Transabdominal intestinal ultrasound parameters as intermediate treatment targets in a treat-to-target strategy for moderate-to-severe ulcerative colitis
Article information
Abstract
Background/Aims
This study aimed to investigate the prognostic value of early post-induction intestinal ultrasound (IUS) f indings in predicting long-term clinical outcomes among patients with moderate-to-severe ulcerative colitis (UC).
Methods
This retrospective, single-center study consecutively enrolled patients with moderate-to-severe, left-sided or extensive UC. Clinical endpoints were assessed at the end of follow-up or 1 year after induction therapy (for patients with at least 1 year of follow-up) and categorized as clinical remission (Short Clinical Colitis Activity Index [SCCAI] ≤2) or non-remission (SCCAI >2). Patients who experienced a negative disease course before the evaluation point were classified as clinical non-remission.
Results
A total of 56 patients were included. The bowel wall thickness (BWT; 5.1±1.7 mm vs. 6.0±1.4 mm; P=0.032) and Milan ultrasound criteria (MUC; 8.3±3.2 vs. 9.9±2.2; P=0.029) evaluated at early follow-up IUS were lower for patients reaching long-term clinical remission. Patients with BWT <5 mm on early follow-up IUS (83% vs. 45%; P=0.006) or MUC <6.2 (100% vs. 45%; P<0.001) had a significantly higher rate of long-term clinical remission. Kaplan-Meier analysis revealed a lower cumulative probability of a negative disease course in patients with BWT <5 mm (P=0.025) or MUC <6.2 (P=0.025). Cox regression analysis identified BWT ≥5 mm as an independent predictor of a negative disease course.
Conclusions
BWT <5 mm and MUC <6.2 may serve as intermediate targets for IUS-guided “treat-to-target” strategy, offering a practical approach to improve long-term clinical remission in patients with moderate-to-severe UC.
INTRODUCTION
Ulcerative colitis (UC) is a chronic, relapsing form of inflammatory bowel disease characterized by continuous mucosal inflammation of the colon and rectum. Patients with UC often experience impaired quality of life, frequent relapses, and substantial healthcare utilization [1]. The “treat-to-target” (T2T) strategy, which emphasizes objective disease assessment to achieve sustained remission, has been widely recommended in UC management [2]. Within this framework, a key challenge is the timely identification of patients who are unlikely to achieve long-term remission, thereby enabling earlier optimization of treatment strategies.
Transabdominal intestinal ultrasound (IUS) has emerged as a compelling, noninvasive assessment tool for disease monitoring in inflammatory bowel disease. While its utility in assessing transmural inflammation, monitoring therapeutic response, and guiding T2T strategies in Crohn’s disease is well-established, the evidence base supporting its application in UC remains comparatively limited [2]. Nevertheless, growing evidence suggests that IUS parameters—such as bowel wall thickness (BWT), vascularity, and composite indices like the Milan ultrasound criteria (MUC)—can effectively reflect endoscopic activity and treatment response in UC [3,4].
Evidence has demonstrated that early IUS findings are predictive of short-term outcomes in UC. For instance, a reduction in BWT at 6 weeks following treatment induction has been shown to predict endoscopic remission between 8 to 26 weeks post-induction in moderate-to-severe UC [5]. Additionally, a BWT reduction at 48 hours after administration of intravenous corticosteroid is associated with treatment response in severe cases [6]. Conversely, persistently elevated BWT and MUC scores at early assessment portend a higher risk of an unfavorable disease course in severe UC [7,8]. However, the prognostic value of early IUS for long-term outcomes in patients with moderate-to-severe UC remains unclear. Therefore, further investigation is warranted to define its role in guiding long-term treatment optimization and advancing ultrasound-guided T2T strategies for moderate-to-severe UC.
To address this gap, we conducted a retrospective study to determine whether early post-induction IUS parameters can predict sustained clinical remission, endoscopic remission, and negative disease outcomes during extended follow-up in patients with moderate-to-severe UC.
