Abstract
Background
We assessed whether differential efficacy of early combined immunosuppression (ECI) in comparison with conventional management (CM) is present in patients with Crohn disease (CD) according to disease location.
Methods
In this posthoc analysis of the Randomized Evaluation of an Algorithm for Crohn’s Treatment trial, the effect of ECI vs CM modified by disease location (isolated-colonic vs ileal-dominant) in terms of time to first complication (hospitalization, surgery, or disease-related complications—presence of a new abscess, fistula, or stricture; serious worsening of disease activity; extraintestinal manifestations) was analyzed using a marginal Cox proportional hazard model to account for cluster randomization. Factors adjusted included practice size, country, and other covariates selected in a backward logistic regression analysis with the first composition as outcome and P < 0.10.
Results
Of the 1969 patients with CD, 435 had isolated colonic CD (ECI n = 257, CM n = 178) and 1534 had ileal CD (ECI n = 817, CM n = 717). Over 24 months there was a significant differential impact for ECI vs CM for reducing the risk of a CD-related complication between patients with colonic CD and ileal CD (colonic CD hazard ratio [HR] = 0.51; 95% CI, 0.30-0.85 vs ileal CD HR = 0.79; 95% CI, 0.57-1.10; P = 0.033). No difference was identified between ECI vs CM for reducing the risk of surgery (colonic HR = 0.52 vs ileal HR = 0.74; P = 0.468) or hospitalization (colonic HR = 0.77 vs ileal HR = 0.83; P = 0.806).
Conclusions
In this posthoc analysis of the Randomized Evaluation of an Algorithm for Crohn’s Treatment trial, symptom-based ECI was associated with greater efficacy for reducing the risk of CD-related complications in patients with colonic disease location relative to ileal disease location.
Keywords: ileal Crohn disease, early combined immunosuppression, treat-to-target
INTRODUCTION
Treatment of patients with Crohn disease (CD) early in the disease course using tumor necrosis factor (TNF) antagonist therapy in combination with azathioprine has been shown to be more effective for achieving corticosteroid-free clinical remission and endoscopic remission than using TNF-antagonist monotherapy.1 In keeping with this concept, the Randomized Evaluation of an Algorithm for Crohn’s Treatment (REACT) cluster randomized trial showed that the use of early combined immunosuppression (ECI) with symptom-based monitoring and treatment adjustment resulted in a reduced risk of hospitalization, surgery, and CD-related complications compared to conventional “step care”–based management (CM).2
It is now recognized in clinical practice that patients with an increased baseline risk of disease-related complications are most likely to benefit from the use of ECI-based treatment algorithms.3, 4 However, notwithstanding that several clinical phenotypes are associated with an increased risk of complications, the possibility of differential response is less well established. Ileal disease involvement has been associated with an increased risk of development of CD-related complications across population-based cohort studies, inception cohorts, and posthoc analyses of clinical trials.5-8 These patients may therefore represent a population most likely to benefit from ECI. A treat-to-target approach based upon symptoms may be suboptimal for ileal disease given that clinical symptoms are generally poor surrogates for inflammation and that this lack of association between symptoms and inflammation is likely stronger in patients with ileal involvement than in those with isolated colonic disease, leading to delays in diagnosis and appropriate management.5, 9, 10 Therefore, it is possible that the beneficial effects of ECI utilizing clinical disease activity index–based treatment adjustments would be less apparent in patients with ileal disease than in those with isolated colonic disease.
To test this hypothesis, we performed a posthoc data analysis from the REACT trial that compared the relative efficacy of ECI on time-to-complication in patients with isolated colonic CD with that in patients with ileal-dominant disease.
