Terminal Ileum Thickness During Maintenance Therapy Is a Predictive Marker of the Outcome of Infliximab Therapy in Crohn Disease

Abstract

Background

Mucosal healing has been associated with long-term response to therapy for Crohn disease (CD). However, little is known about the significance of terminal ileum (TI) transmural thickness in predicting clinical outcomes.

Methods

In this retrospective observational cohort study, we examined the association of an index ultrasonographic assessment of TI thickness during the maintenance phase and the subsequent clinical outcome of CD in a cohort of patients treated with infliximab (IFX). Treatment failure was defined as treatment discontinuation because of lack of efficacy, a need for dose escalation, or surgery. Clinical response was defined as treatment continuation in the absence of any of the aforementioned failure criteria.

Results

Sixty patients with CD receiving IFX therapy were included in the study. The patients were followed for a median of 16 months (5-24 months) after an index intestinal ultrasound. Thirty-eight patients (63.3%) maintained response to the therapy and 22 patients (36.6%) failed the treatment, with a mean follow up of 10.5 months (6.5-17 months) vs 9.25 months (1-10.25 months), respectively. On univariate analysis, the only variables differing between treatment response and failure were a TI thickness of 2.8 vs 5 mm (P < 0.0001) and an IFX trough level of 6.6 vs 3.9 µg/mL (P = 0.008).

On multivariable analysis, only a small bowel thickness of ≥4 mm was associated with the risk of treatment failure (odds ratio, 2.9; 95% CI, 1.49-5.55; P = 0.002).

Conclusions

Our findings suggest that transmural thickness of ≥4 mm can predict subsequent treatment failure in patients with CD treated using IFX, indicating transmural thickness <4 mm as a potential novel valuable therapeutic target.

Introduction

Higher infliximab trough levels have been correlated with favorable outcomes in CD, and lower levels have been correlated with disease exacerbation and loss of response to infliximab (IFX) treatment.^1-3

Intestinal ultrasonography (IUS) is a noninvasive imaging procedure that has recently been widely adopted for bedside assessment of Crohn disease (CD). It is performed regularly in many European countries (especially Italy, Germany, Spain, and the Scandinavia region) and in Australia, Canada, and Israel. In addition to sonographic assessment of bowel wall layers, IUS can facilitate the visualization of possible intestinal and extraintestinal pathologies (such as lumen stenosis, abscesses, and fistulas).^4-8 As such, IUS is one of the diagnostic modalities recommended by the European Crohn’s and Colitis Organization for follow-up of treatment in CD.⁹ The lack of radiation and ready availability make ultrasound an attractive technique, particularly in a disease that affects young patients who may require serial studies. However, the technique is operator-dependent,¹⁰ and considerable experience is required to obtain maximal accuracy.

Transmural healing (a decrease in thickness of the bowel wall [BWT] to ≤3 mm) is an important treatment goal that is associated with improvement of all clinical outcomes and better long-term results even after discontinuation of biological therapy.^{11, 12} Magnetic resonance enterography¹¹ and IUS are employed mainly to compare transmural thickness with mucosal healing; ¹² however, the association of IUS parameters during maintenance therapy with clinical outcomes is very restricted.

Accordingly, we decided to investigate the association between the terminal ileum (TI) thickness measurement by IUS during maintenance therapy and the likelihood of therapeutic failure.

METHODS

Patients

This is a retrospective cohort study of patients with CD who received regular IFX treatment at Sheba Medical Center between January 2016 and March 2019. The IFX trough levels and anti-IFX antibody levels were routinely assessed.

We included adult patients with CD who were treated with scheduled maintenance IFX therapy and had had at least 1 IUS examination after week 14 of therapy. Patients were excluded if the terminal ileum was not involved, if they had small bowel or ileocecal resection and if they did not have IFX trough levels (TL) obtained within 3 months of performance of the IUS.

The study was approved by the Sheba Medical Center ethics committee. The patients’ consent was waived.

Clinical Scores and Outcomes

Clinical status and disease activity were assessed using the Harvey-Bradshaw Index, which defines clinical remission as a score of ≤4 points and response as a 3-point decrease.^{13, 14}

The primary outcome of this study was the persistence of maintenance IFX treatment vs treatment failure, as assessed by the need for CD-related surgery, discontinuation because of lack of efficacy, or IFX dose escalation.

