STUDY PROTOCOL

Etrolizumab for the Treatment of Ulcerative Colitisand Crohn’s Disease: An Overview of the Phase3Clinical Program

William J. Sandborn . Severine Vermeire . Helen Tyrrell .

Azra Hassanali . Stuart Lacey . Swati Tole . Amanda R. Tatro on behalf of The Etrolizumab Global Steering

Committee

Received: January 31, 2020 / Published online: May 22, 2020� The Author(s) 2020

ABSTRACT

Introduction: Etrolizumab is a next-generationanti-integrin with dual action that targets twopathways of inflammation in the gut. A robustphase 3 clinical program in ulcerative colitis (UC)and Crohn’s disease is ongoing and will evaluatethe efficacy and safety of etrolizumab in well-de-fined patient populations in rigorous trials thatinclude direct head-to-head comparisons againstapproved anti-tumor necrosis factor alpha agents(anti-TNF). The etrolizumab phase 3 clinical pro-gram consists of six randomized controlled trials(RCTs; UC: HIBISCUS I and II, GARDENIA,LAUREL, HICKORY; Crohn’s disease:

BERGAMOT) and two open-label extension trials(OLEs; UC: COTTONWOOD; Crohn’s disease:JUNIPER) evaluating patients with moderately toseverely active UC or Crohn’s disease.Methods: In the UCRCTs, patients are randomlyassigned according to each protocol to receiveetrolizumab, adalimumab, infliximab,orplacebo.InBERGAMOT, patients are randomly assigned toreceive etrolizumab 105 mg, etrolizumab210 mg,or placebo. The primary outcomes for the UCRCTs are Mayo Clinic score-based clinicalresponse, remission, and clinical remission; forBERGAMOT, theco-primaryoutcomesareclinicalremission (based on abdominal pain and stoolfrequency) and endoscopic improvement (basedon the Simple Endoscopic Score for Crohn’s dis-ease). The OLEs will primarily assess long-termefficacy and safety. Secondary and exploratoryendpoints include endoscopy, histology, qualityof life, and biomarkers at various timepoints.Discussion: The etrolizumab phase 3 clinicalprogram is the largest andmost comprehensive ininflammatory bowel disease, enrollingmore than3000 patients. The program explores both induc-tion and maintenance regimens. HIBISCUS I andII and GARDENIA are among the first head-to-head trials in UC against an anti-TNF and are thefirst registrational trials making that comparison.This program will also help address unansweredclinical questions on evaluation of treatmenteffects and treatment selection across a range ofpatientswith varying treatmenthistories using anextensive repository of patient samples and data.

The authors William J. Sandborn and Severine Vermeirecontributed equally to this work.

Digital Features To view digital features for this articlego to https://doi.org/10.6084/m9.figshare.12006219.

W. J. SandbornUniversity of California, San Diego, La Jolla, CA,USA

S. VermeireUniversity Hospital Leuven, Leuven, Belgium

H. Tyrrell � S. LaceyRoche Products Limited, Welwyn Garden City, UK

A. Hassanali � S. Tole � A. R. Tatro (&)F. Hoffmann-La Roche AG, Global ProductDevelopment Medical Affairs, 4070 Basel,Switzerlande-mail: amanda.tatro@roche.com

Adv Ther (2020) 37:3417–3431

https://doi.org/10.1007/s12325-020-01366-2

TrialRegistration: ClinicalTrials.gov:HIBISCUS I(NCT02163759), HIBISCUS II (NCT02171429),GARDENIA (NCT02136069), LAUREL (NCT02165215), HICKORY (NCT02100696), COTTON-WOOD (NCT02118584), BERGAMOT (NCT02394028), JUNIPER (NCT02403323).

Keywords: Anti-integrin; Crohn’s disease;Etrolizumab; Inflammatory bowel disease;Ulcerative colitis

Key Summary Points

Inflammatory bowel disease, broadlyencompassing ulcerative colitis (UC) andCrohn’s disease, is a set of chronic,relapsing/remitting gastrointestinaldiseases with long-term effects onpatients’ quality of life and well-being.

Current systemic therapies, includingcorticosteroids, immunosuppressants, andanti-tumor necrosis factor alpha agents,are not effective for many patients andincrease the risk of opportunisticinfections, non-Hodgkin lymphoma, andother undesired side effects.

Etrolizumab is an investigational next-generation anti-integrin therapywith a dualaction targeting two pathways ofinflammation in the gut for the treatment ofmoderate-to-severe UC andCrohn’s disease.

Etrolizumab selectively inhibits a4b7 andaEb7 to control both trafficking ofimmune cells into the gut and theinflammatory effects on the gut lining.

The etrolizumab phase 3 clinical trialprogram is the largest ([3000 patients)and most comprehensive registrationalprogram in UC and Crohn’s disease thatwas developed not only to characterizethe safety and efficacy of etrolizumab butalso to help advance the field byaddressing unanswered clinical questionsrelated to evaluation of treatment effectsand treatment selection through anextensive repository of patient samplesand data generated.