METHODS
Data Source
The REACT trial (March 2010–October 2013) consisted of 41 practices across Canada and Belgium, which were randomized to either ECI (n = 22) or CM (n = 19). Each site recruited up to 60 consecutive patients with CD, with no restrictions based on disease activity or treatment exposures. Treatment effects of the country and practice size (≤100 patients with CD treated annually) were balanced between groups through a computer-generated minimization procedure. The algorithm used was a complex multidrug approach to management, and therefore masking was not possible. To minimize the risk of bias, outcome measures were objective, and to avoid the potential for treatment contamination, individual practices were unaware of the assignment of other practices.11 The ECI treatment sites were provided with education on the algorithm for treatment. Patients who failed to achieve clinical remission (defined as a Harvey Bradshaw Index [HBI] score ≤4) after initiation of corticosteroids (4 weeks in Canada, 12 weeks in Belgium) received combination therapy with a TNF antagonist and antimetabolite (thiopurines or methotrexate). Every 12 weeks thereafter, disease activity was assessed and patients were progressively escalated through a stepwise approach (escalation of the TNFα antagonist, switching the antimetabolite, switching the TNFα antagonist, and finally surgery) for persistently active disease (HBI > 6) until clinical remission was achieved. Practitioners assigned to CM were unaware of the ECI algorithm details, and treatment was conducted for patients per usual practice. In this posthoc analysis, the total number of patients included with available data on disease location was 99% (n = 1969/1995) of the original trial.
Exposure
We evaluated the independent association between disease location and clinical outcomes, with disease location categorized as isolated colonic or ileal-dominant (isolated ileal or ileocolonic) involvement. Isolated ileal and ileocolonic disease locations were grouped into the ileal-dominant category based on established differences between them and isolated colonic CD for clinical presentation, disease-related risk, underlying pathology, and response to therapy.5 Disease location was historic based on recorded phenotype, and repeat phenotyping was not repeated at the time of enrollment.
Outcomes
The primary outcome of interest for the current analysis was time to the first occurrence of the composite of hospitalization, surgery, or CD-related complication (presence of a new abscess, fistula, or stricture; serious worsening of disease activity; extraintestinal manifestations). Secondary outcomes of interest were time to the first hospitalization, surgery, and complication, and the proportion of subjects in corticosteroid-free remission at 12 and 24 months after randomization. We also evaluated rates of exposure to corticosteroids, antimetabolites, TNF antagonists, or the combination of antimetabolites and TNF antagonists over the 24-month period of the study in patients with isolated colonic and ileal-dominant CD by treatment assignment to either the ECI or CM treatment algorithm groups.
Statistical Analysis
Time to the first occurrence of the composite and times to first hospitalization, surgery, and complication were analyzed using marginal Cox proportional hazards models to account for cluster effects.12 Corticosteroid-free remission at 12 and 24 months after randomization were analyzed using the modified Poisson regression to adjust for cluster effects.13 All models contained the main effects of intervention (ECI vs CM), disease location (isolated colonic CD vs ileal-dominant CD), and their interactions, adjusting for practice size and country. Other covariates included in the models were selected from a pool of variables including age, sex, weight, smoking status, duration, prior surgery, fistulizing disease, HBI score, corticosteroid use, antimetabolite and/or TNF antagonist use, SF-36 physical component score, SF-36 mental component score, EQ-5D calculated index score, and the EQ-5D visual analog of self-rated health. To select appropriate covariates, backward logistic regression analyses with respective binary outcomes—that is, composite events, hospitalization, and remission—were performed at a statistical significant level of P < 0.1. All analyses were done with SAS version 9.4 (SAS Institute Inc, Cary, NC).
RESULTS
The distribution of isolated colonic (n = 435, 22.1%), isolated small bowel (n = 662, 33.6%), and ileocolonic (n = 872, 44.3%) disease location were consistent with prior large-scale real-world studies and generally in keeping with most phase 3 clinical trial study demographics.5-8 Baseline demographics of patients with isolated colonic and ileal-dominant CD, stratified by enrollment in the CM or ECI treatment algorithm groups, are presented in Table 1. Patients with isolated colonic CD less often had prior CD-related surgery; however, the proportions of patients entering the trial while receiving corticosteroids, antimetabolites, TNF antagonists, or combination therapy were similar.
TABLE 1.
