Pharmacokinetics, Safety and Efficacy of Intravenous Vedolizumab in Paediatric Patients with Ulcerative Colitis or Crohn’s Disease: Results from the Phase 2 HUBBLE Study

Abstract Background and Aims To date, there are no systematic pharmacokinetic [PK] data on vedolizumab in paediatric inflammatory bowel disease [IBD]. We report results from HUBBLE, a dose-ranging, phase 2 trial evaluating the PK, safety and efficacy of intravenous vedolizumab for paediatric IBD. Methods Enrolled patients [aged 2–17 years] with moderate to severe ulcerative colitis [UC] or Crohn’s disease [CD] and body weight ≥10 kg were randomized by weight to receive low- or high-dose vedolizumab [≥30 kg, 150 or 300 mg; <30 kg, 100 or 200 mg] on Day 1 and Weeks 2, 6 and 14. Week 14 assessments included PK, clinical response and exposure–response relationship. Safety and immunogenicity were assessed. Results Randomized patients weighing ≥30 kg [UC, n = 25; CD, n = 24] and <30 kg [UC, n = 19; CD, n = 21] had a baseline mean [standard deviation] age of 13.5 [2.5] and 7.6 [3.2] years, respectively. In almost all indication and weight groups, area under the concentration curve and average concentration increased ~2-fold from low to high dose; the trough concentration was higher in each high-dose arm compared with the low-dose arms. At Week 14, clinical response occurred in 40.0–69.2% of patients with UC and 33.3–63.6% with CD in both weight groups. Clinical responders with UC generally had higher trough concentration vs non-responders, while this trend was not observed in CD. Fourteen per cent [12/88] of patients had treatment-related adverse events and 6.8% [6/88] had anti-drug antibodies. Conclusions Vedolizumab exposure increased in an approximate dose-proportional manner. No clear dose–response relationship was observed in this limited cohort. No new safety signals were identified.

VDZ on day 1, and at Weeks 2, 6, and 14 Within each indication (UC and CD), two weight groups and two VDZ dosing groups (low and high dose) In paediatric patients with UC or CD, VDZ exposure increased in an approximately dose-proportional manner between low and high VDZ doses Ulcerative colitis 16.4 (16.9)

Introduction
Paediatric ulcerative colitis [UC] and Crohn's disease [CD] share many features with the adult forms of the illnesses. However, children with UC and CD also have age-related complications related to growth, development, nutrition, bone mineral density accretion and psychosocial needs. 1 Unlike biological therapy in adults, where weight-based dosing is linear, in young children with a body weight <30 kg, a higher dose per kilogram may be required to achieve similar trough concentration levels. 2 This highlights the importance of dose-ranging studies in paediatric inflammatory bowel disease [IBD] to determine the optimal dose in all paediatric age groups. The IBD treatment landscape has improved for children and adults since the development of biological therapy worldwide. In the USA, for example, several biologics are approved by the US Food and Drug Administration [FDA] for adults, two of which are also available for children. Intravenous [IV] infliximab [REMICADE, Janssen Biotech, Inc.] was the first to receive FDA approval for adults with UC and CD, and later for paediatric patients with IBD; 3 subsequently, subcutaneous [SC] adalimumab [HUMIRA, AbbVie, Inc.] was FDAapproved for both adult and paediatric patients with UC and CD. 4 IV and SC ustekinumab [STELARA, Janssen Biotech, Inc.] are FDA-approved for use only in adult, but not paediatric patients with UC and CD [currently in phase 3 trials in paediatric patients with CD and UC]. 5 Oral tofacitinib [XELJANZ, Pfizer, Inc.] and SC golimumab [SIMPONI, Janssen Biotech] are both FDA-approved for adult patients with UC, whereas phase 3 trials in paediatric patients with UC are ongoing. 6,7 IV vedolizumab [ENTYVIO, Takeda Pharmaceuticals U.S.A., Inc.] is indicated for adults with moderately to severely active UC and CD, 8 and paediatric phase 3 trials in UC 9 and CD 10 are now underway. Real-world studies evaluating off-label use of biologics, such as ustekinumab 11,12 and vedolizumab, [13][14][15] in the paediatric population suggest positive outcomes, but the safety and efficacy findings await confirmation from clinical trial data.
In previous adult studies of IV vedolizumab, positive exposure-response relationships for clinical response and remission were demonstrated, and these relationships appeared to be steeper in patients with UC than in those with CD. 16 However, an effective and safe dose of IV vedolizumab in children has yet to be determined. Data from a previously developed population pharmacokinetic [PK] model based on vedolizumab phase 3 study results in adult patients with UC or CD were used to select the dose and weight groups evaluated in this paediatric study. 17 Here, we now report results of HUBBLE, a paediatric phase 2 study of IV vedolizumab, a gut-selective monoclonal anti-α 4 β 7 -integrin antibody evaluating PK, safety and efficacy of IV vedolizumab in paediatric patients with UC and CD.

