 1. Ascherio A. Environmental factors in multiple sclerosis. generation. Nature 2013;504:451–455. Expert Rev Neurother 2013;13:3–9. 19. Haghikia A, Jorg S, Duscha A, et al. Dietary fatty acids 2. Dendrou CA, Fugger L, Friese MA. Immunopathology of directly impact central nervous system autoimmunity via multiple sclerosis. Nat Rev Immunol 2015;15:545–558. the small intestine. Immunity 2015;43:817–829.

Neurology: Neuroimmunology & Neuroinflammation 7 20. Furusawa Y, Obata Y, Fukuda S, et al. Commensal 30. Kaz AM, Wong CJ, Dzieciatkowski S, Luo Y, Schoen RE, microbe-derived butyrate induces the differentiation of Grady WM. Patterns of DNA methylation in the normal colonic regulatory T cells. Nature 2013;504:446–450. colon vary by anatomical location, gender, and age. Epi- 21. Smith PM, Howitt MR, Panikov N, et al. The microbial genetics 2014;9:492–502. metabolites, short-chain fatty acids, regulate colonic Treg 31. Macfarlane S, Macfarlane GT. Regulation of short-chain cell homeostasis. Science 2013;341:569–573. fatty acid production. Proc Nutr Soc 2003;62:67–72. 22. Haghikia A, Duscha A, Berg J, et al. Role of fatty acids in 32. Lauer K. Environmental risk factors in multiple sclerosis. multiple sclerosis: therapeutic potential of propionic acid Expert Rev Neurother 2010;10:421–440. (P1.374). Neurology 2016;86. 33. Chen J, Chia N, Kalari KR, et al. Multiple sclerosis pa- 23. Polman CH, Reingold SC, Banwell B, et al. Diagnostic tients have a distinct gut microbiota compared to healthy criteria for multiple sclerosis: 2010 revisions to the controls. Sci Rep 2016;6:28484. McDonald criteria. Ann Neurol 2011;69:292–302. 34. Miyake S, Kim S, Suda W, et al. Dysbiosis in the gut 24. Bernardo D, Durant L, Mann ER, et al. Chemokine (C-C microbiota of patients with multiple sclerosis, with a strik- motif) receptor 2 mediates dendritic cell recruitment to the ing depletion of species belonging to clostridia XIVa and human colon but is not responsible for differences IV clusters. PLoS One 2015;10:e0137429. observed in dendritic cell subsets, phenotype, and function 35. De Filippis F, Pellegrini N, Vannini L, et al. High-level between the proximal and distal colon. Cell Mol Gastro- adherence to a Mediterranean diet beneficially impacts the enterol Hepatol 2016;2:22–39.e5. gut microbiota and associated metabolome. Gut 2016;65: 25. Mann ER, Landy JD, Bernardo D, et al. Intestinal den- 1812–1821. dritic cells: their role in intestinal inflammation, manipu- 36. BererK,MuesM,KoutrolosM,etal.Commensal lation by the gut microbiota and differences between mice microbiota and myelin autoantigen cooperate to trig- and men. Immunol Lett 2013;150:30–40. ger autoimmune demyelination. Nature 2011;479: 26. Badami E, Sorini C, Coccia M, et al. Defective differen- 538–541. tiation of regulatory FoxP31 T cells by small-intestinal 37. Louveau A, Smirnov I, Keyes TJ, et al. Structural and dendritic cells in patients with type 1 diabetes. Diabetes functional features of central nervous system lymphatic 2011;60:2120–2124. vessels. Nature 2015;523:337–341. 27. Kurtzke JF. Rating neurologic impairment in multiple 38. Delves PJ, Roitt IM. Roitt’s Essential Immunology, 12th sclerosis: an expanded disability status scale (EDSS). Neu- ed. Hoboken: Wiley-Blackwell; 2011. rology 1983;33:1444–1452. 39. Odoardi F, Sie C, Streyl K, et al. T cells become licensed 28. Abraham C, Cho JH. Inflammatory bowel disease. N Engl in the lung to enter the central nervous system. Nature J Med 2009;361:2066–2078. 2012;488:675–679. 29. Glebov OK, Rodriguez LM, Nakahara K, et al. Distinguish- 40. Atarashi K, Tanoue T, Oshima K, et al. Treg induction by ing right from left colon by the pattern of gene expression. a rationally selected mixture of Clostridia strains from the Cancer Epidemiol Biomarkers Prev 2003;12:755–762. human microbiota. Nature 2013;500:232–236.

8 Neurology: Neuroimmunology & Neuroinflammation Aquaporin-4 antibodies in patients treated with natalizumab for suspected MS

Anna Gahlen, MD ABSTRACT Anne-Kathrin Trampe, Objective: To evaluate (1) the frequency of aquaporin-4 antibody (AQP4-ab)-seropositive cases PhD among patients treated with natalizumab (NAT) and previously diagnosed with MS (MSNAT)in Steffen Haupeltshofer, a nationwide cohort, (2) the clinical course of NAT-treated AQP4-ab–seropositive neuromyelitis MSc optica spectrum disorder (NMOSD) patients (NMONAT), (3) AQP4-ab titers in NMONAT and AQP4- Marius Ringelstein, MD ab–seropositive NMOSD treated with other immunotherapies (NMOIT), and (4) immune mecha- Orhan Aktas, MD nisms influencing disease activity in NMONAT. Achim Berthele, MD Methods: MSNAT serum samples were retrospectively screened with a cell-based assay for AQP4- Brigitte Wildemann, MD IgG and titers determined by ELISA. The annualized relapse rate (ARR) and disability progression Ralf Gold, MD were assessed. Serum levels of proinflammatory cytokines (interleukin [IL]-1b, IL-4, IL-6, IL-8, IL- Sven Jarius, MD 10, IL-17, IL-21, and interferon [IFN]-g) and the chemokine CXCL-10 of NMONAT patients Ingo Kleiter, MD identified in this (n 5 4) and a previous study (n 5 5) were measured by cytometric bead array and ELISA.

NAT Correspondence to Results: Of the 1,183 MS patients (851 female, median 9 NAT infusions), only 4 (0.33%; 3 Prof. Kleiter: female, 1 male) had AQP4-IgG. Of these, 2 fulfilled the 2006 NMO criteria and all met the 2015 [email protected] NMOSD criteria. The ARR was higher in NMONAT vs MSNAT (p 5 0.0182). All 4 NMONAT patients had relapses and 2 had an increase of disability. AQP4-ab titers were higher in NMONAT (n 5 9) vs NMOIT (n 5 13; p 5 0.0059). IL-8, IL-1b, and IFN-g serum levels were significantly higher, and CXCL-10 was significantly lower in NMONAT vs NMOIT. Conclusions: Misdiagnosis of NMOSD with MS is rare. NAT was not able to control disease activity in NMONAT patients, who had higher serum levels of AQP4-IgG and proinflammatory cytokines than patients with NMOSD treated with other immunotherapies. Neurol Neuroimmunol Neuroinflamm 2017;4:e363; doi: 10.1212/NXI.0000000000000363

GLOSSARY AQP4-ab 5 aquaporin-4 antibody; ARR 5 annualized relapse rate; CBA 5 cell-based assay; EAE 5 experimental autoimmune encephalomyelitis; EDSS 5 expanded disability status scale; HEK 5 human embryonic kidney; JCV 5 JC virus; IL 5 interleukin; IFN 5 interferon; NAT 5 natalizumab; NMOSD 5 neuromyelitis optica spectrum disorder.

Neuromyelitis optica (NMO) is a relapsing autoimmune CNS disease which mainly affects the optic nerves and spinal cord and often leads to severe disability.1 The detection of a serum antibody targeting the astrocytic water channel aquaporin-4 antibody (AQP4-ab)2 led to the definition of revised diagnostic criteria3 and to renaming of the entity as NMO spectrum disorder (NMOSD).4 Although distinction between MS and NMOSD was facilitated by AQP4-ab testing, there is a substantial overlap between clinical phenotypes of MS and NMOSD, which caused diagnostic uncertainty or misdiagnosis, in particular prior to the availability of AQP4-ab testing. Moreover, current AQP4-ab assays differ significantly with regard to sensitivity.1 Distinguishing MS and NMOSD is of high clinical relevance, since Supplemental data at Neurology.org/nn From the Department of Neurology (A.G., A.-K.T., S.H., R.G., I.K.), St. Josef-Hospital, Ruhr-University Bochum; Department of Neurology (M.R., O.A.), Medical Faculty, Heinrich-Heine-University Düsseldorf; Department of Neurology (A.B.), Klinikum Rechts der Isar, Technische Universität München; and Molecular Neuroimmunology Group (B.W., S.J.), Department of Neurology, University of Heidelberg, Germany. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 optimum treatments for the 2 diseases differ. NAT treatment time in years) in MSNAT and NMONAT patients; Most treatments found to be beneficial in MS the ARR prior NAT, calculated as the number of attacks within the last 12 months before start of NAT; and the expanded have been suggested to be ineffective or even disability status scale (EDSS) score in NMONAT patients during to cause disease exacerbation in NMOSD.5 and after NAT treatment. The clinical course of NMONAT pa- Previously, we have shown that natalizu- tients was further evaluated by a standardized questionnaire and mab (NAT), a very effective therapy for relaps- telephone interview of treating neurologists. The latter included detailed data about immunotherapies, MRI presentation, ing MS, had detrimental effects in 5 patients relapses, and disability progression before, during, and after who had been misdiagnosed with MS and NAT treatment. were treated with NAT prior to establishment As control group for studies of AQP4-ab titers, a prospective – cohort of AQP4-ab–seropositive NMOSD patients not treated of the correct diagnosis of AQP4-ab with NAT was tested. That cohort was matched for age, sex, 6 seropositive NMOSD. Although this study disease duration, and disease activity (acute disease vs remission) suggested treatment failure, it could not rule against the identified cohort of NMONAT patients. Patients with out that other patients with known or unde- an apheresis therapy in the last 60 days were excluded. Active disease was defined as relapse 660 days from the day of serum 7 tected NMOSD might benefit from NAT. sampling. To rule out the influence of immunotherapies when In the current study, we investigated a large testing antibodies and cytokines, we divided the control group set of serum samples from NAT-treated pa- into NMOSD without immunotherapy (NMOno IT) and IT tients with the diagnosis of MS (MSNAT) for NMOSD with immunotherapy (NMO ). the presence of AQP4-IgG to identify AQP4- Serological studies. The presence of serum AQP4-IgG was – evaluated by a standardized immunohistochemical cell-based ab seropositive patients with NMOSD ’ NAT assay (CBA) according to the manufacturer s protocol (AQP4- (NMO ). We hypothesized that some CBA; Euroimmun, Lübeck, Germany).10 Staining of each sample NAT AQP4-ab–seropositive NMO patients was evaluated with a fluorescence microscope (EuroStar; Euro- might have been misdiagnosed with MS and immun) at 3200 magnification and 1,000 ms exposure time in that NAT might turn out to be unable to comparison with appropriate negative and positive controls. Positive and ambiguous samples were sent in a masked fashion for control disease activity in those patients. In confirmation to 2 independent reference laboratories (Euro- addition, we investigated immune mecha- immun; Molecular Neuroimmunology Group, University nisms possibly underlying disease activity in Hospital Heidelberg, Germany). NAT Serum AQP4-ab concentrations were evaluated by a human NMO patients and compared serologic M23-AQP4–based ELISA according to the manufacturer’s markers with control NMOSD patients not protocol (anti-AQP4-Ab ELISA version 2; RSR, Cardiff, UK). treated with NAT. The cut-off for positive values was 3.0 U/mL. To evaluate the presence of AQP4-IgG reactive only with the M1 isoform of – METHODS Standard protocol approvals, registrations, human AQP4, additionally, a second, M1-AQP4 based ELISA was used (anti-AQP4-ELISA version 1; RSR; cut-off 3.0 and patient consents. Ethical approval was obtained from the U/mL).11 Absorbance was read at 450 and 405 nm by an ELISA institutional ethics review boards of the Universities of Bochum plate reader (Emax; Molecular Devices, Sunnyvale, CA). (no. 4390-12) and Düsseldorf (nos. 3419 and 3738). Prospec- Cytokines (interleukin [IL]-1b, IL-4, IL-6, IL-8, IL-10, IL- tively evaluated patients provided written informed consent. A 17, IL-21, and interferon [IFN]-g) and the chemokine CXCL-10 waiver for retrospective analysis of serum samples and unblinding were evaluated by a multiplex bead-based immunoassay (cyto- of NMONAT patients to get further clinical information from metric bead array) according to the manufacturer’s protocol treating doctors was issued by the institutional ethics review (BD Biosciences, Heidelberg, Germany). Samples were acquired board, since previous data suggested deleterious effects of NAT on a flow cytometer, and data were analyzed using FCAP Array treatment in AQP4-ab–seropositive NMO patients.6,8 software v3 (both BD Biosciences, Franklin Lakes, NJ). IL-8 and Patients. We retrospectively evaluated blinded frozen serum CXCL-10 additionally were evaluated by ELISA according to the samples from MSNAT patients stored in a serum depository at manufacturer’s protocol (IL-8 Quantikine ELISA, CXCL-10 the St. Josef Hospital Bochum.9 The sera were initially sent for Quantikine ELISA; both R&D, Minneapolis, MN). For IL-8, analysis of anti-NAT-abs with neutralizing activity. The study serum was diluted 1:10, and for CXCL-10, serum was diluted population consisted of a nationwide cohort of patients with 1:2. Absorbance was read at 650 and 450 nm. All samples were relapsing-remitting MS treated at all levels of care. We screened stored at 280°C under standard conditions and investigated all samples obtained between February 2007 and August 2009. In simultaneously in the respective assay. case of several samples per patient, only the first eligible sample Statistical analysis. ARR between groups was compared by the was considered. Inclusion criteria were $6 infusions of NAT 2-sided Mann-Whitney U test and ARR prior/during/after NAT within at least 6 months, absence of anti-NAT-abs, availability of and all serologic tests by the Kruskal-Wallis test with Dunn clinical data for calculation of the annualized relapse rate (ARR), correction for multiple comparison using Prism v6 (Graph Pad, and availability of sufficient sample volume. Samples seropositive San Diego, CA). for AQP4-IgG were unblinded, and new serum samples were requested from the treating physician. The following clinical data were assessed: the ARR during RESULTS Description of study population. We NAT therapy (total number of attacks divided through the total screened a serum depository, which contained

2 Neurology: Neuroimmunology & Neuroinflammation 2,787 serum samples from 2,245 individual patients 1:1,000). Four of the samples identified by CBA were treated with NAT for suspected relapsing-remitting positive for AQP4-IgG by ELISA as well. Patient 1 MS. A total of 1,183 samples from 1,183 patients with weak staining by CBA (titer 1:32) was negative fulfilled the inclusion criteria (figure 1). The clinical for AQP4-IgG of the M23 isoform by ELISA. To rule characteristics of included patients were median age out that the negative ELISA result was caused by the 40 years, 71.9% female, median disease duration 8 use of the M23-AQP4 isoform in the AQP4-ELISA, years, median NAT infusions 9, median ARR prior to we tested that sample in addition in a second ELISA NAT 2, and median EDSS at the time of serum which uses human M1-AQP4 as antigenic substrate; sampling 3.5. Prior to NAT, 143 patients (12.1%) however, the sample was negative as well. A follow-up had received an immunosuppressive therapy at least sample from this patient taken 4.7 years after the first once and 920 patients (77.8%) an immunomodula- sample was again weakly positive in the CBA (titer tory therapy. 1:10) but negative in the 2 ELISAs. Given the repeatedly negative M1- and M23-ELISA results and Frequency of AQP4-abs in NAT-treated patients with the fact that the patient had clinical symptoms suspected MS. Of 1,183 patients, 5 (0.4%; 4 female) compatible with secondary-progressive MS rather were positive for AQP4-IgG by a CBA (table 1). than NMOSD (see below), patient 1 was finally The results were confirmed in 2 independent labo- classified as AQP4-IgG–seronegative and excluded ratories, which use the same AQP4-transfected from further analysis. Follow-up samples were also human embryonic kidney (HEK) 293 cells as test available from patient 4. This patient was positive by substrate. Signal strength at standard 1:10 starting CBA and ELISA both in a sample taken 10 months dilution was intense in 3 and weak in 2 patients. The after the start of NAT and in a sample obtained 4.9 median AQP4-ab titer was 1:320 (range 1:10– years later during fingolimod therapy. For patient 2, an earlier sample taken after 5 infusions of NAT ex- isted in the serum depository. Of interest, it was Figure 1 Flowchart of study population and assays negative by CBA, but the sample initially used for the study taken 7 months later after 14 infusions of NAT was strongly positive by both CBA and ELISA.

Disease course and diagnosis of AQP4-ab–seropositive patients treated with NAT. Epidemiological and clinical features of the 5 AQP4-ab–seropositive patients are summarized in table 2. Median age was 39.7 years and median disease duration was 8.1 years. Four patients had previously received immunomodulatory therapies, 2 additionally immunosuppressants. Four patients were seropositive for JC virus (JCV)-abs. At the time of serum sampling, the median number of NAT infusions was 12. In total, patients were treated with NAT for median 2.8 years (range 1.0–7.3 years). Reasons for NAT withdrawal were diagnosis of NMOSD in 2 patients (after 1.0 and 2.2 years, respectively), seropositivity for JCV-abs in 2 patients (after 2.8 and 4.2 years, respectively), and clinical deterioration in 1 patient (after 7.3 years). The diagnosis of NMOSD had to be rejected on clinical grounds in 1 female patient (1; AQP4-IgG positive in CBA and negative in ELISA). She had a history of severe and incompletely remitting attacks with double vision, sensory defects, and paresis of the left leg, but no episodes of optic neuritis. Brain MRI was always typical for MS with supratentorial and in- fratentorial lesions. Unfortunately, no spinal MRI A total of 2,787 samples sent for analysis of antinatalizumab antibodies (NAT-ab) were was performed throughout the clinical course. During screened. Samples from 1,183 patients with a diagnosis of MS were evaluated for therapy with IFN-beta-1b and glatiramer acetate 2–3 AQP4-IgG by a cell-based assay (CBA). Positive samples were retested with further CBAs and ELISAs. NMOSD was diagnosed according to the 2015 criteria. AQP4-ab 5 aquaporin- relapses per year occurred with relapse-associated 4antibody;NMOSD5 neuromyelitis optica spectrum disorder. disability up to EDSS 7.5. After 2 infusions of

Neurology: Neuroimmunology & Neuroinflammation 3 (figure e-1 at Neurology.org/nn). He had 2 relapses Table 1 Results of AQP4-ab testing during 4.2 years of NAT therapy and an increase of

AQP4-ELISA short spinal lesions. After switching to fingolimod, he AQP4-CBA AQP4-CBA AQP4-CBA titer (M23) Bochum, Patient Bochuma Euroimmun Euroimmun U/mL had 3 further relapses (1 optic neuritis and 2 episodes of myelitis) and subsequently, still under the assump- First sample tion of MS, received alemtuzumab. Four and 5 b 1 Positive Positive 1:32 Negative months later, 2 severe relapses with longitudinally 2 Positive Positive 1:1,000 800 extensive transverse myelitis occurred, which left the 3 Positive Positive 1:320 175 patient wheelchair-bound and finally established the

4 Positive Borderline 1:10 10.53b diagnosis of NMOSD.

5 Positive Positive 1:1,000 800 Patient 2 (female) had been initially diagnosed with relapsing-remitting MS because of recurring ep- Follow-up sample isodes of sensory disturbances and double vision.

1c Weak staining Borderline 1:10 Negativeb IFN-beta-1b was ineffective. She later had 2 episodes

4d Positive Positive 1:100 52.72 of myelitis, but no optic neuritis. The only available MRI showed short spinal cord lesions and multiple Abbreviation: AQP4-ab 5 aquaporin-4 antibody. supratentorial brain lesions. She had 1 relapse while Cell-based immunofluorescence assays (CBAs) for AQP4-abs were tested in 2 different laboratories. AQP4-ELISA detecting the M23 isoform (negative ,3.0 U/mL). on NAT therapy with a stable EDSS. NAT was dis- a No titer available. continued due to JCV-ab seropositivity after 2.8 b Also negative in M1-AQP4 ELISA. years. Subsequently, she was treated with glatiramer c 4.7 years after first sample (still on natalizumab therapy). acetate and had 2 mild relapses with stable EDSS d 4.9 years after first sample (during fingolimod therapy). and 2 new small spinal lesions at the last MRI follow-up. After switching to dimethyl fumarate, mitoxantrone, she was switched to NAT, which was she was relapse and progression free for the remaining given for a total of 7.3 years. During this time, 1 follow-up of 2.8 years. relapse with monoparesis of the left arm occurred. ARR in NAT-treated patients with NMOSD. To evaluate After some years of NAT treatment, walking gradu- the clinical efficacy of NAT in NMOSD, we compared ally deteriorated, she had to use a walker, and later the ARR before, during, and after NAT therapy. The became wheelchair-bound, which is why NAT was ARR of the 4 identified AQP4-ab–seropositive suspended. Subsequent therapies with dimethyl NMOSD patients decreased from median 2.0 (range fumarate and rituximab failed to control progres- 2.0–3.0) before, to 0.735 (0.36–2.27) during, and sion. The treating neurologist was convinced that 0.55 (0.30–1.0) after NAT therapy (the latter this patient suffers from secondary-progressive MS. including time without immunotherapy and on im- Two female AQP4-ab–seropositive patients (3 munotherapies recommended for NMOSD or not). and 5) had typical NMO according to the 2006 The ARR clearly decreased in 2 patients with an early Wingerchuk criteria;3 both patients also met the 2015 diagnosis of NMOSD and change from NAT therapy Wingerchuk criteria.4 Clinicoradiological features to mitoxantrone and rituximab, respectively (patients 3 included recurring episodes of myelitis and optic and 5), whereas it did not change in the 2 patients neuritis as well as longitudinally extensive lesions on switched to immunomodulatory therapies used for MS spinal cord MRI. Both patients had ongoing disease (patients 2 and 4) (table 2). activity (ARR 1.0 and 2.27, respectively) and Next, we examined all patients from our NAT- a worsening of the EDSS (0.5 and 1.5, respectively) treated cohort stratified for the AQP4-ab status. We during NAT treatment. After the failure of NAT, found a significant difference between AQP4-ab– both patients were switched to recommended seropositive NMOSD (median ARR 0.7, range NMOSD therapies (mitoxantrone and rituximab, 0.2–2.2) and AQP4-IgG–seronegative MS (median respectively). In both patients, the ARR decreased ARR 0, range 0–6.6; p 5 0.0182) (figure 2). Adding and the EDSS improved by 1.0 until the end of the 5 NAT-treated AQP4-ab–seropositive NMO follow-up. patients previously identified6 to the analysis, the Two further AQP4-ab–seropositive patients did ARR in NMOSD still was significantly worse not meet the 2006 criteria for NMO, but the revised (AQP4-ab seropositive median ARR 2.4, range 0.2– 2015 criteria for NMOSD when treated with NAT. 6.0; p , 0.0001 vs AQP4-ab negative). Patient 4 (male) had an initial episode of optic neuritis and 3 attacks of myelitis. The initial MRI showed Serological findings in AQP4-ab–seropositive NMOSD several short spinal cord lesions and, in addition, 2 patients treated with NAT. Since AQP4-IgG are path- supratentorial brain lesions adjacent to the lateral ogenic, we investigated the AQP4-ab titer in the 4 ventricle, all compatible with the diagnosis of MS NMONAT patients identified in the current study and

4 Neurology: Neuroimmunology & Neuroinflammation Table 2 Characteristics of aquaporin-4 antibody–seropositive NAT-treated patients

Patient 1 Patient 2 Patient 3 Patient 4 Patient 5

Sex FFFM F

Age,a y 39.7 39.3 25.4 42.3 49.3

Disease duration,a y 10.0 3.4 8.1 1.2 13.0

Previous treatmentsb IFN-b1b, GLAT, MITOX IFN-b1b IFN-b1a, IFN-b1b, GLAT None AZA, MITOX, IFN- b1b

NAT infusions,a n 914191012

Total duration of NAT treatment, y 7.3 2.8 2.2 4.2 1.0

Reason for NAT withdrawal Disease progression JCV-ab positive NMO diagnosis JCV-ab positive NMO diagnosis

Attack manifestations typical for NYYY Y NMOSD

MRI LETM NA N Y Y Y

MRI cerebral lesions YYNAY Y

ARR prior to NATc 2.0 3.0 2.0 2.0 2.0

ARR during NAT 0.14 0.36 2.27 0.48 1.0

Relapses during NAT 1d 15 2 1

EDSS start of NAT 5.0 4.5 4.0 1.5 5.5

EDSS end of NAT 7.0 4.5 4.5 1.5 7.0

Follow-up,a y 8.1 8.2 5.0 8.6 4.0

Treatment after NAT and duration DMF (0.8), RTX (1.0) GLAT (1.2), DMF (2.8) MITOX (NA) FTY (4.4), ALEM (0.5) RTX (4.0) of treatment,b y

Relapses after NAT 0d 2 (2 on GLAT) 3 (0 on MITOX) 5 (3 on FTY, 2 on ALEM) 2 (2 on RTX)e

ARR after NAT 0d 0.30 0.60 1.0 0.50

Last EDSS 7.5 4.5 3.5 8.0 6.0

Final clinical outcome Progressive Stable Improved Progressive Improved

Final diagnosis SPMS NMOSD NMOSD NMOSD NMOSD

Abbreviations: ALEM 5 alemtuzumab; ARR 5 annualized relapse rate; AZA 5 azathioprine; DMF 5 dimethyl fumarate; EDSS 5 expanded disability status scale; FTY 5 fingolimod; GLAT 5 glatiramer acetate; IFN 5 interferon; JCV 5 JC virus; LETM 5 longitudinally extensive transverse myelitis; MITOX 5 mitoxantrone; N 5 no; NA 5 not available; NAT 5 natalizumab; NMOSD 5 neuromyelitis optica spectrum disorder; RTX 5 rituximab; SPMS 5 secondary-progressive MS; Y 5 yes. a Relative to first serum sample. b Treatments appear in chronological order. c At least 12 months before the start of NAT. d Secondary-progressive (SP) disease course from year 6 after the start of NAT. e Mild relapses in the first year of RTX treatment.

