Elsevier

Clinical Radiology

Volume 72, Issue 7, July 2017, Pages 590-597
Clinical Radiology

There is good agreement between MR enterography and bowel ultrasound with regards to disease location and activity in paediatric inflammatory bowel disease

https://doi.org/10.1016/j.crad.2017.02.008Get rights and content

Highlights

  • There is good agreement between ultrasound and MRI in paediatric IBD.

  • Fewer technical difficulties were reported with ultrasound than MRI.

  • Using histology as the gold standard, both MRI and US show high specificity.

  • We find ultrasound is well tolerated in children and advocate its use.

Aim

To investigate concordance of bowel ultrasound and magnetic resonance enterography (MRE) in identifying active disease in children with inflammatory bowel disease.

Materials and methods

The imaging of children with inflammatory bowel disease who had undergone bowel ultrasound and MRE within 30 days were retrospectively reviewed, from January 2009 to November 2015. Ultrasound was without oral contrast medium; MRI was conducted with patients unsedated with oral contrast medium and gadolinium. Imaging data included bowel thickness, markers of activity, and complications. Endoscopy and biopsy reports were also reviewed.

Results

Forty-nine patients (median age 14 years, 33 male) met the inclusion criteria, and 31 children also had endoscopy within 30 days. Active inflammation was seen in 17.6% of bowel segments at ultrasound and 17.3% at MRE. There was good agreement between ultrasound and MRE on the location and activity of disease (Cohen's kappa 0.75, 95% confidence interval [CI]: 0.66–0.83). One patient had an inflammatory phlegmon detected at MRE only; there was no other significant discrepancy in identifying complications. In patients with histopathology, MRE, and ultrasound demonstrated high specificity 85.1% (77.9–90.6) and 86.6% (79.6–91.8) at the bowel segment level. Technical difficulties, including poor tolerance of oral contrast medium and movement, were more common in MRE.

Conclusion

There was good concordance between MRE and ultrasound for disease location and activity, and fewer technical difficulties with ultrasound. Bowel ultrasound is useful in children, and its use is advocated.

Introduction

Inflammatory bowel disease (IBD) presents a significant global health burden in children. The annual incidence of new Crohn's diagnoses in children is estimated at 0.2–8.5 per 100,000 and for ulcerative colitis 0.5–4.3 per 100,000,1, 2 and the rate of new diagnoses is rising internationally.3

Paediatric IBD differs from adult disease in several aspects, including a significantly higher incidence of colonic involvement in cases of early-onset IBD,4 and a high risk of growth failure, affecting up to 40% of Crohn's disease patients.5 Assessment of the small bowel is recommended at the time of diagnosis for all children with Crohn's disease, unclassified IBD, and suspected ulcerative colitis with atypical symptoms in accordance with the revised Porto criteria6 and can also prove useful in monitoring response to therapy.7

Magnetic resonance enterography (MRE) is increasingly used for assessment of the small bowel in children due to its good sensitivity and specificity compared with reference standards8, 9 and as part of an increasing move away from examinations using ionising radiation, such as barium follow-through examinations, due to concerns regarding the impact of cumulative radiation dose in children with IBD.10

High-quality MRE images in children can be more difficult to obtain than in adults, with poor tolerance of oral contrast medium, movement artefact, and limitations of spatial resolution in small patients particular challenges. This has led some centres to perform a large proportion of their MRE examinations in children <10 years of age under general anaesthesia,11 with its inherent additional risks and economic impact.

Ultrasound has also been investigated for assessment of the small bowel and has the benefits of avoiding ionising radiation, potentially higher spatial resolution than MRE, and inherently dynamic assessment of bowel loops. Although performing bowel US with oral contrast medium has been described as decreasing interobserver variability and increasing sensitivity,12 results without specific oral contrast medium are still good13 and, in the experience of the present authors, is better tolerated by young patients.

Although a large trial is currently underway comparing MRE and ultrasound versus reference standards in adults for assessment of small bowel disease in IBD, there has been limited investigation directly comparing the accuracy of MRE and ultrasound for this indication in children.14, 15, 16, 17

In 2008, the present authors began increasing the use of bowel ultrasound, having noted its excellent patient tolerability and suitability for problem-solving in cases where MRE had been equivocal or degraded, often due to motion artefact. In these early years of establishing the bowel ultrasound service, both ultrasound and MRE were routinely performed as part of the standard of clinical care.

In practice, bowel ultrasound was found to be especially useful in children, and therefore, the present study was undertaken to compare the accuracy of bowel ultrasound versus MRE and histology in this cohort of patients.

