Abstract
Introduction
Ulcerative colitis (UC) is a diffuse, nonspecific inflammatory disease of unknown etiology. Although corticosteroids are essential for inducing remission in moderate to severe UC, steroid dependence and refractoriness remain significant clinical obstacles. Predicting which patients will develop steroid dependence or refractoriness remains challenging. Furthermore, difficult-to-treat (D-to-T) cases – those unresponsive to advanced therapies – have emerged as additional therapeutic concerns. This study aimed to identify factors associated with refractory UC, including steroid-dependent, steroid-refractory, and D-to-T presentations.
Methods
A total of 216 patients with UC who received treatment at Sapporo Medical University Hospital between April 2017 and December 2023 were retrospectively analyzed. Patients were classified into four groups: steroid-naive (SN), steroid-free (SF), steroid-dependent (SD), and steroid-refractory (SR). D-to-T cases were identified within the SD and SR groups. Patient background characteristics were obtained from electronic medical records. Variables analyzed included sex, age of onset, disease duration, alcohol consumption, smoking status, disease phenotype, extraintestinal manifestations (EIMs), and cytomegalovirus (CMV) and Clostridioides difficile infection status.
Results
Significant differences were observed in the prevalence of EIM and CMV reactivation across the four groups (p < 0.001 for EIM; p < 0.001 for CMV). The prevalence of EIM was significantly higher in the SR group compared with the SN group (45.5% vs. 7.7%, p = 0.002) and the SF group (45.5% vs. 4.5%, p < 0.001). CMV reactivation was more frequent in the SR group than in the SN group (36.4% vs. 1.3%, p < 0.001) and the SF group (36.4% vs. 4.5%, p < 0.004). Multivariable analysis revealed that, in comparison with the SN group, the SD group was independently associated with EIM (odds ratio [OR] = 4.59, 95% confidence interval [CI]: 1.35–15.6) and CMV reactivation (OR = 12.5, 95% CI: 1.31–119). Compared to the SF group, the SD group was associated with EIM (OR = 5.71, 95% CI: 1.44–22.7). The SR group was independently associated with EIM (OR = 12.8, 95% CI: 2.51–64.9) and CMV reactivation (OR = 65.1, 95% CI: 4.39–964) compared to the SN group. Compared to the SF group, the SR group was associated with EIM (OR = 16.4, 95% CI: 2.95–90.8) and CMV reactivation (OR = 17.0, 95% CI: 2.03–142). There are also significant associations between D-to-T status and both EIM (p < 0.01) and CMV reactivation (p = 0.037).
Conclusions
Among patients with UC, the presence of EIM and CMV reactivation was significantly associated with steroid dependence, steroid refractoriness, and resistance to biologic and molecular-targeted therapies.
Keywords: Ulcerative colitis, Steroid, Steroid dependence, Refractory, Difficult-to-treat, Extraintestinal manifestations, Cytomegalovirus
Introduction
Ulcerative colitis (UC) is a diffuse, nonspecific inflammatory disease of unknown etiology [1]. Although its precise cause has not been determined, several factors – such as genetic predisposition, high-fat or high-sugar diets, and intestinal dysbiosis – are considered to contribute to disease development [1–3]. The treat-to-target approach for patients with UC emphasizes clinical remission, endoscopic healing, restoration of quality of life, and prevention of disability [4]. The mainstay of induction treatment for patient with active UC are 5-aminosalicylic acid and corticosteroids [5, 6]. Steroids are effective for achieving remission, but a subset of patients becomes steroid-dependent or steroid-refractory, posing ongoing clinical challenges. While the development of advanced therapies (ATs) has improved outcomes for patients with refractory UC [7], some remain unresponsive even to biologics or molecular-targeted agents. Moreover, it remains difficult to predict which patients will develop steroid dependence or refractoriness. A better understanding of the clinical characteristics of patients with refractory UC may facilitate improved disease management and personalized treatment strategies. This study aimed to identify factors associated with refractory UC, including steroid-dependent (SD), steroid-refractory (SR), and difficult-to-treat (D-to-T) cases.
Methods
Patients
This single-center, retrospective observational study included 216 patients with UC who received treatment at Sapporo Medical University Hospital between April 2017 and December 2023. Patients with UC were identified through electronic medical records. Diagnosis was confirmed based on previous physician assessments and clinical and endoscopic findings at our institution, in accordance with the Ministry of Health, Labour and Welfare’s guidelines for inflammatory bowel disease (IBD) [1]. Patients were categorized into four groups: steroid-naive (SN), steroid-free (SF), steroid-dependent (SD), and steroid-refractory (SR). Definitions were as follows: (1) SF status referred to patients who responded to steroids and maintained steroid discontinuation; (2) steroid-dependent referred to patients with response to steroids but relapse if dose reduced <10 mg/day within 3 months or relapse within 3 months of discontinuation; (3) steroid-refractory indicated active disease despite 0.75–1 mg/kg/day oral prednisolone for 4 weeks or failure of intravenous corticosteroids for ≥1 week [1, 8]. Two patients who were classified as SD based on their response to budesonide were excluded from the analysis. Clinical remission was defined as either (1) rectal bleeding = 0 and stool frequency = 0 based on the Mayo score or (2) a partial Mayo score <3 with no individual subscore >1 [4]. We compared the four groups based on baseline characteristics, including sex, age at onset, disease duration, alcohol use, smoking history, disease phenotype, and extraintestinal manifestations (EIMs). Laboratory and endoscopic parameters assessed at diagnosis included hemoglobin, platelet count, serum albumin, C-reactive protein (CRP), Mayo score, and Ulcerative Colitis Endoscopic Index of Severity (UCEIS). Cytomegalovirus (CMV) reactivation was defined as a positive result from mucosal polymerase chain reaction with a cutoff value of >1,650 copies/µg, as reported by Yamakawa et al. [9] and immunostaining of biopsy specimens from colonic mucosa. Clostridioides difficile infection was diagnosed by detecting glutamate dehydrogenase and A/B toxins in stool samples. 5-ASA intolerance was determined by the attending physician based on symptoms such as fever, diarrhea, or abdominal pain following 5-ASA administration. We performed multiple comparisons across the four groups to identify variables that showed significant differences. We also defined refractory UC as encompassing both the SD and SR groups [1]. In contrast, D-to-T was defined as failure of two or more biologics or molecular-targeted therapies with different mechanisms of action, based on the International Organization for the Study of Inflammatory Bowel Disease (IOIBD) [10]. We analyzed the association between clinical factor and ATs, including biologic and molecular-targeted therapies, in patients with refractory UC.
