ARTICLE OPEN ACCESS CLASS OF EVIDENCE

Vascular safety of erenumab for migrainepreventionDavid Kudrow MD Julio Pascual MD PhD Paul K Winner DO David W Dodick MD Stewart J Tepper MD

Uwe Reuter MD Frank Hong MD PhD Jan Klatt MD Feng Zhang MS Sunfa Cheng MD

Hernan Picard MD PhD Osa Eisele MD MPH Julie Wang DPM Jonathan N Latham PharmD and

Daniel D Mikol MD PhD

Neurologyreg 202094e497-e510 doi101212WNL0000000000008743

Correspondence

Dr Kudrow

dbkudrowearthlinknet

AbstractObjectiveTo examine the cardiovascular cerebrovascular and peripheral vascular safety of erenumabacross migraine prevention studies

MethodsVascular adverse events (AEs) and blood pressure data were integrated across 4 double-blindplacebo-controlled studies of erenumab and their open-label extensions in patients with chronicor episodic migraine Subgroup analyses were conducted by acute migraine-specific medicationuse and number of vascular risk factors at baseline Standardized search terms were used toidentify vascular AEs (cardiovascular cerebrovascular or peripheral) An independent com-mittee adjudicated whether targeted events were vascular in origin

ResultsIn placebo-controlled studies 2443 patients received placebo (n = 1043) erenumab 70 mg(n = 893) or erenumab 140 mg (n = 507) subcutaneously once monthly Regardless of acutemigraine-specificmedication use or vascular risk factors at baseline AE incidence was similar acrossthe placebo and erenumab treatment groups Hypertension AEswere reported for 09 (placebo)08 (erenumab 70 mg) and 02 (erenumab 140 mg) of patients Vascular AEs which weresimilar across double-blind and open-label treatment generally were confounded with plausiblealternative etiologies In 18 patients with events reviewed by the independent committee 4 eventswere positively adjudicated as cardiovascular in origin 2 deaths and 2 vascular events All 4positively adjudicated cardiovascular events occurred during open-label erenumab treatment

ConclusionSelective blockade of the canonical calcitonin gene-related peptide receptor with erenumab formigraine prevention had a vascular safety profile comparable to that of placebo over 12 weekswith no increased emergence of events over time Further study of long-term safety of ere-numab in patients with migraine is needed

Clinicaltrialsgov identifiersNCT02066415 NCT02456740 NCT01952574 NCT02483585 NCT02174861 andNCT01723514

Classification of evidenceThis analysis provides Class II evidence that for patients with migraine erenumab does notincrease the risk of vascular AEs

MORE ONLINE

Class of EvidenceCriteria for ratingtherapeutic and diagnosticstudies

NPuborgcoe

From California Medical Clinic for Headache (DK) Santa Monica CA Department of Neurology (JP) University Hospital Marques de Valdecilla and IDIVAL Santander SpainPremiere Research Institute (PKW) Nova Southeastern University West Palm Beach FL Department of Neurology (DWD) Mayo Clinic Phoenix AZ Department of Neurology(SJT) Geisel School of Medicine at Dartmouth Hanover NH Department of Neurology (UR) Charite Universitatsmedizin Berlin Germany Novartis East Hanover NJ (FH) andBasel Switzerland (JK) Amgen Inc (FZ SC HP OE JW DDM) Thousand Oaks CA and PharmaScribe LLC (JNL) Atlanta GA

Go to NeurologyorgN for full disclosures Funding information and disclosures deemed relevant by the authors if any are provided at the end of the article

The Article Processing Charge was funded by Amgen Inc

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 40 (CC BY-NC-ND) which permits downloadingand sharing the work provided it is properly cited The work cannot be changed in any way or used commercially without permission from the journal

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology e497

Migraine affects over 1 billion people worldwide1 and15ndash20 of Americans2 More than 25 of adults with mi-graine are candidates for preventive therapy3 but fewer than30 of candidates receive it45 Over 80 of patients withchronic migraine discontinue oral preventive therapy withinthe first year6 There is an unmet need for an effective well-tolerated therapy to prevent migraine

Calcitonin gene-related peptide (CGRP) plays a key role inmigraine pathophysiology78 Monoclonal antibodies haverecently been developed that bind to either CGRP or thecanonical CGRP receptor to prevent migraine BecauseCGRP can mediate vasodilation9 inhibition of the CGRPpathway could theoretically attenuate compensatory vasodi-lation during ischemic conditions but the relative importanceof the CGRP receptor pathway compared with other vaso-dilatory pathways during ischemia (eg myocardial) has notbeen established9ndash11 In addition patients with migraine havean increased risk of vascular events including stroke andmyocardial ischemia1213 and acute migraine-specific medi-cations such as triptans and ergotamine have known vaso-constrictive effects1415 Thus it is important to examinevascular safety particularly over the longer term in patientstreated with therapies that block the effects of CGRP

Erenumab (in the United States erenumab-aooe) is a fullyhuman monoclonal antibody that specifically targets andblocks the canonical CGRP receptor to prevent migraine16

Administered subcutaneously once monthly erenumab hasbeen shown to be effective for migraine prevention17ndash20 Thispooled analysis of vascular safety in clinical studies of erenu-mab for migraine prevention included vascular (cardiovas-cular cerebrovascular or peripheral) adverse events (AEs)that were reported by investigators events that were adjudi-cated across all studies by an independent committee ofclinical experts and a pooled analysis of blood pressure (BP)measurements This report also includes results of a 24-hourambulatory BP monitoring study that assessed the potentialcardiovascular effects of erenumab in healthy controls

MethodsThe primary research questions for this pooled analysis wereto examine if the rates of vascular (cardiovascular or cere-brovascular) AEs were higher in the erenumab group vs theplacebo group of controlled clinical studies both overall andin subgroups of patients at a higher risk of vascular AEs as wellas the effect of erenumab treatment on BP This analysis

provides Class II evidence that for patients with migraineerenumab does not increase the risk of vascular AEs

DesignInformation about AEs was collected as reported by the pa-tient either spontaneously or in response to the investigatorrsquosnondirected questioning per standard procedures in clinicaltrials Regulatory authorities in the regions where the trialswere conducted reviewed study protocols including AE datacollection methods before implementation Reported AEsand BP results were integrated for 12 weeks of double-blindtreatment across 4 placebo-controlled migraine preventionstudies of erenumab administered subcutaneously oncemonthly For long-term safety AE data were integrated forany exposure to erenumab in these 4 studies and their open-label extensions

The pivotal studies testing the doses of erenumab that areapproved for migraine prevention (70 mg and 140 mg oncemonthly) were NCT02066415 (n = 667) which enrolledpatients with chronic migraine17 and NCT02456740 (Studyto Evaluate the Efficacy and Safety of Erenumab [AMG 334]in Migraine Prevention [STRIVE] n = 955) which enrolledpatients with episodic migraine18 The supportive studiesNCT01952574 (n = 483)19 and NCT02483585 (Study toEvaluate the Efficacy and Safety of Erenumab [AMG 334]Compared to Placebo in Migraine Prevention [ARISE] n =577)20 also enrolled patients with episodic migraine but inthe placebo-controlled periods of these studies 70 mg was thehighest dose tested Safety data for erenumab doses less than70 mg which were investigated in the placebo-controlledperiod of supportive study NCT01952574 but are not ap-proved for migraine prevention were excluded from this in-tegrated safety analysis All but 1 of the studies had a 12-weekplacebo-controlled period the episodic migraine STRIVEstudy had a 24-week placebo-controlled period Patientsassigned to an erenumab arm received the same dose (70 or140 mg) throughout double-blind treatment

Patients completing the double-blind placebo-controlledperiods were continued in an active dose-blinded treatmentperiod (STRIVE study) or open-label erenumab treatment(all other studies) For the episodic migraine studies fol-lowing completion of the double-blind placebo-controlledperiod the active treatment or open-label treatment period waspart of the main study protocol In NCT01952574 open-labelerenumab is being maintained for up to 256 weeks (5 years)21

Patients who completed the chronic migraine study could entera separate open-label extension study (NCT02174861)

GlossaryAE = adverse event ARISE = Study to Evaluate the Efficacy and Safety of Erenumab (AMG 334) Compared to Placebo inMigraine Prevention BP = blood pressure CGRP = calcitonin gene-related peptide DBP = diastolic blood pressure SBP =systolic blood pressure MedDRA = Medical Dictionary for Regulatory Activities NSAID = nonsteroidal anti-inflammatorydrug STRIVE = Study to Evaluate the Efficacy and Safety of Erenumab (AMG 334) in Migraine Prevention

e498 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

During the active treatment or open-label periods patientsreceived only 70mg (ARISE study) or 70mg or 140mg (otherstudies)