METHODS
1. Study Design and IRB Approval
This single-center retrospective cohort study was conducted at Peking Union Medical College Hospital (PUMCH). The Institutional Review Board of PUMCH (IRB No. I-24PJ2496) approved the study protocol. The requirement for informed consent was waived by the IRB. In this study, week 0 was defined as the initiation of induction therapy and served as the anchor point for all subsequent time points. Baseline assessments were performed after the onset of disease activity and prior to treatment initiation (no more than 4 weeks before initiation of induction therapy). Early follow-up IUS was conducted between 8 and 18 weeks after induction therapy initiation. This time window reflects the heterogeneous response kinetics of different induction regimens in real-world clinical practice, in which different regimens exhibit a distinct course for the completion of induction therapy and achieving meaningful improvement detectable by IUS. Notably, the median timing of early IUS assessment was 12 weeks, a commonly used time point for early clinical evaluation. Long-term endpoints were assessed at ≥ 6 months (Fig. 1).
The overview of this study design. IUS, intestinal ultrasound.
2. Patients
We retrospectively enrolled consecutive patients with moderate-to-severe UC treated at PUMCH between January 2018 and December 2024. Inclusion criteria were as follows: (1) age ≥ 14 years; (2) established UC diagnosis according to the Chinese Clinical Practice Guideline for Ulcerative Colitis Management (2023, Xi’an) [9]; (3) disease extent classified as left-sided colitis (E2) or extensive colitis (E3) according to Montreal classification [10]; (4) moderate-to-severe disease activity, defined as a Mayo endoscopic subscore (MES) ≥ 2; (5) completion of baseline IUS and endoscopic evaluation within 4 weeks prior to induction therapy (week 0), regardless of the induction regimens; (6) early follow-up IUS performed between 8 and 18 weeks after initiation of the induction regimens; and (7) minimum post-induction follow-up of 6 months. Patients were excluded for: (1) history of bowel resection; (2) presence of gastrointestinal malignancy; and (3) induction treatment for UC compromised by baseline concurrent infections.
3. Data Collection
The following data were extracted from the PUMCH electronic medical record system: (1) demographic characteristics: age at induction therapy, age at diagnosis, gender, and body mass index (BMI); (2) UC disease characteristics: disease duration, disease extent according to Montreal classification, MES, prior treatments, extraintestinal manifestations, and concurrent intestinal infections at baseline (e.g., Clostridioides difficile infection confirmed by positive fecal enzyme immunoassay and nucleic acid amplification testing, cytomegalovirus colitis confirmed by immunohistopathology); (3) Laboratory parameters: C-reactive protein, erythrocyte sedimentation rate, hemoglobin, white blood cell count, platelet count, and albumin; and (4) medications: induction and maintenance medical treatment agents.
4. Baseline and Early Follow-up Transabdominal IUS Examination
The baseline IUS examination was conducted within 4 weeks prior to the initiation of induction therapy (week 0), and early follow-up IUS was performed between 8 and 18 weeks post-induction (Fig. 1). These procedures were executed by a team of radiologists, each with more than 3 years of experience in performing transabdominal IUS. Scans were conducted using Aixplorer (Supersonic Imagine, Aix-en-Provence, France) and IU22 (Philips, Eindhoven, Netherlands) with Curve and Linear probes (2–5 MHz, 3–12 MHz). All radiologists had undergone standardized training and followed the recommendations of the European Federation of Societies for Ultrasound in Medicine and Biology [11]. Machine settings were optimized for bowel assessment, and the velocity for color Doppler imaging was set at 5 cm/sec.
Two radiologists (Q.Z. and Z.W.) blinded to the clinical and endoscopic outcomes independently reviewed all images/videos. For each patient, the following parameters of the most severely affected bowel segment (defined as the segment with the greatest BWT) were identified: (1) BWT (mm): measured at the thickest point of the segment; (2) bowel wall stratification: categorized as clear, blurred, or loss; (3) Limberg score: assigned on a scale of 0–4 based on the degree of bowel wall thickening and vascularity [12]; (4) MUC: MUC=1.4 × BWT (mm)+2 × bowel wall flow (BWF), where BWF is coded as “1” if vascularity is present and “0” if absent [13]. Discrepancies were resolved by a third expert (W.L.), with consensus if needed.
