Laparoscopic-Assisted Modified Posterior Sagittal Anorectoplasty for Rectobulbar Urethral Fistula of Anorectal Malformation: A Prospective Study.

Introduction: Laparoscopic anorectoplasty (LAARP) is useful for the management of rectoprostatic urethral fistula (RPUF), due to easier rectal mobilization, avoidance of posterior sagittal muscular incision, and shorter hospital stay. However, its role in rectobulbar urethral fistula (RBUF) is still debated as there is a chance of urethral diverticulum (UD), due to incomplete dissection. Laparoscopy-assisted modified posterior sagittal anorectoplasty (LAMPSARP) utilizes advantages of LAARP combined with fistula dissection using small sagittal incision preserving puborectalis. The present study compares the results of LAMPSARP with LAARP for correction of RBUF associated with anorectal malformations (ARMs). Materials and
Methods: All male ARM with RBUF presenting in a tertiary center in Central India (January 2014-January 2016) were included. Low male ARM, RPUF, rectovesical fistula, and congenital pouch colon were excluded. They were randomized into LAARP and LAMPSARP groups. Complications were assessed in terms of anal stenosis, mucosal prolapse, and UD. Kelly's scoring and Krickenbeck scoring were used to assess continence, and visible anal cosmesis scale (VACS) was used to assess wanal cosmesis. Results were statistically analyzed using a comparison of means and 2 × 2 contingency tables.
Results: Fifty-six colostomized patients with RBUF (26 LAARP, 30 LAMPSARP) were included. Mean operative duration in the LAARP group and LAMPSARP group was 42 ± 10 min and 56 ± 12 min, respectively (P < 0.0001). On mean follow-up of 4.5 years, mucosal prolapse (53.9%) and UD (15.38%) were significantly higher in LAARP group, while anal stenosis was similar. All three, Kelly's score, Krickenbeck score, and VACS, were better (P < 0.05) in the LAMPSARP group.
Conclusion: Laparoscopy-assisted modified posterior sagittal approach is better for RBUF and offers better surgical outcome. Copyright:

INTRODUCTION

Laparoscopic anorectoplasty (LAARP) proposed by Georgeson et al. in 2000 is becoming a new gold standard for high anorectal malformation (ARM) with rectoprostatic urethral fistula (RPUF) and rectovesical fistulae.[1] LAARP's advantages are better rectal mobilization, reduced bleeding, smaller incisions, avoidance of posterior sagittal muscular incision, and shorter hospital stay. However, it is not popular for the management of rectobulbar urethral fistula (RBUF) due to difficulties in the management fistula deep in the pelvis, the risk of the urethral diverticulum (UD) due to residual stump, and uncertain long-term outcomes of the technique.[234] The laparoscopy-assisted modified posterior sagittal anorectoplasty (LAMPSARP) technique is previously shown to be advantageous for dissection of longer and lower rectovestibular fistulae in female ARM by using posterior sagittal assistance without dividing sphincters.[5] We evaluated the LAMPSARP approach for RBUF and presented long-term results to assess its advantages over the total laparoscopic approach in terms of operative and long-term outcome.

MATERIALS AND METHODS

This is a prospective cohort study including all patients of male ARM with RBUF who underwent colostomy at neonatal age and presented between January 2014 and January 2016 in a tertiary institute in Central India. All low malformation (treated with cutback anoplasty), RPUF (managed with LAARP), rectovesical fistula (managed with LAARP), and congenital pouch colon (managed with open pouch excision) were excluded from the study.

