Inflammatory Bowel Diseases Affect the Phenotype and Disease Course of Coexisting Immune-Mediated Inflammatory Diseases: A Systematic Review With Meta-Analysis

Abstract

Background

It is unclear whether inflammatory bowel diseases (IBDs) affect the phenotype and severity of co-occurring immune-mediated inflammatory diseases (IMIDs). We aimed to investigate the characteristics of IMIDs in relation to co-occurring IBD.

Methods

We conducted a systematic review of Medline and EMBASE databases from inception to September 2020. We identified studies reporting the phenotype, severity, or disease course of IMIDs among patients with or without co-occurring IBD. A meta-analysis was conducted using random effects models.

Results

The electronic search yielded 13 220 studies that we narrowed down to 73 eligible studies for full-text review, including 42 on primary sclerosing cholangitis, 12 on axial spondyloarthropathies, and 8 studies on psoriasis. In primary sclerosing cholangitis, IBD was associated with less frequent involvement of extrahepatic bile ducts (risk ratio [RR], 0.50; 95% confidence interval [CI], 0.33-0.75), longer liver transplantation–free survival (hazard ratio, 0.70; 95% CI, 0.60-0.82), and no increased risk of cholangiocarcinoma (RR, 0.88; 95% CI, 0.59-1.31). Patients with axial spondyloarthropathies and co-occurring IBD were characterized by an increased risk of dactylitis (RR, 2.06; 95% CI, 1.24-3.42), a lower Bath Ankylosing Spondylitis Radiology Index (mean difference [MD] = -2.28; 95% CI, -3.26 to -1.30), and better Schober’s test results (MD = 1.07; 95% CI, 0.64-1.49). Psoriasis and co-occurring IBD was associated with reduced disease severity (RR, 1.41; 95% CI, 1.02-1.96) and less frequent presentation in nails (RR, 0.14; 95% CI, 0.05-0.42), with no apparent impact on psoriatic arthritis (RR, 0.94; 95% CI, 0.27-3.31).

Conclusions

This systematic review with meta-analysis found IBD is associated with a distinct disease phenotype among the IMIDs investigated. Our findings emphasize the importance of multidisciplinary approaches to patients with co-occurring IMIDs and IBD.

Introduction

Inflammatory bowel diseases (IBDs), including ulcerative colitis (UC) and Crohn’s disease (CD), are chronic, immune-mediated, and disabling diseases that are associated with significant physical morbidity and impairment of health-related quality of life.^1-3 Immune-mediated inflammatory diseases (IMIDs), such as IBD, affect approximately 10% of the Western population, and the incidence is increasing globally.^4,5 Among IMIDs, IBD is unique for increasing the risk of other concomitant IMIDs, affecting around 1 in 5 patients.⁶

The co-occurrence of other IMIDs in patients with IBD is associated with an increased risk of IBD-related surgeries and the need for biologic therapies.^6,7 There is also some evidence that the phenotype and severity of UC and CD might be unique and milder in the presence of co-occurring primary sclerosing cholangitis (PSC).^7,8 Conversely, the clinical impact of IBD on co-occurring IMIDs remains largely unknown.⁹ We aimed to analyze the impact of IBD on the phenotype, severity, and outcomes of IMIDs in a systematic review and meta-analysis.

Methods

This systematic review with meta-analysis was conducted according to the MOOSE (Meta-analyses Of Observational Studies in Epidemiology) guidelines.¹⁰ The MOOSE checklist is provided in Supplementary Data Content 1. The review protocol was registered in PROSPERO (CRD42019124415) prior to study initiation.

Search Strategy

In line with the recommendations of PRISMA and AMSTAR 2,^11,12 we searched for studies published in 2 databases, Medline (PubMed) and EMBASE, through September 2020, to identify studies that described the phenotype and disease course of IMIDs in the presence of co-occurring IBD. We used combinations and variations of search terms, as described in Supplementary Data Content 2. Furthermore, additional studies were sought out in the bibliographies of the studies already found, as well as in the congress abstracts for rheumatological, dermatological, hepatological, and gastrointestinal conferences taking place between 2017 and 2020 (Figure 1).

