Management of fourth degree obstetric perineal tear without colostomy using non - stimulated gracilis - our experience over eleven years.

BACKGROUND: Although gracilis muscle transposition for faecal incontinence has been well-described method, its literature for use in obstetric perineal tear without colostomy is sparse. In this study, we have tried to analyse its use in fourth-degree obstetric perineal tears. PATIENTS AND METHODS: A total of 30 patients with recto-vaginal fistula with faecal incontinence secondary to obstetric perineal tear were retrospectively studied between February 2003 and May 2014. The recto-vaginal fistula was explored, dissected and identification of sphincters was done using muscle stimulator. Fistula closure was done followed by sphincter repair, vaginal tightening procedure and single gracilis transposition. None of the patients had covering colostomy. Faecal incontinence was assessed pre- and post-operatively by digital rectal examination (single examiner), Park's score and Corman's score in all cases and using barium hold and transperineal ultrasonography, manometric studies in a few cases. The outcome was measured at an average follow-up of 8.8 months (7-24 months).
RESULTS: As per Park's score 26 patients had Grade I continence, two had Grade II and two patients had Grade III continence. Corman's score improved from fair to excellent in 26 patients. The patients in whom manometry was performed showed a remarkable rise in both resting and squeeze pressures. Two patients developed post-operative infections in upper 1/3 thigh incision site and three patients at gluteal region scar site.
CONCLUSION: Satisfactory continence following gracilis muscle could be achieved.

INTRODUCTION

Reconstruction of the anal sphincter was first described in 1946 by Pickrell et al.,[1] reporting their results in four children with faecal incontinence, with a follow-up study of further cases reported shortly thereafter.[2] The subsequent literature, while useful in defining the surgical technique, is mostly concerned with faecal incontinence, which is not obstetric in origin. The first reference to the use of this technique in obstetric practice was not until 1979,[3] although a defect of the anal sphincter following vaginal delivery is the most common cause of faecal incontinence in women.[4] The incidence of anal sphincter tears at delivery varies from 0.6% to 6%[45] with nearly 50% of women developing anal incontinence after primary repair.[45678] In one study, all patients with anal incontinence following primary repair had persistent anal sphincter defects.[4] However, 35–41% of all primipara had identifiable sphincter defects at follow-up investigation by endosonography,[89] although most of these did not affect anal function.[9] Analysis of the long-term effects of anorectal function after third-degree obstetric tears as a result of obstetric injury suggested that anal function deteriorates further over time and with subsequent vaginal deliveries.[10]

Use of the gracilis muscle to correct obstetric perineal tears is reported to have good results.[11121314151617181920]

This study reports the results of treatment of faecal incontinence of obstetric origin by exploration of the recto-vaginal fistula, reconstitution of the posterior vaginal wall and the anterior wall of the ano-rectum, using a single non-stimulated gracilis reconstruction of the anal sphincter without colostomy in thirty female patients.

PATIENTS AND METHODS

Between February 2003 to May 2014, thirty female patients with an age range of 21–59 years (mean: 36.7) with complete posterior vaginal tears into the rectum and faecal incontinence as a result of obstetrical injury were treated by exploration of the recto-vaginal fistula, fistula closure, sphincter repair, vaginal tightening procedure and single gracilis reconstruction of the anal sphincter by gracilis transposition. The intention was to recreate the sphincters, separate the vaginal mucosa from the anal wall and provide soft tissue interposition using gracilis muscle. Patients had an average follow-up of 8.8 months (3–24 months). The patients underwent thorough preoperative counselling.

All the patients were assessed preoperatively by Park's score[21] and Corman's score [Table 1]. Digital rectal examination (DRE) was performed by the author pre- and post-operatively in all the patients. The patients were examined in left lateral and lithotomy position [Figure 1]. Anal sphincter tone was assessed at rest, on squeeze and on thigh adduction[22] [Table 2].

Table 1

Pre- and post-operative scores

Figure 1

Clinical presentations. (a) Anorectal fistula, (b) complete tear, (c) gaping anus, (d) anal scar with faecal matter

Table 2

Anal sphincter tone assessment with DRE

Transperineal ultrasonography and barium enema was done pre-operatively in all patients. Transperineal ultrasound was performed to demonstrate the anal sphincter thickness[10] [Table 3]. Barium enema was done using a thin paste of 50% barium prepared in 300 ml normal saline. The patient was advised rotatory movements in supine position on table and walking for 20–30 m. Lateral and supine films were taken with a marker at anal verge[18] [Table 4]. Barium holding time was noted in the patients.

Table 3

Transperineal ultrasonography for thickness in mm with follow-up in months

Table 4

Barium hold enema

Six patients underwent rectal manometry to assess resting and squeeze pressures [Table 5].

