Introduction

Unlike Hirschsprung’s disease (HD), on which the pathogenesis is well-established and numerous reports have been published [1, 2], the pathogenesis of isolated hypoganglionosis (IHG), an “allied disease of HD” (ADHD), has not yet obtained global consensus due to the dearth of reports on IHG [3]. As revealed in a nationwide survey in Japan [4], which identified 104 cases of congenital IHG (CIHG) but only 8 cases of acquired IHG (AIHG) over 10 years, CIHG and AIHG are rare diseases. In view of its rarity, there is a lack of defining histological criteria, and therefore, histological criteria for the diagnosis of IHG need to be established.

We previously reported IHG as a distinct entity, using conventional hematoxylin and eosin (H&E) staining, and proposed two categories: CIHG and AIHG [5, 6]. AIHG is histopathologically characterized by degeneration of ganglion cells, decrease in the number of ganglion cell, and gliosis in Auerbach’s plexus with preservation of size [5]. However, degenerating ganglion cells sometimes resemble glial cells, and the identification of ganglion cells on HE stains is often difficult due to variation in morphology in relation to neural maturation and variable section quality and depending on the plane of section through a ganglion cell body [7].

Some authors have claimed that immunohistochemical staining for S-100, bcl2, c-kit, and CD56 is useful for the definitive pathological diagnosis of ADHD, which includes entities such as chronic intestinal pseudo-obstruction (CIPO), immaturity of ganglia, intestinal neuronal dysplasia (IND), and IHG [8–10]. However, no studies comparing histological characteristics and immunohistochemical staining patterns of HD, CIHG, and AIHG have been reported.

Hu C/D protein is routinely used as a neuronal marker in both the central and peripheral nervous systems, including the enteric nervous system [11]. Hu C/D staining facilitates recognition of individual neurons [7, 12]. CD56 is a neural cell surface adhesion glycoprotein (NCAM), involved in adhesion between neural cells and their processes and in the initial formation of the neuromuscular synapse [10, 13, 14].

In the present study, we set out to define criteria for a histopathological diagnosis of AIHG, using histological parameters and immunostaining patterns for Hu C/D and CD56, allowing to differentiate between AIHG and CIHG.

Patients and methods

Patients and tissue specimens

Paraffin-embedded intestinal specimens obtained from 79 patients diagnosed with HD (n = 49), CIHG (n = 25), and AIHG (n = 5) from January 1996 to December 2015 were collected at our department and from other hospitals, as part of clinical and histopathological consultation. Histopathological diagnoses were confirmed by revision of HE-stained sections and immunohistochemical staining for Hu C/D and CD56.

Histological definitions

HD was defined by the absence of ganglion cells and Auerbach’s plexus in the aganglionic distal segment (AGS) and the presence of a normal number of ganglion cells with normal size of Auerbach’s plexus in the proximal normoganglionic segment (NGS). CIHG was defined by a reduced number of ganglion cells in Auerbach’s plexus and a reduced plexus area and plexus length [15–17]. AIHG was defined as a reduced number of ganglion cells with neuronal degeneration in Auerbach’s plexus but preservation of plexus size [5].

Histological procedure

Formalin-fixed paraffin-embedded 4-μm tissue sections of full-thickness intestinal samples were stained for microscopic evaluation with H&E. For immunohistochemical staining, after inhibition of endogenous peroxidase, sections were exposed to primary antibodies at 4 °C overnight, followed by streptavidin–biotin–peroxidase (Nichirei, Tokyo, Japan). We used 3,3′-diaminobenzidin as chromogen and hematoxylin as counterstained. As primary antibodies, we used anti-Hu C/D (16A11, monoclonal, 1:400; Life Technologies, Carlsbad, USA) and anti-CD56 (1B6, monoclonal, 1:50; Leica, Wetzlar, Germany).

Histological assessment

In H&E-stained sections, ganglion cells are characterized by large nuclei with prominent nucleoli, surrounded by granular, amphophilic cytoplasm [18]. Auerbach’s plexus is characterized by ganglion cells, glial cells, and/or nerves positioned between the circular and longitudinal layers of the muscularis propria. The ganglion cells stained positive for Hu C/D as earlier reported [7], using as criteria for staining specificity dark brown granular staining of a nucleated perikaryon, covering the nucleus or encircling at least 50% of the nuclear circumference. In the absence of an unambiguous nucleus, the cell was not included. For evaluation of the size of Auerbach’s plexus by CD56 staining, only the positive area between the circular and the longitudinal layer of the muscularis propria was considered because nerve fibers, nerve bundles, smooth muscle fibers, and natural killer cells also stain for CD56 [14].