Study population
This study enrolled male and female paediatric patients [aged 2-17 years; body weight ≥10 kg] with moderate to severe UC or CD who failed to respond or were intolerant to corticosteroids, immunomodulators and/or tumour necrosis factor [TNF] antagonists. Moderately to severely active UC was defined as a complete Mayo score of 6-12, a total of Mayo subscores of stool frequency and rectal bleeding of ≥4, and an endoscopy subscore of ≥2. 18 Moderately to severely active CD was defined as a simple endoscopic score for CD [SES-CD] of ≥7, the CD activity index [CDAI] components of average daily abdominal pain score of >1, and total number of liquid/very soft stools >10 for the 7 days before the first dose of study drug. 19 Figure 1]. An interactive web response system was used for patient randomization and for dispensing medication ID number of the investigational drug. The high and low doses were selected on the basis of simulations performed using a population PK model derived from IV vedolizumab adult phase 3 data. 17

Study endpoints and assessments 2.3.1. PK and immunogenicity
The primary endpoints evaluated were PK parameters at Week 14 including area under the serum concentration curve [AUC], average concentration [C avg ] and observed serum concentration at the end of the dosing interval [C trough ]. Blood samples for PK assessment were collected post-dose on Day 1 and at Weeks 1 and 2, between Weeks 2 and 6, at Weeks 6, 10, 14 and 22, and at the final safety visit 18 weeks after the last dose of the study.
Vedolizumab serum concentrations were measured using a validated sandwich enzyme-linked immunosorbent assay, as previously described. 17 Immunogenicity of vedolizumab was measured on the basis of the development of anti-drug antibodies [ADAs]. Blood samples were obtained within 30 min before dosing on Day 1 and at Weeks 2, 6, 14 and 22/ early termination visit, and at any unscheduled visit for a patient who experienced a serious adverse event [AE] or disease exacerbation. Serum titres of ADAs were determined using a validated drug-tolerant electrochemiluminescence assay. 21

Efficacy
Disease activity was assessed at screening and during the study for efficacy using the Mayo and paediatric UC activity index [PUCAI] scores in UC, and the CDAI, paediatric CD activity index [PCDAI] and SES-CD scores in CD. [18][19][20] For the clinical components of the scores, patients with UC or CD, or their parents/legal guardians made daily entries into a diary, and laboratory results [CD only] were collected during screening and the study period. Flexible sigmoidoscopy/colonoscopy assessed by a central reader was used to assess the endoscopic component of the Mayo score in patients with UC and the SES-CD score in patients with CD.
Secondary efficacy endpoints included the proportion of patients with UC who achieved clinical response at Week 14-defined as a reduction in complete Mayo score of ≥3 points and ≥30% reduction from baseline, with an accompanying decrease in rectal bleeding subscore of ≥1 point or absolute rectal bleeding subscore of ≤1 point-and in CD as a ≥70-point decrease from baseline in CDAI score. Exploratory

Vedolizumab exposure-response relationship
The IV vedolizumab exposure-response relationship was assessed at Week 14. This relationship was explored by evaluating C trough by clinical response in each treatment group, for each indication and weight group separately. C trough by clinical responsiveness was defined as a ≥3-point decrease and ≥30% reduction from baseline in complete Mayo score for UC, and a ≥70-point decrease from baseline in CDAI score for CD. According to paediatric measures, C trough by clinical responsiveness was defined as a ≥20-point decrease from baseline in PUCAI score for UC and ≥15-point decrease from baseline in PCDAI score for CD.

Safety
Safety assessments of all AEs and serious AEs that occurred throughout the study were conducted. AEs  The youngest patient in the ≥30 kg category was aged 8 years at enrolment. according to the treatment they were randomized to receive or treatment that they actually received, respectively. The PK set was defined as all patients who received at least one dose of study drug and had at least one measurable concentration of vedolizumab. All statistical analyses were conducted using SAS version 9.2 or higher.

Sample size calculation
The planned sample size was 80 patients including 40 with a body weight of ≥30 kg and 40 with a body weight of 10 to <30 kg. Randomization caps were implemented to ensure that the sample size for each dose regimen was a minimum of nine patients with UC and nine patients with CD per weight group. A sample size of nine patients was expected to have ≥80% power to establish 95% confidence intervals [CIs] that were between 60% and 140% of the geometric mean estimates for total clearance after IV vedolizumab administration for each dose, indication and weight group, assuming the inter-patient variability for clearance in the paediatric population was similar to that in the adult population [% coefficient of variation ≤36.6%]. The sample size was based on industry guidance and was not adjusted for anticipated dropouts. 22,23 The planned sample size was intended to allow for descriptive analysis of exposure and efficacy for each indication and dosing group.