INTRODUCTION

Inflammatory bowel disease (IBD) is a set ofchronic, relapsing/remitting gastrointestinaldiseases with long-term effects on patients’quality of life and well-being. The two majortypes of IBD are ulcerative colitis (UC) andCrohn’s disease. Commonly used systemictherapies, including corticosteroids (CS),immunosuppressants (IS), and anti-tumornecrosis factor alpha agents (anti-TNFs), are noteffective for many patients and increase the riskfor opportunistic infections, non-Hodgkinlymphoma, and other undesired side effects[1–3]. More recently, the anti-integrin vedoli-zumab was approved for the treatment ofpatients with moderate-to-severe UC or Crohn’sdisease who do not respond to conventional orTNFa inhibitor therapy, and the anti-inter-leukin (IL)-12/IL-23 antibody ustekinumab wasapproved for the treatment of moderate-to-sev-ere UC and Crohn’s disease [4, 5]. However,there is a clear need for safer treatment optionsthat limit the side effects of systemic immuno-suppression and that have long-term efficacy ina broad patient population. Clinical research toimprove the management of IBD is progressing,but even as new treatment options are beingevaluated and approved for use, significantclinical questions remain unanswered. Forexample, with the emergence of new treatmentoptions, evidence for comparative efficacy iscritical to guide registration labeling and todefine the optimal treatment paradigm. Fewcomparative efficacy trials in the postmarketingsetting have been performed [6, 7]. Efforts tooptimize treatment selection in the light ofemerging clinical evidence are confounded bydifferences and/or incomplete characterizationof clinical trial populations and the variable useof efficacy metrics/scales. The multitude ofendpoints such as patient-reported outcomes(PROs), endoscopy, histology, and biomarkersare poorly correlated, and indeed, the method-ology of evaluating these endpoints varies fromstudy to study. In addition, while short-termsymptomatic improvement is important toboth patients and physicians, evaluating long-term outcomes is critical given the chronic

3418 Adv Ther (2020) 37:3417–3431

nature of the disease. Therefore, it would beuseful to conduct a comprehensive clinicalprogram to generate data that can be used tosystematically evaluate and correlate variousshort- and long-term efficacy and disease out-comes to inform optimal treatment selection.

IBD is characterized by a dysregulation of theimmune system in genetically susceptible indi-viduals in response to commensal microbiotaand other environmental triggers [8]. Anti-in-tegrins, such as etrolizumab and vedolizumab,are a new class of agents that selectively inhibitthis lymphocyte trafficking to and within thelarge and small intestines while avoiding broad-spectrum immunosuppression. Inhibition ofthe interaction between the integrin a4b7 andits ligand mucosal vascular addressin cell adhe-sion molecule 1 has been shown to be effectivein both UC and Crohn’s disease [9, 10].

Etrolizumab is a next-generation anti-inte-grin with dual action that targets two pathwaysof inflammation in the gut. Unlike vedolizu-mab, which targets only a4b7, etrolizumabselectively inhibits a4b7 and aEb7 to controlboth trafficking of immune cells into the gutand the inflammatory effects on the gut lining(Fig. 1). Etrolizumab uniquely inhibits aEb7-ex-pressing lymphocytes residing in the gut thathave been shown to exhibit an inflammatoryphenotype in patients with UC [11, 12]. Inhi-bition of the aE integrin, therefore, targets notonly the aEb7-expressing inflammatory cellspresent in the gut mucosa before the initiationof therapy but also those aEb7-expressinginflammatory cells that may continue to trafficinto the gut mucosa via the a4b1:vascular celladhesion molecule 1 pathway that is notinhibited by b7 or a4b7 antagonists and hasbeen shown to potentially play an importantrole in Crohn’s disease [13].

Results from the phase 2 study EUCALYPTUShave demonstrated a benefit of etrolizumabtreatment over placebo in patients with mod-erate-to-severe UC [14]. A number of patientsfrom EUCALYPTUS have now received morethan 5 years of treatment with etrolizumabthrough enrollment in the open-label extension(OLE) phase 2 study SPRUCE. A robust phase 3clinical program in UC and Crohn’s disease isongoing and aims to evaluate the efficacy and

safety of etrolizumab in well-defined patientpopulations in rigorous trials that include directhead-to-head comparisons against otherapproved biologics. Given the wealth of clinicaltrial and real-world data available from the anti-TNF agents, infliximab and adalimumab werechosen as the comparators of choice for theetrolizumab clinical trial program versus newerbiologics, for which the clinical evidence andtherapeutic experience are relatively morerecent and more limited. Further, these clinicalstudies will evaluate historical clinical end-points as well as several newer endpoints.Together, these studies will not only assess theefficacy of etrolizumab but will also provide acomprehensive data set to enhance future trialdesigns for IBD by allowing better understand-ing of the performance of and associationsacross various new endpoints and identifyingstudy inclusion criteria that may facilitate bettermeasurement of treatment effects. Herein, weprovide an overview of the comprehensivephase 3 clinical program of etrolizumab in UCand Crohn’s disease.

METHODS

Study Designs

The etrolizumab phase 3 clinical program isdesigned to evaluate safety and efficacy inpatients with moderately to severely active UCor Crohn’s disease who have had inadequateresponse or intolerance to prior CS, IS, and/oranti-TNFs. The program consists of six multi-center, prospective randomized controlled trials(RCTs) and two OLE studies (Figs. 2, 3). The OLEstudies will provide many years of data, which iscritical given the chronic nature of the disease.

HIBISCUS I and II, GARDENIA, and LAURELare investigating patients with UC who are anti-TNF-naive: HIBISCUS I and II are identicalinduction trials evaluating etrolizumab head tohead against an active comparator, adali-mumab, and placebo; GARDENIA is a mainte-nance study evaluating etrolizumab against anactive comparator, infliximab; LAUREL is amaintenance trial evaluating etrolizumabagainst placebo; HICKORY is an induction/

Adv Ther (2020) 37:3417–3431 3419

maintenance trial evaluating etrolizumab versusplacebo in anti-TNF-experienced patients withUC. Patients from the five UC RCTs may be eli-gible to roll over to open-label treatment inCOTTONWOOD.

In Crohn’s disease, BERGAMOT is an induc-tion/maintenance trial evaluating anti-TNF-naive and -experienced patients. Registrationrequirements for new medications in Crohn’sdisease were in flux at the time of study con-ception; hence, BERGAMOT was designed withan exploratory cohort 1, which, in collabora-tion with the US Food and Drug Administration(FDA), European Medicines Agency (EMA), andclinical experts in IBD, helped inform endpointselection for pivotal cohort 3 [15, 16]. Eligiblepatients from BERGAMOT may be able to rollover to open-label treatment in JUNIPER.

