Headache
`© 2018 American Headache Society
`
`Research Submissions
`
`ISSN 0017-8748
`doi: 10.1111/head.13446
`Published by Wiley Periodicals, Inc.
`
`The Effect of Beginning Treatment With Fremanezumab on
`Headache and Associated Symptoms in the Randomized Phase
`2 Study of High Frequency Episodic Migraine: Post-Hoc
`Analyses on the First 3 Weeks of Treatment
`
`Stephen D. Silberstein, MD; Alan M. Rapoport, MD; Pippa S. Loupe, PhD; Ernesto Aycardi, MD;
`Mirna McDonald, MS; Ronghua Yang, PhD; Marcelo E. Bigal, MD, PhD
`
`Background.—Migraine has a substantial impact on daily living, affecting productivity and quality of life for patients and
`their families. Patients frequently discontinue preventive medications in part because of a delay in headache and symptom relief
`due to the long dose titration procedures necessary for some migraine preventives.
`Objective.—To evaluate the efficacy of fremanezumab, a selective monoclonal CGRP ligand antibody, during the first 3
`weeks of therapy in patients with high-frequency episodic migraine (HFEM) to relieve migraine headaches and associated
` symptoms and to reduce use of acute migraine medications.
`Methods.—In a multicenter, randomized, double-blind, placebo-controlled, phase 2 study, patients with HFEM who met
`inclusion criteria and were 80% compliant with daily headache diary entry were randomized and treated once every 28 days for
`3 months with either placebo or fremanezumab 225 or 675  mg.  Compared to placebo, both doses of fremanezumab significantly
`reduced the primary endpoint of the HFEM study, change in the number of migraine days in month 3 relative to baseline.
`Herein, we performed post-hoc analyses to assess the efficacy of each dose during the first 3 weeks of treatment to reduce
`migraine headache parameters, associated migraine symptoms, and the consumption of acute migraine medications.
`Results.—The sample consisted of 297 study participants. Compared to placebo, decreases in migraine days were seen dur-
`ing the first week of therapy for both fremanezumab doses with least square mean (LSM) differences between fremanezumab
`225  mg and placebo of −0.93 (95% CI: −1.36, −0.49) and between 675  mg dose and placebo of −1.02 (95% CI: −1.46, −0.58),
`both P  <  .0001. This benefit was maintained through the second week of therapy for the 225 and 675  mg doses, respectively,
`(−0.76 (95% CI: −1.11, −0.40) P  <  .0001, −.79 (95% CI: −1.15, −0.44) P  <  .0001) and the third week of therapy (−0.64 (95%
`CI: −0.97, −0.30) P  =  .0003 and −0.64 (95% CI: −0.98, −0.30) P  =  .0003). Likewise in the first week, patients recorded
`reductions in associated migraine symptoms such as nausea, vomiting, photophobia, and phonophobia, which continued through
`weeks 2 and 3. There were also reductions in days with acute medication use to treat migraine for the 225 and 675  mg fre-
`manezumab doses compared to placebo. In the first week, LSM differences between 225  mg and placebo were −1.02 (95% CI:
`−1.39, −0.64) and between 675  mg and placebo were −1.06 (95% CI: −1.39, −0.64) P  <  .0001); for the second and third weeks
`(−1.01 (95% CI: −1.14, −0.55) P  <  .0001; −.90 (95% CI: −1.04, −0.44) P  <  .0001; −.91 (95% CI: −0.92, −0.34) P  <  .0001;
`and −.83 (95% CI: −0.84, −0.26) P  =  .0002), respectively.
`Conclusion.—Fremanezumab treatment resulted in a rapid preventive response in patients with HFEM, with reductions seen
`in several headache parameters and migraine symptoms within the first week after therapy initiation and continuing during the
`
`From the Jefferson Headache Center, Thomas Jefferson University, Philadelphia, PA, USA (S. Silberstein); David Geffen School
`of Medicine at UCLA, Los Angeles, CA, USA (A. Rapoport); Global Research and Development Teva Pharmaceuticals Ltd,
`Overland Park, KS, USA (P. Loupe); Xenon Pharmaceuticals Inc, Boston, MA, USA (E. Aycardi); Global Research and
`Development Teva Pharmaceuticals Ltd, West Chester, PA, USA (M. McDonald and R. Yang); Purdue Pharma, Stamford, CT,
`USA (M. Bigal).
`
`Address all correspondence to P. S. Loupe, Global Research and Development, Teva Pharmaceuticals Ltd, 11100 Nall Avenue, Overland
`Park, KS 66211, USA, email: pippa.loupe@tevapharm.com
`Accepted for publication October 9, 2018.
`
`1
`
`EX2181
`Eli Lilly & Co. v. Teva Pharms. Int'l GMBH
`IPR2018-01423
`
`1
`
`

