Endoanal magnetic resonance imaging of fistula‐in‐ano: a comparison of stir with gadolinium‐enhanced techniques

Abstract

To compare a STIR sequence with gadolinium-enhanced techniques on endoanal magnetic resonance (MR) imaging of fistulas-in-ano by correlating the findings with those at surgery. Twenty-two consecutive patients with clinical suspicion of perianal sepsis were studied using an endoanal coil followed immediately by a phased array coil. T1-weighted precontrast and postcontrast and STIR images in transverse and coronal planes were produced with each coil and analysed by noting the presence and site of a collection, primary track, the position of any internal opening, and subcutaneous or supralevator extension. An "expert" and also a "trainee" radiologist assessed the images. Operative findings were similarly recorded. The Fisher exact test was used to compare imaging with surgery. Interobserver variation was calculated using a kappa statistic. Of 22 patients with suspected fistulas, 8 were simple, 4 were complex, and 3 were superficial sinuses. Five had no anal pathology, 1 had anal excoriation, and 1 had a polyp. At surgery, 6 intersphincteric, 1 transsphincteric, 8 extrasphincteric, no supralevator collections, and 9 internal openings were noted. The overall sensitivity and specificity for detecting these were 75% and 64%, respectively, for STIR imaging, and 58.3% and 62.8% for gadolinium-enhanced imaging. There was good agreement between the "trainee" and the "expert" in the interpretation of images (kappa=0.7). A STIR sequence is more sensitive overall than gadolinium-enhanced techniques on endoanal magnetic resonance imaging of fistulas-in-ano because of increased sensitivity in detecting the internal opening. A combination of endoanal and phased array techniques using STIR imaging sequences is valuable preoperative assessment in both simple and complex cases.