BACKGROUND: Fistulotomy plus primary sphincteroplasty for complex anal fistulas is regarded with scepticism, mainly because of the risk of postoperative incontinence.

OBJECTIVES: The aim of this study was to evaluate safety and effectiveness of this technique in medium-term follow up and to identify potential predictive factors of success and postoperative continence impairment.

DESIGN AND SETTING: This was a prospective observational study conducted at a tertiary care university hospital in Italy.

PATIENTS: A total of 72 patients with complex anal fistula of cryptoglandular origin underwent fistulotomy and end-to-end primary sphincteroplasty; patients were followed up at 1 week, 1 and 3 months, 1 year, and were invited to participate in a recent follow-up session.

MAIN OUTCOME MEASURES: Success regarding healing of the fistula was assessed with 3-dimensional endoanal ultrasound and clinical evaluation. Continence status was evaluated using the Cleveland Clinic fecal incontinence score and by patient report of post-defecation soiling.

RESULTS: Of the 72 patients, 12 (16.7%) had recurrent fistulas and 29 patients (40.3%) had undergone seton drainage before definitive surgery. At a mean follow-up of 29.4 (SD, 23.7; range, 6–91 months, the success rate of treatment was 95.8% (69 patients). Fistula recurrence was observed in 3 patients at a mean of 17.3 (SD, 10.3; range, 6–26) months of follow-up. Cleveland Clinic fecal incontinence score did not change significantly (p = 0.16). Eight patients (11.6% of those with no baseline incontinence) reported de novo postdefecation soiling. None of the investigated factors was a significant predictor of success. Patients with recurrent fistula after previous fistula surgery had a 5-fold increased probability of having impaired continence (relative risk = 5.00, 95% CI, 1.45–17.27, p = 0.02).

LIMITATIONS: The study was limited by potential single-institution bias, lack of anorectal manometry, and lack of quality of life assessment.

CONCLUSIONS: Fistulotomy with end-to-end primary sphincteroplasty can be considered to be an effective therapeutic option for the treatment of complex anal fistulas, with low morbidity, a high rate of success even at long-term follow-up, and a very low rate of postoperative major fecal incontinence, although minor impairment of continence (postdefecation soiling) may occur. Caution should be used in selecting patients with a history of recurrent fistula and fecal incontinence.
 

Surgical management of complex anal fistula should take 2 major endpoints into consideration: healing of the fistula and preservation of sphincter function and stool continence.^1–3 None of the various surgical procedures presently used represents an ideal standard—not even less invasive procedures such as ligation of the intersphincteric fistula tract (LIFT),⁴,⁵ anal fistula plug placement,^6–9 or the endorectal advancement flap.¹⁰ Whereas successful treatment of the fistula often involves an increase in postoperative fecal incontinence, sphincter-sparing techniques are often disappointing because of the risk of recurrence.² Cutting setons, fistulotomy, and fistulectomy are suitable only for low, simple anal fistulas because of the high risk of postoperative continence impairment, which may result in major fecal incontinence or, more often, in soiling after defecation.¹¹,¹² For this reason, Westerterp et al¹³ described the position of the surgeon choosing a surgical treatment for a high anal fistula as being like that of Ulysses in Homer’s Odyssey—“between Scylla and Charybdis.”

Although Parkash et al¹⁴ first described fistulotomy plus end-to-end primary sphincteroplasty for the treatment of complex anal fistulas in 1985, this procedure still tends to be regarded with scepticism. The few studies published on this topic have had methodological limitations or only short-term follow up. Moreover, the indications for this surgical approach have not been clearly defined.

The aim of this prospective study was to evaluate the safety and effectiveness of fistulotomy plus end-to-end primary sphincteroplasty with medium-term follow up and to identify factors potentially predictive of success or of major or minor postoperative fecal incontinence.
 

PATIENTS AND METHODS

Patients

Between July 2004 and November 2011 a selected series of patients affected by complex anal fistula of cryptoglandular origin were included in a prospective study to evaluate treatment with fistulotomy plus end-to-end primary sphincteroplasty (FIPS) at the Department of Surgical Sciences, Catholic University, Rome. The study was approved by the institutional ethics committee, and all patients gave written informed consent.