Table 5

Manometry

Surgical technique

Bowel preparation was done the day before the surgery. Under combined spinal and epidural anaesthesia, with a urinary catheter in situ, the patient was placed in the exaggerated lithotomy position with head low. A line joining the adductor tightness/tubercle and the midpoint of the knee joint line was marked. Two fingerbreadths below and parallel to the above line three incisions were marked. The fourth incision was marked on the upper third leg encircling the tibial tuberosity [Figure 2]. After infiltration, single gracilis was harvested along its length to include 1–2 cm of adjoining tibial periosteum beyond the insertion [Figure 3]. The harvested gracilis was tunnelled into the perineum through a subcutaneous tunnel, the pivot point being the pedicle with a cm of fat around it.

Figure 2

Markings

Figure 3

Harvest of gracilis with distal periosteum

Double opposing skin flaps were marked at the junction of vaginal and anal mucosae [Figure 4].

Figure 4

Perineal markings

Meticulous dissection was performed to separate the anorectal and vaginal mucosae, keeping close to the latter. Once the dissection reached the virgin tissue above the recto-vaginal tear, blunt finger separation of the vagina and rectum was carried out up to 8–9 cm, to the full length of the operator's index finger [Figure 5 and Video 1]. Two horizontal curvilinear incisions were marked on the buttocks 1.5–2.0 cm posterior to the anus and overlying the ischial tuberosities. Tunnels were made superior to the ano-coccygeal raphe by blunt finger dissection from these incisions. Tunnels were also made on either side of anal canal [Video 2]. The posterior vaginal mucosa was trimmed judiciously [Figure 6]. The anal fistula was then repaired with interrupted 3/0 Vicryl sutures. The anorectal sphincter tissue was identified and confirmed by use of a muscle stimulator on either side of the tear and repaired end to end using interrupted 2/0 Vicryl. The gracilis muscle was then passed twice in a clockwise direction so that bulk of the muscle was between the critical areas of fistula. A periosteal stitch with no.1 prolene was taken through ischial tuberosity under direct vision using an illuminated retractor. This prolene stitch was taken through the contralateral ischial tuberosity in 26 patients. The donor's thigh was taken out from lithotomy position and held in adduction. The prolene suture of ischial periosteal stitch was weaved through the gracilis tendon and tension adjustment was done, in adduction by assessing the anal tone [Figure 7 and Video 3].

Figure 5

Dissection of anorectal and vaginal mucosa

Figure 6

Trimming and repair of vaginal mucosa

Figure 7

Tension adjustment

None of the patients in this series were subjected to the colostomy.

Post-operative management

The patients were mobilised the following day with instructions to avoid extreme abduction of the thighs. Liquid diet was given in the first 48 h. Soft diet was advised for 5 days, regular diet thereafter. Stool softener was started from 5th day and continued for 3 months.

From 5th day onwards, the patients were instructed to do pelvic floor strengthening exercises and thigh adduction exercises 5 times every hour (anal sphincter contracted for 10 s, and subsequently thigh adducted for 10 s). Confirmation regarding the proper execution of these exercises was carried out by rectal examination by the operating surgeon on the 7th day. Intercourse was not advised for 3 months and squatting was not permitted for 3 months. No electrical stimulation of the muscle was done post-operatively.

Patients were followed up every week for a month, and once a month for 3 months. At each follow-up, patients were questioned about (i) faecal continence (ii) perineal/vaginal soiling and (iii) ability to hold and duration. Per rectal examination was carried out as previously described. Barium hold enema and transperineal ultrasound were done at 3 months of follow-up.

RESULTS

In this series 26 of the thirty patients had a satisfactory outcome and were a continent at an average follow-up of 8.8 months. The average age of the patients in the study was 36 years. Of the four patients who had unsatisfactory outcome average age was 39.7 years. The post-delivery duration in this series was ranging from 1.5 to 35 years with an average of 12.8 years.

In 26 patients, the pre-operative Park's score Grade IV improved to Grade I post-operatively. Two patients had Grade III and 2 patients Grade II Park's score post-operatively. Corman's score in 26 patients also improved from fair to excellent.

The DRE in 26 patients was noted to shift from lax and non-gripping to normal and gripping, at rest and squeeze, respectively. The distiguising finding on adduction of the thigh was very distinct constricting feel to the examining finger. We believe this was due to the contraction of the gracilis sling. It was noted that the feel to the examining finger at rest, squeeze and on adduction was weak in patients with a poor score.

The transperineal ultrasound, which measured thickness of external anal sphincter ranged from 1.3 to 3.2 mm (average 3.1 mm) preoperatively. Post-operatively, it was done in 11 patients only due to socioeconomic reasons and the range was from 4 to 11 mm (average 10.1 mm).