All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.
Statistical Analysis
All statistical analyses were performed using EZR, a modified version of R Commander that includes functions frequently applied in biostatistics [11]. The Kruskal-Wallis test and one-way analysis of variance were used to evaluate differences among groups. For post hoc analysis, the Steel-Dwass and Tukey methods were employed. A p value <0.05 was considered statistically significant.
Covariates associated with steroid-dependent and steroid-refractory status were analyzed using univariable and multivariable logistic regression. Due to extensive missing data, clinical and endoscopic activity scores were excluded from these analyses. Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs).
Associations with D-to-T status were evaluated using the chi-square test. All authors had access to the study data and approved the final manuscript.
Results
A total of 216 patients with UC were included. The background characteristics of the patients are shown in Table 1. The percentages of females were 54.6%. The median age of disease onset was 33 years old and the median of disease duration was 120 months. The most common disease phenotype was extensive colitis. EIM was observed in 25 (11.6%) of all patients. CMV reactivation was observed in 13 (6.0%) of all patients.
Table 1.
Characteristics of the patients with UC
| | All (n = 216) |
|---|---|
| Male/female ratio, n (%) | 98 (45.3)/118 (54.6) |
| Age of disease onset, median (IQR), years | 33 (21–47) |
| Disease duration, median (IQR), months | 120 (66–216) |
| Alcohol, n (%) | 48 (22.1) |
| Smoking, n (%) | |
| Current | 23 (10.6) |
| Former | 34 (15.7) |
| Never | 115 (53.2) |
| Unknown | 44 (20.4) |
| Disease phenotype, n (%) | |
| E1 (proctitis) | 45 (20.8) |
| E2 (left-side colitis) | 38 (17.6) |
| E3 (extensive colitis) | 121 (56.0) |
| Non-classifiable (NC) (atypical or discontinuous distribution) | 2 (0.9) |
| EIMsa, n (%) | 25 (11.6) |
| Primary sclerosing cholangitis | 8 |
| Arthralgia | 6 |
| Uveitis | 1 |
| Erythema nodosum | 7 |
| Thrombosis | 5 |
| Bronchiolitis | 1 |
| Hemoglobin, median (IQR), g/dL | 12.5 (10.6–13.8) |
| Platelet, median (IQR), ×104/μL | 30.9 (26.2–37.6) |
| Albumin, median (IQR), g/dL | 3.8 (3.4–4.3) |
| CRP, median (IQR), mg/dL | 0.26 (0.1–0.72) |
| Mayo score, median (IQR) | 6 (4–9) |
| UCEIS, median (IQR) | 4 (3–5) |
| 5-ASA intolerance, n (%) | 35 (16.2) |
| C. difficile infection, n (%) | 5 (2.3) |
| CMV reactivation, n (%) | 13 (6.0) |
CRP, C-reactive protein.
aThree patients have multiple EIMs.
The Characteristics of the Four Groups
We have shown the characteristics of the four groups in Table 2. The number of patients were 78 in the steroid-naïve group (SN group), 66 in the steroid-free group (SF group), 61 in the steroid-dependent group (SD group), and 11 in the steroid-refractory group (SR group), respectively.
Table 2.
Characteristics of the patients for SN group, SF group, SD group, and SR group
| Group | |||||
|---|---|---|---|---|---|
| SN (n = 78) | SF (n = 66) | SD (n = 61) | SR (n = 11) | p value | |
| Male/female ratio, n (%) | 31 (39.7)/47 (60.3) | 31 (47.8)/35 (52.2) | 32 (52.5)/29 (47.5) | 4 (36.4)/7 (63.6) | 0.449 |
| Age of disease onset, median (IQR), years | 35 (22–48) | 34 (21–48) | 30 (20–39) | 45 (26–53.5) | 0.197 |
| Disease duration, median (IQR), months | 114 (60–207) | 132 (69–252) | 132 (84–228) | 120 (78–234) | 0.412 |
| Alcohol, n (%) | 15 (19.2) | 14 (20.9) | 17 (27.9) | 2 (18.2) | 0.982 |
| Smoking, n (%) | 0.08 | ||||
| Current | 7 (9) | 5 (7.5) | 9 (14.8) | 2 (18.2) | |
| Former | 6 (7.7) | 10 (14.9) | 15 (24.6) | 3 (27.3) | |
| Never | 42 (53.8) | 35 (52.2) | 34 (557) | 4 (36.4) | |
| Unknown | 23 (29.5) | 17 (25.3) | 3 (4.9) | 2 (18.2) | |
| Disease phenotype | 0.01 | ||||
| E1 (proctitis) | 24 (30.8) | 12 (17.9) | 9 (14.8) | 0 (0) | |
| E2 (left-side colitis) | 13 (16.7) | 15 (22.4) | 8 (13.1) | 2 (18.2) | |