5. Endpoint Evaluation
The primary endpoint was long-term clinical remission assessed at the end of follow-up (for patients with 6–12 months of follow-up) or at 1 year post-induction (for patients with follow-up exceeding 1 year), defined according to the Short Clinical Colitis Activity Index (SCCAI) [14]. The SCCAI was determined by 2 gastroenterologists (L.Y. and Y.Z.) blinded to the IUS results based on clinical records. Clinical remission was defined as SCCAI ≤ 2 [14], and clinical non-remission was defined as SCCAI > 2 or the occurrence of a negative disease course before the assessment time point. Negative disease course was defined as the occurrence of at least one of the following outcomes during the entire follow-up period: treatment escalation, corticosteroid use, UC-related hospitalization, or colectomy. Treatment escalation was defined as either optimization of the current therapy (re-induction, interval shortening, or dose intensification due to relapse) or switching to another drug with a different mechanism of action. The secondary endpoint was endoscopic remission assessed according to MES by colonoscopy performed between 22 and 54 weeks post-induction. The MES was independently assessed by 2 gastroenterologists (L.Y. and Y.Z.) blinded to the IUS results, based on endoscopic images. Endoscopic remission was defined as MES=0 in all segments. Discrepancies for SCCAI and MES were adjudicated by a third gastroenterologist (Y.L.). If disagreement persisted, the final interpretation was reached by consensus.
6. Predictive Variables
The primary predictive variables of interest were IUS parameters measured at early follow-up, including BWT and the MUC, assessed in the most severely affected bowel segment. These variables were prespecified as predictors and were evaluated for their association with predefined clinical and endoscopic endpoints.
7. Statistics
Statistical analysis was performed using SPSS software (version 27.0, IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean±standard deviation or median with interquartile range (IQR), depending on data distribution, and were compared using the Student t-test or Mann-Whitney U test, as appropriate. Categorical variables are presented as frequencies and percentages and were compared using the chi-square test or Fisher exact test. Time-to-event data were analyzed using Kaplan-Meier survival curves, and differences between groups were assessed using the log-rank test. All statistical tests were two-sided, and P<0.05 was considered statistically significant. For logistic regression analysis, variables identified as significant in univariate analysis were considered for multivariate regression. To avoid multicollinearity, only one variable among highly correlated variables (such as BWT and MUC) was included in one model. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cutoff values of early follow-up IUS parameters (including BWT and MUC) for predicting long-term clinical remission. Optimal cutoff points were identified using the Youden index (sensitivity + specificity−1).
Interobserver agreement for the IUS parameters (BWT, stratification, vascularity, Limberg grade), SCCAI, and MES was evaluated between 2 independent radiologists. Weighted Cohen’s kappa (κ) coefficients were calculated for categorical variables, and intraclass correlation coefficients (ICC) using a two-way random-effects model were calculated for continuous variables. Agreement was interpreted as poor (κ or ICC < 0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), or excellent (> 0.80).
RESULTS
1. Study Population
A total of 56 consecutive patients who met the inclusion criteria between January 2018 and December 2024 were enrolled in the study. The median disease duration was 3 years (IQR, 1–9 years). The median time of the early follow-up IUS examination was 12 weeks after the initiation of induction therapy (week 0) (IQR, 9–14 weeks). The median follow-up time was 14 months (IQR, 10–21 months) (Table 1). Twenty-four patients (43%) were male. The median age at induction therapy was 44.5 years (IQR, 32–54 years). According to the Montreal classification, 18 patients (32%) were classified as left-sided colitis (E2), and 38 (68%) as extensive colitis (E3). A total of 37 patients (66%) had a history of prior corticosteroid treatment, and 26 (47%) had previously received biologic therapy. Concurrent intestinal infections were reported in 6 (11%) of the cohort at baseline, while extraintestinal manifestations were present in 3 patients (5%) (Table 1). At baseline clinical assessment, all patients had an SCCAI ≥ 5 (Table 1). Thirteen patients (23%) exhibited moderate activity (MES=2), while 43 (77%) with severe activity (MES=3). At the end of follow-up or 1 year after induction therapy (for those with follow-up ≥ 1 year), a total of 32 patients (57%) had clinical remission according to SCCAI. The median time for assessing clinical remission was 12 months post-induction (IQR, 11–12 months). Thirty-seven patients (66%) underwent endoscopic evaluation between 22 and 54 weeks (median 36 weeks from baseline: IQR, 23–44 weeks), among whom 8 patients (22%) reached complete endoscopic remission.
