Transmural Healing Assessed by Combination of Fecal Calprotectin and Intestinal Ultrasonography Is Associated With Reduced Risk of Bowel Damage Progression in Patients With Crohn’s Disease

Abstract

Background

Fecal calprotectin (Fcal) and intestinal ultrasonography (IUS) could be used as noninvasive tools to monitor mucosal and transmural healing, respectively, in Crohn’s disease (CD). We assessed the agreement between Fcal and IUS to detect active CD and investigated their complementary to predict long-term CD outcomes.

Methods

In this prospective study, we consecutively included CD patients with concomitant IUS and Fcal testing within 7 days. Patients were divided into 4 groups: Transmural healing (TH; both normal), IUS healing (Fcal > 100 µg/g but normal IUS), biochemical remission reflecting mucosal healing (MH; Fcal < 100 µg/g but abnormal IUS), and no healing (abnormal Fcal and IUS). The primary endpoint was active CD. The secondary endpoints were time to bowel damage progression, time to relapse-related drug discontinuation, and patients’ acceptability (10-points acceptability numerical scale).

Results

Among the included 112 patients, 44.6% (50/112), 12.5% (14/112),16.1% (18/112), and 26.8% (30/122) achieved TH, IUS healing, biochemical remission, and no healing, respectively. The agreement between IUS and Fcal to detect an active CD was poor (71.4%, κ-coefficient = 0.41 ± 0.09). Transmural healing was associated with a reduced risk of bowel damage progression compared to no healing (P < .0001) contrary to IUS healing (P = .15) or biochemical remission (P = .84). Transmural healing was associated with a lower risk of relapse-related drug discontinuation than MH (hazard ratio [HR] = 0.09 [0.02-0.45], P = .003), IUS healing (HR = 0.10 [0.02-0.60], P = .001), or no healing (HR = 0.09 [0.018-0.04], P = .002). IUS was better accepted than Fcal testing (9.6 ± 0.8 vs 7.9 ± 2.3; P < .0001, 10-points range-acceptability numerical scale).

Conclusions

Transmural healing, evaluated by the combination of noninvasive and well-accepted tools such as Fcal and IUS, is associated with improved long-term outcomes and could be used to monitor patients with CD in daily practice.

Introduction

Crohn’s disease (CD) is an inflammatory bowel disease (IBD) that can lead to altered quality of life and significant disability for the patients.¹ Due to chronic and transmural inflammation, CD can evolve to bowel damage including stricture and fistula requiring intestinal resection.^2–4 The therapeutic goals have dramatically changed over the last decade in order to attempt to modify the natural history of CD. While clinical remission is a mandatory but insufficient target, the combination of clinical and endoscopic remission has become the recommended objective in patients with CD in the era of advanced therapies.^5,6 Achieving endoscopic remission focusing on mucosal evaluation is associated with more favorable outcomes such as higher rate of sustained clinical remission and reduced risk of surgery or hospitalization.^7–10 However, the need for repeated colonoscopies could restrict the use of this strategy in daily practice due to its limited acceptability.¹¹ Moreover, transmural healing (TH) is becoming a very promising endpoint, as it seems to be associated with better outcomes than endoscopic remission^12,13 and is better accepted than colonoscopy by CD patients.¹¹ Transmural healing is mostly assessed by MRI that it is able to perform a concomitant evaluation of the colon and the small bowel, and demonstrated its reliability to assess transmural inflammation,^14–16 to detect extramural complications, as well as to monitor therapeutic efficacy.^14,16,17 However, the use of MRI procedures is heterogeneous across the world and could be restricted by the cost or the accessibility depending on the area of the world and the organization of the healthcare system. To adhere to the guidelines on treat-to-target strategies and tight monitoring,¹⁸ developing and validating noninvasive tools is an urgent need. Fecal calprotectin (Fcal), better accepted than colonoscopy,¹¹ is now recognized as a validated noninvasive biomarker of endoscopic inflammatory activity in CD.^18,19 Intestinal ultrasonography (IUS) is also a noninvasive tool that could be an interesting alternative to assess transmural disease activity in patients with CD.^20–22 However, the agreement and the potential complementarity of these noninvasive monitoring tools have been poorly investigated so far.

