Biologics for prevention of postoperative Crohn’s disease recurrence: A retrospective cohort study in China

To the Editor: Patients with Crohn’s disease (CD) often undergo bowel resection due to complications. Indeed, bowel resection is not a curative approach and has relatively high rates of postoperative complications and recurrence.^[1] The 10-year risk of having a second resection after the first is 35%, although more recent studies suggest that this may have decreased to closer to 30%.^[2] How to prevent postoperative recurrence (POR) has become a major concern.

Studies have shown that mesalazine and azathioprine do not effectively prevent POR and biologics have a better preventive effect. The relative efficacies of different biologics remain controversial, suggesting that further trials are needed.

This study aimed to evaluate the effectiveness of prophylactic therapy in preventing POR after intestinal surgery in patients with CD in a real-world setting and to compare vedolizumab, ustekinumab, and infliximab for effectiveness, which is currently a gap in China.

This study retrospectively analyzed patients with CD treated at Jinling Hospital, Medical School of Nanjing University, from March 2021 to August 2023. Inclusion criteria were: (1) Meeting the diagnostic criteria for CD in the 2023 Chinese national clinical practice guideline on diagnosis and management of Crohn’s disease;^[1] (2) Undergoing intestinal surgery, including partial resection of the intestine, colon, and rectum; intestinal perforation repair surgery; stoma, etc.; (3) Undergoing biologics treatment according to the recommended dosage and modality within 6 months after surgery for at least 3 months. Exclusion criteria were: (1) Treatment limited to simple anal fistula surgery, appendectomy, or endoscopic dilatation; (2) missed visits or lack of clinical data. Ethical approval (No. 2022DZKY-048-02) and informed consent were obtained.

The baseline clinical data were collected through the hospital’s electronic medical record system. The primary outcome was clinical recurrence at the end of follow-up. Disease severity was assessed clinically by the Crohn’s Disease Activity Index (CDAI), which defines clinical recurrence as a CDAI >150 and a CDAI increase of 100.^[1] When a case relapsed, the treatment plan was changed according to the specific situation, including reoperation, switching to another biologic, co-administering immunosuppressants or hormones, re-induction, etc. The secondary outcome was the status of endoscopy recurrence. Endoscopic POR was considered with a Rutgeerts score ≥i2, and endoscopic remission with a Rutgeerts score <i2.^[3]

SPSS Version 27 (IBM Inc., Armonk, NY, USA) was used for statistical analysis. Continuous variables were presented as median (Q₁, Q₃), and categorical variables as frequency and percentage. Continuous variables were compared by the Kruskal–Wallis test. Categorical variables were compared by the chi-squared test, Fisher’s exact test, or Kruskal–Wallis test. The Scheffe test was used for post hoc analysis. The same indicator at different time points was compared by the Wilcoxon signed rank test (continuous variable) or the McNemar test (categorical variable).

Kaplan–Meier curves were generated to assess differences in survival for all subgroups by the log-rank test. Multivariable analysis was performed using Cox regression. Any independent variable significantly associated with the dependent variables in univariate analysis (P <0.05) was entered into multivariate analysis.

A total of 345 patients were enrolled and 62 were finally included, with 17 administered vedolizumab after surgery, 28 administered ustekinumab, and 17 administered infliximab [Supplementary Figure 1, http://links.lww.com/CM9/C393].

Supplementary Table 1, http://links.lww.com/CM9/C393 shows the baseline characteristics of the three cohorts. Overall, of the 62 patients included, 34 (54.8%) were men, median age at the initial postoperative biologic use was 37.50 (27.75–52.25) years, and median disease duration was 5.00 (2.00–9.25) years. The indication for the most recent surgery in most patients included stenosis or obstruction (38/62, 61.3%), followed by perforation (17/62, 27.4%).

When comparing treatment groups, patients administered vedolizumab included more women (male percentages of 29.4% [5/17], 67.9% [19/28], and 58.8% [10/17] in the vedolizumab, ustekinumab, and infliximab, respectively [P = 0.039]) and were older at diagnosis (age at diagnosis: vedolizumab vs. ustekinumab, P = 0.013; vedolizumab vs. infliximab, P = 0.054; ustekinumab vs. infliximab, P = 0.953) compared with the ustekinumab and infliximab groups. Patients in the infliximab group had enhanced exposure to immunosuppressants compared with the vedolizumab and ustekinumab groups (immunosuppressant use percentages of 17.6% [3/17], 42.9% [12/28], and 58.8% [10/17] in the vedolizumab, ustekinumab, and infliximab groups, respectively, P = 0.047). In terms of sickness behavior, patients treated with infliximab were more likely to show enhanced penetration than those administered ustekinumab and vedolizumab (B3 [penetration]: vedolizumab, 41.2% [7/17]; ustekinumab, 21.4% [6/28]; infliximab, 58.8% [10/17], P = 0.047).