Patients were included if they had a high transsphincteric fistula (tract crossing more than 30% of the external anal sphincter) or if they had a low transsphincteric fistula and were at risk for postoperative fecal incontinence (anterior fistula in women, recurrent fistula, or history of fecal incontinence). Patients affected by Crohn’s disease, intersphincteric or low transsphincteric or suprasphincteric fistulas were excluded. Patients affected by acute perianal sepsis were treated by using a loose seton for drainage before definitive surgery and were included in the study after the complete resolution of the acute sepsis.
 

Baseline Assessments

All patients were evaluated clinically, and the following data were collected for each patient: demographic characteristics, current/recent smoking behavior, fistula type, seton placement, duration of seton drainage, previous anorectal surgery, recurrent fistula, and continence status.

Fistulas were evaluated with a 3-dimensional endoanal ultrasound (3D-EAUS) imaging device (model 2202; BK Medical, Herlev, Denmark) equipped with a 360° rotating endoprobe (model 2050; BK Medical). The internal opening, primary tract, secondary tract, abscesses, and horseshoe extension were evaluated as previously described.¹⁵ Along with clinical evaluation, 3D-EAUS was also used to evaluate resolution of perianal sepsis after seton drainage.

Continence status was evaluated using the Cleveland Clinic Florida (CCF) fecal incontinence score¹⁶ and patient report of postdefecation soiling.
 

Surgical Procedure

Two enemas were given the evening before the surgical procedure as bowel preparation; no antibiotics were given either as prophylaxis or postoperatively. All procedures were performed under general anesthesia in the lithotomy position and conducted by the same experienced colorectal surgeon (C.R.). Patients were treated in a day surgery setting or with a postoperative hospital stay.

Setons, if present, were removed. The internal opening and the fistula tract were identified by hydrogen peroxide injection and by fistula probes (Fig. 1A). A complete fistulotomy of the primary tract was performed from the external to the internal opening, with section of both anal sphincters (Fig. 1B). The primary fistulous tract peripheral to the external sphincter was then excised using a scalpel; also, any secondary tracts or residual cavities were excised. The primary tract passing through the sphincters was curetted in order to remove any granulation tissue. The internal opening was removed at the level of the mucosal surface. Thereafter, an end-to-end primary sphincteroplasty was performed using a series of 2 to 4 interrupted 2-0 Vicryl (Ethicon Endo-Surgery, Cincinnati, OH) stitches encompassing both sphincter stumps and the entire wall of the fistula tract (Fig. 1C). Finally, anal mucosa and submucosa were closed with a 3-0 Vicryl (Ethicon Endo-Surgery) continuous suture. The external part of the perianal wound was left open to permit serum drainage (Fig. 1D).

During the postoperative period, stool softeners and analgesics were given as needed. Patients were advised to avoid strong physical activity during the first 2 postoperative weeks.
 

Follow-up and Evaluation of Success

Follow-up visits were scheduled at 1 and 2 weeks and 1, 3, 6, and 12 months after the operation. In addition, all patients were recalled for a follow-up session in November 2011, at which all patients were examined by clinical and physical evaluation, and a 3D-EAUS examination was performed.

Success was defined as the absence of drainage or abscess formation, closure of the external opening, complete healing of the wound, and absence of sphincter dehiscence at the last follow-up session.
 

Statistical Analysis

To assess the factors potentially predictive of success or postoperative continence disturbances, the following baseline variables were considered: age, sex, smoking, type of fistula tract, recurrent fistula, previous anorectal surgery, use of seton drainage, and postdefecation soiling.

