Hirschsprung’s disease (HD) is one of the most common congenital disorders of the lower gastrointestinal tract, with an incidence of approximately 1:5000 live births[1,2]. Children with HD usually present within the first six months of their life with symptoms related to bowel obstruction, such as delayed first passage of meconium, abdominal distension, and bilious vomiting[2,3]. However, later presentations are also recognized, where intractable constipation and Hirschsprung-associated enterocolitis may be more prominent[1,3]. The underlying congenital defect of HD is a complete absence of ganglion cells in the most distal part of the gastrointestinal tract. Due to this aganglionosis, there is a lack of peristaltic movement in the affected bowel with loss of smooth muscle relaxation, resulting in functional constriction. In around 80% of patients, the aganglionic segment is located in the rectosigmoid area, but it may also cover a longer segment, up to the total colon or even the small bowel in rare cases[2-4]. The current method of establishing aganglionosis is by contrast enema, anorectal manometry, and rectal biopsies[2,4-6]. Of these options, rectal biopsies are considered to have the highest diagnostic specificity[4,6,7].
STANDARD SURGICAL REPAIR AND NEUROPATHOLOGICAL ANALYSIS
Surgical resection of the aganglionic segment remains the gold-standard treatment of HD[2,8]. The aim of surgery is to resect the aganglionic bowel and pull-through the bowel segment that contains a normal enteric nervous system. Therefore, it is important to check intraoperatively in which part ganglion cells are present. This is usually performed by neuropathological review of a frozen biopsy of the whole bowel circumference during open surgery or multiple biopsies during laparoscopic surgery. The most widely applied techniques to identify ganglionic and aganglionic bowel segments are hematoxylin and eosin (HE) and acetylcholinesterase (AChE) histochemistry[4,5]. Using HE staining, no ganglion cells are identified in the affected aganglionic bowel. The AChE enzyme histochemistry usually shows abundant hypertrophic extrinsic nerve fibers in the aganglionic segment[4,5]. However, both techniques have their limitations in the visualization of ganglion cells and extrinsic nerve fibers. For that reason, more and more laboratories additionally use enzyme histochemistry for the detection of ganglion cells, such as lactate dehydrogenase (LDH), nicotinamide adenine dinucleotide (NADH) tetrazolium reductase reactions, and/or succinate dehydrogenase (SDH). Most ganglion cells are stained using any of these three enzymatic techniques. Therefore, negative LDH, NADH and/or SDH staining is an additional indication of aganglionosis[10,11]. There is growing evidence that immunohistochemistry against calretinin may also be supportive in detecting aganglionic segments[12,13]. In general, it is argued that the dissection level should be at least five centimeters proximal to the biopsy location were ganglion cells were confirmed to be sure that ganglionic bowel is pulled through[4,8]. However, the transition zone may extend even longer in patients with a longer aganglionic segment, which requires a longer resection margin.
Since the first description of surgical treatment of HD, which consisted of resection of the aganglionic bowel and pulling through the ganglionic bowel by Swenson, various surgical techniques have been proposed. In recent decades, the main advances in surgery for HD have been the performance of primary repairs instead of two-staged or three-staged procedures as well as the utilization of laparoscopic and transanal techniques[16,17]. Today, the Duhamel pull-through and the endorectal pull-through are the most commonly performed procedures, without obvious preference for one of these techniques in terms of complications and outcomes. One of the long-term complications after successful surgery for HD is Hirschsprung-associated enterocolitis, but that is beyond the scope of this review.
POSTOPERATIVE BOWEL DYSFUNCTION AND TREATMENT
Despite surgical removal of the aganglionic segment, postoperative bowel function is not always favorable. Bowel dysfunction after surgery for HD can be clinically divided into constipation and fecal incontinence-related symptoms. Both are known to have a considerable impact on the physical and mental well-being of children, adolescents, and adults with HD[19-21].