We, therefore, conducted a cross-sectional prospective study to assess the agreement between Fcal and IUS to detect active CD and to investigate the complementary of these tools to predict long-term CD outcomes.

Patients and Methods

Ethical Considerations

The study was performed in accordance with the Declaration of Helsinki, Good Clinical Practice, and applicable regulatory requirements. The study was approved by a local Ethics Committee (IRB00013412, “CHU de Clermont Ferrand IRB #1,” IRB number 2022-CF024) with compliance with the French policy of individual data protection. All authors had access to the study data. All the authors reviewed and approved the final manuscript.

Study Design

We performed a prospective cross-sectional observational study during 2 periods (due to the transient absence of our sonographer), from February 2022 to April 2022 and from December 2023 to April 2024, in 1 IBD referral center. In routine practice, Fcal testing is performed to monitor patients with CD. We consecutively included all the patients with CD (≥18 years old) coming to the IBD center for a routine visit with scheduled IUS during our 2 periods of inclusion. To reduce intraindividual variation, the stools were collected in the morning within 7 days before or after IUS. Patients were instructed to transport the stool samples in a dedicated container at room temperature. Fecal samples were immediately transferred upon patient arrival to the Clermont-Ferrand Hospital Biochemistry Laboratory. Calprotectin was measured using a quantitative immunoturbidimetry test (LIAISON, Diasorin Laboratories), according to the manufacturer’s instructions. Laboratory personnel, who were blinded to IUS results, performed the analyses.

IUS Procedures

The IUS was performed by a 2-year-experienced IBD sonographer (from 30 to 50 procedures by month) blinded by clinical and biochemical results (Fcal) using a dedicated ultrasound system (Aixplorer Ultimate, Supersonic imagine). The IUS report included a segmental evaluation of jejunum, nonterminal ileum, terminal ileum (last 30 cm), cecum/right colon, transverse colon, left/sigmoid colon, and rectum focusing on the following items: Bowel thickness (≥3 mm), vascular pattern (color Doppler signal), bowel wall layer stratification, mesenteric fat wrapping, lymph nodes, and complications such as stricture, abscess, or fistula. All these items were defined according to international guidelines.^23,24

Definitions of Healing

Symptomatic CD was defined using PRO-2 (patients-reported outcomes) as either stool frequency >3 and/or abdominal pain >1.¹⁸ Intestinal ultrasonography was considered abnormal in the presence of at least one of the following CD-related lesions: Bowel thickening (>3 mm), abnormal vascular pattern (color Doppler signal score >0), loss of bowel wall layer stratification, mesenteric fat wrapping, or enlarged lymph nodes and/or CD-related complications (ie, stricture, fistula, or abscess).²⁵ Fecal calprotectin was considered as elevated in case of values above 100 µg/g. This threshold was based on our expertise,^26–28 manufacturer’s recommendations, and clinical relevance^29–32 (cutoff value recommended for mild lesions such as early postoperative recurrence) to optimize the sensitivity of the test. Biochemical remission, indirectly reflecting mucosal healing, was defined as a normal value of Fcal (<100 µg/g) but active CD according to IUS procedure. Intestinal ultrasonography healing was defined as no sign of active inflammation on IUS but an elevated level of Fcal. Transmural healing was considered for the patients with a combination of normal Fcal (<100 µg/g) and normal IUS.³³ No healing was defined by an abnormal level of Fcal and CD-related lesions visualized on IUS.

Study Outcomes

Primary endpoint was active CD described either as abnormal IUS as aforementioned Fcal > 100 µg/g. Secondary endpoints were (1) progression of bowel damage defined as previously^12,14,16 by the need for bowel resection, the occurrence of new stricture or fistula, or worsening of pre-existing stricture (an increase of pre-stenotic dilation or occurrence of major obstructive symptoms) or fistula (new abscess or new fistula tract); (2) relapse-related drug discontinuation defined as therapeutic escalation with a change of medication owing to CD relapse (stool frequency >3 and/or abdominal pain >1 confirmed by an objective marker of inflammation such as endoscopy, imaging, or Fcal); and (3) monitoring tool acceptability assessed using a 10-point acceptability numerical scale.^11,34,35

Data Collection

Study data were collected from the medical records, were anonymized, and managed using an electronic data capture tool hosted at Clermont-Ferrand University Hospital (REDCap, Research Electronic Data Capture).³⁶ REDCap is a secure, web-based application designed to support data capture for research studies, providing (1) an intuitive interface for validated data entry; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data download to common statistical packages; and (4) procedures for importing data from external sources.