Continuous data were analyzed as means (with SD and range) and were compared using the paired samples t test; categorical data were analyzed as frequencies and percentages and were compared using the χ² test, the Fisher exact test, or the McNemar test as necessary. To assess factors potentially predictive of success and postoperative impairment of continence, a relative risk ratio (RR) was calculated for each characteristic by dividing the probability of success or impairment of continence in patients with the characteristic by the corresponding probability in patients without the characteristic. A p value of less than 0.05 was considered to be statistically significant. For the analysis of the literature, weighted averages were calculated as follows: ∑(single study average × study cohort size)_1,2…n / pooled cohort size. This was done to minimize the effect of the different study cohort sizes on the calculated averages and to provide an overall value for the outcome measures evaluated. Analyses were carried out with SPSS version 17.0 software for Windows (IBM-SPSS, Chicago, IL, USA).
 

RESULTS

Baseline and Operative Data

A total of 72 patients were enrolled in the study. Their baseline characteristics are presented in Table 1. As shown, the primary fistula tract was high anterior transsphincteric in 29 patients (40.3%) and high posterior transsphincteric in 43 patients (59.7%). In 29 patients who underwent loose seton drainage before the FIPS procedure, the mean duration of drainage until resolution was 4.6 months. A total of 26 patients (36.1%) had undergone previous anorectal surgery: 12 patients (16.7%) were being treated for recurrence after previous fistula surgery, and 14 patients had undergone other procedures.

As also shown in Table 1, the overall preoperative CCF fecal incontinence score at baseline was 0.1. A total of 69 patients (95.8%) were fully continent with no sphincter lesions demonstrated on 3D-EAUS. Three patients (4.2%) had varying degrees of incontinence: all 3 were incontinent to gas, with 2 (2.8%) also reporting postdefecation soiling. One of these patients was a woman affected by a recurrent fistula after a fistulotomy and following a hemorrhoidectomy and a lateral internal sphincterotomy for anal fissure (preoperative CCF score = 2). The second was a woman with a history of difficult vaginal delivery (CCF score = 3). The third was a man with an idiopathic continence impairment (CCF score = 4). The 2 women had a lesion of the anal sphincters, whereas the man had no sphincter lesion.

The mean duration of the surgical procedure was 20.9 (SD, 2.8; range, 12–26) minutes, and the mean length of postoperative stay was 1.3 (SD, 0.8; range, 0–4) days. None of the patients experienced any degree of local or systemic sepsis or fever. In 1 male patient (1.4%), a sphincter dehiscence occurred 5 days after the surgery: he underwent a second FIPS procedure 1 month after the first procedure.
 

Outcome

No patient was lost to follow-up. Outcome data are presented in Table 2. The overall success rate as of the last follow-up visit (November 2011) was 95.8%; 3 patients (4.2%) had recurrence within a mean of 17.3 months after the procedure. The first recurrence was in a male patient with a high posterior transsphincteric fistula who had an abscess formation 6 months after the operation. At clinical evaluation and 3D-EAUS, a low posterior intersphincteric fistula was detected at the site of the previous surgical scar. He underwent a second FIPS procedure, and the fistula healed without continence impairment. Another recurrence was observed 26 months after surgery in a female patient who was treated for a high posterior transsphincteric fistula recurrent to multiple previous surgical procedures. The failure of FIPS was treated by placing a loose seton and then performing an endorectal advancement flap. Recurrence also took place in a male patient 20 months after FIPS was performed for a high anterior transsphincteric fistula: he was treated with an endorectal advancement flap with complete healing.

The mean overall CCF incontinence score in the 72 patients did not vary significantly, changing from a preoperative 0.1 (SD, 0.6; range, 0–4) to 0.2 (SD, 1.1; range, 0–8) at last follow-up (p = 0.16). As shown in Table 2, postdefecation soiling was reported in 10 patients (13.9%) at follow-up, representing a significant increase from the 2 patients (2.8%) with postdefecation soiling at baseline (p = 0.008). Among the 10 patients with postdefecation soiling at follow-up, de novo soiling occurred in 8 patients (11.6% of the 69 patients who had full continence and no sphincter defects before surgery). The 2 women with sphincter defects who had incontinence to gas and postdefecation soiling at baseline continued to have postdefecation soiling at follow-up. In addition, status worsened to major fecal incontinence in 1 patient (1.4%): the male patient who was incontinent only to gas before surgery and had no sphincter lesion at baseline became incontinent also to liquid stools, although he never reported having postdefecation soiling.
 