Constipation was reported in 8% to 71% of patients who underwent pull-through surgery for HD. This widely varying prevalence of constipation may be explained by differences in follow-up duration and definitions of constipation. Nevertheless, a notably high rate of 22% to 33% of patients still suffer from constipation in adulthood[20,22-24], and there seems to be no clear improvement with ageing[20,25,26].
Although the causes of persistent obstructive symptoms following pull-through surgery for HD are numerous, the first differentiation is between anatomical and functional etiologies. Various reasons for the obstructive complaints have been shown, including mechanical obstruction (e.g., twisting or adhesions), persistent aganglionosis or hypoganglionosis, internal sphincter achalasia, motility disorder, and functional megacolon. Furthermore, it is known that patients with HD lack the rectoanal inhibitory reflex (RAIR), which cannot be restored by surgery and thus makes them prone to postoperative constipation[7,29]. Based on these observations, the following work-up for the diagnosis and treatment of postoperative obstruction can be recommended (Figure 1). First, following a thorough medical and surgical history, a rectal exam and contrast enema should be performed to exclude mechanical obstruction requiring revision surgery. Second, anorectal manometry may be helpful to test the presence of different autonomic reflexes. Anorectal manometry may also show functional causes for the obstructive complaints, such as pelvic dyssynergia[28,30]. Third, the proximal resection part of the HD specimen may be reviewed again by an experienced neuropathologist and/or a new rectal biopsy at the anastomotic site may be obtained to look for persisting aganglionosis or a transition zone, which might require a redo pull-through in specific cases. When neither of these two examinations show irregularities, a botulinum toxin injection may be given, which can be repeated if it works[28,31]. If botulinum toxin has no effect, the child should take part in a motility work-up program. This includes a high-fiber diet, laxatives, prokinetic agents, enemas, psychological support, biofeedback training and/or pelvic physical therapy. For intractable constipation, bowel management using transanal or (antegrade) colonic irrigation may be advised. Unfortunately, the impact of a structured work-up on persistent constipation has not yet been evaluated in a randomized setting.
Figure 1 Work-up algorithm for persistent constipation and fecal incontinence after pull-through surgery for Hirschsprung’s disease.
Postoperative fecal incontinence is described in between 1% to 50% of HD patients. In contrast to constipation, there is a clear decreasing prevalence with increasing age[20,25,26,32]. Nevertheless, 9% to 19% of adult patients still suffer from fecal incontinence[20,22-24,33], and 16% to 48% from soiling[20,22-24,34].
Fecal continence requires anal sensation and control of the anal sphincter muscles. Therefore, in the case of fecal incontinence after a pull-through procedure, the integrity of the anal canal and sphincters should be checked[31,35]. This can be performed during physical examination with or without sedation, performing anorectal manometry or endorectal ultrasound (Figure 1)[31,35,36]. In addition to sphincter damage, it is important to determine if there is either colonic hypomotility leading to overflow incontinence, or colonic hypermotility with loss of rectal reservoir function[4,31,35,36]. A contrast enema or colonic manometry can be used to disclose the presence of either hypomotility or hypermotility[31,35]. Both conditions require a completely different treatment. In the case of hypomotility, the child is in fact pseudo-incontinent and the management should be directed towards prevention of constipation[31,35]. In contrast, when the child suffers from hypermotility, the goal of treatment is to slow down colonic transit and thicken stool consistency, including a constipating diet, loperamide, bulking medication and/or retrograde enemas[4,31,35,36]. If social continence is not achieved, colonic irrigation or stoma formation may be considered as the final treatment option[4,35].
NEUROPATHOLOGICAL FINDINGS IN THE GANGLIONIC, PROXIMAL BOWEL
Persistent bowel dysfunction following successful surgery raises the issue that the presence of ganglion cells does not necessarily indicate that the proximal, ganglionic colon of HD patients is completely “healthy” and thus functional. Recent studies have confirmed structural abnormalities in the proximal, ganglionic colon of humans and animals with HD, which can be divided into: (1) Histopathological abnormalities of enteric neural cells; (2) Imbalanced expression of neurotransmitters and neuroproteins; (3) Abnormal expression of enteric pacemaker cells; (4) Abnormalities of smooth muscle cells; and (5) Abnormalities within the extracellular matrix (Figure 2).
