Ligation of Intersphincteric Fistula Tract Versus Rectal Advancement Flap in the Treatment of Complex Anal Fistula

Material and method

Design:

Phase III, single centre, randomised, open-label, parallel group trial controlled with active comparator, with blind evaluation of results by a non-commercial third party. This study was conducted in 72 patients who had undergone complex anal fistula surgery and met the inclusion criteria.

The study was conducted by the Colo proctology Unit of the General and Digestive Surgery Department of Hospital General Universitario Reina Sofia, Murcia, Spain, from February 2013 to June 2016, and approved by the hospital's Independent Ethics Committee (IEC). The study was conducted pursuant to the principles of the Declaration of Helsinki (Seoul, October 2008) and the Good Clinical Practice (GP) standards of the European Economic Community's task force on the efficacy of medicinal substances (1990) and all legislation currently applicable in Spain (Royal Decree 223/2004 of 6 February).

Patients with a diagnosis or suspected diagnosis of anal fistula were referred to the Coloproctology Office. At the first visit, demographic data and a clinical history was obtained, followed by a physical examination of the anal region, with rigid anoscopy, ordering an endoanal ultrasound. At a second visit, if the diagnosis of complex anal fistula was confirmed, after providing information about the study, the patient was invited to participate, signed the consent form, and underwent the pre-operative Wexner scale. Patients were allocated a randomisation code for LIFT or RAF on the day of surgery. After surgery, controls were performed on days 7 and 14, by means of a symptoms diary to be completed by each patient, and 1, 3, 6 and 12 months after the surgery, with the Wexner scale calculated in the last control of each patient.

Complex anal fistula was defined as a fistula at risk of affecting continence involving more than 30% of the sphincteric apparatus. Infection of surgical site at or near the incision in the 30 days following surgery (28). Haematoma was defined as accumulation of blood in the surgical area, with bluish or violet colouring of the affected area; surgical wound dehiscence was defined as the spontaneous separation of the edges of the wound, in the absence of surgical site infection.

The inclusion criteria were patient diagnosed with transsphincteric (medium or high) or suprasphincteric cryptoglandular anal fistula, over 18 years of age, with no abscess at the time of surgery, who agreed to participate and signed the informed consent form, and was capable transsphincteric understanding and following study instructions. The exclusion criteria were minor patient, diagnosed with, or with a suspected diagnosis of, Crohn's disease, malignant anal tumour, tuberculosis or suppurative hidradenitis, a history of radiotherapy of the ano-perineal region, language difficulties that prevent comprehension of the study, or refusal to participate.

Each patient's data were collected in the case report form and transferred to a database for their statistical analysis by an analyst unrelated to the study, preserving their anonymity and the treatment group to which they belonged.

Study objectives:

The primary objective was to show the non-inferiority in therapeutic efficacy of LIFT versus RAF in the surgical treatment of CAF. The primary efficacy endpoint was the proportion of subjects in each treatment group without fistula recurrence 12 months after surgery.

The secondary objectives were to show the non inferiority of LIFT versus RAF in therapeutic safety:

• Evaluating functionality, both before and after surgery, in both treatment groups, through anal continence, using the Wexner scale.
• Comparing post-operative complications in both treatment groups: infection of surgical site, haematoma and surgical wound dehiscence.
• Comparing time to fistula relapse in both treatment groups, by a survival analysis.
• Analysing hospital management factors (duration of surgery and hospitalisation) in both treatment groups.

Randomisation and surgeries:

The subjects were randomised 1:1 to receive LIFT or RAF surgery. Patient randomisation was by randomisation in balanced six-cell blocks; the sequence was custodied by the Teaching Secretary, unrelated to the study. The surgeon requested each patient's randomisation code on the day of surgery. LIFT or RAF.

The patients were operated by three surgeons from the Coloproctology Unit, whose learning curve was based on unifying experimental technique (LIFT) criteria in six patients before the start of the study. All the surgeries used epidural anaesthesia.

