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

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.


Introduction
Ulcerative colitis (UC) is a chronic immune-mediated disease affecting the colonic mucosa [1]. Until now, systemic corticosteroids are recommended as induction therapy for moderate to severely active disease [2,3]. Still, the likelihood of a UC patient receiving corticosteroids within the first five years after diagnosis is 30-50% [4,5]. In historical cohorts from the 1960s with limited sample size, remission rates of around two-thirds to three-quarters were reported [6][7][8], but new data on efficacy and factors influencing and predicting response are rare. These newer studies have mainly focused on acute severe colitis and factors influencing the necessity of a rescue therapy or colectomy [9][10][11][12], while investigations on patients with moderately active UC are lacking.
Here we present the clinical results of an investigator-initiated, prospective, multicenter trial originally performed to evaluate predictive microbial alterations before initiation of systemic corticosteroids in patients with active UC [13]. We assessed clinical and biochemical response and aimed to determine demographic, disease-specific, and clinical factors influencing and predicting response to treatment.

Study Population
From May 2018 to December 2020, patients with active UC were recruited from 18 study centers in Austria. Patients were eligible if they were suffering from active disease and were scheduled for treatment with oral or intravenous systemic corticosteroids. The diagnosis of UC needed to be either established before study inclusion or confirmed later in patients with suspected UC at inclusion. The diagnosis was made according to the local practice. Active UC was defined as Lichtiger score ≥ 4 and severely active colitis as Lichtiger score > 10 [14,15]. Exclusion criteria included bacterial, viral, or parasitic infections and a diagnosis of Crohn's disease. Stable ongoing medications could be continued during the study period. Corticosteroid dosing and tapering were performed in accordance with the clinical practice at each study center.

Data Collection
Patient characteristics and the Lichtiger score were assessed, and serum and stool samples were collected at baseline (time of steroid initiation) and after 4 weeks of corticosteroid therapy. Further follow-up was evaluated with telephone interviews 3 and 12 months after corticosteroid initiation for assessment of steroid dependence and colectomy rate. Patients refractory to corticosteroids could have an unscheduled study visit to start rescue therapy with infliximab or a calcineurin inhibitor and were labeled as nonresponders at week 4.

Primary Endpoint
The primary endpoint was Lichtiger response to systemic corticosteroids after 4 weeks of treatment. Lichtiger response was defined as a decrease in the Lichtiger score of ≥50% from baseline to week 4 [15] ( Table 1). The Lichtiger score is determined by eight variables: diarrhea, nocturnal stools, visible blood in stool, fecal incontinence, abdominal pain/cramping, general well-being, abdominal tenderness, and need for antidiarrheals. The score ranges from 0 (no activity) to 21 points (maximal activity) [15,16].
Steroid dependence after three months (defined as unable to reduce corticosteroids below the equivalent of prednisolone 10 mg per day or as recurrent active disease after stopping steroids [24]) and colectomy rates after 12 months were assessed.
Baseline patient demographics, disease characteristics, medical therapies, and biomarkers associated with response to corticosteroids were examined. The accuracy of predicting response to treatment at baseline with available biomarkers and the Lichtiger score was evaluated.

Statistical Analysis
Patient characteristics were reported as absolute and relative frequencies for categorical data and as median and range for numerical data. Comparisons between groups were carried out using Mann-Whitney U, Kruskal-Wallis, chi-square, or Fisher exact tests as appropriate. A p-value of 0.05 or less was considered statistically significant. Logistic regression analysis was performed to identify factors associated with nonresponse to corticosteroids. Variables with a p-value of <0.2 in the univariable analysis were included in the multivariable model. The accuracy of prediction of response to corticosteroids was assessed using receiver operating characteristic (ROC) curves. All statistical analyses were conducted using R version 4.2.3 [25].

Clinical Efficacy of Systemic Corticosteroids in Active UC
A total of 103 patients were included in the main efficacy analysis. Despite clinical improvement to corticosteroids, seven patients received additional biologics between baseline and week 4 (Table S1). These seven patients were excluded in a sensitivity analysis (n = 96) ( Figure S1). Patient characteristics of both cohorts at baseline are shown in Table S2.
baseline and week 4 (Table S1). These seven patients were excluded in a sensitivity analysis (n = 96) ( Figure S1). Patient characteristics of both cohorts at baseline are shown in Table  S2.
A total of 75 (73%) patients achieved Lichtiger response and 59 (57%) received Lichtiger remission at week 4 ( Figure 1A). Patients not achieving Lichtiger response had higher past smoking rates (78% vs. 46%, p = 0.005) and lower hemoglobin levels  Table S3). . 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. . 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. Fecal calprotectin and CRP levels at baseline and at week 4 correlated with the Lichtiger score at the same time points, while fecal LCN-2 only correlated with the clinical score at week 4 ( Figure S2).

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.

