Changes Over Time in the Lémann Index and the Inflammatory Bowel Disease Disability Index in a Prospective Cohort of Patients With Crohn’s Disease

Abstract

Background

Crohn’s disease (CD) is a progressive, destructive, and disabling disorder. Our study aimed to assess changes over time in the Lémann index (LI) and the Inflammatory Bowel Disease Disability Index (IBD-DI) in a cohort of CD patients.

Methods

This was a single-center prospective cohort study of 130 consecutive CD patients with a follow-up of at least 4 years. The LI 1 and the IBD-DI 1 questionnaires were assessed in 2016 and again between September 2020 and October 2021 (LI 2 and IBD-DI 2).

Results

Of the 130 patients with assessment of both LI 1 and IBD-DI 1, 61 had calculation of the LI 2 and 98 patients answered the IBD-DI 2 questionnaire, with a median time between the 2 evaluations of 4.2 years. The LI increased for 16 (26%), decreased for 26 (43%), and remained unchanged for 19 (31%) patients. The median LI did not change over time (9.6 vs 9.3; P = .14). Clinical disease activity was significantly associated with bowel damage progression. A high initial LI (>7.9) was not associated with CD progression (surgery, drug dose escalation, or change of biologic). The IBD-DI decreased for 59 (60.2%), increased for 37 (37.8%), and remained unchanged for 2 (2%) patients. The median IBD-DI decreased significantly over time (23.2 vs 21.4; P = .006). There was no correlation between the 2 indexes.

Conclusions

This is the first prospective cohort study assessing changes over time in both the LI and the IBD-DI in CD patients. After 4 years, the LI appeared to be stable and the IBD-DI decreased, with no correlation between the 2 indexes.

Graphical Abstract

Open in new tabDownload slide

Introduction

Crohn’s disease (CD) is a chronic, relapsing, and destructive inflammatory bowel disease (IBD). Persistent inflammation in CD can lead to structural bowel damage such as fistulas, strictures, or abscesses,¹ which are associated with an increased risk for surgery.² The disease burden is high in CD, which can affect physical, psychosocial, and emotional dimensions and impair quality of life and daily activities.³

Over the past 20 years, the treatment landscape for CD has expanded with the emergence of several effective biologics including agents blocking tumor necrosis factor α (anti-TNFα), adhesion molecules, interleukin-12/23, interleukin-23p19, and small molecule drugs.⁴ Short- and long-term treatment targets have also evolved, with the increased evidence that the treat-to-target strategies may change the disease course of CD, prevent bowel damage progression and disability, and improve the quality of life of patients.^5,6

The Lémann index (LI) is a scoring system recently developed and validated to quantify the cumulative bowel damage over time in CD by combining endoscopic and iconographic data and surgical resection history in a standardized manner.^7,8 The Inflammatory Bowel Disease Disability Index (IBD-DI) is a validated questionnaire assessing IBD-related disability and its severity.⁹ Both disability and bowel damage in patients with CD are associated with disease activity and long-standing disease.^9,10 Previously, we investigated both the LI and the IBD-DI in a cohort study of 130 CD patients and showed no correlation between functional disability and bowel damage except in the subgroup of patients with perianal location.¹¹ Few studies have evaluated the progression of bowel damage measured by the LI in longitudinal cohorts.^12-14 To date, no study has investigated the combined progression of both bowel damage and functional disability in a cohort of CD patients.

We aimed to describe the change of bowel damage, assessed by the LI, and functional disability, assessed by the IBD-DI, in a cohort of 130 unselected CD patients over 4 years, and to identify associated factors.

Methods

Selection of Patients

From August and November 2016, 130 CD patients were consecutively included at the tertiary gastroenterology department of Lille University Hospital.¹¹ All patients performed investigations for the calculation of the LI and answered the IBD-DI questionnaire during this period (IBD-DI 1 and LI 1).

These patients were followed up prospectively until October 2021 and were included in the present study if they updated the IBD-DI questionnaire (IBD-DI 2) and/or performed the additional investigations in the framework of routine care for the calculation of the LI (LI 2), both between September 2020 and October 2021. The protocol was approved by the local ethics committee (reference no. DEC21-292).

