The study adhered to the principles of the Declaration of Helsinki, the International Conference on Harmonisation/Good Clinical Practice, Good Clinical Practice for Trials on Drugs (Ministry of Health, Labour and Welfare Ordinance No. 28, 27 March 1997), and the AstraZeneca K.K. policy on Bioethics and Human Biological Samples. The study protocol was approved by the institutional review board at each investigational site, and all patients provided written informed consent prior to participation (or their legal guardians if aged under 20 years). This study was registered at ClinicalTrials.gov (NCT01959165).

This was a multi-center, randomized, double-blind, placebocontrolled, parallel group, phase II study to evaluate the efficacy and safety of multiple doses of 21 mg and 70 mg abrilumab, and a single dose of 210 mg of abrilumab in Japanese patients with moderate-to-severe UC. The study was conducted from 21 November 2013 to 15 October 2016. The study design is shown in Fig. 1.

The study consisted of 3 study periods. In the 12-week double-blind period, the patients were randomized using an interactive web response system to receive placebo or multiple doses of 21 mg or 70 mg, or a single dose of 210 mg of abrilumab in a ratio of 1:1:1:1.

In the placebo and abrilumab 21 mg and 70 mg groups, patients received the investigational product by subcutaneous (SC) injection on day 1, and at weeks 2, 4, and 8. In the abrilumab 210 mg group, patients received abrilumab 210 mg SC on day 1 and placebo at weeks 2, 4, and 8. Patients were stratified according to prior anti-TNF-α agent use and participation in an immunophenotyping and absolute counting PD assay sub-study, with 40% to 60% of patients with any prior anti-TNF agent use permitted in the study.

Randomization codes were allocated strictly and sequentially as patients became eligible. After randomization, an unblinded pharmacist or designee in each site prepared the investigational products according to a preparation manual and provided them to blinded site staff under blinded conditions.

All documents/information indicating patients’ treatment details were handled confidentially. The double-blind period was followed by a 36-week open-label period in which all patients received abrilumab 210 mg SC every 12 weeks, and then a 28-week safety/104-week PML follow-up period.

A randomized, double-blind design was selected to minimize bias. Placebo was selected as the control drug to evaluate the efficacy of abrilumab precisely, but concomitant use of 5-aminosalicylic acid, oral prednisolone (or equivalent), and AZA or 6-MP (or equivalent) was permitted in the double-blind period on the condition that the dosage and administration remained unchanged during the period. The open-label period was established to evaluate the efficacy in persistent response and the long-term safety of abrilumab, and to provide patients with continued treatment opportunity from an ethical point of view. The 28-week safety follow-up period was established to evaluate late adverse events (AEs), and its duration was determined as ≥ 5 times the elimination half-life of abrilumab 210 mg from a global phase I Study 20101261 (33.8 ± 6.74 days) (data on file). The PML follow-up period was established as in the overseas phase II study [17]. The dose of abrilumab for the Japanese patients in this study was based on previous study 20110259 (data on file), which showed that the PK and PD of abrilumab were similar in Japanese and non-Japanese patients, meaning no dose adjustments were required.

The inclusion criteria were: age 18 to 65 years at screening; a diagnosis of UC established ≥3 months before visit 2 (week 0); moderate-to-severe active UC (total Mayo score of 6 to 12 with a rectosigmoidoscopy score ≥2 measured in the screening period) prior to visit 2; and inadequate response, loss of response, or intolerance to immunomodulators and/or anti-TNF-α agents. Patients taking AZA or 6-MP could be included if the treatment had been initiated ≥12 weeks prior to visit 2 and the dosage had been stable for ≥ 8 weeks prior to visit 2, or if the treatment had been discontinued ≥8 weeks before visit 2. Similarly, patients taking 5-aminosalicylic acid and/or oral prednisone or equivalent up to 20 mg/day were included if the dosage had been stable for ≥2 weeks prior to visit 2. Patients also required to have normal neurological examination findings, no history of tuberculosis, and a negative tuberculosis test at screening to be included.