The trials started in May 2014 and will rununtil enrollment is complete at centers acrossAsia, Australia, Europe, North America, andSouth America. All procedures involving

humans are in accordance with the standards ofall ethics committees and institutional reviewboards and with the 1964 Helsinki declarationand its later amendments or comparable ethicalstandards. Informed consent will be obtainedfrom all individual patients included in thestudy.

Patients

Key eligibility criteria for all trials are having adiagnosis of UC or Crohn’s disease for at least3 months before screening, having moderatelyto severely active disease, and having had aninadequate response or intolerance to priortreatment with CS, IS, and/or anti-TNF. In UC,moderately to severely active disease is definedas a Mayo Clinic score (MCS) of 6–12 with acentrally read endoscopy subscore C 2 and nosubscore\1. In Crohn’s disease, moderately toseverely active disease is defined by the clinicalassessment of a Crohn’s disease activity index

Fig. 1 Etrolizumab dual mechanism of action. IEL intraepithelial lymphocyte, MAdCAM-1 mucosal vascular addressin celladhesion molecule 1, VCAM-1 vascular cell adhesion molecule 1

3420 Adv Ther (2020) 37:3417–3431

(CDAI) score of 220–480, as well as a simpleendoscopic score for Crohn’s disease (SES-CD)C 7 or C 4 for isolated ileitis. The pivotalcohort 3 of BERGAMOT also requires a stoolfrequency score (SF) C 6 or SF[3 and abdom-inal pain score (AP)[ 1. Patients should remainon stable doses of oral 5-aminosalicylic acid (5-ASA), oral CS [equivalent to B 30 mg/day pred-nisone (B 20 mg/day allowed for BERGAMOT)],budesonide (B 9 mg/day), probiotics, and IS(azathioprine, 6-mercaptopurine, methotrex-ate) for the induction period; patients receivingoral CS are required to taper CS during main-tenance therapy. Essentially, tapering of CSoccurred at the end of induction with two ratebands: 5 mg/week for a CS dose of 10 mg/dayand 2.5 mg/week for a CS dose B 10 mg/day.

The taper was expected to be completed withinapproximately 8 weeks after induction and atleast 9 months before the primary endpointassessment. Complete inclusion and exclusioncriteria for each individual study will be fullydetailed in future reports.

Interventions and Assessments

In the UC studies (Fig. 2), patients randomlyassigned to etrolizumab receive 105 mg every4 weeks. Depending on the protocol, patientsmay alternatively be randomly assigned toreceive (1) adalimumab subcutaneously perapproved dosing schedule (160 mg at week 0,80 mg at week 2, and 40 mg at weeks 4, 6, and8); (2) infliximab intravenously per approved

Fig. 2 Ulcerative colitis trial designs. anti-TNF anti-tumor necrosis factor alpha agent,MCSMayo Clinic score,OLI open-label induction, RB rectal bleeding score.*Patients who achieved a C 3-point decrease and 30%

reduction in MCS and C 1-point decrease in RB or anabsolute RB of 0 or 1 are randomly assigned to themaintenance arms

Adv Ther (2020) 37:3417–3431 3421

dosing schedule (5 mg/kg at weeks 0, 2, and 6,and then every 8 weeks thereafter); or (3) pla-cebo. The consideration to use adalimumab andinfliximab as active comparators in the induc-tion and maintenance phases of these studieswas based on their approval and wide use asstandard-of-care therapy at the time these trialswere designed.

In the Crohn’s disease study (Fig. 3), patientswere randomly assigned to receive inductiontherapy for 14 weeks of (1) subcutaneouslyadministered etrolizumab 105 mg every4 weeks; (2) subcutaneously administered etro-lizumab 210 mg every 4 weeks; or (3) placeboand then, depending on induction response,maintenance dosing with (1) etrolizumab105 mg every 4 weeks or (2) placebo. A higherdose of etrolizumab was included in theCrohn’s disease induction study because of thelack of phase 2 data and in order to explore anydose-ranging effects; notably, aE expression ishigher in the ileum and proximal colon, whichis frequently involved in Crohn’s disease, rais-ing the possibility that a higher dose may bewarranted to optimize efficacy in Crohn’s dis-ease [17]. The BERGAMOT trial was initiated

when clinical trial endpoints were in transitionfrom the historical registrational endpoint ofCDAI to endoscopy and PROs. The lack of evi-dence to quantify benefit from the more recentendoscopic and patient-reported measuresmade it challenging to formulate clinical/sta-tistical assumptions for a head-to-head superi-ority or non-inferiority trial; therefore, theBERGAMOT study employed a placebocomparator.

All treatments were double blinded unlessotherwise noted; HIBISCUS I and II, GAR-DENIA, and BERGAMOT used a double-dummydesign to ensure masking of treatment or dose,respectively.

Assessments done before randomization andat regular intervals throughout the study periodincluded physical examination, neurologicassessment, electrocardiograph, hematologyand serum chemistries, serum for pharmacoki-netic analysis of etrolizumab, antidrug antibodytests, stool sample analysis, biopsies (histology),immunohistochemistry, and colonoscopy.PROs, including the Inflammatory Bowel Dis-ease Questionnaire, EuroQoL-5D, UC-PRO tool(‘‘signs and symptoms’’ and ‘‘systemic systems’’

Fig. 3 Crohn’s disease trial designs. *Patients who achieved a C 70-point reduction in Crohn’s disease activity index scorefrom baseline are again randomly assigned to the maintenance arms

3422 Adv Ther (2020) 37:3417–3431

modules), Crohn’s disease (CD)-PRO tool(‘‘signs and symptoms’’ and ‘‘systemic systems’’modules), components of CDAI (SF, AP, generalwell-being) and components of MCS [SF andrectal bleeding score (RB)], are collected dailyusing electronic PRO devices. The etrolizumabprogram uses the UC-PRO Signs and Symptoms(SS) and CD-PRO/SS, the first PRO tools toundergo a rigorous development process out-lined by the FDA, with input from patients andclinical experts [18, 19].

