Pathology of Enteric Neuromuscular Disorders

Fig. 17.1
Diagnosis of Hirschsprung disease by rectal suction biopsy. (a) An adequate suction rectal biopsy should be ≥2 mm in greatest dimension and contain a generous sample of submucosa. (b) Hirschsprung disease is diagnosed based on absence of ganglion cells in exhaustive histological sections of an adequate biopsy and the presence of crowded abnormally large caliber submucosal nerves (arrows). (c) Hypertrophic submucosal nerves (arrows) have Glut1-immunoreactive perineuria similar to extra-enteric (extrinsic) nerves. Glut1 also labels erythrocytes in vessels (v). Scale bars = 40 μm

The diagnosis of HSCR is firmly established when a distal rectal biopsy with adequate submucosa shows aganglionosis and unequivocal nerve hypertrophy. However, most honest pathologists will admit to some degree of nervousness rendering a diagnosis based solely on the H&E findings in some cases. Reasons for consternation are many. Nerve hypertrophy is not a consistent finding and the distinction between adequate and inadequate submucosal sampling is arbitrary. Ganglion cells , particularly the immature ones found normally in neonates, can be difficult to distinguish from reactive endothelial cells or lymphocytes. Inflammation, not uncommon in the setting of a constipated infant who has undergone rectal examination, barium enema, and possibly other diagnostic procedures, can obscure ganglion cells. The diagnostic challenge is compounded by artifacts like compression or tissue dessication, which compromise histological resolution. Some of these problems are reduced by careful handling and expedient transportation/fixation. Unless enzyme histochemistry (see below) is planned, biopsies can be fixed at the bedside and sent to the laboratory in fixative. Any unfixed biopsies should be sent to the laboratory in a sealed container on a saline-moistened telfa pad and promptly fixed or frozen by laboratory staff.

The many challenges working with H&E-stained sections have led to many proposed ancillary histopathological approaches to evaluate rectal biopsies. Several papers have been published which tout immunohistochemistry to detect neuronal markers (e.g., PGP9.5) to facilitate recognition of ganglion cells [12], but very few laboratories employ these methods because in most cases ganglion cells , when present are fairly abundant and readily identified by H&E staining. When ganglion cells are rare, finding them requires evaluation of many sections from a given biopsy, which would require immunostains on an impractically large number of sections and/or destaining and immunostaining H&E sections with equivocal ganglion cells. In contrast, two widely utilized ancillary approaches, acetylcholinesterase (AChE) histochemistry and calretinin immunohistochemistry, detect changes in mucosal innervation which complement information gleaned from H&E sections.

Use of AChE histochemistry as a diagnostic tool for HSCR was pioneered by Meier-Ruge in the 1970s [13]. AChE is expressed on the membranes of cholinergic nerves from pelvic autonomic ganglia, which enter the distal rectum and project rostrally through all layers of the bowel wall. In the normal mucosa, these nerves are slender and sparse (Fig. 17.2a). However, in aganglionic rectum mucosal AChE-positive nerves are thick and concentrated (Fig. 17.2b) [14]. In experienced laboratories, AChE immunostaining alone appears to be a fairly sensitive and specific diagnostic approach [15, 16]. However, performance and interpretation of AChE histochemistry requires regular practice. False negative results from biopsies of premature infants or term babies less than 3 weeks of age are particularly problematic because, as with submucosal nerve hypertrophy, the density, coarseness, and extent of mucosal AChE-positive innervation increase with age [16, 17].


Fig. 17.2
Ancillary staining methods for Hirschsprung disease. (a, b) Acetylcholinesterase histochemistry highlights cholinergic nerve twigs which are sparse in normal mucosa (a) but abundant (arrows) in the mucosa overlying aganglionic rectal tissue. (c, d) Calretinin-immunoreactive ganglion cells (arrowhead) and mucosal neurites (arrows) are present in a biopsy of ganglionic rectum (c) but absent in mucosa overlying aganglionic rectal tissue (d). Round immunoreactive structures in (d) are mast cells. Scale bars: 50 μm

Multiple factors restrict use of AChE histochemistry to specific centers. Some practices, particularly those with small pediatric volumes, cannot justify the expense and effort required to maintain a histochemical assay that is used relatively infrequently. AChE histochemistry also necessitates frozen tissue, typically an additional suction biopsy, because the enzymatic activity is lost when tissues are formalin-fixed and embedded in paraffin. Acquisition of additional tissue and appropriate handling can be impediments, especially if biopsies are performed in remote clinics .