Predictive Factors

None of the factors hypothesized to e predictive of success in this study was significantly related to outcome (Table 3). Patients with placement of a seton for drainage before surgery appeared to have a lower probability of success (ie, higher incidence of recurrence) than patients without seton placement, but this effect was not significant (RR, 0.90; 95% CI, 0.79–1.01; p = 0.06).

When factors potentially predictive of de novo onset of postdefecation soiling were evaluated, patients with recurrent fistula after previous fistula surgery had a 5-fold increased probability of having impaired continence (RR, 5.00; 95% CI, 1.45–17.27; p = 0.02). None of the remaining factors were significant (Table 4).


TABLE 4. Potential predictors of de novo postdefecation soiling
 

DISCUSSION

In this prospective study of 72 patients with complex anal fistulas of cryptoglandular origin treated with fistulotomy plus primary sphincteroplasty and followed for a mean of 29.4 months, the long-term success rate was 95.8% for healing of the fistula. De novo impairment of fecal continence occurred in 11.6% of those with no previous impairment and major fecal incontinence occurred in 1.4% of the total. The presence of recurrent fistula after previous fistula surgery significantly increased the risk of postdefecation soiling during follow-up, but none of the other factors hypothesized to be predictive of success or postdefecation soiling had significant effects.

The patients in the present study comprised a homogeneous group with complex fistulas who came to the operation without acute perianal sepsis, which had been resolved either spontaneously or by a draining seton. The surgical procedure consisted of fistulotomy, “partial fistulectomy” (limited to the fistula tract peripheral to the external anal sphincter), and end-to-end primary sphincteroplasty with mucosal-submucosal suture. In particular, the excision of the portion of fistula tract passing through the sphincters was avoided in order to minimize further possible damage of sphincters that may occur with a “total fistulectomy.” Thereafter, the interrupted absorbable stitches, which encompassed both sphincter stumps and the entire wall of the fistula tract, aimed to both restore sphincter continuity and obliterate any dead space within the remaining fistula tract. Even if the muscle division and primary suture could theoretically compromise its baseline strength, and therefore its physiology, data from previous series¹⁷,¹⁸ and the present study indicate that FIPS would have only a slight negative impact in this regard.

Clinical signs of fecal incontinence were evaluated in detail in the present series. However, the lack of anorectal manometry data represents a real limitation. In this regard, other series noted that anal pressures did not change significantly after this surgical procedure, indicating that anal pressures were not related to good clinical and functional outcome.¹⁷,¹⁸ Furthermore, the current study was conducted at only a single institution, which may not be generally representative. Nevertheless, in this series, FIPS appeared easy and safe. Total reconstruction was obtained in almost all patients (except 1 patient who had a dehiscence of the sphincter reconstruction and required a further sphincter suture, which was successful). Wound healing was rapid and a very good appearance of the anal verge and perianal area was obtained. Considering the possible impacts of a long wound-healing period (requiring special management) and anal deformities (with consequent symptoms related to skin alterations) secondary to both “traditional” fistulotomy and fistulectomy, these details are not negligible.

FIPS has also demonstrated a low risk of morbidity in other published series (Table 5).^18–22 In studies that adopted an overlapping sphincteroplasty,¹⁷,²³,²⁴ no dehiscence was reported (Table 5). However, it should be taken into account that, when performed to treat fecal incontinence due to anal fistula surgery, overlapping sphincteroplasty can cause bleeding or infective complications in up to 25% of the patients.²⁵

Recurrence can take place at various times after the FIPS procedure, sometimes after several years, as demonstrated in the present series, thus confirming the difficulty of definitively assessing recurrence rates after FIPS as well as other procedures.²⁶ It is worth noting that management of the recurrences after FIPS were conservative and successful in all 3 patients (re-do FIPS in one, endorectal advancement flap in 2 patients).

