RPS 901 - Inflammatory bowel disease (IBD): an ongoing challenge in imaging

Author Block: F. Maccioni, L. Busato, A. Valenti, S. Cardaccio, A. Longhi, M. Barletta, S. Oliva, M. Aloi, C. Catalano; Rome/IT
Purpose: Several MRI parameters have been validated as expressions of Crohn's disease activity, using colonoscopy as the gold standard (GS). However, the only reliable GS for Crohn's disease activity is the pathological specimen, which is rarely available. This study was conducted to select the MRI parameters most indicative of disease activity, compared with several clinical reference GSs, but also independently of a GS, through a cluster analysis.
Methods or Background: We retrospectively analysed 60 paediatric patients with Crohn's disease at onset and at follow up, who underwent MREnterography , ileo-colonoscopy and clinical-laboratory examinations in a 30-day time interval. The sixteen different MRI parameters most indicative of intestinal inflammation were intracorrelated and correlated with the Simple Endoscopic Score for Crohn's Disease, with the pCDActivity score, foecal calprotectin and C-reactive protein. Statistical analysis performed included cluster analysis and Cramer's analysis.
Results or Findings: Significant results were obtained using cluster analysis, which identified two clusters with severe (Group 1) and mild-to-moderate (Group 2) disease. Eight out of sixteen MRI parameters varied significantly within the clusters: the length, DWI, T2 fat suppressed wall signal, T2 fat suppressed fat signal, arterial-phase contrast enhancement, venous-phase contrast enhancement, late-phase contrast enhancement, stratified wall enhancement, and comb sign (p<0.001). No significant association was found between MRI, clinical, laboratory or endoscopic parameters, except a mild association between foecal calprotectin and arterial, venous
and late-phase contrast enhancement.
Conclusion: The cluster analysis identified eight MRI parameters that significantly increased in mild to severe Crohn’s disease. By grading these nine parameters, an effective non-invasive activity score for disease monitoring can be developed and validated in paediatric patients.
Limitations: This is a single-centre retrospective analysis. A deeper analysis of the missed correlation between MRI, clinical, and endoscopic parameters is required.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: No information provided by the submitter.

Author Block: J. Ritter¹, M. Nickel¹, M. El Husseini¹, P. H. Gouder¹, F. Nensa², J. S. Kirschke¹, R. Braren¹; ¹Munich/DE, ²Essen/DE
Purpose: Chronic inflammatory bowel diseases such as ulcerative colitis (UC) and Crohn's disease (CD) are associated with malabsorption disorders. These carry the potential to alter body parameters such as bone density and fat-muscle composition. The aim of this study was to detect and quantify potential changes in the age- and gender-specific parameters of bone density and body composition.
Methods or Background: A total of 433 patients (n=153 UC, n=280 CD) with chronic inflammatory bowel disease were investigated retrospectively. The patient groups studied were age-correlated against each other and a reference cohort (n=811). All patients received a CT scan within the clinical setting. Bone marrow density (BMD) values were extracted using an custom-made segmentation and classification algorithm. Body tissue composition was assessed in a subcohort of 116 patients (35 UC, 81 CD) using a deep learning–based volumetric tissue classification system.
Results or Findings: Significantly reduced BMD compared to a reference cohort was seen in almost all age and gender groups in both UC and CD (p-value range: 1.35E-08 to 0.03). Within the subcohort, the body composition showed significantly lower muscle mass in male UC patients and gender-independent in patients with CD. The proportion of adipose tissue, on the other hand, did not significantly change. In the subgroup of CD patients, significantly lower muscle mass was found with short-term cortisone use, and among continuously-medicating female patients. However, earlier cortisone use did not lead to significant changes in muscle mass.
Conclusion: Chronic inflammatory bowel diseases are associated with loss of BMD and muscle mass. The latter are mainly observed in patients with Crohn's disease on recent cortisone therapy.
Limitations: The limitation of the study is the limited availability of therapy data leading to a relatively small subcohort.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of the University Hospital rechts der Isar, Technical University of Munich (Number 87/18S).

