|Home | About | Journals | Submit | Contact Us | Français|
To assess the modifications in the technique of laparoscopic-assisted anorectal pull-through (LAARP) practiced at our institute and to analyze the postoperative outcome and associated complications.
A retrospective study was done to analyze the results for LAARP procedure done for high anorectal malformations (ARMs) from January 2001 to May 2016. A total of 68 patients had undergone LAARP, with 62 male and 6 female children. Staged procedure was done in 55 patients and one child with rectovestibular fistula of 5 months of age had a single-stage procedure. Technical modifications such as traction over the fistula which helps in identification and dissection of the puborectalis muscle, dividing the fistula without ligation, railroading with Hegar's dilators over the suction cannula which creates adequate pull-through channel, have helped save time and make the procedure simpler. The patients were followed up with clinical evaluation and continence scoring.
All the patients withstood surgery well. One patient with rectovesical fistula required conversion to gain adequate length of the distal rectum, for whom the colostomy was closed and relocated at splenic flexure level. The complications were mucosal prolapse (8 cases), anal stenosis (5), adhesive obstruction (2), distal rectal necrosis (3), and urethral diverticulum (2). The progress has been satisfactory and weight gain is adequate. 71.15% patients had good continence on follow-up.
LAARP procedure is safe for high ARMs with good continence and correctable side effects. It has completely replaced posterior sagittal anorectoplasty procedure for high anomalies in our practice.
Laparoscopically assisted anorectal pull-through (LAARP) for high anorectal malformations (ARMs) reduces the amount of posterior dissection and avoids dividing and weakening external sphincters required for accurate placement of the bowel into the muscle complex and diminishes perirectal scarring. For high ARMs, LAARP has been the procedure of choice in our department from 2000 and has replaced posterior sagittal anorectoplasty (PSARP).[2,3] It has gone through several technical modifications, which has made the procedure simpler, decreased the operative time, and improved postoperative results. The present study is intended to discuss the technical modifications of LAARP and analyze the postoperative outcome and complications associated with our technique.
This is a retrospective study analyzing LAARP performed at the Department of Pediatric Surgery, Coimbatore Medical College Hospital, Coimbatore, from January 2001 to May 2016. Out of 68 patients who underwent LAARP (62 males and 6 females) with age group ranging from 2 months to 6 years, staged procedure was done in 67 patients and primary pull-through in one child (rectovestibular fistula).
All neonates with high ARMs underwent sigmoid loop colostomy at birth, LAARP at 3-6 months of age and colostomy closure after 8 weeks.
Pressure augmented distal colostogram was done to identify the type of fistula. Rectovesical and rectoprostatic fistulas have been suitable for LAARP. After adequate bowel preparation, intravenous antibiotic cover was given an hour before surgery.
If the length of the distal bowel loop was not adequate, dismantling the colostomy to gain additional length was performed.
Under endotracheal general anesthesia with caudal block, distal loop wash was given and the contents were emptied completely before draping the abdomen.
Nasogastric decompression was done. Bladder was catheterized with silicone Foleys catheter of appropriate size and in case of difficulty it was done with the aid of a cystoscope.
The child was placed in the supine Trendelenburg position with legs widely spread (frog position) and pelvis elevated with sandbag or folded towel. The surgeon stood to the right and the assistant to the left at the head end, while the second surgeon (required at a later stage of the operation) worked from the perineum. Pneumoperitoneum was created with CO2 (8–12 mm of Hg) by closed method. The abdomen was accessed by three ports, one 5 mm umbilical port or supraumbilical port for 30° telescope and two 5 mm or 3 mm accessory working ports one at right lumbar region and another at left lumbar above the colostomy [Figure 1a]. For comfort, an additional 3 mm suprapubic port or a stay suture which would help in retraction of the bladder was done initially but given up in our later cases as we gained experience. The first step of LAARP was visualization of the anatomy and identification of any associated anomalies [Figure 1b].
Dissection was commenced at the level of the peritoneal reflection. The terminal branches of the sigmoid and superior rectal arteries were divided to gain adequate bowel length. In certain cases for want of length the inferior mesenteric artery is divided using vessel sealing devices or bipolar artery provided the marginal vessels at the colostomy site are intact. The distal rectal peritoneal attachment was released with bipolar scissor dissection or hook cautery and was continued anteriorly and laterally on the rectal wall taking care not to injure the ureter and genital structures.
