A Systematic Review on Long-Term Efficacy Outcome Measures in Crohn’s Disease Patients

Abstract

Background

The traditional long-term treatment goal of Crohn’s disease [CD] is maintenance of corticosteroid-free clinical remission. Additional treatment targets, such as biochemical, endoscopic and patient-reported remission, are advocated. The relapsing–remitting nature of CD provides a challenge to the timing of target assessment. Cross-sectional assessment at predetermined moments disregards the health state in between measurements. In this systematic review, we provide an overview of outcomes used to assess long-term efficacy in clinical trials in CD.

Methods

A systematic search of the PubMed and EMBASE databases was performed to identify clinical trials in luminal CD reporting on maintenance treatment [strategies] since 1995. Two independent reviewers selected eligible articles for full text retrieval, and assessed if long-term corticosteroid-free clinical, biochemical, endoscopic or patient-reported efficacy outcomes were used.

Results

The search resulted in 2452 hits and 82 articles were included. Clinical activity was used in 80 studies [98%] as the long-term efficacy outcome, and in 21 [26%] of these concomitant corticosteroid use was taken into account. C-reactive protein was used in 32 studies [41%], faecal calprotectin in 15 studies [18%], endoscopic activity in 34 studies [41%] and patient-reported outcome in 32 studies [39%]. In seven studies, clinical, biochemical, endoscopic activity and the patient’s perspective were measured. In most studies cross-sectional measures or multiple measurements over time were used.

Conclusion

In none of the published clinical trials in CD was sustained remission on all treatment targets reported. Cross-sectional outcomes at predetermined times were widely applied, leading to lack of information regarding sustained corticosteroid-free remission for this relapsing–remitting chronic disease.

1. Introduction

Crohn’s disease [CD] is a chronic inflammatory bowel disease characterized by a relapsing, remitting course.^1–3 The traditional treatment goal of CD is induction and maintenance of corticosteroid [CS]-free clinical remission. Conventional medical therapies—even the most modern—aiming to dampen the inflammatory and immediate disease burden, however, generally fail to improve the long-term course of CD. It has been hypothesized that current treatment modalities are insufficiently effective, suboptimally dosed or given in an inappropriate order. Indeed, patients in clinical remission may still have ongoing mucosal inflammation, which may cause disease progression.⁴ Therefore according to many peers, treatment goals should be shifted from reactive strategies based on patient symptoms towards repeated objective evaluation of inflammation and striving for endoscopic or even histological remission. Meanwhile, the patient’s perspective, elusively captured in patient-reported outcome measures, has become a major endpoint of many trials in CD patients.^1,4–7

In clinical trials, assessing the efficacy of treatment strategies and interventions for CD, a variety of outcome measurements at different time points have been used.^5,8,9 Ileocolonoscopy is broadly advocated to be the gold standard to assess mucosal inflammation. However, frequent and repeated measurements over longer periods of time are not feasible, due to the invasive nature of the bowel cleaning and the procedure. Available clinical activity indices do not adequately correlate with mucosal inflammation and, therefore, are not suited to be used as a proxy for measuring sustained endoscopic remission. Similarly, most clinical activity indices were not validated as a patient-reported outcome, and thus do not sufficiently take the patient’s perspective into account. In addition, sustained CS-free remission is warranted to prevent drug-related side effects.

Accordingly, evidence- and consensus-based therapeutic goals for treat-to-target therapies are recommended in the STRIDE-criteria, an Initiative of the International Organization for the Study of IBD [IOIBD]. These long-term targets include endoscopic healing, normalized quality of life [QoL] and absence of disability, in addition to symptomatic remission, normalization of C-reactive protein [CRP] and calprotectin. Nevertheless, prospective studies [inductive evidence] to corroborate the alleged beneficial effects of this strategy are still lacking. Similarly the Food and Drug Administration [FDA] and European Medicines Agency [EMA] recommend to measure the patient’s perspective and to use objective outcomes of disease, such as biomarkers and endoscopy, for the development and registration of new medicinal products.^5,10

When evaluating longitudinal treatment efficacy, the relapsing–remitting nature of CD posts a challenge to the timing of outcome measures. Cross-sectional outcome measurements at predetermined times may be insensitive to disease activity in between these measurements as well as for corticosteroid use and quality of life, since patients most commonly experience different health states over time. This is especially an issue in long-term outcome studies.