Calretinin immunohistochemistry is another ancillary method to resolve changes in mucosal innervation that correlate with aganglionosis, and has been adopted by many laboratories as an alternative or complement to AChE histochemistry. Calretinin is a calcium-binding protein expressed in a subset of submucosal and myenteric ganglion cells, including muscularis mucosae and lamina propria neurites from intrinsic neurons (Fig. 17.2c) [18]. Aganglionic bowel is devoid of calretinin-immunoreactive mucosal innervation (Fig. 17.2d), except in a 1–2 cm region immediately distal to ganglionic bowel, where neurites from the latter extend into the mucosa of the aganglionic segment [19]. Calretinin immunohistochemistry can be performed on formalin-fixed paraffin-embedded sections, so no additional biopsies or special tissue processing is required. More importantly, several studies have demonstrated equivalent or superior diagnostic specificity and sensitivity to AChE histochemistry [2022], although rare situations exist when calretinin immunohistochemistry may be misleading (e.g., very short-segment aganglionosis) [19, 23].

“Full-Thickness ” Rectal Biopsy

Full-thickness biopsies require anesthesia and may be associated with slightly higher rate of complications that suction biopsies, but are indicated in certain situations. One frequent use is to exclude HSCR in a toddler or older patient, particularly if prior suction biopsies yielded inadequate submucosa. Other indications for full-thickness biopsy include results from suction biopsies that suggest possible very short-segment HSCR and evaluation of a patient with chronic obstructive symptoms months to years after HSCR surgery to exclude transition zone pull through. Since full-thickness biopsies are performed under anesthesia, the surgeon is able to visualize the anorectal transition and any prior surgical anastomosis boundary and establish the exact location of the biopsy. Whether truly full-thickness or not, any longitudinal incisional biopsy should be oriented by the surgeon to designate the proximal and distal ends because a transition between ganglionic and aganglionic bowel may be evident along the length of a 2–4-cm-long “strip” biopsy and provide definitive evidence for very short segment HSCR. In a patient, whose obstructive symptoms persist long after a pull-through procedure, punch biopsies taken at four quadrants just proximal to the anastomosis line may be used to exclude features of transition zone pull through (e.g., partial circumferential aganglionosis, hypoganglionosis, or nerve hypertrophy), which may involve only portions of the bowel circumference [2426]. Most full-thickness biopsies are large enough to be divided and freeze a small portion including mucosa and submucosa for histochemical staining, if indicated .

Diagnosis of Intestinal Neuronal Dysplasia by Rectal Biopsy

Rectal biopsy is the principal diagnostic procedure for isolated intestinal neuronal dysplasia type B (IND). IND was first described by Meier-Ruge in patients with symptoms of Hirschsprung disease but ganglion cells in their rectal biopsies [27]. The diagnostic criteria have evolved with time, but remain based on counts of submucosal ganglion cells, as identified by enzymatic histochemical staining for lactate dehydrogenase and/or succinate dehydrogenase activities [28]. The latter, like AChE histochemistry , are performed on frozen sections, but the quantitative analysis requires a standardized section thickness (15 μm) and is subject to significant observer bias [29]. Formalin-fixed paraffin-embedded biopsies are not adequate. An overabundance of “giant” submucosal ganglia (>8 ganglion cells/ganglion in an appropriately stained section) is the primary diagnostic feature of IND. However, the proportion of giant ganglia appears to change with age [30], and the formal diagnostic criteria are not considered valid for patients under a year of age [28].

IND has also been reported in ganglionic bowel proximal to the aganglionic segment in HSCR, albeit mostly in patients less than a year of age, a finding that some studies suggest may portend a worse outcome after pull-through surgery [3137]. Recently, we used paraffin sections, immunostaining for the neural marker Hu C/D, and colonic tissue from autopsy control infants (no history of dysmotility) to establish diagnostic criteria for IND-like submucosal hyperganglionosis (IND-SH) [38]. Based on these criteria, IND-SH (deviations >3 standard deviations from controls) were observed at the proximal surgical margin of 15 % of patients with short-segment HSCR, up to 15 cm proximal to the aganglionic segment.