Outcomes

For the UC RCTs (Table 1), the primary end-points are based on clinical response (C 3-pointdecrease and 30% reduction in MCS and C 1-point decrease in RB or an absolute RB of 0 or 1),remission (MCS B 2, with individual sub-scores B 1 and an RB of 0), or clinical remission(MCS B 2 with individual subscores B 1) at thetime points indicated in the individual proto-cols. The primary outcome measures for COT-TONWOOD are long-term efficacy (based onthe partial MCS assessed at 12-week intervals),remission and endoscopic remission atweek 108, and the incidence and severity ofadverse events.

The co-primary endpoints for BERGAMOTare clinical remission (unweighted AP B 1 andSF B 3) and endoscopic improvement (C 50%reduction in SES-CD from baseline) at weeks 14and 66 (Table 2). The primary outcome mea-sures for JUNIPER are long-term efficacy (basedon clinical remission assessed at 12-week inter-vals), endoscopic remission [SES-CD B 4 (B 2for patients with ileal disease) with no seg-ment[1] at week 108, and the incidence andseverity of adverse events.

Secondary and exploratory endpointsinclude other endoscopic measures, histology,quality of life, safety, etrolizumab pharmacoki-netics, and biomarkers assessed at various timepoints in the respective protocols and will helpcharacterize the comprehensive efficacy andsafety profile of etrolizumab, as well as anyexposure–response correlations. Exploratoryendpoints aim to assess the relationshipbetween different endpoints, the ability to use

biomarkers to predict and assess treatmentresponse, and the potential to use baseline andassessment data to personalize care for patientswith IBD. In addition, the etrolizumab studiesevaluated the frequency and severity of allserious and nonserious adverse events thatoccurred during the conduct of the study.

Statistical Analyses

A detailed statistical analysis plan will be pro-duced for each study and finalized before theprimary analysis. Briefly, sample sizes for eachstudy indicated in Figs. 2 and 3 were calculatedto provide greater than 80% power at the two-sided 5% significance level for the primary effi-cacy endpoints based on previously observedactive and placebo rates. All formal statisticalcomparisons for categorical data will use theCochran–Mantel–Haenszel test statistic, adjust-ing for appropriate stratification factors. Con-tinuous endpoints will be analyzed using ananalysis of covariance model with the appro-priate stratification factors and the baselinevalue of the studied measure as a covariate. Forall analyses, the point estimate, 95% confidenceintervals, and P value will be reported.

STRENGTHS AND LIMITATIONS

Etrolizumab for Patients with Moderate-to-Severe UC or Crohn’s Disease

The etrolizumab phase 3 clinical program is thelargest and most comprehensive registrationalprogram in UC and Crohn’s disease and willrecruit more than 3000 patients with IBD. Thisprogram will characterize the effect of etrolizu-mab on numerous outcomes, including multi-ple clinical, endoscopic, histologic, PRO, andquality-of-life indices; an ambitious biomarkerdiscovery program is also included. An exten-sive safety database with years of data will alsoresult from the two OLE studies and will beuseful in confirming the expected favorablesafety profile of etrolizumab.

Adv Ther (2020) 37:3417–3431 3423

Table 1 Ulcerative colitis trials

Trial Treatments Patients Primary endpoint(s) Key secondaryendpoint(s)

Expectedcompletiondate

HIBISCUS I and II

induction

Etrolizumab

105 mg vs

PBO vs

ADA

Anti-TNF-

naive

Remission vs PBO at

week 10

Remission vs ADA at

week 10

July 1, 2020,

and May 25,

2020,

respectively

GARDENIA

induction and

maintenance

Etrolizumab

105 mg vs

IFX

Anti-TNF-

naive

Both clinical response

at week 10 and

clinical remission at

week 54

Clinical response at

week 10

Sustained clinical

response at both

week 10 and

week 54

Clinical remission at

weeks 10 and 54

Sustained clinical

remission at both

week 10 and

week 54

April 28, 2020

LAUREL

maintenance

Etrolizumab

105 mg vs

PBO

Anti-TNF-

naive

Remission at week 62

among responders at

week 10

Clinical remission at

week 62 among

clinical remitters at

week 10

July 21, 2020

HICKORY

induction and

maintenance

Etrolizumab

105 mg vs

PBO

Anti-TNF-

experienced

Remission at week 14

Remission at week 66

among responders at

week 14

Clinical response at

week 14

Clinical remission at

week 66 among

responders at

week 14

July 17, 2020

3424 Adv Ther (2020) 37:3417–3431

Which Is the Optimal First-Line IBDTherapy Option?

Data from this program will help cliniciansmake informed treatment decisions to optimizepatient outcomes. Collectively, the trials studya large range of patients, including patientstreated with etrolizumab, infliximab, adali-mumab, or placebo. Anti-TNFs have been amainstay in first-line treatment of moderate-to-severe UC and Crohn’s disease; although newtreatments have been approved, cliniciansattempt to navigate treatment choice usingindirect comparisons that are quite limited giventhe disparate methodologies used across trials.To date, HIBISCUS I and II and GARDENIA willbe the first phase 3 registration trials in UC togenerate efficacy data for a new agent head tohead against adalimumab and infliximab.

Planned subgroup analyses, including strati-fication by disease severity, disease location,and treatment history, may also provideinsights into optimal treatment selection forpatients. A key population of interest is anti-

TNF-experienced patients, a difficult-to-treatpopulation with limited options and uncom-pelling clinical data to guide therapeutic deci-sions. Preliminary analyses from the open-labelinduction cohort of HICKORY have suggestedthat treatment with etrolizumab is associatedwith an improvement in clinical, biomarker,endoscopic, and histologic outcomes in thishard-to-treat population [20, 21]. The etrolizu-mab program also includes the assessment ofvarious biomarkers to characterize their relationto disease prognosis, prediction of response totherapy, and measurement of response. Collec-tively, data from this program will not onlysupport the efficacy and safety of etrolizumabbut also will provide rich data for physicians todecide when and how to best use etrolizumab.