Bipolar scissors was used once the dissection reached the bladder neck to avoid lateral damage to the pelvic nerves beyond which dissection was done subserosally [Figure 1c]. Anterior dissection was stopped on identifying prostate/lower end of the uterus that is usually visualized as an indentation through the posterior vaginal fornix. The distal colon was mobilized up to the fistulous connection. The higher the fistula, the easier was the dissection. Dissection was stopped at the point of coning or narrowing, which occurs just before the fistulous communication. It is always ideal to stay posterior to the fistulous site during dissection. The puborectalis sling was identified before fistula division and traction on the fistula helped in posterior midline dissection to create the pull-through channel [Figure 1d]. A small cuff of fistulous tissue was left behind to prevent injury to the vaginal or urethral wall and to avoid injury to the pelvic nerve plexus, but care is taken to avoid leaving behind a large cuff. Fistulous division was done with bipolar scissors instead of monopolar to avoid injury to the nerve plexus.[2,3]
With cephalic traction on the fistula, both bellies of pubococcygeus can easily be identified in relation to the urethra. The classic anatomic arrangement of the puborectalis, resembling a “sling-shot,” [Figure 2a] can often be appreciated. The contractility of the levator ani muscle and center of its two bellies was identified by the laparoscopic muscle stimulator. Conventional diathermy in a low setting current can be used just as effectively as a laparoscopic muscle stimulator. When there is insufficient muscle mass to clearly ascertain the pubococcygeus, the midline was identified based on the position of the distal end of the divided fistula and the urethra. Externally, the anal area of the perineum was mapped out using transcutaneous electrostimulation (muscle stimulator with 100–150 milliamps current) if not done previously. The area of maximal contraction and ventrocephalad elevation of the perineum was noted with simultaneous contraction of the puborectalis. The anterior and posterior limits of this anal area were marked and a 12 mm vertical midline incision was made at the proposed site of anal orifice. The intrasphincteric plane was created from below to the level of the levator sling using laparoscopic back light as guide from above [Figure 2b]. The dissected intrasphincteric plane was dilated with serial Hegar's dilators up to 10–12 mm size and railroaded over the suction cannula between the two bellies of the pubococcygeus muscle in the midline.
A 10 mm trocar was inserted through the dilated tract into the peritoneal cavity [Figure 2c]. The distal end of the rectum was grasped using an endo-Babcock clamp and pulled onto the perineum through the puborectalis sling taking care not to twist the bowel and ensuring perfect hemostasis [Figure 2d]. Anoplasty was done with 4-0 vicryl stitches. Finally, the rectum was retracted cephalad and secured to the presacral fascia with 2-0 vicryl seromuscular sutures two on either side of the rectum lengthening the skin lined anal canal (only if there was redundant bowel).
Our patients were followed up for a minimum period of 6 months and were evaluated postoperatively according to Pena's postoperative continence scoring.
Fifty-eight cases have completed all three stages, eight cases are awaiting colostomy closure, and all are on regular follow-up. The age-wise distribution, site of fistula, associated anomalies, and complications in our patients are shown in Table 1.
We had 17 patients in the age group of 1–6 months, 26 patients between 7 and 12 months, 19 patients between 1 and 2 years, and 6 patients between 3 and 6 years of age.
All the patients tolerated surgery well. Four patients required conversion due to problems in gaining enough length for the distal rectum. In one patient with rectovesical fistula, colostomy was closed and relocated at splenic flexure. In three other patients, colostomy was dismantled and closed and abdominoperineal pull-through of the proximal colon was done. In addition, there were two cases of rectal serosal tear that were left to heal spontaneously.
With regard to the location of the fistulous site, 57 had rectoprostatic, 2 had rectovaginal, 5 rectovesical, and 4 were rectovestibular. (In our institution, most of the female ARMs were of anovestibular type that were managed perineally. We had only four children with rectovestibular fistulas and two children with rectovaginal fistulas who underwent LAARP).
Mucosal prolapse was the most common complication (8 cases) in our series, which required mucosal trimming. This can be prevented by taking seromuscular sutures through rectum to presacral fascia while placing cephalad tension laparoscopically. Anal stenosis occurred in five cases and was treated with posterior triangular anoplasty. Two patients developed adhesive intestinal obstruction which was managed conservatively in one patient and another patient required laparotomy and adhesiolysis. There was a peroperative spill of distal rectal contents particularly barium in these two patients. This could be avoided if on-table distal loop wash is given after induction. Distal rectal necrosis of 1.5 cm of rectum near the neoanus occurred in one patient, and in two other patients, there was distal rectal ischemic stricture. Peroperatively, distal vascular branches to the rectum were sacrificed for adequate mobilization. In patients mandating revision, PSARP was done. Urethral diverticulum occurred in two patients, probably due to leaving behind a long stump of rectourethral fistula, where one was managed conservatively and the other required fistulous tract dilatation to free its contents. These patients are asymptomatic on follow-up. Two cases had transient neurogenic bladder that were managed conservatively (catheterization), asymptomatic at present and on regular follow-up.
