Predicting relapse after stopping anti-TNF

Predicting relapse after stopping anti-TNF

October 18, 2022

Issue 20

Clinical Question

Can we predict relapse after anti-TNF cessation in CD?

Editor’s Bottom Line

Relapse of Crohn’s disease following withdrawal of anti-TNF is common, and remains difficult to predict.


Pauwels RWM, van der Woude CJ, Nieboer D, et al. Prediction of Relapse After Anti-Tumor Necrosis Factor Cessation in Crohn’s Disease: Individual Participant Data Meta-analysis of 1317 Patients From 14 Studies. Clin Gastroenterol Hepatol. 2022;20(8):1671–86.e16.


Researchers in multiple countries performed this meta-analysis of 14 cohort studies, including a total of 1,317 Crohn’s disease (CD) patients. All of the patients had achieved remission using an anti-tumor necrosis factor (TNF) agent, subsequently discontinued that drug and were followed for at least 12 months.

Results showed 632 patients experienced disease relapse a median 13 months after anti-TNF discontinuation, leading to a pooled one-year relapse rate of 38%. At two years, 52% of patients had experienced disease relapse.  

The investigators applied a previously studied prediction model developed as part of the diSconTinuation in CrOhn’s disease patients in stable Remission on combined therapy with Immunosuppressants (STORI) trial and found it did not discriminate well between relapsers and non-relapsers. Using the current dataset, the researchers developed a new predictive model that incorporated a number of risk factors. These included the presence of clinical symptoms at the time of discontinuation (Hazard Ratio [HR], 2.2; 95% Confidence Interval [CI], 1.2–4), CD diagnosis at 16 years of age or younger (HR, 1.5 vs. 17–40 years of age; 95% CI, 1.11–1.89), an absence of concomitant immunosuppressant use (HR, 1.4; 95% CI, 1.18–172), smoking (HR, 1.4; 95% CI, 1.15–1.67), use of a second line anti-TNF agent (HR, 1.3; 95% CI, 1.01–1.69), upper gastrointestinal tract disease (HR, 1.3; 95% CI, 0.96–1.79), use of adalimumab (HR, 1.22 vs. infliximab; 95% CI, 0.99–1.50), younger age at treatment cessation (HR, 1.2 for every 10 years; 95% CI, 1–1.33), higher C-reactive protein levels (HR, 1.04 per doubling of levels; 95% CI, 1–1.08), and longer disease duration (HR, 1.07 for every 5 additional years of disease; 95% CI, 0.98–1.17). Visible endoscopic inflammation at the time of anti-TNF discontinuation was not linked with higher risk of relapse.

The new prediction model had moderate discriminative ability (C-statistic, 0.59). For patients with three or fewer of these risk factors, the model had was 79% sensitive and 39% specific in predicting a roughly 22% risk of relapse within one year of discontinuation. For patients with more than three of these risk factors, the model was 37% sensitive and 78% specific in predicting a roughly 42% risk of relapse within one year.


Study Design: Meta-analysis

Funding: Not disclosed

Allocation: Not applicable

Setting: Multicenter

Level of Evidence: 2a