What Is the Optimal Method to AssessDisease Activity and Discern TreatmentEffect?

The clinical trial landscape of IBD is constantlyevolving to find the optimal method to assess

Table 1 continued

Trial Treatments Patients Primary endpoint(s) Key secondaryendpoint(s)

Expectedcompletiondate

COTTONWOOD

open-label

extension

Open-label

etrolizumab

105 mg

Anti-TNF-

naive and

anti-TNF-

experienced

Long-term efficacy

(pMCS)

Remission at week 108

Endoscopic remission

at week 108

Incidence and severity

of AEs

N/A August 16,

2025

Remission is defined as MCS B 2, with individual subscores B 1 and RB of 0. Clinical remission is defined as MCS B 2with individual subscores B 1. Clinical response is defined as C 3-point decrease and 30% reduction in MCS and C 1-point decrease in RB or an absolute RB of 0 or 1ADA subcutaneously administered adalimumab (160 mg at week 0, 80 mg at week 2, and 40 mg at weeks 4, 6, and 8), AEadverse event, anti-TNF anti-tumor necrosis factor alpha agent, IFX intravenously administered infliximab (5 mg/kg atweeks 0, 2, and 6, and then every 8 weeks thereafter), MCS Mayo Clinic score, N/A not applicable, PBO placebo, pMCSpartial Mayo Clinic Score, RB rectal bleeding score

Adv Ther (2020) 37:3417–3431 3425

disease activity and discern the treatment effectof new agents. At the time the BERGAMOTstudy was initiated, there was a lack of clarity onregulatory endpoints for new agents in Crohn’sdisease. The CDAI has historically been plaguedwith high placebo remission rates, limiting itsability to discern treatment effect [22]. Prelimi-nary analyses from BERGAMOT exploratorycohort 1 observed similarly high rates of CDAI-based remission with placebo [23]. Healthauthorities have since indicated that evaluationof new agents in Crohn’s disease should includePROs and endoscopic assessment of mucosalinflammation [15]. As no validated PRO existsfor Crohn’s disease, the PRO2, an index using a

weighted composite of AP and SF based onCDAI, has been suggested for use in clinicaltrials [24]; however, initial analyses from BER-GAMOT observed similarly high PRO2 remis-sion rates with placebo, complicatingassessment of treatment effect. Unweightedmeasures of AP and SF may better representclinically meaningful improvements and pro-vide a better assessment of impact of therapy.Assessment of cohort 1 from BERGAMOT sup-ported the move from CDAI-based endpoints tothe current co-primary endpoints of clinicalremission (unweighted AP B 1 and SF B 3) andendoscopic improvement with an observeddrop in placebo rates [23].

Table 2 Crohn’s disease trials

Trial Treatments Patients Primary endpoints Key secondary endpoints Expectedcompletiondate

BERGAMOT

induction

and

maintenance

Etrolizumab

105 mg and

210 mg vs

PBO

Anti-TNF-

naive and

anti-TNF-

experienced

Clinical remission at

week 14

Clinical remission at

week 66 among

clinical responders at

week 14

Endoscopic remission

at week 14

Endoscopic remission

at week 66 among

clinical responders at

week 14

Clinical remission at week 66

among clinical remitters at

week 14

Endoscopic improvement at

week 66 among patients who

achieved endoscopic

improvement at week 14

CS-free clinical remission at

week 66

CS-free clinical remission for

24 weeks at week 66

April 27,

2021

JUNIPER

open-label

extension

Open-label

etrolizumab

105 mg

Anti-TNF-

naive and

anti-TNF-

experienced

Long-term efficacy

(clinical remission)

Endoscopic remission

at week 108

Incidence and severity

of AEs

N/A October 23,

2025

Clinical remission is defined as an unweighted AP B 1 and SF B 3. Clinical response is defined as a C 70-point reductionin CDAI from baseline. Endoscopic improvement is defined as a C 50% reduction in SES-CD from baseline. Endoscopicremission is defined as SES-CD B 4 (B 2 for patients with ileal disease) with no segment[ 1AE adverse event, AP abdominal pain score, anti-TNF anti-tumor necrosis factor alpha agent, CDAI Crohn’s disease activityindex, N/A not applicable, PBO placebo, SES-CD simple endoscopic score for Crohn’s disease, SF stool frequency score

3426 Adv Ther (2020) 37:3417–3431

In alignment with the focus on PROs byhealth authorities, Roche and Genentech, incooperation with a small consortium, haveworked to develop the UC-PRO and CD-PRO[18, 19]. The UC-PRO and CD-PRO are modularinstruments that were designed to comprehen-sively assess the signs, symptoms, and impact ofUC and Crohn’s disease, capturing the experi-ence from the perspective of the patient. Theseare the first PRO tools that have undergone arigorous development process outlined by theFDA, with input from both patients and clinicalexperts. Health authorities are currentlyreviewing this tool and are in the midst ofevaluating its use for PRO assessment; in fact,the EMA has issued a letter of support encour-aging data sharing and further studies to vali-date this novel tool [25]. The signs andsymptoms modules of the UC-PRO and CD-PRO(UC-PRO/SS and CD-PRO/SS) are included assecondary endpoints across the etrolizumabprogram—data from these can help supportwidespread use in clinical trials and in practice.

Objective assessment of inflammation is alsoimportant when discerning treatment effect.Central reading of endoscopies has emerged asthe gold standard to reduce local reader bias andreduce placebo rates. However, there is noconsensus on the optimal number of centralreaders, the inclusion of local readers, and theoptimal adjudication method. Endoscopiesfrom cohort 1 of BERGAMOT were read by alocal reader and two central readers, allowingfor the systematic evaluation of the perfor-mance characteristics of five different endo-scopy reading models. Preliminary analysis hassuggested that in Crohn’s disease, models withtwo readers provide the greatest discriminationin discerning treatment effect, and a model thatincludes at least one central reader and a localreader with consensus among readers deter-mined on a sliding scale does so with the leastrequirement for a third reader to reach con-sensus [26]. Analyses further detailing differ-ences between endoscopy read models and localversus central readers will be confirmed usingdata from BERGAMOT; these insights can notonly support the assessment of efficacy of etro-lizumab but also inform central readingmethodology.