5 previously identified NMONAT patients.6 As IL-1b (median 7.5 vs 2.8 pg/mL, p 5 0.0002), and a comparator, we collected serum samples from pa- significantly lower levels of CXCL-10 (median 32.8 tients with NMOSD who were not treated with NAT vs 160.6 pg/mL, p 5 0.0091) in NMONAT vs and divided this group into patients with immuno- NMOIT patients, whereas no differences were found therapy (NMOIT,n5 13) and untreated patients between NMONAT and NMOno IT patients. Levels of (NMOno IT,n5 9) (table e-1). We found signifi- IL-4, IL-6, IL-10, IL-17, and IL-21 were generally cantly higher AQP4-ab titers in NMONAT (median very low and did not differ significantly (figure 3 and 479.1 U/mL, range 10.5–799.2 U/mL) vs NMOIT not shown). (median 30.7 U/mL, range 0–510.1 U/mL; p 5 0.0058), but no difference to NMOno IT (median DISCUSSION We examined AQP4-ab in patients 124.71 U/mL, range 0–799.2 U/mL) (figure 3A). treated with NAT for suspected MS and determined Using the same experimental groups, we assessed their frequency and relevance for the clinical course. whether alterations in proinflammatory cytokines We found 4 of 1,183 patients (0.33%) to be and the chemokine CXCL-10 might contribute to unequivocally AQP4-ab seropositive. Before being increased disease activity in NMONAT patients (figure identified by our study, 2 of the 4 AQP4-ab– 3, B–H). We found significantly higher levels of seropositive patients already had been diagnosed with serum IL-8 (median 2.8 vs 0 ng/mL, p 5 0.0001), NMOSD and fulfilled classic diagnostic criteria,3 IFN-g (median 5.9 vs 1.7 pg/mL, p 5 0.0019), and leaving 2 unrecognized AQP4-ab–seropositive patients.

Neurology: Neuroimmunology & Neuroinflammation 5 patients with NMO never develop AQP4-IgG, we Figure 2 Annualized relapse rate during natalizumab therapy cannot rule out that more patients were falsely diagnosed with MS. Therefore, it is formally not possible to calculate assay specificity. However, given the low prevalence of NMOSD in Europe,16 it is highly likely that the majority of patients included in the MSNAT cohort had indeed MS. This would correspond to a very high specificity (1 false-positive result among .1,000 samples tested, or $99.9%) of the CBA used here. Our results are in line with a previous report, showing a false-positive rate of a fixed transfected CBA of 0.1% in a Northern Californian MS population.17 The high specificity of AQP4-abs for NMOSD was also confirmed in a recent meta-analysis.18 The clinical response to NAT therapy was dependent on the AQP4-ab status. As expected, the majority of patients with MS had no relapses, whereas none of the NMONAT patients was relapse free. Taking also the previous literature into account, it is apparent that NAT is not beneficial for most patients with NMOSD.6–8,19–24 A review of 19 patients published so far revealed that the majority of patients with AQP4-ab–seronegative patients with relapsing-remitting MS were compared with AQP4-ab– NMOSD treated with NAT for at least 3 months seropositive patients with confirmed NMOSD* and additionally 5 previously identified** NAT- treated patients with NMO.6 Whisker box plots show minimum, Q1, median, Q3, and maximum deteriorated (table e-2). Moreover, early appearance values. Mann-Whitney U test. AQP4-ab 5 aquaporin-4 antibody; ARR 5 annualized relapse of severe attacks with atypical cerebral manifestation rate; NAT 5 natalizumab; NMOSD 5 neuromyelitis optica spectrum disorder. was noted in a subset of NMONAT patients.6,21,22,24 As already suggested,6 we now could verify that Both had the final diagnosis of NMOSD according to AQP4-ab titers are indeed higher in NMONAT the 2015 criteria,4 which illustrates that NMOSD is patients as compared to NMOIT patients. We also only rarely misdiagnosed as MS nowadays when the found an increase of the AQP4-ab titer in 1 patient clinical course is followed over several years. during NAT therapy. Although it might be hypo- The AQP4-ab status of the 5 patients who were thesized that this elevated AQP4-ab serum level con- seropositive in the screening CBA was analyzed in tributes to relapses in NMONAT patients,15 the several additional, methodologically independent underlying mechanisms remain speculative. NAT assays to reduce the risk of center or assay bias.11,12 results in a rise in proinflammatory cytokines and Only 1 patient was classified as probably false posi- distinct T- and B-cell subsets in the periphery.25–27 tive for AQP4-IgG (0.08%). Previous reports sug- These changes of the peripheral immune milieu gested that NAT-containing serum might cause might promote high AQP4-ab titers and subsequent a false-positive signal in the AQP4-ab CBA, owing disease activity. Indeed, we found a higher level of to cross reactivity of NAT with AQP4-expressing proinflammatory cytokines, particularly IL-8, IFN-g, HEK cells used in this assay.13,14 Our results suggest and IL-1b, in the sera of NMONAT patients in that this might occur—if at all—only very rarely. comparison with NMOIT. An increase of T-helper However, since AQP4-abs are mainly IgG1 and (Th)17-, Th1- and Th2-related cytokines and che- NAT is an IgG4, it was proposed to use anti- mokines in the serum and particularly the CNS of human-IgG1 as a secondary antibody in the patients with NMOSD has been reported.28 Particu- CBA.13 Another solution in NAT-treated patients larly, IL-8 and IL-1b were shown to be increased in would be to add an independent confirmation the CSF of patients with NMOSD.29,30 CSF IL-8 method, e.g., an AQP4-ab ELISA. correlates with disease activity, EDSS, and length of It is unknown whether all 1,178 AQP4-ab– spinal cord lesions in NMOSD.28,31 seronegative patients were correctly diagnosed with The AQP4-ab and cytokine levels in NMONAT MS prior to NAT treatment. Our study revealed patients did not differ significantly in comparison a wrong diagnosis in at least 4 patients who were with untreated patients with NMOSD. This indi- AQP4-ab seropositive. Given that the sensitivity of all cates that NAT, unlike immunotherapies recommen- immunoassays is limited for reasons inherent to the ded for NMOSD, is not able to reduce methods used, that AQP4-ab serum levels vary proinflammatory cytokines which might be relevant substantially over time,15 and that 10%–20% of for the control of disease activity.

6 Neurology: Neuroimmunology & Neuroinflammation Figure 3 Titers of aquaporin-4 IgG and of proinflammatory cytokines in aquaporin-4 antibody–seropositive neuromyelitis optica spectrum disorder patients

Serological studies in NAT-treated patients with NMOSD (NMONAT,n5 9) and control NMOSD patients treated with immunotherapies (NMOIT,n5 13) or not (NMOno IT,n5 9). (A) Titer of AQP4-IgG measured by ELISA, serum was diluted 1:10. Some samples were saturated at the upper detection limit of 800 U/mL. Shown are individual patients and the median of each group. (B) Titer of IL-8 measured by ELISA, serum was diluted 1:10. (C–H) Other cytokines measured by cytometric bead array, serum was undiluted. Shown are individual patients and median. The Kruskal-Wallis test was used for statistical analysis. *Patients treated with tocilizumab were excluded from IL-6 analysis. AQP4 5 aquaporin-4; IFN 5 interferon; IL 5 interleukin; NAT 5 natalizumab; NMOSD 5 neuromyelitis optica spectrum disorder.

The downregulation of serum CXCL-10 in same route and mode of action for migration to the NMONAT patients seems counterintuitive, since CNS, but of interest, NAT treatment increases CXCL-10 might contribute to the recruitment of peripheral Th17 responses27,40 and CCR6 CXCR31 inflammatory cells to the CNS in expression of CD41 T-cells,26,40 which potentially NMOSD.32,33 CXCL-10 is elevated in the CSF but could worsen Th17-mediated diseases such as not serum of patients with NMO, MS, and other NMOSD. neuroinflammatory diseases.31,34,35 It was shown that Limitations of our study include the retrospec- CXCL-10 levels in the CSF (and to a lesser degree in tive evaluation of the clinical course and the low the serum) were reduced after 1 year of NAT treat- amount of serum available for serologic studies, ment for MS.36 preventing analysis of further inflammatory Although NAT reduces VLA-4 expression on markers. T-cells, B-cells, and monocytes/macrophages and We identified only few cases of AQP4-ab– thereby prevents migration of these cells to the seropositive NMOSD among patients with CNS,37 not all lymphocyte subsets are equally suspected MS, suggesting that the diagnostic criteria affected. Studies in the animal model experimental in daily practice are sufficient to discriminate between autoimmune encephalomyelitis (EAE) revealed that MS and NMOSD. Our data do not support routine Th17 but not Th1 cells can invade the CNS in the assessment of AQP4-abs in typical patients with MS. absence of VLA-4 expression and cause primarily However, in cases with recurrent optic neuritis or supraspinal infiltrates.38 It was proposed that differ- myelitis, opticospinal lesions on MRI, and particu- ential integrin expression promotes the targeting of larly when persistent disease activity occurs despite different Th-cell subsets to distinct niches of the optimal MS therapy, AQP4-ab testing should be CNS.38 In EAE, Th17 cells migrate to the CNS performed. As shown in previous studies, NAT had viathechoroidplexusthrough specific interaction limited efficacy in NMOSD, probably due to periph- of their chemokine receptor CCR6 with CCL20.39 eral inflammatory mechanisms induced by NAT It is unknown whether human Th17 cells use the treatment.

Neurology: Neuroimmunology & Neuroinflammation 7 AUTHOR CONTRIBUTIONS 6. Kleiter I, Hellwig K, Berthele A, et al. Failure of natalizu- A.G. analyzed and interpreted data and drafted the manuscript. A.-K.T., mab to prevent relapses in neuromyelitis optica. Arch S.H., and S.J. analyzed and interpreted data and revised the manuscript. Neurol 2012;69:239–245. M.R., O.A., A.B., and B.W. contributed patients and data and revised 7. Govindarajan R, Salgado E. Is it too early to predict the the manuscript. R.G. supervised the study and revised the manuscript. failure of natalizumab in NMO? Arch Neurol 2012;69: I.K. designed the study, obtained funding, analyzed and interpreted data, 1085; author reply 1086. did statistical analysis, and drafted the manuscript. All authors were 8. Barnett MH, Prineas JW, Buckland ME, Parratt JD, involved in revising the manuscript for intellectual content. Pollard JD. Massive astrocyte destruction in neuromyelitis ACKNOWLEDGMENT optica despite natalizumab therapy. Mult Scler 2012;18: – The authors thank all patients and the treating neurologists, particularly 108 112. Alexander Sieke, for their contributions and Christa Kraushaar-Sczesni, 9. Trampe AK, Hemmelmann C, Stroet A, et al. Anti-JC Marion Frickmann, Britta Haas, Anna Eschlbeck, and the Nikon Imaging virus antibodies in a large German natalizumab-treated Center at the University of Heidelberg for expert technical assistance. multiple sclerosis cohort. Neurology 2012;78:1736– 1742. STUDY FUNDING 10. Jarius S, Probst C, Borowski K, et al. Standardized Part of the study was funded by an unrestricted research grant of Biogen method for the detection of antibodies to aquaporin-4 Germany to I.K. The work of S.J. was supported by research grants from based on a highly sensitive immunofluorescence assay Merck Serono and from the Dietmar-Hopp-Stiftung to the Department employing recombinant target antigen. J Neurol Sci of Neurology, University of Heidelberg (Prof. Brigitte Wildemann). 2010;291:52–56. 11. Jarius S, Franciotta D, Paul F, et al. Testing for antibodies DISCLOSURE to human aquaporin-4 by ELISA: sensitivity, specificity, A. Gahlen received travel funding from Sanofi Genzyme. A.-K. Trampe and direct comparison with immunohistochemistry. and S. Haupeltshofer report no disclosures. M. Ringelstein received travel – funding from Novartis, Bayer Healthcare, Biogen Idec, Genzyme, Teva, J Neurol Sci 2012;320:32 37. and Merz. O. Aktas received travel funding and/or speaker honoraria 12. Granieri L, Marnetto F, Valentino P, et al. Evaluation of from Bayer, Novartis, Biogen, Merck Serono, Teva, Genzyme, Roche, a multiparametric immunofluorescence assay for standard- and Almirall; is an academic editor for PLoS ONE; consulted for Novar- ization of neuromyelitis optica serology. PLoS One 2012; tis; and received research support from Novartis, Bayer, Biogen, German 7:e38896. Ministry of Science, German Research Foundation, Heinrich and Erna 13. Cohen M, De Seze J, Marignier R, Lebrun C. False Schaufler-Foundation, and Walter and Ilse Rouse-Foundation. A. positivity of anti aquaporin-4 antibodies in natalizumab- Berthele received travel funding and/or speaker honoraria from Bayer treated patients. Mult Scler 2016;22:1231–1234. Healthcare, Biogen, Merck Serono, Mylan, Novartis, Roche, and Teva 14. Sanchez Gomar I, Diaz Sanchez M, Ucles Sanchez AJ, and received research support from Bayer Healthcare. B. Wildemann et al. An immunoassay that distinguishes real neuromyeli- served on the scientific advisory board for Novartis, Merck Serono, and Sanofi Genzyme; received travel funding and/or speaker honoraria from tis optica signals from a labeling detected in patients Biogen, Merck Serono, Novartis, Teva, and Sanofi Genzyme; and receiving natalizumab. BMC Neurol 2014;14:139. received research support from Bundesministerium für Forschung und 15. Jarius S, Aboul-Enein F, Waters P, et al. Antibody to Technologie, Dietmar Hopp Stiftung, Klaus Tschira Stiftung, Biogen, aquaporin-4 in the long-term course of neuromyelitis Merck Serono, Novartis, Teva, and Sanofi Genzyme. R. Gold served on optica. Brain 2008;131:3072–3080. the scientific advisory board for Teva Laquinimod DSMB; received 16. Pandit L, Asgari N, Apiwattanakul M, et al. Demographic speaker honoraria from Biogen, Genzyme, Teva, Merck Serono, Bayer and clinical features of neuromyelitis optica: a review. Schering, Ozgene, and Novartis; is on the editorial board for Therapeutic Mult Scler 2015;21:845–853. Advances in Neurologicals Disorders, Aktuelle Neurologie, and Experimental 17. Pittock SJ, Lennon VA, Bakshi N, et al. Seroprevalence of Neurology. S. Jarius reports no disclosures. I. Kleiter received travel fund- aquaporin-4-IgG in a northern California population rep- ing and/or speaker honoraria from Bayer Healthcare, Roche, and Merck; consulted for Bayer Healthcare, Chugai, Roche, and Shire; and received resentative cohort of multiple sclerosis. JAMA Neurol – research support from Biogen, Chugai, Diamed, and Affectis. Go to 2014;71:1433 1436. Neurology.org/nn for full disclosure forms. 18. Jarius S, Wildemann B. Aquaporin-4 antibodies (NMO- IgG) as a serological marker of neuromyelitis optica: Received January 24, 2017. Accepted in final form April 17, 2017. a critical review of the literature. Brain Pathol 2013;23: 661–683. REFERENCES 19. Kleiter I, Hellwig K, Berthele A, et al. Is it too early to 1. Jarius S, Wildemann B, Paul F. Neuromyelitis optica: predict the failure of natalizumab in NMO?-reply. Arch clinical features, immunopathogenesis and treatment. Clin Neurol 2012;69:1085–1086. Exp Immunol 2014;176:149–164. 20. Jacob A, Hutchinson M, Elsone L, et al. Does natalizumab 2. Zekeridou A, Lennon VA. Aquaporin-4 autoimmunity. therapy worsen neuromyelitis optica? Neurology 2012;79: Neurol Neuroimmunol Neuroinflamm 2015;2:e110. 1065–1066. doi: 10.1212/NXI.0000000000000110. 21. Jurynczyk M, Zaleski K, Selmaj K. Natalizumab and the 3. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti development of extensive brain lesions in neuromyelitis CF, Weinshenker BG. Revised diagnostic criteria for neu- optica. J Neurol 2013;260:1919–1921. romyelitis optica. Neurology 2006;66:1485–1489. 22. Kitley J, Evangelou N, Kuker W, Jacob A, Leite MI, 4. Wingerchuk DM, Banwell B, Bennett JL, et al. Interna- Palace J. Catastrophic brain relapse in seronegative tional consensus diagnostic criteria for neuromyelitis opti- NMO after a single dose of natalizumab. J Neurol Sci ca spectrum disorders. Neurology 2015;85:177–189. 2014;339:223–225. 5. Kleiter I, Gold R. Present and future therapies in neuro- 23. Lee DH, Laemmer AB, Waschbisch A, et al. Neuromye- myelitis optica spectrum disorders. Neurotherapeutics litis optica presenting with relapses under treatment with 2016;13:70–83. natalizumab: a case report. J Med Case Rep 2014;8:155.

8 Neurology: Neuroimmunology & Neuroinflammation 24. Kornberg MD, Newsome SD. Unmasking and provoking 33. Shimizu F, Nishihara H, Sano Y, et al. Markedly increased severe disease activity in a patient with NMO spectrum IP-10 production by blood-brain barrier in neuromyelitis disorder. Neurol Neuroimmunol Neuroinflamm 2015;2: optica. PLoS One 2015;10:e0122000. e66. doi: 10.1212/NXI.0000000000000066. 34. Narikawa K, Misu T, Fujihara K, Nakashima I, Sato S, 25. Krumbholz M, Meinl I, Kumpfel T, Hohlfeld R, Meinl E. Itoyama Y. CSF chemokine levels in relapsing neuromye- Natalizumab disproportionately increases circulating pre-B litis optica and multiple sclerosis. J Neuroimmunol 2004; and B cells in multiple sclerosis. Neurology 2008;71: 149:182–186. 1350–1354. 35. Ando H, Sato T, Tomaru U, et al. Positive feedback loop 26. Kivisakk P, Healy BC, Viglietta V, et al. Natalizumab via astrocytes causes chronic inflammation in virus- treatment is associated with peripheral sequestration of associated myelopathy. Brain 2013;136:2876–2887. proinflammatory T cells. Neurology 2009;72:1922–1930. 36. Mellergard J, Edstrom M, Vrethem M, Ernerudh J, Dahle 27. Bornsen L, Christensen JR, Ratzer R, et al. Effect of na- C. Natalizumab treatment in multiple sclerosis: marked talizumab on circulating CD41 T-cells in multiple scle- decline of chemokines and cytokines in cerebrospinal rosis. PLoS One 2012;7:e47578. fluid. Mult Scler 2010;16:208–217. 28. Uzawa A, Mori M, Kuwabara S. Cytokines and chemo- 37. Wipfler P, Oppermann K, Pilz G, et al. Adhesion mole- kines in neuromyelitis optica: pathogenetic and therapeu- cules are promising candidates to establish surrogate tic implications. Brain Pathol 2014;24:67–73. markers for natalizumab treatment. Mult Scler 2011;17: 29. Yanagawa K, Kawachi I, Toyoshima Y, et al. Pathologic 16–23. and immunologic profiles of a limited form of neuromye- 38. Rothhammer V, Heink S, Petermann F, et al. Th17 lym- litis optica with myelitis. Neurology 2009;73:1628–1637. phocytes traffic to the central nervous system indepen- 30. Uzawa A, Mori M, Arai K, et al. Cytokine and chemokine dently of alpha4 integrin expression during EAE. J Exp profiles in neuromyelitis optica: significance of interleukin- Med 2011;208:2465–2476. 6. Mult Scler 2010;16:1443–1452. 39. Reboldi A, Coisne C, Baumjohann D, et al. C-C chemo- 31. Matsushita T, Tateishi T, Isobe N, et al. Characteristic kine receptor 6-regulated entry of TH-17 cells into the cerebrospinal fluid cytokine/chemokine profiles in neuro- CNS through the choroid plexus is required for the initi- myelitis optica, relapsing remitting or primary progressive ation of EAE. Nat Immunol 2009;10:514–523. multiple sclerosis. PLoS One 2013;8:e61835. 40. Benkert TF, Dietz L, Hartmann EM, et al. Natalizumab 32. Chihara N, Aranami T, Oki S, et al. Plasmablasts as migra- exerts direct signaling capacity and supports a pro- tory IgG-producing cells in the pathogenesis of neuromye- inflammatory phenotype in some patients with multiple litis optica. PLoS One 2013;8:e83036. sclerosis. PLoS One 2012;7:e52208.

Neurology: Neuroimmunology & Neuroinflammation 9 Safety/tolerability of the anti-semaphorin 4D antibody VX15/2503 in a randomized phase 1 trial

Christopher LaGanke, ABSTRACT MD Objective: To evaluate the safety, tolerability, pharmacokinetics (PK), and pharmacodynamics (PD) Lawrence Samkoff, MD of VX15/2503 in a randomized, single-dose, dose-escalation, double-blind, placebo-controlled Keith Edwards, MD study enrolling adult patients with MS. Lily Jung Henson, MD Methods: Single IV doses of VX15/2503 or placebo were administered. Ten patients each were Pavle Repovic, MD randomized (4:1 randomization ratio) into 5 ascending dose cohorts of 1, 3, 6, 10, or 20 mg/kg. Sharon Lynch, MD Safety, immunogenicity, PK/PD, MRI, ECG, and lymphocyte subset levels were evaluated. A Dose Lael Stone, MD Escalation Safety Committee (DESC) approved each dose escalation. David Mattson, MD VX15/2503 was well tolerated, and all participants completed the study. Antibody Aaron Galluzzi, MSc Results: treatment–related adverse events were primarily grade 1 or 2 and included urinary tract infec- Terrence L. Fisher, PhD tion (12.5%) and muscle weakness, contusion, and insomnia (each 7.5%). No dose-limiting tox- Christine Reilly, BS icities were observed, and no maximum tolerated dose was determined. One subject (20 mg/kg) Laurie A. Winter, BS experienced disease relapse 3 months before study entry and exhibited a grade 3 (nonserious) John E. Leonard, PhD increase in brain lesions by day 29, possibly related to VX15/2503. Twenty-nine patients ex- Maurice Zauderer, PhD hibited human anti-humanized antibody responses; 5 with titer $100. No anti-VX15/2503 anti-

body responses were fully neutralizing. VX15/2503 Cmax, area under the time-concentration

curve, and mean half-life increased with dose level; at 20 mg/kg, the T1/2 was 20 days. Cellular Correspondence to SEMA4D saturation occurred at serum antibody concentrations #0.3 mg/mL, resulting in Dr. Leonard: [email protected] decreased cSEMA4D expression. At 20 mg/kg, cSEMA4D saturation persisted for $155 days. Total sSEMA4D levels increased with dose level and declined with antibody clearance. Conclusions: These results support the continued investigation of VX15/2503 in neurodegenerative diseases. ClinicalTrials.gov identifier: NCT01764737. Classification of evidence: This study provides Class III evidence that anti-semaphorin 4D antibody VX15/2503 at various doses was safe and well tolerated vs placebo, although an increase in treatment-emergent adverse events in the treatment group could not be excluded (risk difference 20.7%, 95% CI 228.0% to 32.7%). Neurol Neuroimmunol Neuroinflamm 2017;4:e367; doi: 10.1212/NXI.0000000000000367

GLOSSARY AE 5 adverse event; AUC 5 area under the time-concentration curve; BBB 5 blood-brain barrier; CTCAE 5 Common Terminology Criteria for Adverse Event; C-SSRS 5 Columbia–Suicide Severity Rating Scale; EAE 5 experimental autoimmune encephalomyelitis; DESC 5 Dose Escalation Safety Committee; DLT 5 dose-limiting toxicity; EDSS 5 Expanded Disability Status Scale; EOI 5 end of infusion; Gd 5 gadolinium; HAHA 5 human anti-human antibody; MTD 5 maximum tolerated dose; OPC 5 oligodendrocyte precursor cell; PD 5 pharmacodynamics; PK 5 pharmacokinetics; QTc 5 QT interval; SAE 5 serious adverse event; TEAE 5 treatment-emergent adverse event.