After a follow-up of 10.00 (5.00–15.00) months (no significant differences among the three groups, P = 0.862), clinical recurrence occurred in 18 patients (18/62, 29.0%), including 5 in the vedolizumab group (5/17, 29.4%), 4 in the ustekinumab group (4/28, 14.3%) and 9 in the infliximab group (9/17, 52.9%). Kaplan–Meier curve analysis was used to draw survival functions for the outcome of clinical recurrence [Figure 1]. Overall, there was no significant difference among the three groups (P = 0.100). However, comparing between subgroups, cases treated with ustekinumab were less likely to relapse than the infliximab group (P = 0.021), but not better than the vedolizumab group (P = 0.395). The rate of postoperative occurrence was lower in the vedolizumab group than the infliximab group (non-significant, P = 0.356).

Figure 1.

Comparative risk of clinical POR among patients with Crohn’s disease undergoing intestinal surgery and biologics treatment with vedolizumab vs. ustekinumab vs. infliximab. POR: Postoperative recurrence.

Patients underwent endoscopy at an average of 6.8 ± 3.0 months after the first postoperative treatment with biologics. In all, 8 of the 32 patients achieved endoscopic POR (Rutgeerts score ≥i2). There were no significant differences in follow-up time (P = 0.640) and Rutgeerts score (P = 0.130) among the different groups [Supplementary Figure 2, http://links.lww.com/CM9/C393].

Univariate- and multivariate-adjusted hazard ratios (HRs) for POR’s risk factors are shown in Supplementary Table 2, http://links.lww.com/CM9/C393. Multicollinearity was assessed using variance inflation factors (VIF), and no concerning collinearity was identified among the included variables. The relationship between the independent variables and the outcome was tested for equal proportional risk and the results confirmed that the assumption was satisfied for all included variables. In univariate Cox regression analysis, biologic choice, sickness behavior, experience of ileocecal resection, extensive lesion status, and previous immunosuppressive therapy were associated with a different risk of clinical recurrence. In multivariate analysis, significance was maintained for extensive lesion status (HR = 3.145, 95% confidence interval [CI] = 1.042–9.494; P = 0.042) and previous immunosuppressive therapy (HR = 4.005, 95% CI = 1.253–12.804; P = 0.019). Kaplan–Meier curve analysis was used to draw survival functions for the outcome of clinical recurrence for significant risk factors [Supplementary Figure 3, http://links.lww.com/CM9/C393].

In patients without clinical recurrence (patients not reaching follow-up time points or who had clinical recurrence before these time points were excluded), CDAI, C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR) were compared among different biologic groups and follow-up time points (baseline, 3 months, 6 months, 12 months). There were no significant differences in CDAI, CRP, and ESR among different treatment groups at baseline, indicating similar inflammation levels.

Supplementary Table 3, http://links.lww.com/CM9/C393 shows differences among follow-up time points. CRP at 3 months was significantly higher in the vedolizumab group than in the ustekinumab and infliximab groups (P <0.001). Other differences were not statistically significant.

This retrospective analysis demonstrated that vedolizumab and ustekinumab were superior to infliximab, but only the difference between the ustekinumab and infliximab groups was significant. Patients administered immunosuppressants or who had extensive lesions were more likely to have clinical POR. CDAI, ESR, and CRP decreased significantly after 3 months of biologic treatment.

Several studies comparing different biologics in preventing POR have reported controversial results. In another study, no differences in endoscopic POR risk were observed between anti-tumor necrosis factor (TNF) prophylaxis and the other groups: anti-TNF, 40.2%; vedolizumab, 33%; ustekinumab, 61.8%.^[4] The variation in results could be attributed to the fact that the majority of patients included in this study were either in remission or experiencing mild symptoms. In this study, after correcting for confounding factors in multifactorial COX regression analysis, there was no significant correlation between the choice of postoperative biologic and clinical recurrence outcomes, meaning ustekinumab may show trends toward effectiveness but further studies are needed to confirm this. Additionally, the follow-up period may not be sufficient and the sample size is limited.

Clinical risk factors are useful to stratify patients in terms of POR risk and guide preventive strategies. Factors considered to predict “difficult disease control” include combined lesions, extensive lesions (lesions involving the intestinal segment accumulated >100 cm), esophagogastroduodenal lesions, young age at onset, and the need for steroid therapy at the initial disease onset.^[1] In the present study, of the five examined risk factors, patients were more likely to have POR only with extensive lesions. In addition, cases requiring immunosuppressant therapy throughout the course of the disease were more likely to recur. The timing of immunosuppressant administration may be a relevant factor, and more severe or poorly controlled disease may require the use of immunosuppressants.

This study analyzed the actual postoperative use of biologics in patients with CD, providing data for the efficacy of different biologics in the prevention of POR in China. The current study provided certain guidance and might be used as a reference for further clinical promotion and application. Since the definitions of clinical POR and Rutgeerts score have never been formally unified, all current diagnostic modalities had limitations.^[5] Our findings get only a preliminary conclusion due to the lack of a control group treated with traditional medication. As this was an observational, non-interventional study, data on endoscopy were only available if it is deemed necessary by the patients’ physician. Therefore, some of the endoscopic POR data are missing and the results may be affected by inconsistent evaluation time.

In conclusion, this study demonstrated that biologics (vedolizumab, ustekinumab, and infliximab) could effectively prevent POR. Ustekinumab might show a better preventive efficacy than infliximab. Patients administered immunosuppressants or with extensive lesions are more likely to undergo clinical POR.

Supplementary Material

cm9-139-1095-s001.docx ^{(627.1KB, docx)}