What Clinical Trial Endpoints AreClinically Relevant?

The translation of clinical trial endpoints toguide decisions in clinical practice is key; cor-relations between clinical, endoscopic, andhistologic outcomes and patient prognosis haveyet to be fully elucidated.

Improvement in the endoscopic appearanceof the intestinal mucosa is recognized as animportant goal of therapy for its associationwith improved clinical outcomes [27]. However,specifics about this recommendation wererecently under discussion. In patients with UC,initial evidence supported that patients achiev-ing a Mayo endoscopic score (MES) of 0 or 1 hada similarly reduced rate of colectomy comparedwith those with an MES of 2 or 3 [28]. Morerecently, in a cohort with a median follow-up of48 months, a MES of 0 was associated with asignificantly lower rate of colectomy than a MESof 1 [29]. The UC RCTs of the etrolizumab pro-gram include endoscopic improvement (MES =0 or 1) and endoscopic remission (MES = 0) asendpoints. Combined with long-term enroll-ment in the COTTONWOOD OLE, these datamay provide insights into the clinical relevanceof an MES = 0 compared with MES = 1. InCrohn’s disease, the long-term impact ofachieving endoscopic and deep remission wasexplored in patients in the CALM trial [30]. Thestudy found that patients in clinical remission(CDAI\150) did not have lower rates of thecomposite endpoint of major adverse outcomesreflecting Crohn’s disease progression (newinternal fistula/abscess, stricture, perianal fis-tula/abscess, Crohn’s disease hospitalization, orCrohn’s disease surgery); however, patients whoachieved endoscopic remission [Crohn’s diseaseendoscopic index of severity (CDEIS)\4 withno deep ulcerations] or deep remission(CDAI\150, CDEIS\4 with no deep ulcera-tions, and no steroids for at least 8 weeks) weresignificantly less likely to have a major eventover time than those who did not. Similarly,BERGAMOT includes endoscopic remission askey secondary endpoint and, together withlong-term enrollment in JUNIPER OLE, willexplore the implications of endoscopic end-points on patient prognosis.

Adv Ther (2020) 37:3417–3431 3427

Emerging evidence suggests the importanceof histologic normalization, specifically theresolution of neutrophilic infiltration, as a keyendpoint that is associated with better clinicaloutcomes in patients with UC [31, 32]. A varietyof histologic scoring systems have been sug-gested for use, but evaluation of each scoringsystem has been largely limited by the smallsample sizes and/or retrospective nature of thestudies in which they have been assessed [33].Recently, newer indices have undergone formalevaluation and are in use in the etrolizumab UCtrials. The UC trials include endpoints of his-tologic remission, defined as a Nancy Histolog-ical Index (NHI) score of 0 or 1 [34]. The NHIwas selected for formal assessment as it offers aneasily interpreted index that concisely capturesthe ‘‘absence of neutrophils,’’ the most impor-tant factor to establish histologic remission, asan NHI score of 0 or 1 [35]. There will also be theopportunity to assess other scales such as theRobart’s Histopathology Index [36] and todetermine which components of the scores areassociated most strongly with long-term out-comes. Conversely, in Crohn’s disease, consen-sus about histology is much less clear owing todifferences in disease pathophysiology, presen-tation patterns of inflammation within thebowel regions (i.e., skip lesions), and theimmaturity of scoring indexes. Data fromBERGAMOT will allow for the exploration ofthese questions.

A Vast Biobank of Patient Samples

The breadth of this program will not onlycomprehensively evaluate the safety and effi-cacy of etrolizumab but can also help addressunanswered clinical questions. The etrolizumabclinical program will generate an extensiverepository of samples (blood, biopsy, and stool)from more than 3000 patients, collected longi-tudinally throughout the trials, and data gen-erated from these trials will allow for theinvestigation of additional measures of prog-nosis and response. Data from a broad patientpopulation will be available to explore the cor-relation of disease activity assessments, evaluatethe effect of various endpoints on long-term

outcomes, and validate new tools to predicttreatment outcome and improve patientprognosis.

In summary, the etrolizumab phase 3 clinicalprogram is ongoing and will comprehensivelycharacterize etrolizumab and inform its place intherapy. Although there remain many gaps inknowledge in the clinical landscape of IBD,information generated by this large programwill improve the ability of clinical trials toidentify effective treatments in IBD and willprovide evidence to help guide clinicians tooptimally care for their patients.

ACKNOWLEDGEMENTS

We thank all the investigators and patientsinvolved in the etrolizumab clinical program,without whom we would be unable to addressthese important clinical questions.

Funding. These studies and this publication,including the journal’s Rapid Service and OpenAccess fees, are funded by F. Hoffmann-LaRoche, Ltd.

Medical Writing Assistance. Third-partymedical writing support was provided byAndrew Occiano, PharmD (ApotheCom, SanFrancisco, CA), and was funded by F. Hoff-mann-La Roche, Ltd.

Authorship. All authors meet the Interna-tional Committee of Medical Journal Editors(ICMJE) criteria for authorship for this article,take responsibility for the integrity of the workas a whole, and have given their approval forthis version to be published.

Authorship Contributions. All authors(WJS, SV, HT, AH, SL, ST, ART) were members ofthe writing group and participated in develop-ment of the report, agreed on the content,reviewed drafts, and approved the final version.

