Long-term clinical and endoscopic efficacy of vedolizumab and therapeutic persistence in patients with ulcerative colitis in China: a multicenter observational study
Article information
Abstract
Background/Aims
Limited data are available on the long-term efficacy of vedolizumab (VDZ) in Asian patients with ulcerative colitis (UC), particularly regarding endoscopic outcomes.
Methods
In this multicenter, retrospective, real-world study, UC patients treated with VDZ were evaluated for steroid-free remission, endoscopic remission (Mayo endoscopic score ≤ 1), and mucosal healing (Mayo endoscopic score = 0) at weeks 52, 104, and 156. Treatment persistence was analyzed using Kaplan-Meier curves.
Results
Among 239 patients, steroid-free remission was achieved in 59.4%, 52.9%, and 40.3% at weeks 52, 104, and 156, respectively. Endoscopic remission and mucosal healing rates were 53.3% and 32.2% at week 52, 39.8% and 27.5% at week 104, and 34.8% and 26.2% at week 156. The probability of treatment persistence was 76.8%, 63.3%, and 56.4% at these time points. Anti-tumor necrosis factor (anti-TNF) exposure was independently associated with a higher risk of failing to achieve long-term VDZ persistence (hazard ratio [HR], 2.754; 95% confidence interval [CI], 1.478–5.135; P= 0.001), whereas achieving early steroid-free remission after induction was an independent protective factor (HR, 0.480; 95% CI, 0.235–0.978; P= 0.043). VDZ-related adverse events occurred in 7.1% of patients, with no serious events reported.
Conclusions
VDZ demonstrated long-term clinical and endoscopic efficacy with a high safety profile in Chinese UC patients. Notably, anti-TNF-naive patients with early steroid-free remission were more likely to maintain long-term treatment persistence.
INTRODUCTION
Ulcerative colitis (UC) is a chronic, nonspecific inflammatory disease affecting the colon and rectum. Typical symptoms include abdominal pain, diarrhea, rectal bleeding, and tenesmus, all of which significantly impair patients’ quality of life [1]. In recent years, both the incidence and prevalence of UC have continued to rise across Asia, making it a growing public health concern [2]. Vedolizumab (VDZ), a humanized monoclonal antibody targeting the α4β7 integrin, offers a new treatment option for UC patients [3]. Evidence from the pivotal GEMINI trials showed that VDZ can induce and maintain clinical remission, and long-term extension studies have supported its durable efficacy and favorable safety profile [4,5]. Despite these advances, most existing data come from Western populations, and evidence from Asian patients, particularly regarding long-term clinical and endoscopic outcomes, remains limited [6-9]. To address this gap, we conducted a multicenter real-world study to evaluate the long-term effectiveness of VDZ, aiming to provide additional evidence for its clinical use in Asian UC patients.
METHODS
1. Study Design
This retrospective, observational, real-world cohort study was conducted at 3 centers in China (the Sixth Affiliated Hospital of Sun Yat-sen University, Guangzhou, Guangdong; Sir Run Run Shaw Hospital of Zhejiang University School of Medicine, Hangzhou, Zhejiang; the First Affiliated Hospital of Sun Yatsen University, Guangzhou, Guangdong). Patients with UC who received at least 1 dose of VDZ from October 2020 onward were considered for inclusion, and follow-up data were collected through December 2024, which served as the final data lock point. This study was approved by the Ethics Committee of the Sixth Affiliated Hospital, Sun Yat-sen University (2022ZSLYEC-326). The requirement for informed consent was waived by the Ethics Committee.
2. Inclusion and Exclusion Criteria
The inclusion criteria were as follows: (1) patients had a confirmed diagnosis of UC based on current guidelines [10,11]; (2) patients prescribed VDZ; (3) patients were ≥ 18 years old; (4) patients with moderate-to-severe activity according to the Mayo score, or with mild activity but refractory to traditional drugs. The exclusion criteria were as follows: (1) patients in clinical remission at baseline; (2) patients with insufficient clinical data to evaluate efficacy at baseline and week 52.
