A New Approach to Cloaca: Laparoscopic Separation of the Urogenital Sinus

Patient population and data definitions

We conducted a review of 9 children with cloacal malformations who underwent operative repair that included laparoscopy for both rectal mobilization and UG separation. These repairs were performed at three institutions between 6/1/2016 and 5/30/2019 by 2 primary surgeons (D.S. and B.D.). All patients had previously undergone diverting colostomy. Clinical parameters reviewed included demographics, comorbid conditions, operative duration, estimated blood loss, intraoperative and postoperative transfusion requirements, common channel, urethral and vaginal lengths, gynecologic anomalies, intraoperative and perioperative complications, postoperative length of stay, 30-day readmissions, and postoperative urethral, vaginal, and rectal issues. Vaginal stenosis was defined as inability to accommodate a size 6 Hegar dilator at the time of colostomy reversal.

Ethics and institutional review

This study was approved by the Institutional Review Board at our institution (IRB-P000032845).

Operative technique

The procedures started with cystoscopy for confirmation of anatomic relationships and anomalies, including common channel, urethral and vaginal lengths, presence or absence of vaginal septum, bladder size, and position and number of ureteral orifices. In 1 patient, a ureteral stent was placed in an ectopic ureter that came off of the common channel. Using cystoscopy, a Foley catheter was placed for bladder decompression and to help with identification of the common channel during laparoscopy.

Each patient underwent a circumfrential sterile preparation, from nipples to toes, which allowed for turning between prone, supine, and lithotomy positions as needed. The peritoneal cavity was accessed at the umbilicus and insufflation was maintained between 8 and 12 mmHg. Three other abdominal access sites were created under direct visualization in the right upper quadrant, right lower quadrant, and left upper quadrant.

Diagnostic laparoscopy was performed to survey and confirm the anatomy obtained from preoperative imaging, anatomy of the Mullerian structures, length of distal colon and rectum from the mucus fistula, and rotation of the bowel. Pertubation of the Fallopian tubes in certain instances was utilized to examine patency of the Mullerian structures and to help define the distal insertion of the (hemi)vaginas into the common channel. As previously described, a 3 French feeding tube was cannulated into the fimbriae and small amounts of saline are injected to fill the fallopian tube, uterus, and vagina. ¹⁰

Blood supply to the distal colon was maintained by preserving the inferior mesenteric artery. Circumferential dissection along the rectum without violation of the wall was performed to preserve intramural blood supply. To help with anterior rectal dissection, a transabdominal monofilament stay suture was placed to retract the uterus anteriorly. Monopolar hook cautery and thermal vessel sealer were used during the initial rectal mesenteric dissection for speed and hemostasis; however, the distal dissection of the common wall with the vagina was performed sharply to minimize any thermal spread. Dissection was continued distally until the rectum narrowed at the insertion to the common channel (or in patients with two hemivaginas, to the vaginal septum). The position of the fistula relative to the levators was noted. If the rectum tapered down to a narrow fistula above the levators, it was generally sharply transected, flush with the posterior vagina. The opening was then repaired laparoscopically with absorbable sutures. Otherwise, to avoid leaving a diverticulum, the rectum was left in situ for later flush division during the perineal dissection.

Laparoscopic urogential separation was performed using sharp dissection to separate the common wall between the bladder and Mullerian structures to avoid any thermal damage (Fig. 1A), especially for the critical and difficult separation of the distal common wall. If a uterine traction stitch was used to aid rectal dissection, this was adjusted for UG separation. With the 30° laparoscope and magnification, visualization during the low pelvic dissection was ideal and allowed identification and protection of the ureters throughout their course and they could be traced to their insertion in the trigone (Fig. 1B). Dissection was carried distally until the vagina(s) entered the common channel. Upon completion of the separation, the bladder, ureters, and Mullerian structures could be clearly visualized (Fig. 1C).

OPEN IN VIEWER

FIG. 1.

Laparoscopic view (A) after opening of the peritoneum between the bladder (easily identified by the presence of Foley catheter balloon) and vagina, (B) visualizing and protecting the ureter during dissection, and (C) after dissection is complete. Bl, bladder; Va, vagina, Ur, ureter; Ut, uterus.

In cloaca reconstruction, achieving adequate mobility of the vaginas to reach the perineum is often a limiting factor. To facilitate adequate vaginal mobilization, the cardinal ligament was divided but sufficient blood supply was maintained by preservering the uterosacral ligament. The round ligaments often needed to be divided as well to ensure adequate mobilization, but we attempted to preserve at least one round ligament if possible.

With the abdominal mobilization complete, the remainder of the dissection and reconstruction was performed in the prone position with the hips elevated. A posterior sagittal incision centered in the midline was carried through the soft tissue until the previously dissected structures were encountered. Rectum or vagina was encountered first, usually predictable from preoperative imaging. The dissection was completed posteriorly with sharp dissection and each of the individual structures was separated flush off the common chanel to avoid leaving a diverticulum in the common channel.

Since the common channel was left as the urethra, a distal urethroplasty was only required, and the common channel was closed over the Foley catheter. To minimize the risk of recurrent fistulization, the urethroplasty was buttressed with a layer of fat and we ensured that healthy vaginal tissue was in apposition to the region of the closed common channel. If vaginal mobilization was not sufficient to reach the perineum, interposition grafts could be performed as needed. The perineal body and the anorectoplasty were then performed in the standard manner, centering the neoanus in the sphincter complex. The colostomy in most cases was left untouched.

Feeding began on postoperative day (POD) 1. Anal dilation was started on POD 14. The Foley catheter was left in place until POD 14 and was removed after a voiding cystourethrogram. If the patient did not spontaneously void, she was at this time deemed safe for intermittent catheterization as needed. No vaginal dilation was used.

