Abstract
Background & Aims:
Little is known about the effects of discontinuing 5-aminosalycilates (5-ASA) therapy for patients with Crohn’s disease (CD) who initiate therapy with an anti-tumor necrosis factor (anti-TNF) agent. We analyzed data from 2 national population cohorts to compare outcomes of patients with CD already on 5-ASA therapy who started treatment with an anti-TNF agent and then either stopped or continued 5-ASA.
Methods:
The primary outcome was any adverse clinical event, defined as a composite of new corticosteroid use or CD-related hospitalization or surgery. We collected data from the Truven MarketScan health claims database (United States, US) and the Danish health registers. Our analysis included patients with CD who started anti-TNF therapy after at least 90 days of oral 5-ASA therapy. Patients were classified as stopping 5-ASA if therapy was discontinued within 90 days of starting anti-TNF. We performed multivariable Cox regression models controlling for demographics, clinical factors, and healthcare utilization. Adjusted hazard ratios (aHR) with 95% CI values were calculated, comparing stopping 5-ASA with continuing 5-ASA.
Results:
A total of 3178 patients with CD were included in our final analysis (2960 in the US and 218 in Denmark). Stopping 5-ASA after initiating anti-TNF therapy was not associated with an increased risk of adverse clinical events in the US cohort (aHR, 0.89; 95% CI, 0.77–1.03; P=.13) or in the Danish cohort (aHR, 1.13; 95% CI, 0.68–1.87; P=.63). Similar results were obtained from sensitivity analyses of concomitant immunomodulator use and duration of 5-ASA treatment before initiation of anti-TNF therapy.
Conclusions:
In a retrospective analysis of 2 national databases, we found that stopping 5-ASA therapy for patients with CD who were starting anti-TNF therapy did not increase their risk of adverse clinical events. These results should be validated in a prospective study.
Keywords: Inflammatory bowel disease, mesalamine, IBD treatment
INTRODUCTION
Crohn’s disease (CD) is an inflammatory bowel disease (IBD) characterized by chronic, relapsing inflammation in the gastrointestinal tract that can result in bowel damage and disability.1 European and United States (U.S.) guidelines on the treatment of CD recommend against using 5-aminosalicylates (5-ASA) for the induction or maintenance of remission.2,3 Despite this, 5-ASA are commonly used in the treatment of CD in real world practice.4–6 There is great interest in understanding de-escalation strategies in CD in order to decrease costs and mitigate potential side effects. However, to date, most research on de-escalation of therapy in CD has focused on stopping immunosuppressive agents.7 In ulcerative colitis (UC), stopping 5-ASA after starting anti-tumor necrosis factor agents (anti-TNF) has no impact on clinical outcomes.8 Further, concomitant 5-ASA treatment with anti-TNF and vedolizumab does not appear to impact clinical remission and endoscopic healing rates.9.10 After a CD patient requires escalation to biologics, the benefit of continuing or withdrawing 5-ASA is currently unknown. This is of significant interest to patients and payers as 5-ASA agents are costly and often require taking multiple pills a day. We therefore aimed to compare clinical outcomes in CD patients already on oral 5-ASA who started an anti-TNF and then either continued or discontinued oral 5-ASA agents utilizing two large national population-based cohorts.
METHODS
Study Populations
We utilized two large national databases, one from the U.S. and the other from Denmark. The U.S. cohort was derived from the Truven Health MarketScan® Commercial Database which is the largest multi-payer health claims databases in the U.S. containing over 450 million patient-level entries per year. This database includes de-identified individual-level healthcare claims information that reflects real-world treatment patterns. The MarketScan database includes enrollment data from large employers and health plans. For this study, all enrollees from 2007 through 2016 were included. The International Classification of Diseases (ICD), 9th and 10th Revision, codes were used to identify patients with at least one code for CD (555.x and K50.x) included. Other inclusion criteria were initiation of anti-TNF biologic therapy (i.e., infliximab, adalimumab, or certolizumab), continuation of biologic therapy for at least 90 days, use of 5-ASA for at least 90 days prior to initiation of biologic therapy, and at least 30 days of post-biologic follow-up time. Patients with known prior history of bowel resection were excluded. Medication prescriptions were identified using string searches of medication names.