Data Collection

The following data were collected for each patient from the electronic patient’s medical record at inclusion: age, age at diagnosis, sex, disease duration, CD location and phenotype according to the Montreal classification,¹⁵ previous intestinal resection (date, location, length), smoking status, extraintestinal manifestations, previous and current IBD therapies (including combination therapy [ie, the combination of biologic and immunosuppressants]), treatment optimization, and date of introduction and discontinuation of each treatment. Clinical disease activity (assessed by the Harvey-Bradshaw index or the physician global assessment), morphologic data (magnetic resonance imaging [MRI]), and endoscopic data were prospectively collected. Laboratory variables (such as C-reactive protein [CRP], fecal calprotectin, hemoglobin, hematocrit, white blood cell count, albumin, platelet count, and mean corpuscular volume) were retrospectively collected at the 2 time points.

Evaluation of Disability Using the IBD-DI

The functional disability was assessed previously for the 130 CD patients by the IBD-DI questionnaire (IBD-DI 1),¹¹ and again at a second time point separated by at least 4 years between September 2020 and October 2021 (IBD-DI 2), during a face-to-face interview by the IBD specialist nurse.

The validated version of the IBD-DI comprises 14 items⁹: 1 for overall health, pain, regulating defecation, body image, looking after one’s health, work and education, arthralgia, number of liquid stools; and 2 for sleep and energy, affect, and interpersonal activities. Each item (except for arthralgia) was scored from 0 to 4 points (no limitation, mild limitation, moderate limitation, severe limitation, extreme limitation). The score is calculated by the following formula: score × 100/ (px4), where p represents the number of answered items. The total score ranges from 0 to 100. IBD-DI can be separated into 4 subgroups according to its severity: no disability (0-20 points), mild disability (20-35 points), moderate disability (35-50 points), and severe disability (50-100 points).⁹

The evolution of the disability was calculated by the difference between IBD-DI 2 and IBD-DI 1 (ΔIBD-DI).

Calculation of the LI

Bowel damage was assessed for each patient using the LI according to the recently published protocol.⁸ The LI is divided into 4 subscores for each organ (upper tract, small bowel, colon/rectum, and anus). Each organ is divided into segments (3 segments for the upper tract; 20 segments of 20 cm for the small bowel; 6 segments for the colon/rectum; 1 segment for the anus). The severity of bowel damage was calculated for each segment. Fistulizing lesions or strictures are rated from 0 (none) to 3 (severe). The LI ranges from 0 to 140 (theoretical maximum value corresponding to resection of the entire digestive tract) and is calculated with coefficients assigned to each organ and segment, according to the published protocol updated in 2021.⁸

The LI was calculated based on MRI, physical examination of the perineum, and colonoscopy. Surgical history was collected from surgical or pathologist reports. Additional investigations based on disease location or symptoms were collected: upper endoscopy (for upper tract involvement) or pelvic MRI (for perianal involvement). MRIs were reviewed by a radiologist (M.A.) and a gastroenterologist (L.C.). All investigations were performed within 1 year. Calculation of the LI score was performed by using Microsoft Excel from the published protocol (Supplementary Figure 1).⁸

The LI 1 assessed previously for the 130 patients¹¹ was recalculated for each patient with the updated protocol published in 2021 and the LI 2 was calculated at a second time point separated by at least 4 years. The change in the LI between the 2 time points was calculated by the difference between LI 2 and LI 1 (ΔLI).

Study Objectives

The main objective of our study was to assess changes over time in the LI and the IBD-DI in an initial cohort of 130 CD patients followed up prospectively.

Secondary objectives were (1) to identify factors associated with the improvement of disability, (2) to identify factors associated with the progression of bowel damage, (3) to study the impact of the LI on CD progression (defined as surgery, drug dose escalation, or changes of biologic for loss of response), and (4) to evaluate the correlation between the 2 indexes at the 2 time points separated by at least 4 years.

Statistical Analysis

Descriptive statistics were used to analyze baseline characteristics. Continuous data were described as median (interquartile range [IQR]) or mean ± SD. Discrete data were determined as percentages. Percentages were compared using the chi-square test. Paired t tests were used to assess evolution of quantitative variables. The Kaplan-Meier method was used to assess the occurrence of CD complications over time. Survival was compared by the log-rank test. To analyze factors associated with LI and IBD-DI change over time, we used the Mann-Whitney test for median comparisons and means of bivariate analyses of variance. A P value <.05 was considered to be statistically significant. The correlation between LI and IBD-DI was assessed by Spearman’s correlation test. Calculations were performed using NCSS 9 software, NCSS, LLC.

