[Evaluation of three-dimensional endoanal endosonography of perianal fistulas and correlation with surgical findings]

Abstract

Objective: This study aims to assess the accuracy of three-dimensional endoanal ultrasound (3D-US), two-dimensional ultrasound (2D-US) and physical examination (PE) for the diagnosis of perianal fistulas and correlate the results with intraoperative findings.

Materials and methods: A prospective, observational study with consecutive inclusion of patients was performed between December 2008 and August 2009. Twenty-nine patients diagnosed with a perianal fistula due to undergo surgery were included. All patients underwent PE, 2D-US and 3D-US, and the results were compared to intraoperative findings. The examinations were repeated with hydrogen peroxide instilled through the external opening.

Results: Internal opening (IO): no significant differences with regards to the number of IO diagnosed by PE and 2D-US or 3D-US (P>0.05). Primary tract: good concordance between 3D US and surgery (k=0.61), and this was higher than any of the other techniques used (PE: k=0.41; 2D-US: k=0.56). Secondary tracts: both 2D and 3D-US show good concordance with surgery (86%, k=0.66; 90%, k=0.73, respectively). Abscesses/cavities: The ultrasound examinations showed a moderate concordance with surgery (k=0.438, k=0.540, respectively).

Conclusions: 3D-US shows a higher diagnostic accuracy than 2D-US when compared with surgery to estimate primary fistula height in transphincteric fistulas. 3D-US shows good concordance with surgery for diagnosing primary and secondary tracts and a high sensitivity and specificity for diagnosis of the IO. There was a tendency to overestimate fistula height with 2D-US as shown by the lower specificity of 2D-US for the diagnosis of high transphincteric fistulas and lower sensitivity of 2D-US for low transphincteric fistulas.