Comparison of the Efficacy of the Various Treatment Modalities in the Management of Perianal Crohn’s Fistula: A Review

Crohn's disease (CD) is a transmural inflammatory bowel disease (IBD) that can affect any part of the gastrointestinal (GI) tract. With the disease's progression, adhesions and transmural fissuring, intra-abdominal abscesses, and fistula tracts may develop. An anal fistula (or fistula-in-ano) is a chronic abnormal epithelial lined tract communicating the anorectal lumen (internal opening) to the perineal or buttock skin (external opening). The risk of fistula development varies from 14%-38%. It can cause significant morbidity, which adversely impacts the quality of life. It is mostly believed that an anal crypt gland infection causes anal abscesses, leading to fistula development. Crohn's disease's pathogenesis involves Th1 and Th17 hypersensitivity due to an unknown antigen within the intestinal mucosa. Evidence to support this review was gathered via the Pubmed database. Search terms used were combinations of "Perianal fistula," "seton," "immunotherapy." Studies were reviewed and cross‐referenced for additional reports. Setons are surgical thread loops passed from the external to the internal opening of the fistula tract and exteriorized through the anorectal canal, facilitating abscess drainage and inciting a local inflammatory reaction, thus promoting the resolution of the fistula. Biologicals such as anti-tumor necrosis factor (TNF) antibody (infliximab, adalimumab, certolizumab), anti-IL-12/23 (ustekinumab), and anti-α₄β₇ integrin antibody (vedolizumab) have been approved for Crohn's disease targeting the Th1/Th17-mediated inflammation. Other therapeutic modalities are fistulotomy, cyanoacrylate glue, bioprosthetic plugs, mucosal advancement flap, ligation of inter-sphincteric fistula tract (LIFT), diverting stoma, proctectomy, video-assisted anal fistula treatment (VAAFT), and fistula laser closure (FiLaC). Our review found that chronic seton therapy should be the primary approach, especially if the patient has a perianal abscess. It has a low incidence of re-intervention, recurrent abscess formation, and side-branching of the fistulous tract, with preservation of the fistulous tract's patency and cost-effectiveness. The major disadvantage of seton therapy is the discomfort and time to achieve stability. Among the biologicals, infliximab is the only therapy which has a statistically significant effect on the healing rate of perianal Crohn's fistula compared to placebo, but the major disadvantage associated with anti-TNF as sole therapy is high re-intervention rate, prolong maintenance therapy, high recurrence rate, and severe side effects. We hypothesize that the two aspects should be addressed concurrently to increase the fistula healing or closure rate. First, the seton should be used as initial therapy to maintain tract patency to allow abscess drainage and minimize the intestinal flora colonization within the tract mucosa, thereby leukocytic infiltration and propagation of inflammation within the tract. The second aspect that has to be considered is that we should target the initial stimulation of the Th1/Th17 mediated hypersensitivity instead of a factor/cytokine involved in the inflammation mediation. Although the unknown antigen triggering such hypersensitivity is not clear, we could target the RAR-related orphan receptor γ (RORγ)-T (transcription factor involved in activation of Th17 cells) and the T-bet (transcription factor involved in activation of Th17 cells) within the GI mucosa by a novel target immune therapy.


Introduction And Background
The doctrine of fistula-in-ano treatment is to close the fistula tract without compromising continence [1]. The data from the National Health Interview Survey (NHIS) in 2015 revealed an estimated 3.1 million, or 1.3%, of US adults have inflammatory bowel disease (IBD), with a significant increase in prevalence and hospitalization rate from 1999 [2,3,4]. The prevalence also differed significantly among several sociodemographic factors [2]. The mean hospitalization costs were $11,345 for Crohn's disease (CD). It increased annually by 3% from 2003 to 2008, although unchanged between 2008 to 2014 [5].