Semaphorins are a family of soluble and transmembrane proteins serving as axonal-guidance factors and other functions in the development and regeneration of the CNS.1 They also participate in vascular growth, tumor progression, and the activation and migration of immune and Supplemental data inflammatory precursor cells. at Neurology.org/nn Author affiliations are provided at the end of the article. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by Vaccinex, Inc. Coinvestigators are listed at Neurology.org/nn. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology 1 Semaphorin 4D (SEMA4D) is a 300-kDa autoimmune encephalomyelitis in multiple transmembrane protein predominantly ex- rodent models.13 pressed on T cells, but also expressed on VX15/2503, a high-affinity humanized monocytes, professional antigen-presenting monoclonal anti-SEMA4D antibody derived cells, platelets, and oligodendrocytes.2 Cellular from MAb 67-2, blocks the interaction activation stimulates increased expression of between SEMA4D and its three receptors.13– cSEMA4D. In addition, the extracellular 16 This article describes the results of a phase 1 domain of cSEMA4D can be proteolytically study evaluating the safety and tolerability of cleaved from the cell surface yielding a 240- VX15/2503 in patients with MS; no similar kDa, homodimeric soluble form of the protein trials have been described. We conducted this (sSEMA4D)3; both forms are biologically study to evaluate VX15/2503 as a potential active.4 Finally, although SEMA4D functions therapeutic agent for MS and, possibly, other primarily as a ligand, it may also function as neurodegenerative diseases. a receptor, signaling through its cytoplasmic 5 domain. METHODS Study drug. VX15/2503 was produced by Cata- Three cellular receptors have been identi- lent Pharma Solutions (Madison, WI) and vialed by Ajinomoto fied for SEMA4D. Plexin-B1 (PLXNB1), Althea, Inc. (San Diego, CA)14,16; generic and proprietary names a high-affinity receptor, is expressed on den- have not been assigned. A matched placebo was supplied for comparison of safety observations (see appendix e-1 at dritic and endothelial cells, oligodendrocytes, Neurology.org/nn). astrocytes, and neurons.6 SEMA4D engage- Study design. This phase 1 study was a single-dose, dose- ment with PLXNB1 induces activation and escalation, randomized, double-blind, placebo-controlled trial migration of endothelial cells; it also induces enrolling adult patients diagnosed with relapsing or progressive growth cone collapse in neurons, apoptosis of MS for at least 1 year as defined by the McDonald criteria.17 The primary protocol-specified objective was to determine the safety neural precursor cells, and process extension and tolerability of VX15/2503 in patients with MS; secondary 7–9 collapse and apoptosis of oligodendrocytes. and exploratory objectives were to characterize the single-dose Plexin-B2 (PLXNB2), a SEMA4C receptor pharmacokinetics (PK), pharmacodynamics (PD), and immu- expressed on keratinocytes, has intermediate nogenicity of VX15/2503 (see appendix e-1). No interim analysis was planned, and no changes were made to study objectives or affinity for SEMA4D but can activate trial design after study initiation. SEMA4D-positive gd T cells aiding epithelial The study was conducted at 11 US clinical centers. Each of repair.10 Finally, CD72 is a low-affinity the 5 dose cohorts included 10 patients; ascending dose levels SEMA4D receptor that influences B-lymphocyte of 1, 3, 6, 15, and 20 mg/kg were used based on results of completed toxicology and clinical studies.14,16 Enrolled participants 11 maturation. were randomly assigned (4:1 randomization ratio, drug to pla- MS is a chronic neuroinflammatory disease cebo) on study day 1 to receive a single IV dose of either VX15/ characterized by blood-brain barrier (BBB) 2503 or placebo; each cohort contained 8 antibody-treated patients. The randomization schedule was software generated; breakdown, localized myelin destruction, patient treatment assignments were communicated using an and progressive neuronal degeneration. interactive voice response system. This system also prescribed, SEMA4D-induced signaling cascades induce for each patient, specific numbered vials of investigational agent. glial activation, neuronal process collapse, Vital signs were measured before, during, and after study drug infusion. Intracohort patient dosing was staggered by at least 24 inhibit migration and differentiation of oligo- hours after the first 4 patients in the prior cohort had been dosed; dendrocyte precursor cells (OPCs), and dis- study drug administration to the remaining patients in a cohort rupt endothelial tight junctions forming the was staggered by at least 2 hours between patients. No more than BBB. Because SEMA4D mediates both 3 patients were infused in any 1-day period; all patients were monitored for 24 hours after dosing. On study day 2, serum 12 inflammatory responses and demyelination, chemistry, hematology, PK/PD, and lymphocyte enumeration it is a potential target for treatment of neuro- samples, among others, were obtained from each subject. Partic- degenerative diseases.6 The murine anti- ipants also received an in-person or telephonic clinical assessment on day 4. Safety follow-up periods of 6 weeks (cohort 1), 8 weeks SEMA4D antibody MAb 67-2 blocks (cohorts 2 and 3), 12 weeks (cohort 4), or 27 weeks (cohort 5), SEMA4D binding to OPC in vitro and re- based on expected durations of cSEMA4D saturation, were spec- duces semaphorin-mediated apoptosis13;it ified for collection of comprehensive safety, PK, and PD samples. also promotes OPC migration to the site of Escalation to the next dose cohort proceeded only after the DESC (consisting of three independent neurologists with clinical lesions, repairs lysolecithin-induced demyelin- expertise in MS, the sponsor’s representatives, and the study ation in vivo, and attenuates experimental medical monitor) determined by data review that the previous

2 Neurology: Neuroimmunology & Neuroinflammation dose level was safe and well tolerated. Cohort data reviews were with alemtuzumab, rituximab, mitoxantrone, total lymphoid irra- blinded and included adverse events (AEs), ECGs, vital signs, and diation, bone marrow transplantation, or T cell or T-cell receptor clinical laboratory and lymphocyte subset results through day 15, vaccination. Patients were also excluded if they had received any and human anti-humanized antibody (HAHA)18 and PK/PD experimental agent within 6 months before dosing, or within data; the committee made no requests for review of unblinded a period equivalent to 5 half-lives of the agent (whichever was data. longer); or were currently involved in any other research study, or had undergone any major surgical procedure within 4 weeks Stopping rules. The infusion for any patient was to be stopped before dosing. Participants were also excluded if they had a history if the patient experienced a possibly drug-related serious AE of congestive heart failure, symptomatic ischemia, conduction (SAE) or a possibly drug-related significant non–serious AE abnormalities uncontrolled by conventional intervention, or which, in the opinion of the principal investigator, study medical myocardial infarction within 6 months before dosing, or had monitor, or sponsor’s medical representative, warranted discon- a clinically significant ECG finding at screening, as determined tinuation of the study for that patient. The occurrence of any of by the investigator. Finally, patients with sensitivity to VX15/ the following drug-related safety events also warranted DESC 2503 or its excipients, with a known or suspected HIV or hep- evaluation and possible subject discontinuation: one or more atitis B or hepatitis C infection, or with a known or suspected SAEs; a severe AE or an elevation in a protocol-specified liver allergy to gadolinium (Gd) or other contraindications to brain enzyme occurring in $2 participants at the same dose level; or an MRI were excluded. increase in corrected QT interval (QTc) of 60 milliseconds over baseline on 2 consecutive days or any QTc .500 milliseconds in Safety assessments. In addition to determining the safety and $2 patients in the same dose cohort (see appendix e-1 for tolerability of VX15/2503 in patients with MS, we also character- additional details). ized the single-dose PK, PD, and immunogenicity of VX15/ 2503. Safety assessments included AEs and SAEs; immunoge- Standard protocol approval, registrations, and patient nicity (Ig levels; HAHA); enumeration of T/B lymphocytes and consents. Institutional review board approvals for the study pro- natural killer cells; clinical laboratory evaluations (hematology, tocol, amendments, and informed consent documents were ob- coagulation parameters, serum chemistry, and urinalysis); vital tained before use in the study; written informed consent was signs; 12-lead ECGs; physical examination findings; and obtained from study participants before the initiation of study Columbia–Suicide Severity Rating Scale (C-SSRS); these assess- procedures. This study was conducted in accordance with the ments were used in the determination of an maximum tolerated Declaration of Helsinki, International Conference on Harmoni- dose (MTD) by the DESC (see appendix e-1 for additional zation (ICH) guidelines, and applicable portions of the United information). States Code of Federal Regulations. The ClinicalTrials.gov AE severity was rated according to the Common Terminol- identifier (NCT01764737) was obtained before study initiation. ogy Criteria for Adverse Events (CTCAE) version 4.03; AEs that Study participants. Men or women 18–60 years of age pre- were possibly or probably related to the study drug and were viously diagnosed with MS for at least 1 year as defined by the ongoing at the time a patient completed the study or discontin- McDonald criteria, and who had an Expanded Disability Status ued early were followed until resolution or until stabilized. All Scale (EDSS) score of 0–6.5, inclusive, at screening were eligible treatment-emergent AEs were reported to the Food and Drug for enrollment if they met the following criteria: a body mass Administration. index of 18–32 kg/m2, inclusive; willingness to undergo and PK/PD assays. Blood samples were collected at screening, pre- had no contraindications to brain MRI; willingness to use a med- dose, after the end of infusion (EOI), and on days 1, 2, and 4 and ically acceptable method of contraception throughout the study then periodically thereafter; PK/PD samples were analyzed as period and for 6 months after the dose of study drug, unless described.16 No CSF samples were collected for analysis. PK patient was surgically sterile or postmenopausal. Before study parameters were estimated for the PK population using non- drug administration, male patients had to agree to defer from compartmental analyses. Standard PK parameters (Cmax;CL; donating sperm for 6 months, and women of childbearing poten- Tmax; and effective half-life, etc.) were determined by cohort and tial were required to have a negative serum pregnancy test at time point. Analyses of area under the time-concentration curve screening (confirmed via baseline urinalysis). Finally, participants (AUC) and Cmax to assess dose proportionality were performed as on stable treatment regimens of interferon b, glatiramer acetate, or described.19 PD evaluations for peripheral blood T-cell dalfampridine were allowed to enroll. cSEMA4D expression and saturation levels as well as for serum Exclusionary criteria were an MS relapse that did not stabilize sSEMA4D levels were performed as described,16,20 using samples within 30 days before the start of screening; any clinically signif- collected at PK time points (see supplemental materials). As icant cardiac, endocrine, hematologic, hepatic, immunologic, VX15/2503 half-life increased with dose level, final cSEMA4D metabolic, urologic/gynecologic, pulmonary, neurologic, psychiatric, saturation samples were collected on day 43 (cohort 1), day 57 or renal conditions; a history of relevant clinically significant allergic (cohorts 2 and 3), and day 85 (cohorts 4 and 5), respectively; or anaphylactic reactions; any other clinically significant major cohort 5 subjects showing persistent cSEMA4D saturation had disease that, as assessed by the investigator, could have posed a risk additional samples collected on days 155 and 190 for analysis, as to patient safety or interfered with the study evaluations, procedures, necessary. or completion; any clinically significant laboratory value outside the normal range at screening; or abnormal hematologic, renal, or Statistical analysis. Because the primary objective of this study hepatic function based on laboratory tests. was assessment of the safety and tolerability of VX15/2503, sta- Pregnant or breastfeeding women were excluded. Patients tistical analyses of typical efficacy variables were not performed. who had received treatment with any MS disease-modifying ther- Results from the PK population were used for summary statistics apy other than interferon b or glatiramer acetate within 3 months and determination of mean serum VX15/2503 concentrations. before dosing were also excluded, as were subjects who had Descriptive statistics (number of patients, mean, SD, percent received natalizumab, daclizumab, or fingolimod for any indica- coefficient of variation, median, minimum, maximum, and the tion within 6 months before dosing, or had any prior treatment 25th and 75th percentiles) were used to summarize the serum

Neurology: Neuroimmunology & Neuroinflammation 3 Figure 1 Disposition of study participants enrolled in the single ascending dose phase 1 study of VX15/2503

This flow diagram depicts the disposition of the participants in this single-dose, dose-escalation study of intravenously administered VX15/2503. Forty participants were treated with VX15/2503 and 10 with placebo; none discontinued the study. The median infusion time for all patients was 60 minutes, with durations ranging from 45 to 170 minutes.

concentrations at each scheduled time point by treatment. Geo- Each patient received one infusion. The mean metric mean values were presented for Cmax,CL, and AUC body weight at screening for the treated population values. (n 5 40) was 77.7 kg (range 51.7–115 kg); the mean Disease activity assessments. Biological activity was assessed total dose administered was 617.9 mg (range 50.0– in an exploratory manner using brain MRI and EDSS data.21 1,940 mg). Serum sample analysis indicated that no EDSS measurements were performed at baseline, on day 29 placebo patients received active drug. Dose escalation and at time points specified for PK assessments. Brain scans were proceeded sequentially to 20 mg/kg VX15/2503, the performed at baseline and study day 29 with and without Gd; all MRIs were centrally analyzed and evaluated for the number of highest planned dose level; no dose-limiting toxicities Gd-enhancing lesions, the number of new and existing un- were observed, and no MTD was determined. enhancing T1 lesions, the number of new and enlarging T2 lesions, and total volume of T1 and T2 lesions (see appendix e-1). Safety assessments. Treatment-emergent adverse events (TEAEs) were reported for 26 of 40 (65%) VX15/ RESULTS Demographics and participant disposition. 2503-treated patients (table 1); no dose-dependent Fifty patients were enrolled and randomized pattern was evident in TEAE incidence. TEAEs for (figure 1); the sample size was not based on statistical 18 of 26 antibody-treated patients (69.2%) and for 7 considerations. Baseline characteristics were similar of 10 placebo-treated patients were considered unre- between treatment and placebo groups (see tables lated to treatment. Treatment-related TEAEs were 1S and 2S; appendix e-1). Twenty-nine (58%) of noted for 8 (20%) antibody-treated patients and 3 the subjects were women and 44 (88%) were Cau- placebo patients; no participants experienced a TEAE casian; 5 patients (10%) were African American. The leading to a fatal outcome, early termination, or dis- median participant age was 48 years (range 22–60); continuation. TEAEs most commonly reported by the median time since MS diagnosis for the study VX15/2503-treated patients were for urinary tract population overall was 12.0 years (range 2–40 years). infections (5 patients; 12.5%), and muscular weak- Forty-seven patients (94%) had received prior MS ness, contusion, and insomnia (3 subjects each; therapy. The mean EDSS score at baseline for the 7.5%). Three subjects (7.5%) experienced grade 3 treated population (n 5 40) was 3.8 (SD 5 2.0) vs an TEAEs after VX15/2503 infusion, and 23 partic- EDSS of 5.1 (SD 5 1.7) for the placebo population ipants (57.5%) experienced grade 1 or 2 events. The (n 5 10). The first participant was enrolled on severe TEAEs included one SAE (foot fracture November 29, 2012, and the last subject’s final visit requiring hospitalization and surgery; 3 mg/kg; was on November 14, 2014. There were no early unrelated); fatigue (1 patient; 6 mg/kg; nonserious; terminations or discontinuations; all participants unrelated); and CNS lesion (1 patient; 20 mg/kg; were analyzed in the safety and exploratory nonserious; possibly related). The latter participant populations. experienced a grade 3 increase from baseline to day 29

4 Neurology: Neuroimmunology & Neuroinflammation Table 1 Summary of treatment-emergent AEs for the safety population after a single dose of VX15/2503 or placebo

VX15/2503, mg/kg

AEs, N (%) Placebo (N 5 10) 1.0 (N 5 8) 3.0 (N 5 8) 6.0 (N 5 8) 10 (N 5 8) 20 (N 5 8) Total (N 5 40) Total (N 5 50)

Patients reporting AE 4 (40.0) 7 (87.5) 5 (62.5) 3 (37.5) 4 (50.0) 7 (87.5) 26 (65.0) 30 (60.0)

Patients reporting TEAE 4 (40.0) 7 (87.5) 5 (62.5) 3 (37.5) 4 (50.0) 7 (87.5) 26 (65.0) 30 (60.0)

Patients with treatment-related TEAEa 2 (20.0) 2 (25.0) 0 (0.0) 1 (12.5) 1 (12.5) 4 (50.0) 8 (20.0) 10 (20.0)

Patients with serious TEAEb 0 (0.0) 0 (0.0) 1 (12.5) 0 (0.0) 0 (0.0) 0 (0.0) 1 (2.5) 1 (2.0)

Patients with treatment-related 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (12.5) 1 (2.5) 1 (2.0) TEAE; CTCAE Grade ‡3c

Most frequent TEAEd

Urinary tract infection 0 (0.0) 1 (12.5) 2 (25.0) 0 (0.0) 1 (12.5) 1 (12.5) 5 (12.5) 5 (10.0)

Muscular weakness 0 (0.0) 3 (37.5) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 3 (7.5) 3 (6.0)

Contusion 1 (10.0) 1 (12.5) 1 (12.5) 0 (0.0) 1 (12.5) 0 (0.0) 3 (7.5) 4 (8.0)

Insomnia 0 (0.0) 1 (12.5) 1 (12.5) 0 (0.0) 0 (0.0) 1 (12.5) 3 (7.5) 3 (6.0)

Pain in extremity 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (12.5) 1 (12.5) 2 (5.0) 2 (4.0)

Upper respiratory tract infection 0 (0.0) 0 (0.0) 1 (12.5) 1 (12.5) 0 (0.0) 0 (0.0) 2 (5.0) 2 (4.0)

Fatigue 0 (0.0) 0 (0.0) 0 (0.0) 1 (12.5) 0 (0.0) 1 (12.5) 2 (5.0) 2 (4.0)

Headache 1 (10.0) 0 (0.0) 0 (0.0) 1 (12.5) 0 (0.0) 1 (12.5)e 2 (5.0) 3 (6.0)

Rash 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (25.0) 2 (5.0) 2 (4.0)

Fall 1 (10.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (12.5) 1 (12.5) 2 (5.0) 3 (6.0)

Abbreviations: AE 5 adverse event; CTCAE, Common Terminology Criteria for Adverse Events; Gd, gadolinium; TEAE 5 treatment-emergent adverse event. A summary of the AEs arising after a single dose of VX15/2503 or placebo. Each dose cohort contained ten randomized patients, 8 of whom were treated with VX15/2503. a Possibly, probably, or definitely related to study treatment; CTCAE version 4.0. b Foot fracture resulting from automobile accident requiring hospitalization and surgical repair; unrelated to study treatment. c Grade 3 increase in Gd-enhancing lesions; baseline to day 29; nonserious; possibly related to study treatment. d Reported for incidence $5% of VX15/2503-treated participants; TEAEs were grade 1 or 2. e Grade 3; unrelated.

of the number of T1 Gd-enhancing lesions from 11 various populations after VX15/2503 administra- to 27, and an increase in the EDSS score from 3.5 to tion, regardless of the dose. 4.0. Although the lesion increase was considered possibly related to VX15/2503 treatment, this patient Immunogenicity. HAHA responses were confirmed experienced a relapse 3 months before study entry beginning on day 8 for 25 of 40 VX15/2503- and exhibited active disease at screening. treated patients (62.5%), and on day 15, or later, In general, there were no clinically important for another 4 patients. Five of these 29 patients changes in physical examination findings, vital signs, (17.2%) (cohorts 1–3) exhibited HAHA responses serum chemistry, hematology, coagulation parame- requiring serum dilutions of $100; the remainder ters, or urinalysis values after administration of exhibited low-titer responses. One of these 5 patients VX15/2503. Similarly, C-SSRS results revealed no (cohort 2) exhibited a T-cell SEMA4D saturation suicidal ideation, regardless of the dose level. No value of 2% on day 43, vs 88–95% for the 7 other changes in ECG results occurred that suggested cohort 2 patients, possibly reflecting a delayed a possible treatment effect. Central evaluation of HAHA (neutralizing) response. Cellular SEMA4D ECG results noted transient, clinically nonsignifi- desaturation occurred on day 57 for the remaining cant shifts from normal to abnormal for 3 of 40 pa- antibody-treated cohort 2 participants. No other tients (7.5%) treated at 3, 6, or 10 mg/kg dose level HAHA-positive patients exhibited similar effects on of VX15/2503; no dose correlation was apparent, cSEMA4D saturation. and no similar shifts were noted for placebo- TEAEs reported on day 1 possibly reflective of treated patients. Last, flow cytometric analyses of infusion-related reactions included 1 placebo-treated lymphocytes and lymphocyte subsets showed no patient (mild dizziness/flushing and mild hyperhidro- obvioustrendsinchangesinthelevelsofthese sis; possibly related) and 4 VX15/2503-treated

Neurology: Neuroimmunology & Neuroinflammation 5 roughly 30 days after infusion at the 1 mg/kg dose Figure 2 Semi-log plot of mean VX15/2503 serum concentrations vs time after a single infusion level. With the increasing VX15/2503 dose level, longer periods of cSEMA4D saturation were observed, with $155 days required after infusion for full cSEMA4D desaturation for patients treated with 20 mg/kg VX15/2503. Mean saturation values for samples collected from placebo-treated participants ranged from zero to 13% (median 0.5%; not shown). Cellular SEMA4D expression levels decreased to approximately 50% after VX15/2503 binding to the cell surface receptor and its internalization and returned to baseline expression levels after antibody clearance, as described previously.16 Levels of total sSEMA4D (free and antibody bound) in the sera of antibody-treated patients increased with VX15/2503 dose level and peaked at roughly day 30 (not shown) primarily because of complexed sSEMA4D (see appendix e-1). Soluble SEMA4D levels returned to baseline after antibody clearance; mean sSEMA4D levels reached to baseline by day 43 for the 1 mg/kg dose cohort and day 155 for the 20 mg/kg dose cohort. Placebo-treated pa-

Mean serum VX15/2503 concentrations are shown vs scheduled sampling days. Data tients showed no similar PD effects. shown are for antibody-treated patients in each of the 5 dose cohorts; samples from pla- MS disease activity. cebo-treated subjects were assayed but not reported because of the absence of antibody Although efficacy was not formally in these samples. Sampling times were end of infusion, 4 and 8 hours on days 1 2, 4, 8, evaluated, EDSS assessments at day 29 remained 15, 29, 43, 57, 71, 85, 155, and 190; time points beyond day 29 were successively applied essentially unchanged compared with baseline follow- 6 to cohorts 2 through 5. Data shown represent mean values SD. Serum assays were ing single doses of VX15/2503 from 1 to 20 mg/kg. performed in triplicate using a validated method.15,16 Similarly, no effects on MS disease activity were evident based on brain MRI results (appendix e-1 and patients; 1 each in cohort 1 (mild nausea; probably table 3S). related; low-titer response) and cohort 4 (moderate blood pressure elevation 1 hour after infusion; unre- DISCUSSION Single IV infusions of VX15/2503 lated; low titer), and 2 VX15/2503-treated (low-titer) were well tolerated by the patients with MS enrolled patients in cohort 5 (subject 1: mild headache and in this study. The highest planned dose of 20 mg/kg mild rash; both unrelated, and subject 2: moderate was evaluated, and no MTD was determined. Analy- myalgia; unrelated). sis of the VX15/2503-treated population for TEAE PK and PD. Serum antibody concentration data incidence and severity revealed no safety trends. No derived from analysis of samples from the 40 clinically meaningful changes were reported for VX15/2503-treated patients are shown in figure 2; physical examination, laboratory parameters, or target-mediated disposition appears evident at lower C-SSRS results for antibody-treated patients. serum antibody concentrations. Mean half-life and Although 29 VX15/2503-treated patients developed

Cmax values (table 2) increased linearly with increas- HAHA responses, most were low-titer responses and ing dose level, ranging from roughly 4 days at none exhibited fully neutralizing antibodies. 1 mg/kg VX15/2503 to approximately 20 days Pharmacologic effects were similar to those re- 14,16 at 20 mg/kg; AUC0–N values, however, increased at ported previously for VX15/2503. Lymphocyte a rate greater than proportional to the dose. levels generally remained unchanged from baseline, VX15/2503 serum concentrations of #0.3 regardless of the antibody dose level. The threshold mg/mL, attained with single doses of VX15/2503 $ for cSEMA4D saturation of human T cells was #0.3 1.0 mg/kg, produced complete T-cell SEMA4D sat- mg/mL, as previously reported.16 Furthermore, as uration; cSEMA4D desaturation occurred with observed previously, expression of cSEMA4D on VX15/2503 clearance from the periphery (figure 3). peripheral blood T lymphocytes declined for all The mean cSEMA4D saturation value at the EOI for VX15/2503-treated patients because of receptor all 5 dose levels was 88.0% (range 68.5%–98.7%). internalization and returned to baseline levels Cellular SEMA4D saturation fell with antibody clear- after antibody clearance.16 Also, serum levels of total ance and reached approximately 20% saturation at sSEMA4D increased with the dose level consistent

6 Neurology: Neuroimmunology & Neuroinflammation Table 2 PK parameters derived from noncompartmental analysis of VX15/2503 PK population data

VX15/2503 dose level, mg/kg

Parameter 1.0 (N 5 8) 3.0 (N 5 8) 6.0 (N 5 8) 10 (N 5 8) 20 (N 5 8)

AUC0–N,h$ng/mL

n 788 8 8

Mean (SD) 1,479,141.5 5,912,533.8 14,078,216.2 31,895,922.7 89,606,735.2 (365,733.97) (971,930.82) (2,736,884.47) (7278649.05) (34,207,160.30)

Geometric mean 1,437,307.3 5,844,994.6 13,838,405.9 31,146,683.8 83,803,601.7

Median 1,677,427.6 5,669,043.0 14,292,817.0 32,527,928.8 83,058,364.6

Cmax, ng/mL

n 888 8 8

Mean (SD) 19,761.9 (7,607.82) 45,455.2 (4,563.71) 96,961.7 (15,362.86) 168,827.8 (39,726.11) 274,287.0 (65,226.10)

Geometric mean 18,592.2 45,241.3 95,947.0 164,901.3 267,814.8

Median 18,301.6 46,018.5 93,620.3 156,083.1 263,206.9

Tmax,h

n 888 8 8

Mean (SD) 5.1 (8.25) 3.7 (3.49) 2.2 (3.03) 1.7 (1.32) 5.2 (4.05)

Geometric mean 2.4 2.5 1.5 1.4 3.5

Median 1.2 1.6 1.1 1.3 5.5

t1/2,d

n 788 8 8

Mean (SD) 3.7 (1.61) 6.1 (1.87) 10.1 (2.68) 14.9 (5.95) 20.0 (4.82)

Geometric mean 3.5 5.8 9.8 14.1 19.4

Median 2.9 6.4 9.5 12.9 19.8

CL, mL/h

n 788 8 8

Mean (SD) 50.5 (8.39) 38.5 (11.01) 35.2 (13.22) 23.6 (7.67) 19.7 (5.49)

Geometric mean 49.9 37.1 33.3 22.4 19.0

Median 50.2 38.3 32.2 23.0 18.2

Abbreviations: AUC0–N 5 area under the concentration-time curve extrapolated to infinity; CL 5 clearance; Cmax 5 maximum serum concentration; t1/2 5 antibody half-life; PK 5 pharmacokinetics; Tmax 5 time of maximum serum concentration. A summary of the PK parameters determined after a single IV administration of VX15/2503. Summary statistics for each PK parameter were determined for each dose level. Geometric mean values were presented for Cmax, CL, and AUC values. Analyses of AUCs and Cmax to assess dose proportionality for VX15/2503 were performed using a power model.20 No subjects randomized to receive placebo were treated with the study drug. Half-life in days is derived from the data for half-life in hours by dividing the values by 24.

with the increased half-life of the VX15/2503- The findings from this study suggested that sSEMA4D complex vs those of the soluble ligand; VX15/2503 was well tolerated by this patient popu- these levels also declined after antibody clearance.15,16,22 lation. Use of this anti-SEMA4D antibody represents A semi-log plot of serum antibody concentration a potential therapeutic strategy for subjects with neu- vs time suggested that VX15/2503 clearance was pri- rodegenerative disease. Nonclinical studies indicate marily due to target-mediated disposition (figure 2), that this approach may promote survival, differentia- as is commonly observed for monoclonal antibodies tion, and migration of OPCs, all negatively affected targeting cellular ligands.14 The effective half-life by SEMA4D, potentially promoting remyelination increased linearly from approximately 4 days at 1 of damaged neurons.13 The possible clinical benefits mg/kg to roughly 20 days at 20 mg/kg VX15/2503; of VX15/2503 will be evaluated in future studies

Cmax values also increased linearly over this dose enrolling patients with MS. range. AUC0–168 values, however, exhibited a slightly Anti-SEMA4D treatment may also reverse dam- greater than dose proportional increase with the age to the BBB,13 and nonclinical studies using the increasing dose level. YAC-128 mouse model of Huntington disease2

Neurology: Neuroimmunology & Neuroinflammation 7 participants and participated in data acquisition and study coordination. Figure 3 Mean percent cellular SEMA4D saturation vs time after VX15/2503 Mr. Galluzzi performed the statistical analyses. Mrs. Reilly and Ms. Winter administration participated in data acquisition and the review and revision of the manuscript. All authors exercised editorial control of the content of the manuscript and approved the final version. Vaccinex, Inc. provided financial support for conduct of the study, data acquisition and analysis, and the writing of the manuscript.

