Power of 3D Transrectal Ultrasonography and Magnetic Resonance Imaging in Diagnosis of Cryptogenic Rectal Fistulae

Abstract

Background. The only radical curative treatment for rectal fistulae is surgery. The choice of surgery requires precise characterisation of the fistulous tract. The most common instrumental methods for rectal fistula diagnosis are transrectal ultrasonography (TRUS) and magnetic resonance imaging (MRI).Objectives. Comparative assessment of the diagnostic power of 3D TRUS and MRI techniques in revealing cryptogenic anal fistulae with respect to intraoperative examination.Methods. The study enrolled 92 patients with rectal fistulae aged 27 to 66 years. Fistulous opening was external in 47 (51.1%) and obliterated in 45 (48.9%) patients. The average patient age was 42.7 ± 15.9 years. Surgery for acute paraproctitis 14 to 32 days prior to examination was in history of 58 (63.1%) patients. All patients had preoperative subsequent 3D TRUS and MRI compared with intraoperative examination results.Results. The rate of correct fistulous tract type diagnosis verified with intraoperative revision was 96.7% (89/92) with 3D TRUS and 82.6% (76/92) with MRI (p = 0.0027). The error rate of 3D TRUS estimation of external sphincter involvement was 1.1% (1/92), sensitivity 96.6%, specificity 93.5%, overall accuracy 94.5%. The MRI error rate was 21.7% (20/92), with a statistically significant difference for sensitivity and overall accuracy (p < 0.0001). The rate of correct estimation of internal fistulous localisation in “anorectal clock” was 97.8% (90/92) with 3D TRUS and 90% (81/90) with MRI (p = 0.0342). Internal fistula was not detected with MRI in 2/92 (2.2%) cases, which explains the deviation. Intraoperative revision identified total 113 abscesses. The rate or correct abscess estimation was 97.3% (110/113) with 3D TRUS and 74.7% (71/95) with MRI. MRI failed to detect abscess in 18/113 (15.9%) cases (p < 0.0001).Conclusion. 3D transrectal ultrasonography is statistically superior over magnetic resonance imaging in estimating internal fistula localisation in “anorectal clock”, fistulous type, as well as the fistulous tract location relative to external sphincteric tissue in patients with transsphincteric anal fistulae. Estimation of pararectal and intramural abscesses was also significantly different.