Author Block: Y. Wang¹, Z. Luo², Z. Zhou¹, X. Shen¹, X. Wang¹, B. Huang², S-T. Feng¹, X. Li¹; ¹Guangzhou/CN, ²Shenzhen/CN
Purpose: The purpose of this study was to develop a radiomics model combining VAT and bowel features to improve the predictive efficacy of infliximab (IFX) therapy on the basis of bowel model.
Methods or Background: This retrospective study included 231 CD patients (training cohort, n=112; internal validation cohort, n=48; external validation cohort, n=71) recruited from two tertiary centres. A machine-learning VAT model and bowel model were developed separately to identify CD patients with primary nonresponse to IFX. A comprehensive model consisting of VAT and bowel radiomics features was further established to verify whether features extracted from VAT would improve the predictive efficacy. Area under the curve (AUC) and decision curve analysis were used to compare the prediction performance. Clinical utility was compared using integrated differentiation improvement (IDI).
Results or Findings: VAT model and bowel model exhibited comparable performance for identifying patients with primary nonresponse in internal [AUC of VAT model vs bowel model, 0.737 (95% CI, 0.590-0.854) vs 0.832 (95% CI, 0.750-0.896)] and external validation cohort [AUC of VAT model vs bowel model, 0.714 (95% CI, 0.595-0.815) vs 0.799 (95% CI, 0.687-0.885)], with relatively good net benefit. Comprehensive models adding VAT into bowel model yielded a satisfactory predictive efficacy in both internal [AUC, 0.840 (95% CI, 0.706-0.930)] and external validation cohort [AUC,0.833 (95% CI, 0.726-0.911)], with significantly improved predictive efficacy (IDI=4.2% and 3.7%; both P<0.05).
Conclusion: VAT has an effect on IFX treatment response and improves the performance for identification of CD patients at high risk of primary nonresponse to IFX therapy.
Limitations: An initial limitation was the use of CTE rather than MRE to develop radiomics models. In future studies, CT-based radiomics framework may facilitate artificial development in the field of MR through transfer learning. Besides, the radiomics signatures extracted from single-phase images underutilised all of the information. Lastly, sample size is limited.
Funding for this study: Funding was received from the National Natural Science Foundation of China [grant numbers 82070680, 82072002, 82270693, 82271958].
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board of The First Affiliated Hospital of Sun Yat‐sen University (No. [2020]351), and the need for informed patient consent was waived.

Author Block: L. Huang, R. Zhang, X. Shen, Y. Wang, X. Wang, Z. Li, Z. Li, X. Li; Guangzhou/CN
Purpose: Bowel damage (BD) significantly impacts Crohn's disease (CD) patient's prognosis, showing varied magnetic resonance enterography (MRE) features. This study seeks to investigate the potential mechanisms underlying MRE features of BD using multi-scale data and develop an optimal diagnostic model for BD.
Methods or Background: 230 CD patients were prospectively recruited and categorized into BD (n=103) and non-BD (n=127) using the Lémann index. All patients underwent MRE. Faecal 16S rRNA gene sequencing, and faecal and blood metabolomics were conducted. Relationships between MRE findings, gut microbiota, and faecal/blood metabolites were analysed using causal mediation analysis. Diagnostic models for BD were constructed using gut microbiota, faecal metabolites, blood metabolites, and MRE findings alone or in combination, with their performance assessed using ROC analysis.
Results or Findings: Seven MRE features, including penetration, bowel thickness, and perienteric T2WI signal, differed significantly between BD and non-BD patients. The components of gut microbiota and faecal/blood metabolites were distinct between the two groups. Causal mediation analysis revealed that Blautia may promote intestinal penetration accompanied by perienteric inflammation (i.e., perienteric T2WI signal) through faecal arachidonic acid and blood ceramide (d18:2/23:1). Individual component models, including intestinal microbiota, faecal metabolites, and blood metabolites, yielded AUCs of 0.588 (95% CI: 0.453 - 0.713), 0.671 (95% CI: 0.538 - 0.787), 0.640 (95% CI: 0.533 - 0.737), respectively. Incorporating MRE into these factors to construct a multi-scale model increased the AUC by 0.708 (95% CI: 0.582 - 0.823).
Conclusion: BD-induced MRE findings are partially attributable to microbial and metabolite factors, providing novel insights into the possible mechanisms driving these MRE alterations. The inclusion of MRE enhances diagnostic performance of the multi-scale model for BD, rendering it a promising tool for BD diagnosis.
Limitations: The diagnostic model lacks validation in an external dataset.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Ethics Committee of the First Affiliated Hospital, Sun Yat-sen University (No. IIT-2021[215], Guangzhou, China).

Author Block: V. Válek, T. Rohan, L. Prokopova, D. Bartusek, M. Smela, S. Konecny, J. Husty, V. Zboril, P. Kavrikova; Brno/CZ
Purpose: This study sought to compare the sensitivity and specificity of ultrasound and magnetic resonance (MRI) enterography in the evaluation of inflammatory changes in Crohn's disease (CD) in patients with different macroscopic findings on colonoscopy.
Methods or Background: This retrospective study analysed 47 consecutive patients with CD who underwent contrast enhanced MRI enterography, expert bowel ultrasound and colonoscopy within 1 month at the University Hospital between 2018 and 2022. Based on the endoscopic findings, patients were divided into four groups: 1 (ulcers), 2 (ulcers and aphthae below 5mm), 3 (aphthae below 5mm, oedema, erythema), and 4 (aphthae below 5mm). In all patients, four segments of the colon and terminal ileum were evaluated (together 215 segments).
The sensitivity and specificity of ultrasound and MRI in the evaluation of inflammatory changes were compared in all analysed locations (Fisher's exaxt test). Macroscopic colonoscopic findings were the gold standard.
Results or Findings: Group 1 achieved significantly better sensitivity and specificity in the assessment of inflammatory activity in CD than the other groups (p<0.001), with sensitivity and specificity on MRI 90% and 91% and 88% and 87% on ultrasound. The combination of ultrasound and MRI findings increased sensitivity and specificity to 94% and 93% (p=0.02; 0.003). In the independent colon evaluation, ultrasound achieved higher sensitivity (89%, 81%, 58%, 60%) compared with MRI (83%, 63%, 33%, 20%) in all groups. The difference was significant in group 1 and 2 (p<0.001) and nonsignificant in groups 3 and 4 (p=0.622; 0.576).
Conclusion: A combination of MRI and bowel ultrasound is useful in the evaluation of bowel activity in patients with CD. Ultrasound can contribute to the assessment of disease activity, especially in the colon, compared to MRI.
Limitations: The study was limited by being retrospective, and by its small patient cohort.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board of the University Hospital, Brno.