STUDY FUNDING This phase 1 study was funded solely by Vaccinex, Inc. Vaccinex, Inc., a private Delaware corporation, has patent rights based on inventions described in this publication and is developing a humanized anti- SEMA4D antibody for clinical use. Dr. Fisher, Dr. Zauderer, and Dr. Leonard, and Mrs. Reilly are employees of or have ownership interest in Vaccinex, Inc. Ms. Winter was a Vaccinex employee at the time she participated in data acquisition.

DISCLOSURE C. Laganke reports no disclosures. L. Samkoff received publishing royalties from John Wiley & Sons. K. Edwards received travel funding and/or speaker honoraria from Biogen, Genzyme, and Novartis; is employed by MS Center of Northeastern NY; has consulted for Biogen, Genzyme, EMD Serono, and Novartis; served on the speaker’s bureau for Biogen and Genzyme; and received research support from Actelion, Biogen, Envivo, Eisai Inc, Genentech, Genzyme, Hoffmann-La Roche, Novartis, Pfizer, and Vaccinex. L.J. Henson served on the scientific advisory board for Genzyme; received travel funding and speaker honoraria from Genzyme; was employed by Piedmont, Swedish Health Services; consulted for BioMarin and Mitsubishi Tanabe; served on the speaker’s bureau for Genzyme, Novartis, Teva, Pfizer, and Serono; received research support from Genzyme, Biogen, Novartis and Opexa; her spouse received research support from Parexel; holds stock in Merck and Amgen; and served as an expert witness. P. Repovic consulted for and served on the speaker’s bureau for Biogen, Sanofi Genzyme, Teva, EMD Serono, Percent VX15/2503 saturation of cSEMA4D on peripheral T lymphocytes is shown for anti- and Novartis; also consulted for Anvil Biosciences; served on the speaker’s body-treated participants in each dose cohort. The scheduled sampling times for each dose Bureau for Pfizer and Acorda; and received research support from No- cohort were as described in the legend of figure 2. Data shown represent mean values 6 SE; vartis, NIH, and The Guthy-Jackson Charitable Foundation. S. Lynch flow cytometric assays were performed in triplicate and data analyzed as previously received speaker honoraria from MS-Cure; received research support described.16,20 T-cell saturation values below 20% were considered unsaturated; mean pre- from Novartis, Alexion, Acorda, Actelion, Biogen, Sanofi, Roche, dose saturation values ranged between 0.7% and 18%. Genentech, Chugai, MedImmune, Vaccinex, Sun Pharma, Teva, Genyzme, MedDay, Opexa, NIH, and National Multiple Sclerosis Society. L. Stone is editor-in-chief for International Journal of MS Care suggest that repeated anti-SEMA4D treatment slows and received research support from National Multiple Sclerosis Society. brain volume loss in these mice vs sham-treated ani- D. Mattson served on the scientific advisory board for Acorda and Bio- mals.2 Finally, anti-SEMA4D treatment may reverse gen; received travel funding and/or speaker honoraria from Teva, Biogen, glial cell and astrocyte activation.2,6,13 Based on these EMD Serono, Pfizer, Acorda, Novartis, Genzyme, Bayer, and Questcorp; served on the speaker’s Bureau for Teva, Biogen, EMD Serono, Pfizer, results and those presented here, we initiated a phase Acorda, Novartis, Genzyme, Bayer, and Questcor; received research 2 clinical study evaluating VX15/2503 in patients with support from Teva, Biogen, MedImmune, Roche, ONO, Acorda, prodromal and early-manifest Huntington disease Genentech, Novartis, Genzyme, Sanofi-Aventis, Actelion, Vaccinex, Alkermes, and Indiana Clinical Translational Science Institute/NIH. A. (NCT02481674). Galluzzi has been employed by Premier Research, Chiltern International, and PRA Health Sciences. T.L. Fisher is employed by and holds stock or AUTHOR AFFILIATIONS stock options in Vaccinex, Inc. C. Reilly has been employed by Vaccinex From the North Central Neurology Associates (C.L.), Cullman, AL; Uni- Inc. L. A. Winter has been employed by Vaccinex and Ortho Clinical versity of Rochester Medical Center (L.S.), NY; MS Center of Northeastern Diagnostics. J.E. Leonard holds patents for Therapeutic application of New York (K.E.), Latham; Piedmont Healthcare (L.J.H.), Atlanta, GA; chimeric antibody to human lymphocyte restricted differentiation antigen Swedish Neuroscience Institute–MS Center (P.R.), Seattle, WA; Univer- for treatment of B Cell lymphoma, Therapeutic application of chimeric sity of Kansas Medical Center (S.L.), MO; Cleveland Clinic (L.S.), OH; and radiolabeled antibodies to human B lymphocyte restricted differen- Indiana University Medical Center (D.M.), Indianapolis; PRA Health tiation antigen for treatment of B cell lymphoma, and Chimeric anti- Sciences (A.G.), Lenexa, KS; and Vaccinex, Inc. (T.L.F., C.R., L.A.W., CD20 antibody, expression and use of anti-CD20 antibodies as in vitro J.E.L., M.Z.), Rochester, NY. or in vivo purging agents in patients receiving bone marrow transplantation or peripheral blood stem cell transplant; is Senior Vice President, AUTHOR CONTRIBUTIONS Development; and holds stock/stock options in Vaccinex. M. Zauderer Dr. Leonard wrote the draft manuscript and participated in its revision. holds a patent for Use of anti-semaphorin 4D antibody to treat cancer Dr. Fisher, Dr. Leonard, and Dr. Zauderer made substantive intellectual and Use of anti-semaphorin 4D antibody to treat neuroinflammatory/ contributions to the design and conceptualization of this study; partici- neurodegenerative diseases; has been employed by and holds stock or stock pated in the analysis and interpretation of the data; and revised the man- options in Vaccinex, Inc. Go to Neurology.org/nn for full disclosure forms. uscript. Dr. LaGanke, Dr. Samkoff, Dr. Edwards, Dr. Henson, Dr. Repovic, Dr. Lynch, Dr. Stone, and Dr. Mattson provided study Received November 8, 2016. Accepted in final form April 28, 2017.

8 Neurology: Neuroimmunology & Neuroinflammation REFERENCES semaphorin CD100: a novel mechanism for regulating B 1. Hota P, Buck M. Plexin structures are coming: opportu- cell signaling. Immunity 2000;13:621–631. nities for multilevel investigations of semaphorin guidance 12. Kumanogoh A, Kikutani H. The CD100-CD72 interac- receptors, their cell signaling mechanisms, and functions. tion: a novel mechanism of immune regulation. Trends Cell Mol Life Sci 2012;69:3765–3805. Immunol 2001;22:670–676. 2. Southwell AL, Franciosi S, Villanueva EB, et al. Anti- 13. Smith E, Jonason A, Reilly C, et al. SEMA4D compro- semaphorin 4D immunotherapy ameliorates neuropa- mises blood-brain barrier, activated microglia, and inhibits thology and some cognitive impairment in the remylenation in neurodegenerative disease. Neuro Biol Dis YAC128 mouse model of Huntington disease. Neuro 2015;73:254–268. Biol Dis 2015;76:46–56. 14. Leonard JE, Fisher TE, Winter L, et al. Nonclinical safety 3. Suzuki K, Kumanogoh A, Kikutani H. Semaphorins and evaluation of VX15/2503: a humanized IgG4 anti- their receptors in immune cell interactions. Nat Immunol SEMA4D antibody. Mol Can Therap 2015;14:964–972. 2008;9:17–23. 15. Fisher TL, Reilly CA, Winter LA, et al. Generation and 4. Zhu L, Bergmeier W, Wu J, et al. Regulated surface preclinical characterization of an antibody specific for expression and shedding support a dual role for semaphor- SEMA4D. MAbs 2016;8:150–162. in 4D in platelet responses to vascular injury. Proc Natl 16. Patnaik A, Weiss GJ, Leonard JE, et al. Safety, pharma- Acad Sci USA 2007;104:1621–1626. cokinetics and pharmacodynamics of a humanized anti- 5. Hall KT, Boumsell L, Schultze JL, et al. Human CD100, semaphorin 4D antibody, in a first-in-human study of a novel leukocyte semaphorin that promotes B-cell aggre- patients with advanced solid tumors. Clin Cancer Res gation and differentiation. Proc Natl Acad Sci USA 1996; 2015;22:827–836. 93:11780–11785. 17. Polman CH, Reingold SC, Banwell B, et al. Diagnostic cri- 6. Okuno T, Nakatsuji Y, Moriya M, et al. Roles of teria for multiple sclerosis: 2010 revisions to the McDonald Sema4D-plexin-B1 interactions in the central nervous sys- criteria. Ann Neurol 2011;69:292–302. tem for pathogenesis of experimental autoimmune enceph- 18. Bourdage J, Cook C, Farrington DL, Chain JS, Konrad alomyelitis. J Immunol 2010;184:1499–1506. RJ. An affinity capture elution (ACE) assay for detection 7. Giordano S, Corso S, Conrotto P, et al. The semaphorin of anti-drug antibody to monoclonal antibody therapeutics 4D receptor controls invasive growth by coupling with in the presence of high levels of drug. J Immunol Methods Met. Nat Cell Biol 2002;4:720–724. 2007;327:10–17. 8. Basile JR, Gavard J, Gutkind JS. Plexin-B1 utilizes RHOA 19. Smith BP, Vandenhende FR, DeSante KA, et al. Confi- and ROK to promote the integrin-dependent activation of dence interval criteria for assessment of dose proportion- AKT and ERK, and endothelial cell motility. J Biol Chem ality. Pharm Res 2000;17:1278–1283. 2007;282:34888–34895. 20. Fisher TL, Seils J, Reilly C, et al. Saturation monitoring 9. Giraudon P, Vincent P, Vuaillat C, et al. Semaphorin of VX15/2503, a novel semaphorin 4D-specific antibody, CD100 from activated T lymphocytes induces process in clinical trials. Cytometry B Clin Cytom 2016;90: extension collapse in oligodendrocytes and death of imma- 199–208. ture neural cells. J Immunol 2004;172:1246–1255. 21. Kurtzke JF. Rating neurologic impairment in multiple 10. Witherden DA, Watanabe M, Garijo O, et al. The sclerosis: an expanded disability status scale (EDSS). Neu- CD100 receptor interacts with its plexin B2 ligand to rology 1983;33:1444–1452. regulate epidermal gammadelta T cell function. Immunity 22. Gordon MS, Margolin K, Talpaz M, et al. Phase 1 safety 2012;37:314–325. and pharmacokinetic study of recombinant human anti- 11. Kumanogoh A, Watanabe C, Lee I, et al. Identification of vascular endothelial growth factor in patients with CD72 as a lymphocyte receptor for the class IV advanced cancer. J Clin Oncol 2001;19:843–850.

Neurology: Neuroimmunology & Neuroinflammation 9 Olfactory dysfunction in patients with primary progressive MS

Felix A. Schmidt, MD ABSTRACT Matthew B. Maas, MD Objective: We tested the hypothesis that olfactory function is more impaired in patients with pri- Rohat Geran, MD mary progressive MS (PPMS) than that in relapsing-remitting MS (RRMS). Charlotte Schmidt, MD Methods: Standardized olfactory testing was performed in 32 patients with PPMS, 32 patients Hagen Kunte, MD with RRMS, and 32 healthy controls (HCs). Patients with olfactory dysfunction due to an alterna- Klemens Ruprecht, MD tive primary etiology were excluded. The validated olfactory testing method yielded individual Friedemann Paul, MD scores for olfactory threshold, odor discrimination, and odor identification, along with a composite Önder Göktas, MD Threshold Discrimination Identification (TDI) score. Lutz Harms, MD Results: Olfactory dysfunction was identified in 27 (84%) patients with PPMS, 10 (31%) patients with RRMS, and 1 (3%) HC. While age and sex were similar between PPMS and HCs,

Correspondence to theTDIscoreandallolfactorysubscoresweresignificantlyworseinpatientswithPPMScom- Dr. Schmidt: pared with HCs (all p , 0.001). After adjustment for differences in age, sex, Expanded Disability [email protected] Status Scale (EDSS), and disease duration, odor discrimination, odor identification, and the compositeTDIscorewereworseinpatientswithPPMSvsRRMS(p 5 0.03, 0.04, and 0.02, respectively). Neither age, sex, EDSS, nor disease duration was significantly associated with the composite TDI score. Conclusions: Olfactory dysfunction was more frequent and severe in PPMS compared with RRMS, independent of disease duration and overall disability status. Further research on cellular level differences in olfactory neural pathways may lead to new insights about disease pathogenesis in MS. Neurol Neuroimmunol Neuroinflamm 2017;4:e369; doi: 10.1212/NXI.0000000000000369

GLOSSARY EDSS 5 Expanded Disability Status Scale; HC 5 healthy control; PPMS 5 primary progressive MS; RRMS 5 relapsing- remitting MS; TDI 5 Threshold Discrimination Identification.

MS is a chronic inflammatory demyelinating disease of the CNS, which typically follows a relapsing-remitting disease course (relapsing-remitting MS [RRMS]). However, approximately 15% of all patients with MS initially present with a primary progressive disease course (primary progressive MS [PPMS]) characterized by steadily increasing neurologic disability without recovery.1 Some evidence suggests that—albeit part of the same underlying pathobiology—PPMS presents a less inflammatory course of MS.2 Different studies have reported olfactory dysfunction in patients with MS at rates of 20%– 40% in cohorts comprised mostly of RRMS.3 The underlying pathophysiologic distinctives leading to PPMS vs RRMS are not well understood. Differences in the pattern of disease symptomatology may identify opportunities to better understand disease pathophysiology. Exploring for differences in olfactory dysfunction between PPMS and RRMS has been

From the Clinical and Experimental Multiple Sclerosis Research Center (F.A.S., R.G., H.K., K.R., F.P., L.H.), Department of Neurology, NeuroCure Clinical Research Center (F.A.S., F.P.), Department of Psychiatry (C.S.), Department of Audiology and Phoniatrics (Ö.G.), Charité– Universitätsmedizin Berlin (F.P.), Germany; Department of Neurology (M.B.M.), Feinberg School of Medicine, Northwestern University, Chi- cago, IL; MSB Medical School Berlin (H.K.), Germany; and Experimental and Clinical Research Center (F.P.), Max Delbrück Center for Molecular Medicine, Berlin, Germany. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 identified by other experts as a gap in the lit- the Fisher exact test for categorical variables and the Student t test erature and potential area for future research.3 for continuous variables after confirming normality. There were no differences in age or sex between the PPMS and HC cohorts, As such, we sought to test the hypothesis that although differences were identified between the PPMS and olfactory sense impairment is present in pa- RRMS cohorts. To isolate the effect of PPMS vs RRMS pathol- tients with PPMS, and that olfactory dysfunc- ogy, we created linear regression models for each olfactory mea- tion is more severe in PPMS than in RRMS. sure that included terms for MS subtype (PPMS vs RRMS), age, sex, EDSS, and disease duration. Statistical analyses were performed using SPSS 23.0 (IBM SPSS Statistics, Armonk, NY). METHODS Subjects. Standardized olfactory function testing The study was approved by the medical ethics committee of was performed in patients diagnosed with PPMS and RRMS ac- Charité–Universitätsmedizin Berlin and was conducted in cording to the 2010 McDonald criteria4 and in age-matched accordance with the Declaration of Helsinki and its currently healthy controls (HCs). Predefined exclusion criteria were as applicable version and applicable German laws. All participants follows: age younger than 18 or older than 70 years, pregnancy, provided written informed consent to participate in the study. olfactory disorders with a known different etiology (post- traumatic, postinfectious, sinonasal, infections of the upper RESULTS We studied 96 subjects, 32 with PPMS, respiratory tract, allergies, tumors treated with chemotherapy or 32 with RRMS, and 32 HCs. The demographics radiation, depression, and Parkinson or Alzheimer disease), and and clinical characteristics of study participants are patients taking medications that could cause olfactory dysfunction (e.g., amitriptyline, methotrexate, and doxycycline), shown in table 1. In the PPMS cohort, 26 (81%) including patients who had received corticosteroid treatment were hyposmic and 4 (13%) anosmic, whereas 13 within the last 3 months since corticosteroids can also affect the (41%) patients with RRMS were hyposmic and olfactory function.5 We used the Expanded Disability Status none were anosmic. Olfactory test results of the 3 Scale (EDSS) to measure physical disability. cohorts and univariate differences in those character- Olfactory testing. The Threshold Discrimination Identifica- isticsaresummarizedintable1andpresentedinthe tion (TDI) Test was used for orthonasal olfactory testing.6 The figure. Of 16 possible points, the observed range of olfactory threshold (T) was measured using 48 Sniffin’ Sticks with olfactory function scores was 0–12 for odor thresh- a 16-stage dilution series of n-butanol. The discrimination test old, 5–16 for odor discrimination, and 1–16 for (D) was performed with 48 Sniffin’ Sticks of different smell qualities to test the distinction of smells. Everyday odors were odor identification. identified with the identification test, which consisted of 16 Snif- There were no significant differences in age or sex fin’ Sticks. A TDI value of less than 16 points was defined as between the PPMS and HC groups. Patients with anosmia, up to 31 points as hyposmia, and a value above 31 PPMS were found to have diminished odor thresh- points as normosmia.6 old, odor discrimination, and odor identification, re- Data analysis. Olfactory function measures in patients with sulting in a lower composite olfactory function score PPMS were compared with HCs and patients with RRMS using compared with HCs. Univariate analysis also showed

Table 1 Patient characteristics

p Value (PPMS p Value (PPMS p Value (RRMS Diagnosis PPMS RRMS Controls vs RRMS) vs controls) vs controls)

No. of cases 32 32 32

Age, y 53.4 6 9.3 35.5 6 9.3 51.9 6 17.6 ,0.001 0.68 ,0.001

Sex (female) 13 (40.6) 22 (68.8) 17 (53.1) 0.044 0.45 0.31

Odor threshold 5.3 6 2.6 6.8 6 2.2 7.8 6 1.4 0.016 ,0.001 0.034

Odor discrimination 9.3 6 2.2 12.4 6 2.3 12.3 6 1.4 ,0.001 ,0.001 0.84

Odor identification 10.9 6 3.7 13.0 6 1.5 14.6 6 0.7 0.004 ,0.001 ,0.001

Composite (TDI) 25.4 6 6.1 32.2 6 4.2 34.8 6 2.1 ,0.001 ,0.001

Qualitative olfactory status

Normal 2 (6) 19 (59) 31 (97) ,0.001 ,0.001 ,0.001

Hyposmic 26 (81) 13 (41) 1 (3)

Anosmic 4 (13) 0 (0) 0 (0)

EDSS 4.9 6 2.1 2.6 6 1.8 NA ,0.001 NA NA

Disease duration, y 11.3 6 8.4 5.6 6 5.9 NA 0.002 NA NA

Abbreviations: EDSS 5 Expanded Disability Status Scale; NA 5 not applicable; PPMS 5 primary progressive MS; RRMS 5 relapsing-remitting MS; TDI 5 Threshold Discrimination Identification. All values are presented in mean 6 SD or n (%) unless otherwise specified.

2 Neurology: Neuroimmunology & Neuroinflammation function after adjustment for important covariates, Figure Olfactory function test results: differences of Threshold Discrimination Identification composite score and all individual olfactory subscores whereas a standardized measure of disease disability between 3 cohorts (EDSS) and disease duration did not emerge as significant contributor to olfactory function. Previous studies investigating olfactory function in patients with MS primarily included patients with RRMS. A decreased odor threshold was found in patients with clinically isolated syndrome and early disease stages of RRMS, whereas odor discrimination and odor identification were more often affected in later disease stages.3 A single study has reported olfactory assessments in patients with PPMS by testing odor identification alone and found mild impairment (13% affected).7 We have now demonstrated a significantly worse olfactory function in PPMS as compared to RRMS and further provide a model that

Results are displayed with mean and SDs. Controls 5 healthy controls; PPMS 5 primary adjusts for the significant baseline differences in dis- progressive MS; RRMS 5 relapsing-remitting MS; TDI 5 Threshold Discrimination ease duration and EDSS to show that the differences Identification. are not merely reflective of general disease burden but seem to be particular to disease subtype.7 It is impor- that patients with PPMS had lower odor threshold, tant that we also found a much higher rate of olfac- discrimination, identification, and composite TDI tory impairment than in the prior cited study, score compared with patients with RRMS, although suggesting that our testing method may be more patients with PPMS also had worse EDSS and longer sensitive. disease duration as well as differences in age and sex. The reasons for olfactory dysfunction in patients Multivariate models that adjusted for age, sex, with MS are not well understood.3 A human autopsy EDSS, and disease duration found that PPMS diag- cohort reported demyelination in parts of the olfac- nosis was still associated with worse odor discrimina- tory pathway in 71% of pathologically confirmed MS tion, identification, and composite TDI score cases.8 Decreased olfactory bulb volume is correlated compared with RRMS after adjustment for covari- with increased MS lesion load in the olfactory brain, ates, although no difference in odor threshold. There both of which are linked to impaired olfactory func- were no associations between any other patient char- tion.3,9 Olfactory dysfunction is sometimes recog- acteristics and olfactory function measures except nized in the context of acute relapses as well, that higher EDSS was associated with diminished implicating inflammatory damage.10 Observed differ- odor discrimination (table 2). ences between PPMS and RRMS, independent of other disease severity measures, suggest that PPMS DISCUSSION This study shows that patients with may uniquely affect the olfactory brain tissue more PPMS have significantly decreased olfactory function than RRMS. compared with HCs. Moreover, in comparison with Although the olfactory testing techniques used a cohort of patients with RRMS, PPMS diagnosis here were robust, we did not use MRI to measure was an independent predictor of worse olfactory the lesion load and atrophy of the olfactory pathway

Table 2 Linear regression models for olfactory function measures

Olfactory measure Threshold Discrimination Identification Composite (TDI)

PPMS diagnosis 20.86 (22.71 to 0.99), p 5 0.36 21.87 (23.54 to 20.20), p 5 0.03a 22.32 (24.55 to 20.10), p 5 0.04a 25.00 (29.05 to 20.95), p 5 0.02a (vs RRMS)

Age, y 20.04 (20.11 to 0.04), p 5 0.31 20.04 (20.10 to 0.03), p 5 0.28 0.02 (20.07 to 0.11), p 5 0.61 20.06 (20.22 to 0.11), p 5 0.49

Female sex 20.41 (21.76 to 0.94), p 5 0.54 0.25 (20.97 to 1.47), p 5 0.68 20.06 (21.68 to 1.56), p 5 0.94 20.19 (23.14 to 2.77), p 5 0.90

EDSS 20.19 (20.54 to 0.16), p 5 0.28 20.35 (20.67 to 20.03), p 5 0.03a 20.1 (20.52 to 0.33), p 5 0.66 20.64 (21.42 to 0.14), p 5 0.10

Disease 0.07 (20.03 to 0.17), p 5 0.16 0.04 (20.04 to 0.13), p 5 0.32 0.00 (20.12 to 0.11), p 5 0.94 0.11 (20.10 to 0.33), p 5 0.30 duration, y

Abbreviations: EDSS 5 Expanded Disability Status Scale; PPMS 5 primary progressive MS; RRMS 5 relapsing-remitting MS; TDI 5 Threshold Discrim- ination Identification. a Significant.

