Anal fistula: Levovist-enhanced endoanal ultrasound: a pilot study.

PURPOSE: The aim of the study was to investigate the usefulness of the contrast agent Levovist in ultrasound assessment of anal fistula.
METHODS: Fifteen patients (11 females, mean age 46) with a diagnosis of anal fistula were assessed by physical examination, conventional ultrasound, Levovist-enhanced ultrasound, and surgery. Levovist was injected a cannula into the fistula. The results of physical examination, conventional ultrasound and Levovist-enhanced ultrasound were compared with surgical findings as criterion standard.
RESULTS: At physical examination, three intersphincteric fistulas and two sinuses were diagnosed. Using conventional ultrasound, five intersphincteric and five transsphincteric fistulas were found; four fistulas and one sinus were not detected. Levovist-enhanced ultrasound revealed one sinus, five intersphincteric, seven transsphincteric, and one extrasphincteric fistulas; only one fistula was not detected. At surgery, three intersphincteric, seven transsphincteric, and two sinuses were found; however, the extrasphincteric fistula detected by Levovist was missed. Compared with physical examination, Levovist-enhanced ultrasound and surgery were significantly favorable in the diagnosis of anal fistula (P < 0.05 in chi-squared test and Fisher's exact probability test). The concordance rate of surgery with conventional ultrasound was 69 percent (9/13) and with Levovist-enhanced ultrasound was 77 percent (10/13). However, because the extrasphincteric fistula was missed at surgery, the accuracy of Levovist-enhanced ultrasound was in fact 85 percent (11/13) if surgical finding was not used as the standard. The internal opening was detected at physical examination in 2 patients (13 percent), with conventional ultrasound in 4 patients (27 percent), with Levovist-enhanced ultrasound in 9 patients (60 percent) and during surgery in 11 patients (85 percent). Consistently, Levovist-enhanced ultrasound and surgery were significantly better than physical examination in the diagnosis of internal opening (P < 0.05). One secondary extension and two sphincter defects were detected by both types of ultrasound. The extension was not confirmed during surgery. No patients developed recurrence or nonhealing of wound. One patient developed incontinence to flatus and one developed a perianal hematoma.
CONCLUSION: Levovist-enhanced ultrasound is better at assessing anal fistula than physical examination and conventional ultrasound. However, a future trial comparing Levovist, hydrogen peroxide, and magnetic resonance imaging is needed to establish which is the most cost-effective preoperative imaging technique to use.