Disclosures. WJS: research grants fromAtlantic Healthcare Limited, Amgen, Genen-tech, Gilead Sciences, Abbvie, Janssen, Takeda,Lilly, Celgene/Receptos; consulting fees from

3428 Adv Ther (2020) 37:3417–3431

Abbvie, Allergan, Amgen, Arena Pharmaceuti-cals, Avexegen Therapeutics, BeiGene, Boeh-ringer Ingelheim, Celgene, Celltrion, Conatus,Cosmo, Escalier Biosciences, Ferring, Forbion,Genentech, Gilead Sciences, Gossamer Bio,Incyte, Janssen, Kyowa Kirin PharmaceuticalResearch, Landos Biopharma, Lilly, OppilanPharma, Otsuka, Pfizer, Precision IBD, Progen-ity, Prometheus Laboratories, Reistone, RitterPharmaceuticals, Robarts Clinical Trials [ownedby Health Academic Research Trust (HART)],Series Therapeutics, Shire, Sienna Biopharma-ceuticals, Sigmoid Biotechnologies, Sterna Bio-logicals, Sublimity Therapeutics, Takeda,Theravance Biopharma, Tigenix, TillottsPharma, UCB Pharma, Ventyx Biosciences,Vimalan Biosciences, Vivelix Pharmaceuticals;and stock or stock options from BeiGene, Esca-lier Biosciences, Gossamer Bio, Oppilan Pharma,Precision IBD, Progenity, Ritter Pharmaceuti-cals, Ventyx Biosciences, Vimalan Biosciences;WJS spouse: consultant, stock options fromOpthotech, Progenity; employee, stock optionsfrom Oppilan Pharma, Escalier, Precision, Ven-tyx Biosciences, Vimalan Biosciences. SV: grantsfrom Abbvie, Pfizer, Takeda, Janssen; consultantfor Abbvie, Pfizer, Takeda, Janssen, Prodigest,Progenity, Mundipharma, Amgen, Hospira,Genentech/Roche, Gilead, Galapagos, Ferring,Tillotts Pharma, Arena Pharmaceuticals. HT andSL are employees of Roche. AH and ST areemployees of Genentech. ART is an employee ofGenentech and has stock options with Roche.

Compliance with Ethics Guidelines. Allprocedures involving humans are in accordancewith the standards of all ethics committees andinstitutional review boards and with the 1964Helsinki declaration and its later amendmentsor comparable ethical standards. Informedconsent will be obtained from all individualpatients included in the studies. The lists ofapplicable ethics committees will be madeavailable when the primary analyses of thesestudies are reported.

Data Availability. Data sharing is notapplicable to this article as no data sets weregenerated or analyzed during the current study.

Open Access. This article is licensed under aCreative Commons Attribution-NonCommer-cial 4.0 International License, which permitsany non-commercial use, sharing, adaptation,distribution and reproduction in any mediumor format, as long as you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons licence, andindicate if changes were made. The images orother third party material in this article areincluded in the article’s Creative Commonslicence, unless indicated otherwise in a creditline to the material. If material is not includedin the article’s Creative Commons licence andyour intended use is not permitted by statutoryregulation or exceeds the permitted use, youwill need to obtain permission directly from thecopyright holder. To view a copy of this licence,visit http://creativecommons.org/licenses/by-nc/4.0/.

REFERENCES

1. Stein RB, Hanauer SB. Comparative tolerability oftreatments for inflammatory bowel disease. DrugSaf. 2000;23(5):429–48.

2. Lemaitre M, Kirchgesner J, Rudnichi A, et al. Asso-ciation between use of thiopurines or tumornecrosis factor antagonists alone or in combinationand risk of lymphoma in patients with inflamma-tory bowel disease. JAMA. 2017;318(17):1679–86.

3. Ford AC, Peyrin-Biroulet L. Opportunistic infec-tions with anti-tumor necrosis factor-alpha therapyin inflammatory bowel disease: meta-analysis ofrandomized controlled trials. Am J Gastroenterol.2013;108(8):1268–76.

4. Sands BE, Sandborn WJ, Panaccione R et al (2019)Ustekinumab as induction and maintenance ther-apy for ulcerative colitis. N Engl J Med 381(13):1201–1214

5. Na SY, Moon W (2019) Perspectives on current andnovel treatments for inflammatory bowel disease.Gut Liver 13(6):604–616

6. Schreiber S, Peyrin-Biroulet L, Loftus EV Jr, et al.OP34 VARSITY: A double-blind, double-dummy,randomised, controlled trial of vedolizumab versusadalimumab in patients with active ulcerative col-itis. J Crohns Colitis. 2019;13(suppl_1):S612–S613.

Adv Ther (2020) 37:3417–3431 3429

7. Colombel JF, Sandborn WJ, Reinisch W, et al.Infliximab, azathioprine, or combination therapyfor Crohn’s disease. N Engl J Med. 2010;362(15):1383–95.

8. Lamb CA, O’Byrne S, Keir ME, Butcher EC. Gut-se-lective integrin-targeted therapies for inflammatorybowel disease. J Crohns Colitis. 2018;12:S653–S668.

9. Feagan BG, Rutgeerts P, Sands BE, et al. Vedolizu-mab as induction and maintenance therapy forulcerative colitis. N Engl J Med. 2013;369(8):699–710.

10. Sandborn WJ, Feagan BG, Rutgeerts P, et al. Vedo-lizumab as induction and maintenance therapy forCrohn’s disease. N Engl J Med. 2013;369(8):711–21.

11. Tew GW, Hackney JA, Gibbons D, et al. Associationbetween response to etrolizumab and expression ofintegrin aE and granzyme A in colon biopsies ofpatients with ulcerative colitis. Gastroenterology.2016;150(2):477–87.e9.

12. Lamb CA, Mansfield JC, Tew GW, et al. Aeb7 inte-grin identifies subsets of pro-inflammatory colonicCD4? T lymphocytes in ulcerative colitis. J CrohnsColitis. 2017;11(5):610–20.