Baseline Demographic and Clinical Characteristics of Enrolled Patients with Moderate-to-Severe Ulcerative Colitis
2. Baseline IUS Parameters and Correlation with Endoscopy
With regard to IUS, at baseline, the mean BWT was 7.3±2.2 mm. The corresponding mean MUC was 11.9±3.4 (Table 1). A Limberg score ≥ 2 was observed in 47 patients (84%). Among all patients, 15 (27%) exhibited clear wall stratification, 19 (34%) showed blurred stratification, and 22 (39%) had stratification loss. In the analysis for baseline IUS examination, BWT and MUC in the most severely affected segment showed significant correlation with MES (BWT: ρ= 0.411, P=0.002; MUC: ρ=0.419, P=0.001). Also, a significant change of BWT (7.3±2.2 mm vs. 5.5±1.6 mm; P<0.001), MUC (11.9±3.4 vs. 9.0±2.9, P<0.001), Limberg score (P<0.001), and bowel wall stratification (P<0.001) from baseline IUS examination to the early follow-up was observed (Fig. 2, Supplementary Table 1).
Representative intestinal ultrasound images of the colon in active inflammation (A, B) and remission (C). (A) Transverse colon during active inflammation, showing increased blood flow signals within the bowel wall (white arrows). (B) Transverse colon during active inflammation, demonstrating bowel wall thickening (red arrow). (C) Ascending colon in remission, with normal bowel wall appearance (white arrow).
3. Early IUS Predictors for Long-Term Clinical Remission
At baseline clinical assessment, there were no significant differences in demographics, disease activity, or IUS parameters between patients who achieved and those who did not achieve long-term clinical remission (Table 2). At early follow-up IUS assessment, both BWT (5.1±1.7 mm vs. 6.0±1.4 mm; P=0.032) and MUC (8.3±3.2 vs. 9.9±2.2; P=0.029) were significantly lower in patients who achieved long-term clinical remission compared to those who did not (Table 2, Fig. 3). For the other IUS parameters, neither Limberg score nor the wall layer stratification showed statistical difference between patients with clinical remission compared to those without (Table 1). In multivariate analysis, BWT per mm increase (model 1: odds ratio [OR], 0.67; 95% confidence interval [CI], 0.46–0.99; P=0.043) and MUC each unit increase (model 2: OR, 0.79; 95% CI, 0.63–0.98; P=0.035) of the most affected segment at early follow-up IUS predicted long-term clinical non-remission (Table 2). Notably, there were also statistically significant differences in sex (model 1: OR, 3.94; 95% CI, 1.12–13.93; P=0.033; model 2: OR, 4.11; 95% CI, 1.14–14.79; P=0.030) and disease duration (model 2: OR, 0.89; 95% CI, 0.80–1.00; P=0.047) (Table 2).
Univariate and Multivariate Logistic Regression Analyses to Predict Clinical Remission at Median 12 Months Post-Induction Therapy
Difference for BWT (A) or MUC (B) at early follow-up IUS between patients with and without clinical remission. **P<0.01, ***P<0.001. BWT, bowel wall thickness; MUC, Milan ultrasound criteria; NS, not significant.
4. Optimal BWT and MUC Cutoff at Early Follow-up to Predict Long-Term Clinical Remission
A BWT cutoff of 5 mm at early follow-up predicted clinical remission with 88% specificity (95% CI, 69%–96%) and 47% sensitivity (95% CI, 31%–64%) (area under the ROC curve [AUROC], 0.673; 95% CI, 0.533–0.813; P=0.015) (Supplementary Fig. 1). In the entire study cohort, 38 (68%) patients showed a BWT ≥ 5 mm at early follow-up IUS and 17 (45%) of them reached clinical remission, while 15 (83%) of the 18 patients showing BWT < 5 mm achieved clinical remission (P=0.006) (Fig. 4A, Supplementary Table 2). Correspondingly, a MUC of 6.2 at early follow-up predicted clinical remission with 100% specificity (95% CI, 86%–100%) and 38% sensitivity (95% CI, 23%–55%) (AUROC, 0.646; 95% CI, 0.502–0.790; P=0.047) (Supplementary Fig. 1). A total of 44 patients (79%) showed a MUC ≥ 6.2 at early follow-up IUS and 20 of them (45%) reached clinical remission, while all of the patients (n=12, 100%) showing MUC < 6.2 achieved clinical remission (P<0.001) (Fig. 4B).