FIGURE 1: The sequence of events leading to perianal fistula formation
The pathogenesis of Crohn's disease involves Th1 and Th17 hypersensitivity due to an unknown antigen (possibly enteric floral antigens) within the intestinal mucosa. Increased production of transforming growth factor-β (TGF-β) and Interleukin-6 (IL-6) is responsible for the commitment of naive T-helper cells (Th0 cells) to Th17 cells, while IL-12 is required for differentiation of a Th0 cell into a Th1 cell. The production of IL-21 and IL-23 is responsible for the maintenance and expansion of the Th17 cells, while tumor necrosis factor (TNF) mediates the inflammation [14]. In the inflammatory infiltrate, IL-12, TNF, and IL-13 induce epithelial-to-mesenchymal transition and upregulation of matrix metalloproteinases, leading to tissue remodeling and fistula formation [15].

Classifications of perianal fistula
Park's fistula classification [16] is based on the fistula's relation to the external sphincter muscle and is a more anatomically precise classification (Table 1, Figure 2). The American Gastroenterological Association (AGA) Classification [17] for perianal fistulas is based on the complexity ( Table 2), and is widely used as an empirical classification and includes a physical examination of the perianal area. Classification of the fistula based on four types, based on the relation of the fistula to the sphincter muscle; aka Park's Classification)

Park's Classification
Type Name Comments I Superficial Fistula tract does not traverses through any sphincter or musculature II Intersphincteric (most common) Fistula tract traverses between the internal and external anal sphincter through the intersphincteric plane III Transsphincteric Fistula tract traverses through the external anal sphincter IV Suprasphincteric Fistula starts in the intersphincteric plane and then passes upward to a point above the puborectalis muscle, and then laterally over this muscle and downward between the puborectal and levator muscles into the ischiorectal fossa V Extrasphincteric Fistula passes from the perineal skin through the ischiorectal fossa and levator ani muscle, and finally penetrates the rectal wall. May arise from trauma, foreign body, pelvic abscess, or cryptoglandular abscess   Setons are surgical thread loops passed from the external to the internal opening of the fistula tract and exteriorized through the anorectal canal, facilitating abscess drainage and inciting a local inflammatory reaction [18], thus promoting the resolution of the fistula. Surgeons usually prefer setons when the patient has a concomitant perianal abscess since it also allows incision and drainage. In addition, antibiotics (metronidazole or ciprofloxacin) are also supplemented, which further promotes healing [15].
The biologicals, e.g., anti-TNF antibody (infliximab, adalimumab, certolizumab), anti-IL-12/23 (ustekinumab), and anti-α₄β₇ integrin antibody (vedolizumab) have been approved for Crohn's disease targeting the Th1/Th17 mediated inflammation and the diapedesis of the leukocytes in the intestinal mucosa respectively. With superficial and low inter-sphincteric fistulas, fistulotomy is often the choice, especially in the absence of active proctitis and a failed antibiotic therapy or any local therapy. Other modalities to manage the perianal Crohn's fistula are cyanoacrylate glue, bioprosthetic plugs, mucosal

Seton Placement
Setons have been considered the mainstay of surgical management for most fistulae-in-ano. There are mainly two techniques for seton placement based on the knot's tightness and the cutting nature. The first is the cutting or tight seton technique, in which a non-absorbable thread is inserted into the fistula tract and exteriorized through the anorectal canal, which incites inflammation and fibrosis. The setons are subsequently tightened, causing gradual, controlled cutting of the sphincter (staged fistulotomy) while allowing the reactive fibrosis to occur, ensuring the sphincter integrity [19].
In a study done by Raslan et al. with 28 patients with complex perianal Crohn's fistula, a 90.2% healing rate was noted by the end of the study (one year), with a recurrence rate of 9.8%. A direct correlation between the healing time and the distance from the anal verge was also observed [13]. There have been many studies indicating complications associated with this technique, including prolonged perianal pain and incontinence. Although the rate of the incontinence associated varied, some reported minor damage 15.7% to flatus, 5.9% to liquid stools, and no incontinence to solid stool [13]; others indicated a significant incontinence rate of 20-67% [20] and 58% (103/178) [21] of varying degrees.