Ambulatory BPwas assessed in a phase 1 study (NCT01723514)that enrolled 32 healthy controls and 16 patients with migrainewho received erenumab (70 or 140 mg) or placebo sub-cutaneously once monthly for 12 weeks22

Standard protocol approvals registrationsand patient consentsAn institutional review board or independentethics committee approved each study in this analysisThe studies included in this analysis were registeredat clinicaltrialsgov NCT02066415 NCT02456740NCT01952574 NCT02483585 NCT02174861 andNCT01723514 Patients provided written informed con-sent to participate in each study

Eligibility criteriaEach of the migraine prevention studies had an initialscreening period (up to 3 weeks) followed by a 4-weekprospective baseline period to confirm study eligibility beforethe patient was enrolled and randomized Key inclusion cri-teria at screening were age 18 years or older (up to 60 or 65years) and a history of migraine with or without aura for atleast 12 months The chronic migraine study includedpatients with at least 15 headache days per month during thebaseline period of which at least 8 were migraine days Theepisodic migraine studies included patients with fewer than 15headache days per month during the baseline period of whichat least 4 were migraine days

In the episodic migraine studies use of medications for acuteheadache before screening was restricted as follows acutemigraine-specific medications (triptan or ergot) were limitedto fewer than 10 days per month analgesics or nonsteroidalanti-inflammatory drugs (NSAIDs) to fewer than 15 days permonth and opioids or butalbital-containing medications tofewer than 4 days per month In the chronic migraine studythere was no restriction for prior use of acutemigraine-specificmedications analgesics or NSAIDs however in the prior 3months patients could not have used opioids onmore than 12days or butalbital on more than 6 days In each migraineprevention study after randomization patients could useacute migraine-specific medications analgesics or NSAIDs asneeded to treat acute migraine

Patients were excluded from the migraine prevention studiesif they had myocardial infarction stroke TIA unstable an-gina or coronary artery bypass surgery or another re-vascularization procedure within 12 months prior toscreening Medically stable patients who had any of theselisted conditions more than 12 months prior to screeningcould participate in the studies Patients were excluded fromstudies NCT02066415 and NCT01952574 if they hadpoorly controlled hypertension (systolic BP [SBP] above

150ndash160 mm Hg or diastolic BP [DBP] above 90ndash100 mmHg) and patients with any unstable medical condition wereexcluded from each study Healthy controls in the ambula-tory BP study had no history of hypertension hypotensionor vascular disease

Safety assessmentsAEs that were new or worsened after the first dose of studytreatment were recorded through 12 weeks after the last dosein studies with doses up to 70 mg (the ARISE study and studyNCT01952574 before protocol amendment to include higherdoses) and 16 weeks after the last dose in studies with dosesup to 140 mg For each AE the Medical Dictionary forRegulatory Activities (MedDRA) was used to determine thepreferred term and system organ class AE severity was gradedwith Common Terminology Criteria for Adverse Events(CTCAE) criteria version 403 A serious AE was defined asan AE that met at least 1 of the following criteria fatal life-threatening required inpatient hospitalization or pro-longation of existing hospitalization resulted in persistent orsignificant disabilityincapacity congenital anomalybirthdefect or another medically important serious event

Because of the acknowledged interest in the safety of this newclass of medication an independent committee was estab-lished to review potential vascular events The adjudicatorswho were clinical experts in cardiovascular and cerebrovas-cular disease adjudicated each potential event based on pre-specified definitions Events that were referred to theadjudicators included the following death acute myocardialinfarction or hospitalization for unstable angina event non-fatal stroke or TIA coronary revascularization procedurehospitalization for hypertension hospitalization for peripheralartery disease event or revascularization procedure for pe-ripheral artery disease Adjudicators were blinded to treat-ment assignment throughout the adjudication process

In each study resting BP was measured at monthly studyvisits before each administration of study treatment Thephase 1 study also monitored BP continuously for 24 hoursstarting at 900 AM on day minus2 (ie 2 days before the first doseof study treatment) and on days 8 36 and 64 (ie 7 days aftereach dose of study treatment)

Statistical analysisAnalyses of AEs were conducted for 2 pools of safety dataThe ldquo12-week placebo-controlledrdquo pool was the integratedsafety dataset from 12 weeks of double-blind treatment (or inthe STRIVE study the first 12 weeks of double-blind treat-ment) The ldquoany erenumab exposurerdquo pool included AEs thatoccurred during exposure to either double-blind or open-labelerenumab treatment through the data cutoff for the 5-yearopen-label extension study (NCT01952574) or completionof the study (all other studies) Integrated analyses of AEs forthe 12-week placebo-controlled pool were conducted byassigned treatment (placebo erenumab 70 mg or erenumab140 mg) Integrated analyses for any erenumab exposure were

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e499

summarized by actual treatment received (erenumab 70 mgor erenumab 140 mg) when the AEs occurred A single pa-tient in the any erenumab exposure pool could have receivedboth erenumab 70 mg and 140 mg because of the secondrandomization for the active treatment phase in the STRIVEstudy or because of study protocol amendments inNCT02174861 and NCT01952574 that increased the open-label dose of erenumab from 70 mg to 140 mg for all patientsIf a patient experienced the same AE when he or she receivedeach dose of erenumab then this AE was counted for both70 mg and 140 mg Exposure-adjusted incidence rates of AEsfor any erenumab exposure pool were calculated by adjustingfor the length of exposure to erenumab

For the AE analyses in the 12-week placebo-controlled pool 1subgroup analysis examined the incidences of AEs separately byuse or nonuse at baseline of acute migraine-specific medi-cations (triptans or ergot) Another subgroup analysis exam-ined the incidences of AEs separately by the number of vascularrisk factors at baseline (none 1 or ge2) using 7 categories forvascular risk factors (table 1) Each subgroup analysis includedincidences of any AE any AE by severity any serious AE anyAE leading to discontinuation of study treatment and death

To analyze vascular AEs standardized MedDRA queries withnarrow search terms were used to determine the overallincidences in each treatment group of the 12-week placebo-control pool in the following categories ischemic CNS vas-cular conditions ischemic heart disease peripheral arterialdisease or hypertension The same queries were used to

analyze vascular AEs for any erenumab exposure The resultsof each search were summarized descriptively Positively ad-judicated vascular events were listed

The integrated data analysis of the 12-week double-blind poolalso summarized resting SBP and DBP at monthly study visitsTo evaluate BP in a manner that was clinically relevant theproportion of patients with an increase from baseline ofge10mmHg in DBP or ge20mmHg in SBPwas evaluated at themonthly visits For the phase 1 study of ambulatory BP hourlySBP and DBP were summarized for each treatment group

Data availabilityQualified researchers may request data from Amgen clinicalstudies Complete details are available at amgencomdatasharing

ResultsPatient dispositionDuring 12 weeks of double-blind treatment across the 4 mi-graine prevention studies 2443 patients with migraine re-ceived placebo (n = 1043) erenumab 70 mg (n = 893) orerenumab 140 mg (n = 507) Most of these patients (94erenumab 92 placebo) completed 12 weeks of double-blindtreatment In NCT01952574 an additional 213 patients whoreceived lower doses of erenumab 7mg (n = 108) or 21 mg (n= 105) were not included in the pooled safety analysis for 12weeks of double-blind treatment but could be analyzed forAEs during any exposure to erenumab if they switched to70 mg for the open-label phase

For the analysis of AEs during any exposure to erenumab2499 patients received at least 1 dose of erenumab 70 mg (n= 2128) or at least 1 dose of erenumab 140 mg (n = 1223)A single patient in any study except ARISE could have re-ceived both erenumab 70 mg and 140 mg Thus consideringall erenumab exposure during double-blind or open-labeltreatment 852 patients received both doses of erenumab1276 received only 70 mg and 371 received only 140 mgThe total exposure to either dose of erenumab was 2639patient-years

Patient characteristicsBaseline patient characteristics use of acute migraine-specificmedications and vascular risk factors were balanced across thetreatment groups (table 2) Upon entry into the respectivestudy mean patient age in each treatment group was 413ndash418years and 833ndash850 of patients were female Mean diseaseduration was approximately 21 years in each treatment groupand the mean number of monthly migraine days in eachtreatment group was 103ndash119 At baseline acute migraine-specific medications were used by 641ndash669 of patients ineach treatment group Almost all of the acute migraine-specificmedications used were triptans (lt1 were ergot-based)Medications for management of hypertension were used atbaseline by 37ndash65 of patients in each treatment group