(A) The proportion of patients reaching clinical remission among those with early follow-up BWT ≥5 mm and BWT <5 mm. (B) The proportion of patients reaching clinical remission among those with early follow-up MUC ≥6.2 and MUC <6.2. BWT, bowel wall thickness; MUC, Milan ultrasound criteria.
5. Association between Early Follow-up BWT or MUC and a Negative Disease Course
A total of 21 patients (38%) experienced negative disease courses during the follow-up period (median, 14 months; IQR, 10–21 months). Kaplan-Meier analysis revealed that patients with BWT ≥ 5 mm at early follow-up IUS had a significantly higher risk of a negative disease course compared to those with BWT < 5 mm (log-rank, P=0.025) (Fig. 5A). During the follow-up period, 47% of patients with BWT ≥ 5 mm experienced a negative disease course with a mean event-free survival time of 21 months (95% CI, 15–27 months), compared to 17% of patients in the BWT < 5 mm group, whose mean event-free survival was 34 months (95% CI, 28–41 months) (Fig. 5A). Cox regression revealed BWT ≥ 5 mm at early follow-up IUS independently predicting a negative disease course (HR, 7.07; 95% CI, 1.48–33.82; P=0.014) (Table 3, Supplementary Table 3). Correspondingly, patients with MUC ≥ 6.2 at early follow-up IUS had a significantly higher risk of negative disease course compared to those with MUC < 6.2 (log-rank, P=0.025) (Fig. 5B). During follow-up, 46% of patients with MUC ≥ 6.2 experienced negative disease course, with a mean event-free survival of 22 months (95% CI, 16–28 months), compared to only 8% of those with MUC < 6.2, whose mean event-free survival was 37 months (95% CI, 32–44 months) (Fig. 5B).
Kaplan-Meier curves for the cumulative probability of negative disease courses in patients with early follow-up (A) BWT ≥5 mm or BWT <5 mm and (B) MUC ≥6.2 or MUC <6.2. BWT, bowel wall thickness; MUC, Milan ultrasound criteria.
Univariate and Multivariate Cox Regression Analyses for Predicting a Negative Disease Course
6. Early Predictors of IUS for Long-Term Endoscopic Remission
There was no significant difference in BWT at the early follow-up IUS between patients who achieved endoscopic remission and those who did not (5.5±2.5 mm vs. 5.5±1.2 mm; P=0.955). MUC demonstrated comparable findings, with no significant difference between endoscopic remission and non-remission (9.2±4.2 vs. 8.9±2.4; P=0.869) (Supplementary Table 4).
7. Reliability Analysis
For BWT, an almost perfect agreement was achieved, with an ICC of 0.953 (95% CI, 0.931–0.966). Both the Limberg score and bowel wall stratification demonstrated good agreement, with κ values of 0.727 (95% CI, 0.645–0.810) and 0.617 (95% CI, 0.505–0.730), respectively. Excellent inter-rater reliability was also demonstrated for both the SCCAI and the MES, with ICC of 0.944 (95% CI, 0.928–0.951) and 0.960 (95% CI, 0.950–0.969), respectively.
DISCUSSION
The T2T strategy is now a cornerstone for UC management. However, despite its adoption, heterogeneity in response to induction therapy underscores the critical need for early predictors of long-term outcomes [15]. In this real-world study, we demonstrated that an early IUS assessment (median 12 weeks post-induction) provided modest but meaningful prognostic information for long-term clinical remission (median 12 months) in patients with moderate-to-severe UC. Specifically, both a BWT < 5 mm and an MUC score < 6.2 at early follow-up were independently associated with a higher rate of clinical remission and a lower risk of a negative disease course. These findings indicate that these IUS parameters may provide useful guidance as objective, intermediate targets within a T2T approach.
BWT is consistently recognized as the most reliable IUS parameter for assessing UC activity and therapeutic response. In severe UC patients, a > 20% reduction in BWT at 48 hours post intravenous corticosteroid initiation can accurately identify treatment responders [6], while a BWT ≥ 4 mm at this early timepoint predicts an increased 1-year colectomy risk [7]. Most recently, a multicenter cohort study showed in newly diagnosed UC, a transmural remission (BWT ≤ 3.0 mm without color Doppler signal) at 3 months after diagnosis increased the likelihood of steroid-free clinical remission at 12 months [16]. Focusing on moderate-to-severe UC, our study extends prior findings in several important aspects. Rather than restricting analysis to a single induction regimen, we evaluated a heterogeneous, real-world population receiving diverse therapies and assessed early follow-up IUS at approximately 12 weeks after induction therapy, a time point that reflects routine clinical reassessment. Within this context, we demonstrated that BWT alone independently predicts long-term clinical remission without incorporation of additional ultrasound parameters.