Results

Patients’ Characteristics

Among the 130 CD patients eligible for the study,¹¹ 101 patients had investigations for the calculation of LI 2 and/or answered the IBD-DI 2 between September 2020 and October 2021. Twenty-nine patients were excluded because of loss of follow-up (n = 14), death (n = 1), or missing data (n = 14). Of the 101 patients included in our analysis, 98 completed the IBD-DI 2 questionnaire (population DI2), 61 patients performed the examinations for the calculation of the LI 2 (population LI2), and 58 patients performed both examinations necessary for the calculation of the LI 2 and completed the IBD-DI 2 questionnaire (population DI + LI2) (Figure 1).

More than half of the 101 CD patients had a long-standing disease (>15 years), 35% had perianal involvement, and 56% had a history of resection at inclusion. A total of 40% of patients had a clinical activity at assessment 2, and 1 patient was on steroids. A total of 65% of patients were previously exposed to 1 class of biologic and 21% were exposed to more than 1, and 18% had no immunosuppressant or biologic therapy at the time of inclusion. Nine (8.9%) patients received a short course of steroids during the study period (median time on steroids 1.75 [IQR, 1 to 3] months). Baseline demographic and clinical characteristics of the different populations are presented in Table 1. There was no statistically significant difference between populations. Laboratory variables at the first and the second time points are presented in Supplementary Table 1.

Evolution of the Disability by the IBD-DI

Change of the IBD-DI

Ninety-eight of the 130 patients completed the IBD-DI questionnaire at the 2 time points (population DI2). The mean follow-up between the 2 IBD-DI questionnaires was 4.2 (IQR, 3.9 to 4.9) years. At assessment 1, 39.8% (39 patients) had no disability, 27.5% (27 patients) had mild disability, 16.3% (16 patients) had moderate disability, and 16.3% (16 patients) had severe disability. At assessment 2, 49% (48 patients) had no disability, 31.6% (31 patients) had mild disability, 10.2% (10 patients) had moderate disability, and 9.2% (9 patients) had severe disability (Figure 2A). The median IBD-DI 1 was 23.2 (IQR, 19.6 to 30.4; extreme values: 0 to 71.4) and the median IBD-DI 2 was 21.4 (IQR, 10.7 to 33.9; extreme values: 0 to 64.3), with a significant improvement in the median IBD-DI of 1.8 points (P = .006) (Figure 3A). The IBD-DI decreased in 59 (60.2%) patients, increased in 37 (37.8%) patients, and remained stable in 2 (2%) patients. The median ΔIBD-DI was −3.57, corresponding to an improvement of 3.57 points (IQR, −5.4 to +16.1; extreme values: −35.71 to 51.79).

Factors associated with the improvement of the IBD-DI

Younger median age at diagnosis (21.2 years vs 24.9 years; P = .01), dose escalation of biologic (39.3% vs 16%; P = .016), and combination therapy at the time of the IBD-DI 2 (21.3% vs 5.4%; P = .03) were significantly associated with the improvement of functional disability. In contrast, sex, CD duration, LI 1, new resection, treatment modification, steroids exposure during the study period, extradigestive manifestations, number of previous biologics, and clinical or biological activity of CD were not significantly associated with the improvement of the IBD-DI score (Table 2).

The mean CRP and fecal calprotectin levels in patients with improved or stable IBD-DI were not significantly different from patients with worsening of IBD-DI (3.1 mg/L vs 6 mg/L; P = .16; and 209 mg/kg vs 335 mg/kg; P = .8, respectively).

Progression of Bowel Damage by the LI and Associated Factors

Evolution of the LI over time

Sixty-one of the 130 (46.9%) patients had the investigations for the calculation of the LI 2 (population LI2). The investigations were performed for the majority of patients within 3 months (85%), and extended to 12 months for 9 patients (15%). All patients had entero-MRI and a perineal examination. Sixteen of the 28 patients with colonic involvement had a colonoscopy. Access to endoscopy was reduced during the COVID-19 pandemic at the time of the study resulting in missing data. We integrated fecal calprotectin to assess disease activity in patients with a history of colonic involvement at LI 1 who did not undergo endoscopy at LI 2. One of the 3 patients with a history of upper involvement had an upper endoscopy and 19 of the 23 patients with anal involvement had a pelvic MRI.

