Early Experience With Single-Stage Transanal Endorectal Pull Through For Rectosigmoid Hirschsprung ’ s Disease

Address for correspondence: Dr. Manoj Krishna Shrestha Consultant Paediatric Surgeon Kathmandu Model Hospital Kathmandu, Nepal E-mail: shresthamanoj@hotmail.com Tel: +977-9849982899 1Dr. Manoj Krishna Shrestha, MBBS, MS, Consultant Paediatric Surgeon, Kathmandu Model Hospital, Kathmandu, Nepal, 2Dr. Mamata Sherchan, MBBS, Medical officer, Siddhartha Woman and Children Hospital, Butwal, Rupandehi, Nepal, 3Dr. Bhim Gopal Dhoubhadel, MBBS, Medical Officer, Siddhartha Woman and Children Hospital, Butwal, Rupandehi, Nepal, 4Dr. Ranga Bahadur Basnet, MBBS, MD (Path), Senior Consultant Pathologist, Kathmandu Model Hospital, Kathmandu, Nepal. Abstract


Introduction
H irschsprung's disease (HD) is congenital aganglionosis of distal intes ne.It is the commonest cause of intes nal obstruc on in neonates with an incidence of 1 in 4000-5000 newborns.Rectosigmoid HD comprises 75-80% of all HD 1 .
The diagnosis of HD is based on clinical presenta on, radiological fi ndings, anorectal manometry and rectal biopsy 2 .Delayed passage of meconium, features of intes nal obstruc on in neonatal period, cons pa on/obs pa on in exclusively breast-fed babies warrant HD.Barium enema (BE) fi nding of radiological transi on zone or rectosigmoid index <1 supports diagnosis of HD 3 .The diagnosis is confi rmed by rectal biopsy 2 .
The principle of treatment of HD is resec on of aganglionic segment of the gut, pull-through of ganglionic proximal gut and its anastomosis with the anus.Staged-opera ons i.e., colostomy followed by any one of the abdominal pull-through opera ons and colostomy closure, have been standard treatment.None of the pull through opera ons is devoid of complica ons.There is risk of damage to sacral nerves responsible for fecal and urinary incon nence as well as sexual dysfunc on with Swenson's procedure, occurrence of fecaloma in the retained aganglionic segment in Duhamel procedure and high incidence of cons pa on with Soave procedure 4 .Mul stage procedures involve high morbidity and mortality of the pa ent and prolonged psychological stress of parents.Hence there has been a recent trend towards minimally invasive single-stage primary pull-through for HD 5 .
Dela Torre-Mondragon and Ortega-Salgado in 1998 fi rst reported that pull-through opera on can be performed transanally 6 .Many reports have been published confi rming safety and feasibility of transanal endorectal pull-through (TEPT) 7,8 .TEPT represents the latest development in the concept of minimally invasive surgery for short segment HD. Laparoscopy or mini-laparotomy can be incorporated in case of long segment HD 9,10 .
The aim of this study is to evaluate the effi cacy and safety of single-stage TEPT for rectosigmoid HD in diff erent pediatric age group in our setup.

Materials and Methods
Over a 5 year-period (July 2008 -July 2013), all the children who were clinically suspected HD underwent BE and punch rectal biopsy to confi rm HD.Twenty children aged 22 days to 7 years (17 boys and 3 girls) with biopsy proved HD and BE documented short segment HD were included in the study.Children who presented with colostomy, who needed laparotomy and colostomy, who could not be sa sfactorily decompressed by rectal irriga on and needed colostomy and who have long segment disease in BE were excluded from the study.
All the pa ents were approached as per standard treatment protocol 2 shown in Figure 1 (Fig 1).All clinically suspected pa ents underwent plain X-ray abdomen in erect posture to rule out pneumoperitoneum.Those who had pneumoperitoneum underwent emergency laparotomy, mul ple biopsies to confi rm aganglionic and ganglionic segments and diver ng or leveling colostomy.These pa ents were excluded from the study.
Those pa ents who had no pneumoperitoneum were subjected to barium enema.We performed barium enema a er ruling out pneumoperitoneum at presenta on on admission if no rectal manipula on like per rectal digital examina on or rectal enema or even suppository had been done.The barium enema was delayed at least for 24 hours if any rectal manipula on was present in order to lessen false posi ve or false nega ve fi ndings.
Conserva ve management consisted of NPO, IV fl uid, IV an bio cs and rectal wash out.Rectal wash out with warm normal saline was started a er the barium enema was completed.We performed punch rectal biopsy on those pa ents who improved with the conserva ve treatment.The biopsy proved and barium enema documented short segment Hirschsprung's disease were included for the study.Those who could not do well with conserva ve treatment and underwent laparotomy, mul ple biopsies and colostomy were excluded from the study.
The pa ents' demographics, diagnos c work up, opera ve fi ndings and postopera ve fi ndings were recorded.Preopera ve prepara on: Adequate hydra on and intravenous second genera on cephalosporin and metronidazole were used as prophylac c an bio cs in all pa ents.Preopera ve bowel prepara on consisted of rectal washout with warm normal saline daily.Breast fed babies were allowed breast milk ll 6 hours before opera on.Older children were kept on low residual diet ll 24 hours and clear liquids ll 6 hours before opera on.