The exclusion criteria were rectal disease only (i.e., within 10 cm of the anal verge); toxic megacolon; CD; history of subtotal colectomy with ileorectostomy or colectomy with ileoanal pouch, Koch pouch, or ileostomy for UC; planned bowel surgery within 12 weeks from visit 2; stool positive for Clostridium difficile toxin at screening; primary sclerosing cholangitis; history of gastrointestinal surgery within 8 weeks of visit 2; malignancy or underlying immunocompromised conditions; participation in another clinical trial within 30 days prior to screening; abnormal laboratory tests at screening, including white blood cell count (<3×10⁹/L), hemoglobin (<100 g/L), or liver tests; pregnancy or lactation (or a planned pregnancy within 7 months of study completion); males or females unwilling to use effective contraception for the duration of and 7 months after finishing the investigational product (the partners of male patients were required to use 2 forms of contraception, and male patients were not permitted to donate sperm); and any other patient considered unsuitable for inclusion by the investigators. Patients with exposure to the following treatments were also excluded: cyclosporine A, tacrolimus, or mycophenolate mofetil within 1 month prior to visit 2; anti-TNF-α agents within 2 months prior to visit 2 or during 5 halflives (drug elimination time), whichever was longer; leukocytapheresis or granulocytapheresis within 1 month prior to visit 2; prior exposure to any drugs that target α4β7 integrins or the mucosal addressin cell adhesion molecule pathway, including abrilumab; use of topical (rectal) aminosalicylic acid agent (e.g., mesalamine) or topical (rectal) steroid within 2 weeks prior to visit 2; intravenous or intramuscular corticosteroids from 2 weeks prior to screening and during the screening period; live attenuated vaccine within 1 month prior to Visit 2 or plans to receive any live attenuated vaccine during the study; and any antibiotics, antivirals, or antifungals for treatment of infection (intravenous within 30 days prior to visit 2, oral within 14 days prior to visit 2).

In the double-blind period, patients in the placebo and abrilumab 21 mg or 70 mg groups received the respective investigational product by SC injection on day 1 and at weeks 2, 4, and 8. In the abrilumab 210 mg group, patients received abrilumab 210 mg SC on day 1, followed by placebo SC at weeks 2, 4, and 8. In the open-label period, all patients received abrilumab 210 mg SC at week 12 and every 12 weeks until and including week 48. In all treatment groups, patients received 3 injections per dose (1 mL/syringe; total 3 mL per dose). The placebo was identical in appearance to abrilumab. All SC injections during both treatment periods were administered into different sites on the patient’s anterior abdominal wall, thigh, or upper arm.

The primary endpoint was clinical remission at week 8 defined as a total Mayo score ≤ 2 points, and with no individual subscore >1 point. The secondary and exploratory outcome variables were the following: induction of response at week 8 as assessed by the total Mayo score, whereby response is defined as a decrease ≥3 points and 30% in total Mayo score compared to baseline (visit 1), and with an accompanying decrease in the subscore for rectal bleeding of ≥ 1 point or with an absolute subscore for rectal bleeding of 0 or 1; mucosal healing at week 8 as assessed by rectosigmoidoscopy, defined as an absolute subscore for rectosigmoidoscopy of 0 or 1; response at week 12 as assessed by the partial Mayo score (PMS), defined as reduction by ≥ 2 points and 25% in PMS compared to baseline (visit 1); sustained response rates at week 24 in patients who achieved response at week 12 by PMS; the safety and tolerability of abrilumab 21 mg, 70 mg, and 210 mg through 48 weeks of dosing exposure and for 28 weeks after ceasing dosing; PK evaluation (serum abrilumab concentrations); the PD effects of abrilumab on circulating lymphocytes, as measured by flow cytometry and immunophenotyping and receptor occupancy assay coupled with absolute count measurements (the IPAC assay assessed free and total α4β7 on naive CD4 T cells as well as changes in α4β7-high central memory CD4 T cell absolute counts [cells/μL]); and the proportion of patients who developed anti-drug antibodies to abrilumab.

Patient-reported outcomes (PROs) were measured using the Inflammatory Bowel Disease Questionnaire [18], the Patient Global Rating of Change, and UC PRO, and were recorded by patients on an electronic PRO device. Data were collected on electronic case report forms by the study investigators and transferred to Medidata Rave^®, a web-based data capture system.

Disease remission and response rates were estimated using the Mayo score [19], a composite index of 4 items (stool frequency, rectal bleeding, rectosigmoidoscopy findings, and physician’s global assessment). Each item is graded semi-quantitatively on a score of 0 to 3 for a maximal total score of 12. Patients recorded symptoms in the Mayo daily symptom diary.

AEs were recorded from the time of randomization to the end of the follow-up period, and serious AEs (SAEs) were recorded from the time of informed consent. AE preferred terms were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 19.1. AEs were graded using the National Cancer Institute CTCAE version 4.0, and investigators judged and recorded the likelihood of relationship of AEs to study treatment, with reference to a causality guide. Neurological AEs potentially consistent with PML were referred to an adjudication committee, which included several specialists not involved in the study. Other safety assessments included laboratory assessments, tuberculosis testing (including chest radiograph), pregnancy tests, neurological examination, electrocardiogram, and stool samples for C. difficile toxin.