A distal loopogram and voiding cystourethrogram in separate settings were done before anorectoplasty for locating the level of fistula. The corrective procedure was performed after 12 weeks of colostomy and when the baby's weight was a minimum of 5 kg. Stoma closure was done 6–8 weeks after anorectoplasty. All patients were operated on at one center by the same pediatric surgeons’ team of two surgeons having experience in laparoscopic surgery of more than 10 years. Before the intervention, computer-generated randomization was done into two groups by secretarial staff. Patients were divided into two groups; LAARP group and LAMPSARP group. The LAARP was performed as a standard procedure described by Georgeson et al.[3]

The technique of laparoscopy-assisted modified posterior sagittal anorectoplasty

LAMPSARP was performed in two major steps; laparoscopic and perineal, as described by Liem et al. which is explained with the help of schematic diagrams as shown in Figure 1.[5] Patient was catheterized. In the laparoscopic step, one trocar is placed at the umbilicus for the laparoscope (5 mm), after insufflation (8–10 mmHg pressure and flow rate of 1 L/min); hook and grasper are placed at the right and left iliac fossa for instruments (3 mm) with trocar-less stab technique. A percutaneous bladder suspension stitch was used to expose the site of dissection. A window is created in the mesorectum, and dissection around the rectal pouch is continued 5–10 mm below the peritoneal reflection. The anterior dissection is done under the advantage of magnification to free the rectum from the prostatic urethra, seminal vesicles, and vas deferens. The posterior dissection is done in the presacral space as low as possible. The rectal mesentery is mobilized adequately to achieve enough length of the mobilized rectum for pull-through. Pneumoperitoneum is deflated, trocar and instruments are removed, and port sites are closed. In the perineal step, the position of the patient is changed to the prone jack-knife and modified PSARP is performed. It involves a 3–4-cm posterior sagittal skin incision in midline starting from the coccyx. The incision is deepened in the upper part, and the posterior-most part of pubococcygeal raphe is divided without division of muscle complex. The coccyx is not routinely removed (a deviation from Liem's orginal description as rectal pouch is easily identified without removal of coccyx in RBUF). The puborectalis is retracted downward without dividing. The midline dissection is continued to the rectal pouch, which is easily identified as laparoscopic posterior dissection has already been done. The rectal pouch is detached from the anterior surface of the sacrum bluntly using a peanut dissector. Separation of the rectal pouch is started from the lower part of the prostatic urethra by meticulous dissection anterior to the rectum and continued distally to the fistula. The fistula is divided and repaired with the 4-0 polyglactin-910 suture (Vicryl®, Johnson and Johnson Private Limited; Jogeshwari (E), Mumbai - 400 060). The site of the neoanus and the center of the external sphincter are identified using a muscle stimulator, and a tunnel is created underneath an undivided puborectal sling part of muscle complex connecting with the presacral space, which is dilated with Hegar's dilators (sizes 6–12) [Figure 2]. The rectal pouch is pulled through the tunnel and then sutured to the external sphincter and then to the skin; however, the V-Flap described by Liem was not used. The rectum is sutured to the upper border of the sphincter complex, and then, the incision is closed in two layers.

Figure 1

(a-c) Schematic diagram explaining laparoscopic-assisted modified posterior sagittal anorectal pull-through

Figure 2

Creation of tunnel underneath puborectalis without division (skin between proposed anus and modified posterior sagittal incision has been divided in this image for demonstration purpose)

The outcome was assessed in terms of operative duration and hospital stay, postoperative complications, continence, and cosmesis. Complications were assessed by a senior resident who was not involved in the surgical management, in terms of anal stenosis, mucosal prolapse, urethral stricture, and UD (which were assessed 6 monthly by perineal and abdominal ultrasound as a screening tool for UD), and only positives were confirmed on computed tomography (CT)/magnetic resonance imaging (MRI) and urethroscopy. Kelly's score and score based on Krickenbeck classification of postoperative results [Table 1] were used to assess the continence after 1 year of completion of all stages for at least 4 years. Postoperative continence scores were assessed by a paramedical team member of the incontinence clinic under the supervision of senior resident involved in the assessment, at least thrice with a minimum interval of 6 months, and the best score out of all the assessments was allotted to the patient. The anal cosmesis was assessed after 1 year of completion of all stages, using the visual anal cosmesis scale (VACS) which was a Likert scale ranging from 1 to 4 (1- poor, 2 - fair, 3 - good, 4 - excellent) as assessed by the parents [Table 1]. Results were statistically analyzed using a comparison of means by a t-test and comparison of frequency by the Chi-square test using 2 × 2 contingency tables. A P < 0.05 was considered statistically significant.