The MUC score, which integrates 2 of the most reliable and reproducible IUS parameters (BWT and BWF), mitigates operator-dependence [17,18]. Previous multicenter studies have shown that the MUC score outperforms the MES in predicting colectomy [19], and that a cutoff value of 6.2 effectively discriminates endoscopic activity [20]. Furthermore, an MUC ≤ 6.2 at baseline IUS examination or 12 weeks post-biologic initiation has been associated with a more favorable disease course and subsequent endoscopic remission (at 18–26 weeks), respectively [8,21]. Complementing this evidence, our study demonstrates that an early follow-up BWT < 5 mm and MUC < 6.2 independently predict higher clinical remission rates at about 1 year and a lower risk of negative disease course over 1.6 years in moderate-to-severe UC patients receiving diverse induction therapies. Collectively, these findings may position early BWT and MUC as practical, clinically meaningful decision-support markers and early prognostic indicators (rather than a treatment target equivalent to IUS transmural healing criteria) for an IUS-guided T2T strategy in UC.
In contrast to Crohn’s disease, clinical symptoms in UC generally correlate well with endoscopic activity [22-24]. However, in the present study, although early IUS parameters were significantly associated with long-term clinical remission, no statistically significant association was observed with endoscopic remission. While this finding appears inconsistent with previous reports, several factors may plausibly explain this discrepancy. First, the statistical power for endoscopic analyses was markedly limited. Only 37 patients underwent follow-up endoscopy, and merely 8 achieved endoscopic remission, substantially reducing the ability to detect meaningful associations. Second, the study population represented a difficult-to-treat cohort with uniformly moderate-to-severe disease and a high prevalence of prior exposure to biologics (34%) and corticosteroids (66%), resulting in a low overall rate of mucosal healing. In such populations, endoscopic remission may be delayed or incomplete despite symptomatic improvement, weakening correlations between early imaging markers and endoscopic outcomes. Third, the retrospective, real-world design may have introduced selection bias, as patients with persistent or worsening symptoms were more likely to undergo follow-up colonoscopy. Consequently, patients who achieved favorable endoscopic outcomes may have been underrepresented, further attenuating observable associations. Finally, the long interval between early IUS assessment and endoscopic evaluation may also have contributed to the lack of association, as interim treatment modifications and disease fluctuations could dilute the predictive signal of early ultrasound findings. Taken together, these considerations suggest that the absence of a statistically significant association with endoscopic remission in this study likely reflects methodological and population-related constraints rather than a true lack of biological relevance of early IUS parameters.
Also, the modest AUROC values observed in this study should be interpreted in light of its design and objectives. Several factors may have attenuated discriminatory performance: first, the use of a heterogeneous real-world cohort with diverse induction regimens, in contrast to highly controlled trial populations; second, evaluation of long-term outcomes at approximately 1 year rather than shorP=rm endpoints; third, the relatively small sample size. While these factors may limit discrimination, they also enhance the external applicability of our findings to routine clinical practice.
The strengths of this study highlight its clinical relevance and how it builds on prior work. First, conducted in a real-world setting with diverse induction therapies, our findings extend the applicability of early IUS assessment beyond single-agent or protocolized cohorts, supporting generalizability across practice settings. Second, the 12-week post-induction IUS assessment, aligned with routine practice [22], was used to predict long-term outcomes at 1 year, complementing prior studies focused on earlier or mid-term endpoints (26 weeks) and addressing an unmet need in T2T management. Third, blinded evaluation of validated parameters provides real-world support for the prognostic value of BWT and MUC, particularly for unfavorable disease course. Together, by demonstrating the predictive power of IUS under these conditions, our findings support a potential role for IUS as an intermediate assessment within T2T strategies, with possible implications for individualized treatment planning.
