Simplify to improve in capsule endoscopy – TOP 100 is a swift and reliable evaluation tool for the small bowel inflammatory activity in Crohn’s disease

Abstract

Background: Capsule endoscopy is a widely recognized method to study the small bowel, including in patients with Crohn’s disease (CD). The Lewis score (LS) is a valuable tool in this setting, able to assess inflammatory activity. TOP100, a new software tool of the RAPID Reader^®, emerged to assist in the time-consuming capsule reading process, by automatically selecting 100 images that will most likely contain abnormalities.

Aim: Evaluate the agreement between TOP100 and classic reading (CR) in determining LS in the setting of CD.

Methods: Retrospective study including consecutive patients undergoing small bowel capsule endoscopy (SBCE) for suspected or established CD. One experienced reader performed CR and calculated the LS. Another experienced reader, blinded to the CR results, reviewed all SBCE videos using TOP100 and calculated the LS.

Results: One hundred and fifteen patients were included. SBCE detected significant inflammatory activity (LS ≥135) in 64 patients (55.7%). We verified a strong agreement between the two methods of capsule reading (Kappa = 0.83, p < .001), with an agreement on 89.6% of the cases. The agreement was superior in moderate-to-severe inflammatory activity (Kappa = 0.92, p < .001). All cases of moderate-to-severe activity detected by CR were identified by TOP100 as significant inflammatory activity. A good agreement was verified in all tertiles (p < .001).

Conclusions: Although the classical review of the entire video remains the gold standard, the TOP100 has been shown to be a useful tool in assisting the reader in a prompt calculation of LS, in particular for identifying patients with moderate-to-severe inflammatory disease.

Keywords:

• Capsule endoscopy
• software
• TOP 100
• Lewis score
• Crohn’s disease

Author contributions

Freitas M. and Boal Carvalho P. designed the study, carried out data analysis, and drafted the manuscript. Freitas M. and Arieira C. participated on the acquisition of data and performed the statistical analysis. Boal Carvalho P., Rosa B., and Moreira MJ. critically revised the manuscript. Cotter J. critically revised and approved the final version of the manuscript.

Disclosure statement

Rosa B. has consulting services agreement with Medtronic. The remaining authors declare that they have no conflict of interest.