Patients with a simple superficial fistulas and no macroscopic rectal disease can often be treated by surgery alone without the need for medical therapy beyond initial antibiotics. If the fistula does not traverse any sphincter muscle, these patients are good candidates for fistulotomy. For patients in whom the fistula traverses the sphincter muscle, fistulotomy may still be appropriate if the amount of muscle is small and the continence is not already compromised. Healing rates of up to 100 % have been reported [1–3]. The risk associated with primary fistulotomy is poor wound healing, recurrence and incontinence [2]. For patients in whom fistulotomy is not appropriate, a seton can be considered the primary treatment and left for long term drainage [4]. In some patients. the setons can be removed after the perianal sepsis resolves and the fistula tracts will close [1]. Indeed, a healing rate of up to 25 % has been found in the placebo groups in the medical trials discussed below. If the fistula recurs, the patients should be then be treated as if they have a complex fistula (scenario 3).

The presence of active disease in the rectum will significantly compromise the success rate of surgical intervention. After drainage of sepsis and placement of a seton(s), these patients should be evaluated for medical therapy to try and eradicate the inflammation in the rectum. In some cases, medical therapy may also cure the fistula. Placement of a draining seton prior to inititation of medical treatment has been shown to improve the results with anti-TNF therapy [5]. Below is a brief summary of the major classes of drugs used to treat perianal CD.

Antibiotics are useful to help control perianal sepsis and may also decrease pain, but there are no randomized clinical trials (RCT) that show that antibiotics alone can result in fistula closure. Uncontrolled studies show a benefit with the use of metronidazole or ciprofloxacin that is quickly lost on withdrawal of the drug.

As with antibiotics, there are no RCTs that support the use of azathioprine or its derivatives as single therapy for the treatment of PF in patients with CD. Pearson et al. performed a meta-analysis of RCTs using these drugs and in the subset of patients with perianal disease, found a 54 % response with the drugs versus 21 % with placebo [6]. However, these drugs are slow in onset and are rarely used as first line mono drug therapy for fistulizing perianal disease.

The first randomized placebo controlled trial using anti-TNF therapy in fistulizing CD was performed by Present et al. in 1999. They studied 94 patients and compared a 3 dose induction regimen with either 5 or 10 mg/kg of infliximab to placebo. They found a significantly higher number of patients treated with infliximab had a response and or achieved remission [7]. In this study there was no maintenance therapy and the duration of fistula closure was about 3 months. Subsequently, a maintenance study (ACCENT II) was performed taking patients who responded to induction with infliximab and randomizing them to maintenance every 8 weeks with infliximab or placebo. The infliximab maintenance group had a longer time until loss of response as compared to the placebo group [8].

Initial RCTs with adalimumab included a subgroup analysis for patients with fistulizing disease. In CLASSIC-1 and GAIN there was no benefit to adalimumab over placebo [9, 10]. However, in CHARM there was a significant benefit to the use of adalimumab [11].

Results in studies combining antibiotics and anti-TNF agents have been mixed. West et al. combined infliximab with either ciprofloxacin or placebo, and observed a nonsignificant trend toward a better response with concomitant antibiotics [12]. Dewint et al. performed a RCT adding either placebo or ciprofloxacin to adalimumab and found a significant increase in response and remission with combined therapy. However, the added benefit of the antibiotic was lost when it was discontinued [13].

The anti-TNF drugs have convincingly been shown to reduce fistula drainage and induce remission, but not without high cost and risks including infections, infusion reactions and malignancy. Even when fistula tracts are healed with biologic therapy, numerous studies have demonstrated residual tracts by ultrasound suggesting that the track may not be truly healed [14].