Considerable confusion and controversy exists regarding IND. Many have questioned the existence or clinical significance of this histopathological phenotype [30, 3943]. Skepticism is due to many factors including lack of appropriate controls, changes in diagnostic criteria, erroneous extrapolation of diagnostic criteria to H&E-stained paraffin sections, and claims increased abundance of giant submucosal ganglia is a secondary adaptation to downstream obstruction, as opposed to a primary neuropathy. At this time, it seems prudent to regard IND as an “investigational” phenotype in need of research studies with appropriate controls to validate diagnostic features and demonstrate any clinical significance. Certainly the diagnosis should not be rendered based solely on analysis of paraffin sections or outside the context of a reference laboratory, which has performed adequate internal validation studies.

Use of Rectal Biopsies to Diagnose Other Conditions

Rectal biopsy is primarily a procedure to diagnose HSCR or IND, and the deliberate search for other histopathologic etiologies for intestinal dysmotility is best approached with full-thickness or seromuscular biopsies from other parts of the gastrointestinal tract. Suction biopsies and many biopsies designated as “full-thickness” by the surgeon fail to extend into the muscularis propria , and provide no insight into the muscularis propria or myenteric plexus . Those that truly are full-thickness generally sample a very small portion of the muscularis propria and myenteric plexus. Small sample size and the normal hypoganglionic nature of the distal rectum prohibit diagnosis of hypoganglionosis and reduce the likelihood of recognizing key pathological features that are only present in a small subset of neurons or muscle cells (e.g., inclusion disorders). Furthermore, the neuromuscular microanatomy of the distal rectum is different from most of the rest of the intestines and mimics changes considered pathologic in other sites. The muscularis propria of the distal rectum is thickened and separated into discrete bundles by fibrous tissue, which also interdigitates between the muscularis externa and interna around the myenteric plexus. Elsewhere this pattern of fibrosis might suggest visceral myopathy or post-inflammatory scarring. As discussed above, except in infants and young toddlers, large submucosal nerves with extrinsic morphological features (e.g., conspicuous Glut1-immunoreactive perineurium) are normal in the distal rectum, but not more proximally. These nerves are identical in size to, but fewer in number than, the large caliber nerves observed in the transition zone of HSCR. In an infant with HSCR they are regarded by many as an indicator of physiologically abnormal bowel that can cause persistent obstructive symptoms if it is not resected during a pull-through procedure [44, 45]. In the distal rectum of an older patient, occasional large nerves are normal and aganglionosis or prior transition zone pull through should only be suspected if abundant large nerves are present.

Despite these potentially misleading features, occasionally findings in a rectal biopsy done to exclude HSCR actually lead to another specific diagnosis. While insensitive, rectal biopsy has led to accurate diagnosis of conditions associated with inclusions in ganglion cells such as mitochondrial disorders [46, 47] or neuronal nuclear inclusion disease [48]. In principle, diagnostic features of some inflammatory visceral myopathies or neuropathies might also be evident in a true full-thickness rectal biopsy, which samples a generous amount of muscularis propria and myenteric plexus. Although widespread involvement of the intestinal tract is usually present in these very rare disorders, the changes can be patchy and we have no idea how often rectal tissues are affected. More typically multiple sizable laparoscopic or open surgical intestinal biopsies are considered the diagnostic standard for adequate evaluation .

Intestinal Biopsy to Evaluate a Patient with Chronic Pseudo-obstruction

Once HSCR is excluded, most infants either resolve their symptoms or can be managed satisfactorily with dietary/medical therapy. Unfortunately other patients continue to have debilitating dysmotility or acquire chronic intestinal pseudo-obstruction as older children or adults. Clinical findings in many of these patients are difficult to distinguish from true obstruction, and it is not uncommon for them to undergo laparotomy to exclude an anatomic etiology. After anatomic cause is excluded, the recurrent nature of their disorder coupled with a history of prior abdominal exploration prompts concerns about obstruction due to abdominal adhesions, which can lead to multiple laparotomies in some instances. Diversion enterostomy is also a common, particularly for those patients with profound colonic dysmotility. Intestinal biopsy is often considered as part of any of these surgeries or sometimes as a primary diagnostic procedure . Intestinal biopsy to identify ganglionic bowel for either a leveling ostomy or primary pull-through procedure is also an integral component of HSCR management .