Steroid Dependence and Colectomy Rates
One out of the 103 patients included in the study was steroid refractory and required rescue therapy with infliximab. The patient was therefore labeled as a nonresponder at week 4. A total of 13 (15%) out of the 87 patients with data at month 3 were steroid dependent. Steroid-dependent patients at month 3 had higher rates of severely active disease (85% vs. 45%, p = 0.008) and higher leukocyte levels (10.4 (7.0, 19.5) vs. 8.1 (3.8, 23.7), p = 0.006), CRP levels (40 (7, 236) vs. 10 (0, 227), p = 0.006), and a higher Lichtiger score (14 (9, 15) vs. 10 (5, 17), p = 0.008) at baseline (Table S4). Corticosteroid dose at baseline did not correlate with Lichtiger scores and did not influence response to treatment. One (1%) out of the 70 patients with data at month 12 required colectomy. This patient suffered from severely active colitis at baseline.

Influence of Disease Activity on Response to Corticosteroids
To estimate the effect of disease activity on the efficacy of corticosteroids, we compared response rates in patients with (Lichtiger score > 10, n = 48) and without (Lichtiger score ≤ 10, n = 55) severely active colitis (Table S5). Lichtiger response was achieved by 36 (75%) patients with and 39 (71%) patients without severely active colitis at baseline (p = 0.642). Equally, no difference could be observed for combined response (Figure 3). Corticosteroid dosing did not differ significantly between groups (Table S5). Although patients with severely active colitis had significantly higher fecal calprotectin, CRP, thrombocyte, and leukocyte levels at baseline and tended to have higher hospitalization rates, levels of all biomarkers and the Lichtiger score aligned between groups after 4 weeks of treatment ( Figure S3).

Influence of Medical Therapies on Clinical Response to Corticosteroids
Lichtiger response rates were numerically higher and combined response rates significantly higher (p = 0.018) in biologic-naïve patients (n = 75) compared to biologic-experienced (n = 28) patients ( Figure 3). Biologic experience was defined as prior or ongoing biologic therapy. Biologic-experienced patients had longer disease duration (p = 0.003) and higher rates of prior corticosteroid (p < 0.001) and immunomodulator (p = 0.001) exposure (Table S6). The Lichtiger score and biomarkers at baseline were comparable between the two groups, but biologic-naïve patients had a significantly lower Lichtiger score (p = 0.015) at week 4 ( Figure S4). Furthermore, we assessed if treatment outcomes differed according to prior corticosteroid and azathioprine exposure. No significant differences could be observed.

Influence of Medical Therapies on Clinical Response to Corticosteroids
Lichtiger response rates were numerically higher and combined response rates significantly higher (p = 0.018) in biologic-naïve patients (n = 75) compared to biologic-experienced (n = 28) patients (Figure 3). Biologic experience was defined as prior or ongoing biologic therapy. Biologic-experienced patients had longer disease duration (p = 0.003) and higher rates of prior corticosteroid (p < 0.001) and immunomodulator (p = 0.001) exposure (Table S6). The Lichtiger score and biomarkers at baseline were comparable between the two groups, but biologic-naïve patients had a significantly lower Lichtiger score (p = 0.015) at week 4 ( Figure S4). Furthermore, we assessed if treatment outcomes differed according to prior corticosteroid and azathioprine exposure. No significant differences could be observed.

Modelling Response Prediction by Baseline Covariates
Concerning the Lichtiger response, the score itself and biomarkers at baseline showed modest ability to predict response to corticosteroids at week 4. Prediction could be improved by combining those parameters (

Sensitivity Analysis
The sensitivity analysis (n = 96) was conducted to exclude potential bias by seven patients initiating biologics concomitantly to corticosteroids between baseline and week 4 despite clinical improvement. No substantial differences to the primary results could be observed.

Discussion
This prospective, multicenter study highlights the efficacy of systemic corticosteroids in active UC on clinical and biochemical endpoints and suggests that short-term response to treatment is dependent on biologic exposure rather than on baseline disease activity, whereas disease activity impacts longer-term steroid dependence.
Despite the common use of systemic corticosteroids in active UC and the ongoing recommendation of their usage in treatment guidelines [2,3], actual data on the efficacy of this treatment are rare. In the initial studies investigating systemic corticosteroids for the treatment of UC more than 60 years ago, only 24% of patients with mild to moderate colitis and 33% with severe colitis did not achieve clinical remission, but the quality of evidence

Sensitivity Analysis
The sensitivity analysis (n = 96) was conducted to exclude potential bias by seven patients initiating biologics concomitantly to corticosteroids between baseline and week 4 despite clinical improvement. No substantial differences to the primary results could be observed.