`

`2 s
`
`econd and third weeks. Patients also were able to rapidly reduce their use of acute medications to treat migraine attacks.
`The trial is registered at Clinicaltrials.gov as NCT02025556.
`
`Month 2018
`
`Key words: migraine, fremanezumab, TEV-48125, calcitonin-gene-related peptide (CGRP), migraine preventive
`medication
`
`Abbreviations: CGRP calcitonin-gene-related peptide, CM chronic migraine, EM episodic migraine, HFEM, high-
`frequency episodic migraine, HIT-6, Headache Impact test, IgG2Δa, immunoglobulin G2Δa, ITT, intent-
`to-treat, IWRS, interactive web response system, MMRM, mixed effect model of repeated measurement,
`M/S, moderate-to-severe; MSQ, Migraine-Specific Quality of Life Questionnaire
`
`(Headache: 2018;0:2-11)
`
`INTRODUCTION
`Migraine is a debilitating neurologic disease that
`can transform from an episodic into a chronic form.1-3
`Migraine frequency can be conceptualized as a con-
`tinuum from low-frequency episodic into high-fre-
`quency episodic and then into chronic migraine; one
`goal of preventive treatment is to avoid this trans-
`formation. This is achieved by reducing headache
`frequency and lowering disability and impact.4 Risk
`factors for progression, as well as factors protecting
`transformation from episodic into chronic migraine,
`have been identified.5
`Current preventive treatments for migraine often
`have unwanted adverse events or contraindications,
`limiting their use. In addition, they often require slow
`titration to achieve tolerable dosing and can take up
`to 4–6 weeks at a minimum to provide relief to pa-
`tients.6-9 Adherence to preventive migraine medica-
`tions is low8 and lack of initial efficacy, accompanied
`by adverse events, often results in discontinuation.
`
`Conflict of Interest: Dr. Silberstein reports receiving consulting
`fees from Alder BioPharmaceuticals, Allergan, Amgen,
`Automatic Technologies, Avanir Pharmaceuticals, Curelator,
`Depomed, Dr. Reddy’s Laboratories, electroCore, Eli Lilly,
`eNeura, Insys Therapeutics, Supernus Pharmaceuticals, Teva
`Pharmaceuticals, Theranica BioElectronics, and Trigemina.
`Dr. Rapoport reports that he is on the speakers bureau for
`Amgen, Depomed, Dr. Reddy’s, electroCore, Teva, and has
`been a consultant for Amgen, Autonomic Technologies, Dr.
`Reddy’s, Electrocore, Impax, Teva, and Zosano
`P. Loupe and R. Yang are full-time employees of Teva
`Pharmaceuticals and M. McDonald is a former employee of
`Teva Pharmaceuticals.
`Dr. Aycardi is a full-time employee of Xenon Pharmaceuticals
`Inc.; he was an employee of Teva Pharmaceuticals at the time
`of this study.
`Dr. Bigal is a full-time employee of Purdue. He was a full-time
`employee of Teva and of Labrys Biologicals at the time of this
`study.
`
`Calcitonin-gene related peptide (CGRP) is a
`neuropeptide involved in the pathophysiology of mi-
`graine; it induces trigeminal sensitization and acti-
`vation of second-order sensory neurons in the brain
`stem.10-13 By its effect on trigeminal sensitization,
`CGRP may be a risk factor for increased headache
`frequency, predisposing the progression of migraine
`from an episodic to chronic form.14-16 Its blockage
`may be associated with a rapid decreased frequency
`or remission of migraine, which has been demon-
`strated for chronic migraine.17
`Fremanezumab (Ajovy, Teva Pharmaceuticals,
`Ltd) is a fully humanized immunoglobulin G subclass
`2 (IgG2Δa) monoclonal antibody with two mutations
`in the hinge that reduce effector function.18 This an-
`tibody selectively targets the CGRP ligand and is
`well-tolerated and effective as a preventive treatment
`for episodic and chronic migraine in phase 2 and 3
`studies.19-22 Fremanezumab resulted in significant
`reductions in migraine and headache days within
`the first month of treatment; additionally, as post-
`hoc analyses in the phase 2 chronic migraine study
`suggested, improvement in headache hours occurs as
`early as 3 days after dosing.17
`Herein, we conduct similar analyses in the
`episodic migraine study, to determine the effect of 2
`doses of fremanezumab (225 and 675 mg) during the
`first 3 weeks of therapy on reducing headache pa-
`rameters and associated symptoms in the phase 2,
`high-frequency episodic migraine (HFEM) study.
`
`METHODS
`Study Design.—The methods for the fremanezumab
`phase 2 HFEM study have been previously reported.19
`Briefly, the study was performed from January 2014
`
`2
`
`

`

`Headache
`
`3
`
`to January 2015 in 62 US sites, including headache
`centers, neurology clinics, and primary care facilities.
`The study was conducted in accordance with Good
`Clinical Practice and US FDA guidelines, and was
`registered at clinicaltrials.gov as NCT02025556. All
`protocols were approved by the institutional review
`boards for each site and patients provided written
`informed consent before participating in the study.
`For patients to be considered eligible participants
`in the study they had to be men or women aged 18–65
`years with confirmed diagnoses of migraine as per the
`International Classification of Headache Disorders
`(ICHD III beta).23 Inclusion criteria consisted of hav-
`ing 8 or more headache days per month for at least 3
`months and confirmation of this headache frequency
`during the 28-day study run-in period. Patients had to
`have 8–14 days of headache per month fulfilling one
`of the following criteria (migraine day, headache pre-
`ceded or accompanied by migraine aura, or headache
`relieved by an ergot or triptan). Patients were allowed
`to have used one standard migraine preventive drug
`at stable doses for at least 2 months prior to study
`onset and acute migraine medications up to 14 days
`per month (maximum of 4 days of opioids or barbitu-
`rates per month). Patients were excluded if they used
`onabotulinumtoxinA for migraine or for any medi-
`cal or cosmetic reasons during the 6 months prior to
`study entry. They were excluded if they had discontin-
`ued >2 medication categories or >3 more preventive
`medications (within 2 medication categories) due to
`lack of efficacy or had treatment with an investiga-
`tional drug or device within 30 days of study entry
`or had any prior exposure to a monoclonal antibody
`targeting the CGRP pathway. They were allowed to
`treat a migraine attack as usual.
`This was a multicenter, randomized, double-blind,
`placebo-controlled, parallel-group, 16-week clinical
`trial consisting of a 28-day screening period, and
`a 3-month double-blinded treatment period. Patients
`who met eligibility criteria were trained during
`the 28-day screening period to provide daily head-
`ache and migraine symptom data using an elec-
`tronic headache diary. Diary entry involved patients
`logging into the web-based interactive system (IWRS)
`every day and answering questions on the occurrence
`of headache, migraine, associated symptoms, and
`the use of acute medications during the previous 24
`hours period. The IWRS allowed patients to back
`
`enter 1 day of information and was locked thereafter.
`If patients demonstrated 80% compliance with diary
`entry during the screening period, they were stratified
`based on sex and preventive medication use and then
`randomized into either one of the two fremanezumab
`arms (225 and 675 mg) or placebo arm by study site
`coordinators using the IWRS and eClinical Operating
`System portal. The randomization scheme was devel-
`oped by a staff member of the contract research orga-
`nization managing the study (NCGS, Charleston, SC,
`USA) who had no further role in the study.
`As the HFEM study was conducted without
`prior proof-of-concept for this or other antibodies,
`in agreement with government regulators, an interim
`analysis was initially suggested. For randomization,
`participants were first stratified by sex and use of
`concomitant preventive drugs. Once the stratification
`group was determined, randomization was done by
`block (men using preventive drugs, men not using pre-
`ventive drugs, women using preventive drugs, women
`not using preventive drugs). Participants (n = 192)
`were initially randomized 1:1 to 675 mg or placebo.
`After the fulfillment of this first cohort, participants
`(n = 105) were randomized 9:1 to 225 mg or placebo,
`yielding an approximately 1:1:1 final randomization
`schedule and an interim analysis for the highest dose.
`However, since no safety concerns emerged in the
`chronic migraine study that was being conducted
`in parallel to this HFEM study (and which enrolled
`higher doses) and due to the fast rate of enrollment on
`the current study, it was agreed that the interim anal-
`ysis was no longer justified, since it would not trans-
`fer into any protocol modification while inducing
`the potential for unblinding and Type I error. IRBs
`had access to the protocol and SAP. To help keep the
`blinding, IRBs requested that the patient consent
`forms indicate a randomization of approximately
`1:1:1 ratio.
`Study sites had 2 blinded study coordinators
`at each clinic visit to protect the treatment blind.
`One study coordinator performed the clinical as-
`sessment which included reviewing electronic diary
`data entry procedures and safety assessments with
`patients. The second study coordinator performed
`the treatment administration. Patients were blinded
`to whether they were receiving placebo or one of the
`fremanezumab doses; they all received 4 injections
`at each 28-day treatment visit and the injections
`
`3
`
`

`

`Month 2018
`
`an interaction term, acute medication use and years
`since onset of disease were defined as covariates,
`with patients treated as random. An unstructured
`covariance matrix for repeated observations within
`patients was used and 95% CIs for the least square
`mean differences between each fremanezumab
`group and placebo were constructed. All variables
`were analyzed by the intent-to-treat (ITT) principle,
`which included all randomized participants who re-
`ceived at least one dose of study drug and provided at
`least one measurement. One patient received study
`drug and discontinued prior to providing one mea-
`surement and was not included in the ITT cohort
`analysis. All statistical tests were 2-sided at alpha
`level of 0.05. As these are post-hoc analyses, the re-
`ported P values should be interpreted with caution
`as there were no adjustments for multiplicity. The
`analyses were conducted with SAS (Version 9.2).
`Role of
`the Funding Source.—The analyses
`this paper were designed by all authors,
`of
`some of whom are employees of the funding source,
`Teva Pharmaceuticals Ltd. One of the authors,
`Mirna McDonald, a former employee of Teva
`Pharmaceuticals conducted the statistical analyses.
`All authors were involved in writing the manuscript.
`They had access to all data in the study and there
`were no agreements with the sponsor which would
`preclude the authors’ ability to analyze and interpret
`data and publish manuscripts independently when and
`where they choose.
`
`RESULTS
`Patient Disposition.—The HFEM study was
`conducted in parallel to a study investigating the
`efficacy of fremanezumab in chronic migraine (CM),
`thus screening included 1170 patients with a migraine
`diagnosis. Of these, 264 patients met eligibility criteria
`for CM and were enrolled in the parallel CM study. An
`additional 609 patients did not qualify for either the
`HFEM or CM study and were excluded. A total of
`297 individuals were eligible for the HFEM study and
`were randomized to receive placebo (n = 104), 225 mg
`(n  =  96), or 675  mg (n  =  97). Of the patients in the
`HFEM study randomized into treatment arms, 2
`participants did not receive any study medication and
`so were not included in the ITT or safety cohorts (1
`in the 225 mg group and 1 in the 675 mg group) and one
`
`4 w
`
`ere identical in packaging and appearance re-
`gardless of treatment arm. Each active injection
`contained 225 mg of fremanezumab. Patients in
`the fremanezumab arms received one of two dosing
`strategies: once monthly subcutaneous injections of
`225 or 675 mg for 3 months.
`Outcomes and Statistical Analyses.—The primary
`efficacy endpoint of the phase 2 HFEM study was the
`mean change from baseline in the number of migraine
`days during the third study month. A migraine day was
`defined as a headache day lasting at least 4 consecutive
`hours and meeting the criteria for migraine with
`aura, migraine without aura, or probable migraine;
`it could also be a headache day of any duration
`treated with migraine specific medications, such as
`a triptan or ergot. Key secondary and exploratory
`endpoints included the change from baseline to
`month 3 in the number of headache days and hours
`with headaches of any severity. The number of days
`with moderate-to-severe (M/S) headaches was an
`exploratory endpoint in the HFEM and was defined
`as a day with at least 4 consecutive hours of moderate
`to severe headache. Other exploratory endpoints
`included assessment of treatment-related changes in
`migraine associated symptoms (nausea, vomiting,
`photophobia, and phonophobia) and a change in the
`use of acute medications during the study. Patients
`provided daily information on the use of acute
`medication consumption by responding to questions
`in the e-diary regarding the previous day usage and
`selecting from a list of the most frequently used
`medications to treat migraine attacks. Patients could
`enter the name of a used medication if it was not
`included in the list and could enter information on
`acute medication use at any hour in the e-diary. For
`purposes of these post-hoc analyses, missing weekly
`data were prorated to 7-day rate based on available
`data within that week.
`As one of the questions remaining from the
`original study was how soon after starting freman-
`ezumab did treatment effects become apparent, we
`conducted mixed effect model repeated measure-
`ment (MMRM) analysis to investigate the earliest
`significant separation from placebo. The post-hoc
`analyses included the change from baseline in the
`efficacy variables at week 1, 2, and 3 as the depen-
`dent variable; preventive medication use, sex, visit,
`and treatment as fixed factors, treatment by visit as
`
`4
`
`

`

`Headache
`
`5
`
`Table 1.—Patient Baseline Demographic and Disease Characteristics
`
`Patient Characteristics†‡
`
`Placebo (n = 104)
`
`Fremanezumab
`225 mg (n = 96)§
`
`Fremanezumab
`675 mg (n = 97)§
`
`42.0 (11.6)
`
`40.8 (12.4)
`
`40.7 (12.6)
`
`Age – mean (SD) years
`Sex
`Male
`Female
`Ethnic origin
`85 (82%)
`White
`13 (13%)
`Black or African American
`2 (2%)
`Asian
`0
`Other
`21.1 (14.1)
`Years of migraine
`11.5 (2.2)
`Migraine days/month
`2.9 (.6)
`Migraine days/week
`M/S headache days/month¶
`9.8 (2.7)
`2.4 (.7)
`M/S headache days/week
`12.4 (2.3)
`Headache days/month
`3.1 (.6)
`Headache days/week
`53.7 (45.7)
`M/S headache hours/month
`13.4 (11.5)
`M/S headache hours/week
`82 (49.3)
`Headache hours/month
`20.5 (12.4)
`Headache hours/week
`5.8 (4.3)
`Days with nausea or vomiting/month
`1.4 (1.1)
`Days with nausea or vomiting/week
`7.6 (4.7)
`Days with photophobia and phonophobia/month
`1.9 (1.2)
`Days with photophobia and phonophobia/week
`10.4 (3.6)
`Days using acute medications/month
`Days using acute medications/week
`2.6 (.9)
`†All P values for treatment comparisons of demographic characteristics were >.05.
`‡Data are mean (SD) of days or hours per month or per week, or number (percentage) of patients.
`§The intent-to-treat mixed effect model of repeated measurement (ITT MMRM) analyses included 95 patients from fremanezumab
`225 mg and 96 from fremanezumab 675 mg groups as one patient from each group withdrew from the study prior to drug treatment.
`¶M/S, moderate-to severe.
`
`12 (12%)
`92 (88%)
`
`9 (9%)
`87 (91%)
`
`74 (77%)
`19 (20%)
`1 (1%)
`0
`18.9 (12.9)
`11.5 (1.9)
`2.9 (.5)
`10.0 (3.1)
`2.5 (.8)
`12.6 (3.1)
`3.1 (.8)
`45.9 (27.2)
`11.5 (6.8)
`76.1 (36.7)
`19.0 (9.2)
`6.2 (4.1)
`1.5 (1.0)
`7.8 (4.3)
`1.9 (1.1)
`10.4 (3.6)
`2.6 (.9)
`
`15 (15%)
`82 (85%)
`
`74 (76%)
`18 (19%)
`1 (1%)
`0
`16.9 (12.3)
`11.3 (2.2)
`2.8 (.6)
`9.6 (2.9)
`2.4 (.7)
`12.5 (2.7)
`3.1 (.6)
`48.4 (27.9)
`12.1 (7.0)
`80.4 (36.6)
`20.1 (9.2)
`5.9 (4.0)
`1.5 (1.0)
`7.6 (4.2)
`1.9 (1.0)
`9.8 (4.0)
`2.4 (1.0)
`
`patient was lost to follow-up after the first injection of
`study medication so there were no diary records; she
`was included in the safety cohort but not in the ITT
`cohort of the 225 mg group.
`As shown in Table 1, baseline demographics and
`clinical disease characteristics were similar across
`treatment arms with a mean (SD) of 11.4 (2.1) migraine
`days per month. Compliance with diary entry was
`high throughout the study and similar across groups;
`for instance, during the first treatment month, 91% of
`placebo, 89% of 225 mg, and 84% of 675 mg patients
`provided daily entries at ≥80% compliance rate (≥22
`days per 28 day treatment period).
`
`Efficacy.—Headache Parameters.—The study has
`been described in detail.19 As displayed in Figure 1 and
`Table 2, the number of migraine days for patients
`treated with fremanezumab was reduced compared to
`those treated with placebo in the first week of therapy
`for both fremanezumab doses; LSM differences
`between placebo and active treatment at week 1 for the
`225  mg fremanezumab dose is −0.93 migraine days
`per week (95% CI: −1.36, −0.49); and for the 675  mg
`dose −1.02 migraine days per week (95% CI: −1.46,
`−0.58), both P  <  .0001. This benefit was maintained
`through the second and third weeks of therapy (all
`P < .001). Likewise there were consistent reductions in
`
`5
`
`

`

`6
`
`Month 2018
`
`Fig. 1.—Reductions in headache parameters (mean ± SE) during the first 3 weeks of fremanezumab treatment.
`
`weekly M/S headache days and days of headache of
`any severity within the first 3 weeks of fremanezumab
`administration (all P < .01). For M/S headache hours and
`headache hours of any severity, the same pattern of
`efficacy emerged; the mean number of headache hours
`per week for 225 and 675  mg fremanezumab groups
`decreased by 6.50 and 8.99 hours in week 1, 5.69 and
`6.52 hours in week 2, and 4.90 and 5.96 hours in week 3
`compared to decreases of 0.78, 1.08, and 1.90 hours per
`week in the placebo group (all P < .01).
`
`Migraine associated symptoms: Nausea, Vomiting,
`Photophobia, and Phonophobia.—In addition to
`questions concerning headaches, we also assessed
`whether
`fremanezumab
`caused
`reductions
`in
`symptoms associated with migraine such as nausea,
`vomiting, photophobia, and phonophobia (Table 2
`and Fig. 2).
`Compared to patients on placebo, decreases from
`baseline in the numbers of days per week with nausea
`or vomiting occurred in the first week of therapy for
`
`6
`
`

`

`7
`
`change from baseline at weeks 1, 2, and 3 between the placebo and the fremanezumab dose groups. M/S refers to moderate to severe.
`between placebo and active treatment group. P values and 95% confidence intervals are from the mixed effect model of repeated measurement (MMRM) analyses conducted on the
`†Parameter values provided include the change from baseline in least square means (LSMs) for each treatment group and in bold font, the LSM difference (SE) in change from baseline
`
`P = .0002
`(−0.84, −0.26)
`− 0.55 (.15)
`−0.83 (.13)
`P = .0047
`(−0.73, −0.13)
`−0.43 (.15)
`−0.77 (.13)
`P = .069
`(−0.49, 0.02)
`−0.24 (.13)
`−0.54 (.11)
`P = .0002
`(−9.03, −2.89)
`−5.96 (1.56)
`−7.86 (1.33)
`P = .0042
`(−5.70, −1.07)
`−3.39 (1.18)
`−4.55 (1.00)
`P = .0011
`(−0.88, −0.22)
`−0.55 (.17)
`−0.95 (.14)
`P = .0028
`(−0.79, −0.17)
`−0.48 (.16)
`−0.93 (.14)
`P = .0003
`(−0.98, −0.30)
`−0.64 (.17)
`−1.17 (.15)
`
`P < .0001
`(−0.92, −0.34)
`−0.63 (.15)
`−0.91 (.13)
`P = .0044
`(−0.73, −0.14)
`−0.43 (.15)
`−0.77 (.13)
`P = .0146
`(−0.57, −0.06)
`−0.32 (.13)
`−0.62 (.11)
`P = .0017
`(−7.95, −1.85)
`−4.90 (1.55)
`−6.80 (1.38)
`P = .0094
`(−5.36, −0.75)
`−3.06 (1.17)
`−4.22 (1.03)
`P = .0002
`(−0.96, −0.31)
`−0.63 (.17)
`−1.03 (.15)
`P = .0005
`(−0.86, −0.24)
`−0.55 (.16)
`−1.00 (.14)
`P = .0003
`(−0.97, −0.30)
`−0.64 (.17)
`−1.16 (.15)
`
`−0.28 (.13)
`
`−0.34 (.13)
`
`−0.31 (.11)
`
`−1.90 (1.34)
`
`−1.16 (1.01)
`
`−0.40 (.14)
`
`−0.45 (.14)
`
`−0.52 (.15)
`
`P < .0001
`(−1.04, −0.44)
`−0.74 (.15)
`−0.90 (.13)
`P = .0004
`(−0.84, −0.25)
`−0.54 (.15)
`−0.81 (.13)
`P = .0274
`(−0.56, −0.03)
`−0.30 (.13)
`−0.59 (.11)
`P = .0002
`(−9.87, −3.17)
`−6.52 (1.70)
`−7.60 (1.42)
`P = .0025
`(−6.28, −1.35)
`−3.82 (1.25)
`−4.49 (1.05)
`P = .0002
`(−0.99, −0.32)
`−0.66 (.17)
`−0.96 (.15)
`P = .0009
`(−0.87, −0.22)
`−0.55 (.16)
`−0.92
`P < .0001
`(−1.15, −0.44)
`−0.79 (.18)
`−1.24 (.15)
`
`P < .0001
`(−1.14, −0.55)
`−0.84 (.15)
`−1.01 (.13)
`P = .0003
`(−0.85, −0.26)
`−0.56 (.15)
`−0.82 (.13)
`P = .005
`(−0.64, −0.11)
`−0.38 (.13)
`−0.67 (.12)
`P = .0009
`(−9.03, −2.36)
`−5.69 (1.69)
`−6.77 (1.46)
`P = .0032
`(−6.16, −1.25)
`−3.70 (1.25)
`−4.38 (1.08)
`P < .0001
`(−1.11, −0.44)
`−0.78 (.17)
`−1.08 (.15)
`P < .0001
`(−1.01, −0.38)
`−0.70 (.16)
`−1.07 (.14)
`P < .0001
`(−1.11, −0.40)
`−0.76 (.18)
`−1.20 (.16)
`
`−0.16 (.13)
`
`−0.26 (.13)
`
`−0.29 (.11)
`
`−1.08 (1.42)
`
`−0.68 (1.05)
`
`−0.30 (.15)
`
`−0.37 (.14)
`
`−0.44 (0.15)
`
`P < .0001
`(−1.39, −0.64)
`−1.02 (.19)
`−1.06 (.15)
`P < .0001
`(−1.17, −0.46)
`−0.81 (.18)
`−0.89 (.15)
`P = .0022
`(−0.84, −0.19)
`−0.51 (.17)
`−0.68 (.13)
`P < .0001
`(−13.36, −4.62)
`−8.99 (2.22)
`−8.20 (1.75)
`P = .0005
`(−9.15, −2.56)
`−5.85 (1.67)
`−4.90 (1.32)
`P < .0001
`(−1.34, −0.50)
`−0.92 (.21)
`−1.09 (.17)
`P = .0003
`(−1.11, −0.33)
`−0.72 (.20)
`−0.98
`P < .0001
`(−1.46, −0.58)
`−1.02 (.22)
`−1.38 (.18)
`
`P < .0001
`(−1.39, −0.64)
`−1.02 (.19)
`−1.07 (.16)
`P < .0001
`(−1.18, −0.47)
`−.82 (.18)
`−0.91 (.15)
`P = .0061
`(−0.78, −0.13)
`−0.46 (.17)
`−0.62 (.14)
`P = .0036
`(−10.85, −2.14)
`−6.50 (2.21)
`−5.71 (1.78)
`P = .0062
`(−7.88, −1.31)
`−4.60 (1.67)
`−3.64 (1.34)
`P < .0001
`(−1.34, −0.50)
`−0.92 (.21)
`−1.08 (.17)
`P < .0001
`(−1.22, −0.44)
`−0.83 (.20)
`−1.08 (.16)
`P < .0001
`(−1.36, - 0.49)
`−0.93 (.22)
`−1.28 (.18)
`
`use
`medication
`
`−0.05 (.15)
`
`Days with acute
`
`phonophobia
`and
`photophobia
`
`−0.08 (.15)
`
`Days with
`
`−0.17 (.13)
`
`vomiting
`nausea/
`Days with
`
`any severity
`headaches of
`
`0.78 (1.73)
`
`Hours of
`
`.96 (1.3)
`
`M/S headache
`
`hours
`
`Headache days−0.16 (.17)
`
`−0.26 (.16)
`
`M/S headache
`
`days
`
`−0.36 (.18)
`
`Migraine days
`
`Headache
`
`(n = 96)
`675 mg
`
`(n = 95)
`225 mg
`
`(n = 104)
`Placebo
`
`(n = 96)
`675 mg
`
`(n = 95)
`225 mg
`
`(n = 104)
`Placebo
`
`(n = 96)
`675 mg
`
`(n = 95)
`225 mg
`
`(n = 104)
`Placebo
`
`Parameters†
`
`Change from Baseline to Week 3
`
`Change from Baseline to Week 2
`
`Change from Baseline to Week 1
`
`Table 2.—Change from Baseline in the Occurrence of Headache Parameters and Migraine Symptoms During the First 3 Weeks of Fremanezumab Treatment
`
`7
`
`

`

`8
`
`Month 2018
`
`difference −0.30 (95% CI: −0.56, −0.03) P = .0274) and
`less so in the third week (LSM difference −0.24 (95%
`CI: −0.49, 0.02) P = .0696).
`Similarly, decreases in the number of days per
`week with photophobia and phonophobia occurred
`within the first week for the fremanezumab 225
`and 675 mg doses (LSM differences of −0.82 (95%
`CI: −1.18, −0.47), and −0.81 (95% CI: −1.17, −0.46)
`both P < .0001). These decreases continued in the
`second week (P = .0003 and P = .0004) and third
`week (P = .0044 and P = .0047) of fremanezumab
`treatment.
`in Acute Medication Use.—There
`Reduction
`was also a difference between the placebo and
`fremanezumab groups in the number of days of
`acute medication use per week in the first 3 weeks
`(Table 2 and Figure 2). The use of acute medications
`decreased for the fremanezumab 225 and 675 mg doses
`compared to placebo during the first week (LSM
`differences of −1.02 (95% CI: −1.39, −0.64), and −1.02
`(95% CI: −1.39, −0.64) both P < .0001); second week
`(LSM differences of −0.84 (95% CI: −1.14, −0.55) and
`−0.74 (95% CI: −1.04, −0.44) both P  <  .0001); and
`third week (LSM differences of −0.63 (95% CI: −0.92,
`−0.34) P  <  .0001 and −.55 (95% CI: −0.84, −0.26)
`P = .0002).
`Safety.—The safety results for the HFEM have
`been fully described earlier.18 The most common
`adverse events were injection site reactions (pain,
`erythema, and/or pruritus) and there were no relevant
`changes in vital signs or blood pressure between
`groups and no treatment emergent anti-drug antibody
`response occurred.
`
`DISCUSSION
`Fremanezumab treatment, in a post-hoc analysis
`of the phase 2 HFEM study, resulted in a rapid pre-
`ventive response in patients with episodic migraine.
`Improvements in multiple headache parameters and
`migraine symptoms, as well as acute medication use
`were seen in the first few weeks of treatment. Monthly
`doses of fremanezumab, 225 and 675 mg, demon-
`strated decreases in the number of migraine days and
`other headache parameters in the first week of treat-
`ment. These findings are consistent with the phase 2
`chronic migraine study17 in which patients on freman-
`ezumab had decreases in headaches within the first 3
`days of treatment.
`
`Fig. 2.—Reductions in migraine symptoms (mean ± SE)
`during the first 3 weeks of fremanezumab treatment.
`
`fremanezumab 225 and 675 mg doses (LSM differ-
`ences of −0.46 (95% CI: −0.78, −0.13) P = .0061 and
`−0.51 (95% CI: −0.84, −0.19) P = .0022, respectively).
`Decreases continued in second (LSM difference −0.38
`(95% CI: −0.64, −0.11) P = .005) and third week (LSM
`difference −0.32 (95% CI: −0.57, −0.06) P = .0146)
`for the 225 mg fremanezumab dose. For the 675 mg
`dose, decreases continued in the second week (LSM
`
`8
`
`

`

`Headache
`
`9
`
`Collectively these data suggest that the onset of
`effect for fremanezumab as a preventive treatment
`for migraine may correct some of the issues associ-
`ated with currently available oral migraine preventive
`treatments. Oral therapies often require gradual dose
`titration and their efficacy may not be apparent for
`several months; this along with their adverse event
`profile may lead to patients discontinuing treat-
`ment before their therapeutic effect is apparent.7,24
`Fremanezumab, with its rapid onset of effect, to-
`gether with a low rate of adverse events, may improve
`patient adherence and compliance with a preventive
`medication.
`In this study, patients at baseline had an average
`of 11 migraine days per month, or 3 migraine days
`per week. Within the first week of treatment, patients
`on fremanezumab reported a decrease in 1 migraine
`day per week compared to patients on placebo; thus
`patients rapidly began to feel better. A reduction of
`1 migraine day per month, and a 30% reduction in
`headache frequency from baseline have previously
`been suggested as 2 of the criteria necessary for a
`clinically meaningful effect for a preventive treatment
`for chronic migraine.25 The study was not designed
`to assess the clinical meaningfulness of changes in
`migraine days and the other headache parameters.
`Despite this, it is reassuring that both doses of fre-
`manezumab met these 2 criteria in the first week of
`therapy. The effects seen on migraine days and the
`other headache parameters continued through the
`second and third week of the first month and were
`maintained during the entire 3 months of the study.19
`We examined the early effects of fremanezumab
`on associated migraine symptoms, since they are
`a major contributor to the burden of migraine.26,27
`Many patients have self-reported problems with tak-
`ing oral preventive migraine medications because of
`nausea and vomiting.27 Photophobia and phonopho-
`bia are very common and are thought to result from
`enhanced sensitivity to light and sound occurring
`in both the premonitory (prodrome) and headache
`phases of migraine.28 Treatment-related reductions in
`associated migraine symptoms occurred in the first
`few weeks and continued throughout the three month
`study (all comparisons between fremanezumab dose
`groups and placebo were significantly different with
`the exception of reduction in days of nausea or vomit-
`ing for the 675 mg dose, P < .069).19
`
`Fremanezumab also reduced the use of acute mi-
`graine medications during the study. Patients with fre-
`quent migraine attacks carry a nearly daily burden,29
`and since excessive use of acute medication is a risk
`factor for migraine worsening,30 the early decrease in
`acute care medication consumption is important.
`This study has limitations. First, the analyses were
`not defined a priori, since quick onset of action was not
`anticipated. Given that these were post-hoc analyses,
`adjustments for multiple comparisons are not appropri-
`ate31 and the findings should be considered exploratory
`in nature. Another limitation is that we are unable to pro-
`vide specific drug information on the decreases in acute
`medications. We also lack quality of life measures, such
`as the Migraine-Specific Quality of Life Questionnaire
`(MSQ) and impact data, such as the Headache Impact
`Test