The mean follow-up between the 2 evaluations of the LI was 4.2 years. In this population of 61 patients, the median LI 1 was calculated at 9.6 (IQR, 1.3 to 17.1; extreme values: 0 to 57) with the updated protocol (Supplementary Table 2). The LI was increased in 16 (26.2%), decreased in 26 (42.6%), and remained stable in 19 (31.2%) patients (Figure 2B). The median LI 2 was 9.3 (IQR, 1.3 to 15.5; range 0 to 59), with a nonsignificant decrease of 0.3 points (P = .14) (Figure 3B). The median of the ΔLI was 0 (IQR, −1.25 to 0.1; extreme values: −21 to +25). ΔLI variations between +5 and −5 points concerned 80% of the population.

Factors associated with bowel damage progression

The only factor significantly associated with bowel damage progression was the persistence of clinical disease activity (Harvey-Bradshaw index ≥4 and/or physician global assessment >1) at the second assessment (68.8% vs 33.3%; P = .01). Particularly, disease duration, time to the introduction of a first biologic, active smoking, CD location (including anal location), phenotype, steroid exposure during the study period, and previous intestinal resection were not significantly associated with bowel damage progression (Table 3).

The mean CRP level in the LI 2 population was 8.8 ± 20.2 mg/L (range 0.3 to 106 mg/L) at assessment 1 and 5.3 ± 8 mg/L (range 1 to 46 mg/L) at assessment 2. The mean CRP level in patients with improved or stable LI was not significantly different from that in patients with a worsening of the LI (4.8 mg/L vs 6.6 mg mg/L; P = .57).

Impact of the initial LI on CD progression

The impact of the LI on CD progression was studied in the overall cohort (n = 130 patients). During the follow-up of the 130 patients, 1 patient died (severe pneumocystis on methotrexate monotherapy) and 11 (8.5%) patients were lost to follow-up. The median event-free survival was 4.1 (IQR, 1.3-4.5) years. The survival without CD progression (defined as surgery, drug dose escalation, or a change in biologic for loss of response) was 52% at 5 years (Figure 4A). Thirty-six (27.7%) patients required a change in biologic, 37 (28.5%) patients required drug dose escalation, and 9 (6.9%) patients underwent surgery. Seven patients developed a new CD complication (fistula for 2 patients and strictures for 5 patients) between the 2 assessments.

When we investigated the impact of the initial LI in the overall cohort (n = 130), CD progression was not statistically associated (relative risk ratio, 1.25; 95% confidence interval, 0.74-2.12; P = .4) with a high LI score at the first time point (defined by a LI 1 greater than the median: >7.9) (Figure 4B). Characteristics of the population according to the LI at baseline (higher or lower than 7.9) are presented in Supplementary Table 3. The proportion of patients previously exposed to 1 biologic was significantly lower in the subgroup of patients with LI ≤7.9 at baseline (47.7% vs 27.7%; P = .018). In contrast, the proportion of patients previously exposed to more than biologics were significantly higher in the subgroup of patients with LI >7.9 (n = 2 biologics: 21.5% vs 38.4%; P = .03; n ≥ 3 biologics: 7.6% vs 20%; P = .04).

Change of both the IBD-DI and LI in Population DI + LI2 and Evaluation of its Correlation Over Time

Population DI + LI2 represents the 58 patients who had an evaluation of the 2 indexes at the 2 time points. In this population, the median IBD-DI 1 was 23.2 (IQR, 14.7 to 42.4; extreme values: 0 to 71.4) and the median IBD-DI 2 was 17.0 (IQR, 10.7 to 32.1; extreme values: 0 to 50), with a significant improvement in the median IBD-DI of 6.3 points (P = .0037). The IBD-DI decreased in 37 (63.8%), increased in 20 (34.5%), and remained stable in 1 (1.7%) patient. The median of variation of the IBD-DI was −4.5 (IQR, −16.1 to 4.9; extreme values: −51.8 to 27.8).

The median LI 1 was calculated at 9.3 (IQR, 1.1 to 15.9; extreme values: 0 to 57) and the median LI 2 was 9 (IQR, 1.1 to 15.2; range 0 to 59). The LI was increased in 16 (27.6%), decreased in 24 (41.4%), and remained stable in 18 (31%) patients. The median of variation of the LI was 0 (IQR, −1.2 to 0.1; extreme values: −25 to 21.4).