We performed a systematic review following PRISMA [Preferred Reporting Items for Systematic Reviews and Meta-Analyses] guidelines to provide an overview of the applied outcomes and time points used to measure long-term efficacy of maintenance [drug] therapy or treatment strategies in randomized controlled trials and the accompanying open-label extensions in CD patients. These outcomes form the backbone of currently advocated treatment strategies, in particular those striving for treat-to target, such as deep remission or mucosal healing. Our aim was to explore the extent to which existing clinical trials in CD incorporated clinical activity, biochemical activity, mucosal inflammation, the patient’s perspective, CS withdrawal and the relapsing–remitting disease course in their evaluation of long-term efficacy.

2. Methods

2.1. Search strategy for identification of studies

A structured literature search was conducted up to October 6, 2021 using PubMed and Embase to identify all randomized clinical trials and the accompanying open label extension studies reporting on long-term efficacy of [medical] interventions in patients with luminal CD. Both MeSH-terms and truncated text-words including ‘Crohn’s disease’, ‘methotrexate’, ‘purines’ [tioguanine, mercaptopurine, azathioprine, allopurinol], ‘monoclonal antibodies’ [infliximab, adalimumab, certolizumab, ustekinumab, vedolizumab, golimumab, etanercept], ‘tumor necrosis factor-alpha/antagonists and inhibitors’, ‘randomized controlled trial’ and similar keywords were used.

With regard to the emergence of new therapeutic options for CD and changes in treatment goals, the search was restricted to studies published since 1995 to ensure this review is relevant to currently used treatment strategies and interventions.

2.2. Criteria for study selection

Two authors [LMJ and RHC] independently screened the titles and abstracts from the search strategy and identified eligible studies for full text retrieval. Full texts were reviewed when articles met all of the following criteria: [1] randomized controlled trial or open label extension study; [2] English or Dutch language; [3] adult patients with luminal CD, either active or in remission; and [4] reported on long-term [i.e. ≥48 weeks] efficacy outcomes of medical maintenance therapy [including antimetabolites, immunosuppressives and biologicals], surgical interventions, treatment strategies or management algorithms.

We excluded studies assessing prebiotic and probiotic therapies, faecal microbiota transplantation, nutritional therapy and alternative medicine. Articles not assessing luminal CD disease activity, but, for example, only fistulizing CD, were excluded. Conference abstracts without complete study description and separately published post hoc analyses were excluded.

Studies not meeting these criteria were excluded from this systematic review. In case of dispute, a third reviewer [MJP] was consulted.

2.3. Data collection

Data extraction was independently performed recording the following features of the included studies: author, year of publication, aim of the study, study population, disease activity at baseline, intervention[s] under review, study duration/follow-up, and efficacy outcome measures with their definitions and timing.

We assessed whether and when long-term [i.e. >48 weeks] clinical, biochemical and endoscopic efficacy outcomes and patient-reported outcomes were used. We categorized the outcomes as clinical, biochemical, endoscopic, patient reported or composite. Composite endpoints were defined as outcome definitions that included a predefined selection of clinical, biochemical, endoscopic and/or patient-reported outcomes.

Subsequently, these outcomes were labelled as primary, secondary, other predefined or other reported. The ‘other predefined’ outcomes consisted of tertiary outcomes, exploratory outcomes and unspecified outcomes. The ‘other reported’ outcomes included post hoc analysis and outcomes reported only in the results section of an article, without being mentioned in the methods. Outcomes that were categorized as ‘other reported’ were not included in this review, as they were unspecified in the methods or not part of the initial design of the study [post hoc analysis].

Additionally, data were collected on the timing of outcome measurements, that is the moments of measurement during follow-up, the precise time points when outcomes were assessed, if specified, and the actual reported time points of outcome measurements in the results sections of the articles.

Based on the actual reported outcomes, we labelled the outcomes as ‘cross-sectional’, ‘panel’, ‘longitudinal’ [sustained, durable and per trimester], ‘time to’ or ‘Kaplan–Meier’. Cross-sectional was used for the outcomes measured and reported at just one time point. Panel was used to describe an outcome that was measured and reported at multiple points over time. The outcomes categorized as longitudinal were sustained remission, durable remission and per trimester. These were categorized in line with the definitions used in the studies. Sustained remission was characterized by remission of disease at each time point. The number of time points used to evaluate sustained remission varied from five to as few as two. Durable remission was characterized by remission during at least 80% of all visits including the last visit. Per-trimester outcomes provided an overview of trimesters spent in remission.