Discussion
This prospective, multicenter study highlights the efficacy of systemic corticosteroids in active UC on clinical and biochemical endpoints and suggests that short-term response to treatment is dependent on biologic exposure rather than on baseline disease activity, whereas disease activity impacts longer-term steroid dependence.
Despite the common use of systemic corticosteroids in active UC and the ongoing recommendation of their usage in treatment guidelines [2,3], actual data on the efficacy of this treatment are rare. In the initial studies investigating systemic corticosteroids for the treatment of UC more than 60 years ago, only 24% of patients with mild to moderate colitis and 33% with severe colitis did not achieve clinical remission, but the quality of evidence from these data is limited due to the restricted sample size and the lack of validated clinical scores and biomarkers to measure disease activity [6][7][8]. Furthermore, these data were assessed before the availability of advanced therapies for the treatment of inflammatory bowel diseases. Recent studies investigating the use of corticosteroids in acute severe UC reported treatment failure rates of 26 to 33% [9][10][11]. Another study including patients with newly diagnosed moderate to severe UC found a treatment failure rate of 37% after 3 months [12]. Although we studied a slightly different patient collective including patients with all kinds of baseline disease activity, the rates of clinical nonresponse of around 25% according to the Lichtiger score in our study are in line with published studies.
Interestingly, baseline disease activity did not seem to influence short-term response in our cohort. Even in the subgroup of patients with severely active colitis according to a Lichtiger score > 10, we found high clinical response rates of 75% according to the Lichtiger score. Ben-Horin et al. recently reported similar response rates in patients with acute severe colitis seven days after treatment induction [9]. Of note, the definition of severe colitis (Lichtiger score > 10 vs. Lichtiger score ≥ 10) and Lichtiger response (a ≥50% drop in Lichtiger score vs. a drop of >3 points) were stricter in our cohort.
Rates of steroid dependence after three months (15%) and colectomy rates within 12 months (1%) were low in our cohort. In patients suffering from acute severe colitis, longterm colectomy rates even in responders to intravenous corticosteroids or infliximab were previously reported to be 25% [26]. This difference may be explained by the inclusion of patients with mild and moderate disease activity and by the shorter follow-up in our cohort. Patients with higher disease activity at baseline had an increased risk of becoming steroid dependent and may therefore profit from an early introduction of advanced therapies in clinical practice to avoid relevant side effects of long-term corticosteroid therapy. As shown in our study, 35% of patients already experienced side effects to corticosteroids within 4 weeks of therapy.
According to our data, biologic exposure seems to be negatively associated with response to corticosteroids. Overlapping cofactors such as longer disease duration and higher rates of pretreatment with corticosteroids and immunomodulators in biologicexposed patients may bias these findings. However, when analyzed by itself, response rates did not differ regarding pretreatment with either corticosteroids or immunomodulators. The lower treatment efficacy of biologics and small molecules in UC patients who were previously exposed to biologics, especially TNF alpha antibodies, has been observed for the most advanced therapies [27][28][29][30]. We are not aware of this already being reported for corticosteroids. Together, this suggests that biologic-exposed patients represent a subgroup of UC patients that are especially difficult to treat.
Factors associated with Lichtiger and combined nonresponse to corticosteroids were prior smoking and lower baseline hemoglobin levels. The association of lower hemoglobin levels indicating treatment failure to corticosteroids in UC patients has been reported previously [11]. Furthermore, Ardizzone et al. published that female sex was associated with better outcomes three months after initiation of corticosteroids, a finding we could not confirm [12].
Fecal calprotectin has been shown to predict one-year sustained response and mucosal healing at treatment week 8 or after induction therapy of biologics [31][32][33]. In patients with acute severe colitis receiving intravenous corticosteroids, a baseline calprotectin level above 800 µg/g was able to predict the need for rescue therapy with an acceptable sensitivity of 80% but a weak specificity [11]. An ulcerative colitis endoscopic index of severity (UCEIS) > 6 at admission and fecal calprotectin levels >1000 µg/g on day 3 were independent predictors of steroid failure in acute severe colitis according to a prospective, single-center study [34]. We found that a combination of the Lichtiger score and biomarkers (CRP, fecal calprotectin, and fecal LCN-2) was of limited use to predict response, and the use of a multivariable model identified by regression analysis at baseline improved prediction only slightly.
The limitations of our study include the restricted sample size, the lack of data on endoscopic or histological endpoints, and the lack of a placebo group. Not all patients had a follow-up available until month 12, and data on dose adaptations of concomitant therapies are lacking. Another important limitation is the absence of a uniform dosing and tapering scheme of corticosteroids, but this may not have impacted results substantially because starting doses did not differ according to baseline disease activity and did not impact response rates. Further, this may not have affected rates of steroid dependence at month 3 because even longer tapering schemes due to higher baseline doses should lead to doses under 10 mg after 12 weeks if dosing had not been increased again or the reduction was stopped.
The strength of our study comprises the prospective, multicenter design with realworld data from a cohort with active UC with various disease activities. Further, the study provides high-quality evidence for corticosteroid usage in the biologic era and suggests for the first time the negative association of biologic exposure with clinical and biochemical response to corticosteroids.
In conclusion, our study confirms a good short-term efficacy of systemic corticosteroids on symptom relief and biomarker endpoints in UC patients independent of disease activity and indicates a negative influence of biologic exposure on response rates. Patients with a high baseline disease activity had a higher risk of steroid dependence, supporting early concomitant introduction of immunomodulators or advanced therapies in this subset of patients.

Institutional Review Board Statement:
The study was approved by the research ethics committees at the Medical University of Graz (EK 29-316 ex 16/17) and local ethic committees of participating hospitals and is registered at clinicaltrials.gov (NCT03460847). The study was performed in accordance with the ethical standards laid down in the Declaration of Helsinki and its amendments.
Informed Consent Statement: Written informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.