The Danish cohort was drawn from three nationwide Danish health registers, the Danish National Patient Register (NPR), the Danish Prescription Register, and the Danish Central Personal Registration System) from 2003 through 2014 using a unique personal identification number assigned to each Danish citizen at birth for linkage. The NPR includes records on all discharges from hospitals in Denmark since 1977 and outpatient contacts since 1995. For this study, we collected information about CD diagnoses, gastrointestinal surgery, biologic therapy, and hospital admissions from the NPR. Information about use of biologic therapy, 5-ASA, immunomodulators (azathioprine, 6-mercaptopurine, methotrexate) and corticosteroids was collected from the Danish Prescription Register which was established in 1995 and includes information on all prescriptions filled by Danish residents. Information about death and emigration was collected from the Danish Central Personal Registration System. Data from the Registers was available through December 31st, 2014. Patients with CD were identified using ICD-8 (563.00–563.09) and ICD-10 codes (K50). Anti-TNF biologic and 5-ASA therapies were identified using Anatomical Therapeutic Chemical (ATC) codes and procedure codes. Inclusion criteria were identical to the U.S. cohort.
Outcomes
Our primary outcome was a composite of adverse clinical events defined as need for new corticosteroid therapy, CD-related hospitalization or surgery. The need for corticosteroid therapy was defined as a new prescription of oral steroid at least 90 days after the initiation of biologic therapy. Hospitalizations were defined as any admission with CD as the primary or secondary diagnosis. If CD was coded as the secondary diagnosis, the primary diagnosis had to be related to gastrointestinal symptoms, such as nausea, vomiting, abdominal pain, diarrhea, constipation, or gastrointestinal bleeding. Hospitalizations were counted if the admission occurred at least 1 day after the initiation of anti-TNF therapy. Intestinal surgeries were identified in the MarketScan database by ICD-9/10 and Common Procedural Terminology (CPT) codes in any procedure positions. These included codes for small bowel resections and colon resections as well as NPR operation codes.
Variables
CD patients who discontinued 5-ASA treatment within 90 days after initiation of anti-TNF therapy were classified as having discontinued 5-ASA while those who remained on 5-ASA treatment for at least 90 days after initiation of anti-TNF therapy were classified as having continued 5-ASA. Additional variables that were analyzed in both cohorts included age, sex, duration of 5-ASA treatment before anti-TNF initiation, and health care utilization prior to initiation of biologic therapy. Baseline health care utilization was assessed as a measure of CD severity prior to biologic therapy and included prior corticosteroid use (remote defined as 365 to 90 days prior to anti-TNF therapy or recent defined as within 90 days prior to anti-TNF therapy), number of hospitalizations in the year before starting anti-TNF (categorized as 0, 1–2, or > 2), number of emergency department visits in the year before starting anti-TNF (categorized as 0, 1, or > 2) and prior intestinal surgery (dichotomous variable). Data on the duration of IBD diagnosis was available in the NPR but not MarketScan, and was thus included as a potential confounder in analyses of the former but not the latter.
Statistical Analyses
Descriptive analyses were used to evaluate the baseline characteristics of CD patients who discontinued or continued 5-ASA. The Student t test was used to compare continuous variables and the chi-squared test was used to compare categorical variables. The Wilcoxon rank sum test was used to compare continuous variables with a skewed distribution. The incidence rate of events was calculated as the number of adverse clinical outcomes divided by the person time. Cox regression and Kaplan-Meier methods were used to compare the risk of adverse clinical events between CD patients who discontinued versus continued 5-ASA therapy. Baseline was defined as the date of anti-TNF initiation. No violations of the proportional hazards assumption were identified. Outcomes were evaluated individually and as the composite primary outcome. Multivariable models were adjusted for age, sex, duration of 5-ASA treatment, and health care utilization (corticosteroid use, hospitalization, emergency department visits, and prior intestinal surgery) prior to the initiation of anti-TNF therapy. Multivariable analyses of the Danish cohort also included duration of CD diagnosis, which was not available in the MarketScan database. Adjusted hazard ratios (aHR) and 95% confidence intervals (CI) are reported with an aHR greater than 1 indicating an increased risk of the event in the 5-ASA discontinuation group compared to continuation. Follow-up time for the MarketScan data was limited to 3 years due to the relative lack of longitudinal follow-up data beyond that timeframe. Maximum follow-up time in the Danish cohort was 9 years. Subgroup analyses evaluated the effect of concomitant immunomodulator use (prescribed within 30 days before or after initiation of biologic therapy) on the risk of adverse clinical events. Sensitivity analyses were performed using different criteria for the number of days (60 and 180) on 5-ASA prior to anti-TNF therapy. Statistical significance was defined as a two-sided α of less than 0.05. Statistical analyses were performed using SAS 9.4 (Cary, North Carolina), Stata SE 14.2 (College Station, Texas), and Stata MP 15.1 (College Station, Texas).
Ethics
The study protocol related to the use of the MarketScan data was deemed exempt by the Institutional Review Board of Stanford University. An access and linkage permission was obtained from the Danish Data Protection Agency (J.no. 2008–54-0472) for the Danish cohort.
