An immunohistochemical study of glial fibrillary acidic (GFA) protein and S-100 protein in the colon affected by Hirschsprung's disease

Diagnosis of Hirschsprung disease by analyzing acetylcholinesterase staining using artificial intelligence

OBJECTIVES

Classical Hirschsprung disease (HD) is defined by the absence of ganglion cells in the rectosigmoid colon. The diagnosis is made from rectal biopsy, which reveals the aganglionosis and the presence of cholinergic hyperinnervation. However, depending on the method of rectal biopsy, the quality of the specimens and the related diagnostic accuracy varies substantially. To facilitate and objectify the diagnosis of HD, we investigated whether software-based identification of cholinergic hyperinnervation in digitalized histopathology slides is suitable for distinguishing healthy individuals from affected individuals.

METHODS

N = 190 samples of 112 patients who underwent open surgical rectal biopsy at our pediatric surgery center between 2009 and 2019 were included in this study. Acetylcholinesterase (AChE) stained slides of these samples were collected and digitalized via slide scanning and analyzed using two digital imaging software programs (HALO, QuPath). The AChE-positive staining area in the mucosal layers of the intestinal wall was determined. In the next step machine learning was employed to identify patterns of cholinergic hyperinnervation.

RESULTS

The area of AChE-positive staining was greater in HD patients compared to healthy individuals (p < 0.0001). Artificial intelligence-based assessment of parasympathetic hyperinnervation identified HD with a high precision (area under the curve [AUC] 0.96). The accuracy of the prediction model increased when nonrectal samples were excluded (AUC 0.993).

CONCLUSIONS

Software-assisted machine-learning analysis of AChE staining is suitable to improve the diagnostic accuracy of HD.

Automatic ganglion cell detection for improving the efficiency and accuracy of hirschprung disease diagnosis

Histopathologic diagnosis of Hirschsprung's disease (HSCR) is time consuming and requires expertise. The use of artificial intelligence (AI) in digital pathology is actively researched and may improve the diagnosis of HSCR. The purpose of this research was to develop an algorithm capable of identifying ganglion cells in digital pathology slides and implement it as an assisting tool for the pathologist in the diagnosis of HSCR. Ninety five digital pathology slides were used for the construction and training of the algorithm. Fifty cases suspected for HSCR (727 slides) were used as a validation cohort. Image sets suspected to contain ganglion cells were chosen by the algorithm and then reviewed and scored by five pathologists, one HSCR expert and 4 non-experts. The algorithm was able to identify ganglion cells with 96% sensitivity and 99% specificity (in normal colon) as well as to correctly identify a case previously misdiagnosed as non-HSCR. The expert was able to achieve perfectly accurate diagnoses based solely on the images suggested by the algorithm, with over 95% time saved. Non-experts would require expert consultation in 20–58% of the cases to achieve similar results. The use of AI in the diagnosis of HSCR can greatly reduce the time and effort required for diagnosis and improve accuracy.

Presence of intramucosal neuroglial cells in normal and aganglionic human colon

The enteric nervous system (ENS) is composed of neural crest-derived neurons (also known as ganglion cells) the cell bodies of which are located in the submucosal and myenteric plexuses of the intestinal wall. Intramucosal ganglion cells are known to exist but are rare and often considered ectopic. Also derived from the neural crest are enteric glial cells that populate the ganglia and the associated nerves, as well as the lamina propria of the intestinal mucosa. In Hirschsprung disease (HSCR), ganglion cells are absent from the distal gut because of a failure of neural crest-derived progenitor cells to complete their rostrocaudal migration during embryogenesis. The fate of intramucosal glial cells in human HSCR is essentially unknown. We demonstrate a network of intramucosal cells that exhibit dendritic morphology typical of neurons and glial cells. These dendritic cells are present throughout the human gut and express Tuj1, S100, glial fibrillary acidic protein, CD56, synaptophysin, and calretinin, consistent with mixed or overlapping neuroglial differentiation. The cells are present in aganglionic colon from patients with HSCR, but with an altered immunophenotype. Coexpression of Tuj1 and HNK1 in this cell population supports a neural crest origin. These findings extend and challenge the current understanding of ENS microanatomy and suggest the existence of an intramucosal population of neural crest-derived cells, present in HSCR, with overlapping immunophenotype of neurons and glia. Intramucosal neuroglial cells have not been previously recognized, and their presence in HSCR poses new questions about ENS development and the pathobiology of HSCR that merit further investigation.

