Factors Associated with Response to Systemic Corticosteroids in Active Ulcerative Colitis: Results from a Prospective, Multicenter Trial

Abstract

Background: Among patients with ulcerative colitis, 30-50% receive corticosteroids within the first five years after diagnosis. We aimed to reconsider their effectiveness in the context of the biologic era.

Methods: In this prospective, multicenter study, patients with active ulcerative colitis (Lichtiger score ≥ 4) were eligible if initiating systemic corticosteroids. The primary endpoint was clinical response (decrease in the Lichtiger score of ≥50%) at week 4. Secondary endpoints included combined response defined as clinical response and any reduction in elevated biomarkers (CRP and/or calprotectin). Steroid dependence was assessed after three months.

Results: A total of 103 patients were included. Clinical response was achieved by 73% of patients, and combined response by 68%. A total of 15% of patients were steroid-dependent. Activity of colitis did not influence short-term response to treatment but increased the risk for steroid dependence. Biologic-naïve patients responded better than biologic-experienced patients. Past smoking history (OR 5.38 [1.71, 20.1], p = 0.003), hemoglobin levels (OR 0.76 [0.57, 0.99] for higher levels, p = 0.045), and biologic experience (OR 3.30 [1.08, 10.6], p = 0.036) were independently associated with nonresponse.

Conclusion: Disease activity was not associated with short-term response to systemic corticosteroids but was associated with steroid dependence in patients with active ulcerative colitis. Exposure to biologics negatively affects response rates.

Keywords: inflammatory bowel disease; systemic corticosteroids; ulcerative colitis.

Figure 1

Primary and secondary endpoints. (A) Outcome at week 4 after initiation of systemic corticosteroids according to the Lichtiger score (n = 103). Clinical response was defined as a decrease in the Lichtiger score of ≥50% from baseline to week 4, while clinical remission was defined as a Lichtiger score ≤ 3 at week 4. (B) Outcome at week 4 according to the combined endpoint, including the Lichtiger score, CRP, and calprotectin at week 4 (n = 99). Combined response was defined as a decrease in the Lichtiger score of ≥50% and any reduction in CRP and/or calprotectin from baseline to week 4 in patients with elevated biomarkers (CRP and/or calprotectin) at baseline, while combined remission was defined as a Lichtiger score ≤ 3 and CRP < 5 mg/L and/or calprotectin < 250 mg/kg at week 4.

Figure 2

Changes of biomarkers from baseline to week 4. Boxplots for Lichtiger score and biomarkers at baseline and at week 4 (n = 103) in patients with Lichtiger response (red) and Lichtiger nonresponse (green). Lichtiger response was defined as a decrease in the Lichtiger score of ≥50% from baseline to week 4. The Mann–Whitney U test was used for comparisons between the groups. p-values are provided for comparisons between Lichtiger response and nonresponse groups.

Figure 3

Outcome according to disease activity and biologic exposure. Disease activity: Outcome of patients (n = 103) at week 4 after initiation of systemic corticosteroids in relation to disease activity at baseline according to the Lichtiger score. Severely active colitis was defined as a Lichtiger score > 10 at baseline. Severely active colitis: n = 48, nonseverely active colitis: n = 55. (A) Lichtiger response and (B) combined response. Biologic exposure: Outcome of patients 4 weeks after initiation of systemic corticosteroids (n = 103) according to (C) Lichtiger response and (D) combined response divided into biologic-naive and biologic-experienced patients. Biologic-experienced patients were defined as patients with prior or ongoing biologic therapy. Biologic-experienced patients: n = 28, biologic-naïve patients: n = 75. p-values between groups are indicated above the graphs.

Figure 4

Prediction of response. ROC curves for prediction of response at week 4 before treatment initiation of systemic corticosteroids according to the (A) Lichtiger score and (B) combined endpoint. Biomarkers (lipocalin-2, calprotectin, and CRP) and the Lichtiger score were analyzed separately and together (all four together). For the multivariable model, all variables with a p < 0.2 in the univariable analysis were included.