Benefits of laparoscopic UG separation

This technique affords several advantages over prior methods. One notable benefit is the avoidance of a laparotomy, which would have otherwise been required for UG separation in long common channel cloaca patients. By avoiding TUM, the common channel remains as the functional urethra. This avoids disruption of the anterior blood supply to the urethra and maintains the position of the bladder neck above the levators, potentially improving long-term urinary continence. This may come at the risk of hypercontinence, requiring clean intermittent catheterization until the patient can generate enough intra-abdominal pressure to void. Achieving urinary continence is a known challenge in this population, and in patients with a long common channel and short urethra, UG separation may be superior to TUM mobilization for achieving urinary continence. ⁶ Fares et al. ⁷ propose that enhanced continence may be related to sparing the urethral sphincter and its innervation from manipulation and separation from its fixation to the pubis. In performing UG separation, we have maximized functional urethral length and minimized perisphincteric dissection. The group of patients is currently too young to evaluate continence but functional outcomes will be of interest in the near future. Long-term urodynamic evaluation of these patients will be helpful to evaluate this possible benefit.

Laparoscopic dissection and mobilization also reduce the perineal dissection necessary to complete the procedure. For example, the rectum can usually be mobilized completely through a posterior sagittal dissection, especially if the coccyx is excised. However, starting the mobilization laparoscopically allows for a more limited posterior sagittal incision because the majority of the mobilization was already completed. In addition, we have at times left the perineal body intact, as the entire perineum does not necessarily need to be opened to complete the urethroplasty. We suspect that minimizing the length of the posterior sagittal incision may have benefits for healing and morbidity.

The UG separation can vary in duration and difficulty, based on the vaginal insertion on the common channel. We find the dissection is more challenging in patients who have previously had vaginostomy tubes, due to inflammation secondary to the presence of the tube. Mobilization of the vagina must be performed sharply (we use laparoscopic Metzenbaum scissors) to avoid thermal injury—the size of a bipolar electrocautery device is a limiting factor in its use in this space. We maintain the broad ligament that should preserve the uterine pedicle, but the vaginal arteries are in the dissection plane, and patients are at risk for vaginal stenosis (discussed hereunder). Anecdotally, it seems that the vagina seems to reach the perineum more easily than in an open procedure, and there is less need for vaginal replacements.

Postoperative complications

Although there were no complications encountered intraoperatively, 1 patient did have a delayed diagnosis of a bowel obstruction due to twisting of the rectum, which required operative revision. This complication highlights the importance of maintaining bowel alignment, as with all abdominoperineal pull-through procedures. This is an uncommon complication, so true incidence rates are difficult to establish, but has been seen in laparoscopic, open, and transanal approaches.^11–13

There were several postoperative issues that merit consideration but are unlikely related to the laparoscopic nature of the repair. The patient with the most protracted postoperative course had significant issues related to volume management and underlying chronic renal insufficiency. This patient had significant comorbidities including vesicoureteral reflux, and also underwent a ureteral implantation. We suspect that many of the postoperative challenges encountered were secondary to significant underlying renal pathology. It is unclear whether laparoscopic insufflation affected renal perfusion in this case, as insufflation pressures were generally kept between 8 and 12 mm Hg.

One patient developed a postoperative urethral web after premature displacement of the Foley catheter on POD 2. This was later repaired through a rendezvous direct vision internal urethrotomy. A consideration, employed subsequently by author D.S., to avoid catheter dislodgement is to place a silicone circle stent through the urethra and vesicostomy if one is present. ¹⁴

Vaginal stenosis is a known complication associated with cloacal repair.^15,16 It has been observed in TUM mobilization ¹⁷ and UG separation, without clear evidence of relation to technique. Anecdotally, we have observed that the laparoscopic approach allows for complete mobilization of the Mullerian structures, enabling excellent mobility of the vagina and a low tension perineal–vaginal anastomosis. Only 1 patient required intestinal interposition for vaginal replacement. Owing to small sample size, it is unclear whether the laparoscopic nature of the repair impacts the rate of stenosis compared with traditional techniques. Further study is required to understand whether the stenosis is related to the circumferential vaginal dissection associated with UG separation. Owing to the sensitive nature, postoperative vaginal dilations were not performed and this may contribute to the rate of stenosis. We prefer to defer intervention for stenosis until closer to puberty, when patients are more able to participate in dilations. We recommend a complete examination under anesthesia and vaginoscopy before menarche and before intercourse to ensure timely evaluation of vaginal caliber.

Operative time was long but consistent with the complex nature of these repairs, and interpretation of the operative times reported here is confounded by the presence of other procedures performed during the same anesthetic. In general, it seems that operative times for the separation are faster than the open procedure as the deep pelvis is easier to access with the camera and laparoscopic instruments. In summary, laparoscopy enables minimally invasive separation of UG structures, a procedure that may previously have necessitated laparotomy and extensive posterior sagittal incision and perineal dissection in these long common channel clocal malformations. It enables full evaluation of Mullerian structures and associated anomalies, which is not possible with a perineal-only approach. The visualization and mobilization with a minimally invasive approach are excellent and better than in an open approach. With advancement of optics and instruments available for laparoscopy, the dissection of common planes is made possible. Postoperative recovery was short. Complications and readmissions were not clearly related to the laparoscopic nature of the surgery. From the results seen in our patient cohort, we conclude that complete laparoscopic separation of the rectum, vagina, and urethra in cloacal reconstruction is a novel and appropriate approach in the management of these complex malformations.