Table 3 | Differential diagnosis

Preoperative management

Ahead of any definitive surgical intervention, the bowel above the affected area must be sufficiently decompressed, typically with rectal irrigations or faecal diversion. In neonates and young infants with rectosigmoid disease, rectal irrigations are usually effective, allowing a single-stagepull-throughwithoutadivertingstoma137.Rectalirrigations may be less effective if aganglionosis extends beyond the rectosigmoid colon, as observed in ~28% of patients138. Additionally, in older infants beyond the neonatal period and children with long-standing obstruction, the bowel may be chronically dilated and rectal irrigation may not be effective139. In these patients, a two-stage approach may be recommended to allow for decompression of the proximal bowel, reduce the risk of postoperative complications following pull-through and address any malnutrition present prior to pull-through. About 2–13% of all patients with HSCR have total colonic aganglionosis and a diverting ileostomy is the initial treatment of choice in these patients2,140.

Prior to pull-through105, HAEC must be treated because inflammation can make pathological determination of ganglion cells unreliable and can increase the risk of postoperative complications141. Treatment ofHAECusuallyinvolvesrectalirrigationsandantibiotics,butdiversion should be considered in severe cases or if symptoms do not improve with medical management105.

Surgical treatment

In general, most paediatric surgeons follow a standard approach, unless a new technique that definitively improves outcomes is reco­ gnized. Before starting surgery, several important factors must be taken into consideration — optimal timing of surgery, stages required, best surgical approach and technique, and the extent of aganglionosis in the patient.

Optimal timing of surgery. Whether performing the surgery during the neonatal period increases the risk of complications (for example, anastomotic leakage and stenosis) and whether timing impacts functional outcomes are debated in the field. Although single-stage pull-through is currently safely performed in newborns in some centres142, retrospective studies have suggested that outcomes may be improved if surgery is delayed by 1–3 months143,144. Moreover, evidence suggests that general anaesthesia during the neonatal period may be associated with neurodevelopmental delay in ~23% of patients. However, prospective studies comparing outcomes based on age at general anaesthesia are lacking, and systematic reviews show no conclusive evidence that a single short anaesthetic exposure in infancy has a detectable effect on neurodevelopment145–147. A survey found that 33% of European surgeons perform pull-through surgery at the time of diagnosis, whereas 67% delay the procedure until 3 months of age or until the infant weighs 5 kg (ref. 148). Additionally, a meta-analysis including 780 patients who underwent pull-through according to the Yancey–Soave technique found that patients undergoing surgery before 2.5 months of age had higher rates of soiling, anastomotic stricture and leakage than those undergoing surgery after this age144. The technical challenges of pull-through and/or difficulties in identifying normal ganglia in younger infants may explain these findings, but further prospective studies are needed. When pull-through surgery is delayed beyond the neonatal period, parents must be taught how to perform rectal decompression with washouts to prevent HAEC.

The timing of pull-through in patients with total colonic aganglionosis is also controversial, with approximately one-third of European surgeons opting for pull-through between 1 and 6 months of age, one-third between 6 and 12 months of age, and one-third after 1 year of age148. The aim of delayed pull-through is to wait until the ileostomy output is formed or even until the child is toilet trained for urine, to minimize the risk of perineal excoriation with an ileoanal or ileorectal pull-through149. However, a meta-analysis showed that the occurrence of perineal excoriation is not correlated with the age at pull-through150. Nonetheless, it is generally agreed that pull-through should be delayed until the ileostomy output is formed to a thicker consistency, and the child develops a satisfactory growth curve without the need for parenteral support151.

Single-stage versus multistage surgery. In patients with shortsegment HSCR, most surgeons prefer a single-stage pull-through procedure as the outcomes are similar to those following a two-stage approach137,152. This approach results in fewer re-admissions and a shorter hospital stay153. On the other hand, in patients with longsegment HSCR in whom rectal washouts are not effective, a diverting stoma in the normally innervated bowel is usually recommended as the initial step154. Ideally, frozen biopsies are performed during surgery to ensure normal innervation at the stoma site. Diversion can be either performed on the colon at the most distal location with normal innervation (levelling stoma) or performed on the distal ileum, proximal to

normally innervated colon (ileostomy). When levelling biopsies are done at the same time, the aganglionic bowel may be resected during this procedure, allowing the operative time during the pull-through to be reduced. In rare cases, a three-stage approach may be used, which involves using an ileostomy to protect the downstream anastomosis after pull-through and delayed ileostomy closure to avoid potential complications related to anastomotic fistula154.