For quantitative analysis, the number of ganglion cells per 1 mm of intestine in longitudinal direction was counted, and the total area of Auerbach’s plexus per 1 mm of intestine was measured. These morphological analyses were performed using the HYBRID CELL COUNT software program (KEYENCE, Osaka, Japan). We examined in HD 36 sections of NGS and 36 of AGS, and furthermore 6 sections for CIHG and 10 for AIHG.

Statistical analysis

Patient demographic data and results of the morphological analyses are presented as the median (range) and the mean ± standard error (SE), respectively. Fisher’s exact test and an analysis of variance (ANOVA) with the Tukey–Kramer method were used for comparisons among more than two groups with the JMP® 11 software program (SAS Institute Inc., Cary, NC, USA). A p-value less than 0.05 was considered to indicate a statistically significant difference and is indicated by a single asterisk.

Results

Hirschsprung’s disease

Demographic data and clinical findings of the 49 HD patients are shown in Table 1. Median age at time of the diagnosis was 150 days (range, 14–18,240). The male to female ratio was 38:11. Of the specimens, 44 concerned a bowel resection specimen and five were biopsies. These were from the distal colon (n = 27), mid colon (n = 14), proximal colon (n = 4), small intestine (n = 2), stomach, jejunum, and ileum (n = 1), and ileum and mid colon (n = 1). Gender distribution and specimen location were significantly different from those of the other groups (p < 0.05).

Table 1 The clinical differentiation of HD, CIHG, and AIHG
Full size table

Histopathologically, in HD, a normal number of Hu C/D-positive ganglion cells and normal size CD56-positive Auerbach’s plexus were found in the normoganglionic segment (NGS), indicating the presence of intrinsic innervation. In the aganglionic segment (AGS) in HD, neither Hu C/D-positive cells nor CD56-positive Auerbach’s plexus were present, but CD56-positive nerve bundles were identified indicating extrinsic innervation (Fig. 1). In NGS, the number of Hu C/D-positive cells was 13.22 ± 1.55 per mm and the total area of CD56-positive Auerbach’s plexus was 24.4 ± 6.0 × 103 μm2 per mm (Fig. 2).

Fig. 1
figure 1

The histopathological findings of HE staining and immunohistochemical staining using Hu C/D and CD56 in each disease. The presence of ganglion cells (white arrowhead) and a normal-sized Auerbach’s plexus (white arrow) was confirmed with all of the staining methods in the NGS of HD patients. In contrast, neither ganglion cells nor Auerbach’s plexus, and CD56-positive nerve bundles were observed in the AGS of HD patients (black arrow). In CIHG, the number of ganglion cells was few (white arrowhead in HE staining and Hu C/D staining), and the size of Auerbach’s plexus was small (HE staining and white arrow in CD56 staining). In AIHG, the ganglion cells were decreased in number and degenerated (white arrowhead in HE staining and Hu C/D staining), but the size of Auerbach’s plexus was preserved (HE staining and white arrow in CD56 staining). The magnification is ×200, and the scale bar indicates 50 μm. HD Hirschsprung’s disease, NGS normoganglionic segment, AGS aganglionic segment, CIHG congenital isolated hypoganglionosis, AIHG acquired isolated hypoganglionosis

Full size image
Fig. 2
figure 2

A quantitative analysis of the number of ganglion cells and the area of Auerbach’s plexus. Compared with the NGS of HD patients, the number of ganglion cells per 1 mm and the total area of CD56-positive Auerbach’s plexus per 1 mm in CIHG are significantly smaller (p < 0.05); in contrast, in AIHG, only the number of ganglion cells per 1 mm was significantly smaller than that in the NGS of HD patients (p < 0.05), while the total area of CD56-positive Auerbach’s plexus per 1 mm was equivalent. *Indicates the significant difference (p < 0.05). HD Hirschsprung’s disease, NGS normoganglionic segment, AGS aganglionic segment, CIHG congenital isolated hypoganglionosis, AIHG acquired isolated hypoganglionosis, N.D. not detected, N.S. not significant

