Background: MRI scanning is an important tool in the diagnosis and assessment of perianal Crohn's disease. We aimed to explore whether measurements of the volume of fistulous tracts on MRI at diagnosis and follow-up correlated with clinical outcomes
Methods: This was a single centre pilot study. Patients with a new diagnosis of perianal Crohn's disease on MRI were identified (n=25). 12 patients had a follow-up MRI for following multimodality treatment were included. Measurements of the length of the tract, width (3 point average), as well as a grading of the severity of associated inflammation (T2 imaging) were made by a specialist radiologist. The volume of the tract was calculated based on an approximation to the volume of a cylinder (πr2). Details of the patient's treatment with biologics, immunomodulators, antibiotics and surgical interventions were collated, as well as an assessment of the clinical response of the perianal disease to treatment (none, partial, or complete healing).
Results: All patients had initiated, or had already received, some treatment for perianal Crohn's disease (thiopurine n=11; antiTNFa n=11; antibiotics n=11; surgical drainage/EUA n=11). In all cases there was a decrease in the grading of inflammation around the tract suggesting some impact of medical therapy. 8 patients had a decrease in tract volume and this was greater than 50% in 7 cases. Of these patients, 1 had a complete clinical response with closure of the fistula, 5 had a partial clinical response with significant symptom improvement but no closure, and 1 patient had no response. In three cases the volume of the tract increased (range 79–214%) despite treatment. Of these patients, 2 experienced no healing whilst the 3rd patient with the smallest increase in volume experienced a partial clinical response.
Conclusions: This pilot study, using simple and readily generalizable technique for MRI measurement of perianal fistula volume in Crohn's disease, appears to suggest that changes in fistula volume on MRI correlate with clinical outcomes. Further prospective work in a larger patient cohort adjusting for additional clinical variables is needed to confirm this observation