Antibiotics: Bacteria may in theory play a role in the appearance and persistence of perianal fistulous disease. Thus antibiotics are sometimes used as first-line therapy for fistula healing. In other cases antibiotics, in view of their prophylactic effects against infections and abscesses, are used as adjuvant (or bridging) therapy. Most of the studies of perianal fistulizing disease treated with antibiotics are uncontrolled and the sample sizes are small. In these studies, both metronidazole[22-24], and ciprofloxacin[25], as well as a combination of the 2 drugs[26], showed an initial beneficial effect on the perianal fistula. Response typically occurs after 6 to 8 wk of treatment and is usually manifest in the form of decreased drainage. Fistula closure is uncommon and symptoms tend to recur after suspending treatment[24].

Recently, a small randomized, double-blinded, placebo-controlled study evaluated ciprofloxacin or metronidazole for the treatment of perianal fistulas in patients with Crohn’s disease[27]. Twenty-five patients were randomized to ciprofloxacin 500 mg (10 patients), metronidazole 500 mg (7 patients) or placebo (8 patients) twice daily for 10 wk. Response (≥ 50% reduction in the number of draining fistulas) at week 10 was seen in 4 patients (40%) treated with ciprofloxacin, 1 patient (14.3%) treated with metronidazole, and 1 patient (12.5%) treated with placebo (P = 0.43). One patient from both the ciprofloxacin and placebo arm and 5 (71.4%) treated with metronidazole dropped out of the study (P < 0.02). This study was probably too small to detect differences between treatment arms.

In two studies, antibiotics were used as an adjuvant or a bridge to other drugs. The use of metronidazole and/or ciprofloxacin induced a response (≥ 50% reduction in the number of draining fistulas) at week 8 with fistula closure occurring in 25% of cases[28]. At week 20, those patients who received additional azathioprine therapy had a better medium-term response (48% vs 15%). It should be pointed out that most of the patients in that study had simple fistulas and that only 9 of the 52 cases were classed as complex fistulas. In a placebo-controlled study, all patients received infliximab (3 induction doses at weeks 6, 8 and 12) and were randomized to receive either 500 mg ciprofloxacin twice daily or a placebo for 12 wk[16]. The response rate (defined as ≥ 50% reduction from baseline in the number of draining fistulas) at week 18 showed a tendency in favor of ciprofloxacin in combination with infliximab compared to infliximab alone (OR: 2.37; 95% CI: 0.94-5.98)[16].

Thiopurines: Azathioprine and 6-mercaptopurine have shown efficacy in the treatment of Crohn’s perianal fistulas. In a meta-analysis of 5 controlled studies, a response (defined as complete closure or decreased drainage) was found in 54% of the patients treated with azathioprine or 6-mercaptopurine compared to 21% in the placebo group (OR: 4.44; 95% CI: 1.50-13.2)[29]. This meta-analysis is limited in that fistula response was a secondary endpoint and not the primary one in all of the studies included. There have been no controlled trials in which the primary endpoint was assessment of the effect of thiopurines on the closure of fistulas in patients with Crohn’s disease.

Anti-tumor necrosis factor (TNF)-α agents: The efficacy of the anti-TNF-α antibody infliximab in fistulizing perianal disease refractory to 3 mo of conventional treatment has been shown in a controlled clinical study[19]. The most favorable outcomes were obtained at doses of 5 mg/kg body weight and 3 induction infusions at 0, 2 and 6 wk. This regimen achieved complete fistula closure (no drainage in 2 visits 4 wk apart) in 55% of the patients compared to only 13% in the placebo group. The mean time to response was 2 wk and the mean duration of response was 12 wk after the last infusion. The ACCENT II study later confirmed the response rate to induction (69% at week 14) in the open-label extension[30]. Responders were randomized to infliximab 5 mg/kg body weight or placebo every 8 wk. At week 54, 36% of the patients on infliximab were able to maintain complete closure compared to 19% of those on placebo (P = 0.009). Similar results have also been reported in clinical practice in a large uncontrolled series[31].