Selection Process

In accordance with Cochrane recommendations, 2 investigators (M.A. and M.D.W.) performed independently a review of the search results based on title and abstract, with disagreements resolved through joint discussion and involvement of a third reviewer (J.B.). Full-text assessment was likewise conducted independently by the same 2 authors (M.A. and M.D.W.). The inclusion criteria were (1) adult or pediatric patients with a diagnosis of an IMID coexisting with IBD, as listed in the protocol; (2) original studies reporting the phenotype, disease characteristics, course, or prognosis of IMIDs among patients with and without co-occurring IBD; and (3) studies conducted as epidemiological studies, either as cross-sectional, prospective, or retrospective case-control or cohort studies. IMIDs were defined as diseases in which autoimmune, autoinflammatory, or inflammatory mechanisms play a vital role in the pathophysiology. The IMIDs frequently occurring alongside IBD were included in this review according to each study’s criteria for the diseases. We included studies that reported the need for certain IMID-related therapy, as we considered this a surrogate marker of severity and disease course; however, this review did not aim to evaluate the effectiveness of IMID therapies in the presence of IBD.

Studies in languages other than English with no translated versions or summary were excluded. In addition, studies assessing outcomes following liver transplantation for PSC or colectomy were excluded. If studies reported insufficient data, the authors were contacted; the studies were included only if the necessary data were provided. Abstracts without full-text publications available were considered eligible given that they were not classified as irrelevant or insufficient.

Data Extraction

We used standardized data extraction forms to collect the following items: year of study, study design, geographical region, and characteristics of the patient cohort, including the number of patients, mean age, gender distribution, smoking history, type of IMID, disease duration, and type of IBD. We extracted all effect measures of disease phenotype, localization, severity, and course from each study, and the effect measures were systematically presented in this review. When studies presented data for patients without a co-occurring IBD, these data were extracted for comparison regardless of whether the patient populations were matched.

Data Synthesis and Analysis

The meta-analysis was based on a conservative inverse-variance weighted random effect model using the DerSimonian and Laird method due to our expectation of a heterogeneous group of patients and study characteristics. Pooled proportions, pooled mean differences, and pooled ratios (risk ratios [RRs], odds ratios [ORs], and hazard ratios [HRs]) with 95% confidence intervals (CIs) were calculated whenever feasible based on 2 or more publications for each outcome.¹³ The different ratio estimates were pooled separately. We used data from patients with IMIDs without evidence of any coexisting IBD within the same cohorts as a reference in each calculation of ratios. According to the Cochrane Handbook for Systematic Reviews of Interventions, heterogeneity was assessed using Cochrane Q and I² statistics, in which I² >50% indicated substantial heterogeneity. Predefined subgroup analysis to explain statistical heterogeneity included stratification according to the type of IBD, study design, and geographical study population. Prior to study initiation, we also planned to investigate whether IBD or the IMID occurred first and the age at onset of the diseases, if these data were available. The threshold of significance was determined as P < .05. Statistical analysis was performed using Cochrane Software Review Manager (Revman 5.4) and R version 4.0.3 (R Foundation for Statistical Computing).

Risk of Bias and Study Quality Assessment

Methodological quality was assessed using the Newcastle-Ottawa Scale (NOS) for nonrandomized trials, a scale that is used exensively.¹⁴ High methodological quality was defined as a score of 7 to 9, whereas moderate- and low-quality studies were defined as having scores of 4 to 6 and 0 to 3, respectively. Nonrandomized data from clinical trials were also assessed using the NOS. The risk of publication bias was assessed statistically using Egger’s regression test, and asymmetry was illustrated via funnel plots whenever at least 10 publications were included in the analysis.

As an indirect and separate measure of the risk of bias, we investigated the findings according to whether the studies were performed in IBD-centered or IMID-centered populations.

Data Availability Statement

All data are included in the article and its Supplementary Data Content.

Patient and Public Involvement

No patients were involved in setting the research question or the outcome measures, planning the study design or interpretation, or writing up the results.

Ethical Considerations

Ethical considerations were not relevant in this systematic review of the literature.

Results

Study Selection and Characteristics

The electronic search yielded 13 220 database results and 56 743 congress abstracts, which were narrowed down to 163 studies after a screening based on study titles and abstracts. Of these, a full-text review identified 73 eligible studies, and the selection process from screening to final inclusion is detailed in Figure 1. The studies included a total of 12 891 patients with IMIDs and coexisting IBD and 339 203 with IMIDs but without IBD. The studies included described PSC (n = 42),^15-56 axial spondyloarthropathies (n = 12),^57-68 psoriasis (n = 8),^69-76 pyoderma gangrenosum (n = 4),^77-80 multiple sclerosis (n = 2),^81,82 hidradenitis suppurativa (n = 1),⁸³ autoimmune pancreatitis (n = 1),⁸⁴ autoimmune hepatitis (n = 1),⁸⁵ asthma (n = 1),⁸⁶ and eosinophilic esophagitis (n = 1).⁸⁷ The data extraction is described in Supplementary Data Content 3.