The second technique is the loose seton technique developed to avoid cutting the anal sphincter, thereby theoretically reducing incontinence risk. It can be either placed as a long-term indwelling seton [22] or used as a two-staged fistulotomy [23]. Indwelling seton inhibits the pus collection and promotes continuous drainage of the abscess, hence usually placed after the abscess drainage. In some cases, it results in closure of the fistula, generally within six to 12 weeks [24]. In a study by Thornton [26]. Incontinence rates associated with loose setons have been significantly lower, varied from 5-17%, compared to cutting setons [27,28]. No patient reported a deterioration in fecal continence after seton insertion in a study by Thornton et al. and Kelly et al.
The two-stage seton fistulotomy is more commonly used. Besides continuous drainage, the loose seton placement also incites a fibrotic reaction, which may lead to primary closure or promote migration of the fistula tract superficially, usually within six to eight weeks, after which fistulotomy or fistulectomy can be safely performed. In a study by Kelly et al., 93% (186/200) underwent two-staged seton fistulotomy, which results in the clearance of fistula in all patients. Only 4% (eight) described minor urgency, and none reported incontinence at follow-up [1]. Although the study conducted by Galis-Rozen et al., in the Crohn's group, two-staged seton fistulotomy, only 59% (10) showed significant clinical improvement while 35% (six) showed no improvement [24].
Recently, many surgeons have stopped using the cutting seton technique due to their association with postoperative discomfort and the increased risk of incontinence. Still, many surgeons prefer to use this technique. Incontinence rates associated with loose setons have been significantly lower (5-17%) compared to cutting setons (20-67%) [27,28]. It is important to note that most studies conducted before 2000 agree that cutting seton is at a more disadvantageous position than loose setons, but no convincing evidence of superior efficacy [29] or reduced sphincter destruction has been found in later subsequent studies. It is supported by the fact that the studies showed no statistical significance or consisted of a small sample size (low power) to confer strong statistical relevance. In a case series of 59 patients, Drager et al. reported no significant difference in function or healing (92%) or recurrence rates between the two methods [30]. However, the loose setons' placement remained the preferred choice of surgeons, and many consider it the gold standard for complex fistula [31], even when used as a combination of medical and surgical treatment. In a broader context, loose seton placement, especially as a two-staged seton fistulotomy for a complex fistula, is cost-efficient, well-tolerated, and efficacious ( Table 4).

Studies Method¥ Patient Characteristics Key Findings with Complications (if any) Limitations
Thornton et al. [22] Case series

Biologicals
The pathogenesis ( Figure 3) of Crohn's disease involves Th1 and Th17 hypersensitivity due to an unknown antigen (possibly enteric floral antigens) within the intestinal mucosa. Increased production of TGF-β and IL-6 is responsible for the commitment of naive Th cells (Th0 cells) to Th17 cells, while IL-12 is required for differentiation of a Th0 cell into a Th1 cell. The production of IL-21 and IL-23 is responsible for the maintenance and expansion of the Th17 cells, while the local production of TNF-α mediates and propagates the inflammation [14]. In the inflammatory infiltrate, IL-12, TNF, and IL-13 induce epithelial-tomesenchymal transition and upregulation of matrix metalloproteinases, leading to tissue remodeling and fistula formation [15]. Therefore any pharmacological agent interrupting this pathway will theoretically resolve the inflammation and the progression of the disease. The antagonist to TNF-α has been approved for the management of Crohn's