13. Zundler S, Fischer A, Schillinger D, et al. Thealpha4beta1 homing pathway is essential for ilealhoming of Crohn’s disease effector T cells in vivo.Inflamm Bowel Dis. 2017;23(3):379–91.

14. Vermeire S, O’Byrne S, Keir M, et al. Etrolizumab asinduction therapy for ulcerative colitis: a ran-domised, controlled, phase 2 trial. Lancet.2014;384(9940):309–18.

15. Williet N, Sandborn WJ, Peyrin-Biroulet L. Patient-reported outcomes as primary end points in clinicaltrials of inflammatory bowel disease. Clin Gas-troenterol Hepatol. 2014;12(8):1246–56.e6.

16. European Medicines Agency. Guideline on thedevelopment of new medicinal products for thetreatment of Crohn’s disease, 15; 2016.

17. Ichikawa R, Lamb CA, Eastham-Anderson J, et al.Alphae integrin expression is increased in the ileumrelative to the colon and unaffected by inflamma-tion. J Crohns Colitis. 2018;12(10):1191–9.

18. Higgins PDR, Harding G, Leidy NK, et al. Develop-ment and validation of the Crohn’s disease patient-reported outcomes signs and symptoms (CD-PRO/SS) diary. J Patient Rep Outcomes. 2017;2(1):24.

19. Higgins PDR, Harding G, Revicki DA, et al. Devel-opment and validation of the ulcerative colitispatient-reported outcomes signs and symptoms

(UC-PRO/SS) diary. J Patient Rep Outcomes.2017;2(1):26.

20. Peyrin-Biroulet L, Rubin D, Feagan B, et al. Etrolizu-mab induction therapy improved endoscopic score,patient-reported outcomes, and inflammatorybiomarkers in patients with moderate to severe UCwho had failed TNF antagonist therapy: results fromthe HICKORY open-label induction (OLI) trial. Pre-sented at 25th Annual Meeting of the United Euro-pean Gastroenterology; October 28–November 1,2017; Barcelona, Spain.

21. Rubin DT, Feagan BG, Peyrin-Biroulet L, et al. Etro-lizumab induction therapy improves histologicoutcomes in anti-TNF-failed patients with ulcerativecolitis: results from the HICKORY open-labelinduction cohort. Gastroenterology. 2018;154:S-1366.

22. Jairath V, Zou G, Parker CE, et al. Systematic reviewwith meta-analysis: Placebo rates in induction andmaintenance trials of Crohn’s disease. AlimentPharmacol Ther. 2017;45(8):1021–42.

23. Sandborn W, Panes J, Jones J, et al. Etrolizumab asinduction therapy in moderate to serve Crohn’sdisease: results from BERGAMOT cohort 1. UnitedEur Gastroenterol J. 2017;5(8):1138–50.

24. Khanna R, Zou G, D’Haens G, et al. A retrospectiveanalysis: the development of patient reported out-comemeasures for the assessment of Crohn’s diseaseactivity.Aliment Pharmacol Ther. 2015;41(1):77–86.

25. European Medicines Agency. Letter of support forthe development of patient-reported outcomestools for use as an endpoint in inflammatory boweldisease (IBD) clinical trials. 2019. https://www.ema.europa.eu/en/documents/other/letter-support-development-patient-reported-outcomes-tools-use-endpoint-inflammatory-bowel-disease_en.pdf.Accessed 8 May 2020

26. Reinisch W, Mishkin DS, Oh YS, et al. P132 Analysisof various central endoscopy reading methodolo-gies in the BERGAMOT exploratory inductioncohort evaluating etrolizumab in Crohn’s disease.Presented at 12th Congress of the European Crohnsand Colitis Organization; February 15–17, 2017;Barcelona, Spain.

27. Peyrin-Biroulet L, Sandborn W, Sands BE, et al.Selecting therapeutic targets in inflammatory boweldisease (STRIDE): Determining therapeutic goals fortreat-to-target. Am J Gastroenterol. 2015;110(9):1324–38.

28. Colombel JF, Rutgeerts P, Reinisch W, et al. Earlymucosal healing with infliximab is associated withimproved long-term clinical outcomes in ulcerativecolitis. Gastroenterology. 2011;141(4):1194–201.

3430 Adv Ther (2020) 37:3417–3431

29. Manginot C, Baumann C, Peyrin-Biroulet L. Anendoscopic Mayo score of 0 is associated with alower risk of colectomy than a score of 1 in ulcer-ative colitis. Gut. 2015;64(7):1181–2.

30. Yzet C, Fumery M, Colombel J-F, et al. OP35Endoscopic and deep remission at 1 year preventsdisease progression in early Crohn’s disease: long-term data from CALM. J Crohns Colitis. 2019;13:S024–S25.

31. Bryant RV, Burger DC, Delo J, et al. Beyond endo-scopic mucosal healing in UC: histological remis-sion better predicts corticosteroid use andhospitalisation over 6 years of follow-up. Gut.2016;65(3):408–14.

32. Christensen B, Hanauer SB, Erlich J, et al. Histologicnormalization occurs in ulcerative colitis and isassociated with improved clinical outcomes. ClinGastroenterol Hepatol. 2017;15(10):1557–644.e1.

33. Mosli MH, Parker CE, Nelson SA, et al. Histologicscoring indices for evaluation of disease activity inulcerative colitis. Cochrane Database Syst Rev.2017;5:CD011256.

34. Marchal-Bressenot A, Salleron J, Boulagnon-RombiC, et al. Development and validation of the NancyHistological Index for UC. Gut. 2017;66(1):43–9.

35. Bryant RV, Winer S, Travis SP, Riddell RH. System-atic review: histological remission in inflammatorybowel disease. Is ’complete’ remission the newtreatment paradigm? An IOIBD initiative. J CrohnsColitis. 2014;8(12):1582–97.

36. Mosli MH, Feagan BG, Zou G, et al. Developmentand validation of a histological index for UC. Gut.2017;66(1):50–8.

Adv Ther (2020) 37:3417–3431 3431