Neurology: Neuroimmunology & Neuroinflammation 3 and compare it with other regions of the brain, which travel funding and/or speaker honoraria from Bayer, Novartis, Biogen may have helped clarify the etiology of the differences Idec, Teva, Sanofi-Aventis/Genzyme, Merck Serono, Alexion, Chugai, MedImmune, and Shire; is an academic editor for PLoS One;associate we observed between the PPMS and RRMS groups. editor for Neurology® Neuroimmunology & Neuroinflammation;con- At this time, techniques to accurately measure lesion sulted for Sanofi Genzyme, Biogen Idec, MedImmune, Shire, and Alex- burden in the small structures comprising the olfac- ion; and received research support from Bayer, Novartis, Biogen Idec, Teva, Sanofi-Aventis/Genzyme, Alexion, Merck Serono, German tory system are not well established or anatomically- Research Council, Werth Stiftung of the City of Cologne, German pathologically validated. Finally, although we con- Ministry of Education and Research, Arthur Arnstein Stiftung Berlin, structed a multivariate model to control for important Arthur Arnstein Foundation Berlin, Guthy Jackson Charitable Founda- disease covariates such as age and overall disability, tion, and National Multiple Sclerosis Society of the USA. Ö. Göktas reports no disclosures. L. Harms served on the scientific advisory board the size of our cohort is relatively small, and we can- for Novartis, Sanofi/Genzyme, Roche, and Biogen and received travel not exclude the possibility that the relationship funding and/or speaker honoraria from Biogen, Merck Serono, Gen- between the MS subgroup and olfactory dysfunction zyme, Teva, Bayer Health Care, Novartis, and Grifols. Go to is mediated or confounded by a variable we did not Neurology.org/nn for full disclosure forms.

measure, or incompletely adjusted for age, sex, EDSS, Received January 23, 2017. Accepted in final form April 23, 2017. and disease duration. We conclude that olfactory dysfunction is a frequent REFERENCES symptom in patients with PPMS, leading to severe 1. Lublin FD, Reingold SC, Cohen JA, et al. Defining the olfactory impairment, and is uniquely more severe in clinical course of multiple sclerosis: the 2013 revisions. – PPMS compared with RRMS. The findings suggest Neurology 2014;83:278 286. 2. Lassmann H, van Horssen J, Mahad D. Progressive mul- that olfactory dysfunction might be a surrogate of neu- tiple sclerosis: pathology and pathogenesis. Nat Rev Neu- rodegeneration in these patients. Studies correlating rol 2012;8:647–656. olfactory function with radiologic and clinical markers 3. Lucassen EB, Turel A, Knehans A, Huang X, Eslinger P. of disease progression would be of interest. Olfactory dysfunction in multiple sclerosis: a scoping review of the literature. Mult Scler Relat Disord 2016;6: AUTHOR CONTRIBUTIONS 1–9. Felix Schmidt: study concept and design, acquisition of data, draft of the 4. Polman CH, Reingold SC, Banwell B, et al. Diagnostic manuscript and figures, and study supervision. Matthew Maas: analysis criteria for multiple sclerosis: 2010 revisions to the McDo- and interpretation of data and critical revision of manuscript for intellec- nald criteria. Ann Neurol 2011;69:292–302. tual content. Rohat Geran and Charlotte Schmidt: acquisition of data 5. Heilmann S, Just T, Göktas O, Hauswald B, Hüttenbrink and critical revision of manuscript for intellectual content. Hagen Kunte, K-B, Hummel T. Effects of systemic or topical adminis- Klemens Ruprecht, Friedemann Paul, and Önder Göktas: critical revision tration of corticosteroids and vitamin B in patients with of manuscript for intellectual content. Lutz Harms: study concept and olfactory loss. Laryngorhinootologie 2004;83:729–734. design and critical revision of manuscript for intellectual content. 6. Hummel T, Kobal G, Gudziol H, Mackay-Sim A. Nor- “ ’ ” STUDY FUNDING mative data for the Sniffin Sticks including tests of odor No targeted funding reported. identification, odor discrimination, and olfactory thresholds: an upgrade based on a group of more than 3,000 DISCLOSURE subjects. Eur Arch Otorhinolaryngol 2007;264:237–243. F. Schmid received speaker honoraria from Genzyme. M. Maas served 7. Silva AM, Santos E, Moreira I, et al. Olfactory dysfunction on the scientific advisory board for Hyperfine Research and received in multiple sclerosis: association with secondary progres- research support from the NIH. R. Geran and C. Schmidt report no dis- sion. Mult Scler 2012;18:616–621. closures. H. Kunte served on the scientific advisory board for Roche, 8. DeLuca GC, Joseph A, George J, et al. Olfactory pathol- Novartis, and Genzyme and received travel funding and/or speaker hon- ogy in central nervous system demyelinating diseases. oraria from Bayer, Biogen, Genzyme, Merck, Mylan, Novartis, Roche, Brain Pathol 2015;25:543–551. and Teva. K. Ruprecht served on the scientific advisory board for 9. Goektas O, Schmidt F, Bohner G, et al. Olfactory bulb Sanofi-Aventis/Genzyme, Novartis, and Roche; received travel funding volume and olfactory function in patients with multiple and/or speaker honoraria from Bayer, Biogen, Merck Serono, Sanofi- sclerosis. Rhinology 2011;49:221–226. Aventis/Genzyme, Teva, Novartis, and Guthy Jackson Charitable Foun- dation; is an academic editor for PLoS One; received publishing royalties 10. Doty RL, Li C, Mannon LJ, Yousem DM. Olfactory from Elsevier; and received research support from Novartis, Merck Se- dysfunction in multiple sclerosis: relation to longitudinal rono, and German Ministry of Education and Research. F. Paul served changes in plaque numbers in central olfactory structures. on the scientific advisory board for Novartis and MedImmune; received Neurology 1999;53:880–882.

4 Neurology: Neuroimmunology & Neuroinflammation Autoimmune episodic ataxia in patients with anti-CASPR2 antibody-associated encephalitis

Bastien Joubert, MD ABSTRACT Florent Gobert, MD Objective: To report paroxysmal episodes of cerebellar ataxia in a patient with anti–contactin- Laure Thomas, MD associated protein-like 2 (CASPR2) antibody-related autoimmune encephalitis and to search for Margaux Saint-Martin, similar paroxysmal ataxia in a cohort of patients with anti–CASPR2 antibody-associated autoim- MSc mune encephalitis. Virginie Desestret, MD, Methods: We report a patient with paroxysmal episodes of cerebellar ataxia observed during auto- PhD immune encephalitis with anti-CASPR2 antibodies. In addition, clinical analysis was performed in Philippe Convers, MD a retrospective cohort of 37 patients with anti-CASPR2 antibodies to search for transient epi- Véronique Rogemond, sodes of ataxia. Paroxysmal symptoms were further specified from the referral physicians, the pa- PhD tients, or their relatives. Géraldine Picard, MSc François Ducray, MD, Results: A 61-year-old man with limbic encephalitis and anti-CASPR2 antibodies developed ste- PhD reotyped paroxysmal episodes of cerebellar ataxia, including gait imbalance, dysarthria, and dys- Dimitri Psimaras, MD metria, 1 month after the onset of the encephalitis. The ataxic episodes were specifically Jean-Christophe Antoine, triggered by orthostatism and emotions. Both limbic symptoms and transient ataxic episodes MD, PhD resolved after treatment with steroids and IV cyclophosphamide. Among 37 other patients with Jean-Yves Delattre, MD, anti-CASPR2 antibodies, we identified 5 additional cases with similar paroxysmal ataxic episodes PhD that included gait imbalance (5 cases), slurred speech (3 cases), limb dysmetria (3 cases), and nys- Jérôme Honnorat, MD, tagmus (1 case). All had concomitant limbic encephalitis. Paroxysmal ataxia was not observed in PhD patients with neuromyotonia or Morvan syndrome. Triggering factors (orthostatism or anger) were reported in 4 patients. Episodes resolved with immunomodulatory treatments in 4 patients and spontaneously in 1 case. Correspondence to Conclusions: Paroxysmal cerebellar ataxia must be added to the spectrum of the anti-CASPR2 Dr. Honnorat: antibody syndrome. Neurol Neuroimmunol Neuroinflamm [email protected] 2017;4:e371; doi: 10.1212/ NXI.0000000000000371

GLOSSARY EA 5 episodic ataxia.

Episodic ataxias (EAs) are a group of hereditary channelopathies whose common feature is the occurrence of paroxysmal episodes of cerebellar ataxia.1 Ataxic episodes usually last a few minutes to a few days and can be triggered by emotions, abrupt movements, exercise, or fever. Depend- ing on which ion channel gene is mutated, additional symptoms, such as neuromyotonia or epilepsy, can occur.2 Conversely, paroxysmal symptoms are rare in patients with antineuronal antibody-associated neurologic disorders, and EAs have not yet been reported in such cases. In this study, we report a patient with paroxysmal episodes of ataxia developed during autoimmune encephalitis with anti-CASPR2 antibodies. To assess the relevance of our case, we retrospectively Supplemental data at Neurology.org/nn From the Centre National de Référence pour les Syndromes Neurologiques Paranéoplasiques (B.J., L.T., V.D., V.R., G.P., F.D., D.P., J.-C.A., J.-Y.D., J.H.), Service de Neuro-Réanimation (F.G.), Hôpital Neurologique, Hospices Civils de Lyon, Bron; Institut NeuroMyoGene INSERM U1217/CNRS UMR 5310 (B.J., F.G., L.T., M.S.-M., V.D., V.R., G.P., F.D., J.H.), University of Lyon–Université Claude Bernard Lyon 1; Service de Neurologie (P.C., J.-C.A.), Hôpital Bellevue, Centre Hospitalier Universitaire de Saint-Étienne; and Département de Neurologie (D.P., J.-Y.D.), Groupe Hospitalier Pitié-Salpêtrière, Assistance Publique–Hôpitaux de Paris, France. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/nn for full disclosure forms. The Article Processing Charge was funded by INSERM ADR05. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology 1 searched for similar episodes of transient ataxia orthostatism, and lasted less than 1 minute. No cer- in a cohort of patients with anti-CASPR2 anti- ebellar symptom was observed between the episodes. bodies. Anti–CASPR2 antibody-related disor- Levetiracetam was switched to lacosamide without ders encompass a wide range of neurologic any improvement of the symptoms. autoimmune syndromes, including autoim- Brain MRI was normal. EEG found an asymmetric temporal lobe slowing. CSF analysis found 16 mune encephalitis, neuromyotonia, and Mor- white blood cells/mm3, a protein level of 65 mg/dL, van syndrome. A recent publication by Van and no oligoclonal band. Anti-CASPR2 antibodies 3 Sonderen et al. showed that up to 77% of were detected (end-point dilution using cell- the patients with such antibodies had at least binding–based assay, 1/40,960, in serum and 3 cumulated core neurologic symptoms, 1/5,120 in CSF). Brain single-photon emission CT including encephalic signs, cerebellar symp- found bilateral frontotemporal hypoperfusion with toms, peripheral nerve hyperexcitability, dys- normal perfusion of basal ganglia and cerebellum. autonomia, neuropathic pain, insomnia, and Full-body CT was normal. Sequencing of the weight loss. Our study may contribute to fur- genes involved in EA types 1 and 2, KCNA1 and ther delineate anti–CASPR2 antibody-related CACNA1A, found that the patient carried a rare polymorphism in intron 39 of CACNA1A (c5843- clinical presentation. 14G.A).5,6 KCNA1 was normal. The patient was METHODS We report a patient with autoimmune encephali- treated with 3 daily injections of 1 g methylprednis- tis, anti-CASPR2 antibodies, and paroxysmal cerebellar ataxia. olonefollowedby6monthlypulsesof1gcyclo- Anti-CASPR2 antibodies were screened in serum and CSF as phosphamide.Thefrequencyoftheataxiaepisodes previously described.4 Positivity of both an immunohisto- decreased immediately; they had totally stopped 5 fluorescent assay on rat brain slices and a cell-based binding assay days after the end of the methylprednisolone pulses. – with HEK-293 transfected cells were needed to confirm the The patient recovered all his cognitive abilities 2 presence of anti-CASPR2 antibodies. A signed patient consent- to-disclose form has been obtained from the patient for a video weeks after the initiation of the treatment and had recording of one of the episodes. his last seizure 3 months after the first visit. At We also studied the clinical files of 37 patients with anti- month 7 of follow-up, the patient remained asymp- CASPR2 antibodies, detected in their CSF or sera at the Centre tomatic and seizure free. At the peak of his disease, de Référence National pour les Syndromes Neurologiques Paranéo- the patient had presented with only 2 core symp- plasiques (Lyon, France) between March 2009 and August 2016, toms as defined by Van Sonderen et al.3 (encephalic to search for similar transient cerebellar symptoms. Thirty-three signs and cerebellar ataxia). of those patients have been previously reported.4 Written informed consent was obtained from all patients with approval of the Institutional Review Board of the Hospices Civils de Lyon. Cohort study. Among the cohort of 37 other patients We selected all patients who had been reported by their referral with anti-CASPR2 antibodies (20 with encephalitis physicians to have symptoms that were both transient and sug- and 17 with neuromyotonia or Morvan syndrome), gestive of cerebellar impairment, i.e., gait imbalance, slurred we retrospectively identified 5 patients with transient speech, or limb dysmetria. Further information was collected symptoms suggestive of cerebellar impairment. All by telephone from the referral physicians, and, when possible, from the patients themselves or their relatives. the patients were men, with a median age of 60 years (range 57–69). Median follow-up (range) was 45.6 RESULTS Index case. A 61-year-old man was hospi- months (19.3–113.6). One patient had a history of talized for evaluation after a tonic-clonic generalized surgically treated thyroid carcinoma. All these 5 pa- seizure. He was still active as a corporate executive, tients had anti-CASPR2 antibodies in both serum and his medical history included high blood pressure, and CSF and had a presentation of autoimmune diabetes, myocardial infarction, and a smoking habit. encephalitis with prominent seizures and amnesia, as No prodromal or postictal symptom was reported, we previously reported.4 Three patients also devel- but the patient reported slight memory impairment, oped mild permanent cerebellar symptoms, con- difficulties to concentrate at work, unusual emotive- comitantly to (2 cases) or after (1 case) the onset of ness, and anxiety over a few days before the seizure. paroxysmal ataxia. The patient was treated with levetiracetam and clo- None of them exhibited neuromyotonia or bazam. However, the cognitive symptoms persisted Morvan syndrome. Two of the 5 patients had $3 and several partial temporal lobe seizures occurred. core symptoms as defined in the study by Van One month after the first seizure, the patient began to Sonderen et al.3 The transient episodes are experience repeated episodes of slurred speech, gait described in the table. They were stereotyped and ataxia, and slight dysmetria of the limbs (see video included, according to the cases, gait imbalance at Neurology.org/nn). These events occurred 3–4 (5/5, 100%), slurred speech (3/5, 60%), limb dys- times a day, were often triggered by emotions and metria (3/5, 60%), and nystagmus (1/5, 20%).

2 Neurology: Neuroimmunology & Neuroinflammation Two patients (40%) reported concomitant sensa- tions of neck stiffness. Myoclonic jerks were not observed in any of the patients. Of interest, the ataxia episodes preceded by 7 months the full development of the encephalitis in 1 patient. The appa- immunoglobulin cyclophosphamide plasmapheresis Treatments before resolution of the episodes rition of the ataxia was not subsequent to a change of antiepileptic medication in any of the patients. The attacks were reported to last from a few minutes to 2 hours, according to the cases, with triggering factors identified in 4 patients (80%; orthostatism in 4 patients and anger in 1 patient). Ataxia episodes occurred during a mean period of 4 months Anti-CASPR2 ab titers Serum CSF (range, 0.5–32.5) and stopped after the initiation of an immunomodulatory treatment in 4 patients (steroids alone, steroids and plasmapheresis, steroids and IV immunoglobulin, or rituximab) and spontaneously in 1 case. Interictal, milder cerebel-

Period during which paroxysmal ataxia occurred lar symptoms were observed in 3/5 cases (60%). Interictal brain MRI was unremarkable in all patients; magnetic resonance spectroscopy was not performed during the episodes. We did not perform genetic studies in these 5 patients. 17 mo later 2 mo 1/10,240 1/10,240 Prednisolone and IV Concomitant 13 mo 1/10,240 1/10,240 Rituximab 7 mo before 33 mo 1/10,240 1/10,240 Methylprednisolone Onset in relation to the onset of the limbic symptoms

DISCUSSION In this study, we report stereotyped episodes of paroxysmal cerebellar ataxia in a patient with anti–CASPR2 antibody-related autoimmune imbalance) imbalance) No 10 mo later 2 wk NA 1/320 None imbalance and inferior limb dysmetria) Permanent cerebellar symptoms also present encephalitis. We also retrospectively identified similar paroxysmal symptoms in 5 of 20 other patients with autoimmune encephalitis and anti-CASPR2 antibodies, suggesting that episodic cerebellar ataxia 7/d Yes (slight 4/d No 1 mo later 3 wk 1/40,960 1/1,520 Methylprednisolone, 5/wk No 4 mo later 4 mo 1/10,240 1/640 Methylprednisolone, 3/d Yes (gait 5/wk Yes (slight – – – – , an interval of 2 wk Frequency at peak 3 could be observed in up to 25% of such patients. As they developed during the course of encephalitis and resolved after immunomodulating therapy in most of the patients, we can speculate an immune origin of these paroxysmal ataxias. More- 1min 1 min 3 over, the association of gait imbalance, slurred , Duration of the episodes 2h , speech, and limb dysmetria during these episodes was highly suggestive of cerebellar impairment; although because of the retrospective collection Yes A few minutes 6 Impact on daily activities No of the data, we cannot completely exclude that the alleged cerebellar symptoms were actually of another nature (e.g., orthostatic hypotension, seizures, or peri-ictal autonomic manifestations). Orthostatism No Orthostatism, anger Triggering factors None No A few minutes 2 episodes with Orthostatism Yes A few minutes 2 Anxiety and orthostatism Paroxysmal neurologic symptoms have already been reported in autoimmune encephalitis. For

not available. instance, faciobrachial dystonic seizures are caused 5 by the activation of the motor cortex in patients with anti–leucine-rich glioma1 (Lgi1) antibody-associated encephalitis.7 Of interest, Lgi1 is a secreted protein dysmetria speech, limb dysmetria, nystagmus, and neck stiffness Symptoms during the episodes slurred speech speech, limb dysmetria, and neck stiffness speech, and limb dysmetria that complexes with CASPR2 and the potassium antibodies; NA channel K 1.1.5,8 More recently, orthostatic myoclo- 5 V nus has been observed in a patient with anti-CASPR2 Characteristics of the ataxia episodes in patients with autoimmune encephalitis and anti-CASPR2 antibodies 19 Gait imbalance and limb 22 Gait imbalance, slurred Follow- up, mo 114 Gait imbalance and 51 Gait imbalance Orthostatism No A few minutes to 46 Gait imbalance, slurred 7 Gait imbalance, slurred antibodies.9 Although permanent ataxia without remissions is a well-described feature of anti–CASPR2 antibody-associated encephalitis, to our knowledge, 69, M 60, M 62, M 57, M Age, y, sex 60, M 61, M Table

Abbreviations: ab The first line refers to the index case. EA has never been reported before in patients with

Neurology: Neuroimmunology & Neuroinflammation 3 autoimmune encephalitis, including for instance pa- to counterbalance the effects of anti-CASPR2 anti- tients with anti-Lgi1 encephalitis.4,3,10,11 bodies. Therefore, this polymorphism may have Anti-CASPR2 antibodies associate with various favored the development of episodic cerebellar symp- neurologic disorders, including limbic encephalitis, toms in the patient. neuromyotonia, and Morvan syndrome.3,4,12–14 In Overall, our findings suggest that transient cere- a recent published cohort of 38 anti–CASPR2 bellar ataxia should be added to the spectrum of antibody-positive patients, 77% of the cases had anti–CASPR2 antibody-related symptoms. Such par- $3 core symptoms (encephalic signs, cerebellar oxysmal symptoms are similar to the symptoms of symptoms, peripheral nerve hyperexcitability, dysau- hereditary channelopathies, suggesting that ion chan- tonomia, neuropathic pain, insomnia, and weight nel dysfunction is involved in the pathogenesis of loss).3 In our previously published cohort of 33 pa- anti-CASPR2 antibody syndromes. tients with anti-CASPR2 antibodies, only 16/33 (48%) of the patients presented with $3 of those AUTHOR CONTRIBUTIONS core symptoms.4 However, patients with limbic Dr. Joubert: study concept and design; acquisition of data; and analysis encephalitis were overrepresented and we found and interpretation. Dr. Gobert: acquisition of data and critical revision $ of the manuscript for important intellectual content. Dr. Thomas and a greater number of patients with 3 core symptoms Ms. Saint-Martin: acquisition of data. Dr. Desestret: critical revision of in the group of patients diagnosed with neuromyoto- the manuscript for important intellectual content. Dr. Convers, nia or Morvan syndrome (13/15, 87%) than in the Dr. Rogemond, and Ms. Picard: acquisition of data. Prof. Ducray, Dr. Psimaras, Prof. Antoine, and Prof. Delattre: critical revision of the group of patients diagnosed with limbic encephalitis manuscript for important intellectual content. Dr. Honnorat: study (3/18, 17%). Only 2/6 (33%) of our patients with supervision; acquisition of data; study concept and design; and critical paroxysmal ataxia had $3 core symptoms, which revision of the manuscript for important intellectual content. reflects that this paroxysmal syndrome is mostly asso- ACKNOWLEDGMENT ciated with limbic encephalitis rather than neuromyo- The authors thank Dr. Cécile Marchal, Dr. Philippe Kassiotis, Dr. Jacques tonia or Morvan syndrome. Testaud, Dr. Lejla Koric, and Dr. Isabelle Lambert, who provided CSF, The resemblance of the patients’ paroxysmal ataxia serum samples, and clinical data for the study. with manifestations observed in hereditary channelopathies such as EA is striking. Of interest, neuromyo- STUDY FUNDING tonia can be observed in patients with EA type 1 and This study is supported by institutional grants from Agence Nationale de la Recherche (ANR-14-CE15-0001-MECANO), Fondation pour la Re- anti-CASPR2 antibodies, suggesting similar ion chan- cherche Médicale (Neurodon2014), and CSL Behring France. nel dysfunctions in both diseases. EA type 1 is due to mutations of KCNA1, a gene coding for the voltage- DISCLOSURE 5,13 B. Joubert, F. Gobert, L. Thomas, M. Saint-Martin, V. Desestret, gated potassium channel KV1.1. The clustering of P. Convers, V. Rogemond, G. Picard, F. Ducray, and D. Psimaras report KV1.1 at the nodes of Ranvier depends on CASPR2, no disclosures. J.-C. Antoine served on the scientific advisory board for and electrophysiologic experiments have suggested an Pfizer; received travel funding and/or speaker honoraria from Pfizer and impairment of voltage-gated potassium channels in Laboratoire Français des Biotechnologies et du Fractionnement; served autoimmune neuromyotonia, implying that anti- as an associate editor for Revue Neurologique; and holds a patent for diagnostic test for anti-CV2/CRMP5 andibody detection. J.-Y. Delattre CASPR2 antibodies might indirectly alter the func- served on the editorial board for The Oncologist; received research support 15,16 tions of KV1.1 at the nodes of Ranvier. However, from Institut National du Cancer; reference center for anaplastic oligoden- previous studies have failed to demonstrate a role of drogliomas. J. Honnorat reports no disclosures. Go to Neurology.org/nn for full disclosure forms. CASPR2 in KV1.1 clustering at the synaptic level, and the exact role of anti-CASPR2 antibodies in the Received March 3, 2017. Accepted in final form April 24, 2017. CNS is unknown.16,17 Nevertheless, our observation supports the hypothesis of immune-mediated ion REFERENCES channel dysfunction in anti-CASPR2 antibody syn- 1. Tomlinson SE, Hanna MG, Kullmann DM, Tan SV, dromes. The polymorphism of CACNA1A in the Burke D. Clinical neurophysiology of the episodic ataxias: index patient is interesting because this gene codes insights into ion channel dysfunction in vivo. Clin Neuro- physiol 2009;120:1768–1776. for the Ca 2.1 subunit of the P/Q type voltage- V 2. D’Adamo MC, Hasan S, Guglielmi L, et al. New insights gated calcium channel that is mutated in patients into the pathogenesis and therapeutics of episodic ataxia 6 with EA type 2. We have no data about the signif- type 1. Front Cell Neurosci 2015;9:317. icance of this polymorphism, whose overall frequency 3. Van Sonderen A, Ariño H, Petit-Pedrol M, et al. The in the general population is estimated at less than 1/ clinical spectrum of Caspr2 antibody-associated disease. – 1,000. However, we can hypothesize that it may pro- Neurology 2016;87:521 528. 4. Joubert B, Saint-Martin M, Noraz N, et al. Character- voke a partial impairment of Ca 2.1.6 We can thus V ization of a subtype of autoimmune encephalitis with speculate that in the case, this polymorphism of anti-contactin-associated protein-like 2 antibodies in CACNA1A may have led to the failure of compensa- the cerebrospinal fluid, prominent limbic symptoms, – tory mechanisms dependent on CaV2.1 and necessary and seizures. JAMA Neurol 2016;73:1115 1124.

4 Neurology: Neuroimmunology & Neuroinflammation 5. Browne DL, Gancher ST, Nutt JG, et al. Episodic 11. Balint B, Regula JU, Jarius S, Wildemann B. Caspr2 anti- ataxia/myokymia syndrome is associated with point muta- bodies in limbic encephalitis with cerebellar ataxia, dyski- tions in the human potassium channel gene, KCNA1. Nat nesias and myoclonus. J Neurol Sci 2013;327:73–74. Genet 1994;8:136–140. 12. Irani SR, Pettingill P, Kleopa KA, et al. Morvan syndrome: 6. Terwindt GM, Ophoff RA, Haan J, et al. Variable clin- clinical and serological observations in 29 cases. Ann Neu- ical expression of mutations in the P/Q-type calcium rol 2012;72:241–255. channel gene in familial hemiplegic migraine. Dutch 13. Lancaster E, Huijbers MG, Bar V, et al. Investigations of Migraine Genetics Research Group. Neurology 1998; Caspr2, an autoantigen of encephalitis and neuromyoto- 50:1105–1110. nia. Ann Neurol 2011;69:303–311. 7. Navarro V, Kas A, Apartis E, et al. Motor cortex and 14. Bien CG, Mirzadjanova Z, Baumgartner C, et al. Anti- hippocampus are the two main cortical targets in contactin-associated protein-2 encephalitis: relevance of LGI1-antibody encephalitis. Brain J Neurol 2016; antibody titres, presentation and outcome. Eur J Neurol 139:1079–1093. 2017;24:175–186. 8. Lai M, Huijbers MGM, Lancaster E, et al. Investigation of 15. Horresh I, Poliak S, Grant S, Bredt D, Rasband MN, Peles LGI1 as the antigen in limbic encephalitis previously E. Multiple molecular interactions determine the cluster- attributed to potassium channels: a case series. Lancet ing of Caspr2 and Kv1 channels in myelinated axons. Neurol 2010;9:776–785. J Neurosci 2008;28:14213–14222. 9. Gövert F, Witt K, Erro R, et al. Orthostatic myoclonus 16. Sinha S, Newsom-Davis J, Mills K, Byrne N, Lang B, associated with Caspr2 antibodies. Neurology 2016;86: Vincent A. Autoimmune aetiology for acquired neuro- 1353–1355. myotonia (Isaacs’ syndrome). Lancet 1991;338:75–77. 10. Becker EBE, Zuliani L, Pettingill R, et al. Contactin- 17. Ogawa Y, Horresh I, Trimmer JS, Bredt DS, Peles E, associated protein-2 antibodies in non-paraneoplastic Rasband MN. Postsynaptic density-93 clusters Kv1 chan- cerebellar ataxia. J Neurol Neurosurg Psychiatry nels at axon initial segments independently of Caspr2. 2012;83:437–440. J Neurosci 2008;28:5731–5739.