At the baseline and over time, no significant correlation between the IBD-DI score and the LI was observed (ρ = 0.067, P = .24, SE ± 0.057; and ρ = 0.125, P = .27, SE ± 0.11, respectively) (Supplementary Figures 2A and 2B).

Discussion

Progression of bowel damage and disability are 2 major issues for patients with CD. The aim of current treat-to-target strategies is to prevent bowel damage and subsequent disability.⁶ The LI quantifies the cumulative bowel damage and the IBD-DI quantifies the disability. These 2 indexes have been identified as valid tools by the Selecting Endpoints for Disease-Modification Trials (SPIRIT) consensus to be used in future disease-modification trials.¹⁶ To our knowledge, our study is the only one to investigate both changes over time in the LI and the IBD-DI in a prospective cohort study of CD patients with a follow-up of more than 4 years. Overall, the present study showed stability of bowel damage and a significant reduction in disability over time, with the absence of correlation between the 2 indexes.

Although preventing functional disability is now an important objective in IBD and should become a major secondary endpoint in clinical trials,¹⁷ its reversal has been poorly explored in patients with CD. We used the IBD-DI to measure functional disability and its progression, as the IBD-disk had not been validated when the study was started. Indeed, the IBD-DI, previously validated in a French population-based study,⁹ had good responsiveness.^18-20 The IBD-DI was validated as a face-to-face questionnaire by an IBD nurse and was found to have very good interinterviewer reliability and moderate intrainterviewer reliability (the intraclass correlation coefficient was 0.91 for interobserver reliability and 0.54 for intraobserver reliability).⁹ This is the first time that the IBD-DI was used in a long survey follow-up cohort study. In our cohort of 98 patients, the median IBD-DI was mild (23.2). We observed an improvement in the IBD-DI in approximately two-thirds of the patients, with a significant decrease in the median IBD-DI of 1.8 points after 4 years at the group level. Interestingly, we show that the number of patients with moderate-to-severe disability decreased over time (32.6% and 19.4% at the first and the second assessments, respectively). While active disease was previously associated with higher disability,^9,11,18-20 we found here that clinical disease activity at the time of assessment was not associated with the progression of disability. We suggest that this result is related to the small proportion of patients in our cohort with current disease activity (39.6%) and the absence of an evaluation of cumulative disease activity. However, our results suggest that more intensive treatment strategies (dose escalation of biologics during the follow-up and current combination therapy exposure) decreased the progression of disability. Consistent with previous publications, disease duration, CD phenotype, CD location (including anal location), and prior surgery did not influence the disability.^19,21 In parallel, our study showed that active clinical disease was the unique factor significantly associated with the progression of bowel damage. These findings reinforce the importance of tight control of CD for both improving disability and preventing bowel damage.⁶

The LI is able to quantify bowel damage and assessed both reversible inflammatory parameters (such as parietal inflammation, mucosal ulcerated lesions, strictures, or fistulas) and irreversible parameters such as surgical sanction. Changes in the LI over time were assessed previously but rarely in longitudinal cohort studies or with a long follow-up. One prospective study performed on 30 patients revealed the reversibility or stability of bowel damage for patients under anti-TNF therapies.¹² In our study, the LI was prospectively assessed for 61 patients at 4 years. We rigorously applied the method used for the calculation of the LI in the validation study.⁹ We found at baseline a median LI of 9 and a median change in the LI of 0, reflecting a slow progression over time of the bowel damage. Interestingly, we showed that a high LI at baseline (>7.9) was not associated with CD progression (surgery, drug dose escalation, or change of biologic). Although previous studies have shown that the LI increases with disease duration, the accumulation of stricturing or penetrating lesions, and surgical resections,^10,11 in our study, long disease duration, the presence of fistula/abscess, stricture or anal location at baseline, and the history of surgery were not significantly associated with an increase of the LI over time. However, few patients underwent surgery during the follow-up (n = 3 in the subgroup of patients with an increase in the LI). Although the change in the LI is only available in a small proportion of patients initially eligible, we observed an improvement or a stabilization of bowel damage in almost two-thirds of patients. These results are consistent with those already published.^12,13

At present, there is no threshold value for the LI to define the progression or improvement of bowel damage. Fiorino et al¹² proposed a cutoff of +0.3 for defining bowel damage progression. In our cohort, 80% of patients had a variation in the LI between −5 and +5 over 4 years, raising the question of the clinical relevance of a small variation of the LI and highlighting the need for a standardized threshold.