| E3 (extensive colitis) | 38 (48.7) | 34 (50.7) | 42 (68.9) | 7 (63.6) | |
| NC (atypical or discontinuous distribution) | 0 (0) | 1 (1.5) | 0 (0) | 1 (9.1) | |
| EIMsa | 6 (7.7) | 3 (4.5) | 11 (18) | 5 (45.5) | <0.001 |
| Primary sclerosing cholangitis | 3 | 2 | 1 | 2 | |
| Arthralgia | 1 | 1 | 4 | 0 | |
| Uveitis | 0 | 0 | 1 | 0 | |
| Erythema nodosum | 0 | 0 | 6 | 1 | |
| Thrombosis | 1 | 0 | 1 | 3 | |
| Bronchiolitis | 1 | 0 | 0 | 0 | |
| Hemoglobin, median (IQR), g/dL | 12.7 (12.1–13.9) | 12.6 (10.1–13.9) | 11.1 (8.5–13.5) | 11.4 (11.4–12) | 0.26 |
| Platelet, median (IQR), ×104/μL | 29.3 (25.4–32.0) | 29.95 (27.1–36.8) | 39.3 (30.4–42.4) | 31.2 (30.9–31.4) | 0.31 |
| Albumin, median (IQR), g/dL | 4.2 (3.8–4.4) | 3.6 (3.3–4.2) | 3.6 (3.3–4) | 3.95 (3.9–4.0) | 0.08 |
| CRP, median (IQR), mg/dL | 0.1 (0.1–0.27) | 0.38 (0.19–1.23) | 0.495 (0.16–2.2) | 0.26 (0.26–5.68) | 0.02 |
| Mayo score, median (IQR) | 4 (2.5–9) | 5.5 (4.3–9) | 9.0 (7–10) | 9.0 (9–9.5) | 0.001 |
| UCEIS, median (IQR) | 3.5 (2–4) | 4.5 (2.8–6) | 5 (4.3–5.8) | 5 (5–5.8) | 0.04 |
| 5-ASA intolerance, n (%) | 6 (7.7) | 13 (20) | 13 (21.3) | 3 (27.3) | 0.09 |
| C. difficile infection, n (%) | 1 (1.3) | 1 (1.56) | 3 (4.8) | 0 | 0.461 |
| CMV reactivation, n (%) | 1 (1.3) | 3 (4.5) | 5 (8.2) | 4 (36.4) | <0.001 |
CRP, C-reactive protein.
aSome patients have multiple EIMs.
Demographic and Clinical Factors (Age of Onset, Smoking, Disease Phenotype)
The median age of onset was higher in the SR group (45 years) than that in the other groups. The proportion of current and former smokers was lower in the SN group than that in the other groups. The most common disease phenotype was extensive colitis in all groups. Furthermore, the proportions of extensive colitis were higher in the SD and SR groups than these in the SN and SF groups.
Disease Activity and Comorbidities (EIM, CMV Reactivation, Mayo Score)
EIM, CMV reactivation, and Mayo scores significantly differed among the four groups. EIMs included primary sclerosing cholangitis (PSC), arthralgia, uveitis, erythema nodosum, thrombosis, and bronchiolitis. EIM prevalence was significantly higher in the SR group compared to the SN group (45.5% vs. 7.7%, p = 0.002) and the SF group (45.5% vs. 4.5%, p < 0.001) (Table 3). The SD group had a higher prevalence of EIM (18%) than the SN and SF groups, although these differences were not statistically significant. Erythema nodosum and thrombosis were more frequent in the SD and SR groups. Erythema nodosum accounted for 54% of all EIMs in the SD group and 20% in the SR group. Thrombosis accounted for 60% of all EIMs in the SR group.
Table 3.
Comparison of each group in EIM
| Group | p value |
|---|---|
| SN vs. SF (7.7% vs. 4.5%) | 0.866 |
| SN vs. SD (7.7% vs. 18%) | 0.255 |
| SN vs. SR (7.7% vs. 45.5%) | 0.002 |
| SF vs. SD (4.5% vs. 18%) | 0.07 |
| SF vs. SR (4.5% vs. 45.5%) | <0.001 |
| SD vs. SR (18% vs. 45.5%) | 0.188 |
The bold values indicate p values of 0.05 or below, demonstrating statistically significant differences.
The incidence of CMV reactivation was higher in the SR (36.4%) and SD (8.2%) groups than in the SN (1.3%) and SF (4.5%) groups. Significant differences were noted between the SR and SN groups (p < 0.001) and between the SR and SF groups (p = 0.004) (Table 4). To identify clinical factors associated with steroid dependence and refractoriness, univariable and multivariable analyses were conducted comparing the SD and SR groups to the SN and SF groups (Tables 5, 6).
Table 4.
Comparison of each group in CMV reactivation
| Group | p value |
|---|---|
| SN vs. SF (1.3% vs. 4.5%) | 0.637 |
| SN vs. SD (1.3% vs. 8.2%) | 0.194 |
| SN vs. SR (1.3% vs. 36.4%) | <0.001 |
| SF vs. SD (4.5% vs. 8.2%) | 0.834 |
| SF vs. SR (4.5% vs. 36.4%) | 0.004 |
| SD vs. SR (8.2% vs. 36.4%) | 0.04 |
The bold values indicate p values of 0.05 or below, demonstrating statistically significant differences.
Table 5.