Infliximab maintenance treatment has been shown to decrease the use of hospital resources (fewer hospitalizations and less need for surgery) in patients with fistulizing Crohn’s disease[32]. Nevertheless, it has been reported that in perianal disease, early relapse was common after stopping infliximab treatment, with only 34% of patient maintaining remission at 1 year[33].

Adalimumab, another anti-TNF-α antibody, may also prove effective in perianal Crohn’s disease. In the CHARM study, 33% of the patients randomized to adalimumab achieved long term complete fistula closure versus 13% in the placebo group (secondary endpoint; placebo vs adalimumab group combined; P = 0.043)[34]. Complete fistula healing was sustained for up to 2 years by most of the patients in the open extension trial ADHERE. In a prospective open-label study in patients with active perianal fistulous disease who stopped responding or developed intolerance to infliximab, adalimumab induction therapy (160 mg at week 0 and 80 mg at week 2) induced complete fistula closure at week 4 in 23% of the cases[35].

Other immunomodulators: Randomized studies designed specifically to assess the efficacy of cyclosporine in the closure of fistulas in patients with fistulizing Crohn’s disease have not been published. However, there are several uncontrolled case series which used continuous cyclosporine infusion in patients who had failed conventional therapy[36]. Many patients showed an initial response and were switched to oral cyclosporine; however the response was rapidly lost on drug withdrawal.

Uncontrolled case series suggested that tacrolimus may be useful in the treatment of perianal disease[37,38]. In a small controlled clinical trial, patients treated with tacrolimus (0.2 mg/kg per day) had a higher response rate (defined as closure of at least 50% of fistulas) at week 4 compared to placebo (43% vs 8%), but no differences were observed in terms of complete fistula closure (10% vs 8%)[21].

In a retrospective study methotrexate was used in patients with fistulizing Crohn’s disease; after 6 mo 44% of the patients had partial or complete fistula closure[39]. An early case series suggested that the antimetabolite agent mycophenolate mofetil could be effective in Crohn’s perianal disease[40]. In a more recent uncontrolled study from the same group mycophenolate mofetil induced a partial response in 7 out of 8 patients with perianal fistulas, but the response was subsequently lost in 5 of these 7 patients for several reasons including side effects[41].

In refractory Crohn’s disease, small uncontrolled series showed that thalidomide may be effective in treating complex perianal fistulas[42,43]. Severe side effects, including neuropathy, were common and limited the long term use of the drug. Lenalidomide, an analogue of thalidomide, with lower toxicity and powerful anti-TNF properties was not effective in active luminal Crohn’s disease[44], and has not yet been tested for perianal Crohn’s disease.

Miscellaneous therapies: A pilot open-label study provided data suggesting granulocyte colony-stimulating factor (GM-CSF) is a safe and potentially effective agent for the treatment of active perianal Crohn’s disease[45]. GM-CSF has been used in a placebo-controlled study in patients with luminal Crohn’s disease, some of whom had draining fistulas at study entry. At 6 mo, 4 out of 8 patients in the GM-CSF group and 2 out of 5 in the placebo group had complete fistula closure[46].

Octreotide, a somatostatin analogue, may have a role in treating Crohn’s enterocutaneous fistulas, but has not been used in perianal disease[47]. The effect of elemental diet on perianal Crohn’s disease has been studied in a small retrospective series. Fistulas improved in some patients but early relapse occurred in almost all the cases[48]. In a review of 22 patients with active and refractory perianal Crohn’s disease treated with hyperbaric oxygen, 73% achieved a response[49]. In a randomized, placebo-controlled trial oral, spherical adsorptive carbon was effective for the control of perianal fistulas in patients with Crohn’s disease (remission rates were 29.6% vs 6.7% for placebo)[50]. There is not sufficient evidence for any of these agents to support their use in patients with Crohn’s perianal fistulas outside of clinical trials.

Summary of medical treatment in current guidelines: Despite methodological limitations in the supporting studies, antibiotics and azathioprine or 6-mercaptopurine are considered first-line therapy in complex perianal disease in the ECCO consensus statement[6], and infliximab is reserved as a second-line treatment in case of failure. In the AGA technical review[4,5] infliximab is recommended for treatment of complex perianal disease along with azathioprine or 6-mercaptopurine and antibiotics for the induction phase. Maintenance is recommended with azathioprine or 6-mercaptopurine, in association with infliximab in some cases.