The mean NOS score was 6.6, with 37 high-quality studies and 35 studies of moderate quality (Supplementary Data Content 4). Egger’s regression test did not reveal publication bias in any of the analyses, and the respective forest plots and funnel plots are presented in Supplementary Data Content 5.

Primary Sclerosing Cholangitis

A total of 42 studies, comprising 8081 PSC patients with co-occurring IBD and 5728 PSC patients without co-occurring IBD, were included.^15-56 They were conducted as retrospective cohort studies (n = 31),^{16,17,19-26,28,30,32-34,37,40,42-46,48-56} prospective cohort studies (n = 9),^{15,18,27,29,31,35,36,41,47} and retrospective case-control studies (n = 2).^38,39 Two studies were population based.^23,56 Four studies were carried out using IBD cohorts.^29,41,43,56 The mean study quality was 6.4, including 23 (54.8%) studies of moderate methodological quality and 18 (42.9%) studies of high quality (Supplementary Data Content 4).

The meta-analysis found that PSC with co-occurring IBD was associated with significantly less frequent involvement of the extrahepatic ducts (RR, 0.50; 95% CI, 0.33-0.75; P < .01; I² = 0%) as compared with patients without co-occurring IBD. However, no differences were found in terms of intrahepatic and extrahepatic duct involvement in combination, small-duct PSC histological severity of PSC, or the risk of developing cirrhosis or features of decompensation among patients with PSC with or without co-occurring IBD (Table 1). Accordingly, the occurrence of small-duct PSC, as well as the risk of developing cirrhosis, was similar among the pediatric and adult patient populations (Supplementary Data Content 5, Figures 3 and 5).^45,54 In addition, only 1 large pediatric study investigated the occurrence of portal hypertension in PSC, finding significantly lower rates among patients with co-occurring IBD (21% vs 31%; P < .01).⁵⁴

The risk of developing malignancies was higher among patients with co-occurring IBD than among patients with isolated PSC (RR, 1.53; 95% CI, 1.01-2.32; P = .04; I² = 46%). The increased risk was ascribed to colorectal cancer (RR, 4.77; 95% CI, 2.50-9.09; P < .01; I² = 0%), but not to cholangiocarcinomas (RR, 0.88; 95% CI, 0.59-1.31; P = .54; I² = 31%), or other hepatobiliary cancers (Table 1). Patients with co-occurring IBD had a lower risk of developing autoimmune hepatitis (RR, 0.54; 95% CI, 0.45-0.65; P < .01; I² = 0%) than did patients with isolated PSC, and this also appeared to be the case among pediatric patients (Supplementary Data Content 5, Figure 10).⁵⁴ Unfortunately, no data were found for the impact of IBD on the risk of malignancies among pediatric patients with PSC. Regarding long-term PSC outcomes, having IBD was not associated with undergoing a liver transplantation (RR, 0.84; 95% CI, 0.50-1.41; P = .52; I² = 60%) or with mortality (RR, 0.74; 95% CI, 0.37-1.50), P = .40; I² = 37%). However, in a separate meta-analysis, patients with coexisting IBD experienced significantly longer liver transplantation–free survival (HR, 0.70; 95% CI, 0.60-0.82; P < .01; I² = 5%). This finding was also confirmed among pediatric patients (HR, 0.63; 95% CI, 0.47-0.85).⁵⁴ Studies performed using IBD cohorts did not report any contradictory findings in terms of the risk of cholangiocarcinomas, liver transplantation, or mortality.