At baseline, we collected anonymized data on patient characteristics (age, smoking, gender), CD characteristics (duration, location, phenotype), prior IBD medications, current medications (dose, interval, concomitant immunosuppressants), clinical (PRO-2) and biological (C-reactive protein [CRP], Fcal) activities, as well as IUS data. During follow-up, we collected data on study outcomes.

Statistical Analysis

Statistical analyses were performed using Stata 15 software (StataCorp). All tests were 2-sided, with a type I error set at 0.05.

Continuous data were expressed as mean and standard deviation or median and interquartile range, according to the statistical distribution. The assumption of normality distribution was checked with a Shapiro–Wilk test. The comparisons between groups were performed using chi-squared or Fisher’s exact tests for categorical data, and using analysis of variance (ANOVA) or non-parametric Kruskal–Wallis test (when the assumptions of ANOVA were not met) for continuous variables. The assumption of homoscedasticity was studied using Bartlett’s test. When appropriate (omnibus P-value less than .05), post-hoc tests for side-by-side multiple comparisons were applied: Tukey–Kramer after ANOVA and Dunn after Kruskal–Wallis test.

Agreement between IUS and Fcal to detect active CD (primary endpoint) was assessed by agreement rate and Cohen’s kappa concordance coefficient. The results concerning the kappa coefficient were interpreted according to recommendations reported by Altman³⁷: <0.4: no agreement, 0.4-0.7: poor agreement, >0.7: moderate to strong agreement.

Time-to-drug discontinuation and time to progression of bowel damage were estimated using Kaplan–Meier method. Log-rank test and proportional-hazards Cox model were used to compare groups side by side and to determine prognostic factors associated with the censored outcomes. The proportional-hazards hypothesis was studied using Schoenfeld’s test and plotting residuals. Results were expressed (when possible) as hazard ratio (HR) and 95% confidence interval.

The values of acceptability numerical scale were compared using Student’s paired test.

Results

Baseline Characteristics of the Patients

A total of 112 patients with CD were enrolled in this study. In our cohort, 26.8% (30/112), 16.1% (18/112), 12.5% (14/112), and 44.6% (50/112) of the patients achieved no healing, biochemical remission, IUS healing, and TH, respectively. Their characteristics at baseline are detailed in Table 1. We did not observe any significant difference across the 4 groups except for disease duration (P = .034) and CD phenotype (P = .033) as well as current use of vedolizumab (P = .008) (Table 1). At baseline, 42% (47/112) of the patients were symptomatic according to PRO-2 with a difference between the 4 groups (P > .001) (Table 1). The mean CRP level was not totally similar across the 4 groups of patients (P = .03) (Table 1). The quality of IUS was considered as poor in 14.3% (16/112) of the patients.

Agreement Between IUS and Fecal Calprotectin to Detect Active Disease (Primary Endpoint)

In our cohort, 42.9% (48/112) of the patients presented with abnormal IUS. We found the following IUS lesions: Parietal thickness >3 mm (n = 52 patients), color Doppler signal abnormalities (n = 27 patients), mesentery fat infiltration (n = 18 patients), loss of bowel wall stratification (n = 20 patients), enlarged lymph nodes (n = 5 patients), stenosis (n = 5 patients), and fistula and abscess (no patient) (Table 2).

On the other hand, 31.3% of the patients (44/112) had a positive test for Fcal (>100 µg/g). The agreement between IUS and Fcal (for a cutoff value >100 μg/g) was 71.4% to detect an active CD with κ-coefficient 0.41 ± 0.09. We performed sensitivity analysis with other cutoff values of Fcal and did not find significantly better agreement with IUS (75.0% with κ-coefficient 0.46 ± 0.09 for Fcal > 150 µg/g; 70.0% with κ-coefficient 0.36 ± 0.08 for Fcal > 250 µg/g).