Demographic Characteristics by Disease Location and Treatment Assignment
| Isolated Colonic CD | Ileal-Dominant CD | |||
|---|---|---|---|---|
| CM (n = 178) | ECI (n = 257) | CM (n = 717) | ECI (n = 816) | |
| Age, mean (y) | 45.5 | 46.7 | 43.8 | 43.4 |
| Male, n (%) | 72 (40.5) | 104 (40.5) | 308 (43.0) | 349 (42.8) |
| Current smoker, n (%) | 26 (14.6) | 53 (20.6) | 136 (19.0) | 218 (26.8) |
| Medications | ||||
| Corticosteroids, n (%) | 29 (16.3) | 47 (18.3) | 124 (17.3) | 157 (19.2) |
| Antimetabolite monotherapy, n (%) | 52 (29.2) | 85 (33.1) | 197 (27.5) | 273 (33.4) |
| TNF-antagonist monotherapy, n (%) | 40 (22.5) | 55 (21.4) | 156 (21.8) | 158 (19.3) |
| TNF-antagonist and antimetabolite combination therapy, n (%) | 26 (14.6) | 25 (9.7) | 90 (12.6) | 103 (12.6) |
| Previous surgery for CD, n (%) | 46 (25.8) | 45 (17.5) | 383 (53.4) | 455 (55.8) |
| Disease duration, mean months | 127.3 | 111.4 | 161.6 | 162.3 |
| Fistula, active, n (%) | 16 (9.0) | 18 (7.0) | 55 (7.7) | 55 (6.8) |
| HBI score, mean | 3.43 | 3.55 | 4.28 | 4.30 |
| SF-36 MCS score, mean | 47.1 | 46.9 | 45.5 | 44.1 |
| SF-36 PCS score, mean | 45.6 | 46.9 | 45.5 | 45.6 |
| EQ-5D score, mean | 0.82 | 0.82 | 0.81 | 0.80 |
EQ-5D, XXX; MCS, mental component score; PCS, physical component score; SF-36, XXX.
Development of Complications
Among patients randomized to the CM arm, the proportion developing CD-related complications (presence of a new abscess, fistula, or stricture; serious worsening of disease activity; extraintestinal manifestations) during the 24-month study period was similar between the isolated colonic (29.2%) and ileal-dominant subgroups (28.2%) (Table 2). However, the proportion requiring surgery (6.2% vs 9.3%) or hospitalization (9.0% vs 15.1%) during the 24-month study period was lower in the isolated colonic subgroup than in the ileal-dominant subgroup. Among patients randomized to the ECI arm, the proportion developing CD-related complications (fistula, abscess, stricture) during the 24-month study period was lower in the isolated colonic subgroup (15.2%) than in the ileal-dominant subgroup (22.4%) (Table 2). The proportion of patients requiring surgery (3.1% vs 6.6%) or hospitalization (6.6% vs 12.4%) during the 24-month study period was also lower in the isolated colonic subgroup than in the ileal-dominant subgroup for the ECI arm.
TABLE 2.