Table 1 Determination of baseline vascular risk factorsfrom reported data

Category Description

Diabetes mellitus History of diabetes mellitus

Hypertension Medical history of hypertension or highblood pressure at screening (ge2 occasionswith systolic gt140 mm Hg or diastolicgt90 mm Hg)

Obesity Body mass index gt30 kgm2

Dyslipidemia Total cholesterol gt200mgdL or low-densitylipoprotein cholesterol gt130 mgdL or high-density lipoprotein cholesterol lt40 mgdLor triglycerides gt150 mgdL or medicalhistory of dyslipidemia

Cigarette use History of smoking

Coronary artery disease Ischemic heart disease (as reported in casereport forms for dedicated vascularmedical history or for general medicalhistory)

Cerebrovascular orperipheral artery disease

Cerebrovascular disease or peripheralartery disease (as reported in case reportforms for dedicated vascular medicalhistory or for general medical history) ormedical history of peripheralvasoconstriction necrosis or vascularinsufficiency

e500 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

Table 2 Patient demographic and clinical characteristics at baseline in 4 migraine prevention studies

Baseline characteristicPlacebo QM(n = 1043)

Erenumab 70 mg QM(n = 893)

Erenumab 140 mg QM(n = 507)

Age y 418 plusmn 111 417 plusmn 112 413 plusmn 112

Female 869 (833) 755 (845) 431 (850)

White 934 (895) 813 (910) 475 (937)

North America 544 (522) 471 (527) 248 (489)

Body mass index kgm2 268 plusmn 58 269 plusmn 58 267 plusmn 60

Disease duration y 207 plusmn 122 208 plusmn 124 205 plusmn 122

History of migraine with aura 481 (461) 429 (480) 231 (456)

History of migraine without aura 925 (887) 788 (882) 450 (888)

Monthly migraine days 111 plusmn 55 103 plusmn 50 119 plusmn 58

Acute headache medication use

Any 1024 (982) 872 (976) 498 (982)

Migraine-specifica 690 (662) 572 (641) 339 (669)

Non-migraine-specific 822 (788) 705 (789) 416 (821)

Any hypertension medication use 68 (65) 33 (37) 22 (43)

History of vascular disorders 77 (74) 59 (66) 50 (99)

No of vascular risk factors

0 310 (297) 233 (261) 148 (292)

1 423 (406) 377 (422) 199 (393)

ge2 310 (297) 283 (317) 160 (316)

Vascular risk factors

High cholesterol levelb 489 (469) 438 (490) 241 (475)

Obesityc 253 (243) 230 (258) 133 (262)

High lipid leveld 250 (240) 227 (254) 122 (241)

History of diabetes mellitus 21 (20) 17 (19) 6 (12)

History of dyslipidemia 93 (89) 69 (77) 31 (61)

History of hypertension 93 (89) 51 (57) 34 (67)

High blood pressure at screeninge 73 (70) 58 (65) 34 (67)

History of cigarette use

Current 64 (61) 67 (75) 41 (81)

Former 114 (109) 93 (104) 51 (101)

Never 430 (412) 437 (489) 227 (448)

Unknown 435 (417) 296 (331) 188 (371)

Abbreviation QM = once monthlyData are presented as mean plusmn SD or number () of patientsa Includes triptan-based or ergot-based medications (lt1 received ergot-based medications)b High cholesterol level is defined as one of the following total cholesterol gt200 mgdL low-density lipoprotein cholesterol gt130 mgdL or high-densitylipoprotein cholesterol lt40 mgdLc Obesity is defined as a body mass index gt30 kgm2d High lipid level is defined as triglycerides gt150 mgdLe High blood pressure at screening is defined as systolic blood pressure gt140mmHgor diastolic blood pressure gt90mmHgmeasured on at least 2 occasions

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e501

In each treatment group 66ndash99 of patients had a historyof a vascular disorder most commonly hypertension(34ndash51) More than 70 of patients in each treatmentgroup had at least 1 vascular risk factor at baseline (table 2)The most common vascular risk factor at baseline was dysli-pidemia (based on high cholesterol high lipids or history ofdyslipidemia) One in 4 patients was obese

Vascular adverse events in targeted searchesIn the targeted searches for vascular AEs during 12 weeks ofdouble-blind treatment investigators reported AEs of hyper-tension or diastolic hypertension for 9 (09) patients receivingplacebo 7 (08) receiving erenumab 70 mg and 1 (02)receiving erenumab 140mg (table 3) An AE of cerebral venousthrombosis was reported for 1 (02) patient in the erenumab140 mg group review of the event did not suggest a causalassociation for erenumab This female patient had multiple riskfactors for venous thrombosis age ge40 years body mass indexgt30 kgm2 a prior history of deep vein thrombosis and pul-monary embolism recurrent heparin-induced thrombocytope-nia and a recent infection (sepsis pyelonephritis and kidneyinfection about 7 weeks before the event) No other ischemicvascular event was reported in any treatment group in thepooled 12-week placebo-controlled analysis

In the targeted searches for vascular AEs during any erenumabexposure (either double-blind or open-label) the exposure-adjusted patient incidence rate for an AE of hypertensionwas 25per 100 patient-years for erenumab 70 mg and 17 per 100patient-years for erenumab 140 mg (table 3) The exposure-adjusted patient incidence rate for an AE of cerebral venousthrombosis was 02 per 100 patient-years for 140 mg includingthe AE described above for the double-blind period and anotherAE during open-label erenumab treatment The latter AE oc-curred in a 38-year-old man who experienced subdural hema-toma cerebral venous thrombosis and facial bone fractures asa result of closed head injury sustained during a fall while hikingHe was found unconscious and under the influence of Ecstasy(34-methylenedioxy-methamphetamine) Two patients had anAE of myocardial ischemia (lt01 per 100 patient-years for 70mg) both of which occurred during open-label erenumabtreatment One of these patients had demand ischemia that wasa positively adjudicated vascular event (see description in thenext section) The other patient had exercise-induced ischemiaand had taken sumatriptan 4 hours before the exercise test Thispatient received a total of 2 doses of erenumab 70 mg with thesecond dose 36 days before the event The dose of sumatriptanon the day of the event was 50 mg and the patientrsquos previousdose of sumatriptan was administered 68 days before the eventOther AEs in the searches for vascular events occurred in only 1patient each Review of the individual events did not suggesta causal association for erenumab

Adjudicated vascular eventsEighteen patients had AEs that fulfilled criteria for adjudica-tion The blinded independent clinical experts adjudicatedthe events as cardiovascular in origin for 4 patients All

occurred during open-label erenumab treatment For eachadjudicated event investigators considered the respective AEsas not related to erenumab treatment adjudicators did notassess the relationship of positively adjudicated events totreatment Details on these 4 events are given below

A 55-year-old woman in the ARISE study had AEs of multi-focal pneumococcal bacteremia life-threatening acute re-spiratory distress syndrome atrial fibrillation myocardial(demand) ischemia and elevated troponin levels on day 135during open-label treatment with erenumab 70 mg

A 54-year-old man in study NCT01952574 with a history ofhypertensive cardiovascular disease died of arteriosclerosis andhypertensive heart disease on day 650 of exposure during open-label treatment with erenumab 70 mg On autopsy there wasevidence of severe coronary atherosclerosis with presence ofalcohol and stimulants (phenylpropanolamine and norpseu-doephedrine) This case was reported previously21

A 44-year-old man in the STRIVE study with body mass indexof 30 kgm2 high triglycerides a history of abnormal ECGfindings (first-degree atrioventricular block intraventricularconduction defect and anomalies of repolarization) and mildmitral prolapse died after 12 months of exposure to erenumabduring open-label treatment with 140 mg He was asymp-tomatic for the duration of the study including during physicalexertion Postmortem genetic assessment revealed a geneticform of arrhythmogenic right ventricular cardiomyopathydysplasia which predisposes to sudden cardiac death23 Con-sistent with this diagnosis postmortem findings included hy-perplastic sclerosis of coronary arteries fat infiltration of theright heart chamber musculature hypertrophy of the left car-diac musculature and dilation of both cardiac chambers Thedirect cause of death was attributed to heart failure

A 64-year-old man in study NCT02174861 with a history ofhypertension and migraine with aura had monocular visualblurring lasting 2 minutes on day 209 of exposure duringopen-label treatment with erenumab 140 mg (10 weeks afterlast dose) The event was reported as an AE of TIA

