Impact of Thiopurine Discontinuation at Anti-Tumour Necrosis Factor Initiation in Inflammatory Bowel Disease Treatment

A Nationwide Danish Cohort Study

Sandra Bohn Thomsen; Ryan C. Ungaro; Kristine H. Allin; Rahma Elmahdi; Gry Poulsen; Mikael Andersson; Jean-Frederic Colombel; Tine Jess

Disclosures

Aliment Pharmacol Ther. 2022;55(9):1128-1138. 

In This Article

Results

A total of 10,352 anti-TNF exposed patients were identified from 2003 to 2018. Of these, 2,630 patients with IBD ([CD] n = 1590, [UC] n = 1040) were exposed to thiopurine prior to anti-TNF initiation (Figure 2). Thiopurine duration pre anti-TNF was 180 days or more in 733 patients (27.9%), 90–179 days in 509 patients (19.4%), 30–89 days in 725 patients (27.6%) and 0–29 days in 663 patients (25.2%). Of the included patients, 979 discontinued thiopurine after anti-TNF initiation and were followed for a median of four years, and 1651 people continued thiopurine and were followed for a median of six years, in total contributing 15,014 person-years of follow-up (PY). The median number of anti-TNF treatments given were 11 (1–25).

Figure 2.

Flowchart of study cohort selection. IBD, Inflammatory bowel disease; anti-TNF, anti-tumour necrosis factor

Most patients who initiated anti-TNF were diagnosed between 17 and 39 years of age (69.4%) whereas few were diagnosed below age 17 (7.8%) (Table 1). Patients were more likely to discontinue (vs continue) thiopurine in the later calendar period (45.8% vs 30.1%, 2015–2018, P < 0.01) compared to the early calendar periods (3.8% vs 8.1%, 2003–2006 and 20.6% vs 28.6%, 2007–2010). Thiopurine discontinuers had less frequently received thiopurine for more than 180 days pre-anti-TNF compared to continuers (20.6% vs 32.2%, P < 0.01). The groups did not differ regarding previous disease severity defined by healthcare utilization metrics nor by sex, age at IBD-diagnosis and IBD-subtype (Table 1).

Impact of Discontinuing Thiopurine on new Corticosteroid use, IBD-related Hospitalization, IBD Related Surgery and Death

For the primary composite outcome (new corticosteroid use, IBD-related hospitalization, IBD related surgery and death) we identified 572 events in the thiopurine discontinue group (IR 276.2 per 1000 PY) and 995 events in the thiopurine continue group (IR 208.2 per 1000 PY). This resulted in a statistically significantly increased risk of the composite primary outcome in thiopurine discontinuers with a crude HR of 1.19 (95% CI, 1.07 to 1.32) and an adjusted HR (aHR) of 1.22 (95% CI 1.10 to 1.36) (Table 2, Figure 3). When analysed per protocol, the risk of the composite outcome (including corticosteroid, hospitalization and major surgery) was still increased in thiopurine discontinuers (aHR 1.18; 95% CI, 1.00 to 1.38) (Table S5). Analyses of the individual components of the primary composite outcome revealed a statistically significantly increased risk of hospitalization (aHR 1.14; 95% CI 1.00 to 1.31), and new corticosteroid use in thiopurine discontinuers (aHR 1.27; 95% CI 1.13 to 1.44) (Table 2, Figure 3). No statistically significant differences were observed when analysing the remaining outcomes of IBD-related surgery and death (Table 2), although the direction of effects was similar to the effects for hospitalizations and new corticosteroid use. Sub-analyses showed that risk of the primary outcome was similarly increased in both CD and UC patients who discontinued thiopurine (Figure 4). Likewise, the remaining sub-analyses of the composite primary outcome showed no evidence of effect modification by diagnosis age, calendar period, sex and pre-anti-TNF thiopurine duration (Figure 4). Sensitivity analyses showed that patients who initially continued thiopurine but were no longer treated with thiopurine one-year post-anti-TNF (n = 334), also had a statistically significantly increased risk of the composite primary outcome, compared to continuers who were still on thiopurine one year after starting anti-TNF (aHR 1.20; 95% CI 1.01 to 1.41) (Table S3). Initial discontinuers who restarted thiopurine one-year post-anti-TNF (n = 236) did not have a statistically significantly increased risk of the composite primary outcome, compared to continuers who were still on thiopurine one year after starting anti-TNF (Table S3).

Figure 3.

Smoothed cumulative incidence of an outcome in patients with inflammatory bowel disease who discontinued vs continued thiopurine after anti-TNF initiation. A. Cumulative incidence of the composite outcome (new corticosteroids, IBD-related hospitalization, IBD-related major surgery, or death). B. Cumulative incidence of IBD-related major surgery. C. Cumulative incidence of hospitalization. D. Cumulative incidence of new oral corticosteroids. TNF - Thio. discont., anti-TNF and thiopurine discontinuer; TNF – Thio.cont., anti-TNF and thiopurine continuer; P-value is the test of interaction between the variable and the treatment groups

Figure 4.

Crude incidence rates and hazard risk of the composite outcome in patients with IBD who discontinued vs continued thiopurine after anti-TNF initiation stratified by sex, diagnosis age, IBD-subtype, calendar time and duration of thiopurine pre-anti-TNF. TNF - Thio.discont., anti-TNF and thiopurine discontinuer; TNF – Thio.cont., anti-TNF and thiopurine continuer; IR; incidence rate, HR, hazard ratio; CI, confidence interval; IBD, Inflammatory Bowel Disease; CD, Crohn's disease; UC, Ulcerative colitis; F, female; M, male; Composite outcome: new corticosteroids, hospitalization, major surgery, or death

For the secondary outcome of cancer, 27 patients who discontinued thiopurine developed cancer (IR 8.8 per 1000 PY) and 91 patients who continued thiopurine developed cancer (IR 13 per 1000 PY) resulting in no significant association with risk of developing any cancer (discontinuers vs continuers, aHR 0.67; 95% CI 0.43 to 1.03) (Table S4). Seven patients who discontinued thiopurine developed NMSC (IR 2.3 per 1000 PY), whereas 43 patients who continued thiopurine developed NMSC (IR 6.0 per 1000 PY), resulting in a statistically significantly increased risk of NMSC in thiopurine continuers (discontinuers vs continuers, aHR 0.39; 95% CI 0.17 to 0.86). When NMSC was excluded from all cancers, there was no increased risk of developing any other cancer (discontinuers vs continuers, aHR 0.89; 95% CI 0.53 to 1.49) (Table S4; Figure 5).

Figure 5.

Cumulative incidence of cancer in patients with IBD who discontinued vs continued thiopurine after anti-TNF initiation. A. Cumulative incidence of all cancers B. Cumulative incidence of all cancers excl. non-melanoma skin cancer. C. Cumulative incidence of non-melanoma skin cancer. TNF - Thio. discont., anti-TNF and thiopurine discontinuer; TNF -Thio.cont., anti-TNF and thiopurine continuer

processing....