Description of Discrepancies Between IUS and Fecal Calprotectin

Among the 14 patients with no IUS lesion but elevated level of Fcal (levels ranging from 100 to 1420 µg/g), the quality of the IUS procedure was poor or intermediate in 5 and 4 patients, respectively, mainly owing to low echogenicity (thickened abdominal wall). Among the patients with an elevated level of Fcal but normal IUS, 1 with Fcal = 1420 μg/g is known for an isolated colonic CD, 1 with Fcal = 328 μg/g (known for previous right colon location), and 1 with Fcal = 202 µg/g (ileal and pancolonic location), had no lesion on concomitant MRI examination. Another 1 with Fcal = 359 μg/g and an inconclusive IUS (poor quality) was known for complicated ileal CD and had an abnormal MRI with 2 ileal stenoses located near the sigmoid. Two other patients were known short postoperative endoscopic recurrence (less than 5 cm extended) on ileocolonic anastomosis (1 with Fcal = 262 µg/g and 1 with Fcal = 508 μg/g). Among the 8 remaining patients with slightly increased levels of Fcal ranging from 100 to 127 µg/g, 7 had prior bowel resection with ileocolonic anastomosis.

On the other hand, among the 18 patients included in the group with normal values of Fcal but abnormal pictures on IUS, 6 patients had proximal ileum involvement with bowel thickness ranging from 3.2 to 5.2 mm (Fcal ranging from 13 to 57 µg/g), including 4 patients with abnormal doppler signal and 3 with loss of bowel wall stratification. The other patients had terminal ileum involvement (1 patient with concomitant pre-terminal ileum involvement) with bowel thickness ranging from 3 to 7.2 mm (Fcal ranging from 13 to 59 µg/g), including 5 patients with abnormal Doppler signal, 4 with loss of bowel wall stratification, 1 with enlarged lymph nodes, and 2 with mesentery fat infiltration.

Progression of Bowel Damage

In our cohort, 8.9% (10/112) of the patients experienced progression of bowel damage during a mean follow-up after the last examination of 10.7 ± 10.1 months, including 3 patients who had to undergo bowel resection. Progression of bowel damage was observed in 16.7% (5/30), 22.2% (4/18), 7.1% (1/14), and 0% (0/50) among the patients with no healing, biochemical remission, IUS healing, and TH, respectively (P = .001). In analyses adjusted on the presence of symptoms (PRO-2), CRP level >5 g/L, active smoking, and CD location or behavior, TH was associated with a lower risk of bowel damage progression than no healing (P < .0001) (Figure 1). In contrast, biochemical remission reflecting mucosal healing did not reduce the risk of bowel damage progression compared to no healing (P = .84). Despite a trend, IUS healing did not reach statistical significance to reduce the risk of bowel damage progression compared to no healing (P = .15) (Figure 1). Side-by-side comparisons showed that the patients with TH (P = .0002) had a decreased risk of bowel damage progression compared to those with biochemical remission reflecting mucosal healing (Figure 1). Despite a clear trend, there was no statistically significant difference between TH and IUS healing (P = .066). Also, there was no difference regarding the risk of bowel damage progression between biochemical remission and IUS healing (P = .23) (Figure 1).

Relapse-Related Drug Discontinuation

During the follow-up, 17.9% (20/112) of the patients required therapeutic escalation with change of medication owing to active CD. Among them, 26.7% (8/30) had no healing, 33.3% (6/18) had biochemical remission reflecting mucosal healing, 28.6% (4/14) were found with IUS healing, and 4.3% (2/46) achieved TH (P = .0008). The risk of relapse-related drug discontinuation was decreased in patients who achieved TH compared to biochemical remission (HR = 0.09 [0.018-0,45], P = .003), IUS healing (HR = 0.11 [0.019-0.59], P = .01), or no healing (HR = 0.09 [0.018- 0.04], P = .002) (Figure 2). However, there was no difference for biochemical remission (HR = 1.05 [0.36-3.04], P = .93) or IUS healing groups (HR = 1.23 [0.38-4.16], P = .73) compared to those with no healing (Figure 2). In addition, we did not observe any significant difference between the groups of patients with mucosal healing and IUS healing (HR = 1.17 [0.33-4.19], P = .80) (Figure 2).

Comparison of Acceptability Between IUS and Fecal Calprotectin Procedures

Among the 112 patients, IUS was better accepted than Fcal testing (9.6 ± 0.8 vs 7.9 ± 2.3; P < .0001) according to a 10-point range-acceptability numerical scale (Figure 3).