Finally, 2 studies assessed the revised Mayo Risk Score, which estimates the 1-year to 4-year probability of survival of patients with PSC. While the first and smallest study did not find any difference among patients with and without IBD (-0.34 ± 0.77 vs 0.23 ± 1.26),⁵¹ a larger study found a significantly lower score among patients with co-occurring UC than among those who only had PSC (0.95 vs 1.69; P < .01), indicating that a lower score was associated with better outcomes.³²

Axial Spondyloarthropathies

A total of 12 studies, comprising 321 patients with co-occurring IBD and 8660 patients without co-occurring IBD, investigated the phenotype and prognosis of axial spondyloarthropathies, including ankylosing spondylitis (n = 8),^57,58,61-66 psoriatic arthritis (n = 1),⁶⁷ and unspecified spondyloarthropathies (n = 3).^59,60,68 Of these, 7 studies were conducted as retrospective cohort studies,^59,61-65,67 4 studies were prospective cohort studies,^57,58,60,68 and a single study was conducted as a retrospective nested case-control study.⁶⁶ Only data for adult patients were available, and only one study was conducted using an IBD cohort, with no signs of conflicting findings in the latter.⁶⁴ The mean NOS score was 6.8, with 7 studies of moderate methodological quality and 5 high-quality studies (Supplementary Data Content 4).

In the meta-analysis, co-occurring IBD was associated with an increased risk of dactylitis (RR, 2.06; 95% CI, 1.24-3.42; P < .01; I² = 0%) but not of enthesitis (Table 2). Furthermore, having IBD was associated with a significantly lower Bath Ankylosing Spondylitis Radiology Index, better Schober’s test results, a smaller fingertips-to-floor distance, and a smaller occiput-wall distance (Table 2). However, one study found an association between IBD, regardless of disease activity, and a patient’s Ankylosing Spondylitis Disease Activity Score (HR, 2.80; 95% CI, 1.43-5.52), worse physical function (HR 1.40; 95% CI, 1.09-1.80), and worse patient global well-being (HR, 1.46; 95% CI, 1.10-1.93).⁵⁷ In contrast, one study also demonstrated an association between IBD and better cervical rotation beyond 70 °C, as well as less sacroiliac and lower back pain.⁵⁹ Another recent study found a Spearman correlation between CD activity index and Bath Ankylosing Spondylitis Radiology Index (r = 0.86, P = .01) and BASFI (r = 0.89, P < .01).^61,64

Finally, axial spondyloarthropathy–related therapy was assessed in 4 studies,^60,62,65,67 and a meta-analysis showed that co-occurring IBD was associated with a higher risk of needing biologic therapies (OR = 2.45; 95% CI, 1.05-4.29; P < .01; I² = 0%). A single study also reported greater use of systemic steroids and nonbiological disease-modifying antirheumatic drugs among patients with co-occurring IBD, but the study did not report the indications for these medications.⁶⁰ Unfortunately, these studies did not investigate whether the use of disease-modifying antirheumatic drugs and biologic therapies has the potential to modulate the association between the phenotypes of IMIDs and co-occurring IBD.

Psoriasis

Eight studies, comprising 680 patients with co-occurring psoriasis and IBD and 222 279 patients with isolated psoriasis, were included in this review.^69-76 Apart from one prospective case-control study,⁷² the studies were retrospective cohort studies, including one population-based cohort study.⁷¹ Only data about adult patients were available. The mean NOS score was 7.4, with 2 studies of moderate methodological quality^75,76 and the remaining 6 of high quality.

The phenotypical presentation and severity of psoriasis with and without co-occurring IBD are summarized in Table 3. Having IBD was associated with a significantly lower risk of psoriasis in the nails (RR, 0.14; 95% CI, 0.05-0.42; P < .01; I² = 0%) but not of psoriatic arthritis (RR, 0.94; 95% CI, 0.27-3.31; P = .93; I² = 75%) when compared with patients without co-occurring IBD. However, the population-based study showed a significantly higher risk of psoriatic arthritis among patients with UC (OR = 3.20; 95% CI, 2.04-5.01). Of note, these were newer studies that stratified the phenotype of psoriasis according to treatment with anti-tumor necrosis factors (anti-TNFs). As such, our meta-analysis incorporates patients with psoriasis not induced by anti-TNFs, while data for anti-TNF–induced psoriasis were not reported. Only one study was conducted using an IBD cohort,⁷² and it confirmed a less frequent occurrence of psoriasis in the nails (IBD: n = 3 of 62 [5%] vs non-IBD: n = 22 of 62 [35%]; P < .0001) (Supplementary Data Content 5, Figures 28-30).