RESULTS
Baseline Patient Characteristics
The U.S. and Danish cohorts had 2,960 and 218 CD patients respectively, who met inclusion criteria (Table 1). In the U.S. cohort, there were 1,044 (35.3%) patients who discontinued 5-ASA and 1,916 (64.7%) patients who continued 5-ASA after initiation of anti-TNF therapy. Those who discontinued 5-ASA were slightly younger than those who continued 5-ASA (39.3 vs. 41.0 years, p<0.01). Prior to anti-TNF therapy, those who discontinued 5-ASA had a shorter duration on 5-ASA treatment (median 305 vs. 495 days, p < 0.01), more frequently had prior intestinal surgery (4.1% vs. 2.2%, p < 0.01), and higher rate of recent corticosteroid use (37.1% vs. 32.2%, p = 0.02). Median follow-up time after initiation of biologic therapy was somewhat less in patients who stopped 5-ASA treatment compared to those who continued (296 vs. 325 days, p<0.01).
Table 1.
Baseline patient characteristics
| United States Cohort | Denmark Cohort | |||||
|---|---|---|---|---|---|---|
| Characteristic | Continue 5-ASA (N = 1916) |
Stop 5-ASA (N = 1044) |
P value | Continue 5-ASA (N = 112) |
Stop 5-ASA (N = 106) |
P value |
| Age at start of anti-TNF therapy, mean years (SD) | 41.0 (14.5) | 39.3 (14.0) | <0.01 | 41.1 (14.6) | 36.2 (13.7) | 0.01 |
| Age at CD diagnosis, mean years (SD) | N/A | N/A | 34.5 (14.6) | 31.1 (13.1) | 0.08 | |
| Sex | 0.76 | 0.7 | ||||
| Male | 959 (50.1) | 529 (50.7) | 56 (50) | 56 (53) | ||
| Female | 956 (49.9) | 515 (49.3) | 56 (50) | 50 (47) | ||
| Year of initiation of anti-TNF therapy, n (%) | <0.01 | |||||
| 2007 – 2010 | 660 (34.4) | 287 (27.5) | N/A | N/A | ||
| 2011 – 2013 | 679 (35.4) | 364 (34.9) | N/A | N/A | ||
| 2014 – 2016 | 577 (30.1) | 393 (37.6) | N/A | N/A | ||
| Year of initiation of anti-TNF therapy, n (%) | 0.3 | |||||
| 2003 – 2008 | N/A | N/A | 38 (34) | 38 (36) | ||
| 2009 – 2010 | N/A | N/A | 41 (37) | 28 (26) | ||
| 2011 – 2012 | N/A | N/A | 19 (17) | 26 (25) | ||
| 2013 – 2014 | N/A | N/A | 14 (13) | 14 (13) | ||
| Pre-anti-TNF disease duration, median years (IQR) | N/A | N/A | 4.7 (1.4,9.4) | 3.2 (0.7, 8.1) | 0.02 | |
| Pre-anti-TNF 5-ASA treatment, median days (IQR) | 495 (404, 633) | 305 (211,363) | <0.01 | 844 (476, 1540) | 272 (192, 548) | <0.01 |
| Pre-anti-TNF corticosteroid use, n (%) 1 | 0.02 | 0.6 | ||||
| None | 694 (36.2) | 359 (34.4) | 35 (31) | 29 (27) | ||
| Remote | 606 (31.6) | 298 (28.5) | 22 (20) | 18 (17) | ||
| Recent (≤ 90 days) | 616 (32.2) | 387 (37.1) | 55 (49) | 59 (56) | ||
| Pre-anti-TNF hospitalizations, n (%) 1 | N/A | 0.8 | ||||
| 0 | 1465 (76.5) | UTR | 56 (50) | 49 (46) | ||
| 1 – 2 | 401 (20.9) | UTR | 42 (38) | 41 (39) | ||
| > 2 | 50 (2.6) | UTR | 14 (13) | 16 (15) | ||
| Pre-anti-TNF emergency department visits, n (%)1 | 0.11 | 0.4 | ||||
| 0 | 1269 (66.2) | 653 (62.6) | 93 (83) | 95 (90) | ||
| 1 – 2 | 589 (30.7) | 360 (34.5) | UTR | UTR | ||
| > 2 | 58 (3.0) | 31 (3.0) | UTR | UTR | ||
| Immunomodulator use ± 30 days from initiation of anti-TNF therapy, n (%) | 0.07 | 0.2 | ||||
| No | 1208 (63.0) | 693 (66.4) | 76 (68) | 74 (70) | ||
| Yes | 708 (37.0) | 351 (33.6) | 36 (32) | 32 (30) | ||
| Pre-anti-TNF intestinal surgery, n (%) | <0.01 | 0.8 | ||||
| No | 1874 (97.8) | 1001 (95.9) | 78 (70) | 64 (60) | ||
| Yes | 42 (2.2) | 43 (4.1) | 34 (30) | 42 (40) | ||
| Follow-up time from initiation of anti-TNF therapy, median days (IQR)2 | 325 (163, 642) | 296 (166, 558) | <0.01 | 199 (101, 410) | 142 (82, 336) | 0.1 |
UTR = unable to report, due to contractual agreements that prohibit reporting of small cell sizes.