disease and its complications. In a study conducted by Present et al. on patients with perianal Crohn's fistula, infliximab IV at 5mg/kg resulted in a ≥ 50% reduction in the number of draining fistula in 68% of the patients as compared to the placebo (p=0.02), while 55% had the fistula closure [33]. In another study by Farrell et al., 70% (23/33) of patients with fistulous disease experienced >50% reduction in their Perianal Disease Activity Index (PDAI) at two weeks with 5mg/kg infusion of infliximab [34]. Similarly, in a prospective cohort study by Luna-Chadid et al., after the third infusion of infliximab (5mg/kg), the response was partial (≥ 50% from baseline) in 26% (27/105) and complete (full closure) in 57% (60/105), i.e., 82% (87/105) were responders with infliximab four weeks post-infusion of the last dose [35]. A randomized controlled trial (RCT) by Sands et al. found the response rate to be 43% (130/305) in perianal Crohn's fistulous patients who received three infusions of Infliximab at 0, two, and six weeks and the median time for the loss of response after receiving three infusions of infliximab was 14 weeks [36]. To conclude that infliximab does have a statistically significant effect on the healing rate as compared to placebo for the perianal Crohn's fistula, but it also requires a prolonged maintenance dose after induction therapy to maintain the remission. A study by Yarur et al. [37] also proves an incremental gain in fistula healing with higher infliximab levels. The area under the curve (AUC) for the association between fistula healing and infliximab levels was 0.82 (p < 0.0001), while the AUC for the association of infliximab levels and fistula closure was 0.69 (p = 0.014). The patients who did not show any improvement or lower infliximab levels in serum were having anti-infliximab antibodies and had a lower chance of achieving fistula healing (OR: 0.04 [95%CI: 0.005-0.3], p < 0.001).
Adalimumab (ADA) is a fully human anti-TNF IgG1 monoclonal antibody that has also been shown to induce and maintain clinical response in active CD not controlled by corticosteroids, immunosuppressants, or both, but the results of its overall effectiveness are questionable, while the effect on fistula improvement has not been proven so far. In an RCT (CLASSIC-I: Clinical assessment of Adalimumab Safety and efficacy Studied as Induction therapy in Crohn's disease) conducted by the Hanauer et al. [38] on perianal Crohn's fistula patients, the rates of remission for the Crohn's disease at week four in the adalimumab 40mg/20mg, 80mg/40mg, and 160mg/80mg groups were 18% (p = 0.36), 24% (p = 0.06), and 36% (p = 0.001), respectively, and 12% in the placebo group. While the rates of fistula improvement at week four in the adalimumab 40mg/20mg, 80mg/40mg, and 160mg/80mg groups were 3/4 (75%), 2/10 (20%), and 1/12 (8%) respectively, and 2/6 (33%) in the placebo group, and the rates of fistula remission at week four in the adalimumab 40mg/20mg, 80mg/40mg, and 160mg/80mg groups were 3/4 (75%), 0/10 (0%), and 0/12 (0%) respectively, and 1/6 (17%) in the placebo group. Therefore adalimumab was not found to be superior to placebo in the management of the perianal fistula treatment. However, with a minimal beneficial effect on the rates of remission in moderate to severe Crohn's disease compared to placebo (note only the highest dose group at week four achieved statistical significance [p = 0.001] as compared to placebo). The CLASSIC-II trial (RCT) by Sandborn et al. [39] was a continuation of the CLASSIC-I trial to observe adalimumab's effect on the maintenance of remission. Group I consisted of 55 patients in remission were re-randomized and received adalimumab 40mg every other week ( [41]. The major limitation of this study is that the many patients receiving placebo ceased treatment because of an adverse event (13.4%) than those receiving adalimumab, which might have increased the contrast between the treatment and placebo groups.