The barium enema hold demonstrated leak in the majority (22 patients) who could not hold at all. Remaining eight patients could hold the enaema for 2–3 min. Post-operatively, it was done in ten patients only due to socioeconomic reasons and the holding time was notably improved between 5 and 12 min.

Manometric studies showed significant improvement in both resting and squeeze pressures [Charts 1 and 2]. The preoperative resting pressure average was 27.83 mm of Hg which improved to 46.1 mm of Hg post-operatively. Furthermore, the pre-operative squeeze pressure average of 38.5 mm of Hg escalated to 90.5 mm of Hg post-operatively.

Chart 1

Resting pressure

Chart 2

Squeeze pressure

Complications and unfavourable outcomes

Scar hyperpigmentation was the commonest finding (26 patients) followed by scar hypertrophy (twenty patients). There was a tiny residual fistula in two patients with intermittent soiling. Gluteal wound infection was noted in three patients and thigh infection in two patients. One patient had deep vein thrombosis. Three patients complained of perineal pain after surgery [Table 6].

Table 6

Complications

DISCUSSION

The results of gracilis muscle transposition for anal incontinence using Pickrell's original operation[12] have been conflicting, and the mechanism of action of gracilis transposition in achieving continence is uncertain. The muscle does not seem to act as a dynamic sphincter but does offer passive resistance to outflow, and it has been suggested that the muscle acts as no more than an inert sling.[2324] Present series deals with cases of the recto-vaginal fistula with incontinence wherein the gracilis has not only acted as a vascularised interpositional tissue between the repaired fistulae but also a distinct contracting ring being felt to the finger on DRE on adduction of the thigh. There can be difficulty in re-establishing the physiological length-tension relationship of the muscle and patients are reported as having to perform awkward movements to achieve continence.[23] Contrary to this, it was noted in our series that none of the patients had to resort to any awkward position or movement to maintain continence or to defecate. Striated muscle is unable to maintain a contraction for a prolonged period[18] and this problem has been addressed by implantation of a neuromuscular stimulator after gracilis transposition.[2526] The literature on gracilis reconstruction for faecal incontinence is dominated by the use of this technique in children, after trauma and after bowel surgery. In our series, we have not used stimulator for gracilis and all these cases are post-delivery recto-vaginal fistulae.

Discussion of post-obstetric cases of faecal incontinence in the literature remains small and mostly includes reports of sporadic cases in larger series of cases of faecal incontinence resulting from other causes. Despite an exhaustive search of the literature, we have managed to find ten cases recorded in which gracilis was used in the reconstruction.[81118192027] This study reports thirty women with post-obstetric faecal incontinence as a result of recto-vaginal tears who were treated with a single gracilis sling with a success rate of 86% (26 patients of thirty having a satisfactory outcome.) relieved entirely of faecal incontinence at all times. All the thirty patients had continence to solid stools. Further, of the four patients who had unsatisfactory outcome two patients had a considerable reduction of this problem with incontinence only to flatus. In this series, we noted that the interposed gracilis conferred adequate vascularised soft tissue interphase over the repaired fistula. The appropriate tension adjustment in adducted thigh by hitching the tendon to the ischial tuberosity gave a unique constricting feel on DRE, post-operatively. Rasmussen (2003) identified poorer results among patients older than 40-year-old when compared to those of younger patients. This study does not show this, as we had nine patients who were above forty of which only one had suboptimal outcome.

The need for a defunctioning colostomy as a preliminary or concurrent step, with the treatment of anal incontinence surgically, is debated. Successful results without faecal diversion have been reported after direct repair, local procedures and gracilis transposition and most surgeons would now agree that a colostomy is not required for treatment of a straightforward anal sphincter injury.[2829303132] This study includes thirty reconstructions after obstetrical injury carried out successfully without colostomy, suggesting that this practice is not necessary.

Various objective methods of assessment of faecal continence have been used[11202333343536373839] and reviewed.[20] These include DRE, barium enema, anorectal manometry, endoanal ultrasound, electromyography, pudendal nerve terminal motor latency and defaecography and transit time of the colon. In this study, DRE, barium enaema and transperineal ultrasound have been used for assessment. Rectal manometry was done preoperatively and post-operatively in only a six cases, as it was previously unavailable. The manometric studies revealed the contraction of the gracilis as a distinct band. It was observed that the duration of squeeze pressures with gracilis contraction was twice that after only sphincter contraction.

The author feels that the use of gracilis in successfully treating these complex defects with an extreme degree of incontinence and psychological problems can be a boon to those faced with this problem.

Videos Available on: www.ijps.org

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.