Within 365 days prior to initiation of biologic therapy.
Defined as the time from initiation of biologic therapy until the end of biologic therapy, loss of follow-up, or administrative censoring.
In the Danish cohort, there were 106 (48.6%) patients who discontinued 5-ASA and 112 (51.4%) patients who continued 5-ASA after initiation of anti-TNF therapy (Table 1). Patients who discontinued 5-ASA were younger (36.2 vs. 41.1 years, p = 0.01) and had shorter disease duration (3.2 vs. 4.7 years, p = 0.02) compared to those who continued 5-ASA. In addition, patients who stopped 5-ASA had lower pre-anti-TNF 5-ASA treatment duration (272 vs. 844 days, p<0.01).
Adverse Clinical Events
In univariable analysis, discontinuation of 5-ASA in CD patients starting anti-TNF was not associated with an increase in adverse clinical events compared to continuing 5-ASA in both the U.S. and Danish cohorts. The proportion of patients experiencing the primary endpoint (composite of need for new corticosteroid therapy, CD-related hospitalization or surgery) over time was not significantly different in both cohorts (Figures 1 and 2). In addition, no significant differences were seen when analyzing the individual components of the composite primary outcome (Figures 1 and 2). One exception was an increased risk of hospitalization among CD patients who discontinued 5-ASA in the Danish cohort (log-rank p=0.01, Figure 2B). After adjusting for age, sex, duration of pre-biologic 5-ASA treatment, CD disease duration (Danish cohort only) and baseline health care utilization in multivariable analysis, no significant differences were seen in the primary outcome in both the U.S. and Danish cohorts when comparing CD patients who discontinued with those who continued 5-ASA after initiating anti-TNF therapy (Table 2). The risk of the individual clinical events that constituted the primary outcome was also similar between the 5-ASA continuation and discontinuation groups in both cohorts with no significant differences observed in adjusted analyses (Table 2).
Figure 1. Proportion of CD patients with adverse clinical events comparing those who continued or stopped 5-ASA in United States cohort.
A. Kaplan-Meier curve for new steroid prescription; B. IBD-related hospitalization; C. surgery; D. composite outcome (new steroid prescription, IBD-related hospitalization, and/or surgery).
Figure 2. Proportion of CD patients with adverse clinical events comparing those who continued or stopped 5-ASA in Danish cohort.
Kaplan-Meier curve for new steroid prescription; B. IBD-related hospitalization; C. surgery; D. composite outcome (new steroid prescription, IBD-related hospitalization, and/or surgery).
Table 2.
Risk of adverse clinical events comparing patients who continue or discontinue 5-ASA therapy after initiating anti-TNF therapy
| Continue 5-ASA | Discontinue 5-ASA | Discontinue versus continue 5-ASA | ||
|---|---|---|---|---|
| IR (95% CI), per 100 person-years | IR (95% CI), per 100 person-years | aHR (95% CI)* | P value | |
| United States Cohort | ||||
| New Steroid Use | 29.8 (27.4 – 32.4) | 29.4 (26.1 – 33.0) | 0.87 (0.74 – 1.03) | 0.12 |
| Hospitalization | 15.8 (14.1 – 17.6) | 16.8 (14.5 – 19.5) | 0.87 (0.70 – 1.08) | 0.21 |
| Surgery | 2.9 (2.3 – 3.7) | 3.2 (2.4 – 4.5) | 0.79 (0.49 – 1.29) | 0.35 |
| Composite | 42.8 (39.8 – 46.0) | 43.9 (39.7 – 48.6) | 0.89 (0.77 – 1.03) | 0.13 |
| Denmark Cohort | ||||
| New Steroid Use | 17.9 (11.9 – 26.9) | 19.5 (12.7 – 29.9) | 1.18 (0.60 – 2.33) | 0.63 |
| Hospitalization | 9.5 (5.7 – 15.8) | 26.9 (18.4 – 39.2) | 2.06 (1.01 – 4.20) | 0.05 |
| Surgery | 8.1 (4.7 – 13.9) | 11.8 (7.0 – 20.0) | 1.13 (0.45 – 2.83) | 0.80 |
| Composite | 39.1 (28.7 – 53.3) | 47.1 (34.4 – 64.5) | 1.13 (0.68 – 1.87) | 0.63 |
In the U.S. cohort, models adjusted for age, sex, duration of pre-biologic 5-aminosalicylate treatment, and baseline health care utilization (hospitalization, emergency department visits, corticosteroid use, gastrointestinal surgery). In the Danish cohort, models adjusted for age, sex, duration of Crohn’s disease at initiation of biologic therapy, duration of pre-biologic 5-aminosalicylate treatment, and baseline health care utilization (hospitalization, emergency department visits, corticosteroid use, gastrointestinal surgery)
We conducted additional analyses to further examine the association of 5-ASA discontinuation with adverse clinical outcomes. When stratifying CD patients by concomitant immunomodulator use at time of anti-TNF initiation, there was no increased risk of the primary outcome following discontinuation of 5-ASA regardless of whether patient was on monotherapy or combination therapy (Table 3). In fact, there was a decrease in risk of the primary outcome as well as hospitalizations and surgeries among patients not on immunomodulators who stopped 5-ASA in the U.S. cohort. No significant differences were seen across outcomes among patients on immunomodulators in the U.S. cohort. There was also no significant difference in the risk of outcomes in the Danish cohort when stratifying by concomitant immunomodulator use with the exception of corticosteroid use being decreased in patients on immunomodulators who stopped 5-ASA. We also examined the primary outcome and individual component outcomes requiring differing times on 5-ASA prior to anti-TNF initiation (30 and 180 days) and found no significant differences (Supplementary Tables 1 and 2).