Event Rate for Endpoints Stratified by Treatment Algorithm and Disease Location
| Isolated Colonic CD | Ileal-Dominant CD | |||
|---|---|---|---|---|
| CM | ECI | CM | ECI | |
| Composite endpoint (hospitalization, surgery, complication; %) | 56/178 (31.5) | 47/257 (18.3) | 228/717 (31.8) | 204/817 (25.0) |
| CD-related complication (%) | 52/178 (29.2) | 39/257 (15.2) | 201/717 (28.0) | 183/817 (22.4) |
| Surgery (%) | 11/178 (6.2) | 8/257 (3.1) | 67/717 (9.3) | 54/817 (6.6) |
| Hospitalization (%) | 16/178 (9.0) | 17/257 (6.6) | 108/717 (15.1) | 101/817 (12.4) |
| Steroid-free remission (%) | ||||
| Month 0 | 110/168 (65.5) | 165/251 (65.7) | 361/696 (51.9) | 430/795 (54.1) |
| Month 6 | 111/147 (75.5) | 171/224 (76.3) | 396/639 (62.0) | 449/710 (63.2) |
| Month 12 | 106/142 (74.7) | 161/213 (75.6) | 368/604 (60.9) | 432/662 (65.3) |
| Month 18 | 77/110 (70.0) | 124/159 (78.0) | 318/507 (62.7) | 362/533 (67.9) |
| Month 24 | 87/118 (73.7) | 130/167 (77.8) | 330/531 (62.2) | 386/540 (71.5) |
Disease Location and Impact of ECI on Development of Complications
In the subgroup of patients with isolated colonic disease, ECI reduced the hazard of CD-related complications (presence of a new abscess, fistula, or stricture; serious worsening of disease activity; extraintestinal manifestations; surgery; and/or hospitalization) with an adjusted hazard ratio [HR] of 0.53 (95% confidence interval [CI], 0.33-0.84). In the subgroup with ileal-dominant disease, ECI also reduced the hazard of CD-related complications but to a lesser extent (adjusted HR = 0.77; 95% CI, 0.58-1.04). There was only weak evidence to suggest that the two HRs were different (P = 0.069). For CD-related complications, there was a significant impact of treatment according to disease location (isolated colonic HR = 0.51; 95% CI, 0.30-0.85 vs ileal-dominant HR = 0.80; 95% CI, 0.58-1.10; interaction P value = 0.035). There was no association between disease location and treatment strategy for either time to surgery or hospitalization (Table 3, Fig. 1).
TABLE 3.
Effect of ECI vs CM Stratified by Disease Location
| Isolated Colonic CD | Ileal-Dominant CD | ||
|---|---|---|---|
| HR (95% CI) | HR (95% CI) | Interaction P | |
| Composite endpoint (hospitalization, surgery, complication)* | 0.53 (0.33-0.84) | 0.77 (0.58-1.04) | 0.069 |
| Time to CD-related complication† | 0.51 (0.30-0.85) | 0.80 (0.58-1.10) | 0.035 |
| Time to surgery‡ | 0.52 (0.19-1.42) | 0.74 (0.53-1.04) | 0.468 |
| Time to hospitalization§ | 0.75 (0.42-1.34) | 0.82 (0.67-0.99) | 0.788 |
| 12 month steroid-free remission¶ | 1.04 (0.54-2.02) | 0.81 (0.55-1.20) | 0.429 |
| 24 month steroid-free remission‖ | 0.97 (0.85-1.10) | 1.10 (1.00-1.21) | 0.070 |
All analyses adjusted for country, practice size, and clustering effect of practice. Additional adjustments made for individual outcomes made based on backward model regression. *Adjusted for age, baseline fistula, baseline HBI (continuous), baseline PCS. †Adjusted for age, baseline fistula, baseline HBI (continuous), baseline PCS. ‡Adjusted for baseline fistula, baseline antimetabolite use, baseline PCS, baseline EQ-5D – visual analogue scale (VAS). §Adjusted for age, baseline PCS, baseline EQ-5D – VAS. ¶Adjusted for sex, previous surgery, baseline antimetabolite use, baseline PCS, baseline MCS. ‖Adjusted for sex, disease duration, previous surgery, baseline TNF-antagonist use, baseline PCS, baseline MCS. Baseline remission forced into steroid-free remission models.
FIGURE 1.
Rates of complications stratified by treatment algorithm and disease location subgroups. A, Hospitalization. B, Surgery. C, CD-related complications. D, Composite of hospitalization, surgery, and CD-related surgery.
Disease Location and Impact of ECI on Corticosteroid-Free Remission
At 12 months, the proportions of patients in corticosteroid-free remission in the ECI and CM arms of the isolated colonic subgroup were similar (75.6% and 74.7%, respectively), whereas in the ileal-dominant subgroup the ECI arm had a numerically higher percentage of patients with corticosteroid-free clinical remission compared with the CM arm (65.3% and 60.9%, respectively). This difference was not statistically significant (isolated colonic relative risk [RR] = 1.04; 95% CI, 0.54-2.02 vs ileal-dominant RR = 0.81; 95% CI, 0.55-1.20; P = 0.429; Table 2). At 24 months, the ECI arms in both the isolated colonic and ileal-dominant subgroups had higher proportions of patients in corticosteroid-free remission as compared with the CM arms in these subgroups (colonic RR = 0.97; 95% CI, 0.85-1.10 vs ileal-dominant RR = 1.10; 95% CI, 1.00-1.21; P = 0.070; Table 2).