3. Data Collection
In our routine clinical practice, patients were scheduled for visits at week 14 and 52, week 104, and week 156 using predefined clinical and objective outcome measures. Endoscopies were conducted by qualified endoscopists at each participating center as part of routine clinical care. Data were extracted from the electronic medical records, including sex, age, active smoking status at baseline, disease duration, symptoms, surgical history, disease location classified by Montreal classification, disease activity based on Mayo score, medication history, endoscopic manifestations, laboratory indexes, therapeutic regimens, and adverse events (AEs). All endoscopic images were independently reviewed by 2 experienced endoscopists blinded to clinical data and treatment outcomes. Discrepancies were resolved through joint reassessment and consensus.
4. Investigated Drug
VDZ was administered intravenously at a dose of 300 mg during weeks 0, 2, and 6 for induction, followed by every 8 weeks for maintenance. In patients with inadequate response or loss of response to the standard regimen, dose escalation was permitted. Dose escalation during the therapy period was at the discretion of the physician, and the selection of dose escalation regimen was made through shared decision-making between the physician and the patient. This could include shortening the medication interval, therapy reinduction, or other approaches.
5. Outcomes and Definitions
The primary outcome of this study was steroid-free remission at week 52. Secondary outcomes included clinical response, endoscopic remission, and mucosal healing at week 52, as well as long-term steroid-free remission and endoscopic efficacy. We also assessed treatment persistence, predictors of long-term therapy continuation, and AEs.
The Mayo score was used to evaluate treatment efficacy, including the partial Mayo score and the Mayo endoscopic score [12]. Clinical remission was defined as a partial Mayo score ≤2 with no individual sub-score exceeding 1. Steroid-free remission was defined as patients in clinical remission without concurrent use of steroids. Endoscopic remission was defined as a Mayo endoscopic score ≤1, and mucosal healing was defined as a Mayo endoscopic score of 0. Disease severity at baseline was evaluated using Mayo score, with mild being defined as a score of 3–5, moderate as a score of 6–10, and severe as a score of 11–12.
To determine the proportion of patients achieving clinical and endoscopic response or remission at week 52, 2 years, and 3 years, patients who discontinued VDZ due to treatment failure were counted as non-responders. For outcomes assessed during the induction phase, induction was defined as the first 14 weeks after treatment initiation. Patients who responded to treatment but did not have sufficient follow-up time were not included in the calculation of the overall response rate at specific time points. Endoscopic efficacy was evaluated only among patients with available endoscopic data. A sensitivity analysis was also performed using multiple imputation to estimate endoscopic outcomes in patients with missing endoscopic assessments but available clinical data.
Treatment persistence was defined as the time from the first VDZ infusion to treatment discontinuation for any reason, including loss of response, AEs, or patient decision. Patients still receiving VDZ at the data lock (December 2024) were censored at their last follow-up or infusion date. For patients lost to follow-up, the censoring time was defined as the date of their last available clinical record.
6. Statistical Analysis
Patient characteristics were described as either mean and standard deviation or median and interquartile range for continuous variables, and as number of cases and proportions for categorical variables. Proportions were compared using the chi-square test or Fisher test. Quantitative variables were compared using the Student t test or its non-parametric counterparts. Missing data were handled using multiple imputation (n=5) to reduce potential bias from incomplete observations. The imputation procedure incorporated relevant baseline covariates and available outcome variables and was performed using the “mice” R package.
The Kaplan-Meier curve was used to describe treatment persistence. To account for competing causes of treatment discontinuation, cumulative incidence functions were estimated using the Fine-Gray method (R package “cmprsk”). Discontinuation due to ineffectiveness and other causes (including personal decision and death) were treated as competing risks, with treatment continuation considered as censored observations.
To identify factors influencing long-term drug persistence, a univariate analysis was first conducted to assess the association of each potential factor with drug persistence. Subsequently, variables with a P-value of less than 0.1 were selected for inclusion in the multivariate Cox regression analysis. The hazard ratios (HRs) and 95% confidence intervals (CIs) for each included variable were also analyzed. For binary outcomes assessed at a predefined time point (e.g., endoscopic remission and mucosal healing at Week 52), logistic regression analyses were performed, and odds ratios (ORs) with 95%CIs were reported. Dose escalation was treated as a time-varying covariate, as it could occur at any point during follow-up. This variable was incorporated into a time-dependent Cox regression model, in which the exposure status was updated at the time of dose escalation for each patient. All other covariates were treated as fixed baseline variables. Multicollinearity among covariates included in the multivariable Cox regression model was assessed using the variance inflation factor (VIF) approach, implemented via the R package “car”. VIF <5 was considered indicative of no significant multicollinearity.