Univariable logistic regression analysis to identify covariates associated with SD and SN/SF in patients with UC
| Covariates | Odds ratio (vs. SN) | Odds ratio (vs. SF) |
|---|---|---|
| Female | 0.598 (0.304–1.18) | 0.803 (0.400–1.61) |
| Onset age | 0.98 (0.958–1.00) | 0.982 (0.962–1.00) |
| Disease duration | 1.000 (0.999–1.010) | 1.000 (0.997–1.000) |
| Alcohol | 1.16 (0.512–2.64) | 1.06 (0.459–2.46) |
| Smoking | 2.28 (1.01–5.14) | 1.65 (0.741–3.66) |
| EIM | 2.64 (0.916–7.61) | 4.62 (1.22–17.5) |
| Hemoglobin | 0.687 (0.484–0.974) | 0.869 (0.678–1.11) |
| Platelet | 1.12 (0.997–1.25) | 1.04 (0.96–1.14) |
| Albumin | 0.407 (0.140–1.19) | 0.935 (0.33–2.65) |
| CRP | 1.32 (0.915–1.90) | 1.10 (0.847–1.42) |
| CMV reactivation | 6.88 (0.781–60.5) | 1.87 (0.429–8.2) |
CRP, C-reactive protein.
Table 6.
Univariable logistic regression analysis to identify covariates associated with SR and SN/SF in patients with UC
| Covariates | Odds ratio (vs. SN) | Odds ratio (vs. SF) |
|---|---|---|
| Female | 1.15 (0.312–4.28) | 1.55 (0.486–6.46) |
| Onset age | 1.020 (0.981–1.06) | 1.010 (0.98–1.05) |
| Disease duration | 1.000 (0.997–1.010) | 1.000 (0.995–1.010) |
| Alcohol | 0.911 (0.165–5.02) | 0.833 (0.15–4.64) |
| Smoking | 4.04 (0.94–17.3) | 2.92 (0.68–12.4) |
| EIM | 10.0 (2.35–42.6) | 17.5 (3.33–91.9) |
| Hemoglobin | 0.702 (0.328–1.5) | 0.981 (0.608–1.58) |
| Platelet | 1.04 (0.836–1.29) | 0.983 (0.823–1.17) |
| Albumin | 0.841 (0.104–6.81) | 2.81 (0.145–54.5) |
| CRP | 1.38 (0.945–2.02) | 1.25 (0.898–1.75) |
| CMV reactivation | 44.0 (4.31–449) | 12.0 (2.22–64.9) |
CRP, C-reactive protein.
Table 5 shows factors associated with the SD group compared to the SN and SF groups. Smoking (OR = 2.28, 95% CI: 1.01–5.14) and hemoglobin level (OR = 0.687, 95% CI: 0.484–0.974) were significantly associated with the SD group compared to the SN group. EIM was significantly associated with the SD group compared to the SF group (OR = 4.62, 95% CI: 1.22–17.5).
Table 6 presents factors associated with the SR group compared to the SN and SF groups. EIM (OR = 10.0, 95% CI: 2.35–42.6) and CMV reactivation (OR = 44.0, 95% CI: 4.31–449) were significantly associated with the SR group compared to the SN group. Compared to the SF group, EIM (OR = 17.5, 95% CI: 3.33–91.9) and CMV reactivation (OR = 12.0, 95% CI: 2.22–64.9) were significantly associated with the SR group.
Next, multivariable analyses were conducted to further evaluate associations with the SD and SR groups (Tables 7, 8). Variables included EIM and CMV reactivation, adjusted for age and sex.
Table 7.
Multivariable logistic regression analysis adjusted for age and sex to identify covariates associated with SD and SN/SF in patients with UC
| Covariates | Odds ratio (vs. SN) | Odds ratio (vs. SF) |
|---|---|---|
| EIM | 4.59 (1.35–15.6) | 5.71 (1.44–22.7) |
| CMV reactivation | 12.5 (1.31–119) | 3.25 (0.665–15.9) |
Table 8.
Multivariable logistic regression analysis adjusted for age and sex to identify covariates associated with SR and SN/SF in patients with UC
| Covariates | Odds ratio (vs. SN) | Odds ratio (vs. SF) |
|---|---|---|
| EIM | 12.8 (2.51–64.9) | 16.4 (2.95–90.8) |
| CMV reactivation | 65.1 (4.39–964) | 17.0 (2.03–142) |
Table 7 shows that, in comparison with the SN group, the SD group was independently associated with EIM (OR = 4.59, 95% CI: 1.35–15.6) and CMV reactivation (OR = 12.5, 95% CI: 1.31–119). Compared to the SF group, the SD group was associated with EIM (OR = 5.71, 95% CI: 1.44–22.7). Table 8 shows the SR group was independently associated with EIM (OR = 12.8, 95% CI: 2.51–64.9) and CMV reactivation (OR = 65.1, 95% CI: 4.39–964) compared to the SN group. Compared to the SF group, the SR group was associated with EIM (OR = 16.4, 95% CI: 2.95–90.8) and CMV reactivation (OR = 17.0, 95% CI: 2.03–142).
We further analyzed refractory UC in the SD and SR groups to assess the association between EIM or CMV reactivation and the use of ATs (Table 9; Fig. 1a, b). Table 9 shows that the number of ATs used was significantly higher in the presence of EIM (1 [1, 2] vs. 3 [1–4], p = 0.004) or CMV reactivation (1 [1, 2] vs. 3 [2–4], p = 0.008) in patients with refractory UC. Moreover, both the time from UC onset to initiation of ATs and the duration of the first AT tended to be shorter in the presence of EIM or CMV reactivation. Figure 1 shows the progression of ATs from induction to remission. Figure 1a and b show data on EIM and CMV reactivation, respectively. Among the ATs, the anti-integrin antibodies did not yield clinical remission, but the anti-TNFα antibodies or JAK inhibitors ultimately yielded clinical remission. All patients who achieved clinical remission with JAK inhibitors received concomitant antiviral therapy.
Table 9.