Additionally, data on CS use were collected. CS use was adequately taken into account when CS withdrawal was part of the outcome definition, since one of the essential treatment goals is remission without CS use. Outcomes reporting medication use were noted as clinical efficacy outcomes.

We registered the review protocol with the International Prospective Register of Systematic Reviews [PROSPERO] database [CRD42020178094 https://www.crd.york.ac.uk/prospero]. This review was reported in accordance with the PRISMA statement [see PRISMA checklist in Supplementary Material].¹¹

3. Results

The search strategy resulted in a total of 2452 hits as depicted in Figure 1. After removing duplicates and screening articles by title and abstract, 96 articles remained for full text assessment. Fourteen articles were excluded for the following reasons: the study duration was too short [four], the reported study was not concerning a randomized controlled trial or open label extension [five] and the study did not comprise luminal CD activity [two]; in addition, one article was a duplicate and for two articles the full text was not available. Finally, 82 articles were included covering randomized clinical trials or accompanying open label extensions using clinical, biochemical, endoscopic and/or patient-reported long-term outcomes to assess CD treatment efficacy.

Figure 1.

Flowchart illustrating published studies included in the systematic review.

3.1. Study characteristics

An overview of the study characteristics, including author, year of publication, study population, disease activity at baseline, intervention[s] under review, follow-up duration and efficacy outcome measures, is presented in Supplementary Table S5. In addition, the inclusion of measurement of biochemical activity using CRP or faecal calprotectin [FC], mucosal inflammation, the patient’s perspective and corticosteroid withdrawal as [part of] a predefined outcome are shown.

Included studies were published between January 1995 and October 2021 with most studies published in 2013 [9/82].The study populations varied from 19¹² to 1982¹³ patients. The majority of the studies included only CD patients, whereas eight studies also included patients with ulcerative colitis and/or other chronic diseases. Disease activity at baseline was most frequently moderate-to-severe active CD, or surgically induced remission. The median follow-up duration was 56 weeks.

In seven studies, published in 2019, 2020 and 2021, long-term clinical, biochemical and endoscopic activity, and the patient’s perspective were included. In three of these studies a CS-free status as part of the outcome was mandated. The proportion of studies using clinical, biochemical, endoscopic or patient-reported efficacy outcome measures is depicted in Figure 2.

Figure 2.

Proportion of CD studies reporting key long-term efficacy outcome measures.

3.2. Clinical outcomes

Clinical activity indices were used in 80 studies [98%] as an outcome to evaluate long-term efficacy, and in five of these studies clinical activity was part of a composite outcome and thus not separately evaluated [see Supplementary Table 1]. Clinical activity was used as [part of] a primary outcome in 42 studies [51%]. Remission, response, recurrence and therapy failure were the most frequently used categories of clinical activity outcomes. To assess these outcomes, Crohn’s Disease Activity Index [CDAI]¹⁴ was used in 63 studies [79%] and Harvey–Bradshaw Index [HBI]¹⁵ was used in 14 studies [18%].

In 20 studies, with year of publication ranging from 2000 to 2021, clinical activity was used as the only long-term outcome measure. Six of these studies were open label extensions with a follow-up duration of at least 104 weeks.

In most studies, requirements for CS use at baseline, such as obligatory tapering or continuation at a stable dose, were applicable. However, when determining clinical disease activity, in 21 studies [26%] concomitant withdrawal of all CS use was a compulsory prerequisite. In most studies no distinction was made between budesonide and prednisone.

3.3. Biochemical outcomes

Biochemical outcomes were used in 33 studies [40%], mostly as secondary or ‘other predefined’ outcome [see Supplementary Table 2]. In two studies, CRP was the primary outcome as part of a composite outcome definition.^16,17 CRP was [part of] an outcome in 32 studies [39%] and FC in 15 studies [18%]. In 12 studies [15%] both FC and CRP concentrations were used as a measure of efficacy, including one composite outcome using both FC and CRP. FC was neither used as a primary outcome nor measured at all as an outcome in the studies published before 2017.