Primary and secondary efficacy analyses
All

PK and immunogenicity analyses
Descriptive statistics were used to summarize serum concentrations of vedolizumab over time and all PK parameters according to dose and weight groups; C trough was also summarized according to response status. The PK parameters were derived by compartmental and/or non-compartmental approaches. This study was not designed or powered for hypothesis testing or comparing different groups. The proportions of patients who were confirmed as ADA positive (persistent [positive ADA on at least two consecutive samples] or transient) or negative were summarized by weight group and treatment group for each indication separately in the FAS. Missing ADA data were not imputed.

Data availability
The datasets, including the redacted study protocol, redacted statistical analysis plan and individual participants' data supporting the results reported in this article, will be made available within 3 months from initial request, to researchers who provide a methodologically sound proposal. The data will be provided after its de-identification, in compliance with applicable privacy laws, data protection, and requirements for consent and anonymization. Data are available upon request via application at https://search.vivli.org.

Study ethics and consent
The patient, or the patient's legally acceptable representative, signed and dated an informed consent/assent form and patient authorization form [if applicable] prior to entering the study. The study was approved by the Institutional Review Board [or equivalent] at each participating centre.

Study population
There were no differences in demographic characteristics between patients with UC and CD [ Table 1]. There were some apparent differences between dose groups in baseline disease characteristics but comparisons are limited by the small sample size [ Concomitant medications were those given prior or during the screening period and ongoing at informed consent and/or post-baseline/Day 1.

PK and immunogenicity
In these paediatric patients with moderately to severely active UC and CD, vedolizumab serum exposure increased in an approximately dose-proportional manner for both weight groups. Consistently higher C trough at Week 14 were observed in the high-dose groups, and Week 14 C avg increased ~2-fold over the dose range from 150 to 300 mg in patients weighing ≥30 kg and from 100 to 200 mg in patients <30 kg [ Figure 2 group] were ADApositive but no patients with CD were ADA-positive. Overall, there was no apparent trend suggesting that positive ADA status had a negative impact on PK parameters or achievement of clinical response at Week 14. However, interpretation of this is limited due to the small number of patients who were positive for ADA.
There was no trend showing that co-administration of immunomodulators was associated with a decrease in ADA rate in patients weighing ≥30 kg for both UC and CD. The impact of immunomodulators on immunogenicity in patients weighing <30 kg could not be evaluated because there was only one positive ADA case in this group.  In the evaluation of vedolizumab exposure-response, in patients with UC, generally a higher C trough was observed in patients who achieved clinical response at Week 14 compared with patients who did not achieve clinical response in each dose arm according to complete Mayo and PUCAI scores. This trend was not observed in patients with CD, based on CDAI and PCDAI scores [adult measures are shown in Figure 3 and paediatric measures in Supplementary

Safety
The overall duration of exposure to vedolizumab was a median of 99.0 days in both weight groups with a range of 98.0-100.5 days. The incidence of AEs was similar between both disease indications in both weight groups: among patients with UC and CD, at least one AE was experienced by 80.0% and 91.3% of patients weighing ≥30 kg, respectively, and 84.2% and 90.5% of patients weighing <30 kg, respectively; most AEs were considered mild and not related to study drug [  The full analysis set included all patients who received at least one dose of study drug according to the treatment they were randomized to receive. CD, Crohn's disease; CDAI, Crohn's disease activity index; CI, confidence interval; UC, ulcerative colitis; VDZ, vedolizumab. a Clinical response, based on the complete Mayo score, was defined as a reduction in complete Mayo score of ≥3 points and ≥30% reduction from baseline with an accompanying decrease in rectal bleeding subscore of ≥1 point or absolute rectal bleeding subscore of ≤1 point; calculation was based on the GEMINI approach. Any patient with missing data for determination of clinical response status was counted as a non-responder. Patients with missing data at  , all of which were mild or moderate in intensity; none were assessed as being related to treatment. None of the hypersensitivity AEs were reported as serious AEs, and no cases of anaphylaxis or infusion-related reactions were reported in either cohort. One patient with hypersensitivity reactions was persistently positive for ADA and was positive for neutralizing ADA. There was no relationship between persistent ADA and hypersensitivity reaction.