Full size image

Congenital isolated hypoganglionosis

Demographic data and clinical findings of 25 CIHG patients are shown in Table 1 and Supplemental Table 1. Median age at diagnosis was 4 days (range 1–1460), age at onset of symptoms was 0 days (range 0–180), and the duration of symptoms until diagnosis was 3 days (range 0–1458). Male to female ratio was 11:13 (unknown in one case). Of the specimens, 23 concerned biopsies and two were bowel resection specimens. These were from the ileum (n = 10), jejunum and ileum (n = 3), ileum and distal colon (n = 3), ileum and mid colon (n = 2), ileum, proximal, and distal colon (n = 2), jejunum, ileum, and distal colon (n = 1), ileum and proximal colon (n = 1), ileum, mid colon, and distal colon (n = 1), mid colon (n = 1), and the distal colon (n = 1). The specimen types were significantly different from those of the other groups (p < 0.05).

Histopathologically, CIHG showed a significantly lower density of myenteric ganglion cells and a reduction in size of the Auerbach’s plexus [5]. Effectively, the number of ganglion cells per mm was extremely low (HE staining and Hu C/D staining) and the size of the Auerbach’s plexus was small (by H&E and CD56 staining) in comparison to those in the NGS in HD (Fig. 1). The number of Hu C/D-positive cells was 2.33 ± 0.84 per mm and the total area of CD56-positive Auerbach’s plexus was 3.54 ± 0.85 × 103 μm2 per mm. These differences were statistically significant (p < 0.05) (Fig. 2).

Acquired isolated hypoganglionosis

Demographic data and clinical findings of 5 AIHG patients are shown in Table 1 and Supplemental Table 2. Median age at diagnosis was 17 years (range 2–30). The precise age at the onset of symptoms was unclear in some cases, but the age distribution varied from the neonatal period to 12 years. The duration of symptoms ranged from 3 months to over 20 years. The male to female ratio was 3:2. Three specimens were from bowel resection and two concerned biopsies. They were from the stomach, small intestine and colon, ileum, mid and distal colon, small intestine and mid colon, mid and distal colon, and distal colon, respectively. Age at diagnosis was significantly different from that of the other groups (p < 0.05). One patient suffered from eosinophilic gastroenteritis followed by AIHG (Supplemental Table 2).

Histologically, by H&E and Hu C/D staining, a decreased number and degeneration of ganglion cells were noted, while by H&E and CD56 staining, the size of the Auerbach’s plexus was preserved (Fig. 1). The number of Hu C/D-positive cells was 1.78 ± 0.36 per mm and the total area of CD56-positive Auerbach’s plexus was 27.13 ± 3.71 × 103 μm2 per mm. The number of ganglion cells in AIHG was significantly lower than that in the NGS in HD (p < 0.05), whereas the size of the Auerbach’s plexus was equivalent to that in the NGS in HD (Fig. 2).

An additional histopathological characteristic of AIHG (patient no. 4 in Supplementary Table 2) was eosinophilic infiltration (white arrowhead) and occasional lack of clarity of the border of the Auerbach’s plexus (Fig. 3 a). The ganglion cells were degenerated and resembled glial cells (white arrow) (Fig. 3 b). Immunohistochemical staining improved the differentiation between AIHG and CIHG (Fig. 1).

Fig. 3
figure 3

The additional histopathological characteristics of AIHG. a The border between Auerbach’s plexus and the muscle layer is unclear, and eosinophilic infiltration (white arrowhead) can be seen. b The ganglion cells are degenerated and resemble glial cells (white arrow). The magnification is ×400, and the scale bar indicates 30 μm. AIHG acquired isolated hypoganglionosis

Full size image

Discussion

Hypoganglionosis was first described by Tiffin et al. in 1940 [19], and Meier-Ruge et al. called hypoganglionosis without aganglionosis as “isolated form” in 1999 [16]. Due to its low incidence and a limited number of studies published, the existence of IHG has been questioned [3]. However, recently, the number of cases of IHG diagnosed has increased, and the entity has gradually come to be established [9, 15, 16]. We previously conducted a nationwide survey of CIHG and AIHG in Japan [4–6]. However, the definitive concept, pathogenesis, etiology, and histopathological criteria to diagnose AIHG remain controversial, as only few reports have been published [20–24].