Association of EIM and CMV reactivation with advanced therapies in patients with refractory UC
| Variable | Refractory UC (n = 72) | With EIM (n = 16) | p value | With CMV reactivation (n = 9) | p value |
|---|---|---|---|---|---|
| Patients treated with any ATs, n (%) | 55 | 13 | | 8 | |
| Number of ATs per patient, median (IQR) | 1 (1–2) | 3 (1–4) | 0.004 | 3 (2–4) | 0.008 |
| Time to first AT initiation, median (IQR), months | 30.5 (17–111.5) | 21 (16–32.75) | 0.126 | 20.5 (12.25–97.25) | 0.52 |
| Duration of the first AT, median (IQR), weeks | 38.5 (13–94.25) | 24 (7.75–76) | 0.552 | 13 (9.5–17.5) | 0.03 |
AT, advanced therapy.
Fig. 1.
Clinical course of refractory UC under ATs. a Clinical course of refractory UC with EIMs under ATs. b Clinical course of refractory UC with CMV reactivation under ATs. AT, advanced therapy; TNF, tumor necrosis factor. *One patient was on antiviral therapy during the same period. **All patients were on antiviral therapy during the same period. ***One patient was on antiviral therapy during the same period.
Lastly, a chi-square analysis was conducted to examine associations with D-to-T cases (Table 10). A total of 12 patients were classified as D-to-T. The analysis revealed significant associations between D-to-T status and both EIM (p < 0.01) and CMV reactivation (p = 0.037).
Table 10.
Association of EIM and CMV reactivation with difficult-to-treat cases in patients with refractory UCa
| | Refractory UC (n = 72) | Difficult-to-treat (n = 12) | p value |
|---|---|---|---|
| EIM | 16 | 9 | <0.01 |
| CMV reactivation | 9 | 4 | 0.037 |
aDifficult-to-treat defined as failure of two or more biologics or molecular-targeted therapies with different mechanisms of action.
Discussion
This study analyzed clinical factors associated with steroid dependence and steroid refractoriness in patients with UC, identifying both EIM and CMV reactivation as significant contributors to these treatment-resistant forms. Additionally, EIM and CMV reactivation appeared to be risk factors for refractoriness to biologics and molecular-targeted therapies. While previous studies have reported that high-dose steroid therapy prior to advanced treatment and extensive colitis are risk factors for patients with D-to-T UC [12, 13], our analysis specifically focused on the association between D-to-T cases and patient background characteristics or comorbidities, including EIM and CMV reactivation. This is the first study to examine the relationship between D-to-T cases and patient background or comorbidities.
First, it is known that 6–40% of patients with IBD develop EIM, which varies by race and region [14]. For instance, the prevalence in European patients with IBD is approximately 36.6% [15], compared to 7.9–10.5% in Asian populations [14, 16, 17]. In our cohort, 11.6% of patients with UC had EIM, which is consistent with existing data. Notably, EIM was significantly more prevalent in the SR group (45.5%) than in the SN (7.7%) and SF (4.5%) groups. There were no significant differences among the SD, SN, and SF groups; however, multivariable analysis demonstrated that EIM was independently associated with the SD group compared to the SN and SF groups. EIM was also significantly associated with D-to-T cases. Patients with EIM tended to receive a greater number of ATs and had a shorter duration of first AT compared to those without EIM. Yiyoung et al. [18] reported that patients with UC who were EIM-positive exhibited lower remission rates and higher relapse frequencies than those who were EIM-negative. EIM has also been linked to colectomy in patients with UC, including pediatric populations [19, 20]. Therefore, our findings regarding the association between EIM and both steroid dependence and refractoriness are consistent with previous reports. The development of EIM is believed to involve immune responses to shared antigens and susceptibility alleles in the human leukocyte antigen (HLA) region [21, 22]. HLA molecules activate adaptive immunity by presenting antigens to CD4+ T cells. Roussomoustakaki et al. [23] reported that the HLA-DR103 genotype in patients with UC was associated with more extensive disease and a higher frequency of EIM. Bhagat et al. [24] demonstrated that colonic epithelial cells and extraintestinal tissues share common peptides. Vavricka et al. [25] found that NF-κB, a transcription factor regulating immune and inflammatory responses, was upregulated in both intestinal and extraintestinal tissues in patients with IBD [26–28]. These findings suggest that immune responses to intestinal antigens contribute to both intestinal and extraintestinal inflammation and that genetic predisposition may also play a role. Furthermore, inflammatory signaling linked to EIM may affect the intestinal mucosa, potentially diminishing the therapeutic efficacy of steroids, biologics, and molecular-targeted therapies.
Second, in this study, CMV reactivation was observed more frequently in the SR group than in the SF group. Prior studies have also reported associations between CMV reactivation and steroid dependence or refractoriness [29, 30], consistent with our findings. Steroids act by binding to glucocorticoid receptor α (GRα), but this effect is attenuated when binding occurs at glucocorticoid receptor β (GRβ) [31]. An elevated GRβ ratio has been associated with steroid refractoriness in patients with UC [32]. GRβ expression increases during CMV reactivation compared to its latent phase [33]. Moreover, CMV infection of THP-1 cells enhances the production of mature IL-1β [34], and IL-1β upregulates both GRα and GRβ expression in nasal polyp tissue, with a greater increase in GRβ [35]. This shift in the GRβ to GRα ratio, induced by IL-1β, may contribute to reduced steroid efficacy. These findings support our conclusion that CMV reactivation contributes to steroid dependence and refractoriness. Additionally, clinical flares related to CMV reactivation may necessitate dose escalation or a switch in biologic agents when antiviral treatment alone is insufficient. Yamakawa et al. [9] suggested that early initiation of antiviral therapy may reduce the risk of colectomy.
In our study, the presence of both EIM and CMV reactivation was associated with primary nonresponse to anti-integrin antibodies, whereas anti-TNFα antibodies or JAK inhibitors were suggested to be potentially effective. Anti-integrin antibodies selectively inhibit lymphocyte infiltration into the gut [36], which may limit their effectiveness in controlling systemic inflammation, including EIM. In contrast, anti-TNFα antibodies and JAK inhibitors exert broad immunosuppressive effects on systemic immune responses. Several reports have demonstrated the efficacy of anti-TNFα antibodies or JAK inhibitors for EIM in patients with UC [37–41].