In vivo detection of Hirschsprung's disease by optical coherence tomography in rats

Hirschsprung's disease (HSCR) is a developmental intestinal obstruction, which is often diagnosed with a repeated biopsy. Optical coherence tomography (OCT) is a noninvasive, real-time imaging modality. This study aims to investigate the feasibility of diagnosis of HSCR, the targeted biopsies of suspicious tissues and the location of operative treatment using OCT. An HSCR Sprague-Dawley (SD) rat model (benzalkonium chloride-treated (BAC-treated)) was used. Colon tissues with BAC-treated and without BAC-treated were imaged using OCT. To establish OCT criteria for identification of HSCR, OCT images were compared with corresponding histology images and muscle layer thickness was measured. Furthermore, attenuation coefficients of OCT signals were calculated to illustrate the differences between tissues with BAC-treated and without BAC-treated. Our results show that OCT images of colon tissues with HSCR are well correlated with histology images. In comparison with a muscle layer without HSCR, the thickness of muscle layer with HSCR is increased significantly. The muscle layer in colon tissues with HSCR for 6 weeks had a higher attenuation coefficient than those without HSCR. However, the attenuation coefficient of those with HSCR for 3 weeks had no obvious change. In conclusion, the study demonstrates for the first time that OCT has the potential for diagnosis, biopsy and location of HSCR in vivo.

Activation of MAP kinase (ERK1/2) in human neonatal colonic enteric nervous system

The aim of this study was to examine mitogen-activated protein kinase (ERK1/2) activation in the human neonatal colonic enteric nervous system. For this, we investigated by immunocytochemistry the cellular localization of phosphorylated ERK1/2 (P-ERK) in a series of normal human colon samples removed from newborns and in patients with intestinal obstruction such as Hirschsprung's disease (HSCR), stenosis and atresia. We checked the presence of P-ERK in the three distinct histological layers of normal colon. Phosphorylated ERK was detected in the colonic mucosa, in the enteric nervous system and in endothelial cells. In the mucosa from normal colon, P-ERK was detected at the upper part of the crypt, while P-ERK activation in epithelial cells is altered in HSCR, stenosis and atresia. In the normal colon, strong P-ERK staining was detected in myenteric and submucosal enteric plexuses. Using confocal microscopy analyses, we observed that P-ERK staining was localized in enteric glial cells and not in enteric neurons. Strong P-ERK staining was also observed in plexuses from stenosis and atresia whereas in HSCR, hypertrophic nerve fibres were not stained.

Complete innervation profile of whole bowel resected at pull-through for Hirschsprung's disease. Unexpected findings

We used Acetylcholinesterase (AchE) staining and neural cell adhesion molecule (NCAM) immunoreactivity to examine such resected lengths to determine the complete innervation profile of resected bowel in Hirschsprung's disease (HD). Resected specimens of colon obtained at pull-through surgery from 15 patients with HD [short type (S-type: n=5), recto-sigmoid type (RS-type: n=5), long type (L-type: n=5)] were sectioned at 1.5 cm intervals and stained conventionally with AchE histochemistry and NCAM immunohistochemistry. The number of positive nerve fibers (PNFs) in the lamina propria and smooth muscle layers was assessed on a scale of 0 to 3 where 0 meant no PNF and 3 meant many PNFs. The three types of HD had different AchE and NCAM innervation profiles, especially the L-type. There were also different AchE and NCAM innervation patterns seen within the same aganglionic or transitional segments of bowel depending on the site of sampling. The mean proportion of transitional segment in resected specimens from RS-type HD was significantly larger than that of S-type HD (P<0.001) although the proportion of aganglionic segments from S- and RS-type HD were almost the same. Our results suggest that the etiology of L-type HD may be different from the etiology of S- and RS-type HD because of different innervation profiles. Because segments of the excised bowel would appear to have different innervation patterns depending on whether the specimen is sampled proximally or distally, it is important to record the exact site of sampling to allow valid comparisons between types of HD to be made.