Discussion

In this study, we observed only a poor agreement between Fcal and IUS to assess disease activity in patients with CD. In addition, our study is the first prospective study reporting that achieving TH, evaluated by the combination of 2 noninvasive tools such as Fcal (as a surrogate marker of endoscopic remission) and IUS (as an alternative to MRI for evaluating transmural inflammation), is associated with favorable outcomes such as decreased progression of bowel damage and relapse-related drug discontinuation. Finally, we confirmed that IUS is better accepted than Fcal testing by patients with CD.

The choice of Fcal cutoff value (100 µg/g) to define disease activity was based on our experience, manufacturer’s instructions, as well as clinical relevance.^26–32,38 We aimed to be able to detect early CD lesions such as postoperative recurrence^28,31,32 but also short ileal lesions.^26,30 As a simple and validated index to determine IUS activity is still discussed in clinical practice, we chose to define IUS activity in the same way as the STARDUST trial.²⁵

In our study, we observed a poor agreement between Fcal and IUS to detect active CD based on the classification suggested by Altman and colleagues.³⁷ In a study conducted by Paredes and colleagues,²⁹ including 105 patients with ileal CD, they reported a significant but moderate correlation between Fcal levels and IUS activity. Unfortunately, no agreement study was performed in this study. Another retrospective study from You et al. included 127 patients with CD and reported a positive correlation between Fcal and quantitative IUS scoring.³⁹ Our team previously reported a poor agreement between Fcal and IUS (kappa coefficient = 0.53) in another sample of 56 patients.⁴⁰ To better understand the results of our agreement study, it could be relevant to analyze what is really assessed using Fcal and IUS. Fecal calprotectin is broadly the reflection of the invasion by neutrophils within the mucosa. Fecal calprotectin is now widely admitted as a surrogate biomarker of endoscopic remission, that is, mucosal healing.^18,19,27 However, the main factors influencing the level of Fcal are the extent and the severity of the disease.²⁶ The translation into clinical practice is that short ileal disease could be not detected by Fcal.³⁰ On another hand, IUS enables transmural evaluation. Previous studies showed a good correlation^33,41 or agreement⁴² between IUS and MRI to assess TH. However, mucosal and transmural inflammation are not totally independent.^12,15,43 IUS has been previously associated with poor to moderate agreement (kappa coefficient ranging from 0.63 to 0.76) with endoscopic findings.^22,33,42 Recently, Yzet and colleagues found interesting performances of IUS and Fcal compared to endoscopy in patients with CD and suggested a potential complementarity between these 2 tools.⁴⁴ Poor agreement but similar performances to assess endoscopic remission could reinforce the idea that IUS and Fcal, giving different but relevant information, are 2 promising noninvasive partners to assess CD activity in daily practice. Our description of discrepancies between IUS and Fcal highlighted potential lower performances for Fcal in some specific situations such as lesions located to proximal small bowel, short-extent lesions regardless of the location including lower part of the rectum even though this kind of lesions are more frequent in the terminal ileum.³⁰ In contrast, IUS could be more effective in some patients owing to the potential impact of patients’ echogenicity including obese patients or at least patients with thickened abdominal wall and difficulty but also in some specific locations such as some parts of the colon transverse colon and rectum).

The concept of bowel damage has been widespread in CD during the last decades.² Our team and others reported that TH is associated with long-term favorable outcomes such as longer time spent in steroid-free remission, lower risk of therapeutic escalation, drug discontinuation for failure, hospitalization, and surgery,^{12–14,45–47} and could be more effective than endoscopic remission to prevent bowel damage progression in patients with CD.^12,13 However, these data were obtained with MRI, which is hitherto considered as the reference to assess TH. Data are now accumulating showing the association between ultrasound TH and reduced risk of relapse,²¹ or long-term mucosal healing or TH.^48,49 In 2019, Castiglione and colleagues reported that IUS TH was associated with a higher rate of steroid-free clinical remission, lower rates of hospitalization, and diminished need for surgery at 1 year compared to mucosal healing and no healing.⁵⁰ A retrospective study from Vaughan et al.⁵¹ found that IUS TH was significantly associated with a reduced risk of therapeutic escalation, and steroid use, but failed to confirm its significant association with hospitalization or bowel resection in multivariable analysis. Zorzi and colleagues reported that responders to IUS after 1 year of anti-TNF therapy have a lower cumulative likelihood of the need for surgery or hospitalization than nonresponders.⁵²