Diagnostic intestinal biopsies from patients with intestinal pseudo-obstruction are performed cognizant that (a) similar clinical findings may be due to numerous etiologies, not all of which have anatomic correlates, (b) diagnostic histopathological features are often patchy and may be missed with inadequate or unlucky sampling, (c) key histological findings can be mimicked or obscured by artifacts associated with improper handling, and (d) many diagnoses have prognostic or genetic implications, but will not significantly affect clinical management. No standard exists with regard to which or how many sites should be biopsied from a patient with pseudo-obstruction. Sometimes manometric data or other clinical findings suggest more severe involvement of one part of the intestinal tract. However, it is prudent to biopsy multiple sites including large and small intestine to gain information about the distribution of pathological changes and their severity/progression. If segmental dilatation is present, at least one biopsy should be from the dilated area and another from bowel immediately downstream.

An international working group recommended full-thickness biopsies at least 1.5 × 1.5 cm, with transverse closure of the surgical defect [49]. In my experience with pediatric patients, biopsies are more often rectangular or ovoid, range from 1 to 1.5 cm in greatest dimension, and provide adequate tissue for evaluation. Priority should be given to obtaining well-oriented, undamaged, formalin-fixed, paraffin-embedded tissue sections. However, it is usually possible to save a 1 mm3 sample of muscularis propria and enclosed myenteric plexus in electron microscopy fixative and snap freeze small full-thickness portions of each biopsy. Pre-surgical coordination between pathologist and surgeon is advisable and the surgeon should procure biopsies with gentle traction (typically applied with a nylon suture) and send the specimen to the laboratory for immediate processing. Orientation of the biopsy so that sections are cut perpendicular to the serosal surface is most important, and a biopsy should ideally be aligned in a true transverse or longitudinal plane.

As with rectal biopsies, H&E-stained sections provide the starting point for histological evaluation of intestinal biopsies. Generally a single slide (1–3 sections per slide) is sufficient; additional sections or special stains are obtained as needed. Trichrome-stained sections provide good contrast between smooth muscle and collagen-rich fibrous tissue, and are often helpful to resolve intramuscular fibrosis. Except as a research tool, immunohistochemistry should be guided by the clinical and H&E findings. Immunohistochemistry can help identify and quantify specific cell types (e.g., neurons, enteric glia, interstitial cells of Cajal, fibroblast-like cells, and smooth muscle) involved in enteric neuromuscular activity, and has been used to distinguish specific subtypes of enteric neurons and/or the distribution of their cell processes. However, for the most part, disease-specific alterations in the density, distribution, or intensity of immunoreactive cells have not been found.

For example, consider CD117 (c-kit) immunohistochemistry . In the gut wall, CD117 is a fairly specific marker for interstitial cells of Cajal (mast cells also express this antigen), which mediate intestinal pacemaker activity. However, CD117-positive interstitial cells are difficult to quantify, especially in tissue sections. Reduced or absent CD117-immunoreactive pacemaker cells have been reported inconsistently in multiple contexts, including diverse conditions (e.g., HSCR, post-inflammatory strictures), as what appears to be a nonspecific secondary change [50, 51]. Nonetheless, many pathologists use CD117 immunochemistry in their work-up of intestinal biopsies from patients with pseudo-obstruction, with no clear idea how results of such analysis should be interpreted. Similarly, poorly understood alterations in the densities of neurochemically defined subtypes of enteric neurons have been reported in patients with slow transit constipation, hypoganglionosis, idiopathic megacolon, transition zone of Hirschsprung disease, and congenital chronic intestinal pseudo-obstruction [52].

Additional Disorders Diagnosed by Biopsy

Some types of intestinal neuropathology that can be diagnosed from biopsies are listed in Table 17.1 and their histopathological features are briefly reviewed in the following sections.

