Magnetic resonance imaging of fistula-in-ano: STIR or SPIR?

Patients with clinically suspected anorectal sepsis were studied using MRI in order to determine if T-2 weighted sequences with fat suppression conveyed any additional benefit over conventional short tau inversion recovery (STIR) sequences. 23 consecutive patients (16 male) undergoing MRI for suspected perianal sepsis were studied prospectively using a 1.0 T whole body system and body coil. Axial and coronal T-1 weighted turbo spin echo sequences were obtained, followed by STIR and T-2 weighted spectral fat saturation inversion recovery (SPIR) sequences. Images were assessed for the presence of sepsis or fistula, and information provided by the sequences compared. Active disease was diagnosed in 17 patients, 14 of whom had fistula-in-ano; one intersphincteric, 10 transsphincteric and three extrasphincteric. Internal openings were identified in all of these 14 patients; anal in 10, rectal in two, and both in two. Diagnosis and fistula classification was possible in all of these 14 subjects on the basis of STIR sequences alone. The anal sphincters and pelvic floor musculature were better resolved by STIR than SPIR, leading to easier and more confident determination of fistula anatomy in eight of the 14 (57%). In no case did STIR sequences fail to resolve inflammation seen subsequently on SPIR, despite reduced track intensity. T-1 weighted sequences were generally non-contributory. Both STIR and SPIR sequences are adequate to classify fistula-in-ano, but classification was easier with STIR due to superior resolution of pelvic floor structures.