Treatment Exposure
The proportion of patients being treated with TNF-antagonist monotherapy at 12 months (isolated colonic CD: n = 173/369, 46.9%; ileal-dominant CD: n = 533/1307, 40.8%) and 24 months (isolated colonic CD: n = 140/294, 47.6%; ileal-dominant CD: n = 463/1106, 41.9%) was higher in the isolated colonic subgroup. The proportion of patients being treated with a TNF-antagonist agent in combination with an antimetabolite agent at 12 months (isolated colonic CD: n = 85/369, 23%; ileal-dominant CD: n = 238/1307, 18.2%) and 24 months (isolated colonic CD: n = 74/294, 25.2%; ileal-dominant CD: n = 199/1106, 18%) was also higher in the isolated colonic subgroup (Table 4).
TABLE 4.
Baseline Medication Use and Treatment Escalation Stratified by Treatment Algorithm and Disease Location
| Isolated Colonic CD | Ileal-Dominant CD | ||||
|---|---|---|---|---|---|
| CM | ECI | CM | ECI | ||
| Proportion of patients on different medications at each time point | |||||
| Steroids | |||||
| Month 0 | 29/178 (16.3%) | 47/257 (18.3%) | 124/717 (17.3%) | 157/817 (19.2%) | |
| Month 12 | 17/152 (11.2%) | 21/217 (9.7%) | 72/628 (11.5%) | 70/679 (10.3%) | |
| Month 24 | 13/123 (10.6%) | 10/171 (5.9%) | 74/550 (13.5%) | 42/556 (7.6%) | |
| Antimetabolite monotherapy | |||||
| Month 0 | 78/178 (43.8%) | 110/257 (42.8%) | 287/717 (40.0%) | 376/817 (46.0%) | |
| Month 12 | 74/152 (48.7%) | 106/217 (48.9%) | 261/628 (41.6%) | 353/679 (52.0%) | |
| Month 24 | 64/123 (52.0%) | 85/171 (49.7%) | 235/550 (42.7%) | 286/556 (51.4%) | |
| TNF-antagonist monotherapy | |||||
| Month 0 | 66/178 (37.1%) | 80/257 (31.1%) | 246/717 (34.3%) | 261/817 (32.0%) | |
| Month 12 | 68/152 (44.7%) | 105/217 (48.4%) | 248/628 (39.5%) | 285/679 (42.0%) | |
| Month 24 | 58/123 (47.2%) | 82/171 (48.0%) | 218/550 (39.6%) | 245/556 (44.1%) | |
| Combination antimetabolite and TNF antagonist | |||||
| Month 0 | 26/178 (14.6%) | 25/257 (9.7%) | 90/717 (12.6%) | 103/817 (12.6%) | |
| Month 12 | 33/152 (21.7%) | 52/217 (24.0%) | 96/628 (15.3%) | 142/679 (20.9%) | |
| Month 24 | 30/123 (24.4%) | 44/171 (25.7%) | 85/550 (15.5%) | 114/556 (20.5%) | |
| Cumulative rates of treatment with different medications | |||||
| Steroids | |||||
| Month 12 | 14.0% | 11.9% | 14.3% | 15.2% | |
| Month 24 | 23.0% | 18.8% | 20.8% | 21.6% | |
| Antimetabolite monotherapy | |||||
| Month 12 | 8.8% | 20.0% | 10.1% | 20.2% | |
| Month 24 | 13.0% | 23.9% | 16.1% | 27.7% | |
| TNF-antagonist monotherapy | |||||
| Month 12 | 15.7% | 22.1% | 12.6% | 21.3% | |
| Month 24 | 20.6% | 26.9% | 16.8% | 27.4% | |
| Combination antimetabolite and TNF antagonist | |||||
| Month 12 | 9.3% | 19.0% | 5.8% | 13.7% | |
| Month 24 | 12.9% | 21.4% | 8.9% | 19.0% |
The cumulative rates of treatment with TNF-antagonist monotherapy in the ECI arm of the isolated colonic and ileal-dominant subgroups were similar at 12 (22.1% vs 21.3%) and 24 months (26.9% vs 27.4%). The cumulative rates of treatment with a TNF-antagonist agent in combination with an antimetabolite agent were higher in the ECI arm of the isolated colonic subgroup than in the ileal-dominant subgroup at 12 months (19.0% vs 13.7%) and 24 months (21.4% vs 19.0%) (Table 4, Fig. 2).