Regarding postoperative incontinence, risks have been shown to vary from 18% to 64% for “lay-open” techniques.²⁷ These data may be the consequence of sphincter division significantly altering the continence mechanisms. However, even when sphincter-saving procedures are used (mainly endorectal advancement flap), postoperative incontinence may vary from 0% to 42%.²⁷ In an attempt to reduce the continence impairment associated with fistulotomy or fistulectomy, Parkash et al¹⁴ introduced the concept of reconstruction of the sphincters immediately after their section. In a retrospective series, these authors showed a high success rate (97.5%) after follow-up periods ranging from 6 to 60 months, with postoperative continence impairment in only 3 of 82 patients (3.7%). However, almost all patients (118 out of 120, 98.3%) had simple anal fistulas. Since then, several studies have shown the results of this technique also for complex anal fistulas, with healing rates varying from 85.7% to 100% (Table 5).^17–24,²⁸,29 On the basis of these data, in 2011 the first German S3 guideline for the treatment of cryptoglandular anal fistulas considered “fistula excision with direct sphincter reconstruction” to be a therapeutic option.³⁰

The 11.6% rate of postoperative de novo minor fecal incontinence (8 of the 69 patients fully continent before surgery) and 1.4% rate of more severe fecal incontinence found in the current study are consistent with the available literature on FIPS, which shows that minor impairment of continence (mainly postdefecation soiling or incontinence to gas) occur in 0% to 21.7% of patients.^17–24,²⁸ A few studies have reported postoperative major fecal incontinence, with rates ranging from 0% to 10.9% (Table 5).¹⁷,¹⁸,^20–24,²⁸

In obstetric anal sphincter injury, primary sphincteroplasty has been used with good results in short- and medium-term follow up.³¹,³² Primary sphincter reconstruction provides a reduction of the overall cost of treatment compared with delayed sphincteroplasty.³³ The results of both primary and delayed sphincteroplasty for obstetric trauma may deteriorate with time.³² Although most primary sphincteroplasties are performed by obstetricians, an improvement of the outcome has been described when the surgical procedure is performed by a colorectal surgeon.³⁴ Usually, all primary sphincter repairs after anal fistulotomy are performed by a colorectal surgeon.

Although none of the potential predictive factors of success vs failure included in the present study was found to have a significant effect, patients who require loose seton drainage before FIPS may have had a higher risk of recurrence (ie, a lower relative probability of success). This factor has also been investigated in other studies but never found to be significantly predictive.³⁵,³⁶ The use of seton drainage, necessary in case of an extensive abscess as first-line treatment of acute perianal sepsis, may reflect the greater severity of such fistulas. In fact, other studies have found fistula complexity (type, horseshoe extension, and nonidentified internal opening) to be a statistically significant predictor of recurrence.³⁵,³⁷

Several studies have investigated whether previous surgery for recurrent fistula might be a risk factor for postoperative incontinence but did not demonstrate a relationship.³⁵,³⁷,³⁸ In our study, preoperative detection of sphincter defects was not related to worsening of continence status after FIPS. However, patients undergoing FIPS for fistula recurrence after previous fistula surgery had a risk 5 times greater than that of patients without recurrence. This seems to contrast with other published studies on FIPS, in which the authors considered such patients to be ideal candidates.¹⁸,²³ However, in a prospective study by Perez et al,¹⁷ the CCF incontinence score significantly improved after surgery both for patients with recurrence and for those with incontinence. Similar results were achieved by Roig et al in 2010.¹⁸ Because of these controversial data, the indication for performing FIPS in patients with recurrent fistula or those with preexisting fecal incontinence needs to be elucidated in further studies.

CONCLUSIONS

FIPS can be considered to be an effective therapeutic option for the treatment of complex anal fistulas, with low morbidity, a high rate of success even at long-term follow-up, and a very low rate of postoperative major fecal incontinence, although minor impairment of continence (postdefecation soiling) may occur. Caution should be used in selecting patients with a history of recurrent fistula and fecal incontinence.