Figure 2 Structural abnormalities in the proximal, ganglionic bowel of patients with Hirschsprung’s disease.
Histopathological abnormalities of enteric neural cells
The neural cells that are present in the proximal, remaining colon following surgery for HD may exhibit abnormalities in their appearance and/or distribution. One of the most frequently described pathological findings in the proximal, ganglionic bowel is intestinal neural dysplasia (IND). IND is inconsistently defined in the literature, but the main feature is giant submucosal ganglia[8,37]. IND has been observed in the proximal colon segment of 20 to 70% of surgically treated patients with HD. Several studies have associated IND or graded features of IND in the proximal, ganglionic colon segment of surgically treated HD patients with persistent postoperative bowel dysfunction[38-41], but others have contested this statement. Other histopathological alterations in the proximal colon segment of patients with HD include myenteric hypoganglionosis and submucosal nerve hypertrophy[8,43,44]. Additionally, several studies reported an increased amount of immature neural cells in the proximal colon of HD patients[45-47], although the clinical consequences of neural immaturity are not clear yet.
Imbalanced neurotransmitter and neuropeptide expression
Enteric neural cells communicate using neurotransmitters and neuropeptides. Peristaltic movements require both contraction and relaxation, which is mediated by a balance between excitatory and inhibitory neurotransmitters and neuropeptides. Disproportionate expression of a specific neurotransmitter or neuropeptide may therefore lead to colonic dysmotility.
The primary inhibitory neurotransmitter of the enteric nervous system is nitric oxide (NO), which is produced by nitrergic neurons that contain NO synthase (NOS). Alterations in the expression of NOS-containing neurons in the proximal, ganglionic bowel have been observed in mouse models of HD[49,50]. In addition, recent studies in humans have shown a relative overabundance of nitrergic neurons in ganglionic bowel of HD patients in comparison to healthy controls[48,50,51]. Two of these studies correlated their findings with functional outcomes of the patients, but were unable to show a significant relation with postoperative bowel function[48,50]. Nevertheless, patients with worse bowel function tended to have a greater proportion of NOS-containing neurons, a finding which requires future research with larger study populations. It has been hypothesized that the overabundance of nitrergic neurons may be an additional expression of neuronal immaturity, as these neurons are the first subtypes to appear during embryonic development. However, it remains unclear if this is a primary or secondary phenomenon.
Neuropeptides that act predominantly as inhibitory neurotransmitters are vasoactive intestinal peptide (VIP), pituitary adenylate cyclase-activating polypeptide (PACAP), galanin, and neuropeptide Y (NPY), all inducing smooth muscle relaxation. In line with the increased presence of inhibitory nitrergic neurons, VIP-immunoreactive neurons were also more frequently found in ganglionic bowel in a mouse model of HD, compared to healthy control mice. However, a more recent human study found that VIP-immunoreactive neurons were as frequently found in the proximal, ganglionic bowel of HD patients as in healthy controls. With regard to PACAP, galanin and NPY, a decreased number of immunoreactive neurons was found in the ganglionic colon in a mouse model of HD compared to healthy control mice.
One of the most widely studied excitatory neurotransmitters of the enteric nervous system is acetylcholine, which is produced by cholinergic neurons containing choline acetyltransferase (ChAT)[48,51]. Although it seems clear that the excessive expression of acetylcholine plays a crucial role in the permanent constriction of the affected, aganglionic segment, a slightly decreased expression in the proximal, ganglionic segment of both mice and humans with HD has been shown. Therefore, an inverse relationship in the expression of NOS and ChAT has been suggested[49,51]; however, future studies are needed as previous investigations consisted of small study populations and a short follow-up time.
