Patients with ulcerative colitis (UC) in deep remission who stopped anti-tumor necrosis factor (anti-TNF) therapy were more than four times as likely to flare over the next two years as patients who continued treatment — 52% vs. 12% — according to results of the randomized BIOSTOP trial, published in the Journal of Crohn’s and Colitis.
The authors concluded that anti-TNF withdrawal “should be restricted to selected, well-informed and highly motivated patients.”
“The BIOSTOP study was a large randomized controlled trial investigating the outcomes 2 years after withdrawal of anti-TNF compared to continued treatment in ulcerative colitis patients in confirmed deep remission,” lead author Ingrid Prytz Berset, MD, of Møre og Romsdal Hospital Trust in Ålesund, Norway, and the University of, told GI & Hepatology News. “We believe that the results of the BIOSTOP study give important and consolidated answers to clinicians and patients, about expected outcomes after anti-TNF withdrawal.”
The prospective, multicenter, open-label noninferiority trial randomized 172 patients with UC at 19 Norwegian hospitals. Participants had received first-line anti-TNF maintenance therapy for at least one year and were required to be in clinical, biochemical, and endoscopic remission before randomization. Remission was defined as a 6-point Mayo score of 1 or less, a Mayo endoscopic score of 0 or 1, and two consecutive fecal calprotectin values below 200 µg/g. Patients were assigned 1:1 to discontinue anti-TNF therapy or continue maintenance treatment and were followed for two years. Patients in either arm with an endoscopically confirmed flare were treated by restarting or optimizing anti-TNF or by optimizing nonbiologic therapy, depending on disease severity.
The primary endpoint was endoscopically confirmed remission at two years, with a prespecified noninferiority margin of 15 percentage points. In the per-protocol analysis, 80.5% of patients who stopped anti-TNF (66 of 82) were in endoscopic remission at two years vs. 96.3% (78 of 81) of those who continued — a 15.8-percentage-point difference that exceeded the margin, so noninferiority could not be claimed. The intention-to-treat analysis was similar (80.2% vs. 96.4%). The authors described the remission rate after withdrawal as significantly lower than with maintenance treatment.
Disease flares occurred in 52% of patients assigned to withdrawal (45 of 86) vs. 12% of those who continued anti-TNF (10 of 86). Median time to flare was 8.5 months in the withdrawal arm and 13.3 months in the maintenance arm. Clinical remission at two years, a secondary endpoint, was 90.4% after withdrawal vs. 98.8% with maintenance — a difference the authors reported as falling within the noninferiority margin.
“In our study significantly more patients had active ulcerative colitis after two years in the anti-TNF withdrawal group compared to the maintenance group. 53% of the withdrawal group and 12% of the maintenance group patients experienced an ulcerative colitis flare during the two years of study follow up,” Dr. Berset said.
Most patients who flared responded to retreatment, according to the study. Of the 45 withdrawal-group patients who flared, 27 restarted anti-TNF, and 20 (74%) regained clinical and endoscopic remission. Thirteen with milder relapses received optimized nonbiologic therapy, and 10 (77%) regained remission. Three patients had anti-TNF retreatment failure and were switched to vedolizumab; two responded, while one with severe colitis and cytomegalovirus infection required colectomy.
“Most ulcerative colitis flares were successfully treated,” Dr. Berset said. “Around 75% of the withdrawal group and 99% of the maintenance group patients achieved new remission after retreatment with the same anti-TNF or non-biologic agents.”
The authors identified no significant predictors of disease activity after anti-TNF withdrawal. Nearly half (48%) of patients assigned to withdrawal remained in steroid-free remission throughout the study. Disease-specific quality-of-life and fatigue scores at two years were significantly worse in the withdrawal arm, the authors reported.
For clinicians considering treatment withdrawal in patients in sustained remission, Dr. Berset emphasized careful patient selection and objective confirmation of remission.
“We conclude that anti-TNF treatment may be withdrawn in selected patients ... especially when safety issues must be considered, and in communities where medication costs are high,” Dr. Berset said. “We strongly recommend confirming deep remission by endoscopy and histology before withdrawal of anti-TNF.”
Dr. Berset also noted that only about one-quarter of participants received concomitant azathioprine [confirm: interview or paper]. At enrollment, 24% of patients were taking azathioprine, and most did not receive it during the trial. The authors observed a nonsignificant trend toward more flares among patients not taking azathioprine (P = .052) but noted the trial was not powered to assess this.
The authors identified several limitations. The COVID-19 pandemic paused and delayed recruitment and prevented some endoscopic follow-up, and pandemic-era guidance recommending against combination therapy in patients in remission reduced azathioprine use, which the authors said may have contributed to the flare rate after withdrawal. The open-label design introduced a risk of patient and study personnel bias. Fewer patients were recruited than planned, which may have affected the study’s power, although the authors said this was unlikely to have affected their conclusions.
The trial was funded by the Norwegian Research Council and Møre & Romsdal Hospital Trust. The funders had no role in study design, data collection, analysis, interpretation, or writing of the report. Several authors reported financial relationships with pharmaceutical and other organizations.