Neurology: Neuroimmunology & Neuroinflammation 5 Clinical/Scientific Notes

Yuichi Hayashi, MD, IVIG TREATMENT FOR REPEATED experienced consciousness disturbance after breakfast. PhD HYPOTHERMIC ATTACKS ASSOCIATED WITH His axillary body temperature could not be obtained, Megumi Yamada, MD, LGI1 ANTIBODY ENCEPHALITIS and he was transferred to our hospital. The staff at the PhD psychiatric hospital denied that he was exposed to Akio Kimura, MD, PhD Hypothermia, defined as a core body temperature of a cold environment or overdosage of drugs before or Takashi Inuzuka, MD, ,35.0°C (,95.0°F), is a life-threatening and emer- after the hypothermic attack. PhD gency situation. Various conditions and diseases can After arriving at our hospital, his core body temper- cause hypothermia, including environmental expo- ature was 33.3°C without piloerection or shivering. Neurol Neuroimmunol His blood pressure was 99/54 mm Hg, and heart rate Neuroinflamm sure, drug intoxication, CNS diseases, and metabolic 2017;4:e348; doi: 10.1212/ abnormalities, such as hypopituitarism, hypoadrenal- was 55 beats per minute. Neurologic examination re- NXI.0000000000000348 ism, hypothyroidism, and hypoglycemia. vealed mild disturbance of consciousness. Results of Leucine-rich glioma-inactivated 1 (LGI1) antibody– routine laboratory tests were normal, except for mild associated encephalitis is an autoimmune-mediated hyponatremia (serum sodium of 133 mEq/L). Endo- encephalitis in which autoantibodies are directed crine tests showed a free T4 level of 0.97 ng/dL, m against a voltage-gated potassium channel (VGKC) thyroid-stimulating hormone level of 0.42 IU/mL, complex protein named LGI1.1 The LGI1 protein is and adrenocorticotropic hormone level of 10.3 pg/mL. widely expressed in the neurons of the CNS and also in Brain MRI and whole-body CT were unremarkable. the hypothalamus.2 Central hypothermia associated Therefore, sepsis, endocrine disease, and exposure to with autoimmune-mediated encephalitis is rare; how- cold environment were excluded. He was diagnosed – ever, it is a potentially life-threatening symptom of with moderate hypothermia associated with anti encephalitis. Here, we describe a patient who experi- LGI1 encephalitis based on the detection of LGI1 anti- enced repeated hypothermic attacks associated with bodies in his serum and CSF, and active rewarming anti–LGI1 encephalitis. was performed until his body temperature returned to normal. His body temperature recovered within 24 Case report. A 56-year-old Japanese man was hours after hospitalization, and his level of conscious- admitted to our hospital with hypothermia and ness ameliorated with the increase of body tempera- consciousness disturbance. He had been diagnosed ture. Hypothermic attacks with mild disturbance of with VGKC antibody encephalitis 1 year earlier based consciousness occurred repeatedly 7 times, and he on the detection of VGKC-complex antibodies and recovered by rewarming in each occasion; however, neurologic symptoms. He was initially admitted to rewarming was not effective for preventing recurrences. our hospital because of seizures and acute con- We administered methylprednisolone pulse therapy sciousness disturbances accompanied with a 6-month which showed no effects. The mean recurrence interval – history of psychiatric symptoms, such as violent of hypothermic attacks was 15.8 (5 37) days. After behavior toward his family, and audiovisual halluci- administration of the first course of IV immunoglob- nations since the summer of 2007. Brain MRI re- ulin (IVIG), the recurrence interval increased to 79 vealed a slightly hyperintense area in the right medial days; and no hypothermic attacks occurred for 5 years temporal region at the initial admission to our hos- after the second course of IVIG treatment. pital. The findings of whole-body CT and routine Discussion. The center of thermoregulation is in the CSF analysis were normal. Corticosteroid treatment hypothalamus. The effector mechanisms for cold enhanced his psychiatric symptoms. He was trans- defense include cutaneous vasoconstriction, piloerec- ferred to a psychiatric hospital because of the con- tion, and heat production by means of shivering. tinuing psychiatric symptoms which were mainly Therefore, the absence of piloerection and shivering suppressed with antipsychotic drugs. Thereafter, he during hypothermia in our case suggested a dysfunc- had stayed in the psychiatric hospital in an air- tion of the hypothalamus thermoregulatory center. conditioned room, and the temperature of which To better understand this clinical problem, we re- was maintained above 20°C. In February 2009, he viewed the reports of 6 patients with VGKC-complex

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology 1 Table Central hypothermia associated with VGKC or LGI1 antibody encephalitis including the present case

Disease duration Core body Severity of Age at between initial symptoms temperature at Realization hypothermia Case hypothermia Underlying of the underlying hypothermia of following AHA no. Study onset, y/sex disease/tumor Underlying drugs disease and hypothermia identification, °C hypothermia 2010

1 Jacob et al., 46/M VGKC-related NA 3–4 mo 33 Hospital Moderate 20083 encephalitis/ thymoma

2 Jacob et al., 72/M VGKC-related NA 5 mo 34.5 NA Mild 20083 encephalitis/ none

3 Jacob et al., 62/F VGKC-related NA 9 y 34.8 NA Mild 20083 encephalitis/ none

4 Jacob et al., 57/M VGKC-related Phenobarbital, phenytoin 1 y 33.4 Hospital Moderate 20083 encephalitis/ none

5 Montiel, et al., 65/F VGKC-related NA 2.5 y NA NA NA 20084 encephalitis/ none

6 Present case 56/M LGI1 antibody Phenytoin, clonazepam, 1.5 y 33.3 Hospital Moderate encephalitis/ haloperidol, quetiapine, none brotizolam, and valsartan

Disturbance of Absence of pilomotor Case consciousness Cardiac erection, and/or How many episodes of Abnormalities of the Immunotherapy after Long-term no. Season at admission abnormalities shivering hypothermic attack hypothalamus on MRI rewarming outcome

1 NA Yes NA NA Twice None First episode: IVIG 1 AZP, Recovered second episode: IVIG 1 CS

2 NA NA NA NA Prolonged 6 mo None PE 1 IVIG Recovered

3 NA None None NA Single Bilateral CS Recovered hyperintensities

4 Autumn None Bradycardia Absent pilomotor Single Bilateral CS Recovered erection and shivering hyperintensities

5 Winter NA NA NA Single None IVIG 1 CS Recovered

6 Winter Mild None Absent pilomotor 7 times None CS was ineffective, IVIG was Recovered erection and shivering effective

Abbreviations: AHA 5 American Heart Association; AZP 5 azathioprine; CS 5 corticosteroids; IVIG 5 IV immunoglobulin; LGI1 5 leucine-rich glioma- inactivated 1; NA 5 not available or not described; PE 5 plasma exchange; VGKC 5 voltage-gated potassium channel; VT 5 ventricular tachycardia.

or LGI1 antibody–associated encephalitis who speculated that the hypothermic attacks were presented with central hypothermia during the related to anti–LGI1 encephalitis. Although disease course, including the current patient repeated hypothermic attacks are rare, it is impor- (table).3,4 Almost all cases, except for a patient with tant that clinicians are aware of these attacks that continued hypothermia, developed single or dual can occur during the course of anti–LGI1 enceph- attacks. Three or more repeated hypothermic alitis, as these attacks can be fatal if undetected or attacks had not been previously reported in pa- overlooked. – tients with anti LGI1 encephalitis. The patients From the Department of Neurology and Geriatrics, Gifu University reported by Jacob et al.3 likely had anti–LGI1 Graduate School of Medicine, Japan. encephalitis; however, encephalitis associated with Author contributions: Dr. Hayashi: study concept and design, writing contactin-associated protein–like 2 (Caspr2) anti- a paper, and patient care and cure. Dr. Yamada: patient care and cure. Dr. Kimura: management for clinical care and cure. Dr. Inuzuka: bodies cannot be ruled out because this type of study supervision and management for clinical care and cure. VGKC-complex antibody can also associate with Acknowledgment: The authors thank Dr. Josep Dalmau for studying limbic encephalitis.5 LGI1 antibody. To the best of our knowledge, some patients Study funding: No targeted funding reported. with anti–NMDAR encephalitis can also develop Disclosure: Y. Hayashi received research support from The Ministry hypothermia.6,7 Themechanismofhypothermiain of Education, Culture, Sports, Science and Technology of Japan. M. Yamada and A. Kimura report no disclosures. T. Inuzaka autoimmune-mediated encephalitis is not clear. received research support from Eizai Co, Dainippon Sumitomo The LGI1 protein is widely expressed in the neu- Pharmaceutical Co, Takeda Pharmaceutical Co, Otsuka Pharmaceutical rons of the CNS and also in the hypothalamus.2 Co, GlaxoSmithKline, The Ministry of Health, Labour and Welfare of Japan, and The Ministry of Education, Culture, Sports, Science and The hypothermic attacks repeatedly occurred Technology of Japan. Go to Neurology.org/nn for full disclosure forms. before effective immunotherapy. Therefore, we The Article Processing Charge was funded by the authors.

2 Neurology: Neuroimmunology & Neuroinflammation This is an open access article distributed under the terms of the 3. Jacob S, Irani SR, Rajabally YA, et al. Hypothermia in Creative Commons Attribution-NonCommercial-NoDerivatives Li- VGKC antibody-associated limbic encephalitis. J Neurol cense 4.0 (CC BY-NC-ND), which permits downloading and shar- Neurosurg Psychiatry 2008;79:202–204. ing the work provided it is properly cited. The work cannot be 4. Montiel P, Sellal F, Clerc C, Richard P, Bataillard M. changed in any way or used commercially without permission from Limbic encephalitis with severe sleep disorder associated the journal. with voltage-gated potassium channels (VGKCs) anti- Received February 13, 2017. Accepted in final form March 10, 2017. bodies [in French]. Rev Neurol (Paris) 2008;164: Correspondence to Dr. Hayashi: [email protected] 181–184. 5. van Sonderen A, Arino H, Petit-Pedrol F, et al. The clinical spectrum of Capr2 antibody-associated disease. Neurology 1. Lai M, Huijbers MGM, Lancaster E, et al. Investigation of 2016;87:521–528. LGI1 as the antigen in limbic encephalitis previously attributed 6. Dalmau J, Tuzun E, Wu H, et al. Paraneoplastic anti-N- to potassium channels: a case series. Lancet Neurol 2010;9: methyl-D-aspartate receptor encephalitis associated with 776–785. ovarian teratoma. Ann Neurol 2007;61:25–36. 2. Irani SR, Pettingill P, Kleopa KA, et al. Morvan syndrome: 7. Tonomura Y, Kataoka H, Hara Y, et al. Clinical analysis of clinical and serological observations in 29 cases. Ann Neurol paraneoplastic encephalitis associated with ovarian teratoma. 2012;72:241–255. J Neurooncol 2007;84:287–292.

Neurology: Neuroimmunology & Neuroinflammation 3 Clinical/Scientific Notes

William Roth, MD ENCEPHALOMYELITIS FOLLOWING DEFINITIVE after the onset of her viral syndrome, she awoke with Charles Tyshkov, MD ZIKA VIRUS INFECTION numbness in both legs extending circumferentially Kiran Thakur, MD from both knees to both feet; the numbness gradually Wendy Vargas, MD extended proximally to the trunk above the umbili- An 18-year-old woman without previous medical is- cus, at which time she presented to the emergency sues was examined at an academic medical center Neurol Neuroimmunol department. On initial evaluation, she was afebrile Neuroinflamm for a complaint of ascending numbness in the legs fol- and with stable vital signs. Her examination was 2017;4:e349; doi: 10.1212/ lowing viral illness. She had taken a trip in early NXI.0000000000000349 remarkable for impairment of fine touch, tempera- spring 2016 to the Dominican Republic, during ture, and pinprick from the T6 dermatome extending which time she was bitten several times by mosqui- distally, and a symmetric gradient fine touch, temper- toes. Approximately 2 weeks following return, she ature, and pinprick deficit was also noted in both developed malaise, nausea, and a nonpruritic macular hands. Right-sided deep tendon reflexes were 31 rash of the trunk, legs, palms, and soles of the feet. throughout, and Babinski sign was positive on the Initial workup was notable for a positive serum Zika left foot. virus PCR (Wadsworth Center, NYC Health Depart- MRI of the cervical, thoracic, and lumbar spine ment). Her symptoms gradually resolved. Six weeks with and without contrast demonstrated multiple T2 hyperintense, nonenhancing lesions in the cervical Figure Multifocal CNS lesions following Zika virus infection and thoracic cords. Brain MRI with and without contrast was significant for multifocal subcortical and cal- losal T2/fluid-attenuated inversion recovery hyperintense lesions, some of which with faint enhancement and restricted diffusion (figure). Lum- bar puncture was performed; opening pressure was 17

cm H2O; red blood cell count was 74; white blood cell count was 10 with 99% lymphocytes; protein was 26 mg/dL; and glucose was 56 mg/dL. CSF Zika PCR was negative, and immunoglobulin M (IgM) levels were indeterminate. Three unique oligoclonal bands were present in the CSF, and myelin basic protein was elevated at 180 mg/dL. Extensive evaluation of alternative etiologies was unremarkable (table). Repeat serum testing performed for Zika virus real-time PCR was negative, and the Zika IgM value was equivocal. She was treated with 3 days of IV methylprednisolone for inflammatory encephalomyelitis with symptomatic improvement.

Discussion. Few reports of Zika virus exist involving the CNS in adolescence and adulthood. Here, we describe a teenage patient with no neurologic history presenting with inflammatory lesions of both the brain and spinal cord, consistent with an acute demy- (A) MRI brain axial FLAIR sequence. Two characteristic lesions, periventricular and splenium elinating process, 47 days following proven Zika virus of corpus callosum. (B) MRI brain axial DWI sequence. Characteristic lesion with diffusion infection by serum PCR. restriction (ADC not shown). (C) MRI thoracic spine sagittal MRI T2 sequence with characteristic lesions. (D) MRI cervical spine sagittal MRI T2 sequence with C3–4 lesion. FLAIR 5 fluid- The long latency period between acute systemic attenuated inversion recovery. Zika virus infection and her neurologic presentation

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 4 weeks of acute viral symptoms.1–5 The prolonged Table Laboratory evaluation latency between the onset of systemic viral symptoms

Reference and development of active inflammatory lesions, range, Laboratory test adults Result however, suggests the possibility of a postinfectious inflammatory response, although it is also possible Infectious studies: serum that her neurologic presentation was not related to Zika virus serum PCR Negative Negative her previous Zika virus infection given the lack of Zika virus serum IgM by MAC-ELISA Negative Equivocal antibody detection of Zika virus in the CSF. Zika virus serum PRNT Negative Positive Our patient’s imaging met the 2010 McDonald

Dengue virus serum PRNT Negative Negative criteria for MS; although a future polyphasic course

West Nile virus serum microsphere Nonreactive Nonreactive would be consistent with MS, a monophasic presenta- immunofluorescence assay tion would be more typical of acute disseminated Arbovirus PRNT Nonreactive Results suggest evidence of encephalomyelitis (ADEM). Encephalomyelitis in the a current or recent infection with Zika virus absence of altered mental status also points to an MS-

T-cell panel like phenotype, although this patient may prove to have no further signs of relapsing-remitting disease. CD4 T cells, cells/mL 393–1489 764 Infection with Epstein-Barr virus (EBV) has m – CD8 T cells, cells/ L 148 788 381 already been proposed as a possible predisposing B cells, cells/mL 61–530 367 factor for developing MS.6 Epidemiologic data Natural killer cells, cells/mL 25–488 176 have shown MS risk as being 10 times greater

CD4/CD8 ratio 0.7–3.6 Normal among individuals who experienced an undiag-

Serum HSV1/2, CMV, VZV, and EBV PCR Negative Negative nosed EBV infection in childhood and at least testing 20-fold greater among individuals who developed Serum EBV-VCA, IgG Negative Positive mononucleosis.6 Exacerbations of both MS and

CSF studies ADEMarealsoknowntobetriggeredbyavariety

Admission CSF profile of other systemic viral infections, which can stim- ulate production of proinflammatory cytokines, Erythrocytes (per microliter) 0–0 136 (tube 1); 74 (tube 4) CD81 T-cell activation, and subsequent central Leukocytes (per microliter) 0–5 15 (tube 1: 98% 7 lymphocytes demyelination. and 2% monocytes); 10 Although Zika virus is not yet among the viruses (tube 4: 99% lymphocytes and 1% monocytes) postulated to play a pathogenic role in the onset or

Protein, mg/dL 15–45 26 exacerbation of central demyelinating syndromes,

Glucose, mg/dL 40–70 56 (serum glucose 121) it is known to cause neurologic disease both directly and by secondary autoimmunity as evi- Zika virus CSF PCR Negative Negative denced by the growing number of cases of congen- Zika virus CSF IgM Negative Equivocal ital birth defects and Guillain-Barré syndrome. Adenovirus, West Nile, CMV, VZV, EBV, Negative Negative Emerging evidence suggests that Zika virus causes enterovirus, HHV-6, Eastern equine encephalitis virus, St. Louis encephalitis virus, a spectrum of neurologic manifestations, and fur- human parechovirus, and CSF PCR testing ther studies are required to fully define this spec- Escherichia coli K1, Haemophilus influenzae, Negative Negative Listeria monocytogenes, Neisseria trum. With the recent international spread of Zika meningitidis, Streptococcus agalactiae, virus, practitioners should be aware of the range of S pneumoniae, and Cryptococcus neoformans/gattii potential neurologic effects of immediate and de-

CSF myelin basic protein 0–5.5 179.6 layed impact of Zika virus infection in both infants and adults. CSF IgG index 0.09–0.25 0.36

CSF immunoglobulins 0–64.6 From the Department of Neurology (W.R., K.T., W.V.), and Department of Pediatric Neurology (C.T., W.V.), Columbia Uni- Urine studies versity Medical Center, New York, NY. Zika virus urine PCR Negative Negative Author contributions: Dr. Roth, Dr. Tyshkov, Dr. Thakur, and Dr. Vargas: clinical care, drafting, and critical revision of the manuscript Abbreviations: CMV 5 cytomegalovirus; EBV 5 Epstein-Barr virus; HSV 5 herpes simplex for intellectual content. virus; IgG 5 immunoglobulin G; IgM 5 immunoglobulin M; MAC-ELISA 5 IgM antibody Study funding: No targeted funding reported. capture ELISA; PRNT 5 plaque-reduction neutralization rest; VCA 5 viral capsid antigen; VZV 5 varicella zoster virus. Disclosure: W. Roth and C. Tyshkov report no disclosures. K. Thakur received research support from NICHD/NIH. W. Vargas has consulted for Alexion Pharma; served on the speakers’ bureau for Teva; is unusual, as previous studies have identified similar and received research support from Teva, National Center for Advancing Translational Sciences, NIH, and National Multiple manifestations including Guillain-Barré syndrome, Sclerosis Society. Go to Neurology.org/nn for full disclosure forms. transverse myelitis, and meningoencephalitis within The Article Processing Charge was funded by the authors.

2 Neurology: Neuroimmunology & Neuroinflammation This is an open access article distributed under the terms of the 3. Soares CN, Brasil P, Carrera RM, et al. Fatal encephalitis Creative Commons Attribution-NonCommercial-NoDerivatives Li- associated with Zika virus infection in an adult. J Clin Virol cense 4.0 (CC BY-NC-ND), which permits downloading and shar- 2016;83:63–65. ing the work provided it is properly cited. The work cannot be 4. Brito Ferreira ML. Neurologic manifestations of arboviruses changed in any way or used commercially without permission from in the epidemic in Pernambuco, Brazil. Conference of the the journal. American Academy of Neurology; April 15–21, 2016; Received December 11, 2016. Accepted in final form March 13, Vancouver, BC, Canada. Abstract. 2017. 5. Parra B, Lizarazo J, Jiménez-Arango JA, et al. Guillain-Barré syndrome associated with Zika virus infection in Colombia. Correspondence to Dr. Roth: [email protected] N Engl J Med 2016;375:1513–1523. 6. Ascherio A, Munger KL. Environmental risk factors for 1. Carteaux G, Maquart M, Bedet A, et al. Zika virus associ- multiple sclerosis: part I: the role of infection. Ann Neurol ated with meningoencephalitis. N Engl J Med 2016;374: 2007;61:288–299. 1595–1596. 7. Cusick MF, Libbey JE, Fujinami RS. Multiple sclerosis: 2. Mecharles S, Herrmann C, Poullain P, et al. Acute myelitis autoimmunity and viruses. Curr Opin Rheumatol 2013; due to Zika virus infection. Lancet 2016;387:1481. 25:496–501.

Neurology: Neuroimmunology & Neuroinflammation 3 Clinical/Scientific Notes

Lekha Pandit, MD, DM, SPONTANEOUS REMISSION LASTING MORE limbs with urinary retention. As reported by the PhD THAN A DECADE IN UNTREATED AQP4 patient, he underwent a myelogram for the same Sharik Mustafa, MD, DM ANTIBODY-POSITIVE NMOSD and was thought to have a tumor in the cervical spinal cord. He was empirically given a course of Neurol Neuroimmunol steroids for 3 weeks following which he gradually Neuroinflamm Neuromyelitis optica spectrum disorder (NMOSD) 2017;4:e351; doi: 10.1212/ associated with aquaporin-4 antibody (AQP4-Ab) regained power in his limbs and remained well NXI.0000000000000351 predominantly targets the optic nerve and spinal for the next 22 years. In 2001, at the age of 36 cord and is characterized by frequent relapses and years, he had left optic neuritis which recovered accrual of irreversible disability.1 Variations in poorly with steroids. Since then, he had myelitis clinical course may include an initial or predomi- averaging 1 attack every 2 years, mostly in the form “ ” nant involvement in extra spinal/optic nerve loca- of a partial cord syndrome. He presented to our tions and a benign course marked by mild attacks center in 2011 where MRI (figure, C and D) and and minimal residual deficits.2,3 However, sponta- AQP4-Ab testing were confirmatory for NMOSD. neous remissions are rare. We are reporting 2 pa- At present, visual acuity in his right eye is 20/200 tients who satisfied the 2015 criteria for NMOSD4 and in the left eye counting fingers at 1 foot dis- and had a unique disease course highlighted by tance. He remains attack free after starting periods of spontaneous remission lasting 19 and immunosuppressants. 23 years. Discussion. The 2 patients described in our report Methods. Patient 1. In 1995, a then 31-year-old had a unique clinical course characterized by long woman developed severe unexplained vomiting for periods of spontaneous remission, 19 and 23 years. nearly a month, followed by a brief period of ataxia, Our first patient had in retrospect an area postre- diplopia, and dysarthria. The following year, she ma syndrome followed by severe bilateral optic had urinary retention for a week without any limb neuritis. Evaluation of later events in patient 2 sug- weakness. In the third year of her illness, she had gests that the first attack was likely to have been sequential optic neuritis staring in the left eye for a tumefactive myelitis which responded to ste- which she received a short course of oral steroids. roids. Both patients had chronic disease spanning Her vision remained poor ever since. In subsequent decades, with severe visual loss and minimal spinal years, she had no other medication except treat- cord dysfunction. ment for hypothyroidism. Nineteen years later, at Wingerchuck et al.5 have highlighted the varied the age of 51 years, she presented to us with optic clinical course of NMO. While 90% of patients neuritis in the right eye. Review of her medical relapse within 5 years after the index event, some records revealed that brain MRI (1995–1997) may have a much longer disease-free period.5 We documented an initially enhancing lesion in the were unable to find published cases in the litera- medulla extending into the high cervical cord. ture that highlighted extended periods of sponta- Visual evoked potentials were abnormal and oli- neous remission similar to our patients. A recent goclonal bands were negative. During her current case report identified an AQP4-Ab–positive ori- admission, her corrected visual acuity was 20/200 ental woman who had 11 episodes of recurrent in both eyes. The rest of neurologic examination unilateral optic neuritis over a period of 22 years.6 was normal. Her recent MRI (figure, A and B) was Localization to the optic nerve and the lengthy consistent with the diagnosis of NMOSD, and disease course were similar to our patients. There optical coherence tomography demonstrated the has been a phenomenal expansion of our knowl- extent of visual loss (figure, E). Cell-based assay for edge of the clinical spectrum of NMOSD in recent AQP4-Ab was positive (EUROIMMUN, Lübeck, years. Our report contributes additional informa- Germany). tion that spontaneous and lengthy remissions may Patient 2. In 1983, at the age of 28 years, this occur and particularly in the early phase of AQP4- male patient had acute onset weakness of lower Ab–mediated NMOSD. A review of history going

(A and B) Hyperintensity (FATSAT axial) and enhancement of the right optic nerve. (C) Resolving LETM (T2W sagittal) in the cervicodorsal spinal cord. (D) Bilateral long segment atrophy of optic nerves. (E and F) OCT images (ring scan data) from both eyes of each patient showing severe retinal nerve fiber layer thinning compared with normative device data (red areas indicate thickness below first percentile). G 5 global; FATSAT 5 fat saturated; LETM 5 longitudinally extensive transverse myelitis; N 5 nasal; NI 5 nasal-inferior; NT 5 nasal-superior; OCT 5 optical coherence tomography; PMB 5 peripapillary macular bundle; T 5 temporal; TI 5 temporal-inferior; TS 5 temporal-superior.

well beyond the conventional period may unearth 1. Jarius S, Ruprecht K, Wildemann B, et al. Contrasting similar cases in other large NMOSD registries and disease patterns in seropositive and seronegative neuromye- provide a valuable cohort of patients for future litis optica: a multicentre study of 175 patients. J Neuro- inflammation 2012;9:14. studies. 2. Collangues N, Cabre P, Marignier R, et al. A benign form of neuromyelitis optica: does it exist? Arch Neurol 2011;68: From the Department of Neurology, KS Hegde Medical Academy, – Nitte University, Mangalore, India. 918 924. 3. Bergamaschi R, Jarius S, Robotti M, et al. Two cases of Study funding: No targeted funding reported. benign neuromyelitis optica in patients with celiac disease. Disclosure: The authors report no disclosures. Go to Neurology.org/nn J Neurol 2009;256:2097–2099. for full disclosure forms. The Article Processing Charge was funded by the authors. 4. Wingerchuck DM, Banwell B, Bennett J, et al. Interna- tional consensus diagnostic criteria for neuromyelitis optica This is an open access article distributed under the terms of the – Creative Commons Attribution-NonCommercial-NoDerivatives Li- spectrum disorders. Neurology 2015;85:177 189. ’ cense 4.0 (CC BY-NC-ND), which permits downloading and shar- 5. Wingerchuk DM, Hogancamp WF, O Brien PC, et al. The ing the work provided it is properly cited. The work cannot be clinical course of neuromyelitis optica (Devic’s syndrome). changed in any way or used commercially without permission from Neurology 1999;53:1107–1114. the journal. 6. Yew CY, Hor YJ, Lim TT, et al. Eleven episodes of recur- Received February 1, 2017. Accepted in final form March 24, 2017. rent optic neuritis of the same eye for 22 years eventually diagnosed as neuromyelitis optica spectrum disorder. Mult Correspondence to Dr. Pandit: [email protected] Scler Relat Disord 2016;10:22–25.