In none of the studies was concomitant CS use taken into account when determining biochemical disease activity, so CS-free biochemical remission was not assessed. In most studies, CRP and FC were reported as panel data, with multiple [sequential] measurements over time.

3.4. Endoscopic outcomes

Endoscopic activity was evaluated in 34 studies [41%] [see Supplementary Table 3]. In 12 studies [35%], endoscopic activity was used as a primary outcome and in 20 studies [59%] as a secondary outcome. Endoscopic outcomes were reported in several forms, such as mucosal healing, endoscopic remission, endoscopic response, endoscopic recurrence or absence of macroscopically visible ulcers, using various endoscopic scores defining disease activity (e.g. Rutgeerts’ postoperative score,¹⁸ Simple Endoscopic Score for Crohn Disease [SES-CD],¹⁹ Crohn’s Disease Endoscopic Index of Severity [CDEIS]²⁰). In 17 studies patients with surgically induced remission were included, of which 15 studies [88%] had endoscopic activity as an outcome.

None of the applied endoscopic scoring systems comprised ongoing or concomitant CS use. Endoscopic activity was most usually described as a cross-sectional outcome, assessed at the end of a study period. In some cases, endoscopic outcomes were reported as panel data with a maximum of two endoscopies during follow-up.

3.5. Patient-reported outcomes

Patient-reported outcome measures [PROMs] were used in 32 studies [39%], mostly as an assessment of quality of life score, validated for IBD patients, the multidimensional Inflammatory Bowel Disease Questionnaire [IBDQ] score [see Supplementary Table 4]. Patient-reported disease activity and absence of disability were not taken into account in any study.

Again, ongoing or concomitant CS use was not part of these patient-reported outcomes. In most studies, the reported PROMs were presented as panel data with multiple measurements over time.

3.6. Composite outcomes

Composite outcomes, combining clinical, biochemical and/or endoscopic activity, were applied in 11 [16%] studies [Table 1]. In six studies,^21–26 a clinical and endoscopic composite outcome was reported. In three studies,^16,17,27 clinical and biochemical composite outcomes were used. Both clinical and endoscopic as well as biochemical and endoscopic composite outcomes were reported in one study,²⁸ while in one study, a clinical, endoscopic and biochemical composite outcome was reported.²⁹

Table 1.