AEs by use of acute migraine-specific medicationsPatients who used acute migraine-specific medications atbaseline had similar incidences of AEs and serious AEs acrossthe placebo and erenumab treatment groups (table 4) Overallpatients using acutemigraine-specificmedications had a slightlyhigher incidence of serious AEs compared with nonusers ofacute migraine-specific medications There was no evidence ofan increased incidence of vascular AEs or serious vascular AEsamong patients using acute migraine-specific medications

AEs by number of vascular risk factorsat baselineRegardless of the number of vascular risk factors at baselineincidences of AEs were similar across the placebo and

e502 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

erenumab treatment groups (table 5) In the subgroup ofpatients with at least 2 vascular risk factors at baseline therewere slightly higher incidences of AEs or serious AEs com-pared with the subgroups of patients with no risk factor or 1risk factor at baseline but incidences were balanced across theplacebo and erenumab groups There was no evidence of anincreased incidence of vascular AEs or serious vascular AEsamong patients with vascular risk factors at baseline

Blood pressure changesIn the integrated analysis of the 4 migraine prevention studiesresting BP readings were similar across the treatment groupsand across the study visits (figure 1) Mean (SD) change inSBP from baseline to week 12 was minus04 (96) mm Hg in theplacebo group minus05 (100) mm Hg in the erenumab 70 mggroup and minus10 (98) mmHg in the erenumab 140 mg groupMean (SD) change in DBP from baseline to week 12 was minus06

Table 3 Treatment-emergent vascular adverse events in migraine prevention studies

CategoryPreferred term

During 12 weeks of double-blind treatment During any exposure to erenumab

Placebo QM(n = 1043)

Erenumab70 mg QM(n = 893)

Erenumab140 mg QM(n = 507)

70 mg QM(n = 2128)

140 mg QM(n = 1223)

Ischemic CNS vascular conditions 0 (00) 0 (00) 1 (02) 0 (00)20682 [00]

4 (03)10151 [04]

Cerebral venous thrombosis 0 (00) 0 (00) 1 (02) 0 (00)20682 [00]

2 (02)10162 [02]

Cerebrovascular disorder 0 (00) 0 (00) 0 (00) 0 (00)20682 [00]

1 (lt01)10154 [lt01]

TIAa 0 (00) 0 (00) 0 (00) 0 (00)20682 [00]

1 (lt01)10163 [lt01]

Ischemic heart disease 0 (00) 0 (00) 0 (00) 5 (02)20662 [02]

0 (00)10164 [00]

Myocardial ischemiaa 0 (00) 0 (00) 0 (00) 2 (lt01)20675 [lt01]

0 (00)10164 [00]

Angina pectoris 0 (00) 0 (00) 0 (00) 1 (lt01)20677 [lt01]

0 (00)10164 [00]

Arteriosclerosis coronary artery 0 (00) 0 (00) 0 (00) 1 (lt01)20678 [lt01]

0 (00)10164 [00]

Blood CPK-MB increased 0 (00) 0 (00) 0 (00) 1 (lt01)20678 [lt01]

0 (00)10164 [00]

Peripheral arterial disease 0 (00) 0 (00) 0 (00) 1 (lt01)20679 [lt01]

1 (lt01)10157 [lt01]

Raynaud phenomenon 0 (00) 0 (00) 0 (00) 1 (lt01)20679 [lt01]

1 (lt01)10157 [lt01]

Hypertension 9 (09) 7 (08) 1 (02) 50 (23)20348 [25]

17 (14)10071 [17]

Hypertension 9 (09) 7 (08) 0 (00) 41 (19)20394 [20]

15 (12)10089 [15]

Diastolic hypertension 0 (00) 0 (00) 1 (02) 0 (00)20682 [00]

1 (lt01)10151 [lt01]

Blood pressure increased 0 (00) 0 (00) 0 (00) 7 (03)20638 [03]

1 (lt01)10159 [lt01]

Blood pressure diastolic increased 0 (00) 0 (00) 0 (00) 1 (lt01)20681 [lt01]

0 (00)10164 [00]

Hypertensive heart diseasea 0 (00) 0 (00) 0 (00) 1 (lt01)20682 [lt01]

0 (00)10164 [00]

Abbreviations CPK-MB = creatine phosphokinasendashmusclebrain QM = once monthlyData for double-blind treatment are presented as number () of patients who received at least 1 dose of 70 mg or 1 dose of 140 mg (the same patient couldhave received both doses) Data for any exposure to erenumab are presented as number () of patientstotal time at risk in years [exposure-adjustedincidence rate per 100 patient-years] where total time at riskwas the time from the patientrsquos first dose of erenumab 70mgor 140mg to onset of the first eventor if no event occurred to the earliest of patientrsquos endof safety follow-up (12 or 16weeks after last dose of 70mgor 140mg respectively) end of study or datacutoff date Each erenumab dose group includes adverse events while the patient received that dose Multiple events of the same preferred term are onlycounted once per patienta Four patients had a positively adjudicated vascular event including 1 patient each with a vascular adverse event of TIA myocardial ischemia andhypertensive heart disease plus 1 patient without a reported vascular adverse event

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e503

(71) mm Hg in the placebo group minus03 (77) mm Hg in theerenumab 70 mg group and minus04 (72) mm Hg in the ere-numab 140 mg group

There were no clinically relevant differences across treatmentgroups for a SBP increase of ge20 mmHg or a DBP increase ofge10 mm Hg from baseline at weeks 4 8 and 12 (table 6)

In the 24-hour continuous BP monitoring study SBP andDBP after 12 weeks of treatment in healthy controls weresimilar in the erenumab 70 mg group and the erenumab140 mg group compared with the placebo group and ere-numab had no effect on the diurnal pattern of BP (figure 2)Similar results were observed for 24-hour continuous BP

monitoring after the first and second monthly doses ofplacebo or erenumab in healthy controls or after each doseof study treatment in patients with migraine (data notshown)

DiscussionDuring 12 weeks of double-blind treatment in 4 clinicalstudies of migraine prevention the incidences of AEs weresimilar across the placebo erenumab 70 mg and erenumab140 mg groups with no evidence of dose dependency forsafety Review of both individual and aggregate AEs did notfind evidence of an association between erenumab treatmentand vascular events including cardiovascular cerebrovascular

Table 4 Patient incidence of adverse events (AEs) during 12 weeks of double-blind treatment in 4 migraine preventionstudies by use of acute migraine-specific medications at baseline

AEs

Users of acute migraine-specific medications Nonusers of acute migraine-specific medications

Placebo QM(N1 = 690)

Erenumab 70 mgQM (N1 = 572)

Erenumab 140 mgQM (N1 = 339)

Placebo QM(N1 = 353)

Erenumab 70 mgQM (N1 = 321)

Erenumab 140 mgQM (N1 = 168)

Any 342 (496) 269 (470) 159 (469) 169 (479) 153 (477) 74 (440)

Any grade ge2 194 (281) 128 (224) 86 (254) 95 (269) 89 (277) 37 (220)

Any grade ge3 24 (35) 21 (37) 9 (27) 9 (25) 9 (28) 7 (42)

Any grade ge4 0 (00) 1 (02) 0 (00) 0 (00) 0 (00) 1 (06)

Serious 13 (19) 12 (21) 5 (15) 3 (08) 3 (09) 0 (00)

Discontinueda 7 (10) 8 (14) 6 (18) 3 (08) 7 (22) 4 (24)

Death 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00)

Abbreviation QM = once monthlyData are presented as number () of patients N1 is the number of patients with usenonuse of an acute migraine-specific medication at baseline Acutemigraine-specific medications include triptan-based or ergot-based medicationsa Investigator reported that the AE led to discontinuation of study treatment

Table 5 Patient incidence of adverse events (AEs) during 12 weeks of double-blind treatment in 4 migraine preventionstudies by number of vascular risk factors at baselinea

AEs

No vascular risk factor 1 vascular risk factor ge2 vascular risk factors

PlaceboQM(N1 = 310)

Erenumab70 mg QM(N1 = 233)

Erenumab140 mg QM(N1 = 148)

PlaceboQM(N1 = 423)

Erenumab70 mg QM(N1 = 377)

Erenumab140 mg QM(N1 = 199)

PlaceboQM(N1 = 310)

Erenumab70 mg QM(N1 = 283)

Erenumab140 mg QM(N1 = 160)

Any 147 (474) 111 (476) 58 (392) 196 (463) 174 (462) 87 (437) 168 (542) 137 (484) 88 (550)