Author Block: S. N. Yılmazer Zorlu, D. Kuru Öz, G. A. Erden; Ankara/TR
Purpose: The objective of this study is to compare the presence of intramural abscesses formed within the internal anal sphincter, which is the extension of the circular muscle layer of the rectum and a distinct craniocaudal course of the perianal fistula within external the anal sphincter (EAS) in patients with and without Crohn's disease (CD).
Methods or Background: Magnetic resonance images (MRI) with possible presence of perianal fistulas obtained between 2011 and 2023 were retrospectively rewieved. Patients with malignancy, ulcerative colitis, haematological disease, and anorectal surgery were excluded. We divided our cohort into two groups according to medical records; those with CD and those without any underlying aetiology. Images were evaluated for both number and type of fistula according to Park’s classification, localisation of internal orifice ,presence of course within EAS, and presence of intramural abscesses.
Results or Findings: Of the 578 patients with perianal fistula, 109 patients had CD while 469 patients did not have any known disease and have been considered idiopathic. There were 19 patients with concomitant intramural abscess (CD:14, idiopathic:5). There was a statistically significant difference between the groups in terms of the presence of intramural abscess(p<0.001). In CD patients, it was statistically observed that the internal orifice was more frequently located in the distal rectum and at the 3 o'clock position relative to the anal clock, compared to the idiopathic group(p<0.001; p=0.04 respectively). There were 60 fistulas (idiopathic:55, CD:5) that extended along the course of the external sphincter. 66.7% of extrasphincteric fistulas were found in CD patients, although other fistula types were primarily seen as idiopathic. Regarding the fistula localization and presence of course within EAS, there was significant difference between the groups (p<0.001).
Conclusion: Despite the rarity, intramural abscess may be indicative of CD.
Limitations: The findings of this paper must be confirmed with studies involving larger patient numbers. Furthermore, comparisons cannot be made with the intraoperative findings.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Ankara University Clinical Research Ethics Committee.

Author Block: J. Rimola Gibert¹, B. Caballol¹, C. Saavedra², S. Rodriguez Gomez¹, A. Fernàndez-Clotet¹, M. C. Masamunt¹, E. Ricart¹, I. Ordás¹; ¹Barcelona/ES, ²Ibague/CO
Purpose: Changes in the 3D volumetry of perianal fistulas measured by MRI could be an imaging biomarker of interest. The objective of the study is to determine the value of volumetric changes in perianal fistulas in patients with Crohn's disease (CD) after medical treatment to predict subsequent outcome.
Methods or Background: This was a retrospective single-center pilot study. We have included CD patients with perianal fistulas who started biological therapy between 2012 and 2020 and have (1) both pre- and post-treatment (3-18 months) pelvic MRI and (2) follow-up after post-treatment MRI of at least 2 to 5 years. According to the last visit, patients were categorised as clinically active or in remission (absence of suppuration). Using specific software we calculated the 3D volumetry of the fistulas, the active component (hypersignal on T2) and the fibrotic component (hyposignal on T2). We compared MRI volumetric changes between both MRIs and calculated the value of MRI volumetric changes in predicting remission using a multivariate stepwise logistic regression analysis.
Results or Findings: Among the 24 patients included, 13 were in perianal remission in the last follow-up. We determined that the % of change of % of active component as predictor of clinical remission (OR 1.06 [1.02-1.11] P=0.008). The logistic regression analyses indicated that the AUC of the ROC curve was 85,3% (p<0.0001). Our results show that a reduction of ≥16% of the % active component of the fistula has a sensitivity of 84.6 and specificity of 81% to predict clinical remission.
Conclusion: Changes in the 3D volumetry of perianal fistulas on MRI have value in predicting clinical evolution. The decrease in the active component could be a therapeutic target.
Limitations: The low number of patients and the follow-up range are the main limitations of this study.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved with the local code HCB/2021/0380.