Composite outcomes

Author	Year	Outcome	Outcome definition	Outcome type	Reported outcomes	CS-free*	Longitudinal
Sandborn	2020	Clinical remission 1.5/1.0 and endoscopic remission	Average daily SF of ≤1.5 and AP score of ≤1.0, with neither worse than the baseline value, SES-CD of ≤4 and a ≥2-point decrease from baseline, with no subscore >1	Secondary	Week: 0, 12/16, 52	0	Panel
		Clinical response and endoscopic response 25%	30% decrease from baseline in SF and/or AP score, with neither parameter worse than patients’ baseline score; ≥25% decrease in SES-CD from baseline	Secondary	Week: 0, 12/16, 52	0	Panel
Roblin	2020	Clinical failure	HBI ≥ 5 associated with FC > 250 µg/g OR the occurrence of adverse events requiring to stop treatment	Other predefined	Month: 0, 3, 6, 9, 12, 15, 18, 21, 24	0	Kaplan–Meier
		Time to clinical failure	See clinical failure	Other predefined	Month: 0, 3, 6, 9, 12, 15, 18, 21, 24	0	Time to
Hisamatsu	2019	Corticosteroid-free triple remission	CDAI < 150, SES-CD ≤ 2 and CRP < 0.10 mg/dl	Other predefined	Week: 52	1	Cross-sectional
Fukushima	2018	Endoscopic and/or clinical recurrence	Rutgeerts’ score i3 or i4 and/or CDAI > 150	Primary	Month: 24	0	Cross-sectional
Feagan	2018	Deep remission	CDAI < 150 and CDEIS ≤ 4 [or ≤2 for patients with isolated ileitis]	Other predefined	Week: 52	0	Cross-sectional
d’Haens	2018	Sustained corticosteroid-free clinical remission	CDAI < 150 from week 22 to 54 and no ulcers at week 54 and no surgery for bowel resection or abscess and no new fistula and complete withdrawal of glucocorticosteroïds	Primary	Week: 54	1	Sustained
Colombel	2017	Deep remission	CDAI < 150, CDEIS < 4 and no deep ulcers, absence of draining fistula, discontinuation of corticosteroids for ≥8 weeks	Secondary	Week: 48	1	Cross-sectional
		Biological remission	FC < 250 μg/g, CRP < 5 mg/L, and CDEIS < 4	Secondary	Week: 48	0	Cross-sectional
Regueiro	2016	Clinical recurrence	CDAI > 200 and a ≥70-point increase from baseline, and Rutgeerts’ score ≥i2 or fistula/abscess development or treatment failure [i.e. initiated prohibited CD medication, had prohibited use of CD medication, or had CD-related surgery]	Primary	Week: 76	0	Cross-sectional
		Clinical recurrence	See primary outcome	Secondary	Week: 104	0	Cross-sectional
		Endoscopic recurrence	Rutgeerts’ score ≥i2 or fistula/abscess development or treatment failure [i.e. initiated prohibited CD medication, had prohibited use of CD medication, or had CD-related surgery]	Secondary	Week: 76	0	Cross-sectional
vande Casteele	2015	Clinical and biochemical remission	HBI ≤ 4 and CRP ≤ 5 mg/L	Primary	Year: 1	0	Cross-sectional
		Durable remission	HBI ≤ 4 and CRP ≤ 5 mg/L throughout the entire randomized maintenance phase	Secondary	Until year 1	0	Durable
Savarino	2013	Endoscopic and clinical recurrence	Rutgeerts’ score i2, i3 or i4; score of 2 on the clinical recurrence grading scale proposed by Hanauer et al; CDAI > 200	Primary	Year: 2	0	Cross-sectional
Yoshida	2012	Clinical remission	CDAI ≤ 150, IOIBD score < 2, and CRP < 0.3 mg/dL	Primary	Month: 12	0	Cross-sectional
		Clinical remission	CDAI ≤ 150, IOIBD score < 2, and CRP < 0.3 mg/dL	Secondary	Month: 36	0	Cross-sectional

^*Corticosteroid use taken into account. 0: no data available on corticosteroid use during efficacy evaluations; 1: corticosteroid-free status use is part of outcome definition of efficacy evaluation; 2: corticosteroid use is reported during evaluation.

Composite outcomes were measured in studies published from 2012 onwards, and were used as primary or secondary outcomes in 8/11 [73%] studies.

Co-primary outcomes instead of composite outcomes were infrequently used. In two studies,^30,31 a clinical and endoscopic co-primary outcome were described, in one study, patient preference and clinical activity were a co-primary outcome,³² a clinical and surgical co-primary outcome were used in another study,³³ while in one study a cross-sectional and time to clinical endpoint were reported as a co-primary outcome.³⁴ Clinical remission at weeks 26 and 56 as a co-primary outcome was also applied in a study published in 2007.³⁵

3.7. Timing of outcome measurements

Five categories for the timing of the reported outcome measurements were identified in the included studies:

1. measuring just once during follow-up: cross-sectional;

2. measuring multiple times and reporting multiple outcomes: panel data;

3. requiring remission at each time point [i.e. sustained remission], remission at a few time points [i.e. durable remission], or per trimester. These we categorized as: longitudinal;

4. reporting a time to a certain outcome: time to; and

5. reporting a survival analysis based on the observed proportions: Kaplan–Meier.

The applicable studies are depicted in Figure 3. Cross-sectional measures or multiple measurements over time, so-called panel data, were most commonly identified.

Figure 3.

Overview of the timing of outcome measures in CD studies.

4. Discussion

This systematic review of the outcomes used to assess long-term efficacy in clinical trials in CD reveals that primarily clinical activity indices have been used as the efficacy outcome. The patient’s perspective, biochemical or endoscopic activity was part of a minority of studies. In none of the evaluated trials was sustained, CS-free, clinical, patient-reported and endoscopic/biochemical remission assessed.