Electrophysiological properties of the aganglionic segment in Hirschsprung's disease

BACKGROUND

In Hirschsprung's disease, the severity of bowel obstruction varies among those patients who have the affected colon of a similar length, suggesting that there is more than a simple aperistaltic obstruction in the pathophysiology of Hirschsprung's disease.

METHODS

A series of our electrophysiological studies of the aganglionic segments from human specimens and rat models were reviewed to obtain an overview of Hirschsprung's disease.

RESULTS

In human studies, a generation of regular spontaneous activity was recorded in both the dilated ganglionic segment and transitional aganglionic region, while the smooth muscle cells of the narrow aganglionic segment were electrically quiescent. According to a pattern of innervation, in the dilated ganglionic segment inhibitory junction potentials associated with or without excitatory junction potentials were observed in all of the examined cells, and these intrinsic nervous inputs were gradually decreased in the transitional region. In the narrow aganglionic segment, only excitatory junction potentials of the extrinsic nervous origin were found in about 20% of the examined cells In rat models, distally increasing tendency of the excitatory nervous inputs was observed in the narrow aganglionic segment.

CONCLUSION

A bowel obstruction in Hirschsprung's disease might be generated due to complex mechanisms involving myogenic and neurogenic abnormalities.

Age-dependent differences in the effects of GDNF and NT-3 on the development of neurons and glia from neural crest-derived precursors immunoselected from the fetal rat gut: expression of GFRalpha-1 in vitro and in vivo

No enteric neurons or glia develop in the gut below the rostral foregut in mice lacking glial cell line-derived neurotrophic factor (GDNF) or Ret. We analyzed the nature and age dependence of the effects of GDNF and, for comparison, those of NT-3, on the in vitro development of the precursors of enteric neurons and glia. Positive and negative immunoselection with antibodies to p75(NTR) were used to isolate crest-derived and crest-depleted populations of cells from the fetal rat bowel at E12, 14, and 16. Cells were typed immunocytochemically. GDNF stimulated the proliferation of nestin-expressing precursor cells isolated at E12, but not at E14-16. GDNF promoted the development of peripherin-expressing neurons (E12 >> E14-16) and expression of TrkC. GDNF inhibited expression of S-100-expressing glia at E14-16. NT-3 did not affect cells isolated at E12, never stimulated precursors to proliferate, and promoted glial as well as neuronal development at E14-16. GFRalpha-1 was expressed both by crest- and non-crest-derived cells, although only crest-derived cells anchored GFRalpha-1 and GFRalpha-2 (GFRalpha-1 >> GFRalpha-2). GDNF increased the number of neurons anchoring GFRalpha-1. GFRalpha-1 is immunocytochemically detectable in neurons of the E13 intestine and persists in adult neurons of both plexuses. We suggest that GDNF stimulates the proliferation of an early (E12) NT-3-insensitive precursor common to enteric neurons and glia; by E14, this common precursor is replaced by specified NT-3-responsive neuronal and glial progenitors. GDNF exerts a neurotrophic, but not a mitogenic, effect on the neuronal progenitor. The glial progenitor is not maintained by GDNF.

Abnormal internal anal sphincter innervation in patients with Hirschsprung's disease and allied disorders

The morphology of the intrinsic innervation of internal anal sphincter (IAS) in Hirschsprung's disease (HSCR) and allied disorders has not been clearly defined. At the time of IAS myectomy, specimens of the IAS were taken from four patients with HSCR, five patients with intestinal neuronal dysplasia (IND), five patients with IAS achalasia, and two patients with hypoganglionosis. Specimens also were taken from five normal controls. The specimens were examined using neural cell adhesion molecule (NCAM) immunohistochemistry, NADPH-diaphorase histochemistry, and acetylcholinesterase (AChE) histochemistry. The number of AChE-positive nerve fibers was markedly increased in the IAS of patients with HSCR, IND, and IAS achalasia compared with controls. NCAM and NADPH-diaphorase activity was absent or markedly reduced in the IAS of patients with HSCR, IND, and IAS achalasia. The IAS of patients with hypoganglionosis show markedly reduced NCAM and NADPH-diaphorase activity and occasional AChE-positive nerve fibers. These findings show that patients with HSCR, IND, hypoganglionosis, or IAS achalasia have abnormal innervation of the IAS and this may contribute to disturbances in gut motility in these conditions.