All statistical tests were two-tailed, with statistical significance defined as P<0.05. Statistical analyses were performed using R software (version 4.4.3, R Foundation for Statistical Computing, Vienna, Austria) and SPSS 26.0 (IBM Corp., Armonk, NY, USA).
RESULTS
1. Study Population
A total of 271 patients diagnosed with UC received VDZ between October 2020 and December 2024. Among them, 239 patients met the inclusion criteria and were included in the study (Fig. 1). The median follow-up duration was 89.6 weeks (interquartile range, 51.7–148.7 weeks). Tumor necrosis factor inhibitor-naive (anti-TNF-naive) patients accounted for 81.2% of the cohort. The severity of UC was categorized as mild in 17.6% of patients, moderate in 60.3% and severe in 22.2% of patients. Baseline characteristics of the study population are summarized in Table 1, and subgroup comparisons between anti-TNF-naive and anti-TNF-exposed patients are shown in Supplementary Table 1.
Flowchart illustrating the selection of patients. UC, ulcerative colitis; VDZ, vedolizumab.
Baseline Characteristics of Patients with UC treated by VDZ
2. Long-Term Clinical Effectiveness in UC
Among the 239 patients who completed the 52-week follow-up, 59.4% (142/239) achieved steroid-free remission, and 68.2% (163/239) achieved clinical response. At 2 years, 52.9% (110/208) of patients achieved steroid-free remission and at 3 years, 40.3% (62/154) achieved steroid-free remission (Fig. 2). Stratified analyses based on baseline disease severity are provided in Supplementary Table 2.
Long-term steroid-free remission, endoscopic remission, and mucosal healing in vedolizumab-treated ulcerative colitis patients at week 52, 104, and 156.
3. Long-Term Endoscopic Effectiveness in UC
A total of 180 patients underwent endoscopy evaluation at both week 0 and week 52, with 53.3% (96/180) achieving endoscopic remission and 32.2% (58/180) achieving mucosal healing. At 2 years, endoscopic remission and mucosal healing were achieved in 39.8% (68/171) and 27.5% (47/171) patients, respectively. At 3 years, endoscopic remission and mucosal healing were achieved in 34.8% (49/141) and 26.2% (37/141) patients, respectively (Fig. 2). Results of endoscopic efficacy after multiple imputations are shown in Supplementary Table 3, for patients with available clinical data but missing endoscopic assessments.
4. Therapeutic Persistence
The probability of continuing VDZ therapy was 76.8% (95% CI, 71.6%–82.4%) at week 52, 63.3% (95% CI, 57.3%–70.0%) in 2 years, and 56.4% (95% CI, 49.7%–64.0%) in 3 years. A total of 94 patients (39.3%) discontinued VDZ therapy. Among them, 79 patients (84.0%) discontinued due to inadequate response, 14 patients (14.9%) for personal reasons, and 1 patient (1.1%) died (Fig. 3). A competing risks analysis distinguishing discontinuation due to loss of efficacy versus other causes is provided in Supplementary Fig. 1.
Three-year therapeutic persistence of vedolizumab (VDZ) in ulcerative colitis patients based on Kaplan-Meier estimates.
5. Dose Escalation
Among the 239 included patients, a total of 39 patients (16.3%) underwent dose escalation during follow-up. Among these 39 patients, 37 (94.9%) received interval reduction, while 2 (5.1%) underwent therapy reinduction. No independent variables were found to be associated with dose escalation (Supplementary Table 4).
6. Predictors of Long-Term Treatment Persistence and Week 52 Endoscopic Outcomes
In the univariate Cox regression analysis, factors including disease duration, anti-TNF exposure, prior steroid use, baseline Mayo score, concomitant immunomodulator therapy at baseline, C-reactive protein >10 mg/L, albumin <35 g/L, and achievement of steroid-free remission or mucosal healing after induction were associated with long-term treatment persistence. Dose escalation was treated as a time-varying covariate in a time-dependent Cox regression model, and it was not significantly associated with VDZ discontinuation (HR, 1.140; 95% CI, 0.600–2.150; P=0.688). In the multivariate Cox regression model, anti-TNF exposure was identified as an independent risk factor for failure to maintain long-term VDZ persistence (HR, 2.754; 95% CI, 1.478–5.135; P=0.001), while steroid-free remission after induction was an independent protective factor (HR, 0.480, 95% CI, 0.235–0.978; P= 0.043) (Table 2). Multicollinearity assessment for variables included in the multivariate model is shown in Supplementary Table 5. All adjusted generalized VIF values were below 2, indicating no significant multicollinearity.