Table 3.
Risk of adverse clinical events comparing patients who discontinue 5-ASA with those who continue 5-ASA therapy, stratified by concomitant use of immunodulator therapy
| Biologic without Immunomodulator | Biologic with Immunomodulator | |||
|---|---|---|---|---|
| aHR (95% CI)* | P value | aHR (95% CI)* | P value | |
| United States | ||||
| Steroid | 0.83 (0.67 – 1.02) | 0.08 | 0.96 (0.72 – 1.28) | 0.78 |
| Hospitalization | 0.69 (0.52 – 0.92) | 0.01 | 1.23 (0.87 – 1.76) | 0.25 |
| Surgery | 0.41 (0.19 – 0.87) | 0.02 | 1.44 (0.75 – 2.76) | 0.27 |
| Composite | 0.80 (0.66 – 0.96) | 0.02 | 1.08 (0.84 – 1.37) | 0.55 |
| Denmark | ||||
| Steroid | 2.15 (0.91 – 5.09) | 0.08 | 0.07 (0.01 – 0.42) | 0.003 |
| Hospitalization | 2.07 (0.77 – 5.56) | 0.15 | 3.29 (0.83 – 13.00) | 0.09 |
| Surgery | 3.15 (0.79 – 12.58) | 0.10 | 0.96 (0.09 – 10.69) | 0.98 |
| Composite | 1.43 (0.73 – 2.79) | 0.33 | 0.81 (0.31 – 2.09) | 0.66 |
DISCUSSION
In this study of two large national databases, we observed that discontinuation of 5-ASA after initiation of anti-TNF therapy in CD patients was not associated with an increase in the risk of adverse clinical outcomes, including corticosteroid use, hospitalization, and surgery. These findings held after controlling for potential confounders and baseline healthcare utilization. Importantly, results were consistent across two independent cohorts from different countries. This observation is clinically relevant as, to our knowledge, there are currently no published data on whether CD patients who initiate anti-TNF therapy can safely discontinue 5-ASA therapy. Given the potential benefits of discontinuing 5-ASA drugs (including decreased costs, reduced risk of medication side effects, and less polypharmacy) coupled with the questionable efficacy of 5-ASA in CD, our data suggest that stopping 5-ASA after patients start anti-TNF is likely safe. However, prospective trials are ideally needed to further establish this as a preferred medication de-escalation strategy in CD.
Current CD guidelines recommend against the use of 5-ASA for the induction and maintenance of remission.2,3 Meta-analyses of clinical trial data have found inconsistent results with 5-ASA in the treatment of CD. For example, Ford et al pooled data from 6 randomized controlled trials and found that 5-ASA did not have benefit of inducing remission over placebo (RR 0.91, 95% CI 0.77–1.06).11 The same study found a possible impact on decreased clinical recurrence in quiescent CD (OR 0.79, 95% CI 0.66–0.95) but this was likely subject to bias as this effect was only seen in per protocol group.11 In addition, a Cochrane review found that neither low nor high dose 5-ASA was effective at inducing response or remission.12 Our study supports these prior publications in demonstrating lack of clinical utility of 5-ASA in the treatment of CD.