Regional differences in the number of neurons in the myenteric plexus of the guinea pig small intestine and colon: an evaluation of markers used to count neurons

BACKGROUND

Subsets of myenteric neurons have been identified. To determine the proportional representation of neurons in each, it is necessary to relate the number of neurons in the subset to that of the complete set. Prior estimates of total numbers of neurons, obtained with many different markers, have varied widely.

METHODS

Markers were compared for counting myenteric neurons in dissected laminar preparations of guinea pig duodenum, jejunum-ileum, and colon; the effect of stretching preparations on these counts was also determined. Markers included the visualization of single-stranded nucleic acid with cuprolinic blue and the immunocytochemical demonstration of neuron specific enolase (NSE), PGP9.5, S-100, and the constitutive expression of a Fos related antigen (FRA).

RESULTS

Neurons could not be counted accurately by demonstrating NSE, PGP9.5, or S-100. The number of neurons detected by demonstrating FRA was consistently less than that determined with cuprolinic blue (approximately 65%). Cuprolinic blue-derived estimates of neuron numbers were higher than most reported in the literature, but comparable to those recently obtained with "a nerve cell body" antiserum. Ganglionic area was found to be stretch independent. The rank order of neurons/cm2 and ganglionic area/ unit resting length was colon > duodenum >> jejunum-ileum; more neurons were found in the myenteric plexus of the colon (7.3 x 10(6)) than in that of the entire small intestine (6.5 x 10(6)).

CONCLUSIONS

Prior studies that have obtained denominators for estimating the proportions of myenteric neuronal subsets with markers that do not reveal the entire population should be re-evaluated. The guinea pig colon contains a surprisingly large number of neurons, the physiological significance of which must be determined.

Perifollicular clear space under skirt-like epithelial structure of human small vellus hair follicle

Human hair follicles, especially small vellus hair follicles, showed characteristic structures as follows: (I) small vellus hair follicles were surrounded with skirt-like epithelial structures or perifollicular connective tissue capsule; (II) perifollicular clear space was present between the outer root sheath and skirt-like structure, or the outer root sheath and perifollicular connective tissue capsule; (III) perifollicular clear space was filled with a loose, alcian blue positive mucinous substance and elongated fibroblasts and mast cells; (IV) perifollicular nerve endings were attached to the follicular wall in a palisading arrangement within the clear space under the skirt; (V) perifollicular nerve end organs and connective tissue capsule were positively immunostained for CD34 and protein gene product 9.5 (PGP9.5). These specialized structures of human small vellus hair follicles seem to correspond to the blood sinus of the sinus hair follicles, where similar structural elements were found.

Nitric oxide synthase distribution in the enteric nervous system of Hirschsprung's disease

BACKGROUND

Hirschsprung's disease is characterized histologically by aganglionosis and functionally by impaired relaxation of the gut. Nitric oxide has recently been described pharmacologically as a major inhibitory mediator in the gut musculature in laboratory animals. The present study hypothesized that NO could be involved in motility disorders of the human gut. The aim was to study the neuronal enzyme synthetizing NO in the ganglionic and aganglionic human gut.

METHODS

Reduced nicotinamide adenine dinucleotide phosphate-diaphorase histochemistry and immunohistochemistry with a specific polyclonal antiserum were used to examine NO synthase distribution in the enteric nervous system in six patients with Hirschsprung's disease and five control patients.

RESULTS

NO synthase was selectively absent in the plexus area and in the musculature of the aganglionic segments, whereas moderate staining was observed in the hypertrophied nerve bundles in the submucosa. In contrast, in the ganglionic segment NO synthase was abundantly present, in a pattern similar to that of the normal colon.

CONCLUSIONS

These findings suggest the involvement of NO in the physiopathology of Hirschsprung's disease.