• Ligation of the intersphincteric fistula tract (LIFT): jackknife position. Location of IFO by injecting a diluted (1/3) solution of hydrogen peroxide through the external fistula orifice (EFO). Opening of the fistula trajectory with a stylus. Curved 2.5 cm incision in the intersphincteric space until trajectory is skeletalised, ligation of internal and external sphincter with 3/0 single thread polyglecaprone 25 absorbable suture (MonocrylTM Plus®), exeresis of the trajectory between the ligations with cold scalpel. Cleansing of distal part with curette and enlargement of the external fistula orifice (EFO). Closure of the muscular intersphincteric plane with loose stitches of 3/0 polyglecaprone 25 single-thread absorbable suture (MonocrylTM Plus®). Closure of intersphincteric skin with simple plaited polyglycolic acid absorbable sutures (SSA90®).
• Rectal advancement flap (CAR): jackknife in case of anterior fistula and lithotomy in case of posterior fistula. Location of IFO by injecting a diluted (1/3) solution of hydrogen peroxide through EFO. Opening of the fistula trajectory with a stylus. Excision of the extrasphincteric component of the fistula and cleansing of sphincteric tract. Design of a flap with electric scalpel, of partial thickness including mucosa-submucosa and part of the fibres of the internal sphincter, with a broad base (twice the apex), sliding in a craniocaudal direction, without tension, to completely cover the internal part of the fistula, closure of the internal orifice with cross stitch and suture of flap with simple 3/0 polyglecaprone 25 absorbable suture (MonocrylTM Plus®).

Peri-operative care and follow-up:

All patients were advised to follow a waste-free diet 48 hours before admission, and were given an enema on the night before surgery. All patients received antibiotic prophylaxis with IV amoxicillin/clavulanic acid 2 g half an hour before the start of surgery. Six hours after surgery, liquid diet commenced, progressing to semi-bland for dinner that night, and bland on the first day after surgery. No postoperative antibiotic treatment was indicated. The ward surgeon, unrelated to the study, discharged the patient based on these criteria: afebrile, controlled pain, appropriate oral tolerance and surgical wound correct.

All patients received 12 months of follow-up, except for one who left the study after 8 months. After discharge, the patient attended for cleaning of the surgical wound on the following two Fridays, with review of the symptom diary (SD) and evaluation of possible complications by the Coloproctology Unit surgeon. The patient attended a scheduled appointment one month after surgery (1st post-op visit), where he was evaluated and data was collected in the case report form (CRF), the SD was collected and possible complications were evaluated (infection of surgical site, haematoma, wound dehiscence). New appointments were scheduled at 3 (2nd post-op visit) and 6 months (3rd post-op visit), where the patient was evaluated and the respective data were collected. The final visit (4th post-op visit) took place at 12 months, where the post-op Wexner scale was applied.

Sample size and statistical analysis:

When the study began, publications reported a relapse rate of 0-33% with rectal advancement flap, with an average of 19% (19) and we can assume that the response rate to surgical treatment was approximately 85% for RAF, while the expected response rate for LIFT was 70%, the average of the rates observed in clinical studies published to date, which ranged from 57% to 94% (19).

With these data, considering response as the absence of fistula recurrence at 12 months, and in order to obtain a power of 80% to reject the null hypothesis (Ho), that the difference between proportions p1 and p2 was lower than the non-inferiority limit, by a normal asymptotic test for proportions, unilateral for two independent samples, considering that the level of significance was 0.5, assuming that the proportion in the reference group was 85%, the proportion in the experimental group was 70%, the proportion of patients in the reference group relative to the total was 50% and the non-inferiority limit was 10% and the expected withdrawal rate was 5%, 36 patients were required for the reference group and 36 for the experimental group, making a total of 72 patients in the study.

The categorical variables were described with mean values and percentages. The continuous variables were described with mean values and standard deviation, or median values and range. The Chi-square or Fisher test was used to compare two qualitative variables, to see whether there was dependence between them. The Bonferroni correction was used for column comparisons. The Student's t-test was used for comparison of the two groups, after checking the normality and homogeneity of variances assumptions with the Kolmogorov-Smirnov and Levene tests, respectively. For the analysis of non-inferiority of the primary and categorical variables, the calculation was through the Farrington-Manning score and Miettinen-Nurminen Score tests, with a level of significance of 0.05 and a non-inferiority limit of 0.10. For the continuous variables, non-inferiority was evaluated by difference of means, with the Mann-Whitney U-test and the Wilcoxon test. All statistical comparisons used a bilateral test with a level of significance of p < 0.05.

The primary study analysis was based on the intention-to-treat population.