2 Neurology: Neuroimmunology & Neuroinflammation Clinical/Scientific Notes

Zuzana Liba, MD, PhD ALEMTUZUMAB AND INTRATHECAL consecutive days) was combined with intrathecal Petr Sedlacek, MD METHOTREXATE FAILED IN THE THERAPY OF administration of methotrexate (MTX) (12 mg for Vera Sebronova, MD RASMUSSEN ENCEPHALITIS a single dose).6 Finally, tacrolimus was switched to Alice Maulisova, MA mycophenolate mofetil because of its side effects Bertil Rydenhag, MD Rasmussen encephalitis (RE) is a rare but devastating and ineffectiveness. Josef Zamecnik, MD unihemispheric brain disorder that often affects chil- The administration of alemtuzumab was com- Martin Kyncl, MD dren.1 The clinical picture is characterized by intrac- plicated by an unexpected severe systemic reaction Pavel Krsek, MD table focal epilepsy and progressive decline of and hyperpyrexia that required intensive care for 21 functions associated with the affected hemisphere.2 days. Sedation, physical cooling, ventilation sup- Neurol Neuroimmunol port, and high doses of IV immunoglobulin were Neuroinflamm Despite its known inflammatory background and 2017;4:e354; doi: 10.1212/ T-cell involvement, immunotherapy appears to slow required to manage this situation. After that, NXI.0000000000000354 rather than halt disease progression, and hemispher- a change in disease course was observed. There were otomy appears to be the only solution for intractable no clinical seizures apart from a fine EPC in the epilepsy.1–4 A potential early therapeutic window has right hand, and overall cognitive performance (in been suggested, and new therapeutic agents have particular psychomotor speed, attention, and ver- become available.1 A monoclonal antibody targeting bal fluency) was improved. Brain MRI revealed CD52 that leads to long-term depletion of lympho- a regression of the inflammatory changes (figure cytes (alemtuzumab) has previously been considered e-1C). An additional dose of intrathecal MTX as a possible treatment option for RE, but clinical was administered to strengthen the positive effect. data are limited.1,5 Unfortunately, a clinical relapse characterized by clustering seizures, accented right-side hemiparesis Case report. A previously healthy, right-handed boy (in terms of muscle weakness) and aphasia occurred developed refractory epilepsy at the age of 7 years. 8 weeks later; however, repeating the intrathecal Based on clinical, electroencephalographic, and neu- MTX led to prompt stabilization. Thus, we contin- roimaging features,3 he was diagnosed with RE of the ued to repeat the intrathecal therapy with gradually dominant hemisphere at the age of 8 years. Unre- extending intervals and clinical stabilization was markable right-sided hemiparesis (pyramidal signs reached for 6 months. Nevertheless, when we but no permanent weakness) and above-average extended the intervals up to a maximum of 3 intellectual capacity were described at that time. months, new flares of the disease occurred (figure Video-EEG showed left-hemispheric epileptiform 1, figure e-1D). Severe lymphopenia in the periph- activity and multiple types of epileptic seizures, includ- eral blood and the absence of lymphocytes in the ing epilepsy partialis continua (EPC) of the right CSF persisted throughout the entire time regardless hand. Brain MRI revealed mild left-hemispheric atro- of clinical relapses. Hemispherotomy was finally phy without any inflammatory signal changes or gado- performed 2 months after the last dose of intrathe- linium enhancement (figure e-1A at Neurology.org/nn). cal MTX, when the boy reached 10 years of age and Blood-brain barrier failure was observed, with no was in a phase of clinical stabilization. At the time pleocytosis in the CSF. Common immunotherapy of surgery, he had preserved function of the right that was escalated as displayed in figure 1 was hand, walked independently, spoke in sentences, applied during the first year of the therapy. Despite and had no seizures apart from an EPC of the right Supplemental data this, brain MRI showed new signal changes (figure hand. Brain MRI showed ongoing atrophy but no at Neurology.org/nn e-1B) and the boy suffered from refractory epilepsy inflammatory signal changes (figure e-1E). Brain and deteriorated. biopsies collected from different lobes showed sim- Thus, further escalation of the immunosuppres- ilar unremarkable inflammatory changes, including sion was applied. According to our previous experi- mild astrogliosis and scattered lymphocytic perivas- ence with a different diagnosis, IV treatment with cular infiltration with CD81 T-cell predominance alemtuzumab (total dose of 0.75 mg/kg for 3 (figuree-1G).Severeneurologicsequelaewere

The timeline of multiple drugs is displayed in the figure. Clinical clusters of epileptic seizures with the need for hospitalization are marked. The concurrent treatment with antiepileptic drugs and anti-infective prophylaxis is not indicated. Common immunotherapy was applied at the beginning as follows: (1) IVMP followed by a slow oral steroid tapper, (2) high doses of monthly repeated IVIG, and (3) tacrolimus—the dose was titrated according its levels. Because of ongoing clinical deterioration and signs of neuroinflammation on MRI, (4) RTX was added. Despite this, the cognitive and motor deterioration slowly progressed. Thus, a novel approach of further escalation of the immunosuppression was applied as follows: (5) IV alemtuzumab, (6) intrathecal MTX with intrathecal steroid. In addition, tacrolimus was switched to (7) mycophenolate mofetil. Finally, (8) hemispherotomy was performed, and the medication has been gradually tapered down. IVIG 5 IV immunoglobulin; IVMP 5 IV methylprednisolone; MTX 5 methotrexate; RTX 5 rituximab.

observed after the neurosurgery, as expected (figure (B.R.), Department of Clinical Neuroscience, Sahlgrenska Academy e-1F). At 18 months after the surgery, the patient at the University of Gothenburg, Sweden. has no functional use of the right hand, hardly Author contributions: Z.L.: concept and design, analysis and interpretation, and critical revision for important intellectual content. walks, and speaks in very simple sentences. He P.S.: concept and design and critical revision for important intellec- is seizure free and attends school with special tual content. V.S. and A.M.: acquisition of data. B.R.: acquisition of assistance. data and analysis and interpretation. J.Z. and M.K.: acquisition of data. P.K.: analysis and interpretation and supervision. Discussion. Extremely demanding therapeutic Acknowledgment: The authors thank Michal Tichy for management effort was performed in an attempt to prevent of neurosurgery and Vladimir Komarek for his support. hemispherotomy of the dominant hemisphere in Study funding: “Neuron for Science Support” foundation. a boy with RE and a mild neurologic deficit. Disclosure: Z. Liba received research support from Neuron for Science Support Foundation. P. Sedlacek, V. Sebronova, A. Maulisova, and Although the cause of RE remains elusive, T cells B. Rydenhag report no disclosures. J. Zamecnik serves as Editor-in- 1 are involved in the pathology. The aim of our Chief for Czecho-Slovak Pathology. M. Kyncl and P. Krsek report combined therapy was to deplete T cells from the no disclosures. Go to Neurology.org/nn for full disclosure forms. The peripheral blood and influence the inflammatory Article Processing Charge was funded by the University Hospital Motol. process behind the blood-brain barrier.5 However, This is an open access article distributed under the terms of the the boy experienced a life-threatening systemic Creative Commons Attribution-NonCommercial-NoDerivatives Li- reaction immediately after alemtuzumab adminis- cense 4.0 (CC BY-NC-ND), which permits downloading and shar- tration and later on, the clinical stabilization ing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from seemed to be dependent on intrathecal MTX. It the journal. was not possible to continue with this therapy Received January 18, 2017. Accepted in final form April 6, 2017. because of its known cumulative side effects and Correspondence to Dr. Liba: [email protected] neurotoxicity. We believe the brain biopsy demonstrated that it was possible to temporarily control 1. Varadkar S, Bien CG, Kruse CA, et al. Rasmussen’s enceph- brain inflammation, but at the cost of inappropriate alitis: clinical feature, pathology and treatment advances. – risks. We hypothesize that our aggressive immu- Lancet Neurol 2014;13:195 205. 2. Olson HE, Lechpammer M, Prabhu SP, et al. Clinical notherapy failed for the following reasons: (1) we application and evaluation of the Bien diagnostic criteria 1,7 missed the early therapeutic window ;(2)the for Rasmussen encephalitis. Epilepsia 2013;54:1753– pathology of RE is more complex, and the immu- 1760. nosuppression was not enough to cure the disease.1 3. Bien CG, Schramm J. Treatment of Rasmussen encephalitis half a century after its initial description: promising pros- From the Department of Pediatric Neurology (Z.L., V.S., A.M., pects and dilemma. Epilepsy Res 2009;86:101–112. P.K.), Department of Pediatric Hematology and Oncology (P.S.), Department of Pathology and Molecular Medicine (J.Z.), Depart- 4. Thilo B, Stingele R, Knudsen K, et al. A case of Rasmussen ment of Radiology (M.K.), Charles University, 2nd Faculty of Med- encephalitis treated with rituximab. Nat Rev Neurol 2009; – icine and Motol University Hospital; Department of Clinical 5:458 462. Psychology (A.M.), Charles University, Faculty of Arts, Prague, 5. Ruck T, Afzali AM, Lukat KF, et al. ALAIN01-Alemtuzumab Czech Republic; and Institute of Neuroscience and Physiology in autoimmune inflammatory neurodegeneration: mechanisms

2 Neurology: Neuroimmunology & Neuroinflammation of action and neuroprotective potential. BMC Neurol 2016; 7.LibaZ,MuthaffarO,TangJ,etal.Rasmussenen- 16:34. cephalitis: response of early immunotherapy in a case 6. Liba Z, Sebronova V, Komarek V, Sediva A, Sedlacek P. of immune-mediated encephalitis. Neurol Neuroimmunol Prevalence and treatment of anti-NMDA receptor enceph- Neuroinflamm 2015;2:e69. doi: 10.1212/NXI. alitis. Lancet Neurol 2013;12:424–425. 0000000000000069.

Neurology: Neuroimmunology & Neuroinflammation 3 Clinical/Scientific Notes

Leana Doherty, MD ANTI-DPPX ENCEPHALITIS: PROMINENT spontaneous downbeat nystagmus (DBN) was seen, Daniel Gold, DO NYSTAGMUS REFLECTED BY EXTRAOCULAR and it was noted that changes in head orientation rel- Lilja Solnes, MD MUSCLE FDG-PET AVIDITY ative to gravity (head tilt and supine position) transi- John Probasco, MD tioned his DBN to UBN, a feature that can be seen 1 Arun Venkatesan, MD, A 43-year-old Caucasian man developed persistent with utricle pathway damage. Overall, in our pa- ’ PhD nystagmus after 10 months of weight loss, diarrhea, tient s case, ocular motor signs would be best ex- insomnia, and 5 months of agitation, memory plained by involvement of the perihypoglossal Neurol Neuroimmunol nuclei (UBN and GEN), paramedian tract cell Neuroinflamm impairment, and vision changes consisting of nystag- 2017;4:e361; doi: 10.1212/ mus and double vision. Ocular motor examination groups, or their connections with the vestibulocere- NXI.0000000000000361 demonstrated left-beating nystagmus (LBN), upbeat- bellum (DBN), and vestibular nuclei (causing SCC ing nystagmus (UBN), and torsional (toward the left and utricle imbalance, right more than left-sided dys- ear) nystagmus components (video at Neurology.org/ function given skew deviation with left hypertropia nn), as well as a left hypertropia from a skew devia- and LBN). tion. There was also gaze-evoked nystagmus (GEN) DPPX, a recently recognized neuronal surface horizontally. General neurologic examination re- autoantigen, is a regulatory subunit of the voltage- vealed postural tremor in his arms, brisk reflexes, mild gated Kv4.2 potassium channel complex expressed dysmetria with finger to nose, and wide-based gait. on neuronal dendrites and soma, dysfunction of – Detailed neurocognitive testing demonstrated deficits which may result in neuronal hyperexcitability.2 4 in recall and attention. He had previously been Kv4.2 channels and DPPX proteins are distributed healthy, and his family history was notable only for throughout the nervous and enteric systems, includ- multiple sclerosis in his brother. Brain MRI showed ing the cortex, cerebellum, brainstem, and myenteric no abnormalities, and CSF studies showed elevated and submucous plexus.2,3,5 The widespread expres- protein. A broad screen for infections, toxins, and sion of DPPX is believed to underlie the broad systemic autoimmune disorders was negative. Periph- manifestations of anti–DPPX-associated disease, eral blood cell counts were normal. Whole-body 18- including neurocognitive deficits, sleep disturbance, fluoro-deoxyglucose PET (18FDG-PET-CT) did not central hyperexcitability, and diarrhea. The majority reveal a malignancy, but showed asymmetric avidity of patients described have gastrointestinal symptoms, of the extraocular muscles (figure). Testing for neu- including diarrhea, constipation, and weight loss, and ronal autoantibodies revealed a serum dipeptidyl- these symptoms as well as neurologic symptoms may peptidase-like protein-6 (DPPX) antibody IgG titer present insidiously, thus making diagnosis challeng- of 1:15,360 (Neuroimmunology Laboratory, Mayo ing. Eye movement disturbances are common. Of 20 Clinic), providing evidence for immune-mediated seropositive anti-DPPX patients, diplopia, oscillopsia, encephalitis. DPPX antibody from CSF was not or blurred vision was reported in 8 patients, with one tested because of inadequate sample quantity before case of upbeat-torsional nystagmus.3 Additional the initiation of treatment. described visual disturbances include saccadic pursuit He was treated with IV methylprednisolone, gaze movements in all directions; spontaneous down- plasma exchange, and rituximab, with improvement beat and gaze-evoked nystagmus, broken pursuit, and in hyperreflexia, gait, memory, sleep, mood, and incomplete suppression of vestibulo-ocular reflex on appetite. His nystagmus and skew deviation also fixation; and broken pursuit and GEN with diplopia Supplemental data improved markedly (video). The UB-torsional nys- on lateral gaze.3,6 Cerebellar and pontine targets by at Neurology.org/nn tagmus (which had an LBN component) seen at pre- autoantibodies resulting in CNS hyperexcitability sentation was believed to result mainly from either may underlie these ocular manifestations. semicircular canal (SCC) pathway imbalance at the Although FDG-PET is often used to assess malig- level of the medulla or, in light of the skew deviation, nancy in patients with autoimmune encephalitis, may disruption of the utricle-ocular motor pathways. In demonstrate abnormalities in brain metabolism, and support of the latter, in follow-up months later, has been reported to reflect muscle hyperactivity in

Neurology.org/nn Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 (D.G.), Russell H. Morgan Department of Radiology and Radiolog- Figure F18-FDG-PET/CT ical Sciences (L.S.), and Johns Hopkins Encephalitis Center (J.P., A.V.), Johns Hopkins University School of Medicine, Baltimore, MD. Author contributions: Leana Doherty: drafting and revising of the manuscript and selecting initial images. Daniel Gold: filming and video acquisition, editing of the manuscript, and video legend. Lilja Solnes: selecting and editing of images. John Probasco: editing of the manuscript. Arun Venkatesan: critical editing of the manuscript. Study funding: No targeted funding reported. Disclosure: L. Doherty reports no disclosures. D. Gold is on the editorial board for Current Treatment Options in Neurology; co-section editor for Neuro-ophthalmology and Otology; and consulted for WellPoint Insurance. L. Solnes reports no disclosures. J. Probasco served on the editorial board and is associate editor for The Neurohospitalist and is Editor-in-Chief for NEJM Journal Watch Neurology. A. Venkatesan served on the scientific advisory board for MedImmune; served as a medical expert for the U.S. Government Vaccine Injury Compensation Program; received research support from NIH; and served as a medical expert for Carnival Cruise Lines. Go to Neurology.org/nn for full disclosure forms. The Article Process- ing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives Li- cense 4.0 (CC BY-NC-ND), which permits downloading and shar- 18 F -FDG-PET/CT demonstrates normal size of extraocular ing the work provided it is properly cited. The work cannot be muscles but markedly asymmetric increased FDG activity changed in any way or used commercially without permission from localizing to the left medial rectus and right lateral rectus the journal. (A and B, arrows), and inferior recti bilaterally (C and D, arrows). Received March 13, 2017. Accepted in final form April 4, 2017.

Correspondence to Dr. Venkatesan: [email protected] other autoimmune neurologic conditions, little is 6 1. Suzuki Y, Matsuda T, Washio N, Ohtsuka K. Transition known of PET findings in anti-DPPX encephalitis. from upbeat to downbeat nystagmus observed in a patient The FDG-PET in our case did not show evidence of with Wernicke’s encephalopathy. Jpn J Ophthalmol 2005; malignancy, but did demonstrate asymmetric metab- 49:220–222. olism in extraocular muscles correlating with the slow 2. Boronat A, Gelfand JM, Gresa-Arribas N, et al. Encephalitis (pathologic) phases of nystagmus. In particular, and antibodies to dipeptidyl-peptidase-like protein-6, a sub- FDG-PET showed increased FDG activity in the left unit of Kv4.2 potassium channels. Ann Neurol 2013;73: 120–128. medial rectus, right lateral rectus, and bilateral infe- 3. Tobin WO, Lennon VA, Komorowski L, et al. DPPX rior recti, correlating with the slow phases of his hor- potassium channel antibody: frequency, clinical accompani- izontal (LB) and vertical (UB) nystagmus, ments, and outcomes in 20 patients. Neurology 2014;83: respectively. Although FDG-PET correlates of nys- 1797–1803. tagmus have not been previously characterized, 4. Piepgras J, Höltje M, Michel K, et al. Anti-DPPX enceph- increased avidity of bilateral medial recti has been alitis: pathogenic effects of antibodies on gut and brain neurons. Neurology 2015;85:890–897. reported in a case of convergence spasm due to brain 5. Clark BD, Kwon E, Maffie J, et al. DPP6 Localization in 7 tumor. brain supports function as a Kv4 channel associated protein. Overall, our case highlights the prominent eye Front Mol Neurosci 2008;1:8. movement abnormalities that can occur in the setting 6. Stoeck K, Carstens PO, Jarius S, et al. Prednisolone and of anti-DPPX encephalitis, points to specific pontine azathioprine are effective in DPPX antibody–positive auto- and medullary nuclei that may be impacted, and immune encephalitis. Neurol Neuroimmunol Neuroinflamm highlights that FDG-PET can demonstrate correlates 2015;2:e86. doi: 10.1212/NXI.0000000000000086. 7. Jeong SH, Oh YM, Kim CY, Kim JS. Bimedial rectus of nystagmus in this disease. hypermetabolism in convergence spasm as observed on pos- From the Department of Neurology (L.D., D.G., J.P., A.V.), De- itron emission tomography. J Neuroophthalmol 2008;28: partments of Ophthalmology, Otolaryngology, and Neurosurgery 217–218.

2 Neurology: Neuroimmunology & Neuroinflammation Clinical/Scientific Notes

Edward Bahou, MD ANCA-ASSOCIATED VASCULITIS favored her right side and fatigued very quickly, Lan Zhou, MD, PhD PREDOMINANTLY PRESENTING WITH SEVERE requiring frequent rest. MYALGIAS Laboratory testing revealed mildly elevated Neurol Neuroimmunol creatinine phosphokinase level of 367 IU/L (normal Neuroinflamm – – 2017;4:e365; doi: 10.1212/ The peripheral nervous system is frequently involved 25 175) and aldolase of 9.1 U/L (normal 1.5 8.1). NXI.0000000000000365 in anti–neutrophil cytoplasmic antibody (ANCA)- C-reactive protein (CRP) and erythrocyte sedimenta- associated vasculitis (AAV), with typical presentations tion rate (ESR) were markedly elevated at 124 mg/L – – such as mononeuritis multiplex, distal sensorimotor (normal 0 5.0) and 79 mm/h (normal 0 20), respec- polyneuropathy, and isolated cranial mono- tively. ANCA titer was elevated at 1:160 (normal , neuropathies.1,2 Here, we report a case of AAV with 1:20), and myeloperoxidase antibody was positive – unusual presentation of predominant severe myalgias. at 186 U/mL (normal 0 19). Antibodies to proteinase The diagnosis of AAV was made by a muscle biopsy 3 and Jo-1 were negative. Urinalysis showed micro- and positive ANCA. scopic hematuria and proteinuria. Chest CT showed a very small lung nodule. Nerve conduction study was Case report. A 63-year-old woman developed rapid normal. EMG showed no abnormal spontaneous onset of soreness in the muscles of buttocks and activity but subtle early recruitment of a few small thighs. Over the course of 3 weeks, the symptoms motor unit potentials in the biceps and deltoid involved her upper arms. Climbing stairs and washing muscles, suggestive of a nonirritable myopathy. A left her hair became difficult because of the pain. She deltoid muscle biopsy showed acute necrotizing denied spine pain or numbness. She denied fever, vasculitis with transmural inflammation and fibrinoid weight change, appetite loss, joint pain, skin rash, necrosis of several small- and medium-sized perimysial or urinary symptoms. She had a history of hypothy- blood vessels (figure). No myopathic changes or roidism, and had mild left foot weakness as a result endomysial inflammation was seen. She was diagnosed of poliomyelitis infection when she was 3 years old. with AAV, and treated with IV infusion of methyl- She did not smoke cigarettes or drink alcohol. prednisolone 1 g/d for 5 days with dramatic improve- She was admitted to an outside hospital, where ment of her symptoms. Subsequent renal biopsy thyroid-stimulating hormone, antinuclear antibody, showed crescentic glomerulonephritis. She was extractable nuclear antigen, and rheumatoid factor discharged on oral prednisone and started on rituxi- were found to be unremarkable. Cervical and lumbo- mab by rheumatology. sacral spine MRI with and without contrast showed mild multi-level degenerative changes. She was Discussion. ANCA are autoantibodies to neutro- discharged without a clear diagnosis. One week later, philic granules and monocytic lysosomes. ANCA she presented to our emergency room for the have been associated with 3 distinct vasculitides, progressive symptoms that significantly affected her which involve inflammation of the small- and function. She could only walk half a city block and medium-sized blood vessels: microscopic poly- could not function as a podiatrist because of the angiitis (MPA), granulomatosis with polyangiitis severe pain and fatigue in the proximal limb muscles. (GPA, previously known as Wegener gran- She was admitted to our neurology service for the ulomatosis), and eosinophilic GPA (previously management. known as Churg-Strauss Syndrome).3,4 Our case Physical examination showed normal mental represents MPA. status, cranial nerve function, sensation, and coordi- The estimated prevalence of AAV is 46–184 per nation. There was severe tenderness to palpation of million.5 AAV is a multisystem disease. Nervous the bilateral deltoid and biceps muscles. Strength system is frequently involved in AAV with peripheral was intact except for the known mild residual weak- neuropathy predominated in each type of AAV.1 ness in the left foot and toes from her prior polio Patients with MPA are usually older with more severe infection. Deep tendon reflexes were 21 except for renal disease along with skin rash and neuropathy.4 the absent ankle jerks. On her gait examination, she Although myalgias are not uncommon in AAV,

H&E stain showed transmural inflammation (arrows) and fibrinoid necrosis (stars) of several perimysial blood vessels with severe occlusions of blood vessel lumen. Bar 5 50 mm.

which have been reported in 48.2% patients with This is an open access article distributed under the terms of the MPA,6 myalgias as a predominant presentation with- Creative Commons Attribution-NonCommercial-NoDerivatives Li- cense 4.0 (CC BY-NC-ND), which permits downloading and shar- out other systemic symptoms or neuropathies typical ing the work provided it is properly cited. The work cannot be of AAV is exceedingly rare. The clinical diagnosis can changed in any way or used commercially without permission from be delayed as seen in our case because of this rare the journal. entity. Our patient did not have objective weakness Received January 24, 2017. Accepted in final form April 17, 2017. and the EMG findings were very subtle, but the mus- Correspondence to Dr Zhou: [email protected]. cle biopsy showed fulminant necrotizing vasculitis. The severe myalgias are most likely due to muscle 1. Zhang W, Zhou G, Shi Q, Zhang X, Zeng XF, Zhang FC. ischemia. Our patient responded very well to the Clinical analysis of nervous system involvement in ANCA- treatment. Our case illustrates that acute, severe, associated systemic vasculitides. Clin Exp Rheumatol 2009; – and progressive myalgias should raise a suspicion for 27:S65 S69. 2. Koike H, Sobue G. Clinicopathological features of neurop- vasculitis, especially in a setting of markedly elevated athy in anti-neutrophil cytoplasmic antibody-associated ESR and CRP, and positive ANCA. Biopsy of a symp- vasculitis. Clin Exp Nephrol 2013;17:683–685. tomatic muscle is essential to establish a tissue diag- 3. Kallenberg CG. The diagnosis and classification of micro- nosis to initiate prompt treatment. scopic polyangiitis. J Autoimmun 2014;48–49:90–93. Department of Neurology, Icahn School of Medicine at Mount Sinai, 4. Yates M, Watts R. ANCA-associated vasculitis. Clin Med – New York, NY. (Lond) 2017;17:60 64. Author contributions: Dr. Bahou and Dr. Zhou participated in the 5. Watts RA, Mahr A, Mohammad AJ, Gatenby P, Basu N, management of the patient; did chart review and data analysis; and Flores-Suarez LF. Classification, epidemiology and clinical drafted the manuscript. Dr. Zhou interpreted the muscle biopsy. subgrouping of antineutrophil cytoplasmic antibody (AN- Study funding: No targeted funding reported. CA)-associated vasculitis. Nephrol Dial Transpl 2015;30 – Disclosure: Dr. Bahou reports no disclosures. Dr. Zhou is an Associate (suppl 1):i14 i22. Editor for Neurology® Neuroimmunology & Neuroinflamma- 6. Guillevin L, Durand-Gasselin B, Cevallos R, et al. Micro- tion. Go to Neurology.org/nn for full disclosure forms. The Article scopic polyangiitis: clinical and laboratory findings in Processing Charge was funded by the authors. eighty-five patients. Arthritis Rheum 1999;42:421–430.