Prior literature investigating the impact of concomitant 5-ASA with immunosuppresives is limited. One prior retrospective study of 186 IBD patients investigated the effect of concomitant 5-ASA treatment with azathioprine at maintaining remission in clinically quiescent disease.13 Over a median follow up of 4.3 years, there was no difference in relapse rates when comparing 5-ASA plus azathioprine with azathioprine alone groups. Another study from Korea, compared concomitant 5-ASA and azathioprine in CD specifically and its impact on need for steroids, biologics and surgery.14 Combination 5-ASA and azathioprine did not result in any improvement in clinical outcomes compared to azathioprine monotherapy. In contrast, a single-center study from Denmark of 537 CD patients treated with first line 5-ASA monotherapy observed that 23% of patients had “5-ASA dependency” which was defined as clinical relapse within one year of stopping 5-ASA treatment with regain of response after re-starting 5-ASA.15 However, rates of relapse on patients who had therapy escalated to immunosuppressive agents were not investigated in this study. Our results add to the current literature and suggest that continuation of previously prescribed 5-ASA does not provide benefit after escalation to more potent anti-TNF biologic medications. We suspect that this is most likely because the anti-inflammatory effects of anti-TNF are more than sufficient for controlling CD with no additive effect from 5-ASA, likely reflecting the higher potency of biologic agents and/or ineffectiveness of 5-ASA in CD. Of note, a prospective clinical trial investigating the impact of stopping 5-ASA in quiescent CD (the STATIC trial) is currently underway in Canada ( NCT03261206).
Since clinical trial evidence supports improved outcomes with combination therapy in CD, we explored if concomitant immunomodulator therapy may modify risk of adverse events after 5-ASA discontinuation.16 In addition, some prior studies have suggested that 5-ASA interacts with immunomodulator metabolism and can increase 6-thioguanine levels by interacting with thiopurine methyltransferase.17–19 The presence or absence of concomitant immunomodulators did not appear to increase the risk of major adverse outcomes when discontinuing 5-ASA in either cohort.
Although rates of composite adverse events were unchanged regardless of 5-ASA discontinuation, in the Danish cohort, hospitalizations were higher among CD patients who discontinued 5-ASA after starting an anti-TNF. It is possible that in Danish gastroenterologists are more likely to stop 5-ASA in patients with increasing symptoms due to presumed lack of efficacy. In this case, 5-ASA discontinuation may be a marker of disease severity with more severe CD patients, who are more likely to require hospitalization, also being more likely to stop 5-ASA. This possible reason is supported by higher rates of pre-anti-TNF hospitalizations, emergency department visits, and surgeries in the 5-ASA discontinuation group from Denmark. Given the smaller sample size of the Denmark cohort, it is possible we were unable to fully control for potential confounders. Future prospective studies should further examine the possible association of CD hospitalization with 5-ASA discontinuation.
This study has several strengths. First, the study included two completely independent population cohorts that permitted cross-validation of results. Although we present the results within the same study, the cohorts were analyzed separately by two separate statisticians to minimize risk of bias. The rigor of this approach and the consistency in the results from two different countries provide greater confidence in the validity of the findings. Second, the use of large population-based cohorts provided the ability to select many patients in the specific clinical scenario of interest as well as improve the precision and power of effect estimates. Third, the source data were derived from national research infrastructures that have well-characterized population databases with robust longitudinal follow-up and complete medication histories. Given the longer-term follow up of patients in these cohorts we were able to observe the progression of medical history over time and capture potential short and long term adverse outcomes following discontinuation of 5-ASA.
We acknowledge several limitations of this study. First, we used databases that relied on administrative health claims to define patient characteristics and outcome. There is therefore the potential for misclassification from improper coding of diagnoses and we had to rely on prescriptions as a measure of patients actually taking medications. However, findings were consistent in both databases and inclusion criteria required CD specific medications so misclassification was likely low. The Danish cohort has been validated and demonstrated at least 97% validity of CD diagnoses comparing charts with diagnosis codes.20 The U.S. database has unfortunately not had similar validation of case definitions but our definition was strict based on ICD-9 code and two CD related medication codes (5-ASA and anti-TNF) so misclassification of cases should be low. Second, both cohorts lack granular CD-specific data, such as laboratory values, disease behavior and location by Montreal classification, smoking status and endoscopic findings that would allow for more detailed control of disease severity. Third, given the retrospective nature of this study, there may have been unmeasured confounders that contributed to the decision about continuing 5-ASA in some patients, which could have attenuated risk of major adverse outcomes.
In addition, we did not include stopping anti-TNF as an endpoint due to insufficient data on reason for discontinuation. Using anti-TNF discontinuation as an endpoint may have misclassified patients as treatment failures when the biologic was stopped for reasons other than failure to improve their CD. In fact, a study from Denmark found that the most common reason for infliximab discontinuation was remission so this would have been an inappropriate endpoint given our aims.21 Last, one potential benefit of 5-ASA is a potential chemopreventive effect against colonic neoplasia; however, there were very few events (< 5) in each cohort, which precluded any meaningful analysis of this outcome.