Substance P (SP) is a well-studied excitatory neuropeptide. Its presence was significantly reduced in the proximal colon of a mouse model of HD in comparison to healthy control mice. However, other authors did not find this difference in both mice and humans. Similarly, no obviously different expression of the neuropeptide calcitonin gene-related peptide (CGRP) was found in the proximal colon of a mouse model of HD in comparison to healthy control mice.
Abnormal expression of enteric pacemaker cells
Coordinated propulsive contractions of smooth muscle cells are dependent on electrical pacemaker activity, which is provided by an intramural network of interstitial cells of Cajal (ICCs). Two other major functions of ICCs are facilitation of muscle innervation and mediation of sensory transmission. The traditional method of visualizing ICCs is by immunohistochemistry of its c-Kit membrane receptor tyrosine kinase[53,54]. By using this technique, a total reduction of ICCs in the proximal, ganglionic colon of HD patients has been observed in comparison with healthy controls. This observation has been contested by others, who did not find an overall difference in the distribution of c-Kit positive ICCs[55-58]. However, a marked variability of ICC values in patients with HD has been noted, which may be a reflection of the heterogeneous character of the disease. Furthermore, two studies linked poor clinical outcomes in patients to very low numbers of ICCs and a low ratio of ICCs to neural innervation[55,59]. More recently, the use of c-Kit has been replaced by a more specific ICC marker: anoctamin-1[53,54,60]. Use of this marker showed a moderate reduction of ICC fibers in ganglionic HD colon, compared to the colon of non-HD controls. These contradicting results are likely to arise from small study populations, but may have also been biased by the fact that the distribution of ICCs can vary with age and location in the gastrointestinal tract[53,54,60]. Therefore, the role of structural ICC abnormalities in persistent bowel dysfunction in patients with HD is still to be elucidated.
Adjacent to ICCs, platelet-derived growth factor receptor alpha positive (PDGFRα+) cells form a second network of pacemaker cells, which works closely together with ICCs and smooth muscle cells to regulate bowel motility. A striking decrease in PDGFRα+ protein expression has been found in ganglionic colon, compared to non-HD controls, which requires future research.
Abnormalities of smooth muscle cells
Coordinated colonic motility requires more than just an intact nerve network, balanced neurotransmitters and neuropeptides, and stimuli from ICCs and PDGFRα+ cells. Smooth muscle cells are the effectors of colonic motility. The chemical and electrical stimuli provoke activity of smooth muscles, leading to peristaltic movements. The response of smooth muscle cells to these stimuli is regulated by a wide variety of ion channels and receptors. Furthermore, smooth muscle cells also need an intact cytoskeleton for maintenance of muscle structural and functional integrity.
With regard to cytoskeletal proteins, there are contradictory reports. On the one hand, a lack of desmin, dystrophin, vinculin, and ε-sarcoglycan has been found in aganglionic bowel, whilst on the other hand, comparable amounts of some of these substances in ganglionic and aganglionic bowel were reported. Unfortunately, little is known about differences in the cytoskeleton between ganglionic HD colon vs healthy colon. The exception to that is a recent study by Zhu et al, which showed that HD patients with an accumulation of neurofilaments in neural cells of their proximal, ganglionic colon had a significantly worse postoperative bowel function one to six years after surgery.
In contrast, reductions of many ion channels and receptors have been reported in the ganglionic bowel of patients with HD in comparison to healthy controls: small-conductance Ca2+ activated K+ (SK3) channels[66,67], different members of the two-pore domain K+ (K2P or KCNK) channels[68,69], voltage-dependent K+ channels[70,71], hyperpolarization-activated nucleotide-gated (HCN) 3 channels, voltage-gated sodium channel type 1β (SCN1B), chloride channel subunit FXYD1, and ryanodine receptors (Ryr). Another factor which is important for the connection between smooth muscles cells and chemical and electric stimuli are gap junctions. The gap junction channels are formed by connexin proteins, of which reductions have been observed in the ganglionic colon of HD patients relative to healthy control tissue. Hence, there is mounting evidence today regarding ion channel and receptor deficiencies in the ganglionic colon of patients with HD. However, if and how all these abnormalities translate into differences in postoperative bowel function is a question that remains to be elucidated.