FIGURE 2.
Rates of treatment exposure stratified by treatment algorithm and disease location subgroups. A, Steroids. B, Antimetabolite. C, TNF antagonist. D, TNF antagonist in combination with an antimetabolite agent.
DISCUSSION
In this posthoc analysis of the REACT clinical trial, the impact of symptom-based ECI treatment adjustments on corticosteroid-free remission at 12 and 24 months was similar between patients with isolated colonic CD and ileal-dominant CD. Patients with isolated colonic CD treated with symptom-based ECI treatment adjustments more often progressed to treatment with TNF-antagonist therapy in combination with an antimetabolite over 24 months, and the use of ECI with symptom-based treatment adjustment had a greater impact on reducing the risk of CD-related complications (presence of a new abscess, fistula, or stricture; serious worsening of disease activity; extraintestinal manifestations) in isolated colonic CD than in ileal-dominant CD. Collectively, these results suggest that symptom-based monitoring to determine treatment adjustment may be less effective to guide therapy and prevent disease-related complications in patients with CD with ileal-dominant disease.
Ileal disease location has consistently been observed to be associated with an increased risk of disease progression and risk of developing an intestinal complication as compared with isolated colonic CD.3-5 Although rates of surgery or hospitalization in the CM arm of our study were lower in patients with isolated colonic CD, the use of symptom-based ECI treatment adjustments resulted in an absolute reduction of approximately 3% for these complications across both disease location subgroups. The rates of CD-related complications were similar between patients with isolated colonic CD and ileal-dominant CD randomized to the CM arm of the study; however, a significantly different risk reduction was observed when using symptom-based ECI treatment adjustments in patients with isolated colonic CD as compared with ileal-dominant CD. Given the consistent absolute numeric reduction in the risk of surgery and hospitalization with ECI across both disease location subgroups despite the higher baseline rate in patients with ileal-dominant disease and the similar baseline rate for the development of CD-related complications in the CM arm of the study among patients with isolated colonic and ileal-dominant CD, it is unlikely that the known association between ileal disease location and disease-related complications influenced our observations.
Two observations from these analyses are noteworthy. Although symptom-based ECI treatment adjustments had a similar impact on achieving corticosteroid-free remission in patients with isolated colonic and ileal-dominant CD, and a similar proportion of patients ultimately were in corticosteroid-free remission at 24 months in these disease location subgroups, the patients with isolated colonic CD had a higher rate of progression to combination therapy with a TNF antagonist and an antimetabolite. A stronger correlation has been observed between symptoms and mucosal disease activity in isolated colonic CD as compared with ileal-dominant CD.5 Algorithmic care is only effective if the monitoring approach used consistently triggers treatment intensification at each step until the target is achieved. Therefore, the observed differences in rates of progression to combination therapy with a TNF antagonist and an antimetabolite despite similar rates of corticosteroid-free remission may simply be a result of the patients with ileal-dominant CD not consistently triggering the algorithm over 24 months because of the poor correlation between symptoms and inflammation. This notion is further supported by the observation that the rates of progression to other steps in the algorithm (steroids, antimetabolite monotherapy, TNF-antagonist monotherapy) were similar between the patients with isolated colonic and ileal-dominant CD randomized to the ECI treatment arm, and the only distinction in treatment escalation was seen when comparing rates of progression with combination therapy.