Our study was the first to explore the correlation between the LI and the IBD-DI over time in CD patients. While prior studies found that the quality of life (measured by the IBD questionnaire) was well correlated with the IBD-DI¹⁹ and also with the LI,²² we found no correlation between the LI and the IBD-DI over time in the same cohort of patients (population DI + LI2). These data were concordant with preliminary results in CD¹¹ but not with studies performed on rheumatoid arthritis patients. Rheumatoid arthritis is also a chronic and progressive disorder, characterized by joint damage on radiologic examinations with a strong relation with functional disability, especially in the later stage of the disease or in less intensively treated patients.^23-25 Therefore, disability for CD patients is a complex phenomenon, which captures the multifactorial dimension of IBD, probably dependent on the phase of the disease but not on the cumulative bowel damage.

This study has several strengths. First, the main strength of our work is the rigorous methodology used for the calculation of the LI.^7,8 Additionally, all MRIs were reviewed by a single trained radiologist to calculate the LI. Second, the IBD-DI was administered face to face by a trained IBD nurse (as in the validation study),⁹ thus limiting intraobserver subjectivity and making sure that there were no missing data. However, our work has certain limitations. First, we extended the time between the examinations for the calculation of the LI to 1 year for 9 patients because of the COVID-19 pandemic at the time of the study. We integrated the value of fecal calprotectin to assess intestinal disease activity for a small proportion of patients with a history of colonic involvement at the LI 1 assessment and with missing endoscopy data (5 patients, all in steroid-free clinical remission). Second, health conditions limited the number of patients initially eligible to half that number (the pandemic led to reduced access to MRI or endoscopy and missed appointments), leading to an inherent bias of selection. This, together with the small number of patients included, explains why we assessed factors associated with changes in indexes and not prognostic factors. Third, the monocentric design of this study, carried out in a tertiary center, could lead to sampling bias. Finally, our study was not planned to analyze CD activity or fluctuations in LI or IBD-DI between these 2 points in time. On the other hand, we analyzed multiple factors that were likely to influence the progression of the LI or the IBD-DI, taking into account previous and new intestinal resections, active smoking, use of steroids, changes of biologic, and dose escalation of the drug for loss of response. Also, we collected CD-related complications between the 2 time points.

Conclusions

Our study evaluated changes in the LI and IBD-DI in a prospective cohort of patients with CD over a long period. After more than 4 years of follow-up, while the LI tended to be stable, the IBD-DI decreased significantly. Clinical disease activity was the only factor associated with bowel damage progression. These findings support the importance of tight control as a treatment target for CD patients. At baseline, a high LI did not impact the progression of CD in terms of drug dose escalation, changes of biologic, or surgery. Finally, our study found no correlation between the LI and the IBD-DI over time. Large prospective studies are needed to further address these issues and examine the impact of treatment strategies on the progression of disability and bowel damage.

Supplementary data

Supplementary data is available at Inflammatory Bowel Diseases online.

Author Contributions

Conception of the study: P.W. and L.C. Design of the study protocol: P.W., M.N., and L.C. Statistical analysis: L.C., A.L., and A.L. Analysis and interpretation of data: L.C., P.W., M.N., and A.L. P.W. and L.C. wrote the article. All authors critically revised the manuscript. The manuscript was approved by all authors.

Funding

None received.

Conflicts of Interest

M.A. has received lecture fees from Boston Scientific, Medtronic, and Takeda. M.N. has received board membership, consultancy, or lecture fees from AbbVie, Amgen, Biogen, Celltrion, Ferring, Fresenius Kabi, Galapagos, Janssen, Lilly, Mayoli-Spindler, MSD, Nordic Pharma, Pfizer, Takeda, and Viatris. P.D. has received personal fees from scientific boards, lectures, or advising activities from AbbVie, Abbott, Amgen, Biocodex, Biofortis, Biogen, BioKuris, Boehringer Ingelheim, Ferring, Fresenius, Galapagos, Intralytix, Janssen, Kitozyme, Lesaffre, MSD, Nestlé, Nogra, Norgine, Pfizer, Roquette, Sandoz, Shire, Takeda, Tillotts, Trenker, and UCB. P.W. has received personal fees from AbbVie, Ferring, Biogen, Janssen. All other authors disclose no conflicts.

Data Availability

No new data were generated or analyzed in support of this research.

References