Thus far, available trials on maintenance therapy and treatment strategy have inconsistently used well-defined long-term outcomes taking into account the relapsing–remitting nature of CD, and the backbone of currently advocated treatment strategies seems to be insufficiently supported by evidence.³⁶

A wide variety of outcomes are available to assess disease activity in CD patients. The efficacy of a treatment [strategy] depends extensively on the choice, definition and timing of outcome measures used.³⁷ The CORE-IBD initiative recently published a minimum set of core outcomes for use in randomized controlled trials of CD patients treated with pharmacological therapies. Patient-reported outcomes, quality of life, endoscopy, biomarkers and safety were considered core domains.³⁸ The use of objective markers for mucosal inflammation, such as endoscopy, imaging or proxy-markers such as CRP and FC, is also recommended by regulatory authorities to evaluate disease activity, since clinical symptoms alone correlate poorly with mucosal inflammation.¹⁰ To prevent structural bowel damage and complications, aiming for deep or endoscopic remission is currently advocated. Assessment of this treatment goal is therefore believed to be pivotal when evaluating efficacy.

In fewer than half of the included studies, biochemical or endoscopic activity was used as an outcome measure of long-term efficacy. The use of biomarkers is especially important as a marker of subclinical mucosal inflammation when endoscopic assessment is not part of the reported outcomes. However, of the 49 studies not using an endoscopic efficacy outcome, as many as 34 studies [69%] did not report a biomarker as an outcome measure. FC was not reported as an outcome before 2017.

Besides the advocated endoscopic remission, part of the treatment goals is to induce and maintain remission with as little use of CS as possible, preferably none. Most of the included studies lacked the prerequisite of a CS-free status [for a longer period of time] when measuring clinical, biochemical and endoscopic remission. Consequently, a patient might meet the terms of the positive or beneficial outcome ‘in [deep] remission’, even though the patient was not or just for 1 day free of CS use. Given the high number of CS patients in the various trials we reported here, this is not an imaginary scenario, but more a common flaw of trial design. Even though withdrawal of CS is believed to be of major clinical importance, in only 11 studies [13%] was a CS-free status as part of their primary or secondary outcome required. Additionally, reporting of CS use is important because CS use can mask symptoms and alter quality of life, consequently thwarting the evaluation of efficacy, or at least its interpretation.

In a few studies, an evaluation of the patient’s perspective of disease control [PROM] was described. Many of these PROMs were validated based on objectively measured disease activity and, thus, usually provide a patient-derived measure of clinical disease activity without discussing the patient’s subjective perceived health.³⁹ Assessment of subjective disease control and aspects of disease that matter most to patients provide valuable information, since a perception gap between physicians and patients concerning their disease state occurs frequently.⁷ Although there is increased focus on patient-centred care, subjective disease control via an adequate PROM has not been used as a primary outcome in any study. In studies in which PROMs were used, only those measuring disease-specific quality of life scores were applied and subjective control or disability was not assessed.

Furthermore, it is important to take into account the relapsing–remitting nature of CD when measuring long-term efficacy. The treatment goal is not only to manage active disease but also to modify the disease course of CD. In the majority of studies, cross-sectional outcomes or multiple measurements over time, so-called panel data, are being used, providing information about activity at the time of examination. This results in loss of information on factual efficacy between measurements, especially for patients with subclinical mucosal inflammation or sustained CS use. The frequencies of assessments decreases in longer duration studies and therefore the information bias increases.

More continuous assessment of disease state is possible by reporting sustained remission, quartiles or trimesters in remission, ‘time to’ or ‘Kaplan–Meier’ outcome measures. However, in addition to the loss of information between measurements, efficacy information is dismissed when follow-up is discontinued in the case of a relapse in ‘time to’ or ‘Kaplan–Meier’ outcome measures. The value of sustained remission depends on the number of time points included in this outcome measure, which can be as few as two. By contrast, a great number of time points included might make this outcome measure too stringent. Moreover, continuous assessment is not feasible for all outcomes, such as endoscopic remission.

Longitudinal, per-patient assessment is detectable in outcomes using durable remission, sustained remission and per-trimester analysis. In these, it is known whether an individual subject is in remission during [most of] the study period. For the other outcome categories, it is unknown whether the outcome was the same for an individual subject at different time points. Efficacy information is only provided for separate time points on a group level.

A limitation of this review is that specific efficacy outcomes measures, such as patient-reported quality of life, were not reported in the initial study reports, but were published later in a separate article, for example as a post hoc analysis. In addition, we included only studies published after the start of the biological era, such that we could focus on outcomes used to measure efficacy of current therapies.