Certolizumab pegol is a pegylated humanized Fab' fragment that binds TNF-α. In an RCT conducted by Sandborn et al., 662 patients with moderate-severe Crohn's disease were assigned randomly to receive either 400mg of certolizumab pegol or placebo subcutaneously at weeks 0, two, and four and then every four weeks. In the overall population, at week six, induction response rates based on Crohn's Disease Activity Index (CDAI) were 35% in the certolizumab group and 27% in the placebo group (p = 0.02); at both weeks six and 26, the response rates were 23% and 16%, respectively (p = 0.02). Through week 26, 30% (14/46) of patients in the certolizumab group had fistula closure compared to 31% (19/61) of patients in the placebo group [42]. So as per the study, certolizumab does have a modest effect in response rates in patients with moderate to severe Crohn's disease, but the use of the certolizumab pegol for the fistula closure in the case of Crohn's does not show any effect in contrast to the placebo (statistically significant finding). The major limitation of the trial was that many patients were also receiving concurrent therapy, so the statistically significant effect of certolizumab therapy presented in the study can not be conclusively established whether it was the effect of the certolizumab or the concurrent therapy or the combination of the therapy. However, the concurrent treatment was proportionately divided into the placebo and the certolizumab group. Schreiber et al. [43] performed a similar RCT; 668 adults with moderate-severe Crohn's disease entered the induction phase. Fourteen percent (58/425; 28 in the certolizumab group and 30 in the placebo group) of patients who responded to induction therapy with certolizumab pegol had draining fistulas at baseline. In the certolizumab group, 54% (15/28) had closure of the fistula (defined as the absence of drainage on gentle compression at any two consecutive visits three weeks apart), compared with 43% (13/30) in the placebo group. Again no statistical significance was noted for the fistula closure with certolizumab pegol therapy as compared to the placebo. However, another subsequent RCT by the same author included 108 adults with draining Crohn's fistula, out of which nonresponders (50/108) at week six to open-label certolizumab therapy were excluded. The responders (the majority 55/58 had perianal fistula) were then randomized to certolizumab pegol 400mg (n = 28) or placebo (n = 30) every four weeks. It was found that at week 26, 36% of patients in the certolizumab pegol group had 100% fistula closure compared with 17% of patients receiving placebo (p = 0.038). The major limitations of the results are the exclusion of the nonresponders and the small sample size to confer any statistical significance. The study also concluded that the protocol-defined fistula closure (≥50% closure at two consecutive post-baseline visits ≥three weeks apart) was not statistically significant (p = 0.069), with 54% and 43% of patients treated with certolizumab pegol and placebo achieving this endpoint, respectively [44].
Vedolizumab is a monoclonal antibody directed against the α₄β₇ integrin on the T cells, inhibiting the T cells' transmigration, limiting the trafficking into the GI mucosa. In an RCT (NCT00783692 trial) conducted by Sandborn et al. [45], 368 patients (group 1) were randomly assigned to receive either IV vedolizumab (300mg; Group 1a; 220 cases) or placebo (Group 1b; 148 cases) at weeks 0 and two, while 747 patients (Group 2) received the open-label vedolizumab at weeks 0 and two. At week six, 461 showed a response to vedolizumab (vedolizumab responders), who were then randomly assigned to receive either IV vedolizumab (300mg) every four weeks ( With the vedolizumab responders, the fistula closure rate was significantly lower (31%) at the end of week 52 in the maintenance phase of the responders. The confidence interval of the absolute risk reduction at week 14 crosses 0, indicating not statistically significant, although the p-value of the ARR was not calculated. Albeit the sample's small size to confer a strong statistical relevance, the modest achievement in fistula closure and its rate was worthwhile. But if we include the vedolizumab non-responders in our calculation, the fistula closure rate at week 52 becomes 12/(39 + 85) = 9.67% (≈ 10%). It would decrease further if we include only Crohn's fistula. The question arises would it be worthwhile to supplement the patient of perianal Crohn's fistula with vedolizumab considering the low response rate (< 9.67%) and severe adverse reaction (24.4%).