This study has several limitations inherent to its retrospective design. First, the modest sample size limited statistical power and constrained the complexity of multivariable models, precluding comprehensive adjustment for clinically relevant factors such as specific induction and maintenance therapies or prior treatment history. In particular, although maintenance therapy was significantly associated with prognosis, treatment selection in real-world practice is largely driven by post-induction disease control and early treatment response. As a consequence, maintenance therapy may reflect a surrogate of early disease status rather than an independent determinant of long-term outcomes, and these interrelated effects could not be fully disentangled in the current analysis. Similarly, given that corticosteroid-based induction accounted for a substantial proportion of the cohort, subgroup analyses stratified by induction therapy were performed (Supplementary Tables 5 and 6). In the corticosteroid-based induction subgroup, the associations between early IUS parameters and 1-year clinical remission were attenuated, likely reflecting the strong shorP=rm effects of systemic corticosteroids in combination with limited sample size. In contrast, in the non-corticosteroid induction subgroup, the association between BWT ≥ 5 mm and negative disease course did not reach statistical significance, primarily due to the small number of events and wide confidence intervals. Second, selection bias is possible, especially for endoscopic outcomes, as patients achieving clinical stability or remission were less likely to undergo repeat endoscopy during follow-up. Consequently, few endoscopic remission events were observed, limiting statistical power for these analyses. Finally, while endoscopy is the gold standard for assessing mucosal inflammation, our study used long-term clinical outcomes as the primary endpoint, which may not fully correlate with endoscopic remission.
In conclusion, this study provides real-world evidence for the potential prognostic utility of early post-induction IUS in a treatment-heterogeneous cohort of patients with moderate-to-severe UC. We found that a BWT < 5 mm or an MUC < 6.2 were associated with higher rate of sustained clinical remission and decreased risk of a negative disease course. These results support BWT and MUC as potential intermediate targets for a T2T strategy, thereby suggesting that IUS may help guide therapeutic decisions and improve long-term disease control in UC.
Notes
Funding Source
This work was supported by Natural Science Foundation of Beijing (grant number: 7252104), CAMS Innovation Fund for Medical Sciences (CIFMS) from Chinese Academy of Medical Sciences (grant number: 2021-I2M-1-062) and National High Level Hospital Clinical Research Funding (grant number: 2022-PUMCH-B-022).
Conflict of Interest
Li Y has received research grants from Natural Science Foundation of Beijing, CAMS Innovation Fund for Medical Sciences and National High Level Hospital Clinical Research Funding. The other authors have no conflicts of interest to declare.
Data Availability Statement
Data analyzed in this study are available from the corresponding author upon reasonable request.
Author Contributions
Conceptualization: Yun L, Wang Z, Zhu Q, Tan B, Li W, Li Y. Data curation: Yun L, Wang Z, Zhu Q, Tan B, Li W, Li Y. Formal analysis: Yun L, Wang Z, Zhu Q, Zhu Y. Funding acquisition: Li Y. Methodology: Zhu Q, Zhu Y, Xu H, Li W, Li Y. Resources: Zhu Y, Tan B, Li Y. Software: Wang Z, Tan B. Supervision: Xu H. Visualization: Yun L, Wang Z. Writing–original draft: Yun L, Wang Z, Li Y. Writing–review & editing: Yun L, Wang Z, Zhu Q, Zhu Y, Xu H, Tan B, Li W, Li Y. Approval of final manuscript: all authors.
Supplementary Material
Supplementary materials are available at the Intestinal Research website (https://www.irjournal.org).
Supplementary Table 1.
Differences between Baseline and Early Follow-up IUS Parameters
Supplementary Table 2.
Sensitivity Analysis to Predict 1-Year Clinical remission Using Alternative BWT Cutoff Values
Supplementary Table 3.
Sensitivity Analysis to Predict Negative Disease Course Using Alternative BWT Cutoff Values
Supplementary Table 4.
Univariate Analysis to Predict Endoscopic Remission
Supplementary Table 5.
Subgroup Univariable Logistic Analysis of 1-Year Clinical Remission Stratified by Induction Therapy
Supplementary Table 6.
Subgroup Cox Univariable Analysis Stratified by Induction Therapy
Supplementary Fig. 1.
Receiver operating characteristic (ROC) curve of early follow-up BWT or MUC to predict clinical remission (n=56). BWT, bowel wall thickness; MUC, Milan ultrasound criteria; AUROC, area under the ROC curve.