Surgical treatment of complex perianal fistulizing disease aims to control sepsis through abscess drainage and intervention in the fistula tracts, including placement of non-cutting setons[51]. Fistulectomy or fistulotomy are rarely indicated in complex fistulas in view of the high rate of proctectomy because of nonhealing or incontinence associated with the procedure[6,51,52]. In severe cases with high fistulas, endorectal flaps are useful[51,53]. In patients with severe refractory disease, diversion with ostomy (loop ileostomy or end sigmoid colostomy) or even proctectomy might be necessary.

In an uncontrolled study carbon dioxide laser ablation has been used as an alternative treatment in patients with perianal Crohn’s disease[54].

The ideal treatment goal for complex perianal fistulas associated with Crohn’s disease is the closure of all the fistulas and the prevention of recurrence. The best outcomes have been achieved in studies using a combination of medical and surgical therapy[55,56].

Surgery may offer some advantages when combined with medical treatment, for example, infliximab. There is concern, however, that use of infliximab may cause abscesses by inducing rapid closure of the fistulas. This problem might be reduced by performing MRI or EUS-guided EUA to detect all fistula tracks and to insert draining setons. Regueiro et al[55] observed that patients who underwent an EUA before infliximab administration were significantly less likely to have fistula recurrence compared to those treated with infliximab alone (44% vs 79%).

Hyder et al[57] investigated such a strategy in patients with perianal Crohn’s disease. After EUA, 12 out of 22 patients required abscess drainage and 17 out of 22 had at least one drainage seton inserted. The short-term efficacy of that strategy was as high as 85% as measured by the PDAI; although the authors noted that long term healing rates were low. Talbot et al[58] reported a similar strategy in patients with complex fistulas-setons were inserted and removed after the second infliximab infusion. Complete healing of the perianal fistula was obtained in 47% of the patients and all showed at least a partial response.

The presence of active proctitis has a negative impact on the outcome of the surgical treatment of Crohn’s fistulas. A pilot study suggested that infliximab treatment has a beneficial additive effect in the multistep treatment to first improve the proctitis before performing surgery in complex perianal Crohn’s disease with active proctitis[59].

In one recent retrospective study, 21 patients with Crohn’s perianal fistulas and symptomatic perianal disease were treated according to a treatment protocol of serial EUS examinations[60]. Surgical and medical therapy was tailored to the results of the EUS findings with seton placement and incision and drainage procedures performed when appropriate. Follow-up EUS examination guided when to remove setons or when to stop infliximab or antibiotics. Median follow-up was 68 wk (35-101 wk). No abscesses developed in any patient. Eighteen out of 21 patients (86%) had complete drainage cessation initially, and 16 out of 21 (76%) had long term cessation of drainage. Eleven (52%) had no persistent fistula activity on EUS. In 7 of these, the fistula remained closed after stopping infliximab or antibiotics while the remaining 4 continued infliximab for mucosal disease. This study showed that EUS-guided combination surgical and medical treatment with infliximab had high short and long term fistula response rates[60].

In a recent randomized prospective study, 10 patients with active Crohn’s perianal fistulas were randomized to either EUS guidance or control[61]. All patients underwent an initial EUS. Patients in the EUS cohort were evaluated by a colorectal surgeon who had access to the EUS findings. The surgeon was blinded to the results of the initial EUS for those in the control group. EUA with seton placement or incision and drainage was done at the surgeon’s discretion, and all patients received optimal medical therapy including antibiotics, azathioprine or 6-mercaptopurine and infliximab. Patients in the control group received further therapy at the surgeon’s discretion without EUS guidance. Those in the EUS cohort had additional EUS evaluations at weeks 22 and 38 with further therapy based on EUS results. One of 5 (20%) in the control group and 4 of 5 (80%) in the EUS group had complete cessation of drainage. In this small study, EUS guidance for combination medical and surgical therapy in perianal Crohn’s disease appeared to improve outcomes.