Predictors of Failure to Achieve Long-Term Therapy Persistence
We also performed additional analyses to identify predictors of endoscopic remission and mucosal healing at week 52 (Supplementary Tables 6 and 7). Briefly, for endoscopic remission at week 52, early endoscopic remission following induction was an independent protective factor (OR, 8.267; 95% CI, 3.705–18.445; P<0.001). Similarly, for mucosal healing at week 52, prior steroid use was an independent risk factor (OR, 0.438; 95% CI, 0.212–0.906; P=0.026), while early mucosal healing following induction was an independent protective factor (OR, 2.167; 95% CI, 1.028–4.569; P=0.042).
7. Safety
During the follow-up period, 17 patients (7.1%) experienced AEs. Among these, 3 patients (1.3%) developed symptomatic enteritis associated with Clostridium difficile. Rashes were observed in 9 patients (3.8%), joint pain in 3 patients (1.3%), and fatigue in 2 patients (0.8%). All these symptoms were mild and no serious AEs were reported that necessitated therapy discontinuation.
DISCUSSION
This multicenter real-world cohort study provided valuable evidence on the long-term clinical and endoscopic efficacy of VDZ in active UC patients, an area where data remain limited. By systematically evaluating outcomes over a 3-year period, we found that VDZ effectively maintained both clinical and endoscopic remission and showed a favorable safety profile.
Since its approval, multiple real-world studies have confirmed the efficacy of VDZ in UC, with clinical remission rates ranging from 36%–66% at week 14 and 47%–64% at week 52 [6,13-20]. Meta-analyses reported a pooled mucosal healing rate of approximately 45% during maintenance therapy [21]. However, data on long-term efficacy beyond 2 years are limited. Gros et al. [22] reported clinical remission rates of 75.7%, 72.4%, and 70.2% at 1, 3, and 5 years, respectively. Another 2-year prospective study comparing VDZ with anti-TNF agents found a propensity score–adjusted remission rate of 43.2% for VDZ [23]. A recent meta-analysis estimated pooled VDZ persistence rates of 73.8% (95% CI, 70.0%–77.6%) at 1 year and 66.9% (95% CI, 56.7%–77.1%) at 2 years [24].
In our study, steroid-free remission was achieved in over 50% of patients at 2 years and in over 40% at 3 years, supporting the durable efficacy of VDZ in UC. Although remission and persistence rates gradually declined with longer follow-up, this trend likely reflects real-world treatment dynamics, including treatment discontinuation, loss of response, and reduced follow-up numbers over time. Nevertheless, a considerable proportion of patients maintained both clinical and endoscopic remission through 156 weeks, underscoring the durable and clinically meaningful benefits of VDZ in long-term management of UC. Differences in long-term remission rates compared with previously reported studies may be explained by variations in baseline disease characteristics and patient selection across cohorts.
Similarly, few studies have reported the long-term endoscopic efficacy of VDZ in UC. A prospective study of 34 patients (median follow-up 3.2 years) reported durable endoscopic remission in 50% at the end of follow-up [25]. A Swedish cohort reported endoscopic remission rates of 28% and 22% at 2 and 3 years, respectively [26]. In our study, endoscopic remission was achieved in 39.8% of patients at 2 years and 34.8% at 3 years, with corresponding mucosal healing rates of 27.5% and 26.2%. These rates, lower than clinical remission, underscore that endoscopic remission and mucosal healing are more stringent measures of disease control.
However, it should be noted that some patients had fewer endoscopic assessments due to sustained clinical remission. Given that endoscopic follow-up is often omitted in clinically stable patients, this may underestimate true endoscopic remission rates. To address this, we performed a sensitivity analysis using multiple imputations for missing data. The imputed results showed slightly higher remission rates, suggesting that the true long-term endoscopic efficacy of VDZ may be greater than observed. Prospective studies with standardized endoscopic follow-up and larger cohorts are needed to confirm these findings.