Table 1

Description of Kelly score, score based on Krickenbeck classification, and visual anal cosmesis scale

Kelly’s score
Criteria	Score
Staining/smearing
None	3
Occasional	2
Constant	1
Accidental defecation/soiling
None	3
Occasional	2
Constant	1
Sphincter contraction
Strong and effective	3
Weak and partial	2
None	1
Score range 3–9, 3: Worst and 9: Best result
Scoring based on Krickenbeck classification of postoperative results
Criteria	Score
Voluntary bowel movements
Yes	2
Feeling of urge, capacity to verbalize, hold the bowel movement
No	1
Soiling
Grade I, occasionally (once or twice per week)	3
Grade II, every day, no social problem	2
Grade III, constant, social problem	1
Constipation
Grade I, manageable by changes in diet	3
Grade II, requires laxative	2
Grade III, resistant to laxative and diet	1
Score range 3–8, 3: Worst and 8: Best result
VACS, based on parents’ response on a Likert scale
Scale	Finding
1	Ugly looking anus, warrants correction
2	Abnormal anus but not ugly, correction demanded
3	Near normal anus, no correction demanded
4	Normal anus, fully satisfied

Scale ranges from 1 to 4, 1: Worst and 4: Best result. VACS: Visual anal cosmesis scale

RESULTS

A total of 157 male patients with ARM were admitted to our center during the study period [Figure 3]. Fifty-six patients of RBUF were included in the study. Twenty-six patients were operated on with LAARP and 30 patients were operated on with LAMPSARP. Two groups were compared for their demographic characteristics, as shown in Table 2, and were found to be statistically comparable. Patients were followed for a median of 5 years (range, 4–6 years). Overall presence of associated anomalies included polydactyly (5, 8.92%), small ventricular septal defect (4, 7.14%), partial sacral agenesis (2, 3.57%), cleft palate (1, 1.78%), and tracheoesophageal fistula (1, 1.78%). Mean operative duration in LAARP and LAMPSARP groups was 42 ± 10 min (30–70) and 56 ± 12 min (40–80), respectively (P < 0.0001). Mean hospital stay was 4 ± 2 days and 4 ± 1 days for LAARP group and LAMPSARP groups, respectively (P > 0.05). There were no intraoperative complications. There was no mortality in either group. The distribution of complications and scores in two groups is shown in Table 3. Mucosal prolapse occurred in 14 (53.9%) patients in LAARP group and 4 (13.3%) patients in LAMPSARP group 2 (P < 0.05). Mucosal prolapse in all patients was managed by local sclerosant injection therapy with sodium tetradecyl sulfate except one in the LAARP group which needed surgical excision. Anal stenosis occurred in 6 (23.1%) patients in LAARP group and 6 (20%) patients in LAMPSARP group (P > 0.05). Anal stenosis was managed with regular anal dilations; however, four patients (two in each group) required anoplasty for its correction. Four patients (15.38%) in the LAARP group developed UD which was identified on ultrasound, confirmed on CT/MRI [Figure 4]. Urethroscopy was not conclusive in any of the patients. UD was found as a complication in four patients of the LAARP group (none in LAMPSARP group), two of them were large and had symptoms of lower urinary tract obstruction and recurrent infection, and hence they required laparoscopic excision. The other two UDs were small and asymptomatic, hence they were kept under observation. None of the patients in the LAMPSARP group developed residual diverticulum (P < 0.05). Outcome as assessed by means of Kelly's score, Krickenbeck score, and VACS was better (P < 0.05) in the LAMPSARP group.