Fistula healing is not possible in a significant percentage of patients with complex fistulizing Crohn’s disease managed according to the currently accepted treatment algorithms[4-6]. In addition, systemic medical treatments may be subject to intolerance or loss of response and surgical treatments such as fistulotomy should be used with caution given the risk of incontinence. Thus there is a therapeutic gap in the management of perianal Crohn’s disease, and a number of local therapies which aim to achieve complete closure are under development. Table 1 summarizes studies of these new local treatments. The following sections will discuss these new local treatments in more detail.

Topical tacrolimus: Topical tacrolimus has been used successfully in the treatment of skin diseases with an immune component such as atopic dermatitis[62]. Casson et al[63] therefore decided to investigate whether an in-house-prepared topical formulation could be beneficial in a series of pediatric patients with different manifestations of Crohn’s disease including one case of perianal fistula; the patient responded to treatment although details of fistula healing were not presented.

The efficacy of topical tacrolimus in perianal Crohn’s disease was recently investigated in a randomized placebo-controlled study[64]. In that study, 19 patients, 12 of whom had fistulizing perianal Crohn’s disease, were randomized to topical tacrolimus (1 mg in 1 g ointment applied twice daily) or placebo for 12 wk. In the case of patients with fistulas, the primary outcome measure was improvement defined as ≥ 50% decrease in actively draining fistulas on 2 consecutive visits. Treatment showed a beneficial effect on anal and perianal ulcerating disease but lacked efficacy in the treatment of fistulizing Crohn’s disease.

Fibrin glue: Instilling fibrin glue into fistulas is a simple and safe procedure which does not preclude the use of other techniques or repeat procedures in the case of failure[65]. Several studies have been published of series of patients treated with fibrin glue and success rates vary from 0% to 80% (Table 1). This variability can be attributed, among other things, to the different types of fistulas treated (simple or complex; cryptoglandular, Crohn’s, or traumatic etiology), and the differences in the definition of healing.

Only one controlled study with patients with Crohn’s disease has compared fibrin glue with surgical treatment not involving fibrin glue. In that study, Lindsey et al[66] randomized patients with simple and complex fistulas to treatment with fibrin glue or conventional treatment (fistulotomy or loose seton placement with or without subsequent flap advancement). For the purposes of the study, complex fistulas were defined as high fistulas, fistulas associated with Crohn’s disease, and low fistulas with compromised sphincters. Both Crohn’s patients with complex fistulas reported healing, in one case after a second procedure. Healing among Crohn’s patients in the other arm was not reported.

Intralesional infliximab: Although systemic infliximab administration is considered one of the more efficacious therapeutic options available for complex perianal fistulas associated with Crohn’s disease, several authors have investigated the efficacy of local application of this drug. The main rationale for this approach is to try and avoid the potential systemic toxicity associated with infliximab. The first study to employ this approach was published by Lichtiger et al[67] in 2001. Nine patients with perianal Crohn’s disease refractory to antibiotics or 6-mercaptopurine were treated with a circumferential and intrafistulous injection of infliximab at 0, 4, and 7 wk. Remission or partial response was achieved in 83% of the patients.

Since then, a number of uncontrolled studies have been conducted to assess the feasibility of local infliximab therapy. Poggioli et al[68] included 15 patients with complex perianal fistulas associated with Crohn’s disease. In 9 of these, intravenous infusion of infliximab was felt to be contraindicated because of fibrostenotic disease. The patients were injected with 15-21 mg of infliximab at the internal and external openings and along the fistula tract. The injections were well tolerated and 10 of the 15 patients achieved healing after 3 to 12 injections. A similar study was reported by Asteria et al[69], although patients were excluded if they had received prior treatment with infliximab. Up to 3 injections of 20 mg of infliximab were made along the fistula tract and at both openings. The efficacy endpoint was reduction in fistula drainage of 50% or more (response) or complete cessation of fistula drainage for at least 4 wk (remission). Overall, 8 patients achieved a response (73%) and, of these, 3 achieved remission (27%). After a longer follow-up (mean 10.5 mo, range 7-18 mo), 6 patients were responders and 4 were in remission.