VDZ exhibited a high rate of therapeutic continuation, which may be attributed to its efficacy and safety profile. Our study indicated that more than half of the patients continued VDZ therapy at 3 years, consistent with previous studies [26-28]. VDZ demonstrated a good safety profile in UC patients, consistent with findings from 2 large safety studies [5,29]. Serious AEs were not found in our study, and the most common AEs were mild allergic reactions and joint pain, with few instances of infusion reactions. It was noted that 1.3% developed symptomatic enteritis associated with Clostridium difficile, but this association required further investigation. Owing to the retrospective nature of our study, the incidence of minor AEs that did not require medical intervention may have been underestimated. Minor or transient events might not have been consistently recorded in medical charts, leading to potential underreporting.
Our findings indicated that anti-TNF exposure was negatively associated with long-term therapy continuation, consistent with previous studies [6,22,30,31]. This suggests that exposure to a previous anti-TNF agent may alter the immunological landscape to reduce the effectiveness of VDZ. Furthermore, early achievement of steroid-free remission following induction was positively associated with long-term outcome, which suggests that proactive measures to achieve early and rapid control of inflammation may contribute to the maintenance of long-term therapy [32]. However, we could not find an association between dose escalation and therapy persistence, which is clinically plausible, as dose escalation generally reflects suboptimal response or secondary loss of response rather than a proactive optimization strategy.
Our study had several strengths. To the best of our knowledge, this is the largest real-world cohort study conducted in the Asian population to evaluate the long-term effectiveness of VDZ in the treatment of UC. We reported 3-year clinical and endoscopic efficacy of VDZ, addressing a gap in the existing literature where such long-term data have been limited. However, there were some limitations that should not be ignored. First, the retrospective observational nature of this study had inherent limitations. Second, the predominant inclusion of patients with moderate activity at baseline may limit the generalization of VDZ effectiveness to severe cases.
In summary, our multicenter, real-world study demonstrated that VDZ was effective in achieving long-term clinical and endoscopic remission in UC patients, accompanied by a high safety profile. The study also highlighted that VDZ was particularly superior in anti-TNF-naive patients who achieved early clinical remission. Nonetheless, further large-scale prospective studies are required to gather additional evidence in real-world settings to validate these findings.
Notes
Funding Source
This study was financially supported in part by National Key R&D Program of China (2023YFC2507300), Key-Area Research and Development Program of Guangdong Province (2023B1111040003), the program of Guangdong Provincial Clinical Research Center for Digestive Diseases (2020B11111 70004), and National Key Clinical Discipline.
Conflict of Interest
Gao X is an editorial board member of the journal but was not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflicts of interest relevant to this article were reported.
Data Availability Statement
Data analyzed in this study are available from the corresponding author upon reasonable request.
Author Contributions
Conceptualization: Huang Z, Yao L, Qiu Y, Chao K, Mao R. Data curation: Huang Z, Yao L, Qiu Y, Liu J. Formal analysis: Huang Z, Li Q. Investigation: Huang Z. Methodology: Huang Z. Resources: Yao L, Cao Q, Chen M, Gao X. Writing–original draft: Huang Z, Yao L, Qiu Y. Writing–review & editing: all authors. Approval of final manuscript: all authors.
Supplementary Material
Supplementary materials are available at the Intestinal Research website (https://www.irjournal.org).
Supplementary Table 1.
Comparison of Baseline Characteristics between Anti-TNF-Naive and Anti-TNF-Exposed UC Patients Treated with VDZ
Supplementary Table 2.
Clinical and Endoscopic Efficacy at Weeks 52, 104, and 156 Stratified by Baseline Disease Activity in Ulcerative Colitis Patients
Supplementary Table 3.
Long-Term Endoscopic Outcomes in VDZ-Treated UC Patients Using Multiple Imputation
Supplementary Table 4.
Predictors of Requirement for Dose Escalation during Follow-up
Supplementary Table 5.
Multicollinearity Analysis of Variables Included in Cox Regression Model
Supplementary Table 6.
Predictors of Achieving Endoscopic Remission at Week 52
Supplementary Table 7.
Predictors of Achieving Mucosal Healing at Week 52
Supplementary Fig. 1.
Competing risks analysis of vedolizumab (VDZ) discontinuation in ulcerative colitis patients.