Grade ge2 91 (294) 61 (262) 33 (223) 110 (260) 83 (220) 49 (246) 88 (284) 73 (258) 41 (256)

Grade ge3 10 (32) 7 (30) 4 (27) 12 (28) 12 (32) 5 (25) 11 (35) 11 (39) 7 (44)

Grade ge4 0 (00) 1 (04) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 1 (06)

Serious 3 (10) 2 (09) 2 (14) 7 (17) 6 (16) 1 (05) 6 (19) 7 (25) 2 (13)

Discontinuedb 4 (13) 8 (34) 3 (20) 4 (09) 6 (16) 4 (20) 2 (06) 1 (04) 3 (19)

Death 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) 0 (00)

Abbreviation QM = once monthlyData are presented as number () of patients N1 is the number of patients with that number of vascular risk factors at baselinea See table 1 for descriptions of the 7 categories of vascular risk factorsb Investigator reported that the AE led to discontinuation of study treatment

e504 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

or peripheral vascular events Vascular ischemic events oc-curred during open-label erenumab treatment and wereconfounded with plausible alternative etiologies

Subgroup analyses were conducted to examine the overallsafety of erenumab in patients with migraine who might be athigher risk of vascular events Approximately two-thirds of thepatients were using an acute migraine-specific medication atbaseline Triptans and ergot-based medications have vaso-constrictive effects that may contribute to a higher risk ofcardiovascular or cerebrovascular AEs with these acutemedications1415 Thus triptans and ergot-based medicationshave relative contraindications for use in patients at high riskof vascular events which includes more than 900000 menand women with migraine in the United States24 In thisanalysis the incidences of AEs were similar between theerenumab and placebo groups supporting that adding ere-numab to an acute migraine-specific medication did not in-crease risk compared with adding placebo to an acutemigraine-specificmedication Overall users of acute migraine-specific medications had a slightly higher incidence of seriousAEs compared with nonusers of acute migraine-specificmedications

Among patients with common vascular risk factors at baselinethe incidences of AEs were similar across the placebo and

erenumab treatment groups Regardless of treatment groupthe incidences of AEs and serious AEs were slightly higher inthe subgroup of patients with at least 2 vascular risk factorsthan in those with 0 or 1 risk factor

Erenumab treatment for 12 weeks also had no relevant effecton BP compared with placebo either for resting measure-ments at monthly study visits in patients with episodic orchronic migraine or for continuous 24-hour ambulatory BPmonitoring after 12 weeks of treatment There were no clin-ically relevant differences in the number of patients witha ge10 mm Hg increase in DBP or a ge20 mm Hg increase inSBP no difference between erenumab and placebo for meanSBP or DBP and no imbalances in outliers for BP This isconsistent with the similar incidence of clinically reported AEsof hypertension between the placebo and erenumab treat-ment groups during the first 12 weeks of the placebo-controlled studies and a similar incidence of hypertensionAEs during long-term erenumab treatment Collectivelythese analyses supported the vascular safety profile for ere-numab across different study periods populations andmethodologies providing a comprehensive examination ofthe safety of erenumab treatment for migraine prevention

Supportive information for these findings was provided by exvivo studies that showed no vasoactive effect of erenumab

Figure 1 Blood pressure by visit in patients with migraine during 12 weeks of double-blind treatment in 4 migraineprevention studies

QM = once monthly

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e505

beyond its blockade of the vasodilatory effect of CGRP2526 Inthose studies which measured arterial segments from brainand coronary vessels exposure to erenumab alone did notaffect vascular contractility Because erenumab is selective forthe canonical CGRP receptor the absence of an effect oferenumab on vascular tone may not be generalizable to otheragents that block CGRP from binding to other receptors inthe calcitonin-like receptor family or to the CGRP liganditself27ndash29

Recent clinical studies provide additional support for thevascular safety of erenumab An exercise treadmill safety studyin 88 patients with coronary artery disease and stable anginashowed no aggravation of myocardial ischemia in the patientswho received erenumab compared with those who receivedplacebo30 Erenumab did not decrease exercise duration asmeasured by the change from baseline in total exercise timecompared to placebo (the primary endpoint of the study)That study enrolled patients with coronary artery disease andstable angina the study population was older (median age65 years) and contained a higher proportion of men (78)than this pooled analysis

A drugndashdrug interaction study in 34 healthy controls foundthat concomitant administration of erenumab 140 mg withthe acute migraine-specific medication sumatriptan had noclinically meaningful or additive effects on resting BP com-pared with sumatriptan alone22 Because the healthy controlsin the latter study received sumatriptan those with elevatedBP or heart rate a prolonged QT interval or a history ofmajor vascular event or major cardiovascular interventionwere excluded from the study

Strengths of this safety analysis in approximately 2500patients with migraine were the size and characteristics of thepatient population and the overall extent of exposure to ere-numab Patient demographics (mostly female and mean ageof approximately 40 years) were representative of patientswith migraine supporting the generalizability of the results toother patients with chronic or episodic migraine Safety resultsfrom the individual studies were also consistent acrosspatients with either chronic or episodic migraine17ndash20 Thenumber of patients was complemented by duration of expo-sure The total exposure to erenumab was 2639 patient-yearsand one of the episodic migraine studies includes open-label

Table 6 Percentage of patients by visit with increases from baseline in blood pressure during 12 weeks of double-blindtreatment in 4 migraine prevention studies

Placebo QM(n = 1043)

Erenumab 70 mg QM(n = 893)

Erenumab 140 mg QM(n = 507)

Week 4

No of patients with SBPDBP result 1022 878 501

SBP increase ge20 mm Hg to SBP le140 mm Hg 15 (15) 14 (16) 8 (16)

SBP increase ge20 mm Hg to SBP gt140 mm Hg 9 (09) 6 (07) 3 (06)

DBP increase ge10 mm Hg to DBP le90 mm Hg 57 (56) 61 (69) 46 (92)

DBP increase ge10 mm Hg to DBP gt90 mm Hg 10 (10) 19 (22) 2 (04)

Week 8

No of patients with SBPDBP result 997 871 497

SBP increase ge20 mm Hg to SBP le140 mm Hg 12 (12) 17 (20) 7 (14)

SBP increase ge20 mm Hg to SBP gt140 mm Hg 3 (03) 6 (07) 3 (06)

DBP increase ge10 mm Hg to DBP le90 mm Hg 58 (58) 65 (75) 47 (95)

DBP increase ge10 mm Hg to DBP gt90 mm Hg 15 (15) 12 (14) 2 (04)

Week 12

No of patients with SBPDBP result 985 860 482

SBP increase ge20 mm Hg to SBP le140 mm Hg 10 (10) 24 (28) 8 (17)

SBP increase ge20 mm Hg to SBP gt140 mm Hg 8 (08) 8 (09) 4 (08)

DBP increase ge10 mm Hg to DBP le90 mm Hg 70 (71) 67 (78) 44 (91)

DBP increase ge10 mm Hg to DBP gt90 mm Hg 12 (12) 12 (14) 7 (15)

Abbreviations DBP = diastolic blood pressure QM = once monthly SBP = systolic blood pressureData are presented as number () of patients except as specified

e506 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

treatment for up to 5 years As noted above the exclusion ofpatients with major vascular events within 12 months beforeeach study was a limitation of the analysis but this is typicalfor most studies of migraine treatment or prevention In ad-dition patients with symptomatic vascular disease for morethan 12 months before screening were allowed to participatein the erenumab studies of migraine prevention Hyperten-sion was not an exclusion criterion in the ARISE or STRIVEstudies and patients could participate in the other studies ifthey did not have poorly controlled hypertension Thus

patients who might be excluded from many migraine pre-vention studies were allowed to enroll in the erenumabstudies

This safety analysis used data from the controlled setting ofclinical trials Additional safety data from real-world analyseswould be useful to confirm the vascular safety of erenumabwith widespread use in a broader patient population A pla-cebo comparator group was not available for the longer-termsafety data in this analysis A limitation of any open-label study

Figure 2 Twenty-four-hour continuous blood pressure in healthy controls after 12 weeks of double-blind treatment ina phase 1 study