Figure 3

CONSORT diagram showing the distribution of included cases in the study

Table 2

Comparison of demographic data between laparoscopic anorectoplasty and laparoscopy-assisted modified posterior sagittal anorectoplasty group

Variable	LAARP group (n=26), n (%)	LAMPSARP group (n=30), n (%)	P
Age at surgery in months (mean±SD)	7.20±3.85	7.06±3.97	0.446
Weight at surgery in kilos (mean±SD)	4.75±1.40	4.57±1.40	0.316
Polydactyly	2 (7.69)	3 (10)	1.000
Ventriculoseptal defect	3 (11.53)	2 (6.66)	0.654
Partial sacral agenesis	1 (3.84)	1 (3.33)	1.000
Cleft palate	1 (3.84)	0	0.464
Tracheoesophageal fistula	0	1 (3.33)	1.000
Mean age of follow-up (months)	56±4	57±5	0.817

LAARP: Laparoscopic anorectoplasty, LAMPSARP: Laparoscopy-assisted modified posterior sagittal anorectoplasty, SD: Standard deviation

Table 3

Comparison of complications and various scores in two groups

Surgical outcome	LAARP (n=26), n (%)	LAMPSARP (n=30), n (%)	P
Anal wound infection	3 (11.53)	1 (3.33)	0.570
Mucosal prolapse	14 (53.84)	4 (13.33)	0.0012*
Anal stenosis	6 (23.07)	6 (20)	0.77
Urethral diverticulum	4 (15.38)	0	0.0258*
Urethral stricture	0	0	-
Kelly’s score (mean±SD)	4.96±0.19	5.7±0.53	<0.0001*
Krickenbeck score (mean±SD)	4.32±0.21	5.6±0.18	<0.0001*
VACS (mean±SD)	1.84±0.67	3.73±0.44	<0.0001*

*Statistically significant. SD: Standard deviation, VACS: Visual anal cosmesis scale, LAARP: Laparoscopic anorectoplasty, LAMPSARP: Laparoscopy-assisted modified posterior sagittal anorectoplasty

Figure 4

Contrast-enhanced computed tomography study showing urethral diverticulum after laparoscopic anorectoplasty in rectobulbar urethral fistula (a – Coronal, b – Transverse)

DISCUSSION

LAARP for high ARM with RPUF allows better rectal mobilization, reduced bleeding, smaller incisions, avoidance of posterior sagittal muscular incision, and shorter hospital stay.[1567] However, there is an ongoing debate regarding the role of LAARP in the RBUF.[8] In RBUF, the rectum shares a long common wall with the urethra, leading to difficulty in dissecting the fistula along the urethra in an infant's narrow pelvis.[2348] Overzealous dissection may lead to urethral injuries; on the other hand, an overcautious approach leads to retaining a long segment of residual fistula leading to UD in up to 18% of cases.[491011] On the contrary, Pena's posterior sagittal anorectoplasty (PSARP) allows complete and safe dissection of the RBUF but is associated with difficult mobilization of extrapelvic rectum/colon and bisection of perineal musculature/puborectalis as its major disadvantages against LAARP.[512] LAMPSARP, described by Liem et al. for female vestibular and male rectourethral fistula, combines the benefit of laparoscopic rectal mobilization and management of fistula directly under vision through the modified posterior sagittal approach.[513] This technique allows easier, accurate, and complete management of fistula without bisecting the perineal musculature/puborectalis. We adopted LAMPSARP for RBUF to avail the advantages of both laparoscopic and Pena's approach, simultaneously avoiding their disadvantages, and compared the outcomes with LAARP for RBUF.