IFX TL Measurement

Serum specimens were accumulated at a trough, before IFX injections. IFX levels in the serum were measured by ELISA as described in earlier research.¹⁵

IUS

The IUS examinations were performed using a BK 3000 ultrasound machine. All examinations were completed by a single skilled physician (D. Carter, >5 years of experience and >1500 IUS examinations performed). All scans were performed using a consistent technique. An initial general abdominal examination was performed using the low-frequency curved array transducer, focused on the detection of inflammation and complications (such as bowel loop dilatation or abscesses). A detailed examination of the bowel wall structure using a high-resolution linear array was then conducted. The overall thickness of the TI was measured in the long axis of the anterior wall. At least 2 measurements were undertaken for each bowel segment, at a minimum distance of 1 cm apart (Fig. 1).

Statistical Analysis

Continuous variables were articulated as the median and interquartile range. The Mann-Whitney U test was used to correlate continuous variables, and the Fisher exact test was used for categorical data. A multivariable logistic regression model was constructed for prediction of treatment failure. All variables with P < 0.1 on univariable analysis were included in the multivariable model.

Receiver operating characteristic analysis cutoffs were determined by the Youden most accurate point. Kaplan-Meier curves were plotted to assess the temporal rate of events, and the log-rank test was computed for the comparison between treatment response and treatment failure. All reported P values were 2-sided, and P < 0.05 was considered statistically significant. All statistics were assessed using SPSS version 24 (IBM Corporation, Armonk, NY).

RESULTS

Demography and Clinical Outcomes

Of 207 patients receiving maintenance IFX (at least 1 dose at or post week 14 after treatment initiation) during the study period, 68 (32.8%) had an available IUS examination. Sixty patients were included in the study after excluding patients with IUS that was not performed during maintenance IFX. Table 1 illustrates the background, clinical, and demographic characteristics of the patients. The mean time between IFX induction and IUS examination was 75 (30-105) months. The median duration of follow-up after IUS was 9.25 months (1-10.25) in the treatment failure group and 10.5 months (6.5-17) in the clinical response group. Three patients had already a single dose escalation before IUS.

Assessment of Treatment Response

Twenty-two (36.6%) patients experienced treatment failure after a median follow-up of 11.5 months (5-16 months). One patient required an increase in the dose of IFX, 18 required another biological treatment, and 3 underwent surgery.

Relationship Between TI Wall Thickness and IFX Therapy Retention

In the treatment response group the median measured TI thickness at the index examination was 2.8 mm (2-3.4 mm), and in the treatment failure group the TI thickness was 5 mm (3.5-6.8 mm; P = 0.002; Fig. 2). Receiver operating characteristic analysis showed that a cutoff value of ≥4 mm provided the most accurate discrimination between the 2 groups, with an area under the curve of 0.84, a sensitivity of 76%, and a specificity of 84% (Fig. 3). Patients with a transmural thickness ≥4 mm had significantly increased risk of treatment failure compared with patients with a transmural thickness <4 mm (18/24 [75%] vs 4/36 [11.1%], respectively; odds ratio, 17.55; 95% confidence interval [CI], 4.0-76; P = 0.02). On multivariable analysis, only a terminal ileum thickness ≥4 mm was associated with the risk of treatment failure (OR, 2.9; 95% CI, 1.49-5.55; P = 0.002) (Table 2). On Cox survival analysis, a terminal ileum thickness <4 mm was significantly associated with the duration of failure-free response (hazard ratio, 5.6; 95% CI, 1.7-17; P < 0.001; Fig. 4).

Relationship Between IFX TL and Treatment Retention

On univariable analysis, there was a significant negative association between IFX TL and treatment failure: 6.6 (0.13-6.97) µg/mL vs 3.9 (3.9-7.8) µg/mL in patients with treatment response and in those with treatment failure, respectively (Fig. 5). However, on multivariable analysis, this finding did not retain statistical association (Table 2).

Relationship Between IFX Serum Level and Intestinal Wall Thickness

There was a significant negative association between IFX TL and TI wall thickness on concurrent IUS. Patients with IFX levels <3.70 µg/mL had significantly higher TI wall thickness than patients with IFX levels >5.9 µg/mL (95% CI, 4.1-6.05; P = 0.004; Fig. 6).

Discussion

In this study, we examined the association between the ultrasonographic measurement of TI thickness during maintenance therapy with IFX in patients with CD and subsequent treatment failure of IFX therapy. We found that a terminal ileum thickness ≥4 mm during maintenance treatment was significantly associated with the risk of IFX therapy failure.

The use of IUS as a significant imaging modality in CD is growing worldwide. Clinicians use IUS for the diagnosis of IBD, for monitoring disease course, and for the diagnosis of luminal and extraluminal complications. This modality is interrater-dependent and expert-dependent and requires significant training.

Moderate rates of interobserver variability for the estimation of disease activity were reported in a cohort of children with CD.¹⁶ Other publications have reported high rates of interobserver agreement between experienced sonographers for the presence/absence of CD lesions and for distinguishing the absence of mild activity or moderate/severe lesions.¹⁷ The use of IUS may be limited because of various technical restrictions, such as excessive abdominal adiposity.