Regarding CMV reactivation, Prosch et al. [42] have reported that TNFα increased the transcriptional activity of human CMV genes. Nakase et al. [43] reported that following the administration of anti-TNFα antibodies in patients with UC complicated by CMV reactivation, mucosal CMV polymerase chain reaction changed from positive to negative. Furthermore, Pillet et al. [44] demonstrated that patients receiving maintenance anti-TNFα antibodies were not at an increased risk of CMV reactivation. These findings are consistent with and support our present observation that anti-TNFα antibodies treatment is effective in patients with UC complicated by CMV reactivation. Since IFN plays a key role in controlling viral reactivation [45], JAK inhibitors may increase the risk of CMV reactivation [46]. In this cohort, all patients with UC complicated by CMV reactivation who achieved clinical remission with JAK inhibitors received antiviral therapy. However, further accumulation of cases is needed to determine the impact of JAK inhibitors on CMV infection and reactivation in patient with active UC.
This study has some limitations. First, the small number of patients in the SR group (n = 11) restricted the number of variables that could be included in the multivariable analysis for steroid refractoriness. As a result, there may be insufficient statistical power and a risk of overlooking potential confounding factors. Additionally, hemoglobin and platelet levels were excluded due to substantial missing data. Second, as most patients at our facility were referred from other hospitals, there may be a selection bias toward inclusion of more cases with EIM or refractory disease. The presence of EIM was determined by each attending physician; thus, the possibility of overdiagnosis or underdiagnosis cannot be excluded.
In conclusion, the present study demonstrated that EIM and CMV reactivation in patients with UC were associated with steroid dependence and refractoriness. Both factors may contribute to the D-to-T condition in UC.
Acknowledgments
Kotaro Akita and Mayuko Erata contributed equally to this work.
Statement of Ethics
This study was approved by the Institutional Review Boards of Sapporo Medical University School of Medicine (IRB No. 302-101). An opt-out informed consent protocol was adapted for the use of participant data for research purposes. This consent procedure was reviewed and approved by the Institutional Review Boards of Sapporo Medical University School of Medicine, Approval No. [302-101], date of decision [01/03/2024].
Conflict of Interest Statement
Hiroshi Nakase was a member of the journal’s Editorial Board at the time of submission. All other authors confirm no conflicts of interest.
Funding Sources
This work was supported by the Health and Labour Sciences Research Grants for research on intractable diseases from the Ministry of Health, Labor and Welfare (MHLW) of Japan (Investigation and Research for intractable Inflammatory Bowel Disease) (Grant No. 20316729, principal investigator: Hiroshi Nakase) and by the Japanese Society for the Promotion of Science (JSPS) KAKENHI (Grant Nos. 24K11158 and JP23K15078, principal investigator: Hiroshi Nakase and Yoshihiro Yokoyama).
Author Contributions
Kotaro Akita: data curation, writing-original draft, and data analysis and interpretation. Mayuko Erata, Yuta Shimomori, Tomoe Kazama, and Hiroki Kurumi: data curation. Yoshihiro Yokoyama and Hiroshi Nakase: writing – review and editing and supervision. Masanori Nojima: data analysis and interpretation.
Funding Statement
This work was supported by the Health and Labour Sciences Research Grants for research on intractable diseases from the Ministry of Health, Labor and Welfare (MHLW) of Japan (Investigation and Research for intractable Inflammatory Bowel Disease) (Grant No. 20316729, principal investigator: Hiroshi Nakase) and by the Japanese Society for the Promotion of Science (JSPS) KAKENHI (Grant Nos. 24K11158 and JP23K15078, principal investigator: Hiroshi Nakase and Yoshihiro Yokoyama).
Data Availability Statement
The data used in this study were obtained from electronic medical records and contain confidential patient information. Due to privacy regulations, the data are not publicly available. Anonymized data may be provided upon reasonable request and approval by the Ethics Committee. Further inquiries can be directed to the corresponding author.