To sum up, among biologicals, infliximab is the only therapy that has a statistically significant effect on the healing rate of perianal Crohn's fistula compared to placebo, but it also requires a prolonged maintenance dose after induction therapy to maintain the remission. The efficacy of adalimumab and certolizumab pegol for the healing or closure of perianal fistula closure has not been established in the study since the statistical significance is lacking compared to placebo. The efficacy of vedolizumab, albeit significantly lower but statistically significant for the remission of the perianal Crohn's fistula, is counterbalanced by the severe adverse reaction associated with its use. Therefore the use of either therapy for the sole purpose of fistula healing or closure in case of Crohn's disease would be futile.

Setons and Biologicals
The reported re-intervention rates in the case of perianal Crohn's fistula with seton drainage were 10-20% The primary outcome measure was the fistula-related re-intervention rate at 1.5 years. It was determined that the seton treatment was associated with the highest re-intervention rate (10/15, versus 6/15 anti-TNF and 3/14 surgical closure patients, p = 0.02), while the re-intervention rates in the PISA cohort were 42%, 48%, and 44%. No substantial differences in PDAI and quality of life between the three treatment groups were observed. It was found that the seton treatment was associated with the highest re-intervention rate (10/15, versus 6/15 anti-TNF and 3/14 surgical closure patients, p = 0.02). It is worth mentioning that chronic seton treatment's inferiority was not observed for any outcome measure in the PISA cohort [10]. The limitation with this study is that the study included a small sample size in the randomized group (44 cases), hence the low power to confer strong statistical relevance, seven patients in Group 1a also received the anti-TNF therapy, four patients in Group 1b also received the additional seton. The sample size needed to detect the statistically significant difference in re-intervention rate with two-sided chi-squared testing equals 42 patients per group, or 126 patients overall (alpha 0.05, power 80% and 5% dropout rate), while the patients in each group did not full-fill the criteria [47] (as suggested by the authors in their first publication). This intermingling of the therapy within the group on a small sample blurred the picture of the re-intervention rate, hence unable to draw firm conclusions ( Table 5).  In the multivariate analysis, none of the studied variables (including concomitant immunosuppressive therapy) correlated with Infliximab's efficacy.

Conclusions
To conclude, chronic seton therapy should be the primary approach, especially if the patient has a perianal abscess, supported by the low incidence of reintervention, recurrent abscess formation, side-branching of the fistulous tract, preservation of the fistulous tract's patency, and cost-effectiveness. The significant disadvantages of seton therapy are the discomfort and the time it takes to achieve stability.
Among the biologicals, infliximab is the only therapy which has a statistically significant effect on the healing rate of perianal Crohn's fistula compared to placebo, but the major disadvantage associated with anti-TNF as sole therapy is high re-intervention rate, prolong maintenance therapy, high recurrence rate, and severe side effects. The efficacy of adalimumab and certolizumab pegol for the healing or closure of the perianal fistula closure has not been established in the clinical trials since the statistical significance is lacking compared to placebo. The efficacy of vedolizumab, albeit significantly lower but statistically significant for the remission of the perianal Crohn's fistula, is counterbalanced by the severe adverse reaction associated with its use. Therefore the use of either therapy for the sole purpose of fistula healing or closure in case of Crohn's disease would be futile.
We hypothesize that the two aspects should be addressed concomitantly to increase the rate of fistula closure. First, the seton should be used as initial therapy to maintain the tract's patency to allow the abscess's drainage and minimize the intestinal flora colonization within the tract mucosa, thereby leukocytic infiltration and propagation of inflammation within the tract. The absorbable seton could be used instead of non-absorbable. The second aspect that has to be considered is that we should target the initial stimulation of the Th1/Th17 mediated hypersensitivity instead of a factor/cytokine involved in the mediation of the inflammation. Although the unknown antigen triggering such hypersensitivity is not clear, we could target the RORγ-T (transcription factor involved in activation of Th17 cells) and the T-bet (transcription factor involved in activation of Th17 cells) within the GI mucosa by a novel target immune therapy.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Intellectual property info: All the images included in the review article. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.