In most cases, the onset of symptoms of HD and CIHG is in the neonatal period or in infants younger than one year of age. Occasionally, the diagnosis is made at a later age. In contrast, AIHG can present at any age, depending to some extent on the underlying etiology (Table 1 and Supplementary Table 2). Neurodegenerative diseases that predispose a patient to enteric ganglion cell loss are usually slowly progressive, and most conditions that underlie AIHG are more common in older children or adults. Chronic constipation or intestinal pseudo-obstruction is typical [25]. Regarding etiology, viral infection [20], Chagas disease [21], a complication due to a Soave procedure [22], and ischemia [26] have been reported. In the present study, infiltration of eosinophils in the Auerbach’s plexus was identified in an AIHG patient diagnosed with eosinophilic enterocolitis (Fig. 3 and Supplementary Table 2). We previously reported [5] that infectious agents may attack enteric ganglion cells, resulting in their degeneration and death.

While H&E staining allows identification of ganglion cells and the Auerbach’s plexus [5], immunohistochemical staining for Hu C/D has proven useful in identifying a decrease in the number of ganglion cells in CIHG and the remaining degenerated ganglion cells. Staining for CD56 allows recognition of preservation of the Auerbach’s plexus, the border of which with the muscle layer can be poorly defined in AIHG (shown in Fig. 3). Phillips et al. [12] described that staining for Hu C/D is particularly versatile because of its high signal-to-noise ratio with very little background. Lin et al. [27] reported that Hu selectively labels neuronal nuclei and cytoplasm, whereas the processes are not stained. Wakamatsu et al. [28] noted that Hu genes/proteins are expressed exclusively in neurons and glial cells. Park et al. [8] showed that CD56 is strongly expressed in all components of the enteric nervous plexus. Nogueira et al. [29] suggested that CD56 is particularly useful in the diagnosis of hypoganglionosis because it is a general neuronal marker staining all nerves. Although the utility of staining for Hu C/D and CD56 in differentiating AIHG from CIHG has not been explored, these markers provide a useful diagnostic tool based on the findings from the present study. Nonetheless, when staining for CD56, morphological features (compact arrangement of neurites, perineurium) are important to distinguish extrinsic hypertrophic nerve bundles from intrinsic nerves of the Auerbach’s plexus because the CD56 antibody labels both [13, 14].

We have found a significant difference in the number of ganglion cells and the size of the Auerbach’s plexus between CIHG and AIHG (Fig. 2). The affected segments of the digestive tract were significantly different between the three groups (p < 0.05). In HD, most samples were from the colon, while in CIHG and AIHG, most samples were from the small intestine and colon (Table 1). Comparison between different intestinal sites is justified based on the report by Amiot et al. who documented equivalent density of ganglia and ganglion cells in the jejunum, ileum, and colon [30].

In the present study, we could not establish an exact cut-off value to distinguish pathological from normal values. One potential confounding factor in our analysis is our use of proximal bowel from HD resections as normal reference. Hypoganglionosis is a well-documented finding in the transition zone oral to the AGS in HD, which can extend over variable distances, and may have been present even at the proximal margin of some of our HD resection specimens [7]. Another confounding variable is chronic distension proximal to the aganglionic segment, which may alter the measured density of ganglia or ganglion cells per unit length. However, we previously measured the number of ganglion cells on H&E-stained sections [5]. Based on the previous report and the present results, we estimate that the cut-off value defining loss of ganglion cells is 2–3 per longitudinal mm of intestine. To establish a cut-off value for the size of the Auerbach’s plexus, further investigations on a larger number of cases are necessary.

The present and earlier published findings clearly establish the histopathological differences between HD, CIHG, and AIHG (Table 2) [5, 25]. However, while CIHG and AIHG can be distinctly differentiated based on the histopathological findings, their etiology, pathogenesis, and prognosis remain to be established through studies on a larger number of cases.

Table 2 Clinical and histopathological differentiation among HD, CIHG, and AIHG
Full size table

In summary, we found immunostaining for Hu/CD and CD56 useful for the definitive histopathological diagnosis of IHG and to distinguish AIHG from CIHG.