Abnormalities of the extracellular matrix
Around the smooth muscle cells lies a three-dimensional structure called the extracellular matrix (ECM). The composition of the ECM varies from tissue to tissue, but mainly consists of two main classes of macromolecules: collagens and glycoproteins. These ECM components serve as a structural skeleton, but also influence migration, proliferation, survival, and/or differentiation of cells. Therefore, it is not surprising that the ECM was found to have a critical role in the development of the enteric nervous system. Various studies have determined imbalances in the composition of ECM in the proximal, ganglionic colon of HD patients.
Collagens are the most abundant proteins in the ECM[76,77]. Among them is collagen IV, a major component of basement membranes surrounding smooth muscle cells in human colon[65,78,79]. Increased amounts of collagen IV have been observed in the aganglionic segment of patients with HD, and in the proximal, ganglionic segment in almost half of these patients in comparison to healthy controls. Similarly, the presence of collagen VI was two to three times greater in the proximal, ganglionic bowel segment of HD patients compared to healthy controls.
Laminins are a large family of glycoproteins, composed of α, β, and γ chains[65,76,77]. An overall increase in laminin concentration was found in the proximal, ganglionic bowel segment of patients with HD in comparison to controls[79,81]. More detailed analysis of the different laminin chains revealed an increased accumulation of the laminin α5 chain in the proximal, ganglionic bowel of almost half the studied HD patients. It has been postulated that the overabundance of laminin may be a reflection of the immature state of HD colon.
Additionally, fibronectin - another family of glycoproteins - showed a marked increase in aganglionic tissue, in comparison to the proximal, ganglionic segment[82-84]. A comparison with healthy controls is lacking to assess the overall presence of fibronectin, but an up-regulated expression of the fibronectin 1 gene has been described in the proximal, ganglionic colon of patients with HD compared to healthy colon. For two other glycoproteins, tenascin and nidogen, the intensity of the immunoreactivity in the proximal, ganglionic colon segment was similar to the colon of non-HD controls[83,86]. Aberrant expression of the above ECM components in the proximal ganglionic bowel, may have an influence on postoperative bowel dysfunction in HD, but this has not yet been proven.
Despite complete surgical resection of the aganglionic colon in patients with HD, bowel dysfunction frequently seems to persist even into adulthood with limited treatment options. At present, there is a growing body of evidence regarding neuropathological abnormalities in the proximal, ganglionic colon of surgically treated HD patients, which may have important future clinical implications. However, most studies are limited by solely reporting laboratory findings, retrospective reporting of functional outcomes based on medical reports, short follow-up periods, and small study populations. Therefore, the current knowledge is too limited to alter surgical management or to extend the standard neuropathological analyses performed.
In the short-term, it will be important to further establish the association between the structural abnormalities of the proximal, ganglionic colon and the postoperative functional outcome. Brooks et al and Zhu et al were the first to associate neuropathological findings to prospectively obtained patient-reported bowel function measures, which is a first step in the right direction. The optimum would be the combination of longitudinal patient-reported functional outcomes after surgery for HD combined with intraoperative and postoperative detailed neuropathological analysis of the proximal, ganglionic colon segment. Thus, new, innovative ways of studying the complex intramural network of different colonic cells should be explored, for example three-dimensional imaging of the enteric nervous system, which may lead to new insights. In this way, it may be possible to predict which patients might benefit from intensified follow-up and bowel management to prevent bowel obstruction and other types of postoperative dysfunction. Eventually, these studies may lead to a different evaluation and treatment of certain HD patients with persistent postoperative constipation and/or fecal incontinence.
In the long-term, it may become clearer whether certain HD patients might benefit from additional surgical resection. This would require further study on the extent of the investigated neuropathological abnormalities as well as the permanent or transient character of the structural deviations. Finally, these findings may represent a target for future therapies, but it is premature to advocate a certain direction of these therapies yet.