Our study has several strengths, including the size of the cohorts, the recruitment of patients from community sites representing routine practice, and the ability to ascertain meaningful differences when using ECI in isolated colonic CD and ileal-dominant CD for reducing the risk of CD-related complications over a 2-year horizon. There are several important limitations worth noting. First, this was a posthoc analysis and the trial was not designed to address the questions posed. Therefore, these results should be considered exploratory and hypothesis-generating. Second, although our cohort size was quite large, the study was not specifically powered to compare the effectiveness of ECI vs CM in disease location subgroups, and point estimates should be interpreted with caution. Finally, endoscopy was not performed in this trial and we were therefore unable to confirm the lack of correlation between symptoms and mucosal disease activity in ileal-dominant CD or that the persistence of mucosal disease activity despite the achievement of clinical remission was a determinant for the differential effect seen with ECI in ileal-dominant CD. Further analyses from the ongoing REACT-2 trial where treatment escalation is based on endoscopic assessments of activity may help us better understand the observations we have made here.
In summary, we observed that a symptom-based ECI treatment escalation resulted in a higher rate of progression to combination therapy with a TNF antagonist and an antimetabolite and greater risk reduction for CD-related complications in isolated colonic CD vs ileal-dominant CD. Further studies are needed to understand how best to monitor and adjust treatment in patients with ileal-dominant CD, possibly with more objective markers of mucosal inflammation and/or direct visualization with colonoscopy, given the known increased risk for disease-related complications in these patients.
Author contributions: Study concept and design: PSD, VJ, SS. Acquisition of data: LWS, RK, GZ, WJS, BGF. Analysis and interpretation of data: PSD, VJ, GZ, LWS, SS. Drafting of the manuscript: PSD. Critical revision of the manuscript and approval of final manuscript: All authors. Guarantor of article: PSD.
Conflicts of interest: Parambir Dulai: research support from Pfizer, Takeda, Janssen, and AbbVie; consulting work for Takeda, Janssen, and AbbVie. Vipul Jairath: consulting fees from AbbVie, Eli Lilly, GlaxoSmithKline, Arena Pharmaceuticals, Genentech, Pendopharm, Sandoz, Merck, Takeda, Janssen, Robarts Clinical Trials, Topivert, and Celltrion; speaker’s fees from Takeda, Janssen, Shire, Ferring, AbbVie, and Pfizer. Reena Khanna: fees for consulting/speaking from AbbVie, Encycle, Janssen, Pendopharm, Pfizer, Robarts Clinical Trials, Shire, and Takeda Canada outside the submitted work. William Sandborn: research grants from Atlantic Healthcare Limited, Amgen, Genentech, Gilead Sciences, AbbVie, Janssen, Takeda, Lilly, and Celgene/Receptos; consulting fees from AbbVie, Allergan, Amgen, Boehringer Ingelheim, Celgene, Conatus, Cosmo, Escalier Biosciences, Ferring, Genentech, Gilead, Gossamer Bio, Janssen, Lilly, Miraca Life Sciences, Nivalis Therapeutics, Novartis Nutrition Science Partners, Oppilan Pharma, Otsuka, Paul Hastings, Pfizer, Precision IBD, Progenity, Prometheus Laboratories, Ritter Pharmaceuticals, Robarts Clinical Trials (owned by Health Academic Research Trust or HART), Salix, Shire, Seres Therapeutics, Sigmoid Biotechnologies, Takeda, Tigenix, Tillotts Pharma, UCB Pharma, and Vivelix; and stock options from Ritter Pharmaceuticals, Oppilan Pharma, Escalier Biosciences, Gossamer Bio, Precision IBD, and Progenity. Brian Feagan: grant/research support from Millennium Pharmaceuticals, Merck, Tillotts Pharma AG, AbbVie, Novartis Pharmaceuticals, Centocor, Elan/Biogen, UCB Pharma, Bristol-Myers Squibb, Genentech, ActoGenix, and Wyeth Pharmaceuticals; consulting fees from Millennium Pharmaceuticals, Merck, Centocor, Elan/Biogen, Janssen-Ortho, Teva Pharmaceuticals, Bristol-Myers Squibb, Celgene, UCB Pharma, AbbVie, Astra Zeneca, Serono, Genentech, Tillotts Pharma AG, Unity Pharmaceuticals, Albireo Pharma, Given Imaging, Salix Pharmaceuticals, Novonordisk, GSK, Actogenix, Prometheus Therapeutics and Diagnostics, Athersys, Axcan, Gilead, Pfizer, Shire, Wyeth, Zealand Pharma, Zyngenia, GiCare Pharma, and Sigmoid Pharma; and speaker’s bureau fees from UCB, AbbVie, and J&J/Janssen. Siddharth Singh: research support from Pfizer and AbbVie; consulting fees from AbbVie and Takeda.