Assessment from day 64 patients received study treatment on days 1 29 and 57

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e507

is contextualization of events that occur when there is nocomparator group Assessment and adjudication of potentialvascular ischemic events by a panel of clinical experts con-firmed only 4 events as being vascular in origin Each occurredduring open-label erenumab treatment when the adjudicatorswere not blinded to the patientrsquos study treatment While theindependent adjudicators did not determine treatment re-latedness the investigators did not consider any of theseevents to be related to treatment BP also seems to have beenunaffected by erenumab treatment If cardiovascular and ce-rebrovascular AEs do not occur more often it might still bespeculated that their severity is increased with blockade of theCGRP receptor In these clinical studies investigators werenot specifically asked to judge whether the study treatmentinfluenced the severity of vascular AEs Given the rarity ofthese events (only 4 events in the whole dataset were deemedcardiovascular in origin) only substantial long-term real-worldsafety data will help address this question Standardized searchesfor AE terms of vascular events used narrow search terms toincrease specificity of the analyses Sensitivity analyses usingbroad search terms identified 2 additional events that were notspecific to vascular injury during any exposure to erenumabT-wave inversion on ECG (n = 1) and increase in blood crea-tine phosphokinase (n = 17) Review of the individual AEs ofincreased blood creatine phosphokinase indicated that thesewere isolated events with no association with cardiac injury(most of them related to intense physical activity)

In this integrated analysis erenumab had a vascular safetyprofile comparable to that of placebo over 12 weeks and theoverall safety profiles for erenumab and placebo were similarregardless of whether patients also used an acute migraine-specific medication or had risk factors for vascular events Avascular safety signal did not emerge with longer-term treat-ment with dose-blinded or open-label erenumab for up to 256weeks Results from ex vivo studies and other clinical studiesprovided support for the vascular safety of erenumab treat-ment Further study of long-term vascular safety of erenumabin patients with migraine is needed

Study fundingThis study was funded by Amgen Inc and Novartis

DisclosureD Kudrow has participated in advisory boards for AmgenNovartis Alder Eli Lilly Biohaven and Promius and hasreceived research support from Amgen-Novartis Alder TevaEli Lilly Biohaven Zosano Allergan Roche-Genentech DrReddyrsquos Laboratories Novartis and UCB J Pascual has re-ceived research support from Allergan and personal fees fromAllergan Stendhal and Amgen-Novartis P Winner is an in-vestigator in clinical trials sponsored by Teva Amgen Gen-entech Novartis Allergan AstraZeneca Biogen Idec Ipsenand Lilly has participated in advisory boards for TevaAmgen Avanir Novartis Allergan Supernus and Lilly andhas been on speakersrsquo bureaus for Allergan Amgen AvanirLilly Promius Novartis and Supernus D Dodick has

received personal fees from Amgen Association of Trans-lational Medicine University Health Network Daniel Edel-man Inc Autonomic Technologies Axsome Aural AnalyticsAllergan Alder Biohaven Charleston Laboratories DrReddyrsquos LaboratoriesPromius Electrocore LLC Eli LillyeNeura Neurolief Novartis Ipsen Impel Satsuma SupernusSun Pharma (India) Theranica Teva Vedanta WL GoreNocira PSL Group Services University of British ColumbiaZosano ZP Opco Foresite Capital and Oppenheimer hasreceived CME fees or royalty payments from HealthlogixMedicom Worldwide Medlogix Communications MednetMiller Medical PeerView WebMD HealthMedscape Cha-meleon Academy for Continued Healthcare Learning Uni-versal Meeting Management Haymarket Global ScientificCommunications Global Life Sciences Global AccessMeetings UpToDate (Elsevier) Oxford University PressCambridge University Press and Wolters Kluwer Health hasstock options for Aural analytics Healint Theranica SecondOpinionMobile Health Epien GBSNocira MatterhornOntologics and King-Devick Technologies has been a con-sultant without fee for Aural Analytics Healint SecondOpinionMobile Health and Epien has a board of directorsposition for Epien MatterhornOntologics and King-DevickTechnologies has a patent without a fee (Botulinum ToxinDosage Regimen for Chronic Migraine Prophylaxis) has re-ceived professional society fees or reimbursement for travelfrom American Academy of Neurology American BrainFoundation American Headache Society American MigraineFoundation International Headache Society and CanadianHeadache Society and has a use agreement through employerfor Myndshft S Tepper has received research support (nopersonal compensation) from Alder Allergan Amgen ATIDr Reddyrsquos ElectroCore eNeura Neurolief Scion Neuro-stim Teva and Zosano is a consultant for Acorda AlderAlexsa Allergan Alphasights Amgen ATI Axsome Thera-peutics Biohaven Charleston Labs DeepBench Dr ReddyrsquosElectroCore Eli Lilly eNeura GLG Guidepoint GlobalGSK M3 Global Research Magellan Rx Management Medi-cxi Navigant Consulting Neurolief Nordic BioTech NovartisPfizer Reckner Healthcare Relevale Satsuma Scion Neuro-stim Slingshot Insights Sorrento Sudler and HennesseySupernus Teva Theranica Trinity Partners XOC andZosano holds stock options from ATI receives royalties fromSpringer and receives a salary from Dartmouth-HitchcockMedical Center and the AmericanHeadache Society U Reuterhas received personal fees from Amgen Alder Allergan Au-tonomic Technologies Eli Lilly ElectroCore MedscapeNovartis StreaMedUp and Teva F Hong is an employeeand stockholder of Novartis J Klatt is an employee andstockholder of Novartis F Zhang is an employee and stock-holder of Amgen Inc S Cheng is an employee and stockholderof Amgen Inc H Picard is an employee and stockholder ofAmgen Inc O Eisele is an employee and stockholder of AmgenInc J Wang is an employee and stockholder of Amgen Inc JLatham received personal fees from Amgen Inc for medicalwriting support D Mikol is an employee and stockholder ofAmgen Inc Go to NeurologyorgN for full disclosures

e508 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

Publication historyReceived by Neurology December 4 2018 Accepted in final formAugust 6 2019

References1 GBD 2016 Headache Collaborators Global regional and national burden of mi-

graine and tension-type headache 1990-2016 a systematic analysis for the GlobalBurden of Disease Study 2016 Lancet Neurol 201817954ndash976

2 US Burden of Disease Collaborators Mokdad AH Ballestros K et al The state of UShealth 1990-2016 burden of diseases injuries and risk factors among US StatesJAMA 20183191444ndash1472

3 Lipton RB Bigal ME Diamond M Freitag F Reed ML Stewart WF Migraineprevalence disease burden and the need for preventive therapy Neurology 200768343ndash349

4 Diamond S Bigal ME Silberstein S Loder E Reed M Lipton RB Patterns ofdiagnosis and acute and preventive treatment for migraine in the United States resultsfrom the American Migraine Prevalence and Prevention Study Headache 200747355ndash363

5 Blumenfeld AM Bloudek LM Becker WJ et al Patterns of use and reasons fordiscontinuation of prophylactic medications for episodic migraine and chronic mi-graine results from the second International Burden of Migraine Study (IBMS-II)Headache 201353644ndash655

6 Hepp Z Dodick DW Varon SF Gillard P Hansen RN Devine EB Adherence to oralmigraine-preventive medications among patients with chronic migraine Cephalalgia201535478ndash488

7 Ho TW Edvinsson L Goadsby PJ CGRP and its receptors provide new insights intomigraine pathophysiology Nat Rev Neurol 20106573ndash582

8 Edvinsson L Haanes KA Warfvinge K Krause DN CGRP as the target of newmigraine therapies-successful translation from bench to clinic Nat Rev Neurol 201814338ndash350

9 Russell FA King R Smillie SJ Kodji X Brain SD Calcitonin gene-related peptidephysiology and pathophysiology Physiol Rev 2014941099ndash1142

10 MaassenVanDenBrink A Terwindt GM van den Maagdenberg AMJM Calcitoningene-related peptide (receptor) antibodies an exciting avenue for migraine treatmentGenome Med 20181010

11 Kurth T Winter AC Eliassen AH et al Migraine and risk of cardiovascular disease inwomen prospective cohort study BMJ 2016353i2610

12 Murinova N Krashin DL Lucas S Vascular risk in migraineurs interaction of en-dothelial and cortical excitability factors Headache 201454583ndash590

13 Mahmoud AN Mentias A Elgendy AY et al Migraine and the risk of cardiovascularand cerebrovascular events a meta-analysis of 16 cohort studies including 1 152 407subjects BMJ Open 20188e020498

14 Wackenfors A Jarvius M Ingemansson R Edvinsson L Malmsjo M Triptans inducevasoconstriction of human arteries and veins from the thoracic wall J CardiovascPharmacol 200545476ndash484

15 Silberstein SD The pharmacology of ergotamine and dihydroergotamine Headache199737(suppl 1)S15ndashS25

16 Shi L Lehto SG Zhu DX et al Pharmacologic characterization of AMG 334 a potentand selective human monoclonal antibody against the calcitonin gene-related peptidereceptor J Pharmacol Exp Ther 2016356223ndash231