Surgical techniques:
The opera on was done under general anesthesia with endotracheal intuba on.A caudal block was used for preemp ve and postopera ve analgesia.The pa ent was placed in lithotomy posi on for older child and suspended lithotomy posi on for infant and small child (Fig 2).The pelvis was slightly raised with a sand bag/rolled towel under the sacrum.The opera on table was slightly inclined to raise the pelvis for be er visualiza on of the anal canal.A bladder catheter was not rou nely inserted.
The anus was retracted with stay sutures to expose the anal canal (Fig. 3).Submucosal injec on of saline or air was used to ease the ini al submucosal dissec on 11 .A circumferen al incision was given on the anal canal mucosa about 5 mm proximal to the dentate line.Mul ple stay sutures were used to hold the mucosal layer as it is dissected circumferen ally.Though ini al mucosal dissec on was diffi cult and slow, it became easier and faster once the correct submucosal plane was found.The perfora ng mucosal vessels were cauterized.The submucosal dissec on was con nued proximally with blunt dissec on by trac on on the mucosal tube and countertrac on on the muscle cuff and cauteriza on of perfora ng vessels.A er approximately 10-15 cm mucosal tube dissec on, there was sudden give way when peritoneal refl ec on is reached (Fig. 4).Then the muscle of the rectum was incised circumferen ally.The muscle cuff was held by two hemostats near midline posteriorly.With the help of a right angle forceps placed posterior to the muscle cuff , it was divided in the midline posteriorly up to the dentate line to prevent cons pa on.The muscular cuff was excised to shorten the cuff length to about 5 cm 12 .This provided more space for opera ve manipula on.The rectosigmoid vessels were either cauterized (in infants) or ligated (in older children) and the rectum and sigmoid colon was gradually pulled down through the anus.The pull through of the colon was con nued at least 5 cm beyond the transi onal zone.The normal colon was dilated, hypertrophied and showed good peristalsis on s mula on (Fig 5).It was cut transversely at this level so that a part of ganglionic colon along with hypoganglionic and aganglionic segment of colon was excised to avoid retained aganglionic segment.Yes we had no facili tes of frozen sec on biopsy.Ganglionic colon was assessed clinically as it was dilated, hypertrophied and had good peristalsis on s mula on.We preferred excising a few inches more of ganglionic segment than having retained aganglionic segment.We discussed this in discussion sec on.It was essen al that the proximal excised end has good vascularity as well.Several fi xa on s tches were placed at seromuscular layer near the cut margin and the muscular cuff just proximal to the anal mucosa to prevent retrac on of pulled through colon.Colo-anal anastomosis was done using 4-0 polyglac n (vicryl) (Fig. 6).Drains were not used.

Postopera ve management
Feeding was started early on demand, usually 3-6 hours a er the opera on star ng with clear fl uid or breast milk and gradually advanced to liquid and so diet as tolerated.Intravenous fl uid was con nued ll 24 and then gradually tapered.Intravenous an bio cs were con nued 72 hours postopera vely followed by oral an bio cs for fi ve days.The pa ent was discharged (usually at POD 4 or 5) when feeds were well tolerated, fl atus and faeces passed regularly, no abdominal distension and/or vomi ng and no fever or any abnormal discharge per anus.
Pa ents were advised to visit for follow up two weeks a er opera on for rectal examina on and rou ne anastomo c dilata on to prevent anastomo c stricture.The parents were taught to introduce the anal dilator beyond the anastomosis site.The dila on schedule is once daily for one month, on alternate days for one month, twice weekly for one month and once a week for three months.Follow up visits were arranged at one month, three months, six months and then yearly to look for late postopera ve complica ons like anastomo c stricture, cons pa on, encopresis, enterocoli s, stooling pa erns, retrac on or prolapse of pulled through colon, residual aganglionosis, etc.We considered our pa ents con nent clinically if they passed stool regularly at least once daily and they remained dry without fecal soiling, We do not have anorectal manometry or video defecography or other means to assess anorectal func on.