Although IUS in the clinical setting is focused mainly on the TI and the colon and thus requires the use of other cross-imaging modalities for disease localization and definition of disease extent, both small intestine contrast ultrasonography and CT enteroclysis have identified lesions and complications in patients with CD with high levels of sensitivity and specificity.¹⁸

The accuracy of IUS is similar to that of magnetic resonance enterography for detection of disease activity and complications in patients with CD patients.^{12, 19-22} The use of IUS has been found to significantly alter clinical decisions in patients with CD, previously based on clinical evaluation and inflammatory biomarkers.^{12, 23-25} The use of IUS can help clinicians detect mural ultrasonographic changes as early as the first 2 weeks of IFX initiation, mainly reduction in bowel wall color on Doppler signal, bowel wall thickness, and the length of the involved segment.²⁶ The IUS procedure is also an accurate imaging modality for the assessment of disease recurrence after surgery and is strongly associated with the Rutgeert score.^{27, 28} Increased BWT is the most significant feature of CD and has served as a valid marker of inflammation.²⁹ Other ultrasonographic signs of disease activity in CD include pseudostratification of the bowel wall, mesenteric and adipose hypertrophy, enlargement of lymph nodes, and enhancement of blood flow in the bowel wall.⁸

The present study is the first to describe an association of BWT detected by IUS during the maintenance phase with subsequent clinical treatment outcomes of IFX in patients with CD. In comparison to previous studies^{12, 25} that showed the accuracy of IUS in the determination of transmural healing and grading disease activity, we found a specific BWT value that may predict future IFX treatment failure. These results are further supported by our group’s recently published study that found a significant association between ultrasonographic transmural BWT and adalimumab therapy retention.³⁰ This finding, if confirmed by future prospective studies, may improve treatment outcomes by facilitating early intervention, eg dose escalation or switching to other biologic treatments.

We found a significant negative association of BWT with IFX TL. The association of higher anti-tumor necrosis factor TL and improved clinical outcome was previously shown in several studies. For example, an adalimumab TL cutoff of 6.7 was associated with optimal inflammatory and endoscopic outcomes in the POETIC study.³¹ Nevertheless, the association of anti-tumor necrosis factor TL and BWT on IUS is a novel finding. The fact that this outcome was not validated in the multivariable analysis may relate to the retrospective design of our study and the modest size of our cohort. Nevertheless, the similar results that were shown for adalimumab support the validity of our observations.³⁰

Various treatment targets are used in the clinical and research setup for CD.^32-34 Our finding supports the use of transmural healing through detection by IUS as a novel therapeutic target in CD. This endpoint can be used in everyday practice and can be an innovative target in studies evaluating treatment strategies in CD.

The study has some limitations. Although all ultrasound examinations were done during the maintenance period, the interval between treatment onset and IUS was variable. A second limitation relates to the fact that only patients with ileal or ileocolonic disease were included, limiting the possibility to generalize these findings to other segments of the intestine. An additional limitation involves the time gap between the performances of IUS studies and IFX level measurements; this limitation relates to the retrospective design of this study. Another limitation regards the exclusion of postoperative patients from the study. We also limited the studied parameters to BWT and did not include other signs of inflammation. Finally, our study included a relatively modest number of patients.

Conclusions

In conclusion, the current study showed that TI transmural thickness ≥4 mm during maintenance IFX therapy was significantly associated with pending treatment failure. If confirmed by other studies, then these findings will support IUS-gauged transmural healing as a novel therapeutic target for CD.

Author contributions: D. Carter and U. Kopylov conceived the study and drafted the manuscript . D. Carter acquired all the IUS pictures. A. Albshesh was involved in study conception, analysis and interpretation of data, and manuscript drafting. B. Ungar, S. Ben-Horin, and R. Eliakim participated in data interpretation and critical revision of the manuscript for important intellectual property. All authors have approved the final draft submitted.

Conflicts of interest: B. Ungar has received consultation fees from Takeda, Neopharm, Janssen, and AbbVie. S. Ben Horin has received consulting and advisory board fees and/or research support from AbbVie, MSD, Janssen, Takeda, and Celltrion. R. Eliakim has received consultant and speaker fees from Janssen, AbbVie, Takeda, and Medtronic.

U. Kopylov has received speaker fees from AbbVie, Janssen, Medtronic, MSD, and Takeda; research support from Takeda, Medtronic, and Janssen; and consulting fees from Takeda, Medtronic, and AbbVie. D. Carter has received speaker’s fees and/or research support from Takeda, Janssen, AbbVie, and Tarp and consultancy fees from Takeda, AbbVie, and Taro.

References

Author notes