17 Tepper S Ashina M Reuter U et al Safety and efficacy of erenumab for preventivetreatment of chronic migraine a randomised double-blind placebo-controlled phase2 trial Lancet Neurol 201716425ndash434

18 Goadsby PJ Reuter U Hallstrom Y et al A controlled trial of erenumab for episodicmigraine N Engl J Med 20173772123ndash2132

19 Sun H Dodick DW Silberstein S et al Safety and efficacy of AMG 334 for preventionof episodic migraine a randomised double-blind placebo-controlled phase 2 trialLancet Neurol 201615382ndash390

20 Dodick DW Ashina M Brandes JL et al ARISE a phase 3 randomized trial oferenumab for episodic migraine Cephalalgia 2018381026ndash1037

Appendix Authors

Name Location Role Contribution

DavidKudrowMD

California MedicalClinic for HeadacheSanta Monica

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

JulioPascualMD PhD

Department ofNeurology and IDIVALUniversity HospitalMarques de ValdecillaSantander Spain

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

Paul KWinnerDO

Premiere ResearchInstitute NovaSoutheasternUniversity West PalmBeach FL

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

David WDodickMD

Department ofNeurology MayoClinic Phoenix AZ

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

Stewart JTepperMD FAHS

Department ofNeurology GeiselSchool of Medicine atDartmouth HanoverNH

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

UweReuterMD

Department ofNeurology ChariteUniversitatsmedizinBerlin Germany

Author Major role in theacquisition of datainterpreted the datarevised the manuscriptfor intellectual content

FrankHongMD PhD

Novartis EastHanover NJ

Author Interpreted the datarevised the manuscriptfor intellectual content

Jan KlattMD

Novartis BaselSwitzerland

Author Interpreted the datarevised the manuscriptfor intellectual content

FengZhangMS

Amgen Inc ThousandOaks CA

Author Analyzed the datainterpreted the datadrafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

SunfaChengMD

Amgen Inc ThousandOaks CA

Author Interpreted the datadrafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

HernanPicardMD PhD

Amgen Inc ThousandOaks CA

Author Interpreted the datadrafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

OsaEiseleMD MPH

Amgen Inc ThousandOaks CA

Author Interpreted the datadrafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

Appendix (continued)

Name Location Role Contribution

JulieWangDPM

Amgen Inc ThousandOaks CA

Author Drafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

JonathanNLathamPharmD

PharmaScribe LLCAtlanta GA

Author Drafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

Daniel DMikolMD PhD

Amgen Inc ThousandOaks CA

Author Analyzed the datainterpreted the datadrafted the manuscriptfor intellectual contentrevised the manuscriptfor intellectual content

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e509

21 Ashina M Dodick D Goadsby PJ et al Erenumab (AMG 334) in episodic mi-graine interim analysis of an ongoing open-label study Neurology 2017891237ndash1243

22 de Hoon J Van Hecken A Vandermeulen C et al Phase I randomized double-blindplacebo-controlled single-dose and multiple-dose studies of erenumab in healthysubjects and patients with migraine Clin Pharmacol Ther 2018103815ndash825

23 Corrado D Link MS Calkins H Arrhythmogenic right ventricular cardiomyopathyN Engl J Med 201737661ndash72

24 Lipton RB Reed ML Kurth T Fanning KM Buse DC Framingham-based cardio-vascular risk estimates among people with episodic migraine in the US populationresults from the American Migraine Prevalence and Prevention (AMPP) studyHeadache 2017571507ndash1521

25 Ohlsson L Haanes K Kronvall E Xu C Snellman J Edvinsson L Erenumab (AMG334) a monoclonal antagonist antibody against the canonical CGRP receptor does

not impair vasodilatory or contractile responses to other vasoactive agents in humanisolated cranial arteries Epub 2019 Jul 31

26 Beltran AER Ramırez AL Bogers AJ et al Pharmacological selectivity of inhibition ofCGRP-induced relaxations by erenumab studied in human isolated internal mammaryartery (abstract MTIS2018-041) Cephalalgia 20183830

27 Poyner D Pharmacology of receptors for calcitonin gene-related peptide and amylinTrends Pharmacol Sci 199516424ndash428

28 Walker CS Hay DL CGRP in the trigeminovascular system a role for CGRPadrenomedullin and amylin receptors Br J Pharmacol 20131701293ndash1307

29 Walker CS Eftekhari S Bower RL et al A second trigeminal CGRP receptor functionand expression of the AMY1 receptor Ann Clin Transl Neurol 20152595ndash608

30 Depre C Antalik L Starling A et al A randomized double-blind placebo-controlledstudy to evaluate the effect of erenumab on exercise time during a treadmill test inpatients with stable angina Headache 201858715ndash723

e510 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

DOI 101212WNL0000000000008743202094e497-e510 Published Online before print December 18 2019Neurology

David Kudrow Julio Pascual Paul K Winner et al Vascular safety of erenumab for migraine prevention

This information is current as of December 18 2019

ServicesUpdated Information amp

httpnneurologyorgcontent945e497fullincluding high resolution figures can be found at

References httpnneurologyorgcontent945e497fullref-list-1

This article cites 29 articles 5 of which you can access for free at

Citations httpnneurologyorgcontent945e497fullotherarticles

This article has been cited by 1 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectionpatient__safetyPatient safety

httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the

Errata

content942310522fullpdf or page

nextAn erratum has been published regarding this article Please see

Permissions amp Licensing

httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

Disputes amp Debates Editorsrsquo ChoiceSteven Galetta MD FAAN Section Editor

Reader response Epidemiology of NMOSD in Sweden from 1987 to2013 A nationwide population-based studyViktoria Papp (Odense Denmark) Thor Petersen (Aarhus Denmark) Melinda Magyari (Copenhagen)

Nils Koch-Henriksen (Aarhus Denmark) Jette Lautrup Frederiksen (Glostrup Denmark)

Finn Sellebjerg (Copenhagen) Egon Stenager (Odense Denmark) and Zsolt Illes (Odense Denmark)

Neurologyreg 2020941048ndash1049 doi101212WNL0000000000009602

We read the nationwide Swedish epidemiologic study of neuromyelitis optica spectrumdisorder (NMOSD) published by Jonsson et al1 with great interest The study revealedsimilar prevalence (104 per 100000 people [CI 085ndash126]) and incidence (079 per 1000000person-years [CI 055ndash103]) of NMOSD to the prevalence (109 per 100000 people [CI081ndash144]) and incidence (070 per 1000000 person-years [CI 046ndash102]) estimates of therecent nationwide Danish study2 and also to the incidence study of NMOSD from the Region ofCentral Denmark (incidence 012 per 1000000 person-years)3 These studies from Scandinaviashowed similar data to each other and also to other white populations and they contrast with theresults of the study from the region of Southern Denmark4 reporting the highest prevalence(44 per 100000 people [CI 31ndash57]) and incidence (40 per 1000000 person-years [CI30ndash54]) so far in a predominantly Caucasian population

Editorsrsquo note Epidemiology of NMOSD in Sweden from 1987 to 2013A nationwide population-based studyIn their retrospective observational cohort study of 294 patients with neuromyelitis opticaspectrum disorder (NMOSD) Drs Jonsson et al summarized the rising incidence rate ofthis condition in Sweden between 1987 and 2013 The investigators postulate that thehigher incidence ratemdash079 per 1000000 person-years from 030 per 1000000 person-yearsmdashmay have been attributed in part to heightened awareness of the disease andgreater availability of MRI and NMO antibody testing These incidence estimates mirrorwhat have been reported in other predominantly Caucasian cohorts In response Drs Pappet al reiterate that although incidence and prevalence estimates are similar between nationsof similar ethnic profiles Southern Denmark has reported some of the highest prevalence(44 per 100000 person-years) and incidence rates (40 per 1000000 person-years) DrsJonsson et al comment that the higher estimates reported by Asgari et al from SouthernDenmark may be partially explained by the prospective nature of their study and the factthat patients from that study were evaluated for NMOSD even if optic neuritis was thepresenting symptom Based on the collective results from these studies it is likely thatNMOSD remains an underrecognized condition and physicians should have a lowerthreshold to evaluate these patients with serum antibody testing and MRI given thetreatment considerations that are specific to NMOSD compared with MS A correction tofigure 5 was published in the September 10 2019 issue of Neurology