In this review, we provide an overview of the outcomes used to measure long-term efficacy of maintenance [drug] therapy or treatment strategies in randomized controlled trials and the accompanying open label extension studies in CD patients. In only a minority of the included studies were mucosal inflammation [or a biochemical marker as a proxy], the patient’s perspective, CS withdrawal and the relapsing–remitting disease course incorporated when evaluating long-term efficacy.^21,40,41 In our opinion, it is remarkable that most treatment guidelines and drug approvals have been based on clinical trials using predominantly symptom-based outcomes, without adequate objective assessment of inflammation. We found limited data on outcomes that are considered to be of essential value to delineate the currently advocated treatment strategy for CD.

Additionally, cross-sectional outcomes at predetermined time points are widely used, leading to loss of factual information in chronic diseases with a relapsing–remitting character, such as CD. Data on longitudinal remission in patients with a relapsing–remitting disease have not been a part of, almost without exception, long-term registration studies. Moreover, in the study design of currently ongoing clinical trials in CD patients, mostly clinical activity status at pre-specified time points as primary endpoints are being used.⁴²

In conclusion, in none of the available trials has sustained remission on all treatment targets been reported, that is CS-free, clinical, biochemical and/or endoscopic, and patient-reported, preferably subjectively, remission.

Therefore, we would like to suggest for future trial design to report more detailed on periods of the above-described in between-measurement points, for example by addressing the number of quartiles during the study period in clinical, biochemical and CS-free remission supplemented with cross-sectional endoscopy and imaging to reflect tight control of mucosal inflammation and disease progression. The clinical disease activity score used should be validated and correlate as accurately as possible with endoscopy [e.g. mobile health index, MIAH], alongside a well-designed and validated tool to assess the patient’s perspective [e.g IBD-control]. Longitudinal outcomes should be incorporated to evaluate actual sustained remission. We suggest defining sustained remission as steroid-free clinical, biochemical and patient-reported remission at each time point [after induction] or as the number of quartiles/trimesters in remission [Figure 4].

Figure 4.

Suggestions for future clinical trial design—longitudinal assessment of corticosteroid-free clinical, biochemical, patient-reported and endoscopic remission. Stars represent the outcome measurements; open stars are optional measurements, depending on the outcomes used for the composite outcome. Per outcome measure, teh figure indicates which information is included [dots, lines in between, and boxes of quartiles], and which information is excluded or unknown [question marks]. Our recommended suggestion is given within the green outline. Abbreviations: FC = faecal calprotectin; IBD-C = IBD-control; mHI = mobile Health Index; MIAH = Monitor IBD at Home; Q = quartile.

Glossary

Abbreviations:

The COSMIN initiative

COnsensus-based Standards for the selection of health Measurement Instruments

CD

Crohn’s disease

CRP

C-Reactive protein

CS

Corticosteroid

EMA

European Medicines Agency

FC

Faecal Calprotectin

FDA

Food and Drug Administration

MeSH

Medical Subject Headings

QoL

Quality of life

PRISMA guidelines

The Preferred Reporting Items for Systematic reviews and Meta-Analyses

PROM

Patient Reported Outcome Measure

STRIDE-criteria

Selecting Therapeutic Targets in Inflammatory Bowel Disease-criteria

Funding

None.

Conflict of Interest

AvB has served as speaker, adviser and/or principal investigator for AbbVie, Arandal, Arena, Celgene, Ferring, Galapagos, Janssen, MSD, Pfizer, Roche, Takeda and TEVA, and received research grants from TEVA, Eurostars funding and ZonMW, all outside the submitted work. MP reports grants and non-financial support from Falk Pharma, grants from the European Commission, grants from ZONMW [Dutch national research fund], grants and non-financial support from Takeda, grants and non-financial support from Johnson and Johnson, grants and non-financial support from Abbvie, non-financial support from Ferring, non-financial support from Immunodiagnostics, and non-financial support from MSD, all outside the submitted work. LJ and RC have no conflicts of interest to disclose.

Author Contributions

LJ and MP conceived the study. LJ and RC collected the data. LJ, RC and MP analysed and interpreted the data. LJ and RC drafted the manuscript. All authors critically reviewed and revised the manuscript for intellectual content. All authors approved the final manuscript.

Data Availability

The data underlying this article are available in the article and in its online supplementary material.