A meta-analysis of the severity of psoriasis demonstrated that the presence of IBD (RR, 1.41; 95% CI, 1.02-1.96; P = .04; I² = 70%), and specifically CD (RR, 1.39; 95% CI, 1.02-1.89; P = .04; I² = 47%), was associated with a milder psoriasis phenotype; however, none of the studies explained their definition of severity (Table 3). The lower risk of severe psoriasis among patients with co-occurring IBD was also demonstrated in the IBD-centered study (IBD: n = 0 of 63 [0%] vs non-IBD: n = 6 of 62 [10%]; P = .004) (Supplementary Data Content 5, Figures 31-33).⁷²

Pyoderma Gangrenosum

Four studies, with a total of 108 patients with co-occurring IBD and 24 patients without co-occurring IBD, described features of pyoderma gangrenosum.^77-80 Only data about adult patients were available. These were all retrospective cohort studies, apart from 1 randomized clinical trial,⁷⁷ and were of moderate (n = 3) to high (n = 1) methodological quality (Supplementary Data Content 4). An overview of their data is presented in Supplementary Data Content 3.

Only one study assessed the characteristics of pyoderma gangrenosum as it related to the presence (n = 12) or absence (n = 24) of co-occurring IBD⁷⁸ and did not report any specific phenotypical or treatment patterns other than a higher occurrence of arthralgia or arthritis among patients with co-occurring IBD (58.3% vs 12.5%; P < .01). The randomized clinical trial investigated the efficacy of infliximab among 18 patients with co-occurring pyoderma gangrenosum and 11 patients with idiopathic pyoderma gangrenosum and found similar response rates between the 2 groups.⁷⁷

Other IMIDs

The impact of IBD on other IMIDs, including multiple sclerosis,^81,82 autoimmune pancreatitis,⁸⁴ autoimmune hepatitis,⁸⁵ asthma,⁸⁶ eosinophilic esophagitis,⁸⁷ and hidradenitis suppurativa,⁸³ is summarized in Table 4. Apart from autoimmune hepatitis, the disease outcomes seem to be less severe in the presence of IBD as compared with patients with isolated IMIDs.

Discussion

To our knowledge, this is the first systematic review with meta-analysis to examine the association between IBD and the phenotype, severity, and outcomes of co-occuring IMIDs. We found patients with IBD to be prone to a distinct phenotype of co-occurring PSC, with longer liver transplantation–free survival. No increased mortality was observed among patients with PSC and co-occurring IBD, but patients were at increased risk of malignancies ascribed to colorectal cancers (but not to cholangiocarcinomas). While the phenotypes of axial spondyloarthropathies and psoriasis in relation to co-occurring IBD were diverse, the meta-analysis indicated either a numerically or statistically significant favorable phenotype and disease severity in the presence of IBD.

Although not yet sufficiently explored, several genetic and environmental factors have been suggested to explain the nonrandom clustering of IBD and IMIDs,⁸⁸ and these factors might also explain the phenotypical distinctness of co-occurring IMIDs. First, genome-wide associations and nonsynonymous single nucleotide polymorphism scans have revealed that overlapping genes across IBD and IMIDs are primarily associated with immunological pathways, including genes related to T cell differentiation, immune cell signaling, and the innate immune response.⁸⁹ Accordingly, patients with IBD and co-occurring IMIDs express distinct immunological features, such as HLA-DRB1 and interleukin (IL)-12B and IL-23R, which are associated with a complicated disease course in IBD and other IMIDs.^90,91 Second, the role of the gut microbiota has been established as an important factor in the pathogenesis and disease course of IBD,⁹² as well as of PSC,⁹³ axial spondyloarthropathies,⁹⁴ and psoriasis.⁹⁵ However, the causal relationship between gut microbial dysbiosis and specific phenotypical presentation or long-term disease outcome is still far from understood.

Several immune-modulating therapies, including biologic therapies, are used to treat both IBD and IMIDs, with the potential to modulate the impact of IBD on the disease course of co-occurring IMIDs. This makes sense given the therapies’ targets of important and common proinflammatory pathways. For instance, ustekinumab, which is a monoclonal antibody against the p40 subunit of IL-12 and IL-23, is approved for the treatment of moderate-to-severe UC, CD, psoriatic arthritis, ankylosing spondylitis, pyoderma gangrenosum, and psoriasis.^96,97 Unfortunately, the current literature does not provide sufficient data to uncover the potential of therapeutics used for IBD to modulate the disease course of co-occurring IMIDs. In addition, UC and CD differ in several aspects, including their pathophysiology, treatment, and epidemiological association with specific IMIDs—which have not been captured in this review—and therefore further research is needed to clarify the true impact of IBD on co-occurring IMIDs. ⁶