Table 17.1
Intestinal neuromuscular pathology in intestinal biopsies


H&E findings

Ancillary pathology

Confirmatory studies

Congenital hypoganglionosis

Sparse myenteric ganglion cells organized as individual cells or pairs with minimal adjacent neuropil

May require multiple levels to confirm impression and determine extent of intestinal involvement; immunohistochemistry with neuronal markers (e.g., Hu C/D, PGP9.5) may help resolve ganglion cells

None (genetic basis unknown)

Degenerative enteric neuropathy (including inflammatory neuropathies )

Reduced density of ganglion cells without significant loss of adjacent neuropil; degenerating neurons; lymphocytic or eosinophilic ganglionitis; rarely neuronal nuclear inclusion disease

Immunohistochemistry to document intra-ganglionic T-cells; consider electron microscopy or immunohistochemistry to characterize intranuclear inclusions

Serology for anti-neuronal antibodies


Ganglioneuromatous or neurofibromatous hyperplasia of enteric plexuses with or without mucosal neuromas
Other clinical features; RET (MEN2B), NF1 , or PTEN (Cowden) mutational analysis

Mitochondrial disorders

Eosinophilic cytoplasmic inclusions (megamitochondria) in ganglion cells; atrophy of muscularis externa

Electron microscopy to document megamitochondria +/− abnormal cristae

Other clinical features; Reduced plasma thymidine phosphorylase activity in MNGIE; Mutational analysis for various established hereditary mitochondriopathies (e.g., Alpers, MNGIE)

Diffuse abnormal layering of small intestinal smooth muscle

Markedly disorganized lamination of muscularis propria (e.g., trilaminar)

Immunohistochemical demonstration of absent Filamin A

Other clinical features: FLNA mutational analysis

Megacystis microcolon intestinal hypoperistalsis syndrome

No consistent histopathological alterations

Possibly abnormal “clumping” of actin-gamma-2 in smooth muscle cells

ACTG2 or MYH11 mutational analysis

Familial visceral myopathy

No consistent histopathological alterations; fibrosis and myocyte vacuolar degeneration are common, but patchy

Possibly abnormal “clumping” of actin-gamma-2 in smooth muscle cells

ACTG2 or MYH11 mutational analysisb

Inflammatory myopathies (visceral leiomyositis)

Dense and usually diffuse lymphocytic or eosinophilic inflammation of muscularis propria

Immunohistochemistry to document intra-ganglionic T-cells

Serology to demonstrate anti-smooth muscle antibodies

aBased in part on the London Classification of gastrointestinal neuromuscular pathology [52]: clinical-histopathologic correlates with established etiologies as determined by consensus of an International Working Group. Excludes morphological abnormalities that may be clearly identifiable, but provide only weak evidence of the pathogenic mechanism and may not be causally related to an observed clinical entity

bFVM has heterogeneous etiologies and ACTG2 or MYH11 mutations are found in only a subset, including many with only nonspecific or negative histopathological findings

Congenital Hypoganglionosis

“Hypoganglionosis” denotes a reduced density of neurons relative to normal. In a literal sense, the term encompasses a wide range of possible abnormalities, including relatively small alterations in neuronal number and/or loss of selective subtypes of neurons. Even in resection specimens, such small changes are impossible to diagnose by simple analysis of H&E-stained sections and very difficult to diagnose reliably even with immunostaining and/or sophisticated types of morphometric analysis. The problem is compounded by the limited sample present in a typical intestinal biopsy, marked variation in the observed numbers of ganglion cells observed in control populations [26, 53], and uncertainty about how distension may affect ganglion cell density. For this reason, many of us only express confidence recognizing moderate-to-severe myenteric hypoganglionosis (to the best of my knowledge, submucosal hypoganglionosis per se has not been described). Myenteric hypoganglionosis can be congenital (“hypogenesis”) or acquired. Acquired forms are neurodegenerative conditions and described in the next section.

The severe and readily recognized form of congenital hypoganglionosis can be diagnosed in H&E-stained sections, provided a generous biopsy of at least one-fourth of the bowel wall circumference is obtained. The essential microscopic features are a predominance of small myenteric ganglia (one or two ganglion cells) with minimal amounts of surrounding neuropil (Fig. 17.3) [54]. Because the entire myenteric plexus is hypoplastic, the laminae of the muscularis propria are closely apposed and ganglion cells are tightly sandwiched between the two muscle layers. Ganglion cell size and cytology may be normal or relatively immature. Submucosal ganglia are usually not affected, and their density often appears to exceed that of myenteric ganglia. Immunostains are not necessary, but neural markers may help resolve immature ganglion cells and exclude aganglionosis. Reduced AChE-positive innervation has been touted as a helpful diagnostic feature, and many published studies of hypoganglionosis are from laboratories that use this technique routinely [55]. Diffuse involvement of the intestinal tract is typical, but similar features may be observed in the transition zone of HSCR. In the transition zone, particularly in short-segment HSCR, hypertrophic extrinsic nerves coexist with hypoganglionosis, whereas hypertrophic myenteric or submucosal nerves are not part of isolated congenital hypoganglionosis. The pathogenesis of congenital hypoganglionosis is unknown, but does not appear to overlap genetically with HSCR [56].