Discussion
In this study, vedolizumab PK parameters were evaluated in a paediatric UC and CD population. The exposure in paediatric patients reported in this study [C trough levels of 9.0-14.7 µg/ mL] can be considered efficacious and within the range of therapeutic levels in adult patients after standard IV induction doses, based on PK data and modelling from previous adult phase 3 studies. 17,24,25 In paediatric patients with moderately to severely active UC and CD, vedolizumab serum exposure increased in an approximately dose-proportional manner for both weight groups, and achieved concentrations similar to those of adults from the previous GEMINI trials. Consistently higher C trough concentrations at Week 14 were observed in the high-dose groups.
In addition, vedolizumab serum concentrations and derived PK parameters at Week 14 were higher for paediatric patients weighing ≥30 kg with UC compared with patients with CD, whereas this difference was not observed for patients weighing <30 kg. However, there was variability in exposure observed across the study population, which was more apparent in C trough than C avg . Indeed, the difference between these UC and CD groups was much more apparent for C trough levels, whereas C avg levels were generally more comparable, suggesting broadly similar PK characteristics between the patients with UC and CD.
The vedolizumab exposure-response relationship results showed that in patients with UC, C trough at Week 14 was generally higher in responders than in non-responders in both treatment arms for both weight groups. However, due to the small sample size, these data do not clearly show that higher dosing in patients with UC was associated with superior clinical outcomes. In patients with CD, there was no clear trend for a higher C trough in responders for both weight groups. No firm conclusion can be drawn regarding the exposure-response relationship in patients with UC or CD due to the small sample size. In patients with UC weighing ≥30 kg, response rates were generally greater in the low-dose vs highdose group, with somewhat overlapping 95% CIs, whereas the high-dose group was favoured over the low-dose group in those weighing <30 kg. This may be due to slight differences in baseline disease characteristics such as higher baseline Mayo score, Mayo endoscopic subscore and C-reactive protein level in the high-dose group of patients with UC weighing ≥30 kg, as well as the higher proportion of patients with severe baseline PUCAI scores in the high-dose group. The baseline disease characteristics in patients with UC weighing <30 kg were generally similar between dose groups. There were no clear differences in efficacy among dose groups in the patients with CD.
The current paediatric UC population in this study and adult population in the GEMINI 1 study had similar disease severity and similar prior exposure to TNF antagonist drugs at baseline. 24 However, with regard to CD, baseline CDAI scores were higher in adults [GEMINI 2] 25 but baseline median faecal calprotectin and C-reactive protein levels were higher in the children in this study, and a greater proportion of children had been previously exposed to TNFα antagonist therapy. It is important to note that CDAI is a disease activity score for adults with CD, which makes translatability to the paediatric population difficult; however, this scoring index has also been previously used for regulatory purposes.  24 In the phase 3 GEMINI 2 study, a reduction of CDAI score by ≥100 points was achieved by 31.4% of adults with CD who were evaluated at Week 6 of induction therapy [300 mg IV vedolizumab]. 25 In contrast, at Week 14 of the current study, a reduction of CDAI score by ≥100 points was achieved in 30.4% of children with CD weighing ≥30 kg [150 and 300 mg IV vedolizumab] and in 47.6% of children with CD weighing <30 kg [100 and 200 mg IV vedolizumab]. Although fixed dosing in this phase 2 paediatric study appeared to be adequate compared with the adult trials, further study is needed on maintenance dosing strategies in the paediatric population, especially in the lowest weight group of <30 kg.
Vedolizumab was generally safe and well tolerated when administered at either 150 and 300 mg in paediatric patients weighing ≥30 kg or at 100 and 200 mg in patients weighing <30 kg, with a comparable safety profile in both weight groups. The overall safety profile observed in this study is generally consistent with the known safety profile of vedolizumab in adults and with the common symptoms of IBD. [24][25][26] This study has some limitations. It was not designed or powered for comparing between-group differences in PK parameters, but for descriptive analysis of exposure and efficacy within each indication and dosing group. Any potential inaccuracies in pre-dose sampling time and in a study of this sample size would lead to variability in the PK parameters, in particular C trough levels, although C avg would be less affected. The small number of paediatric patients, especially in the youngest age group, precluded concrete interpretation of comparisons between disease indication, weight group and/ or dose group for baseline characteristics, safety and clinical response. Additionally, Mayo and CDAI scores were not designed to assess clinical response in children; however, at the time of this phase 2 trial, these measures were recommended by regulatory agencies to enable comparison of paediatric results to previous adult data. Future paediatric IBD studies will probably use paediatric measures [PUCAI and PCDAI] or the SES-CD.
In conclusion, vedolizumab serum exposure increased in an approximately dose-proportional manner in this paediatric population, and vedolizumab exposure, clinical response and safety were considered similar in children relative to adult patients in historical studies. The results from the HUBBLE study, in conjunction with adult data from the GEMINI studies, have been used to determine ideal dosing schedules for further studies in paediatric patients. institutional accounts, from AbbVie, Allergan, AstraZeneca, Janssen, Eli Lilly, Pfizer, Shire and Takeda. B.K. has received research support from Takeda. P.L., S.B., H.K., S.K., W.T. and G.R. are employees of and hold stock/options in Takeda. C.C., W.S., M.R. and R.A.L. were employees of Takeda at the time of this research.