Optimal surgical approach and technique. Several approaches are available for surgical management of HSCR, including transabdominal, transanal, sagittal posterior, laparoscopic, robot-assisted or combined approaches155,156. A study comparing trans-abdominal and transanal approaches found similar rates of postoperative complications, such as incontinence, stricture and soiling, between the two groups157. Proponents of the transanal operation cite potentially lower rates of constipation and HAEC than following the transabdominal operation157, whereas detractors express concerns for excessive stretch and potential devascularization of the anal sphincters158. Laparoscopy-assisted pull-through is a minimally invasive alternative to the open transabdominal approach and has demonstrated similar long-term functional results. The sagittal posterior approach is typically used for redo surgery. The robot-assisted approach is feasible but may have limited benefits in infants owing to the size of the instruments and follow-up remains too short to compare long-term functional outcomes159,160. The robot-assisted approach may be useful in older children or for redo surgery to facilitate rectal dissection in the pelvis156.

Thepull-throughtechniqueusedforHSCRcanvarydependingon the length of aganglionosis and surgeon preference. The three main techniques are the Swenson procedure, the Yancey–Soave procedure

and the Duhamel procedure. The Swenson procedure involves completely removing the aganglionic bowel and then pulling the normally innervatedbowelthroughtocreateacoloanalanastomosis.TheYancey– Soave procedure is an endorectal pull-through, as it involves pulling the normally innervated bowel through a rectal muscular cuff, which is created by dissecting between the muscular layer and the mucosa of therectum.TheDuhamelprocedureinvolvesaretrorectalpull-through and an anastomosis between the normal bowel and the aganglionic rectum left in place. None of these techniques has been shown to be superior to the others, and each has its own advantages and disadvantages (Fig. 5). Prospective randomized controlled trials, with high surgical standard-based studies, on surgical technique and timing, are lacking.

In patients with extensive aganglionosis, resection of the aganglionic colon and ileum may result in high stoma secretion and parenteral nutrition dependency for several months or years, a condition defined asshort-bowelsyndrome161.Severaltechniquesofautologousintestinal reconstructive surgery have been described with the aim of decreasing parenteral nutrition dependency. These include the classic enteroplasties, such as the Bianchi procedure and serial transverse enteroplasty (STEP), which are used in children with short-bowel syndrome of other causes3. Moreover, several techniques have been specifically described in children with HSCR, namely Ziegler’s myectomy–myot- omy, transposition of aganglionic ileum and skipped aganglionic lengthening transposition (SALT). In the latter two procedures, the aganglionic bowel is used to control rapid intestinal transit and high stoma secretion3,162. Finally, intestinal transplantation has also been described for total intestinal aganglionosis; the procedure has a high mortality of 25% but allows weaning off parenteral nutrition in 88% of survivors163.

Ganglionic bowel

Aganglionic bowel

Muscle Anastomosis

External anal sphincter

Fig. 5 | Different surgical techniques for pull-through. The three major surgical approaches to Hirschsprung disease, along with potential advantages and complications are shown. These include the Swenson, Yancey–Soave and Duhamel approaches. Note that each of these approaches can be approached laparoscopically or robotically. Fully transanal approaches are typically Swenson or Yancey–Soave style operations. The main immediate complications of pull-through are a rectal stump leak for Duhamel and an anastomotic

leak for Yancey–Soave and Swenson. The specific mid-term and long-term complications of each pull-through are a rectal spur for Duhamel pull-through, a tight muscular cuff or anastomotic stenosis for Yancey–Soave, and

an anastomotic stenosis for Swenson. Constipation and soiling may develop following all types of pull-through. Adapted with permission from ref. 231, Elsevier.