References
- 1. Nakase H, Uchino M, Shinzaki S, Matsuura M, Matsuoka K, Kobayashi T, et al. Evidence-based clinical practice guidelines for inflammatory bowel disease 2020. J Gastroenterol. 2021;56(6):489–526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Kasapoglu M, Yadavalli R, Nawaz S, Althwanay A, AlEdani EM, Kaur H, et al. The impact of microbiome interventions on the progression and severity of inflammatory bowel disease: a systematic review. Cureus. 2024;16(5):e60786–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Agus A, Denizot J, Thévenot J, Martinez-Medina M, Massier S, Sauvanet P, et al. Western diet induces a shift in microbiota composition enhancing susceptibility to adherent-invasive Escherichia coli infection and intestinal inflammation. Sci Rep. 2016;6:19032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Turner D, Ricciuto A, Lewis A, D’Amico F, Dhaliwal J, Griffiths AM, et al. STRIDE-II: an update on the Selecting Therapeutic targets in Inflammatory Bowel Disease (STRIDE) initiative of the International Organization for the Study of IBD (IOIBD): determining therapeutic goals for treat-to-target strategies in IBD. Gastroenterology. 2021;160(5):1570–83. [DOI] [PubMed] [Google Scholar]
- 5. Mitrev N, Kariyawasam V. Treatment endpoints in ulcerative colitis: does one size fit all? World J Gastrointest Pharmacol Ther. 2024;15(2):91591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Dallegri F, Ottonello L, Ballestrero A, Bogliolo F, Ferrando F, Patrone F. Cytoprotection against neutrophil-derived hypochlorous acid: a potential mechanism for the therapeutic action of 5-aminosalicylic acid in ulcerative colitis. Gut. 1990;31(2):184–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Dai Y, Yang W, Xu L, Pan P, Liu S, Sun Y, et al. Comparative efficacy of different targeted therapies in patients with moderate-to-severe ulcerative colitis: systematic review, network meta-analysis, and mechanistic overview. Pharmacol Res Perspect. 2025;13(3):e70108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Dignass A, Eliakim R, Magro F, Maaser C, Chowers Y, Geboes K, et al. Second European evidence-based consensus on the diagnosis and management of ulcerative colitis. Part 1: definitions and diagnosis. J Crohns Colitis. 2012;6(10):965–90. [DOI] [PubMed] [Google Scholar]
- 9. Yamawaka T, Kitamoto H, Nojima M, Kazama T, Wagatsuma K, Ishigami K, et al. The association between antigenemia, histology with immunohistochemistry, and mucosal PCR in the diagnosis of ulcerative colitis with concomitant human cytomegalovirus infection. J Gastroenterol. 2023;58(1):44–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Parigi TL, D’Amico F, Abreu MT, Dignass A, Dotan I, Magro F, et al. Difficult-to-treat inflammatory bowel disease: results from an International consensus meeting. Lancet Gastroenterol Hepatol. 2023;8(9):853–9. [DOI] [PubMed] [Google Scholar]
- 11. Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant. 2013;48(3):452–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Parigi TL, Massimino L, Carini A, Gabbiadini R, Bertoli P, Allocca M, et al. Prevalence, characteristics, management, and outcomes of difficult-to-treat inflammatory bowel disease. J Crohns Colitis. 2025;19(3):jjae145. [DOI] [PubMed] [Google Scholar]
- 13. Matsuoka K, Igarashi A, Sato N, Mizuno N, Ishii M, Iizuka M, et al. Clinical characteristics of patients with difficult-to-treat ulcerative colitis: a nested case-control study using a Japanese claims database. Intest Res. 2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Yeo MK, Park JH, Kang SH, Moon HS, Sung JK, Jeong HY, et al. What are the risk factors for extraintestinal manifestations in inflammatory bowel diseases? Medicine. 2023;102(9):e33031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Greuter T, Vavricka SR. Extraintestinal manifestations in inflammatory bowel disease: epidemiology, genetics, and pathogenesis. Expert Rev Gastroenterol Hepatol. 2019;13(4):307–17. [DOI] [PubMed] [Google Scholar]
- 16. Rawal KK, Shukla VP, Chikani S, Thakkar M, Ruparelia M, Chudasama RK. Prevalence of extraintestinal manifestations in ulcerative colitis and associated risk factors. Indian J Gastroenterol. 2021;40(5):477–82. [DOI] [PubMed] [Google Scholar]
- 17. Ohta Y, Taida T, Kato J, Ogasawara S, Oyama Y, Mamiya Y, et al. Clinical features focusing on extraintestinal manifestations in Japanese patients with inflammatory bowel diseases: far East 1000. Digestion. 2023;104(4):328–34. [DOI] [PubMed] [Google Scholar]
- 18. Kwon Y, Kim ES, Choe YH, Kim MJ. Increased demand for therapeutic drugs in pediatric ulcerative colitis patients with extraintestinal manifestations. Front Pediatr. 2022;10:853019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Parragi L, Fournier N, Zeitz J, Scharl M, Greuter T, Schreiner P, et al. Colectomy rates in ulcerative colitis are low and decreasing: 10-year follow-up data from the Swiss IBD cohort study. J Crohns Colitis. 2018;12(7):811–8. [DOI] [PubMed] [Google Scholar]
- 20. Orlanski-Meyer E, Aardoom M, Ricciuto A, Navon D, Carman N, Aloi M, et al. Predicting outcomes in pediatric ulcerative colitis for management optimization: systematic review and consensus statements from the pediatric inflammatory bowel disease-ahead program. Gastroenterology. 2021;160(1):378–402.e22. [DOI] [PubMed] [Google Scholar]
- 21. Hedin CRH, Vavricka SR, Stagg AJ, Schoepfer A, Raine T, Puig L, et al. The pathogenesis of extraintestinal manifestations: implications for IBD research, diagnosis, and therapy. J Crohns Colitis. 2019;13(5):541–54. [DOI] [PubMed] [Google Scholar]
- 22. Vavricka SR, Schoepfer A, Scharl M, Lakatos PL, Navarini A, Rogler G. Extraintestinal manifestations of inflammatory bowel disease. Inflamm Bowel Dis. 2015;21(8):1982–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Roussomoustakaki M, Satsangi J, Welsh K, Louis E, Fanning G, Targan S, et al. Genetic markers may predict disease behavior in patients with ulcerative colitis. Gastroenterology. 1997;112(6):1845–53. [DOI] [PubMed] [Google Scholar]