James E Siegler III MD and Steven Galetta MD

Neurologyreg 2020941048 doi101212WNL0000000000009598

1048 Neurology | Volume 94 Number 23 | June 9 2020 NeurologyorgN

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

We would also like to bring attention to figure 5 (Comparison of incidence and prevalence inSweden with other regions) concerning the incidence estimates1 The incidence of NMOSDinAustralia andNewZealand published by Bukhari et al5 is 037 per 1000000 person-years (CI035ndash039) and not 37 per 1000000 person-years (CI 35ndash39) as it is incorrectly shown infigure 5

1 Jonsson DI Sveinsson O Hakim R et al Epidemiology of NMOSD in Sweden from 1987 to 2013 a nationwide population-basedstudy Neurology 201993e181ndashe189

2 Papp V Illes Z Magyari M et al Nationwide prevalence and incidence study of neuromyelitis optica spectrum disorder in DenmarkNeurology 201891e2265ndashe2275

3 Dale GH Svendsen KB Gjelstrup MC et al Incidence of neuromyelitis optica spectrum disorder in the Central Denmark Region ActaNeurol Scand 2018137582ndash588

4 Asgari N Lillevang ST Skejoe HP et al A population-based study of neuromyelitis optica in Caucasians Neurology 2011761589ndash1595

5 Bukhari W Prain KM Waters P et al Incidence and prevalence of NMOSD in Australia and New Zealand J Neurol NeurosurgPsychiatry 201788632ndash638

Copyright copy 2020 American Academy of Neurology

Author response Epidemiology of NMOSD in Sweden from 1987 to2013 A nationwide population-based studyDagur I Jonsson (Stockholm) Olafur Sveinsson (Stockholm) Ramil Hakim (Stockholm) and

Lou Brundin (Stockholm)

Neurologyreg 2020941049ndash1050 doi101212WNL0000000000009601

We are grateful for the comments on our article1 from our Danish colleagues who recentlyconducted a similar study2 applying comparable methods These studies were conducted inparallel and independent from each other in ethnically similar populations strengthening theresults of both studies The study by Asgari3 in 2011 was conducted differently and data werecollected prospectively and involved ophthalmology departments in the Southern Denmarkregion Importantly the study by Asgari implied that there may be an undetected population ofpatients with neuromyelitis optica spectrum disorder (NMOSD) with optic neuritis as an onsetsymptom which is missed This could partly account for the differing results Also Papp et alcorrectly pointed out that in figure 5 of our article1 we mistakenly reported the incidence ofNMOSD in Australia and New Zealand4 as 37 per 1000000 person-years (CI 35ndash39) Thecorrect incidence is 037 per 1000000 person-years (CI 035ndash039) We are sorry for themistake and thankful for the correction The corrected figure does not in any way change theconclusions put forward in our article

NeurologyorgN Neurology | Volume 94 Number 23 | June 9 2020 1049

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

1 Jonsson DI SveinssonO HakimR Brundin L Epidemiology of NMOSD in Sweden from 1987 to 2013 a nationwide population-basedstudy Neurology 201993e181ndashe189

2 Papp V Illes Z Magyari M et al Nationwide prevalence and incidence study of neuromyelitis optica spectrum disorder in DenmarkNeurology 201891e2265ndashe2275

3 Asgari N Lillevang ST Skejoe HP et al A population-based study of neuromyelitis optica in Caucasians Neurology 2011761589ndash1595

4 Bukhari W Prain KM Waters P et al Incidence and prevalence of NMOSD in Australia and New Zealand J Neurol NeurosurgPsychiatry 201788632ndash638

Copyright copy 2020 American Academy of Neurology

Figure Corrected version of figure 5

Incidence (A) andprevalence (B) of neuromyelitis opticaneuromyelitis optica spectrumdisorder in other regions fromavailable publications Horizontal errorbars represent the mean surrounded by a 95 CI

1050 Neurology | Volume 94 Number 23 | June 9 2020 NeurologyorgN

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

Reader response Low hemoglobin and hematoma expansion afterintracerebral hemorrhageZhiyuan Yu (Chengdu China) Jun Zheng (Chengdu China) and Chao You (Chengdu China)

Neurologyreg 2020941051 doi101212WNL0000000000009607

With interest and appreciation I read the article by Roh et al1 which suggested that lowhemoglobin levels could be related to hematoma expansion in intracerebral hemorrhage Al-though previous studies have shown anemia is associated with worse outcomes in thesepatients23 the study by Roh et al is the first one showing that hematoma expansion mediatesthis association However some issues should be noted

The time from onset to baseline CT was relatively long (median 56 hours) which couldinfluence the identification of hematoma expansion Moreover some important results oflaboratory coagulation testing such as prothrombin time and international normalized ratiowere not included in multivariate logistic regression which could make the results not robustenough Therefore their findings still need to be confirmed by further studies

1 Roh DJ Albers DJ Magid-Bernstein J et al Low hemoglobin and hematoma expansion after intracerebral hemorrhage Neurology201993e372ndashe380

2 Kuramatsu JB Gerner ST Lucking H et al Anemia is an independent prognostic factor in intracerebral hemorrhage an observationalcohort study Crit Care 201317R148

3 Zhang S Pan X Wei C et al Associations of anemia with outcomes in patients with spontaneous intracerebral hemorrhage a meta-analysis Front Neurol 201910406

Copyright copy 2020 American Academy of Neurology

Editorsrsquo note Low hemoglobin and hematoma expansion afterintracerebral hemorrhageTo corroborate previous reports that lower hemoglobin levels may be associated withhematoma expansion (HE) and poor functional outcomes after intracerebral hemorrhage(ICH) Drs Roh et al queried their single-center prospective ICH registry of 256 patientsAfter adjustment for known predictors of HE each 1 gdL fall in hemoglobin levels wasassociated with a 20 higher odds of HE and 24 higher odds of severe disability(modified Rankin Scale score 4ndash6) at 3 months Mediation analysis demonstrated that HEcontributed to the poor long-term outcomes seen in patients with lower admissionhemoglobin levels Yu et al address important unmeasured confoundersmdashnotablycoagulopathymdashwhich could have also contributed to adverse radiographic and clinicaloutcomes However as Dr Roh responds patients in this study were excluded if there wasevidence of coagulopathy secondary to systemic disease Furthermore the investigatorsadjusted for the use of therapeutic anticoagulation irrespective of coagulation studiesUltimately both groups agree that these results warrant replication in further studies

James E Siegler III MD and Steven Galetta MD

Neurologyreg 2020941051 doi101212WNL0000000000009604

NeurologyorgN Neurology | Volume 94 Number 23 | June 9 2020 1051

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

Author response Low hemoglobin and hematoma expansion afterintracerebral hemorrhageDavid J Roh (New York)

Neurologyreg 2020941052 doi101212WNL0000000000009608

The comments posited by Yu et al1 appropriately highlight limitations of single-center ob-servational data and the importance of external replication for our findings While adjusting forlaboratory coagulation testing is important to consider it is worth noting that we excludedpatients with laboratory evidence of coagulopathy because of systemic medical disease Inaddition we adjusted for anticoagulation use which should account for prothrombin timeinternational normalized ratio elevations due to a medication effect Subsequently it would beunlikely that the addition of laboratory coagulation markers in a multivariable model wouldaffect the conclusions seen in our cohort

If our findings are replicated the need to study the mechanism(s) behind the findings is importantGiven that underlying medical illness can be associated with lower hemoglobin levels it is plausiblethat the relationship of lower hemoglobin levels with hematoma expansion and poor outcomes seenin our cohort is driven by the underlying disease rather than the hemoglobin level itself Translationalstudies may be the only way to disentangle this potentially confounding factor it is important toconsider this factor before advocating for liberal red blood cell transfusion in intracerebral hemor-rhage as there are known complications associated with unnecessary red blood cell transfusions

1 Roh DJ Albers DJ Magid-Bernstein J et al Low hemoglobin and hematoma expansion after intracerebral hemorrhage Neurology201993e372ndashe380

Copyright copy 2020 American Academy of Neurology

CORRECTION

Vascular safety of erenumab for migraine preventionNeurologyreg 2020941052 doi101212WNL0000000000009310

In the article ldquoVascular safety of erenumab for migraine preventionrdquo by Kudrow et al1

Dr Pascualrsquos affiliation should have been listed as the University of Cantabria The authorsregret the error

Reference1 Kudrow D Pascual J Winner PK et al Vascular safety of erenumab for migraine prevention Neurology 202094e497ndashe510

1052 Neurology | Volume 94 Number 23 | June 9 2020 NeurologyorgN

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited