Journal of Neurology (2020) 267:2851–2864 https://doi.org/10.1007/s00415-020-09835-y ORIGINAL COMMUNICATION Long‑term safety and efficacy of daclizumab beta in relapsing–remitting multiple sclerosis: 6‑year results from the SELECTED open‑label extension study Ralf Gold1 · Ernst‑Wilhelm Radue2 · Gavin Giovannoni3 · Krzysztof Selmaj4 · Eva Kubala Havrdova5 · Xavier Montalban6 · Dusan Stefoski7 · Till Sprenger8 · Randy R. Robinson9 · Sami Fam10 · Jonathan Smith11 · Spyros Chalkias10 · Giorgio Giannattasio10 · Gabriel Lima10 · Wanda Castro‑Borrero10 Received: 12 December 2019 / Revised: 9 April 2020 / Accepted: 9 April 2020 / Published online: 25 May 2020 © The Author(s) 2020 Abstract Objective SELECTED, an open-label extension study, evaluated daclizumab beta treatment for up to 6 years in participants with relapsing multiple sclerosis who completed the randomized SELECT/SELECTION studies. We report final results of SELECTED. Methods Eligible participants who completed 1–2 years of daclizumab beta treatment in SELECT/SELECTION received daclizumab beta 150 mg subcutaneously every 4 weeks for up to 6 years in SELECTED. Safety assessments were evalu- ated for the SELECTED treatment period; efficacy data were evaluated from first dose of daclizumab beta in SELECT/ SELECTION. Results Ninety percent (410/455) of participants who completed treatment in SELECTION enrolled in SELECTED. Within SELECTED, 69% of participants received daclizumab beta for > 3 years, 39% for > 4 years, and 9% for > 5 years; 87% of participants experienced an adverse event and 26% a serious adverse event (excluding multiple sclerosis relapse). No deaths occurred. Overall, hepatic events were reported in 25% of participants; serious hepatic events in 2%. There were no confirmed cases of immune-mediated encephalitis. Based on weeks from the first daclizumab beta dose in SELECT/SELECTION, adjusted annualized relapse rate (95% confidence interval) for weeks 0–24 was 0.21 (0.16–0.29) and remained low on con- tinued treatment. Overall incidence of 24-week confirmed disability progression was 17.4%. Mean numbers of new/newly enlarging T2 hyperintense lesions remained low; percentage change in whole brain volume decreased over time. Conclusions The effects of daclizumab beta on clinical and radiologic outcomes were sustained for up to ~ 8 years of treat- ment. No new safety concerns were identified in SELECTED. Trial registration Clinicaltrials.gov NCT01051349; first registered on January 15, 2010. Keywords Daclizumab beta · Relapsing–remitting multiple sclerosis · SELECTED · Clinical trial Introduction Relapsing–remitting multiple sclerosis (RRMS) is a neuro- degenerative demyelinating disease characterized by acute Sami Fam, Giorgio Giannattasio and Gabriel Lima are former relapses of neurologic symptoms and progressive accumula- employees of Biogen. tion of irreversible disability [4]. The goal of multiple scle- rosis (MS) disease-modifying therapies (DMTs) is to reduce Electronic supplementary material The online version of this the frequency of MS relapses and slow the accumulation of article (https​://doi.org/10.1007/s0041​5-020-09835​-y) contains disability [18]. Effective long-term control of MS disease supplementary material, which is available to authorized users. activity in patients with RRMS is critical for preserving * Ralf Gold function over time [11]. Some adverse events (AEs) may ralf.gold@rub.de only emerge after prolonged courses of treatment. Thus, it is Extended author information available on the last page of the article important to evaluate the efficacy and safety of DMTs over 13 Vol.:(0123456789) 2852 Journal of Neurology (2020) 267:2851–2864 treatment periods longer than those feasible in the initial analysis of SELECTED (data collected as of January pivotal studies. 2014) showed that continuous treatment with daclizumab Daclizumab beta is a humanized monoclonal antibody beta was effective over 3 years of treatment in participants that binds the CD25 subunit of the high-affinity interleukin-2 with RRMS [16]. The yearly incidence of AEs did not (IL-2) receptor and reversibly modulates IL-2 signaling [2]. increase over time, and the safety profile was consistent Blockade of CD25 by daclizumab beta prevents formation with previous observations from SELECT and SELEC- of the high-affinity IL-2 receptor and shifts IL-2 signaling TION [16]. SELECTED was designed to follow partici- to the intermediate-affinity receptor [2, 24]. This results pants who were receiving open-label daclizumab beta in expansion of ­CD56bright natural killer cells, antagonism 150 mg SC every 4 weeks for up to 6 years after comple- of proinflammatory T cells, and reduction in the number tion of SELECTION. of regulatory T cells [2, 24, 26, 30]. Daclizumab beta was In June 2017, a review of daclizumab was initiated by approved in the United States, Europe, Canada, and other the European Medicines Agency (EMA) Pharmacovigilance countries for the treatment of relapsing forms of MS based Risk Assessment Committee (PRAC) following a fatal case on evidence from pivotal clinical trials that showed supe- of fulminant liver failure [10], in addition to four cases of rior efficacy vs placebo and intramuscular interferon beta-1a serious liver injury. Based on PRAC recommendations, on key clinical and radiologic disease outcomes [2, 15, 21] daclizumab beta was restricted to the treatment of patients and a favorable risk–benefit profile at the time of marketing with highly active disease despite a full and adequate course authorization (May and July 2016 in the United States and of treatment with at least one DMT, or with rapidly evolv- Europe, respectively) [9]. In the United States, daclizumab ing severe RMS who are unsuitable for treatment with other was available through a restricted distribution program due DMTs, and required additional monitoring to minimize the to serious safety risks, including liver injury and immune risk of serious liver injury [6, 10]. On March 2, 2018, fol- conditions [29]. Daclizumab beta was formulated to be self- lowing cases of inflammatory encephalitis and meningoen- administered subcutaneously (SC) once monthly [2]. cephalitis, the PRAC initiated an Article 20 referral proce- SELECTED (NCT01051349) was the third and dure for daclizumab beta [8]. On the same day, Biogen and final study in the SELECT trilogy of studies (Fig. 1). AbbVie announced their decision to voluntarily withdraw SELECTED was an open-label, single-arm study that the worldwide marketing authorizations for daclizumab beta, evaluated extended open-label treatment with daclizumab given the limited number of patients being treated and the beta 150 mg every 4 weeks in participants who completed evolving benefit/risk profile [1, 28]. On March 7, 2018, the both SELECT [15] and SELECTION [12]. An interim EMA recommended suspending the marketing authorization Fig. 1  SELECT trilogy study design. Daclizumab beta or placebo the Creative Commons Attribution 4.0 International License (https​:// was administered subcutaneously every 4 weeks in each of the stud- creat​iveco​mmons​.org/licen​ses/by/4.0/) with the following revisions: ies. aGold et al. [15]. bGiovannoni et al. [12]. cGold et al. [16]. Fig- generic name updated to daclizumab beta ure reproduced from Gold et al. [16] distributed under the terms of 13 Journal of Neurology (2020) 267:2851–2864 2853 of daclizumab beta in the European Union and recalling the treatment with daclizumab beta. Women of child-bearing product [7, 28]. potential were required to use effective contraception with Despite the fact that daclizumab beta is no longer a mar- at least one contraceptive method during the study and up keted product for the treatment of RRMS, it is important to 4 months after their last dose of daclizumab beta. Writ- for the safety and efficacy data to be in the public domain, ten informed consent was obtained from each participant to provide additional context on daclizumab beta effects in prior to eligibility evaluation. Participants who experienced RRMS. This report provides the final safety and efficacy a relapse were required to re-consent at the next study visit results from SELECTED after up to 8 years of daclizumab to continue study participation. treatment. SELECTED was terminated early to consolidate ongo- ing extension studies in the daclizumab beta program; participants who were still on treatment as of April 2015 Methods were transitioned to the EXTEND (NCT01797965) long- term extension study, if eligible. EXTEND was an umbrella Study design long-term extension study that included participants from the different ongoing trials, including the phase 3 DECIDE The SELECT trilogy study design has been described pre- trial. Participants who enrolled in EXTEND were considered viously (Fig. 1) [16]. Participants who completed SELECT to have completed SELECTED. EXTEND began in Feb- and SELECTION were eligible for inclusion in SELECTED, ruary 2013. Daclizumab beta dosing subsequently ceased an international, single-arm, open-label extension study in March 2018. Data reported here reflect those collected evaluating 6 years of treatment with daclizumab beta 150 mg within SELECTED or the SELECT trilogy. SC every 4 weeks. SELECTED was conducted across 66 The safety and the intention-to-treat efficacy populations study centers in eight countries; the first participant was consisted of all participants who received at least one dose enrolled/treated on March 31, 2010. The primary objective of daclizumab beta in SELECTED. of SELECTED was to assess the safety of long-term dacli- zumab beta monotherapy in participants with RRMS. Effi- Safety assessment cacy assessment, including number of relapses, assessment of disability progression, number of new/newly enlarging Safety evaluation included all treatment-emergent AEs T2 hyperintense lesions, and annualized percentage change that occurred or worsened in severity from the first dose in whole brain volume (PCBV) were secondary objectives. in SELECTED until up to 6 months after the last dose of The SELECTED protocol was approved by central and local daclizumab beta in SELECTED. ethics committees and the study was conducted in accord- Safety and tolerability assessments included AE monitor- ance with the International Conference on Harmonization ing, physical and neurological exams, vital signs, electrocar- Guideline for Good Clinical Practice [19] and the Declara- diograms, and clinical lab evaluations. Investigators rated tion of Helsinki. AE severity (mild, moderate, or severe) based on protocol guidance and regulatory criteria for a serious AE (SAE). Study participants and procedures An SAE was an AE that resulted in death, persistent or sig- nificant disability/incapacity, or a congenital anomaly/birth Eligibility criteria for SELECT and SELECTION have been defect; required inpatient hospitalization or prolongation of previously reported [12, 15]. Participants who completed existing hospitalization; or was a life-threatening event, in 52 weeks of daclizumab beta treatment in SELECTION the opinion of the investigator. Participants who experienced were eligible to enroll in SELECTED and receive open-label a clinically significant cutaneous AE were referred to a der- daclizumab beta 150 mg SC every 4 weeks for up to 6 more matologist. Liver function testing (alanine aminotransferase years. Participants who permanently discontinued study [ALT], aspartate aminotransferase [AST], and total biliru- treatment because of an AE, had enrolled in any other inves- bin) was performed monthly; results from a liver function tigational study, or who were receiving ongoing treatment test given within the previous 7 days were reviewed by a with any other DMT for MS were ineligible to participate. neurologist and had to be within protocol-defined limits Other participants ineligible to enroll in SELECTED were before a dose of daclizumab beta could be administered. those currently taking valproic acid, carbamazepine, lamo- trigine, or phenytoin; those receiving treatment with isonia- Efficacy assessment zid, propylthiouracil, or nimesulide; and those who had any significant change in their medical status from the previous Relapses were defined as new/recurrent neurological symp- study, including laboratory tests or a clinically significant toms with new objective neurological findings (confirmed condition that, in the opinion of the investigator, precluded by a neurologist) lasting at least 24 h and not associated with 13 2854 Journal of Neurology (2020) 267:2851–2864 fever/infection. Gradually evolving new/recurrent neurologi- on the EDSS from baseline EDSS < 1.0 that was sustained cal symptoms (over months) were considered as disability for 24 weeks) and proportion of participants with 24-week progression, not acute relapses. New/recurrent neurological CDP were evaluated based on Kaplan–Meier product limit symptoms that occurred < 30 days after a protocol-defined method. The date of the first dose of daclizumab beta was relapse were considered part of the same relapse. Expanded used as the start date. Disability Status Scale (EDSS) score was assessed at base- line (defined as the most recent assessment [not occurring during relapse] prior to the first dose of daclizumab beta in Results SELECTED) and every 24 weeks thereafter. Participants who had a ≥ 1.0-point increase in EDSS score from a base- Study participants line EDSS score of ≥ 1.0 or a ≥ 1.5-point increase in EDSS score from a baseline EDSS score of < 1.0 were re-assessed Of 455 participants who completed SELECT and SELEC- 12 and 24 weeks later to determine if disability progression TION, 410 participants (90%) enrolled and were dosed in had occurred. Brain MRI scans were performed at baseline, SELECTED (Fig. 2). Participant characteristics at baseline at weeks 48, 96, 192, 240, and at the end of treatment visit in SELECTED are shown in Table 1. The first participant (week 288), and were read at a central reading institution was enrolled/treated in SELECTED on March 31, 2010, and (Medical Image Analysis Center, Basel, Switzerland) [16]. the date of the last participant visit was August 25, 2016. The original protocol included brain MRI scans at week 144, Overall, 237 (58%) of participants completed the study; of but this assessment was removed in a global amendment to these, 227 transitioned to EXTEND and were considered to the protocol in May 2013. have completed treatment. The number of participants who discontinued study drug each year remained stable through- Efficacy endpoints out the study (10–13% per study year). The mean (median [range]) time on treatment in SELECTED was 40.2 (45.1 Efficacy outcomes were assessed for each participant [0–66]) months (1460.6 patient-years of exposure). In total, based on time from first dose of daclizumab beta received 173 (42%) of participants withdrew from the study and 183 in SELECT or SELECTION and continuing through the (45%) discontinued treatment; the most common reasons for SELECTED study period, for up to 8 years of analyses treatment discontinuation were AEs (89 participants [22%]) (1 year for SELECT, 1 year for SELECTION, and up to and consent withdrawn (55 participants [13%]). The reasons/ 6 years for SELECTED). Efficacy endpoints included time frequency for study withdrawal were similar to those for to first relapse and to 24-week confirmed disability pro- discontinuation. Within the SELECTED study, participants gression (CDP), annualized relapse rate (ARR) by 6-month had received a median (range) of 49 (1–72) doses; 69% of intervals, the number of new or newly enlarging T2 hyperin- participants received daclizumab beta for > 3 years, 39% for tense lesions by yearly intervals, and the annualized PCBV > 4 years, and 9% for > 5 years. Safety and efficacy popula- by yearly intervals. For the analysis of new or newly enlarg- tions for the analysis were the same, comprising all partici- ing T2 hyperintense lesions and annualized PCBV by yearly pants dosed in SELECTED. intervals, data had to be reported for the year prior to the year being assessed. Safety overview and incidence of AEs and SAEs Statistical analyses AEs were summarized by study year (weeks 1–48, weeks 49–96, weeks 97–144, weeks 145–192, and week 193 and ARR was calculated as the total number of relapses divided above) and overall. Year on year, the incidence of AEs, by the total number of days up to the end of the study; the SAEs, and AEs resulting in discontinuation did not increase ratio was multiplied by 365.25. The adjusted ARR was esti- over time and no deaths were reported in SELECTED mated from a Poisson regression adjusted for number of (Table 2). Overall, 87% of participants experienced an AE relapses in the year before study entry. In addition, relapse (20% mild, 54% moderate, and 13% severe), 26% had SAEs rates for individual participants were calculated as the num- (excluding MS relapse), 22% had AEs leading to treatment ber of relapses for that patient divided by the number of days discontinuation (Table 3), and 40% had an AE reported to be they participated in the study; the ratio was multiplied by related to the study drug. Common AEs occurring in 10% of 365.25. Based on these individual relapse rates, the mean participants or more were MS relapse (29%), nasopharyngi- and median were calculated. tis (17%), upper respiratory tract infection (15%), increased The estimated time to 24-week CDP (defined as a ≥ 1.0- ALT (15%), increased AST (12%), headache (11%), urinary point increase on the EDSS from the baseline EDSS ≥ 1.0 tract infection (11%), pharyngitis (10%), and back pain that was sustained for 24 weeks, or a ≥ 1.5-point increase (10%; a listing of AEs with a > 5% incidence are shown in 13 Journal of Neurology (2020) 267:2851–2864 2855 Fig. 2  Participant disposition. Participants who completed SELECT and SELECTION were eligible to enroll in SELECTED. Reasons for withdrawal from the study and discontinuation of the study drug are shown Table 4). The most frequently reported SAEs, excluding MS taken influenza medication containing paracetamol. Two relapse, were urinary tract infection and lymphadenopathy additional participants had ALT/AST > 3 × ULN with con- [in five and six participants, respectively; 1% of participants current bilirubin > 2 × ULN at laboratory assessments dur- for both (Table 5)]. ing hospitalizations that were not captured in the clinical database because they occurred outside of a clinical study Hepatic events site; these cases were medically reviewed. The independent hepatic adjudication committee concluded that none of the Drug-related hepatic AEs were reported in 25%, with serious four cases met the criteria for Hy’s law. However, three of hepatic AEs in 2% (Table 6). Incidence of ALT and AST the four cases were considered by the hepatic adjudication elevations were collected as AEs. Overall, the incidence of committee to be possibly related to study drug. ALT or AST elevations ≥ 3 × upper limit of normal (ULN) was 15%, and the incidence of ALT and AST elevations Cutaneous AEs > 5 × ULN was 9% and 6%, respectively; 4% of participants had ALT or AST levels > 10 × ULN. Two participants had Cutaneous events were reported in 38% of participants, with liver transaminase elevations ≥ 3 × ULN with concurrent serious cutaneous AEs in 4%. Cutaneous SAEs reported in elevation of bilirubin values > 2 × ULN within 3 months more than one participant were urticaria and toxic skin erup- of study drug discontinuation. Both cases were medically tion (three participants each), allergic dermatitis, and psoria- assessed to be confounded by other factors: specifically, one sis (two participants each; Table 6). One serious cutaneous study participant experienced toxic liver disease considered AE was reported as Stevens-Johnson syndrome by the treat- by the investigator to be secondary to valproate treatment ing neurologist, but the diagnosis was not supported by the approximately 2.5 months after discontinuing study treat- case details per the local site dermatologist and the central ment. At the time of this event, the individual had received independent dermatologist assessments, as it was moderate 67 doses of daclizumab beta over 5 years and 2 months of in intensity, localized, lacked any bullous or necrotic skin participation in the SELECT trilogy. The second partici- lesions, and had no areas of skin loss, and there was full pant experienced hepatocellular jaundice with elevated liver thickness of the epidermis [16]. function tests approximately 8 weeks after study treatment Most cutaneous AEs were not injection site reac- discontinuation, following treatment of a cutaneous AE tions. The most common cutaneous AEs were rash (8%), (toxic skin eruption) with herbal supplements and having allergic dermatitis (8%), and eczema (5%). The yearly 13 2856 Journal of Neurology (2020) 267:2851–2864 Table 1  Participant Characteristic Study demographics and clinical population characteristics at SELECTED (N = 410) baseline Previous treatment ­assignmenta Placebo (SELECT); DAC 150 mg (SELECTION) 70 (17) Placebo (SELECT); DAC 300 mg (SELECTION) 66 (16) DAC 150 mg (SELECT); placebo/DAC 150 mg (SELECTION) 69 (17) DAC 150 mg (SELECT and SELECTION) 69 (17) DAC 300 mg (SELECT); placebo/DAC 300 mg (SELECTION) 71 (17) DAC 300 mg (SELECT and SELECTION) 65 (16) Age (years), mean (SD) 38 (9) Female (%) 62 No. of relapses in prior study, mean (SD)b 0.2 (0.5) Range 0–3 EDSS score, mean (SD) 2.7 (1.3) Range 0–6 No. of ­Gd+ lesions, mean (SD) 0.2 (1.0) Range 0–12 Participants with ≥ 1 ­Gd+ lesions, n (%) 40 (10) No. of T2 hyperintense lesions, mean (SD) 46.3 (36.5) Range 0–194 T2 hyperintense lesion volume, m ­ m3, median 3868 T1 hypointense lesion volume, m ­ m3, median 751 Months on treatment, m ­ edianc 45.1 Range < 1–66 Doses, mean (SD) 44.0 (19.4) Median 49.0 Range 1–72 DAC daclizumab beta, EDSS Expanded Disability Status Scale, Gd+ gadolinium-enhancing, SC subcutane- ous, SD standard deviation a Participants in SELECTION who had been treated with daclizumab beta 150 mg or 300 mg in SELECT were randomized to a 6-month placebo washout followed by re-initiation of their previous dosage of dacli- zumab beta or to continue their daclizumab beta dosage in SELECTION b Includes all relapses in SELECTION, whether in the treatment period, randomized washout phase, or fol- low-up period, either confirmed or not confirmed by an independent neurology evaluation committee c Time on treatment, in days, was derived as (date of last dose) − (date of first dose in SELECTED) + 1 Table adapted from Gold et al. [16] distributed under the terms of the Creative Commons Attribution 4.0 International License (https​://creat​iveco​mmons​.org/licen​ses/by/4.0/) with the following revisions: data added for previous treatment assignment incidence of cutaneous events did not increase over time, cutaneous AEs that were severe. Cutaneous events led to and events were reported in 15% of participants during discontinuation of study treatment in 4% of participants. weeks 1–48, 14% during weeks 49–96, 12% during weeks 97–144, 14% during weeks 145–192, and 11% during Infections week 193 and above. Some participants had recurrence of cutaneous AEs across different epochs. The majority Infections were reported in 61% of participants. The most of participants experienced cutaneous AEs that were mild commonly reported AEs of infections (occurring in 10% or or moderate in severity, while seven (2%) participants had more of participants) were nasopharyngitis (17%), upper respiratory tract infections (15%), urinary tract infections 13 Journal of Neurology (2020) 267:2851–2864 2857 Table 2  Summary of AEs by yearly intervals and overall during the SELECTED treatment period Adverse event, n (%) Daclizumab beta 150 mg SC Weeks 1–48 Weeks 49–96 Weeks Weeks 145– Week 193 and Overall (n = 410) (n = 410) (n = 387) 97–144 192 (n = 307) above (n = 250) (n = 343) All AEs 246 (60) 233 (60) 196 (57) 172 (56) 127 (51) 358 (87) Severity of AEs Mild 120 (29) 103 (27) 79 (23) 79 (26) 54 (22) 84 (20) Moderate 113 (28) 118 (30) 104 (30) 82 (27) 63 (25) 222 (54) Severe 13 (3) 12 (3) 13 (4) 11 (4) 10 (4) 52 (13) All SAEs 53 (13) 47 (12) 42 (12) 38 (12) 34 (14) 148 (36) SAEs excluding MS relapse 23 (6) 26 (7) 27 (8) 22 (7) 24 (10) 105 (26) AEs leading to treatment discontinuation 22 (5) 19 (5) 13 (4) 16 (5) 21 (8) 91 (22) Death 0 0 0 0 0 0 AE adverse event, MS multiple sclerosis, SAE serious adverse event, SC subcutaneous Table 3  AEs leading to treatment discontinuation in three or more Table 4  Common AEs (occurring in 5% of participants or more) dur- participants by MedDRA SOC and PT during the SELECTED treat- ing the SELECTED treatment ­perioda ment period AE, n (%) Daclizumab AE, n (%) Daclizumab beta 150 mg SC beta 150 mg SC (N = 410) (N = 410) MS relapse 119 (29) Any AE leading to treatment discontinuation 91 (22) Nasopharyngitis 68 (17) Skin and subcutaneous tissue disorders 15 (4) Upper respiratory tract infection 61 (15) Psoriasis 3 (< 1) Increased ALT 61 (15) Investigations 27 (7) Increased AST 49 (12) Increased ALT 17 (4) Headache 44 (11) Increased AST 8 (2) Urinary tract infection 44 (11) Increased hepatic enzyme 5 (1) Pharyngitis 42 (10) Increased GGT​ 4 (< 1) Back pain 40 (10) Abnormal liver function test 4 (< 1) Respiratory tract infection (viral) 35 (9) Infections and infestations 6 (1) Allergic dermatitis 32 (8) Gastrointestinal disorders 10 (2) Rash 31 (8) Colitis 3 (< 1) Bronchitis 29 (7) Hepatobiliary disorders 11 (3) Diarrhea 25 (6) Liver disorder 4 (< 1) Lymphadenopathy 25 (6) Blood and lymphatic system disorders 8 (2) Viral infection 23 (6) Lymphadenopathy 5 (1) Arthralgia 22 (5) Neoplasms benign, malignant, unspecified 7 (2) Fall 21 (5) Breast cancer 3 (< 1) Eczema 20 (5) AE adverse event, ALT alanine transaminase, AST aspartate transami- Increased GGT​ 20 (5) nase, GGT​gamma-glutamyltransferase, MedDRA Medical Dictionary Respiratory tract infection 20 (5) for Regulatory Activities, PT Preferred Term, SC subcutaneous, SOC Oral herpes 20 (5) System Organ Class Rhinitis 20 (5) AE adverse event, ALT alanine aminotransferase, AST aspartate ami- notransferase, GGT​ gamma-glutamyltransferase, MS multiple sclero- (11%), and pharyngitis (10%). The yearly incidence of sis, SC subcutaneous infections did not increase over time, and infections were a AEs in ≥ 5% of participants by Medical Dictionary for Regula- reported in 34% of participants during weeks 1–48, 32% tory Activities Preferred Term. Participants were counted only once during weeks 49–96, 36% during weeks 97–144, 28% during within each Preferred Term weeks 145–192, and 26% during weeks 193 and above. The 13 2858 Journal of Neurology (2020) 267:2851–2864 Table 5  SAEs occurring in two or more participants during the Table 6  Adverse events of special interest from first dose of dacli- SELECTED treatment p­ erioda zumab beta in SELECTED SAE, n (%) Daclizumab Adverse event, n (%) Total (N = 410) beta 150 mg SC (N = 410) Hepatic ­eventsa AE 104 (25) Any SAE 148 (36) SAE 10 (2) MS relapse 62 (15) Hepatic enzyme increase 2 (< 1) Lymphadenopathy 6 (1) Autoimmune hepatitis 2 (< 1) Urinary tract infection 5 (1) Hepatitis 1 (< 1) Breast cancer 3 (< 1) Liver function test abnormal 1 (< 1) Ulcerative colitis 3 (< 1) Liver disorder 1 (< 1) Toxic skin eruption 3 (< 1) Hepatocellular jaundice 1 (< 1) Urticaria 3 (< 1) Alanine aminotransferase increased 1 (< 1) Autoimmune hepatitis 2 (< 1) Aspartate aminotransferase increased 1 (< 1) Bronchitis 2 (< 1) Hepatic laboratory ­abnormalitiesb Cholelithiasis 2 (< 1) ALT or AST ≥ 3 × ULN 60 (15) Concussion 2 (< 1) ALT or AST > 5 × ULN 37 (9) Dermatitis allergic 2 (< 1) ALT or AST > 10 × ULN 16 (4) Foot fracture 2 (< 1) ALT or AST ≥ 3 × ULN and concurrent bilirubin 2 (< 1)c Gastritis 2 (< 1) > 2 × ULN Herpes zoster 2 (< 1) Hy’s law 0 Increased hepatic enzyme 2 (< 1) Cutaneous ­eventsd Intervertebral disc disorder 2 (< 1) AEs 157 (38) Lower limb fracture 2 (< 1) SAEs 18 (4) Lymphadenitis 2 (< 1) Urticaria 3 (< 1) Pneumonia 2 (< 1) Toxic skin eruption 3 (< 1) Road traffic accident 2 (< 1) Allergic dermatitis 2 (< 1) Upper limb fracture 2 (< 1) Psoriasis 2 (< 1) Wound infection 2 (< 1) Erythrodermic psoriasis 1 (< 1) Photodermatitis 1 (< 1) MS multiple sclerosis, SAE serious adverse event, SC subcutaneous a Angioedema 1 (< 1) SAEs in ≥ 2 patients by Medical Dictionary for Regulatory Activi- ties Preferred Term Dermatitis 1 (< 1) Drug eruption 1 (< 1) Eczema 1 (< 1) majority of infections were mild or moderate in severity, and Erythema nodosum 1 (< 1) 2% were severe. One percent of participants discontinued Seborrheic dermatitis 1 (< 1) treatment due to infections. Stevens–Johnson ­syndromee 1 (< 1) The incidence rate of serious infections was 6%. Seri- Infectionsf ous infections occurring in two or more participants AE 252 (61) included urinary tract infection, pneumonia, bronchitis, SAE 23 (6) herpes zoster, and wound infection (Table 6). There were Urinary tract infection 5 (1) five reports of potential opportunistic infections (1%), Pneumonia 2 (< 1) including vulvovaginal candidiasis, oral candidiasis, fun- Bronchitis 2 (< 1) gal pneumonia, and pulmonary tuberculosis. Pulmonary Herpes zoster 2 (< 1) tuberculosis was reported in a participant from Ukraine Wound infection 2 (< 1) who had received daclizumab beta for 2.5 years (33 total Gastrointestinal infection 1 (< 1) doses). The investigator assessed the AE as severe and not Hepatitis C 1 (< 1) related to daclizumab beta. Treatment was discontinued Infectious mononucleosis 1 (< 1) and the participant withdrew from the study. In six par- Upper respiratory tract infection 1 (< 1) ticipants, daclizumab beta was temporarily interrupted Acute sinusitis 1 (< 1) due to serious infection: herpes zoster in two participants; Appendicitis 1 (< 1) and diverticulitis, wound infection, erysipelas, and pneu- Clostridium difficile colitis 1 (< 1) monia in one participant each. 13 Journal of Neurology (2020) 267:2851–2864 2859 Table 6  (continued) Table 6  (continued) the local dermatologist and the central independent dermatologist as Adverse event, n (%) Total (N = 410) not consistent with Stevens–Johnson syndrome f Diverticulitis 1 (< 1) SOC of Infections and infestations g Erysipelas 1 (< 1) Customized MedDRA search under selected HLTs Furuncle 1 (< 1) h SMQ of Depression and suicide/self-injury HIV infection 1 (< 1) i HLT of Colitis (excluding infective), primary or secondary pathway Peritonsillar abscess 1 (< 1) j Events in the analysis of potential malignancies confirmed upon Pyelonephritis acute 1 (< 1) medical review k Tonsillitis 1 (< 1) The other three non-SAE malignancies were Hodgkin’s disease, Lymphadenopathy ­eventsg penile cancer, and basal cell carcinoma AE 28 (7) SAE 9 (2) Lymphadenopathy 6 (1) Lymphadenopathy events Lymphadenitis 2 (< 1) Lymphoid tissue hyperplasia 1 (< 1) Lymphadenopathy-related events were reported in 7% of par- Depression and suicidal ­ideationh ticipants using a customized Medical Dictionary for Regula- AE 24 (6) tory Activities (MedDRA) search for Preferred Terms under SAE 1 (< 1) selected high-level terms. The most commonly reported lym- Depression and suicide attempt 1 (< 1) phadenopathy-related event was lymphadenopathy (6%); all Inflammatory gastrointestinal ­eventi other lymphadenopathy-related Preferred Terms in the cus- AE 11 (3) tomized search were reported in ≤ 1% of participants. The SAE 6 (1) majority of lymphadenopathy-related events were mild or Ulcerative colitis 3 (< 1) moderate in severity. Severe lymphoid tissue hyperplasia was Colitis 1 (< 1) experienced by one participant. Lymphadenopathy-related Crohn’s disease 1 (< 1) events led to discontinuation of study treatment in 1% of par- Hemorrhagic enterocolitis 1 (< 1) ticipants. The incidence of lymphadenopathy-related SAEs Malignanciesj was 2%. One SAE of lymphadenopathy and one of lymphade- AEk 10 (2) nitis were assessed as related to study treatment by the inves- SAE 7 (2) tigators. The overall incidence of lymphadenopathy-related Breast cancer 3 (< 1) events was stable over time, with no evidence of lymphoma. T-cell lymphoma 1 (< 1) Anal cancer 1 (< 1) Gastrointestinal AEs Clear cell renal cell carcinoma 1 (< 1) Carcinoid tumor pulmonary 1 (< 1) Gastrointestinal AEs (as defined in the MedDRA System Organ Class of Gastrointestinal Disorders) were reported in AE adverse event, ALT alanine aminotransferase, AST aspartate ami- 24% of participants; the majority were mild or moderate in notransferase, HLT high-level term, SAE serious AE, SMQ Standard- ized MedDRA Queries, SOC system organ class, ULN upper limit of severity. The incidence of serious gastrointestinal AEs was normal 3%. The only gastrointestinal event that led to discontinu- a Sublevel SMQ of Drug-related hepatic disorders—comprehensive ation of study treatment in more than one participant was search under the SMQ of Hepatic disorders colitis (three participants). Six (1%) participants reported b Only treatment-emergent lab values are included serious gastrointestinal events: three cases of ulcerative coli- c In one participant, these elevations were noted during a hospitaliza- tis and one case each of colitis, Crohn’s disease, and hemor- tion and assessed as toxic liver disease considered secondary to val- rhagic enterocolitis. Treatment included discontinuation of proate treatment approximately 2.5 months after discontinuing study treatment. In the other case, the participant experienced hepatocellu- study treatment and standard therapies for colitis, including lar jaundice with elevated liver function tests approximately 8 weeks mesalazine, sulfasalazine, corticosteroids, and azathioprine. after study treatment discontinuation, following treatment of a skin The majority of AEs resolved or were stable with no flares event with herbal supplements and having taken influenza medica- following discontinuation of study treatment and/or treat- tion containing paracetamol. Neither was assessed as related to study treatment in the opinion of the investigator. Neither case met Hy’s law ment for the events. A total of 11 participants (3%) had seri- criteria. Two additional participants had ALT/AST > 3 × ULN with ous GI events that could have been immune-mediated colitis; concurrent bilirubin values > 2 × ULN at laboratory assessments dur- six of these 11 participants (1%) reported the following seri- ing hospitalization that were not captured in the clinical database ous events: ulcerative colitis (n = 3), colitis (n = 1), Crohn’s d SOC of Skin and subcutaneous tissue disorders disease (n = 1), and hemorrhagic enterocolitis (n = 1). e The reported case of Stevens–Johnson syndrome was adjudicated by 13 2860 Journal of Neurology (2020) 267:2851–2864 Malignancies Efficacy Potential malignant neoplasms were identified based Efficacy outcomes were assessed based on time from first on a search using the Standardized MedDRA Queries dose of daclizumab beta received in SELECT or SELEC- “malignant or unspecified tumors” and medical review. TION. Daclizumab beta efficacy was maintained in partici- Confirmed malignancies were reported in ten participants pants who continued treatment throughout the study. The (2%): three cases of breast cancer and one case each of adjusted ARR (95% confidence interval [CI]), excluding basal cell carcinoma, anal cancer, pulmonary carcinoid relapses after alternative MS medication for weeks 0–24, tumor, clear cell renal cell carcinoma, Hodgkin’s dis- was 0.21 (0.16–0.29), decreasing to 0.10 (0.04–0.26) beyond ease, penile cancer, and T-cell lymphoma (later revised week 336 (Fig. 3a). The proportion of participants with to lymphadenopathy, based on the clinical course of the 24-week CDP remained stable with extended treatment; patient’s condition and the results of additional diagnos- by the end of SELECTED (up to 8 years of total dacli- tic testing). Of these cases of malignancy, SAEs of anal zumab treatment), 23.9% of participants had 24-week CDP cancer, pulmonary carcinoma, and T-cell lymphoma were (Fig. 3b). considered to be related to study treatment by the investi- The adjusted mean (95% CI) number of new/newly gators. Overall, there was no observed pattern in the type enlarging T2 hyperintense lesions was 1.95 (1.60–2.37) of malignancies. in year 1 and decreased to 1.58 (0.71–3.52) by year 8 of treatment with daclizumab. The mean (median) annualized Other AEs of special interest PBVC was − 0.77% (− 0.63%) in year 1 and 0.32% (0.41%) by year 8 of treatment with daclizumab beta (see Fig. 1 in Autoimmune disorders were reported in 7 (2%) of par- Online Resource 1). ticipants: uveitis and autoimmune hepatitis were reported in two participants each; vitiligo, morphoea, rheumatoid arthritis, and Sjögren’s syndrome were reported in one par- Discussion ticipant each. Serious autoimmune disorders were reported in two (< 1%) participants, and both SAEs were autoim- The SELECT/SELECTION/SELECTED clinical trials mune hepatitis. One case of serious autoimmune hepatitis investigated the safety and efficacy of daclizumab beta in was considered by the investigator to be related to study more than 400 participants with RRMS. These final findings treatment; no action was taken because this participant had from the SELECTED extension study indicate that although been already been withdrawn from the study but was in AEs continued to occur over the extended treatment period, the 6-month monitoring period after stopping treatment. the yearly incidence and severity of AEs remained stable The second case of autoimmune hepatitis was considered throughout the daclizumab beta treatment period and did not not to be related to study treatment; in this case, treatment increase with long-term exposure. No new safety concerns was withdrawn and the participant was withdrawn from were identified in the SELECTED study population. The the study. efficacy benefits seen in a previous 3-year interim analysis Depression and suicidal ideation, identified as AEs of of SELECTED [16] were maintained for up to 8 years of special interest, were reported by 6% of participants. The treatment. most commonly reported events in this category were The overall safety profile of daclizumab beta observed depression (4%) and depressed mood (1%), and the major- in SELECTED was consistent with that reported in the ity of events were mild or moderate in severity. One partici- SELECT and SELECTION studies [12, 15], and in the pant had severe depression and suicidal ideation (two severe 3-year, active-control, phase 3 DECIDE study, which events were recorded in the same participant). There were compared the safety and efficacy of daclizumab beta with no events of depression and suicidal ideation that led to dis- interferon beta-1a in 1841 participants with RRMS [21]. continuation of study treatment. The incidence of depression The overall rates of study withdrawal (42%) and treatment and suicidal ideation SAEs was < 1%. The SAEs of depres- discontinuation (45%) were higher for the daclizumab beta sion and suicide attempt occurred in the same participant, 150 mg group in SELECTED than have been observed in and both SAEs were assessed as not related to study treat- extension studies of other MS DMTs [14, 20]. When com- ment by the investigator. There were no completed suicides. paring the approved dosage across trials, the overall rate of There were no confirmed cases of encephalitis or menin- discontinuation was 22% in the twice daily delayed-release goencephalitis observed or retrospectively identified in dimethyl fumarate group in the 5-year interim analysis of SELECTED. the ENDORSE extension study and 12% in the 0.5 mg 13 Journal of Neurology (2020) 267:2851–2864 2861 Fig. 3  Clinical measures of efficacy in daclizumab beta-treated par- relapses in the year before study entry. Confirmed disability progres- ticipants in SELECTED: a adjusted ARR by 6-month intervals, b sion is defined as ≥ 1-point increase in EDSS score from a baseline 24-week CDP. Efficacy outcomes were assessed from first dose of score ≥ 1 or a ≥ 1.5-point increase from a baseline score = 0, sus- daclizumab beta received in SELECT or SELECTION. Adjusted tained for 24 weeks ARR was estimated from a Poisson regression adjusted for number of fingolimod group in the FREEDOMS extension study. In participants who met threshold ALT/AST elevations, who SELECTED, the most common reasons for withdrawal may or may not have been symptomatic. Although most and treatment discontinuation were as a result of AEs or hepatic AEs were mild or moderate and the majority of consent withdrawn, and the percentage of participants participants experiencing hepatic AEs continued treatment who discontinued treatment for either of these reasons was in SELECTED, hepatic AEs were the most common cause higher in SELECTED (22% and 13%) than in ENDORSE of treatment discontinuation and withdrawal and included (4.5% and 4.9%) or FREEDOMS (2.7% and 6.9%) [14, 20]. increased ALT, AST, hepatic enzyme, or gamma-glutamyl- However, it is important to note that some discontinuations transferase, and abnormal liver function tests. may have been due to protocol-mandated discontinuation in 13 2862 Journal of Neurology (2020) 267:2851–2864 Cutaneous AEs were reported in more than one-third of 144 in SELECTED before this assessment was removed in participants over the course of the SELECTED study and are a protocol amendment. This combined with the early tran- well documented to occur in MS patients treated with dacli- sition of participants in SELECTED to EXTEND explains zumab beta [22, 25]. The incidence reported in SELECTED the variation in numbers of participants with MRI results was similar to that over 2–3 years of daclizumab beta treat- beginning in year 5 of treatment across the SELECTED ment in DECIDE (37%) [21, 22]. trilogy. Also, patients doing less well on treatment or who The incidence of lymphadenopathy in SELECTED were unresponsive may have elected to stop treatment or was consistent with the incidences reported previously in not continue in SELECTED, resulting in potential selection DECIDE and across the daclizumab beta clinical trial pro- bias in favor of those patients who experienced better effi- gram (both 6% for data collected as of November 14, 2014) cacy or better tolerability with daclizumab beta. Although [13, 23]. Most of these events were mild or moderate in conclusions regarding efficacy are limited by the lack of severity and resolved while the participant was on dacli- a placebo control, the low incidence of relapses and MRI zumab beta treatment [23]. Additionally, a case of lymphad- inflammatory lesions over time supports that the treatment enopathy was reported in a patient with MS treated with effects of daclizumab beta in RRMS can be sustained in an earlier intravenous form of daclizumab outside of the the long term. clinical trial setting [25]. Similar to what has been reported In SELECTED, most AEs observed with daclizumab in the clinical trial program, this event resolved following beta were mild or moderate in severity and did not increase discontinuation of treatment. with long-term exposure. There were no observed cases Most infectious, cutaneous, and gastrointestinal AEs of inflammatory encephalitis, meningoencephalitis, fulmi- observed with daclizumab beta were mild or moderate in nant liver failure, or fatal liver injury [8, 27]. The efficacy severity, transient, and generally manageable with standard of daclizumab beta on clinical and radiologic MS dis- therapeutic interventions. The annual incidence of these AEs ease activity outcomes was sustained across yearly treat- in SELECTED after 8 years of treatment with daclizumab ment intervals for up to 8 years. On March 2, 2018, after beta were generally similar to those observed in the first and SELECTED ended, the sponsors voluntarily withdrew second year of treatment in SELECT and SELECTION [12, daclizumab beta from the market, given the limited number 15], respectively, suggesting that up to 8 years of treatment of patients being treated and the evolving benefit/risk pro- with daclizumab beta did not have a negative cumulative file; and on March 7, 2018, the EMA suspended marketing impact on participants with RRMS. Also, an increase in authorization and issued a recall of daclizumab beta [1, 7, rates of malignancy was not observed with long-term use of 10, 28]. The final results from SELECTED contribute to daclizumab beta treatment. the overall understanding of the effects of long-term dacli- Adjusted ARR and the number of new/newly enlarging zumab beta treatment in participants with RRMS in the MRI lesions remained stable and low over up to 8 years of clinical trial setting and represent the concluding chapter daclizumab beta treatment. Annualized percentage change of the SELECT trilogy. in whole brain volume decreased over time on daclizumab beta treatment to levels generally in the range observed in Acknowledgements Open Access funding provided by Projekt DEAL. This study was sponsored by Biogen (Cambridge, MA, USA) and healthy controls [5]. The proportion of participants with AbbVie Inc. (Redwood City, CA, USA). Biogen and AbbVie Inc. pro- 24-week CDP remained low, demonstrating that efficacy vided funding for medical writing support in the development of this of daclizumab beta treatment was maintained for up to manuscript. Katherine Ayling-Rouse (Excel Scientific Solutions, Hor- 8 years of treatment. Overall, the clinical measures of MS sham, UK) wrote the first draft of the manuscript based on input from authors, and Nathaniel Hoover (Excel Scientific Solutions, Fairfield, disease activity evaluated in SELECTED were consistently CT, USA) copyedited and styled the manuscript per journal require- lower than those observed in the placebo group in year 1 of ments. Biogen and AbbVie Inc. reviewed and provided feedback on SELECT [15]. the paper. The authors had full editorial control of the paper, and pro- Long-term extension studies are not without potential vided their final approval of all content. SELECTED Investigators: Czech Republic—Michal Dufek, Radomir Talab, Pavel Stourac, Marta limitations. SELECTED was an open-label study without a Vachova, Denisa Zimova; Germany—Ralf Linker, Patrick Oschmann, placebo control group, similar in design to other open-label Björn Tackenberg, Uwe Zettl; Hungary—Dániel Bereczki, Béla Clem- long-term extension studies [3, 17]. It is important to note ens, Attila Csányi, Tunde Csepany, Péter Diószeghy, András Folyovich, that participants in SELECTED did not receive uniform Péter Harcos, Zita Jobbágy, Etelka Jófejű, Gabriella Kovács, Krisztina Kovács, Zsuzsanna Lohner, Mária Sátori, Attila Valikovics; India— treatment prior to entering the study; participants had been Raja Ram Agrawal, Pahari Ghosh, Thomas Mathew, A. K. Meena, previously randomized to receive either 150 mg or 300 mg Sangeeta Ravat; Poland—Anna Członkowska, Wieslaw Drozdowski, daclizumab beta in SELECT, and some may have under- Waldemar Fryze, Anna Kamińska, Gabriela Kłodowska-Duda, Jan gone a 20-week placebo washout period in SELECTION. Kochanowicz, Krzysztof Selmaj, Zbigniew Stelmasiak, Andrzej Szc- zudlik, Janusz Zbrojkiewicz; Russia—Alexey Boiko, Farit Khabirov, There were fewer participants with MRI data after year 4. Rim Magzhanov, Nadezhda Malkova, Natalia Nikolaevna Maslova, Some participants, but not all, had an MRI scan at week 13 Journal of Neurology (2020) 267:2851–2864 2863 Zahira Mizieva, Irina Poverennova, Semen Prokopenko, Alexey Rozh- Data sharing statement Requests for data supporting this manuscript destvensky, Irina Sokolova, Nikolay Spirin, Olga Vorobeva, Leonid should be submitted to the Biogen Clinical Data Request Portal (https​ Zaslavsky; Ukraine—Nataliya Buchakchyys’ka, Tetyana Kobys, Olex- ://www.bioge​nclin​icald​atare​quest​.com). ander Kozyolkin, Nataliya Lytvynenko, Olena Moroz, Borys Palamar, Valeriy Pashkovskyy, Larysa Sokolova, Olena Statinova, Nataliya Voloshina; United Kingdom—Cris Constantinescu, Clive Hawkins, Open Access This article is licensed under a Creative Commons Attri- Jeremy Hobart, Basil Sharrack, Eli Silber. bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, Author contribution All authors participated in the direction of analy- provide a link to the Creative Commons licence, and indicate if changes ses and interpretation of data, were members of the writing group, were made. The images or other third party material in this article are agreed on the content of the report, reviewed drafts, and approved the included in the article’s Creative Commons licence, unless indicated final version. The data were collected by the study investigators. E-WR otherwise in a credit line to the material. If material is not included in directed the central magnetic resonance imaging reading. JS directed the article’s Creative Commons licence and your intended use is not the statistical analyses. All authors reviewed and approved the final permitted by statutory regulation or exceeds the permitted use, you will manuscript before submission. need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​mmons​.org/licen​ses/by/4.0/. Compliance with ethical standards Conflicts of interest RG: consulting fees from Bayer, Biogen, Merck Serono, Novartis, Sanofi-Aventis, and Teva; editor of Therapeutic References Advances in Neurological Disorders; research support from Biogen, Genzyme, Merck Serono, Novartis, and Teva; E-WR: speaker fees 1. Biogen (2018) Biogen and Abbvie announce the volun- from Actelion, Basilea, Bayer, Biogen, Merck Serono, and Novartis; tary worldwide withdrawal of marketing authorizations for research support from Actelion, Bayer, Biogen, Merck Serono, and ­Zinbryta® (daclizumab) for relapsing multiple sclerosis. http:// Novartis; GGiovannoni: advisory boards for AbbVie Inc., Almirall, inves​tors.bioge​n.com/news-relea​ses/news-relea​se-detai​ls/bioge​ Atara, Biogen, Canbex, Ironwood, Merck, Merck Serono, Novartis, n-and-abbvi ​ e -annou ​ n ce-volun ​ t ary-world ​ w ide-withd ​ r awal ​ . Roche, Sanofi-Genzyme, Synthon, Teva, and Vertex; speaker fees from Accessed 11 Oct 2019 AbbVie, Biogen, Bayer, Genzyme, Merck Serono, Sanofi-Aventis, and 2. Cohan S (2016) Therapeutic efficacy of monthly subcutaneous Teva; co-editor in chief of Multiple Sclerosis and Related Disorders; injection of daclizumab in relapsing multiple sclerosis. Biol research support unrelated to study from Biogen, Genzyme, Ironwood, Targets Ther 10:119–138 Merck Serono, and Novartis; KS: consulting fees from Genzyme, 3. Comi G, O’Connor P, Montalban X, Antel J, Radue EW, Karls- Novartis, Ono, Roche, Synthon, and Teva; speaker fees from Biogen; son G, Pohlmann H, Aradhye S, Kappos L (2010) Phase II EKH: honoraria/research support from Actelion, Biogen, Celgene, study of oral fingolimod (FTY720) in multiple sclerosis: 3-year Genzyme, Merck Serono, Novartis, Roche, and Teva; advisory boards results. Mult Scler 16:197–207 for Actelion, Biogen, Celgene, Genzyme, Merck Serono, Novartis, and 4. Confavreux C, Vukusic S (2014) The clinical course of multiple Roche; supported by the Czech Ministry of Education research project sclerosis. Handb Clin Neurol 122:343–369 PROGRES Q27/LF1; XM: speaker fees/travel expense reimbursement 5. De Stefano N, Stromillo ML, Giorgio A, Bartolozzi ML, Battag- from and steering committee/advisory boards for Actelion, Almirall, lini M, Baldini M, Portaccio E, Amato MP, Sormani MP (2016) Bayer, Biogen, Genzyme, Merck, Neurotec, Novartis, Octapharma, Establishing pathological cut-offs of brain atrophy rates in mul- Receptos, Roche, Sanofi-Aventis, Teva, and Trophos; DS: consult- tiple sclerosis. J Neurol Neurosurg Psychiatry 87:93–99 ing fees from Acorda, Biogen, Serono, and Teva; speaker fees from 6. European Medicines Agency (2017) EMA concludes review of Acorda, Biogen, Elan, EMD Serono, and Teva; royalty payments of Zinbryta and confirms further restrictions to reduce risk of liver licensed drug from Acorda; research support from Biogen, Novartis, damage. https:​ //www.ema.europa​ .eu/docume​ nts/referr​ al/zinbry​ ta- Pfizer, and Serono; TS: speaker/advisory board fees paid to previous articl​ e-20-referr​ al-ema-conclu​ des-review ​ -zinbry​ ta-confir​ ms-furth​ (University Hospital Basel) and current (DKD Helios Klinik Wies- er-restr​ictio​ns-reduc​e-risk_en.pdf. Accessed 11 Oct 2019 baden) employer from Actelion, ATI, Biogen, Desitin, electroCore, 7. European Medicines Agency (2018) EMA recommends immedi- Lilly, Merck Serono, Mitsubishi, Novartis, Sanofi-Genzyme, Roche, ate suspension and recall of multiple sclerosis medicine Zinbryta. and/or Teva; research grants from EFIC Grünenthal, Novartis, Swiss http://www.ema.europ​a.eu/docs/en_GB/docum​ent_libra​ry/Press​ Multiple Sclerosis Society, and Swiss National Science Foundation; _relea​se/2018/03/WC500​24516​7.pdf. Accessed 11 Oct 2019 RRR: employee of and holds stock/stock options in AbbVie; SF, JS, 8. European Medicines Agency (2018) EMA urgently reviewing SC, and WC-B: employees of and hold stock/stock options in Biogen; multiple sclerosis medicine Zinbryta following cases of inflam- GGiannattasio: former employee of Biogen; GL: former employee of matory brain disorders. http://www.ema.europ​a.eu/docs/en_GB/ and holds stock/stock options in Biogen. docum​ent_libra​r y/Press​_relea​se/2018/03/WC500​24489​0.pdf. Accessed 11 Oct 2019 Informed consent All patients provided written informed consent to 9. European Medicines Agency (2018) Zinbryta (daclizumab). https​ participate in the SELECT, SELECTION, and SELECTED studies. ://www.ema.europ​a .eu/en/medic​i nes/human​/ EPAR/zinbr ​y ta. Accessed 11 Oct 2019 Ethical standards The SELECTED protocol was approved by central 10. Faissner S, Gold R (2018) Efficacy and safety of the newer multi- and local ethics committees, and the study was conducted in accord- ple sclerosis drugs approved since 2010. CNS Drugs 32:269–287 ance with the International Conference on Harmonization Guideline for 11. Giovannoni G, Butzkueven H, Dhib-Jalbut S, Hobart J, Kobelt Good Clinical Practice and the Declaration of Helsinki. A full listing G, Pepper G, Sormani MP, Thalheim C, Traboulsee A, Vollmer of Ethics Committees who approved the study is available in Table S1 T (2016) Brain health: time matters in multiple sclerosis. Mult in Gold et al. [16]. Scler Relat Disord 9(Suppl 1):S5–S48 13 2864 Journal of Neurology (2020) 267:2851–2864 12. Giovannoni G, Gold R, Selmaj K, Havrdova E, Montalban X, Radue 21. Kappos L, Wiendl H, Selmaj K, Arnold DL, Havrdova E, Boyko EW, Stefoski D, McNeill M, Amaravadi L, Sweetser M, Elkins A, Kaufman M, Rose J, Greenberg S, Sweetser M, Riester K, J, O’Neill G (2014) Daclizumab high-yield process in relapsing- O’Neill G, Elkins J (2015) Daclizumab HYP versus inter- remitting multiple sclerosis (SELECTION): a multicentre, ran- feron beta-1a in relapsing multiple sclerosis. N Engl J Med domised, double-blind extension trial. Lancet Neurol 13:472–481 373:1418–1428 13. Giovannoni G, Kappos L, Gold R, Khatri BO, Selmaj K, Umans 22. Krueger JG, Kircik L, Hougeir F, Friedman A, You X, Lucas K, Greenberg SJ, Sweetser M, Elkins J, McCroskery P (2016) N, Greenberg SJ, Sweetser M, Castro-Borrero W, McCroskery Safety and tolerability profile of daclizumab in patients with P, Elkins J (2016) Cutaneous adverse events in the randomized, relapsing-remitting multiple sclerosis: an integrated analysis of double-blind, active-comparator DECIDE study of daclizumab clinical studies. Mult Scler Relat Disord 9:36–46 high-yield process versus intramuscular interferon beta-1a in 14. Gold R, Arnold DL, Bar-Or A, Hutchinson M, Kappos L, Havrdova relapsing-remitting multiple sclerosis. Adv Ther 33:1231–1245 E, MacManus DG, Yousry TA, Pozzilli C, Selmaj K, Sweetser MT, 23. Lima G, McCroskery P, Dewar R, Castillo JJ, Holman J, Umans Zhang R, Yang M, Potts J, Novas M, Miller DH, Kurukulasuriya K, Fam S (2016) Characterisation of the lymphadenopathy events NC, Fox RJ, Phillips TJ (2017) Long-term effects of delayed- observed in the daclizumab HYP clinical trials. Mult Scler 22:297 release dimethyl fumarate in multiple sclerosis: Interim analysis of 24. Minocha M, Tran JQ, Sheridan JP, Othman AA (2016) Blockade of ENDORSE, a randomized extension study. Mult Scler 23:253–265 the high-affinity interleukin-2 receptors with daclizumab high-yield 15. Gold R, Giovannoni G, Selmaj K, Havrdova E, Montalban X, process: pharmacokinetic/pharmacodynamic analysis of single- and Radue EW, Stefoski D, Robinson R, Riester K, Rana J, Elkins J, multiple-dose phase I trials. Clin Pharmacokinet 55:121–130 O’Neill G (2013) Daclizumab high-yield process in relapsing- 25. Oh J, Saidha S, Cortese I, Ohayon J, Bielekova B, Calabresi PA, remitting multiple sclerosis (SELECT): a randomised, double- Newsome SD (2014) Daclizumab-induced adverse events in mul- blind, placebo-controlled trial. Lancet 381:2167–2175 tiple organ systems in multiple sclerosis. Neurology 82:984–988 16. Gold R, Radue EW, Giovannoni G, Selmaj K, Havrdova E, Ste- 26. Sheridan JP, Zhang Y, Riester K, Tang MT, Efros L, Shi J, Harris foski D, Sprenger T, Montalban X, Cohan S, Umans K, Greenberg J, Vexler V, Elkins JS (2011) Intermediate-affinity interleukin-2 SJ, Ozen G, Elkins J (2016) Safety and efficacy of daclizumab receptor expression predicts C ­ D56bright natural killer cell expan- in relapsing-remitting multiple sclerosis: 3-year results from the sion after daclizumab treatment in the CHOICE study of patients SELECTED open-label extension study. BMC Neurol 16:117 with multiple sclerosis. Mult Scler 17:1441–1448 17. Goodman AD, Bethoux F, Brown TR, Schapiro RT, Cohen R, 27. Stork L, Brück W, von Gottberg P, Pulkowski U, Kirsten F, Glatzel Marinucci LN, Henney HR 3rd, Blight AR (2015) Long-term M, Rauer S, Scheibe F, Radbruch H, Hammer E, Stürner KH, safety and efficacy of dalfampridine for walking impairment in Kaulen B, Heesen C, Hoffmann F, Brock S, Pawlitzki M, Bopp patients with multiple sclerosis: results of open-label extensions T, Metz I (2019) Severe meningo-/encephalitis after daclizumab of two phase 3 clinical trials. Mult Scler 21:1322–1331 therapy for multiple sclerosis. Mult Scler 25:1618–1632 18. Hart FM, Bainbridge J (2016) Current and emerging treatment of 28. Lancet T (2018) End of the road for daclizumab in multiple scle- multiple sclerosis. Am J Manag Care 22:S159–S170 rosis. Lancet 391:1000 19. ICH Expert Working Group (1996) Guideline for good clinical 29. US Food & Drug Administration (2016) FDA approves Zinbryta practice E6(R1) ICH harmonised tripartite guideline. www.ich. to treat multiple sclerosis. https:​ //www.fda.gov/news-events​ /press​ org/filea​dmin/Publi​c_Web_Site/ICH_Produ​cts/Guide​lines​/Effic​ -annou​nceme​nts/fda-appro​ves-zinbr​yta-treat​-multi​ple-scler​osis. acy/E6/E6_R1_Guide​line.pdf. Accessed 4 Feb 2019 Accessed 10 Oct 2019 20. Kappos L, O’Connor P, Radue EW, Polman C, Hohlfeld R, Selmaj 30. Zhang Y, McClellan M, Efros L, Shi D, Bielekova B, Tang MT, K, Ritter S, Schlosshauer R, von Rosenstiel P, Zhang-Auberson Vexler V, Sheridan JP (2014) Daclizumab reduces CD25 levels L, Francis G (2015) Long-term effects of fingolimod in multiple on T cells through monocyte-mediated trogocytosis. Mult Scler sclerosis: the randomized FREEDOMS extension trial. Neurology 20:156–164 84:1582–1591 Affiliations Ralf Gold1 · Ernst‑Wilhelm Radue2 · Gavin Giovannoni3 · Krzysztof Selmaj4 · Eva Kubala Havrdova5 · Xavier Montalban6 · Dusan Stefoski7 · Till Sprenger8 · Randy R. Robinson9 · Sami Fam10 · Jonathan Smith11 · Spyros Chalkias10 · Giorgio Giannattasio10 · Gabriel Lima10 · Wanda Castro‑Borrero10 1 6 St. Josef-Hospital/Ruhr-University Bochum, 44791 Bochum, Hospital Vall d’Hebron University, Barcelona, Spain Germany 7 Rush University Medical Center, Chicago, IL, USA 2 Medical Image Analysis Center, University Hospital Basel, 8 DKD Helios Klinik Wiesbaden, Wiesbaden, Germany Basel, Switzerland 9 3 AbbVie Inc., Redwood City, CA, USA Barts and The London School of Medicine and Dentistry, 10 Queen Mary University of London, London, UK Biogen, Cambridge, MA, USA 4 11 Department of Neurology, University of Warmia Biogen, Maidenhead, UK and Mazury, Olsztyn, Poland 5 Department of Neurology, First Faculty of Medicine, Charles University, Prague, Czech Republic 13