Section snippets

Materials and methods

A retrospective review was conducted of the imaging of children (<18 years) with Crohn's or indeterminate colitis who had undergone both bowel ultrasound and MRE within 30 days of each other from January 2009 to November 2015. Endoscopy and biopsy, also performed within 30 days of imaging, were reviewed where available. Institutional review board approval was waived as this was a retrospective review.

Demographics

Forty-nine children underwent both MRE and bowel ultrasound within 30 days of each other during the study period, giving 392 bowel segments. Sixteen children were female and 33 male, the median age was 14 years at the time of imaging (range 7–17 years). Twenty-eight of the paired imaging examinations were for new diagnoses of Crohn's or indeterminate colitis, and 21 were for re-assessments of known disease. A subset of 31 of these children also underwent endoscopy with biopsy within 30 days of

Discussion

The results of the present study show good concordance between MRE and ultrasound with regards to the presence, location, and activity of disease in paediatric patients with IBD. Ultrasound was better tolerated than MRE, with no difficulties due to intolerance of intravenous or oral contrast medium, and less reported difficulty with movement artefact. Where correlating contemporaneous histology was available, specificity was good, but sensitivity was low, particularly in the left colon.

References (26)

  • E. Giles et al.

    Systematic review: MRI enterography for assessment of small bowel involvement in paediatric Crohn's disease

    Aliment Pharmacol Ther

    (2013)
  • C.G. Sauer et al.

    Medical radiation exposure in children with inflammatory bowel disease estimates high cumulative doses

    Inflamm Bowel Dis

    (2011)
  • B.J. Mollard et al.

    MR enterography under the age of 10 years: a single institutional experience

    Pediatr Radiol

    (2016)
  • Cited by (16)

    • ACR Appropriateness Criteria® Crohn Disease-Child

      2022, Journal of the American College of Radiology
      Citation Excerpt :

      For the terminal ileum, the sensitivity and specificity for US was 89% and 55%, respectively, compared with MRE of 78% and 46%, respectively. Overall, including all segments within the colon, sensitivity for both modalities was 46% (US and MRE) and specificity was 87% (US) and 85% (MRE) [34]. A study comparing bowel US and MRE in children with 33 patients prospectively showed that US readings are substantially reliable in the colon and terminal and distal ileum (intraclass correlation coefficients [ICC] of 0.79-0.88) but highly unreliable in the mid and proximal small bowel (ICC of 0.0) [35].

    • Point-of-Care Testing and Home Testing: Pragmatic Considerations for Widespread Incorporation of Stool Tests, Serum Tests, and Intestinal Ultrasound

      2022, Gastroenterology
      Citation Excerpt :

      Lack of both IUS oral preparation, resulting in less prominent prestenotic dilatation and standardized stricture definitions,109 may explain the slightly superior diagnostic performance of stricturing disease using MRE.109 In children, both modalities exhibit equal performance for disease presence, location, and disease activity.147 The extensive UK multicenter METRIC study113 applied a construct reference standard model (panel diagnosis) to tackle the lack of a gold standard in small bowel disease.

    • Diagnostic pathways in Crohn's disease

      2019, Clinical Radiology
      Citation Excerpt :

      If both tests are normal, active IBD is essentially excluded. Conversely, once a diagnosis of CD is established, in most hands MRE is superior for disease mapping (i.e., establishing location and activity), and can be employed at this stage; however, this may not be the case at all centres and for all patients; for example, in children and adolescents, there is excellent agreement between US and MRE.29 At first presentation of CD, many patients have imaging findings that overlap with other conditions.

    • Paediatric bowel ultrasound in inflammatory bowel disease

      2018, European Journal of Radiology
      Citation Excerpt :

      Although ultrasound and MRE have been shown to be accurate in diagnosing IBD in children, neither modality has been shown to reliably distinguish between Crohn's and UC [33]. More recently it has been demonstrated that there is good agreement between MRE and ultrasound in paediatric IBD with regards to location and activity [34] and ultrasound has also been demonstrated to detect as much clinically significant disease as MRE [35]. It is generally accepted that ultrasound is a ‘cheaper’ test than MRI.

    • Early onset inflammatory bowel disease – What the radiologist needs to know

      2018, European Journal of Radiology
      Citation Excerpt :

      The role of ultrasound in IBD imaging is a growing field of interest. It has been demonstrated in a large meta-analysis to be of similar sensitivity and specificity to MRI in adults [51], with promising results also in cohorts of children with IBD [52,53]. Ultrasound has many advantages over MRI in paediatric IBD imaging.

    View all citing articles on Scopus
    View full text