Fig. 17.3
Hypoganglionosis . (a) At low magnification, a “string” of very small myenteric ganglia (arrows) is found at the interface between the muscularis interna and externa. (b) Each small ganglion (arrow) is composed of one or two neurons with minimal neuropil. (c) Hu C/D immunostain highlights the sparse myenteric ganglion cell bodies. Scale bars: 50 μm


Dysmotility due to hyperplasia and disorganization of enteric nervous system components is recognized as part of the phenotypic spectrum of at least three hamartoma syndromes —multiple endocrine neoplasia type 2B (MEN2B) , neurofibromatosis type I (NF1 ) , and Cowden syndrome [57]. Ganglioneuromatous hyperplasia can occur with any of the three conditions, whereas intestinal neurofibromas are only associated with NF1 . These lesions can occur anywhere along the length of the bowel and involve mucosa, submucosa, or myenteric plexus, although pseudo-obstruction is most often associated with diffuse lesions that involve extra-mucosal portions of the bowel wall. Ganglioneuromatous enteric lesions have been subdivided into diffuse ganglioneuromas, ganglioneuromatous polyposis, and solitary polypoid ganglioneuroma [58]. Diffuse ganglioneuromas are composed of variable numbers of ganglion cells, glial cells, and nerve processes, and have an infiltrative growth pattern, frequently along exaggerated neural pathways in the myenteric, intramuscular, and submucosal plexuses (Fig. 17.4a–c). Diffuse ganglioneuromas are almost invariably syndromic, and often associated with similar mucosal hamartomas (Fig. 17.4d, e), but mucosal lesions alone do not necessarily imply a syndrome. Solitary polypoid ganglioneuroma is a sporadic mucosal hamartomatous lesion, which only produces dysmotility due to anatomic obstruction or intussusception. Polypoid ganglioneuromas are formed by collections of cytologically mature ganglion cells, glia, and neuropil in the lamina propria, which displace adjacent crypts or glands. The presence of many such lesions constitutes ganglioneuromatous polyposis. A syndromic basis for at least some examples of ganglioneuromatous polyposis has been suggested, but no definite syndrome or genetic association has been identified [57]. While the ganglion cells of these hamartomas are easy to recognize, the network of neural tissue that accompanies them may be difficult to distinguish from surrounding lamina propria or smooth muscle. S100 immunostain highlights the nerve processes and associated glial cells.


Fig. 17.4
Ganglioneuromatous hyperplasia (neurofibromatosis and multiple endocrine neoplasia type 2B). (a) Low magnification image shows a plexiform neurofibroma in the mesentery of the small bowel in a patient with neurofibromatosis. Ganglioneuromatous hyperplasia (arrows) is present in the underlying myenteric plexus (b) and submucosa/mucosa (c). (d, e) A mucosal ganglioneuroma (arrows) in a patient with multiple endocrine neoplasia type 2B is composed of ganglion cell bodies (arrowheads) and surrounding neuropil. Scale bars: (a) 250 μm; (b) 100 μm; (c) 100 μm; (d) 100; (e) 25 μm

Mitochondrial Disorders

Intestinal pseudo-obstruction is a frequent, sometimes severe, and occasionally initial problem for patients with hereditary mitochondrial disease. For patients with severe enteric manifestations, in addition to central nervous system pathology, the term mitochondrial neurogastrointestinal encephalomyelopathy (MNGIE) is used. Similar gastrointestinal dysfunction and pathological findings have been described in patients with mutations in at least three different genes [59], including patients with POLG1 mutations and Alpers syndrome [47]. Histopathological features of mitochondriopathy are multifocal thinning or loss of the muscularis externa and megamitochondria in enteric neurons +/− smooth muscle (Fig. 17.5). In H&E-stained sections megamitochondria are dense, eosinophilic cytoplasmic granules 1–5 μm in diameter. They are only observed in a minority of ganglion cells, sometimes less than 10 %. Less frequently they can be resolved in smooth muscle cells. Electron microscopy can help clarify that these inclusions are giant mitochondria and sometimes resolves abnormal cristae. A thorough neurological examination and other laboratory tests may reveal extra-enteric findings that help confirm the diagnosis.