The importance of recognizing IBD among patients with PSC has been extensively investigated in both adult and pediatric populations. Here, we add to the body of existing evidence a unique phenotype of PSC among adult patients characterized by significantly less inflammation in the extrahepatic bile ducts, a significantly longer liver transplantation–free survival, and no increased risk of cholangiocarcinomas. Together, these features might suggest a distinct phenotype of PSC among adult patients with co-occurring IBD. After all, it is well established that IBD has a unique clinical appearance in the presence of co-occurring PSC.^7,8 A recent meta-analysis also demonstrated an association between the presence of IBD and the risk of recurrence of PSC after liver transplantation,⁹⁸ for which the risk seems to correlate with the disease activity of IBD and inversely correlate with colectomy.⁹⁸ A mechanistic model for this has been suggested in the form of a gut-liver axis involving an increased intestinal permeability due to gut inflammation, which leads to a translocation of bacteria or bacterial products entering the portal-venous system, thereby activating the immune system via dual-homing memory cells.⁹⁹

The limitations of this article include, first, the inability to conduct sensitivity analyses based on the temporal relationship between the development or onset of IBD and IMIDs, and the inability to stratify IBD systematically into CD and UC, owing to a lack of data. As an indirect assessment, we did, however, investigate whether the findings from IBD-centered cohorts differed from those from IMID-centered cohorts and found agreement across these studies, thus supporting the findings in this systematic review. Second, as patients with multiple IMIDs have more health-related contacts, this could result in nonrandom increased surveillance of these patients. While this might confound the finding of increased usage of immunomodulators, it does, nonetheless, reinforce the findings of a distinct disease phenotype of IMIDs in the presence of co-occurring IBD. A third limitation is that our observation of milder IMID phenotypes in the presence of IBD needs to be confirmed in cohorts controlling for the use of biologic therapies, which are commonly prescribed among these patients. The need for further investigation is also apparent given that very few studies indicate worse outcomes of autoimmune hepatitis and autoimmune pancreatitis among patients with co-occurring IBD, pointing toward the need for a more nuanced conclusion. In general, meta-analyses of 2 studies are acceptable; however, given their higher sensitivity to the inherent biases of the studies examined, further confirmation in large, prospective cohorts is desirable. That said, the findings in our meta-analysis about the more prevalent IMIDs were consistent across subanalyses, supporting their reliability.

This is the first systematic review and meta-analysis to investigate the impact of IBD on IMIDs, and its strengths include a large number of methodologically high-quality studies and a systematic approach to distinct IMIDs that were analyzed separately, according to our current paradigms.

Conclusions

The present study provides an important addition to the current literature about the impact of IBD on the phenotype, severity, and disease course of IMIDs. This systematic review with meta-analysis found co-occurring IBD to be associated with a distinct disease phenotype of the IMIDs investigated. This study emphasizes the importance of interdisciplinary collaboration between clinicians from different specialties to better facilitate clinical management, as well as increase our understanding of the diseases. Our findings should encourage further research into the causal relationships between IBD and IMIDs that could lead to advancements in the understanding of isolated, as well as co-occurring, IBD and IMIDs.

Acknowledgments

None.

Author Contributions

Guarantor of the article: J.B. M.A.: study concept design, systematic search, study selection, data extraction, analysis and interpretation of data, and drafting of the manuscript. M.D.W.: Study selection and critical revision of the manuscript. F.B., J.B.S., J.B.: contribution of valuable scientific content, supervision, and critical revision of the manuscript. All authors approved the final version of the manuscript, including the authorship list. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. Transparency declaration: The manuscript’s guarantor (J.B.) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

Funding

None.

Conflicts of Interest

M.A., M.D.W.: None. F.B.: Research grant from Ferring and Tillotts. This does not pertain to the research submitted here. J.B.S.: Research grants from Takeda and the Capital Region Denmark; national coordinator of studies from AbbVie, Arena Pharmaceuticals, Ely Lilly, and Boehringer Ingelheim. None of these pertain to the research submitted here. J.B.: Personal fees from AbbVie, Janssen-Cilag, Celgene, Samsung Bioepis, and Pfizer; grants and personal fees from Takeda, MSD, and Tillotts Pharma; grants from Novo Nordisk Foundation and Bristol-Meyers Squibb. None of these pertain to the research submitted here.

References