- 24. Bhagat S, Das KM. A shared and unique peptide in the human colon, eye, and joint detected by a monoclonal antibody. Gastroenterology. 1994;107(1):103–8. [DOI] [PubMed] [Google Scholar]
- 25. Vavricka SR, Galván JA, Dawson H, Soltermann A, Biedermann L, Scharl M, et al. Expression patterns of TNFα, MAdCAM1, and STAT3 in intestinal and skin manifestations of inflammatory bowel disease. J Crohns Colitis. 2018;12(3):347–54. [DOI] [PubMed] [Google Scholar]
- 26. Hillmer EJ, Zhang H, Li HS, Watowich SS. STAT3 signaling in immunity. Cytokine Growth Factor Rev. 2016;31:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Liu T, Zhang L, Joo D, Sun SC. NF-κB signaling in inflammation. Signal Transduct Target Ther. 2017;2:17023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Wang N, Liang H, Zen K. Molecular mechanisms that influence the macrophage M1-M2 polarization balance. Front Immunol. 2014;5:614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Lv Y, Han F, Jia Y, Wan ZR, Gong LL, Liu H, et al. Is cytomegalovirus infection related to inflammatory bowel disease, especially steroid-resistant inflammatory bowel disease? A meta-analysis. Infect Drug Resist. 2017;10:511–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Gerçeker E, Saygı F, Avcı A, Yuceyar H. Steroid resistance/dependence might be an alarming feature for cytomegalovirus infection among ulcerative colitis patients with increased disease activity. Cureus. 2022;14(10):e30333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Lewis-Tuffin LJ, Cidlowski JA. The physiology of human glucocorticoid receptor beta (hGRbeta) and glucocorticoid resistance. Ann N Y Acad Sci. 2006;1069:1–9. [DOI] [PubMed] [Google Scholar]
- 32. Fujishima S, Takeda H, Kawata S, Yamakawa M. The relationship between the expression of the glucocorticoid receptor in biopsied colonic mucosa and the glucocorticoid responsiveness of ulcerative colitis patients. Clin Immunol. 2009;133(2):208–17. [DOI] [PubMed] [Google Scholar]
- 33. Wang S, Dou Y, Yang H, Ni A, Zhang R, Qian J. Alteration of glucocorticoid receptors and exacerbation of inflammation during lytic cytomegalovirus infection in THP-1 cells. FEBS Open Bio. 2017;7(12):1924–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Xu X, Cai J, Wang X, Lu Y, Guo B, Lai M, et al. Human cytomegalovirus infection activates NLRP3 inflammasome by releasing mtDNA into the cytosol in human THP-1 cells. Microbiol Immunol. 2023;67(6):303–13. [DOI] [PubMed] [Google Scholar]
- 35. Wang Z, Li P, Zhang Q, Lv H, Liu J, Si J. Interleukin-1β regulates the expression of glucocorticoid receptor isoforms in nasal polyps in vitro via p38 MAPK and JNK signal transduction pathways. J Inflamm. 2015;12(1):3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Soler D, Chapman T, Yang LL, Wyant T, Egan R, Fedyk ER. The binding specificity and selective antagonism of vedolizumab, an anti-alpha4beta7 integrin therapeutic antibody in development for inflammatory bowel diseases. J Pharmacol Exp Ther. 2009;330(3):864–75. [DOI] [PubMed] [Google Scholar]
- 37. Peyrin-Biroulet L, Van Assche G, Gómez-Ulloa D, García-Álvarez L, Lara N, Black CM, et al. Systematic review of tumor necrosis factor antagonists in extraintestinal manifestations in inflammatory bowel disease. Clin Gastroenterol Hepatol. 2017;15(1):25–36.e27. [DOI] [PubMed] [Google Scholar]
- 38. Vavricka SR, Gubler M, Gantenbein C, Spoerri M, Froehlich F, Seibold F, et al. Anti-TNF treatment for extraintestinal manifestations of inflammatory bowel disease in the Swiss IBD cohort study. Inflamm Bowel Dis. 2017;23(7):1174–81. [DOI] [PubMed] [Google Scholar]
- 39. Siemanowski B, Regueiro M. Efficacy of infliximab for extraintestinal manifestations of inflammatory bowel disease. Curr Treat Options Gastroenterol. 2007;10(3):178–84. [DOI] [PubMed] [Google Scholar]
- 40. Wang Y, Wan Z, Jin R, Xu T, Ouyang Y, Wang B, et al. Tofacitinib for extraintestinal manifestations of inflammatory bowel disease: a literature review. Int Immunopharmacol. 2022;105:108517. [DOI] [PubMed] [Google Scholar]
- 41. García MJ, Brenes Y, Vicuña M, Bermejo F, Sierra-Ausín M, Vicente R, et al. Persistence, effectiveness, and safety of upadacitinib in crohn’s disease and ulcerative colitis in real life: results from a Spanish Nationwide study (ureal study). Am J Gastroenterol. 2024;120(7):1593–604. [DOI] [PubMed] [Google Scholar]
- 42. Prösch S, Wendt CE, Reinke P, Priemer C, Oppert M, Krüger DH, et al. A novel link between stress and Human Cytomegalovirus (HCMV) infection: sympathetic hyperactivity stimulates HCMV activation. Virology. 2000;272(2):357–65. [DOI] [PubMed] [Google Scholar]
- 43. Pillet S, Jarlot C, Courault M, Del Tedesco E, Chardon R, Saint-Sardos P, et al. Infliximab does not worsen outcomes during flare-ups associated with cytomegalovirus infection in patients with ulcerative colitis. Inflamm Bowel Dis. 2015;21(7):1580–6. [DOI] [PubMed] [Google Scholar]
- 44. Nakase H, Chiba T. TNF-alpha is an important pathogenic factor contributing to reactivation of cytomegalovirus in inflamed mucosa of colon in patients with ulcerative colitis: lesson from clinical experience. Inflamm Bowel Dis. 2010;16(4):550–1. [DOI] [PubMed] [Google Scholar]
- 45. Dağ F, Dölken L, Holzki J, Drabig A, Weingärtner A, Schwerk J, et al. Reversible silencing of cytomegalovirus genomes by type I interferon governs virus latency. PLoS Pathog. 2014;10(2):e1003962. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Chan A, Cockburn E, Huang JY, Segal JP. A case report describing the failure of upadacitinib as sequential medical rescue therapy in treatment resistant acute severe ulcerative colitis.. Ann Surg Case Rep Images. 2024;1(6):1056. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data used in this study were obtained from electronic medical records and contain confidential patient information. Due to privacy regulations, the data are not publicly available. Anonymized data may be provided upon reasonable request and approval by the Ethics Committee. Further inquiries can be directed to the corresponding author.