Fig. 17.5
Mitochondriopathic histopathology . (a, b) H&E- (a) and trichrome- (b) stained sections show near complete effacement of the muscularis externa (me) by fibrous tissue (blue in b) with less severe atrophy of the muscularis interna (mi). (c) Dense eosinophilic granules (megamitochondria) are present in a subset of enteric ganglion cells (arrowhead). Scale bars: (a) 100 μm; (b) 100 μm; (c) 50 μm

Diffuse Abnormal Layering of Small Intestinal Smooth Muscle (X-Linked Pseudo-obstruction)

An X-linked form of familial intestinal pseudo-obstruction was recognized several decades ago and recently shown to be caused by mutations in the Filamin A gene ( FLNA ) [60, 61]. Affected males usually have one or more other congenital anomaly (e.g., cerebral periventricular heterotopias, atrial septal defect, cleft palate) and some are thrombocytopenic. All patients have intestinal malrotation and congenital short small bowel (CSSB ) . Alterations in the density and relative numbers of argyrophilic and argyrophobic ganglion cells have been described, albeit inconsistently, and led to the impression that the disorder is a primary neuropathy [62]. However, stronger evidence now exists for a primary myopathic basis [63]. FLNA is expressed in intestinal smooth muscle, not neurons, and expression is lost in males with FLNA mutations and pseudo-obstruction. Histologic sections of well-oriented biopsies demonstrate diffuse foci of disorganized lamination of the small intestinal muscularis propria, including trilaminar architecture (Fig. 17.6). Colonic biopsies from a teen patient showed a unique pattern of myocyte multinucleation in the innermost layers of the muscularis interna [63]. Although abnormal layering has been observed throughout the small intestine in those few cases with extensive sampling, intact lamination is present in some areas and diagnostic features could be missed with a small biopsy. Therefore, immunohistochemistry and/or mutational analysis should be considered for a male patient with CSSB .


Fig. 17.6
Filamin A-related visceral myopathy (X-linked intestinal pseudo-obstruction). (a) H&E-stained section from an area of abnormal lamination in the small intestinal muscularis propria shows a vaguely trilaminar architecture. (b) Filamin A immunohistochemistry demonstrates dramatic loss of muscular immunoreactivity, in comparison to the diffuse dense cytoplasmic immunoreactivity in a section of normal control bowel (c). Scale bars: 100 μm

CSSB and intestinal malrotation also result from recessive mutations in the autosomal gene, Coxsackie and adenovirus-receptor like membrane protein ( CLMP ) [64]. However, neither pseudo-obstruction nor abnormal smooth muscle lamination is part of the phenotype in CLMP -related CSSB .

Degenerative Enteric Neuropathy

The London classification system for gastrointestinal neuromuscular pathology recognizes neuronal degeneration with or without associated inflammation of ganglia as an etiology for intestinal pseudo-obstruction [52]. Recognition of neuronal degeneration is subjective, and one should be wary about a diagnosis based on subtle cytological changes like nuclear condensation, cytoplasmic hypereosinophilia, cellular vacuolization, or irregular cell contours. Unequivocal forms of neuronal degeneration are associated with one or more of the following: moderate-to-severe hypoganglionosis, lymphocytic or eosinophilic ganglionitis, pathological intranuclear or cytoplasmic inclusions, and nuclear pyknosis or fragmentation. Although inflammatory cells often cluster in the periganglionic space between the muscularis interna and externa, it is rare to find lymphocytes or eosinophils within ganglia. Even in the context of transmural inflammation related to mucosal injury or inflammatory bowel disease, the proportion of inflammatory cells within ganglia is usually small. An exception is in the transition zone of some patients with HSCR, where concentrated intra- or peri-ganglionic eosinophilic inflammation may be present with minimal inflammation elsewhere [65]. As opposed to primary eosinophilic ganglionitis, these HSCR-associated infiltrates are not accompanied by degenerative cytopathology of ganglion cells and are not known to affect neuronal loss, clinical outcome, or motility.

Only gold members can continue reading. Log In or Register to continue

Aug 29, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on Pathology of Enteric Neuromuscular Disorders

Full access? Get Clinical Tree

Get Clinical Tree app for offline access