No formal statistical analysis was conducted; instead, the efficacy and safety of abrilumab were assessed visually using descriptive statistics and plots. Continuous variables were summarized using descriptive statistics (n, mean, SD, minimum, median, and maximum values). The efficacy and safety analysis sets in the double-blind period included all patients who were randomized and received at least 1 dose of investigational product, and the corresponding analysis sets in the open-label period were all patients who had at least 1 dose of open-label abrilumab 210 mg treatment in the open-label period. The PK evaluation was conducted in all patients for whom PK data were available, and the PD evaluation in all patients for whom PD data were available. No formal sample size calculation was performed, but a total of 48 patients were planned for randomization. The number of patients to be recruited into each dose arm (12 patients) was chosen based on the feasibility assessment to allow comparison of Japanese data with the global phase IIb study17 before joining future international phase III studies without significant delay. Data were analyzed using the SAS^® System (SAS Institute Inc., Cary, NC, USA).

Prior to unblinding, the PD analysis set for the double-blind period and the analysis sets for the open-label and safety follow-up periods were defined, as these sets were not defined in the study protocol.

The patient disposition is shown in Fig. 2. In total, 59 patients were enrolled at 23 centers in Japan. Of these, 45 patients were randomized (abrilumab 21 mg: 11 patients, abrilumab 70 mg: 12 patients, abrilumab 210 mg: 9 patients, placebo: 13 patients). Fourteen patients were not randomized because they either did not meet the inclusion criteria or met exclusion criteria. Of the 45 randomized patients, 1 patient in the abrilumab 21 mg group did not receive the investigational product, and of the 44 patients (97.8%) dosed, 43 patients (95.6%) completed the week 8 assessment for the primary analysis and 41 patients (91.1%) completed the double-blind period. Three patients (6.7%) were discontinued from the study in the double-blind period, but all of these patients completed the safety follow-up period. The reasons for discontinuation from the double-blind period were AE, insufficient effect (placebo), and overdose (abrilumab 210 mg) in 1 patient each (2.2%). All 41 patients who completed the double-blind period were enrolled in the open-label period, of whom 26 patients (63.4%) completed the open-label period and entered into the safety follow-up period. Of these, 25 patients (61.0%) completed the safety follow-up period. Fifteen patients (36.6%) were discontinued from the study in the open-label period: 8 patients (19.5%) entered the safety follow-up period and 6 patients (14.6%) completed the safety follow-up period. Reasons for discontinuation of the open-label period included withdrawal of consent (5 patients, 12.2%), AE (4 patients, 9.8%), and other (6 patients, 14.6%) (Fig. 2).

Table 1 compares the baseline characteristics among the 4 treatment groups. The treatment groups were generally wellbalanced for demographic and patient characteristics. All patients were Asian. The mean age was 39.2 years and was somewhat higher in the abrilumab 21 mg group than the other groups. The proportion of male patients was 65.9% overall and was similar across the treatment groups. Most patients in all groups were non-smokers. Mean duration of UC was 5.29 years; this was slightly longer in the abrilumab 21 mg group (mean, 6.75 years) and somewhat shorter in the abrilumab 210 mg group (mean, 3.29 years). There were no notable differences between groups in prior medication. The most commonly used prior medication at baseline in the double-blind period was 5-aminosalicylates (88.6% of all patients). Body weight and body mass index were similar across all treatment groups.

Clinical remission at week 8 was numerically higher in the abrilumab groups (4/31 patients overall [12.9%]: 1/10 patient [10.0%] in the abrilumab 21 mg group, 2/12 patients [16.7%] in the 70 mg group, and 1/9 [11.1%] in the abrilumab 210 mg group) than the placebo group (0/13 patients) (Table 2).

Table 2 also shows the response rates at weeks 8 and 12, the mucosal healing rate at week 8, and the sustained response rates at weeks 12 and 24. The response rate at week 8 was numerically higher than placebo only in the abrilumab 210 mg group, but at week 12 was numerically higher than placebo in both the 70 mg and 210 mg groups. The mucosal healing rate was numerically higher in the abrilumab 70 mg and 210 mg groups as compared with the placebo group. The proportion of patients achieving a sustained response at both week 12 and week 24 was numerically higher in the abrilumab 210 mg group than the placebo group.

The exposure-response relationship for total Mayo score showed a trend across the treatment groups. A trend of reduction in PMS was observed in the abrilumab 70 mg and 210 mg groups at week 8 and week 12. There was no apparent treatment effect of abrilumab on stool frequency up to week 12 as compared with the placebo group, but a numerical trend of improvement in rectal bleeding was observed for abrilumab, while numerical trends of improvement in rectosigmoidoscopy and physician’s global assessment score were observed with abrilumab 70 mg and 210 mg. A treatment effect on PMS and individual Mayo subscores was not apparent with treatment with abrilumab 210 mg given every 12 weeks during the open-label period.

