INTRODUCTION: We report on a patient cohort with dorsal horseshoe fistulas-in-ano. We sought to answer the question of whether these fistulas can be operatively treated, implementing a sphincter-preserving fistulectomy with primary closure of the internal opening, as is done when treating transsphincteric anal fistulas. Long-term clinical course is examined here and operative methods are discussed. METHODS: During the time period from 1985 to 2000, 42 patients (29 men, 13 women)-with an average age of 44 ±11 years were operatively treated for horseshoe fistulas-in-ano originating in cryptoglandular regions. Twenty patients originally had an abscess, which was surgically drained and then a seton was placed in the tract of the fistula. Later, a fistulectomy or curettage of the fistula tract with primary closure of the internal fistula opening was performed in all patients without severing the sphincter muscle. We implemented four different surgical techniques to facilitate this closure: the mucosa-submucosa advancement flap, the rectal wall advancement flap (part or full thickness), the ano-cutaneous advancement flap, and direct closure without any further mobilization. The follow-up averaged 58 months (1-14 years). RESULTS: Thirty-seven of the 42 fistulas (88 percent) healed. In 31 patients, restitution occurred after the first operation, in 4 patients after the second operation and in 2 patients after the third operation. One patient developed a recurrence after the first operation and died from secondary causes before a second operation was performed. The other four patients were listed as unclear, because the time of observation was less than one year. The total recurrence rate of flap procedures is 23 percent (mucosa-submucosa advancement flap, 25 percent; rectal wall advancement flap, 35 percent; anocutaneous advancement flap, 25 percent; direct closure, 0 percent; not significant). Thirty-four (81 percent) of the 42 patients had previously been operatively treated on an average of three times. Twelve patients developed deficits in continence. Eight patients developed minor deficits, which included incontinence for flatus and problems with staining. Four patients became incontinent for liquid stools. There was a significant decrease in manometric resting pressure of 25 percent (from 123 ±40 cm H2O to 91 ±29 cm H2O) and in squeeze pressure of 21 percent (from 262 ±70 cm H2O to 207 ±66 cm H2O). CONCLUSIONS: As in other high anal fistulas, horseshoe-shaped anal fistulas can be operatively treated implementing a fistulectomy combined with any of the above-mentioned forms of closure of the internal fistula opening, with good success rates and acceptable postoperative continence. Sufficient drainage of the retroanal region is of utmost importance. Through these measures, it is possible to avoid severing the sphincter muscle and to prevent an anal canal deformation (keyhole deformity). INTRODUCTION: We report to a patient cohort with dorsal horseshoe fistulas-in-ano. We sought to answer the question of whether these fistulas can be operatively treated, implementing a sphincter-preserving fistulectomy with primary closure of the internal opening, as is done when METHODS: During the time period from 1985 to 2000, 42 patients (29 men, 13 women) -with an average age of 44 ± 11 years were operatively treated for horseshoe fistulas-in-ano originating in cryptoglandular regions. Twenty patients originally had an abscess, which was surgically drained and then a seton was placed in the tract of the fistula. Later, a fistulectomy or curettage of the fistula tract with primary closure of the internal fistula opening was performed in all patients without severing the sphincter muscle. We implemented four different surgical techniques to facilitate this closure: the mucosa-submucosa advancement flap, the rectal wall advancement flap (part or full thickness), the ano-cutaneous advancement flap, and direct closure without any further mobilization. The follow-up averaged 58 months (1-14 years). RESULTS: Thirty-seven of the 42 fistulas (88 percent) healed. In 31 patients, restitution occurred after the first operation, in 4 patients after the second operation and in 2 patients after the third operation. One patient developed a recurrence after the first operation and died from secondary causes before a second operation was performed. The other four patients were listed as unclear, because the time of observation was less than one year. The total recurrence rate of flap procedures was 23% (mucosa-submucosa advancement flap, 25 percent ; rectal wall advancement flap, 35 percent; ano-cut advancement flap, 25 percent; direct closure, 0 percent; not significant). Thirty- four (81 percent) of the 42 patients had previously been operatively treEight patients developed minor deficits, which included incontinence for flatus and problems with staining. Four patients became incontinent for liquid stools. There was a significant decrease in manometric resting pressure of 25 CONCLUSIONS: As in other high anal fistulas, horseshoe-shaped anal (from 262 ± 70 cm H2O to 207 ± 66 cm H2O) fistulas can be operatively treated with a fistulectomy combined with any of the above-mentioned forms of closure of the internal fistula opening, with good success rates and acceptable postoperative continence. Sufficient drainage of the retroanal region is of utmost importance. is possible to avoid severing the sphincter muscle and to prevent an anal canal deformation (keyhole deformity).