Adipose-derived stem cell therapy: Adult stem cell therapy has promising applications in a number of areas of medicine and has no ethical concerns. Given that liposuction is a relatively safe procedure, an appealing source of adult stem cells is lipoaspirate[70]. The stromal cells obtained are subsequently cultured and expanded to produce autologous adipose-derived adult stem cells (ASCs). Trials of ASCs in the treatment of fistulizing Crohn’s disease have delivered the expanded ASCs by injecting them around the fistula opening and directly into the fistula tract.

The first procedure of this type published in the literature was a case report of a 33-year-old woman with Crohn’s disease and a rectovaginal fistula refractory to treatment[71]. ASCs were injected into the rectal mucosa, close to the sutured internal opening. After resection of the posterior vaginal wall and construction of an advancement vaginal flap, the accessory perineal hole was sealed with 2 mL of fibrin glue. One week after the intervention, the wound had completely healed. In 3 mo of follow-up, no recurrence of the rectovaginal fistula was reported.

A subsequent phase I study assessed 9 ASC injection procedures in 4 patients with fistulizing Crohn’s disease[72]. The series included 3 rectovaginal fistulas and 1 perineal fistula. As before, cells were injected into the rectal mucosa, close to the sutured opening and fistula tracks were then filled with fibrin glue. Two of the 3 rectovaginal fistulas and the perineal fistula had healed after 8 wk.

One randomized clinical trial using ASCs has also been conducted. In a recently completed phase II study, 49 patients with perianal fistula-14 of whom had Crohn’s disease-were randomized to receive ASC therapy and fibrin glue or to receive fibrin glue alone (control group)[73]. The primary outcome measure was the proportion of patients with complete fistula closure. An investigator blinded to the treatment confirmed healing by examination of a digital photograph. Five of the 7 patients with Crohn’s disease assigned to ASC therapy achieved healing compared to one of 7 patients in the control group. The difference was not statistically significant but the study was not powered to detect differences in small subgroups such as those with Crohn’s disease. Healing in this trial was defined as the absence of drainage, as well as complete epithelization of external openings. No severe adverse events related to ASCs have been reported when utilized for fistulizing Crohn’s disease.

Fistula plugs: Recently, the use of bioprosthetic plugs made from porcine intestinal submucosa has been tried in patients with perianal fistula. In a prospective study which excluded patients with Crohn’s disease, Johnson et al[74] randomized patients with high transsphincteric or deeper fistulas to either fistula plug or fibrin glue therapy. Of the 10 patients who underwent fibrin glue treatment, 6 (60%) had persistence of one or more fistulas at 3 m compared to 2 out of 15 (13%) of those who underwent the procedure with the fistula plug (P < 0.05). In a subsequent prospective but uncontrolled study, 20 Crohn’s patients with a total of 36 fistula tracts not amenable to fistulotomy were treated with fistula plugs[75]. After irrigation with hydrogen peroxide, each primary opening was occluded with a fistula anal plug. The authors found an overall success rate of 80%, although they noted that patients with complex fistulas with multiple primary openings were less likely to achieve success. Success appeared to be independent of the presence of setons or the use of anti-TNF-α therapy.

Schwandner et al[76] have reported their experience in a series of 19 patients with transsphincteric anorectal fistulas. Seven of these patients had Crohn’s disease. The surgical procedure comprised irrigation of the fistula tract and placement and internal fixation of the anal fistula plug without flap advancement or excision of the fistula tract. Success was defined as closure of both the internal and external openings with no further interventions and absence of abscess formation. Six of the 7 patients with Crohn’s disease achieved success (85.7%).

In another retrospective study, anal fistula plugs were more successful in the treatment of simple anorectal fistulas but were associated with a high failure rate in complex perianal fistulas and particularly in patients with Crohn’s disease (closure rate of 26.6% of fistulas in this group)[77].

In a recent retrospective review reported at the 2008 Digestive Diseases Week conference[78], the use of anal fistula plugs was associated with a lower success rate than previously reported: only 2 of the 22 (15%) Crohn’s disease-associated fistulas healed. In 87% of the procedures the reason for failure was sepsis. These controversial results in an uncontrolled series await confirmation by randomized trials.