In all children, the progress has been satisfactory and weight gain was adequate. Out of 58 patients who had completed all three stages, 15 patients have either incontinence (11) or constipation (4); all the other patients pass formed stools 4–8 times a day initially and now passing 3–4 times per day and have symmetric anal contraction on stimulation. Anal ultrasonogram done in individuals showed a symmetric muscle complex on either sides of pull-through rectum. Distal loopogram done before colostomy closure has shown a good anterior angulation of rectum reflecting accurate placement of rectum within the puborectalis sling. Similarly, magnetic resonance imaging of the pelvis was done in five patients who showed the neorectum placed in the center of levator sling and within the anal sphincter.
Out of 58 patients who have completed all three stages, 52 patients could be evaluated postoperatively according to Pena criteria for assessment of continence for postoperative results [Table 2].
Fifty-two patients were beyond 3 years of age, of which 37 patients (71.15%) had good social continence with a minimum period of 6 months follow-up.
Two patients had sacral anomalies among which one had Grade II soiling and the other had Grade III soiling on follow-up.
Early in the 20th century, an anorectal pull-through procedure was used for high lesions. As surgical techniques improved, endorectal pull-through, abdominoperineal pull-through, and later the sacroabdominoperineal approach came into being. However, in many of these cases, identification and visualization of the levators and external sphincter muscle complex were not possible and the anorectum often was passed “blindly” into its final position. PSARP, popularized by deVries and Pena revolutionized the management of infants with imperforate anus. This approach provided excellent visualization, protection of the urogenital structures, ability to mobilize the bowel sufficiently, and identification of the urinary fistula. Fecal continence improved using this approach but often is less than ideal in patients with high fistulae.[7,8,9,10,11,12]
In an attempt to improve on these results, LAARP uses fundamental concepts learned from decades of high ARM repair incorporating modern technologic advancements. An anatomic reconstruction of the ARM (such as that which results after PSARP) could be achieved with minimal surgical trauma to the continence mechanism.[13,14]
Benefits of the procedure include lack of division of the muscle complex, laparotomy avoided, decreased pain to the patient, and potentially less perineal wound complications. Additional advantages include repair of associated defects at operation (i.e., hernia, identification and repair of cryptorchid testes), superior pelvic visualization not possible with open surgery, and anatomic placement of the pull-through bowel by identifying the central portion of the puborectalis from inside and the external anal sphincter from outside the patient with injury to nerve plexus almost nil.
In our series, various technical modifications were done to the Georgeson procedure. The baby is placed in the frog's position/lithotomy so that the position of the baby remains unchanged. We place a transcutaneous bladder stitch to overcome hindrance of vision even by the decompressed bladder in selected cases. Certain centers advocate colostomy at the umbilical site that is used as an umbilical trocar later, but we prefer a sigmoid loop colostomy in the lateral spinoumbilical line.[16,17] Traction on the undivided fistula helps in defining the slingshot of puborectalis and accurate dissection in the center of it for the pull-through channel under direct vision. Fistula ligation, as proposed, was omitted as the outcome was similar with just division. This avoids one of the most difficult steps in the constrained space training, especially in beginners with inadequate experience with intracorporeal suturing. Conventional diathermy in a low setting current could be used just as effectively as a laparoscopic muscle stimulator for stimulating the puborectalis. Veress needle with expanding sheath is placed through the proposed anal site for accurate creation of pull-through channel. Instead of the above step, we use suction cannula from the perineal end as a guide for accurate dissection. Hegar's dilators up to 12 mm size are successively railroaded with the suction cannula which directs it an accurate angle to dilate the pull-through channel.
These modifications have made the procedure easier and saved time, thereby decreasing morbidity and early postoperative recovery.
LAARP provides excellent visualization of the rectal fistula and surrounding structures. In our experience, dividing the fistula without ligation is safe. LAARP allows accurate placement of the bowel through the anatomical midline and levator sling. Early postoperative recovery, early ambulation, and decreased pain to the patient are seen in LAARP patients. Repair of associated defects at operation (i.e., hernia, identification and repair of cryptorchid testes) is possible. It is minimally invasive and leaves small abdominal and perineal wounds. We have found that LAARP is an effective alternative. Thirty-seven patients (71.15%) are beyond 3 years of age, and their social continence is good with a minimum period of 6 months follow-up. Long-term follow-up is essential for evaluation of final results.
There are no conflicts of interest.