An important consideration with any longitudinal study is immortal time bias. Immortal time is the span of follow-up in which the outcome of interest could not occur due to the exposure definition. We think the potential impact of this is limited in our study. Exposure was defined as stopping 5-ASA within 90 days of starting anti-TNF and the definition of starting anti-TNF treatment was restricted to at least 90 days of anti-TNF before the start of follow up. Therefore, the exposure (5-ASA stop or continue) was determined during the initial 90 days in which patients were required to be on anti-TNF before inclusion in the cohort. Outcomes of interest for our analysis could not have occurred during this initial 90-day anti-TNF treatment period in both groups (stop and continue 5-ASA).
In conclusion, using two population-based databases, we observed that discontinuation of 5-ASA after initiating anti-TNF therapy in CD was not associated with an increased risk of major adverse clinical outcomes. Our findings suggest that 5-ASA can be stopped after escalating therapy to an anti-TNF and provide further evidence that 5-ASA are of limited value in CD. Prospective studies investigating the benefits and risks of 5-ASA discontinuation after biologic initiation are needed to confirm our findings.
Supplementary Material
What You Need to Know.
Background:
We analyzed data from 2 national population-based cohorts to compare outcomes of patients with Crohn’s disease (CD) already on 5-aminosalycilates (5-ASA) therapy who started treatment with an anti-tumor necrosis factor (anti-TNF) agent and then either stopped or continued 5-ASA.
Findings:
Among patients with CD who started anti-TNF therapy, stopping 5-ASA therapy did not increase their risk of adverse clinical events.
Implications for Patient Care:
Patients with CD who are starting anti-TNF therapy appear to be able to safely discontinue 5-ASA therapy, but prospective studies are needed.
Acknowledgments
Grant Support: Access to the Truven Health MarketScan® Commercial Database was provided by the Stanford Center for Population Health Sciences (PHS) Data Core. The PHS Data Core is supported by a National Institutes of Health (NIH) National Center for Advancing Translational Science Clinical and Translational Science Award (UL1 TR001085) and internal Stanford funding. RCU is supported by a Career Development Award from the Crohn’s and Colitis Foundation and an NIH K23 Career Development Award (K23KD111995-01A1). This work was supported in part by the SUCCESS (Sinai Ulcerative Colitis Clinical, Experimental and System Studies) grant.
Disclosures: RCU has served as an advisory board member or consultant for Janssen, Pfizer, and Takeda. Research grants from Abbvie, Boehringer Ingelheim, and Pfizer. JFC has served as an advisory board member or consultant for AbbVie, Amgen, Boehringer-Ingelheim, Arena Pharmaceuticals, Celgene Corporation, Celltrion, Enterome, Eli Lilly, Ferring Pharmaceuticals, Genentech, Janssen and Janssen, Medimmune, Merck & Co., Nextbiotix, Novartis Pharmaceuticals Corporation, Otsuka Pharmaceutical Development & Commercialization, Inc., Pfizer, Protagonist, Second Genome, Gilead, Seres Therapeutics, Shire, Takeda, Theradiag. Speaker for AbbVie, Ferring, Takeda, Celgene Corporation. Stock options: Intestinal Biotech Development, Genefit. Research Grants: AbbVie, Takeda, Janssen and Janssen. JB has served as an advisory board member or consultant for AbbVie, Janssen-Cilag, Celgene, MSD, Pfizer, Samsung Bioepis and Takeda. Research grants from AbbVie, Takeda and Tillots Pharma.
Abbreviations:
- CD
Crohn’s disease
- anti-TNF
anti-tumor necrosis factor alpha
- aHR
Adjusted hazard ratios
- CI
confidence intervals
- U.S.