Epithelial skirt and bulge of human facial vellus hair follicles and associated Merkel cell-nerve complex

Many morphological variations of bulge areas, such as knob-like swellings, were found in extracted human facial vellus hairs. In anagen vellus hair of the face bulge areas including these knobs had a band-like dense aggregation of CAM5.2 (K8, 52.5 kDa) reactive Merkel cells. In telogen hair the bulge became indistinguishable from the clubbed or regressed end of the follicle but Merkel cells continued to be abundant. The epithelial hood at sebaceous gland level showed most commonly a skirt-like structure but variations were also observed; these were bamboo joints, tulip flower, and long apron configurations. Merkel cells were found sparsely in these structures. Palisading stockade-like nerve endings were observed surrounding the follicular epithelium under the skirt and around the bulge areas including the knobs. Merkel cells were sparse in the follicular segment corresponding to the attachment of stockade nerve endings.

Hirschsprung's disease: clinical and experimental observations

Hirschsprung's disease (HD) is a relatively common cause of intestinal obstruction in the newborn. It is characterized by an absence of ganglion cells in the distal bowel beginning at the internal sphincter and extending proximally for varying distances. The etiology of HD-associated enterocolitis remains a complex issue. This study has provided further support for a possible infectious etiology of enterocolitis complicating HD.

Atypical vertebral artery in a patient with an intra-and extraspinal cervical neurenteric cyst

A 15 year old girl experienced Horner's syndrome on the right side, paraesthesia of the right arm and meningitis. CT and MRI showed a cystic lesion on the right side of the cervical spine at C3/C4 which descended with an extraspinal portion down to C6. Histology revealed a complex neurenteric cyst. The ipsilateral vertebral artery showed an atypical course. This abnormal artery besides a partial fusion of the vertebrae C2/C3/C4 suggest a complex malformation at the site of the cyst. Disturbed developmental relationships in this case indicate that pathological blood vessels may represent a risk factor in the surgical treatment of neurenteric cysts.

Distribution of neuropeptide Y-like immunoreactivity in the normoganglionic and aganglionic segments of human colon

The localisation and distribution of neuropeptide Y (NPY)-like immunoreactivity were studied by use of immunohistochemical methods in gut tissues from 19 patients with Hirschsprung's disease, including 4 cases of long segment aganglionosis. In the normoganglionic segment, immunoreactive cell bodies and nonvaricose processes were seen within both myenteric and submucous plexuses. A scarce supply of varicose fibres was found in the lamina propria mucosae, muscularis mucosae and longitudinal muscle layer. NPY fibres were more frequently encountered in the circular muscle layer, although with a weakly immunostaining intensity. In addition, blood vessels in the submucosal connective tissue were surrounded by a typical plexus of varicose, NPY-positive fibres. Immunoreactive endocrine cells could be detected in the colonic epithelium. In the aganglionic segment, numerous nerve fasciculi comprising a small to moderate number of NPY fibres with varicosities were observed throughout the entire layer of the colonic wall. A few varicose, NPY-positive fibres were also contained in the relatively large, hypertrophic nerve fasciculi located in the intermuscular zone and submucosal connective tissue. NPY-immunoreactive fasciculi were more densely distributed in the distal aganglionic segment than in the proximal aganglionic one. On the other hand, the distribution of NPY-positive fibres in long segment aganglionosis was quite different from that in short segment type; in cases of long segment type, no immunoreactive nerve fibres were detected within the circular muscle layer of the proximal aganglionic segment near the oligoganglionic segment and only a few fibres were observed within the hypertrophic nerve bundle of the intermuscular zone.(ABSTRACT TRUNCATED AT 250 WORDS)

Pathological features of multiple endocrine neoplasia type IIb in childhood

The features of two patients with multiple endocrine neoplasia type IIb are described. Patient 1, a 9-year-old boy with marfanoid features, presented with chronic constipation and failure to thrive since infancy. Patient 2, a 12-year-old boy with marfanoid features, presented with a five-year history of persistent cervical lymphadenopathy. In patient 1, the myenteric and submucosal nerve plexuses at all levels of the small and large intestines were comprised of diffusely disorganized, hyperplastic, mature ganglion cells and nonmyelinated nerve fibers. Nerve plexus dissection with morphometric analysis showed marked thickening of the myenteric plexus with a quantitative increase in neural tissue. Patient 2 had a submucosal neuroma of the tongue. Both patients had occult medullary thyroid carcinoma, and patient 2 had cervical lymph node metastases. Both neoplasms showed positive staining for cytokeratin, carcinoembryonic antigen, calcitonin, bombesin, chromogranin, serotonin, and Leu 7. Electron microscopy showed membrane-bound, intermediate-sized, dense-core neurosecretory granules in tumor cells. In patient 2, calcitonin-positive amyloid was present with localization of calcitonin by immunoelectron microscopy to cytoplasmic secretory granules and to extracellular amyloid fibrils. These cases illustrate the potential for missed or delayed diagnosis in multiple endocrine neoplasia syndromes.