Our work confirmed the potential complementarity of Fcal and transmural evaluation. These 2 tools are 2 interesting candidates, thanks to their good acceptability by the patients.¹¹ However, the complementarity of IUS and Fcal has been poorly investigated so far.⁴⁹ In addition, positioning the different noninavsive monitoring tools such as Fcal, MRI, and IUS in the future, both in routine practice and clinical trials, is an important topic that warrants to be further investigated. It is likely that Fcal and IUS could be used more frequently due to better acceptability (for IUS) and lower cost but MRI could stay the reference for assessing TH at baseline and after 6-12 months of treatment due to its easier access to central reading, better evaluation of CD-related complications, and progression of bowel damage assessment. Some initiatives such as the substudy dedicated to IUS (DEEPER-echo) from the DEEPER trial (NCT04973423) are currently ongoing to address the question of whether IUS could give the same information and potentially replace MRI, Fcal, or both, or is complementary to these monitoring tools.

Finally, we confirmed that IUS is better accepted by patients with CD. It is in line with previous data from the ACCEPT study.¹¹ However, the level of acceptability remains high for stool sample collection allowing calprotectin testing.

The main limitations of our study are the absence of a central reading of IUS results, the single-center design and reliance on 1 sonographer that could limit the generalizability of findings, the limited duration of the follow-up, the relatively low frequency of investigated events, and the relatively small sample size for some subgroups. In addition, we chose to not add endoscopic or MRI evaluation to improve patients’ recruitment and because these procedures were not warranted to answer the questions targeted by our study. However, adding these 2 examinations would have provided additional meaningful information for IBD physicians. Our study presents also some strengths. To our knowledge, the complementarity of IUS and Fcal has been poorly investigated hitherto and our work is the first prospective study showing that the combination of IUS TH and Fcal, 2 noninvasive tools, could prevent bowel damage progression in patients with CD. Besides, the comparison of acceptability between these 2 tools could add additional information for IBD physicians.

In conclusion, TH, evaluated by noninvasive and well-accepted tools such as IUS and Fcal, is associated with long-term improved outcomes and could be considered a pragmatic therapeutic target to monitor patients with CD in daily practice. However, additional and independent prospective data with long-term follow-up are needed to confirm our data.

Acknowledgments

We thank CHU Clermont-Ferrand (DRCI) for its recurrent support.

Author Contributions

J.H.: Generation, collection, assembly of data; analysis and interpretation of data; Drafting of the manuscript; Approval of the final version of the manuscript. K.M.: Generation, collection, assembly of data; performing IUS procedure; analysis and interpretation of data; Substantial revision of the manuscript; Approval of the final version of the manuscript. D.C., M.D., M.B.: Generation, collection, assembly of data; Substantial revision of the manuscript; Approval of the final version of the manuscript. B.P.: analysis and interpretation of data; statistical analyses: Substantial revision of the manuscript; Approval of the final version of the manuscript. A.B.: Conception and design of the study; analysis and interpretation of data; Drafting of the manuscript; Approval of the final version of the manuscript.

Funding

None.

Conflicts of interest

A.B.: Consulting fees from Abbvie, Amgen, Arena, Biogen, Celltrion Healthcare, CTMA, Galapagos/AlfaSigma, GutyCare/Resilience Janssen, Lilly, MSD, Nexbiome, Pfizer, Roche, Sandoz, Takeda and Tillotts. Lecture fees from Abbvie, Amgen, Biogen, Celltrion Healthcare Galapagos/AlfaSigma, Janssen, Lilly, Mayoli-Spindler, MSD, Nordic Pharma, Norgine, Pfizer, Roche, Takeda, Tillotts and Vifor Pharma. Research grant from Abbvie, Celltrion Healthcare, Janssen, Lessaffre, Lilly, Pfizer, Sandoz and Takeda. J.H., K.M., D.C., M.D., M.B., B.P.: None. Guarantor or the article: A.B.

References