REFERENCES
- 1. Dulai PS, Siegel CA, Colombel JF, et al. Systematic review: Monotherapy with antitumour necrosis factor α agents versus combination therapy with an immunosuppressive for IBD. Gut. 2014;63:1843–1853. [DOI] [PubMed] [Google Scholar]
- 2. Khanna R, Bressler B, Levesque BG, et al. ; REACT Study Investigators . Early combined immunosuppression for the management of Crohn’s disease (REACT): a cluster randomised controlled trial. Lancet. 2015;386:1825–1834. [DOI] [PubMed] [Google Scholar]
- 3. Siegel CA, Horton H, Siegel LS, et al. A validated web-based tool to display individualised Crohn’s disease predicted outcomes based on clinical, serologic and genetic variables. Aliment Pharmacol Ther. 2016;43:262–271. [DOI] [PubMed] [Google Scholar]
- 4. Guizzetti L, Zou G, Khanna R, et al. Development of clinical prediction models for surgery and complications in Crohn’s disease. J Crohns Colitis. 2018;12:167–177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dulai PS, Singh S, Casteele NV, et al. Should we divide Crohn’s disease into ileum-dominant and isolated colonic diseases? Clin Gastroenterol Hepatol. 2019;17:2634–2643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Subramanian S, Ekbom A, Rhodes JM. Recent advances in clinical practice: a systematic review of isolated colonic Crohn’s disease: the third IBD? Gut. 2017;66:362–381. [DOI] [PubMed] [Google Scholar]
- 7. Peyrin-Biroulet L, Loftus EV Jr, Colombel JF, et al. The natural history of adult Crohn’s disease in population-based cohorts. Am J Gastroenterol. 2010;105:289–297. [DOI] [PubMed] [Google Scholar]
- 8. Thia KT, Sandborn WJ, Harmsen WS, et al. Risk factors associated with progression to intestinal complications of Crohn’s disease in a population-based cohort. Gastroenterology. 2010;139:1147–1155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Vavricka SR, Spigaglia SM, Rogler G, et al. ; Swiss IBD Cohort Study Group . Systematic evaluation of risk factors for diagnostic delay in inflammatory bowel disease. Inflamm Bowel Dis. 2012;18:496–505. [DOI] [PubMed] [Google Scholar]
- 10. Bamba S, Tsujikawa T, Ban H, et al. Predicting mucosal healing in Crohn’s disease using practical clinical indices with regard to the location of active disease. Hepatogastroenterology. 2014;61:689–696. [PubMed] [Google Scholar]
- 11. Kahan BC, Cro S, Doré CJ, et al. Reducing bias in open-label trials where blinded outcome assessment is not feasible: strategies from two randomised trials. Trials. 2014;15:456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Berhane K, Weissfeld LA. Inference in spline-based models for multiple time-to-event data, with applications to a breast cancer prevention trial. Biometrics. 2003;59:859–868. [DOI] [PubMed] [Google Scholar]
- 13. Zou GY, Donner A. Extension of the modified Poisson regression model to prospective studies with correlated binary data. Stat Methods Med Res. 2013;22:661–670. [DOI] [PubMed] [Google Scholar]