In the present study, the rate of the residual diverticulum was 15.38% (7.69% requiring surgical correction) in the LAARP group as compared to none in the LAPMSARP group, substantiating the importance of complete fistula dissection and division close to the urethra, which is aided by LAMPSARP.[451011] Koga et al. suggested the use of cystoscopy for intraoperative fistula measurement during LAARP to avoid such a complication.[10] The LAARP for RPUF is known to have rectal prolapse in 14.08% and a posterior UD in 3.02% of patients.[14] The higher propensity of the UD is peculiar to RBUF, unlike RPUF where fistula management is relatively easier with a laparoscopic approach and has proved to be safe.[15] MRI is more sensitive than cystoscopy and voiding cystourethrogram (VCUG) to detect UD as the urethral side of the fistula gets sealed, and the UD does not communicate with the urethra.[16] We identified all the cases by perineal ultrasound which is a readily available inexpensive tool for follow-up; however, they were further confirmed by further imaging (CT - ‘RI - 1) [Figure 4]. Urethral and perineal sonography is a valuable tool for diagnosis of urethral anomalies due to easy availability, excellent delineation of anatomy with high-frequency probe, lack of radiation exposure, and reduced cost; allowing it to be a good screening tool where a positive finding can be further characterized with higher imaging modalities.[17] There may be potential for malignant transformation due to long-term exposure of colonic epithelium to urine; however, the consensus for surgery on asymptomatic UD is lacking.[161819] All symptomatic UDs need surgical management by posterior sagittal (small UDs) or abdominal approach (large UDs).[16] We managed both of our symptomatic UD with laparoscopic excision as they were large and intra-abdominal.

The other striking finding in the present study was a significantly lower rate of mucosal prolapse after LAMPSARP as compared to LAARP. The higher rate of mucosal prolapse in LAARP as compared to PSARP has been attributed to poor rectal fixation to the pelvic musculature.[20] Few authors recommend rectal anchoring stitch in LAARP to avoid mucosal prolapse, which adds to the complexity of the procedure and may require additional endosuturing skills.[21] This step of surgery becomes easier with modified posterior sagittal incision used in LAMPSARP.[5] It is also evident from our observations that the patients who underwent LAMPSARP had a better outcome in terms of continence as suggested by significantly higher Kelly and Krickenbeck scores (mean follow-up - 4.5 years). The functional outcome in terms of continence scores is found to be better for LAARP as compared to PSARP, which reflects that avoidance of puborectalis division, the precision with which the pulled-through segment can be centered in the pelvic muscle complex, and preservation of anorectal angle, which is supposed to be straight in LAARP as it helps in achieving better continence.[20] Moreover, we strongly feel that assessment of outcome in terms of anal cosmesis, which is appreciated by parents, is also important. Visible anal cosmesis is the parents’ observation which has a direct correlation with their satisfaction after surgery. The importance of cosmesis has been highlighted by Pena in his long-term follow-up studies.[22] We used a Likert scale to evaluate the cosmetic satisfaction of parents and found LAMPSARP to be superior in providing anal cosmesis, which can be attributed to better placement of rectum within sphincter complex, small perineal scar, and decreased incidence of mucosal prolapse. Therefore, our hypothesis that the LAMPSARP provides advantages of both LAARP and PSARP and avoid the drawbacks of both of the techniques. This helps in reducing complications and improving long-term results. For RBUF, laparoscopic dissection using single incision laparoscopy, anterior rectotomy, and pushing the pouch with a canula in to the perineum for perineal fistula dissection have been described to overcome the low fistulae.[23] Liem's approach is easier, safer, and has been found to be comparable to Pena's approach and better than total laparoscopic approach in immediate outcomes;[513] however, the present study establishes better outcomes in terms of continence and cosmesis.

The disadvantage of LAMPSARP is the need for a change of position of the patient during surgery, longer operative duration, and a linear scar below the coccyx, which is however hidden in the cleft. The longer operative time of few extra minutes is not very relevant in the clinical setting to reduce complications. However, the present study has certain limitations about the strength of evidence, heterogeneity, and lack of long-term assessment.

CONCLUSION

Combining laparoscopic assistance and modified posterior sagittal approach of anorectoplasty by LAMPSARP for the repair of RBUF makes the operation technically easy and safe and offers better surgical outcomes in terms of complications, continence, and cosmesis.

Informed consent

The subject gave informed consent for the procedure and acquisition of documentary information. Patients’ anonymity is preserved.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.