Glial fibrillary acidic protein and cartilage

The presence of glial fibrillary acidic protein was tested for in cartilage of bronchi, trachea, pulmonary hamartomas, articular cartilage and chondrosarcomas. The cytoplasm of most chondrocytes in bronchi stained strongly positive, whilst in hamartomas only small foci at the edges reacted. Staining in the trachea was weak. In chondrosarcomas a few cells were positive, but most areas were negative. Articular cartilage was consistently negative. Young chondrocytes expressed the antigen more strongly than mature cartilage. Dedifferentiated cartilage cells such as chondrosarcoma react, in contrast to their cells of origin in articular cartilage.

Distribution and localization of glial fibrillary acidic protein in colons affected by Hirschsprung's disease

The distribution and localization of glial fibrillary acidic (GFA) protein were examined by means of immunohistochemistry in normoganglionic, oligoganglionic, and aganglionic segments of colons from 25 patients with Hirschsprung's disease, including four cases of long segment aganglionosis. In normoganglionic segments, GFA protein-positive glial cells were densely distributed within the myenteric plexus, but sparse in the submucous plexus. Aganglionic segments were completely devoid of glial cells with GFA protein immunoreactivity, coinciding with the lack of enteric ganglia. Instead, GFA protein was found specifically in association with the hypertrophic nerve fasciculi and their branches, which were mainly located in the intermuscular zone and submucosal connective tissue in the distal aganglionic segment of diseased bowels. However, two types of short and long segment aganglionosis differed in the distribution pattern of GFA protein; the extrinsic nerve fasciculi in short segment disease extended toward the normoganglionic segment, but in long segment disease they did not reach this area. A moderate number of GFA protein-positive fasciculi were observed within the circular muscle layer of proximal aganglionic and oligoganglionic parts in short segment aganglionosis, while no immunoreactive fasciculi were encountered within the circular muscle layer of the corresponding parts in long segment aganglionosis. Immunohistochemistry for GFA protein can be of excellent diagnostic value for the aganglionic colon with Hirschsprung's disease, since GFA protein immunohistochemistry discloses exclusively extrinsic, hypertrophic nerve fasciculi, characteristic of the bowel in cases of Hirschsprung's disease.

An immunohistochemical study of somatostatin-containing nerves in the aganglionic colon of human and rat

The distribution of somatostatin-like immunoreactive (SOM-LI) nerves was elucidated immunohistochemically in the gut tissues from patients with Hirschsprung's disease and congenital aganglionosis rats. In the normoganglionic human colon, SOM-LI nerve cell bodies were found to a greater extent in the submucous plexus and to a lesser extent in the myenteric plexus. However, they were rarely observed in both the plexuses of the oligoganglionic segment. SOM-LI nerve fibres were widely distributed in the aganglionic bowel. The circular muscle layer of the distal aganglionic segment was densely innervated by SOM-LI nerve fibres which are probably derived from the extrinsic, hypertrophic nerve bundles. A decreased number of the intramuscular nerves fibres were seen in the proximal aganglionic segment. In the colon and rectum from adult and 21-day-old rats, SOM-LI cell bodies were numerous in both plexuses. On the other hand, enteric neurons were completely lacking from the colon and rectum of congenital aganglionosis rats of 21 days old. No neuronal elements staining for SOM were disclosed in these aganglionic segments of mutant rats. A possible origin and pathophysiological role of the extrinsic nerve fibres containing SOM in the diseased bowel are discussed. It is concluded that SOM-LI nerves in the human distal colon comprise both intrinsic and extrinsic elements, while SOM nerves in the rat colon and rectum are of only intrinsic origin.