2 Neurology: Neuroimmunology & Neuroinflammation Clinical/Scientific Notes

María Sepúlveda, MD* VANISHING SPINAL CORD AFTER VARICELLA- manuscript for intellectual content. All authors have approved final approval of the manuscript. Javier Almeida, MD* ZOSTER VIRUS MYELITIS Study funding: No targeted funding reported. Joan Berenguer, MD Disclosure: M. Sepúlveda, J. Almeida, and J. Berenguer report no Nuria Solá-Valls, MD A 35-year-old woman presented with paraparesis, T7 disclosures. N. Solá-Valls received research support from Instituto de Julia Saura, MD Salud Carlos III, Spain and Fondo Europeo de Desarrollo Regional, sensory level, and urinary retention 5 days after devel- Yolanda Blanco, MD Predoctoral grant for Health Research. J. Saura received travel fund- oping chickenpox. Spinal cord MRI showed a longitu- ing and/or speaker honoraria from Pfizer. Y. Blanco reports no dis- Sara Llufriu, MD dinally extensive myelitis (figure 1). Despite closures. S. Llufriu served on the scientific advisory board for Ixico Francesc Graus, MD treatment with IV methylprednisolone (1 g/d 3 5) and Novartis and received travel funding and speaker honoraria from Albert Saiz, MD 3 Biogen, Teva, Novartis, and Genzyme. F. Graus is on the editorial and acyclovir (10 mg/kg/8 h 3 weeks), the patient board for Lancet; holds a patent for use of IgLON5 as a diagnostic developed complete paraplegia, bilateral arm paresis, Neurol Neuroimmunol text; and receives publishing royalties from Euroimmun. A. Saiz Neuroinflamm and a cervicothoracic sensory level. New MRI showed received travel funding and/or speaker honoraria from and consulted 2017;4:e364; doi: 10.1212/ a cystic-like cervical lesion and patchy signs of sub- for Bayer-Schering, Merck-Serono, Biogen Idec, Sanofi-Aventis, NXI.0000000000000364 Teva, and Novartis. Go to Neurology.org/nn for full disclosure forms. acute hemorrhage with gadolinium enhancement The Article Processing Charge was funded by the Fundació Clinic per from C7 to conus (figure 1). The follow-up MRI la Recerca Biomédica. 18 months later showed severe spinal cord atrophy This is an open access article distributed under the terms of the below C7 with hemosiderin deposit (figure 2). The Creative Commons Attribution-NonCommercial-NoDerivatives Li- cense 4.0 (CC BY-NC-ND), which permits downloading and shar- MRI findings, and the devastating evolution, mirror ing the work provided it is properly cited. The work cannot be the pathologic features described in acute ascending changed in any way or used commercially without permission from necrotizing myelitis.1 the journal.

*These authors contributed equally to this work. Received March 29, 2017. Accepted in final form April 19, 2017. From the Neurology Service (M.S., J.A., N.S.-V., Y.B., S.L., F.G., A.S.) and Radiology Department (J.B.), Hospital Clinic Correspondence to Dr. Saiz: [email protected] deBarcelona;andNeurologyService(J.S.),Althaia,Manresa,Barcelona, Spain. 1. Wiley AD, VanPatten PD, Carpenter PM, Powell HC, Author contributions: M.S., J.A., and A.S.: analysis/interpretation of Thai LJ. Acute ascending necrotizing myelopathy caused the data and drafting and reviewing the manuscript. J.B., N.S.-V., by herpes simplex virus type 2. Neurology 1987;37: J.S., Y.B., S.L., and F.G.: acquisition of data and revising the 1791–1794.

At onset, sagittal MRI shows an extensive cervicothoracic T2 hyperintensity (A.a and A.b). After 3 weeks of therapy, the extensive lesion persists (B.a and B.b) and shows a central cystic-like lesion at C4 level (B.a and B.c, arrows), suggestive signs of subacute hemorrhage (T1 hyperintensity; B.c, arrowhead), and gadolinium enhancement from C7 level (B.d, arrow).

2 Neurology: Neuroimmunology & Neuroinflammation Figure 2 Severe spinal cord atrophy at long term

Follow-up MRI 18 months later reveals a longitudinally severe spinal cord atrophy from C7 to conus (A and B, arrows). Note the filiform aspect of the spinal cord (axial T2-weighted image; D, arrowhead), and the hypointensity in T2 (B), and T2 gradient echo–weighted (C) images suggestive of hemosiderin deposit.

Neurology: Neuroimmunology & Neuroinflammation 3 Clinical/Scientific Notes

Joseph J. Sabatino, Jr., ACUTE LIVER INJURY IN A GLATOPA-TREATED central perivenulitis (figure, B and C). The bile ducts MD, PhD PATIENT WITH MS were intact, and there was no significant ductular Neil J. Mehta, MD reaction. Steatosis, cholestasis, and fibrosis were not Sanjay Kakar, MD present. The patient was treated with IV corticoste- Hepatotoxicity is rarely associated with glatiramer ace- Scott S. Zamvil, MD, roids followed by a several month taper of prednisone tate (GA) treatment of relapsing MS. Here, we report PhD for possible autoimmune hepatitis. The patient’s liver a case of acute liver injury associated with generic GA Bruce A.C. Cree, MD, function tests normalized within 2 months. (Glatopa, manufactured by Sandoz, a Novartis com- PhD, MAS pany, Holzkirchen, Upper Bavaria, Germany). Discussion. GA is a synthetic random copolymer of 4 amino acids (glutamic acid, lysine, alanine, and tyrosine) Neurol Neuroimmunol Neuroinflamm Case report. A 36-year-old woman presented with approved for treatment of relapsing-remitting 2017;4:e368; doi: 10.1212/ unilateral optic neuritis from which she recovered. MS. With more than 2,000,000 patient-years of NXI.0000000000000368 Optic neuritis recurred at age 42. Brain MRI showed exposure to GA, there are 13 published cases of hepa- multiple T2-hyperintense white matter lesions and totoxicity (table e-1 at Neurology.org/nn). Previously a gadolinium–diethylenetriamine penta-acetic acid– reported cases are notable for prior treatment with enhancing lesion. MS was diagnosed. The patient was interferon b, concomitant use of other potentially treated with methylprednisolone 1 g/daily IV for 5 hepatotoxic drugs, and symptom onset 1–8months days. Concomitant medications included norethin- after initiating GA. We report a rare case of acute liver drone/ethinyl estradiol with iron, vitamin D, calcium, injury in the setting of generic GA use. Liver injury and vitamin B12. The patient did not take supple- associated with GA presents with a hepatocellular ments or abuse alcohol. Approximately 3 weeks later, injury pattern without hyperglobulinemia, although the patient was started on generic GA. Influenza and autoimmune hepatitis can occur (Supplemental meningococcal oligosaccharide vaccines were admin- Table). Given our patient’s negative autoimmune istered at the time she began GA. markers, lack of plasma cells on biopsy, and reso- Because of symptoms of injection pain, fatigue, lution of liver injury on a rapid prednisone taper, it nausea, and constipation, the patient discontinued is unlikely that this patient had autoimmune hepa- GA after 13 daily injections. The following day, the titis. The administration of influenza and menin- patient developed worsening nausea with vomiting, gococcal vaccines seems unlikely to be contributory dyspnea with exertion, anorexia, fatigue, dark urine, because neither vaccine is hepatotoxic. Seven years and jaundice. The patient was found to have elevated priortothediagnosisofMS,ourpatienthadahis- liver function tests (aspartate aminotransferase 5 tory of transaminitis, which was attributed to a viral 1,188 U/L, alanine transaminase 5 1,879 U/L, alka- etiology (associated with serologic evidence of prior line phosphatase 5 177 U/L, and total bilirubin 5 hepatitis A and hepatitis E virus exposure), followed 3.6 mg/dL) compared with normal values just 1 by many years of normal liver enzymes. month prior (figure, A). The following laboratory GA is classified as a nonbiological complex drug values were negative or normal: HAV IgM, HBsAg, (NBCD) whose composition and in vivo activity are HBcAg IgG, HCV RNA, HCV antibody, HEV IgM, highly dependent on manufacturing processes.1,2 EBV IgM, CMV IgM, HIV, HSV PCR, urine Systemic toxicities are extremely rare with GA, toxicology screen, acetaminophen levels, ANA, and and there are no guidelines for drug safety monitor- anti–smooth muscle antibody. HAV IgG, HBsAb, ing. In prior clinical trials of Copaxone and generic – Supplemental data HEV IgG, and EBV IgG were positive. Twenty-four GA, the incidence of liver dysfunction was equiva- at Neurology.org/nn hour urine copper (193 mg/24 h [normal range 15– lent to that of placebo-treated patients.3 Since 60 mg/24 h]) was presumed to be elevated secondary approval, over 300 cases of Copaxone-associated to acute liver injury. liver-related abnormalities were reported to the A liver biopsy showed severe portal, interface and United States Food and Drug Administration panacinar, lymphocyte-predominant inflammation, with (US-FDA).Inanimalstudies,transaminaseeleva- confluent necrosis, numerous apoptotic hepatocytes, and tions were observed with chronic high-dose GA.

Liver function tests (panel A) were performed from February 17, 2016 (32 days prior to glatiramer acetate [GA]) to June 13, 2016 (102 days following the start of GA treatment). The gray rectangle represents the 13 days of GA treatment starting at day 0. Alanine transaminase (ALT) (normal range 10–30 U/L) and aspartate aminotransferase (AST) (normal range 6–29 U/L) are depicted on the left y-axis and alkaline phosphatase (alk phos [normal range 33–115 U/L]) on the right y-axis. Liver biopsy (indicated by “Bx” on panel A) on day 20 shows dense portal lymphocytic inflammation (panel B) with interface activity and normal bile ducts and shows lobular lymphocytic inflammation with confluent necrosis (panel C) (hematoxylin and eosin stain, 2003). This histologic picture along with the clinical presentation and temporal profile is consistent with drug-induced liver injury.

Significant hepatotoxicity, nephropathy, and skin possible liver dysfunction and other adverse effects reactions also occurred with protiramer (TV- may be warranted.

5010), a glatiramoid with a higher molecular weight From the Multiple Sclerosis Center (J.J.S., S.S.Z., B.A.C.C.), 4 of the same molar ratio of amino acids as GA. Department of Neurology, Division of Gastroenterology (N.J.M.), Glatopa was approved as a GA biosimilar com- Department of Medicine, and GI-Hepatobiliary Pathology Service pound in 2015 without a requirement for proof of (S.K.), Department of Pathology, University of California San Francisco. either efficacy or safety in clinical trials. Sandoz dem- ’ Author contributions: Dr. Sabatino contributed to interpretation of onstrated Glatopa s equivalence with Copaxone by the data and drafting and revision of the manuscript. Dr. Mehta similarities in chemistry, polymerization, biological, and Dr. Kakar contributed to interpretation of the data and revi- and immunologic properties.5,6 In a study by the US- sion of the manuscript. Dr. Zamvil and Dr. Cree contributed to interpretation of the data and drafting and revision of the FDA using 3 different analytic measures, distinct manuscript. physicochemical differences were found between Co- Acknowledgment: The authors are grateful for Collin Spencer’s help paxone and commercially available copolymer-1.7 in preparing the figure. Teva Pharmaceuticals also found differences in charge Study funding: No targeted funding reported. distribution, molecular density, monomolecular size, Disclosure: J.J. Sabatino received research support from R25 and the existence of a novel polypeptide group in NS00680, NMSS. N.J. Mehta reports no disclosures. S. Kakar 2 received publishing royalties from Elsevier and Springer Publishing. Glatopa compared with Copaxone. Therefore, dif- S.S. Zamvil served on the data safety monitoring board for BioMS, ferences in manufacturing between these NBCDs Teva Pharmaceuticals, Eli Lilly and Com; is a member of the could cause different adverse event profiles. Given clinical advisory board for Myelin Repair Foundation; is deputy ® that generic GA is only recently available for clinical editor for Neurology Neuroimmunology & Neuroinflammation; has a patent pending for Aquaporin-4 peptides and methods for use in MS and is the likely cause of acute liver injury using the same; has consulted for Biogen, Teva, EMD Serono, in the present case report, heightened awareness of Genzyme, Novartis, and Roche; received research support from

2 Neurology: Neuroimmunology & Neuroinflammation Biogen,Teva,NIH,NMSS,andAlexanderM.andJuneL.Mai- Guidance for Glatiramer Acetate Injection. Available at: sin Foundation. B.A.C. Cree consulted for AbbVie, Biogen, EMD regulations.gov/document?D5FDA-2007-D-0369-0395. Serono, Novartis, Sanofi Genzyme, Shire and received research Accessed December 6, 2016. support from Acorda, Celgene, Hoffman La Roche, MedImmune, 3. Carter NJ, Keating GM. Glatiramer acetate. A review of its Novartis, and Teva. Go to Neurology.org/nn for full disclosure use in relapsing-remitting multiple sclerosis and in delaying forms. The Article Processing Charge was funded by the authors. the onset of clinically definite multiple sclerosis. Drugs This is an open access article distributed under the terms of the 2010;70:1545–1577. Creative Commons Attribution-NonCommercial-NoDerivatives Li- 4. Ramot Y, Rosenstock M, Klinger E, Bursztyn D, Nyska A, cense 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be Shinar D. Comparative long-term preclinical safety evalua- changed in any way or used commercially without permission from tion of two glatiramoid compounds (glatiramer acetate, the journal. Copaxone, and TV-5010, protiramer) in rats and monkeys. Toxicol Pathol 2012;40:40–54. Received March 21, 2017. Accepted in final form May 1, 2017. 5. Anderson J, Bell C, Bishop J, et al. Demonstration of equiv- Correspondence to Dr. Cree: [email protected] alence of a generic glatiramer acetate (Glatopa). J Neurol Sci 2015;359:24–34. 1. Cohen J, Belova A, Selmaj K, et al. Equivalence of generic 6. D’Alessandro JS, Duffner J, Pradines J, et al. Equivalent glatiramer acetate in multiple sclerosis: a randomized clinical gene expression profiles between Glatopa and Copaxone. trial. JAMA Neurol 2015;72:1433–1441. PLoS One 2015;10:1–19. 2. Levin PS. Docket No. FDA-2007-D-0369; Product- 7. Rogstad S, Pang E, Sommers C, et al. Modern analytics for Specific Bioequivalence Recommendations; Draft and synthetically derived complex drug substances: NMR, Revised Draft Guidances for Industry; Comments of Teva AFFF-MALS, and MS tests for glatiramer acetate. Anal Pharmaceuticals Industries Ltd. to the Draft Bioequivalence Bioanal Chem 2015;407:8647–8659.

Neurology: Neuroimmunology & Neuroinflammation 3 Clinical/Scientific Notes

David Cuendet, MD AN UNUSUAL CASE OF MILIARY PML-IRIS IN AN in CSF at 965 copies/mL. All the following analyses Rossella Sarro, MD HIV1 PATIENT performed in the CSF were negative or normal: cryp- Laurent Merz, MD tococcal antigen, bacterial, fungal, and acid-fast cul- Mayté Castro Jiménez, tures; PCR for Toxoplasma gondii;CMV,HSV, A 45-year-old female patient recently diagnosed with MD 1 HIV infection (CD4 cell count 64 cells/mm3 and VZV, and Tropheryma whipplei;VDRL;andcytology. Philippe Maeder, MD plasma HIV viral load 1.4 3 106 copies/mL at diag- CSF HIV viral load was not measured. Stool specimen Renaud Du Pasquier, MD nosis) started antiretroviral therapy (ART). At the analysis revealed no Entamoeba histolytica, Giardia lam- Pinelopi Tsouni, MD time, she presented with gastric cryptosporidiosis blia, microsporidies, or Cryptosporidium. and an oral candidiasis, which subsided rapidly on The diagnosis of progressive multifocal encepha- Neurol Neuroimmunol lopathy (PML) was considered. However, because Neuroinflamm treatment. One month after ART initiation, the 2017;4:e370; doi: 10.1212/ patient noticed progressive generalized slowness, the radiologic findings were absolutely not typical NXI.0000000000000370 speech difficulty, and a slight weakness of her right for this disease, a stereotactic brain biopsy was per- arm and leg. She also reported headache, generalized formed. This examination showed diffuse and dense fatigue, night sweats, and weight loss. She was admit- perivascular inflammatory infiltrate (figure, E), scat- ted to the hospital 4 weeks after symptoms onset. The tered atypical astrocytes with hyperchromatic nuclei neurologic examination revealed psychomotor slow- (figure, F, arrowheads), and demyelinization (figure, ing, mild memory and executive dysfunction, men- G). Immunohistochemistry showed CD3 underlin- 1 ingeal signs, dysarthria, mild lower-limb ataxia, and ing dense T-lymphocyte infiltration, mostly CD8 – right faciobrachial weakness (Medical Research and glial cells staining positive for JCV (figure, H Council grade 4). Generalized adenopathy and wast- J). The diagnosis of PML in the setting of immune ing syndrome were also identified. Brain MRI showed reconstitution inflammatory syndrome (IRIS) was es- a pattern of multiple miliary-enhancing parenchymal tablished, and the patient was treated with CCR5 nodules (1–4 mm in diameter) more marked in the antagonist maraviroc with a favorable outcome. brain stem and basal ganglia (figure, A–D). ART treatment was continued. CD41 cell count was 89 cells/mm3 (11.8%), and PML is typically associated with subcortical conflu- plasma HIV viral load had dramatically dropped at 47 ent nonenhancing lesions. However, rarely, PML may “ ” 1 copies/mL. Serological testing revealed past exposure have a miliary presentation. This form is character- to Toxoplasma gondii, Epstein-Barr virus (EBV), and ized by enhancing punctuate lesions on brain MRI, cytomegalovirus (CMV). Cryptococcus neoformans which correspond, histopathologically, to a perivascu- 1 antigen and serologies for rubella virus, herpes sim- lar inflammation composed mostly of CD8 T cells in plex viruses (HSV) 1 and 2, varicella zoster virus contact with JCV-infected glial cells.2 Such a punctate (VZV), Borrelia sp, Bartonella henselae, and hepatitis pattern has been described in natalizumab-associated B and C were negative. Serum Treponema pallidum PML,3 but may be also seen in PML classically asso- hemagglutination assay and venereal disease research ciated with some sort of immunosuppression, such as laboratory test (VDRL) were nonreactive. Peripheral neurosarcoidosis or hematologic diseases.1 A similar blood flow cytometry showed no malignant cells. radiologic pattern was also reported in a patient with Antinuclear antibody (anticytoplasm) titers were pos- myelofibrosis exhibiting a rare mutation of the JCV itive (1/640), but C3 and C4 complement titers were genome.4 To the best of our knowledge, only one within normal limits. Erythrocyte sedimentation rate other case of miliary HIV-related PML has been re- was 110, and C-reactive protein was inferior to 2 ported so far in a patient who was initially diagnosed mmol/L. with EBV encephalitis.5 CSF examination did not reveal any leukocyte or Our patient and the review of the literature sug- erythrocyte; the protein content was within normal gest that miliary PML may occur at an early stage limits (431 mg/L); and the CSF-to-serum glucose ratio of JCV brain infection preceding the emergence of was 0.7. There was no intrathecal IgG synthesis. A typical PML lesions.1 In our case, the presence of quantitative PCR for JC virus (JCV) DNA was positive a strong JCV-specific cellular immune response

T2 axial (A) showing diffuse hyperintensities in basal ganglia and centrum semiovale on both sides, T1 transverse (B) and sagittal (C and D) showing dilated enhancing perivascular spaces after gadolinium enhancement. Neuropathology: brain parenchyma showing diffuse and dense perivascular inflammatory infiltrate (E, hematoxylin and eosin [H&E]), scattered atypical astrocytes with hyperchromatic nuclei (F, H&E, arrowheads), and demyelinization (G, Luxol fast blue). CD3 (H, immunohistochemistry [IHC]) underlines dense T-lymphocyte infiltration, mostly CD81 (I, IHC). JC virus–infected oligodendrocytes (J, IHC, arrowheads) and bizarre astrocytes (J, IHC, arrow).

suggests that miliary PML represents a very early stage License 4.0 (CC BY-NC-ND), which permits downloading and of unmasking PML-IRIS. In conclusion, this case sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from illustrates that a miliary or punctate pattern on brain the journal. MRI should prompt clinicians to rule out PML. Received March 22, 2017. Accepted in final form April 10, 2017. From the Division of Neurology (D.C., M.C.J., R.D.P., P.T.), Department of Clinical Neurosciences, Department of Pathology Correspondence to Dr. Tsouni: [email protected] (R.S.), Department of Infectious Diseases (L.M.), and Department of Radiology (P.M.), Lausanne University Hospital (CHUV), 1. Hodel J, Darchis C, Outteryck O, et al. Punctate pattern: Switzerland. a promising imaging marker for the diagnosis of – Author contributions: D. Cuendet: acquired, analyzed, and drafted natalizumab-associated PML. Neurology 2016;66:1516 the content of this clinical reasoning including medical writing. 1521. P. Maeder and R. Sarro: acquired, analyzed, and drafted the content 2. Wattjes MP, Verhoeff L, Zentjens W, et al. Punctate lesion of the figure and their legends. L. Merz and M.C. Jiménez: revised pattern suggestive of perivascular inflammation in acute the content of this manuscript. R. Du Pasquier: revised the manu- natalizumab-associated progressive multifocal leukoence- script for content and did the supervision of this case report. P. phalopathy: productive JC virus infection or preclinical Tsouni: acquired, analyzed, and revised the content of this manu- PML-IRIS manifestation? J Neurol Neurosurg Psychiatry script and did the supervision of this case report. 2013;84:1176–1177. Study funding: No targeted funding reported. 3. Yousry TA, Pelletier D, Cadavid D, et al. Magnetic reso- Disclosure: D. Cuendet, R. Sarro, L. Merz, M.C. Jiménez, and nance imaging pattern in natalizumab-associated progressive P. Maeder report no disclosures. R. Du Pasquier served on the scientific multifocal encephalopathy. Ann Neurol 2012;72:779–787. advisory board for Biogen, Genzyme, Novartis; received travel funding 4. Tallantyre EC, Paine SM, Sharp CP, Lowe JS, Gran B. and/or speaker honoraria from Sanofi and Genzyme; served on the editorial board for Journal of Neurovirology; and received research Atypical progressive multifocal leukoencephalopathy associ- support from Swiss National Foundation, Swiss Society for Multiple ated with an unusual JC Polyomavirus mutation. Arch Neu- – Sclerosis, and Novartis. P. Tsouni reports no disclosures. Go to rol 2009;66:1021 1024. Neurology.org/nn for full disclosure forms. The Article Processing 5. Gheuens S, Buggy BP, Tlomak W, Wüthrich C, Koralnik Charge was funded by the authors. IJ. A game of viral hide and seek: miliary PML masquerad- This is an open access article distributed under the terms of the ing as EBV encephalitis in an HIV1 patient. Clin Neurol Creative Commons Attribution-NonCommercial-NoDerivatives Neurosurg 2013;115:1861–1863.

2 Neurology: Neuroimmunology & Neuroinflammation Clinical/Scientific Notes

Avi Gadoth, MD DACRYSTIC SEIZURES: A CRY FOR HELP manuscript. Sean J. Pittock: data acquisition and critical revision of manuscript. Jaysingh Singh, MD Study funding: No targeted funding reported. Jeffrey W. Britton, MD A 69-year-old man with a history of myocardial Disclosure: A. Gadoth and J. Singh report no disclosures. J.W. Britton Eoin P. Flanagan, infarction presented with crying-like spells. MRI received research support from Grifols and Mr. and Mrs. David MBBCh Hawk charitable gift for epilepsy research. E.P. Flanagan reports demonstrated abnormal signal in the left mesio- Sean J. Pittock, MD no disclosures. S.J. Pittock and Mayo Clinic have a financial interest temporal lobe, and he was diagnosed with stroke. in patents that relate to functional AQPf/NMO-IgG assays and NMO-IgG as a cancer market. S.J. Pittock consulted for Alexion Neurol Neuroimmunol Over the following 1 year, he developed progres- and MedImmune and received research support from Grifols, Med- Neuroinflamm sive cognitive decline, slow gait, masked facies, Immune, and Alexion, RO1 NS-65829-91. Go to Neurology.org/nn 2017;4:e372; doi: 10.1212/ hypophonic voice, and brief facial and upper NXI.0000000000000372 for full disclosure forms. The Article Processing Charge was funded by extremity spasms (8 per hour) often followed by the authors. brief crying spells lacking emotion. Spell semiol- This is an open access article distributed under the terms of the ogy was consistent with faciobrachial dystonic Creative Commons Attribution-NonCommercial-NoDerivatives Li- 1 2 cense 4.0 (CC BY-NC-ND), which permits downloading and shar- seizures and dacrystic seizures (video at ing the work provided it is properly cited. The work cannot be Neurology.org/nn). Leucine-rich glioma-inactivated-1 changed in any way or used commercially without permission from (LGI1) autoantibodies, detected in serum, led to a diag- the journal. nosis of anti-LGI1 encephalitis. This is the first Received March 28, 2017. Accepted in final form April 23, 2017. report of dacrystic seizures accompanying anti-LGI1 Correspondence to Dr. Gadoth: [email protected] encephalitis.

From the Department of Laboratory Medicine and Pathology (A.G., 1. Irani SR, Michell AW, Lang B, et al. Faciobrachial dystonic E.P.F., S.J.P.) and Department of Neurology (A.G., J.S., J.W.B., E. seizures precede Lgi1 antibody limbic encephalitis. Ann – P.F., S.J.P.), Mayo Clinic, Rochester, MN. Neurol 2011;69:892 900. Author contributions: Avi Gadoth: data acquisition. Jaysingh 2. Blumberg J, Fernandez IS, Vendrame M, et al. Dacrystic Singh: data acquisition, EEG interpretation, and review of seizures: demographic, semiologic, and etiologic insights manuscript. Jeffrey W. Britton: EEG interpretation and review from a multicenter study in long-term video-EEG monitor- of manuscript. Eoin P. Flanagan: critical revision of ing units. Epilepsia 2012;53:1810–1819.

Supplemental data at Neurology.org/nn