Serum abrilumab concentrations showed a dose-proportional increase within the investigated dose range at week 2, and between the abrilumab 21 mg group and the 70 mg group at weeks 4, 8, and 12. Abrilumab treatment led to a reduction in free and total α4β7 levels on naïve CD4+ T cells in the peripheral blood. Maximal reduction in free α4β7 levels compared to baseline was observed at the first post-dose assessment (week 2) and persisted at nearly that level through week 12. No serum antibodies to abrilumab were detected in patients receiving abrilumab.

A numerical trend of improvement in Patient Global Rating of Change was observed in the abrilumab groups compared with the placebo group, and a trend towards improvement in total Inflammatory Bowel Disease Questionnaire score and the Inflammatory Bowel Disease Questionnaire subscores of bowel systems, emotional health, systemic systems, and social function was observed in the abrilumab 210 mg group compared with the placebo group (data not shown).

Regarding UC PRO, no treatment effect of abrilumab was noted in the number of bowel movements, diarrhea, coping activities, emotional impact, and daily impact on life. The abrilumab 210 mg group showed the greatest improvement from baseline in emotional impact and daily impact in life at week 8, but without statistical testing, no significant between-group differences could be determined. Treatment effect on gastrointestinal symptoms and systemic symptoms was observed with abrilumab 210 mg compared with placebo (data not shown).

Forty-four patients received at least 1 dose of study drug and provided safety data. Median duration of exposure ranged from 287.0 to 417.5 days in the abrilumab groups and was 334.0 days in the placebo group. In the double-blind period, fewer patients in the abrilumab groups experienced at least 1 AE compared with those in the placebo group (44.4%-60.0% vs. 69.2%). An SAE was reported in 1 patient (10.0%) in the abrilumab 21 mg group (UC), while 1 patient (11.1%) in the abrilumab 210 mg group reported an SAE with severity grade 3, leading to discontinuation of abrilumab (cerebral infarction). Treatment-related AEs were reported in 2 patients (20.0%) in the abrilumab 21 mg group, 1 patient (11.1%) in the abrilumab 210 mg group, and 2 patients (15.4%) in the placebo group (Table 3).

In the open-label/safety follow-up period, the proportion of patients experiencing at least 1 AE was lower in the abrilumab groups (83.3%-90.0%) than in those who received placebo in the double-blind period (91.7%). No deaths were reported. In the open-label/safety follow-up period, SAEs were reported in 6 patients overall (14.6%). One patient (8.3%) from the original placebo group had 1 SAE of anemia and 1 SAE of chronic sinusitis, and 1 patient (8.3%) had 1 SAE of worsening UC. In patients from the original abrilumab 21 mg group, 1 patient (10.0%) had 1 SAE each of contrast media allergy, UC, polychondritis, and 6th nerve paresis. In patients from the original abrilumab 70 mg group, 2 patients (16.7%) each had 1 SAE of UC and 1 (8.3%) had an SAE of enteritis.

AEs leading to discontinuation of the study drug were reported in 2 patients (16.7%) in the 70 mg group, 1 patient (14.3%) in the 210 mg group, and 1 patient (8.3%) who received placebo in the double-blind period. Treatment-related AEs were reported in 1 to 3 patients (14.3% to 30.0%) in the abrilumab groups and 4 patients (33.3%) in the placebo group (open-label period) (Table 3).

No AEs with severity equal to or greater than grade 4 were reported in either the double-blind or open-label periods. All AEs reported in the double-blind period were CTCAE grade 1 or grade 2 in severity, except for the grade 3 AE (cerebral infarction) in 1 patient (11.1%) from the original abrilumab 210 mg group. In the open-label/safety follow-up period, AEs with CTCAE grade 3 were reported in 1 patient (10.0%) from the original abrilumab 21 mg group (worsening of UC, contrast media allergy, polychondritis, and 6th nerve paresis), 3 patients (25.0%) from the original abrilumab 70 mg group (worsening of UC and enteritis), and 3 patients (25.0%) from the original placebo group (anemia, chronic sinusitis, and UC). Of these, UC and enteritis in the patient from the original abrilumab 70 mg group and anemia in the patient from the placebo double-blind group were judged by the investigator to be related to abrilumab.

There was no PML reported by the adjudication committee during the study. There were no trends indicative of clinically important treatment-related laboratory abnormalities reported in the study.