Results
During the fi ve year study period, 20 children underwent one stage TEPT.There were 17 boys and three girls.The age ranged from 22 days to seven years (mean 18 mo) with 12 infants and 8 older children.Median follow up was 18 months (range 6-36 months).
The opera on me ranged from 105 min to 180 min (mean 120 min).It was longer in older children as submucosal dissec on was more diffi cult in older children compared to infants.Intraopera ve blood loss was 5-40 ml, more in older children.Blood transfusion was needed in none out of 12 infants and 2 out of 8 older children (>10% of es mated blood volume).There was mucosal perfora on in 3 infants during submucosal dissec on.This had no eff ect on infec on or outcome of the opera on.The length of resected bowel was 18-30 cm (mean 25 cm).The opera on events are summarized in Table 1.
Oral feeding consis ng of mother's milk or clear fl uid was allowed 3 hours postopera vely irrespec ve of bowel sound.The feeding increased as the pa ents tolerated.The full feed was achieved by 24-48 hours.A few episodes of vomi ng without abdominal distension were observed in six pa ents which improved with me without medica on Although presence of bowel sound, passage of fl atus or faeces are parameters to start oral feeding, we preferred early enteral feeding as it has several advantages 21 .
Frequency of bowel movement was 2-14 mes/ day which gradually se led to 1-3 mes/day by three months.Transient perianal excoria on occurred in 16 subjects, all improved with medica on and decreased frequency of stool with me.
One pa ent developed cons pa on with encopresis following anastomo c stricture.He improved with anastomo c dilata on, laxa ve and toilet training He was two years old when presented to me, 14 months a er the opera on.We think this was fecal impac on with overfl ow incon nence which improved with dilata on of the stricture.Postopera ve enterocoli s occurred in one infant who improved with gentle passage of fl atus tube, warm NS irriga on and IV an bio cs.One pa ent required diver ng colostomy for anastomo c leakage and peritoni s.One pa ent had retained aganglionic segment shown by scanty ganglion cells in histopathology report of pulled through colon.He underwent re-do pull-through transanally.None had cuff abscess or prolapse or retrac on of pulled through colon.There was no conversion to laparotomy and no mortality.The postopera ve complica ons are summarized in Table 2. Perianal excoria on occurred due to increased frequency of stool complemented by fric on on cleaning the stool and superimposed with fungal infec on.Perianal excoria on was taken care with applica on of barrier like Vaseline or oil, an fungal topical cream and it improved as the frequency of stool decreased with me.

Discussion
The diagnosis and treatment of HD have always been a challenge to pediatricians and pediatric surgeons.Increasing awareness of presenta on and high index of suspicion can lead to early diagnosis.Biopsy-proved aganglionosis and properly performed BE correlated with the level of aganglionosis are prerequisites for TERPT 13 .
The use of complete transanal pull through for HD was fi rst described by Dela Torre-Mondragon and Ortega-Salgado in 1998 6 .Many reports have been published since then from diff erent parts of the world.The safety, feasibility, reproducibility and less invasiveness of the procedure made it popular in recent years 9,14,15,16,17,18 .Ini al submucosal dissec on was slow and diffi cult.It was facilitated by submucosal injec on of air or saline 11 .In our series, we found that the submucosal dissec on was easier in neonates and infants compared to older children.The opera on me and blood loss were less in neonates and infants.This was probably due to more adhesion resul ng from recurrent enterocoli s and bigger vessels in older children.The opera ve me of our series is comparable with several other studies 7,8,9 .
We found increased frequency of bowel movement (2-14 mes/day) a er TEPT which gradually se led to 1-3 mes/day by 3 months.Transient perianal excoria on occurred in 16 subjects; all improved with medica on and decreased frequency of stool with me.Similar fi nding was reported by other authors 7,19 .
We had anastomo c stricture at the beginning of the series.The dilata on of the stricture was diffi cult for the parents as they could not accurately pass the dilator through the stricture.So we started rou ne prophylac c anastomo c dilata on similar to anal dilata on following Pena procedure for ARM to prevent postopera ve anastomo c stricture.Rou ne anastomo c dilata on was easy and parents could perform it correctly.Since then we did not have anastomo c stricture.Rou ne anastomo c dilata on has been recommended by several other authors 7,18,20 .
Enteral feeding 21 was done and there were a few episodes of vomi ng without abdominal disten on in six subjects which improved with me without medica on.This may be because of the side eff ects of the medicines we used.We had mul ple episodes of vomi ng with abdominal distension in two subjects because of enterocoli s in one and anastomo c leakage and peritoni s in the other.They improved with the treatment of the cause.
There was one incidence of retained aganglionic segment.The histopathology of the distal pulled through colon, which was rou nely done to confi rm the presence of ganglion cells in the pulled through colon, showed scanty ganglion cells.Parents were warned that the child might have the previous symptoms and might need second opera on if symptoms recurred.He developed the symptoms a er 4 months and we managed by re-do TEPT.Ideally, frozen sec on biopsy of the distal pulled through colon should be done before colo-anal anastomosis.Since there were no facility of frozen sec on biopsy in the country (except one ins tu on) it was not done.On the other hand, frozen sec on biopsy is not foolproof; it does have false posi ve results 22 .In developing country like ours, we depend clinically to see the dilated, hypertrophied bowel with peristal c movement on s mula on.To be on safe side we resect about 5 cm of the clinically ganglionic segment 7 .
There were no complica ons like cuff abscess, prolapse or retrac on of pulled through colon, adhesive bowel obstruc on as reported in other series 9,23 .

Conclusion
Single-stage TEPT is minimally invasive, safe and eff ec ve procedure for rectosigmoid HD in diff erent pediatric age group in our setup.Moreover it can be learnt quickly and reproducible.

Table 1 :
Opera on events