United States
- 5-ASA
5-aminosalicylate
- NPR
National Patient Register
- ICD
International Classification of Diseases
- CPT
Common Procedural Terminology
- IQR
interquartile range
- SD
standard deviation
- N/A
not applicable
- IR
incidence rate
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Torres J, Mehandru S, Colombel J-F, et al. Crohn’s disease. Lancet 2017;389:1741–1755. [DOI] [PubMed] [Google Scholar]
- 2.Gomollón F, Dignass A, Annese V, et al. 3rd European Evidence-based Consensus on the Diagnosis and Management of Crohn’s Disease 2016: Part 1: Diagnosis and Medical Management. J Crohns Colitis 2017;11:3–25. [DOI] [PubMed] [Google Scholar]
- 3.Lichtenstein GR, Loftus EV, Isaacs KL, et al. ACG clinical guideline: management of crohn’s disease in adults. Am. J. Gastroenterol 2018;113:481–517. [DOI] [PubMed] [Google Scholar]
- 4.Gearry RB, Ajlouni Y, Nandurkar S, et al. 5-Aminosalicylic acid (mesalazine) use in Crohn’s disease: a survey of the opinions and practice of Australian gastroenterologists. Inflamm. Bowel Dis 2007;13:1009–1015. [DOI] [PubMed] [Google Scholar]
- 5.Klag T, Stange EF, Wehkamp J. Management of Crohn’s disease - are guidelines transferred to clinical practice? United European Gastroenterol. J 2015;3:371–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Siegel CA. Refocusing IBD Patient Management: Personalized, Proactive, and Patient-Centered Care. Am. J. Gastroenterol 2018;113:1440–1443. [DOI] [PubMed] [Google Scholar]
- 7.Torres J, Boyapati RK, Kennedy NA, et al. Systematic review of effects of withdrawal of immunomodulators or biologic agents from patients with inflammatory bowel disease. Gastroenterology 2015;149:1716–1730. [DOI] [PubMed] [Google Scholar]
- 8.Ungaro RC, Limketkai BN, Jensen CB, et al. Stopping 5-aminosalicylates in patients with ulcerative colitis starting biologic therapy does not increase the risk of adverse clinical outcomes: analysis of two nationwide population-based cohorts. Gut 2018. [DOI] [PMC free article] [PubMed]
- 9.Singh S, Proudfoot JA, Dulai PS, et al. No Benefit of Concomitant 5-Aminosalicylates in Patients With Ulcerative Colitis Escalated to Biologic Therapy: Pooled Analysis of Individual Participant Data From Clinical Trials. Am. J. Gastroenterol 2018;113:1197–1205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ma C, Kotze PG, Almutairdi A, et al. Concomitant use of aminosalicylates is not associated with improved outcomes in patients with ulcerative colitis escalated to vedolizumab. Clin. Gastroenterol. Hepatol 2018. [DOI] [PubMed]
- 11.Ford AC, Kane SV, Khan KJ, et al. Efficacy of 5-aminosalicylates in Crohn’s disease: systematic review and meta-analysis. Am. J. Gastroenterol 2011;106:617–629. [DOI] [PubMed] [Google Scholar]
- 12.Lim W-C, Wang Y, MacDonald JK, et al. Aminosalicylates for induction of remission or response in Crohn’s disease. Cochrane Database Syst. Rev 2016;7:CD008870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Campbell S, Ghosh S. Effective maintenance of inflammatory bowel disease remission by azathioprine does not require concurrent 5-aminosalicylate therapy. Eur. J. Gastroenterol. Hepatol 2001;13:1297–1301. [DOI] [PubMed] [Google Scholar]
- 14.Kwak MS, Kim K-J, Cheon JH, et al. Comparison of concomitant mesalamine and immunomodulator therapy and immunomodulator monotherapy for crohn’s disease. Gastroenterol. Res. Pract 2018;2018:4826973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Duricova D, Pedersen N, Elkjaer M, et al. 5-aminosalicylic acid dependency in Crohn’s disease: a Danish Crohn Colitis Database study. J Crohns Colitis 2010;4:575–581. [DOI] [PubMed] [Google Scholar]
- 16.Colombel JF, Sandborn WJ, Reinisch W, et al. Infliximab, azathioprine, or combination therapy for Crohn’s disease. N. Engl. J. Med 2010;362:1383–1395. [DOI] [PubMed] [Google Scholar]
- 17.Gilissen LPL, Bierau J, Derijks LJJ, et al. The pharmacokinetic effect of discontinuation of mesalazine on mercaptopurine metabolite levels in inflammatory bowel disease patients. Aliment. Pharmacol. Ther 2005;22:605–611. [DOI] [PubMed] [Google Scholar]
- 18.Graaf P de, Boer NKH de, Wong DR, et al. Influence of 5-aminosalicylic acid on 6-thioguanosine phosphate metabolite levels: a prospective study in patients under steady thiopurine therapy. Br. J. Pharmacol 2010;160:1083–1091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Shah JA, Edwards CM, Probert CS. Should azathioprine and 5-aminosalicylates be coprescribed in inflammatory bowel disease?: an audit of adverse events and outcome. Eur. J. Gastroenterol. Hepatol 2008;20:169–173. [DOI] [PubMed] [Google Scholar]
- 20.Fonager K, Sørensen HT, Rasmussen SN, et al. Assessment of the diagnoses of Crohn’s disease and ulcerative colitis in a Danish hospital information system. Scand. J. Gastroenterol 1996;31:154–159. [DOI] [PubMed] [Google Scholar]
- 21.Larsen L, Drewes AM, Broberg MCH, et al. Changing infliximab prescription patterns in inflammatory bowel disease: a population-based cohort study, 1999–2014. Inflamm. Bowel Dis 2018;24:433–439. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.