Reduced Neuronal Innervation in the Distal End of the Proximal Esophageal Atretic Segment in Cases of Esophageal Atresia with Distal Tracheoesophageal Fistula
- 112 Downloads
Esophageal dysmotility is a common occurence after surgical repair of proximal esophageal atresia (EA) and distal tracheoesophageal fistula (TEF). The etiology of this motility disorder, however, remains controversial. Esophageal dysmotility also is present in isolated TEF or EA before surgery, suggesting a congenital cause. However, there is no information available in the literature with regard to the intramural nervous system of the human esophagus in EA-TEF.
Patients and Methods
We examined the distal end of proximal esophageal atretic segment of neonates undergoing EA-TEF repair for intrinsic neuronal innervation. Using specific antibodies, we studied neuronal markers of specimens from nine cases of EA-TEF and 9 cases of normal esophagus by immunohistochemistry using neurofilament (NF), synaptophysin (SY), S100, and glial cell line-derived neurotrophic factor (GDNF).
In the atretic segment, specimens staining with hematoxylin and eosin showed that there were marked hypoganglionosis and immature ganglion cells in the myenteric plexus. GDNF immunoreactivity in the atretic esophagus were markedly reduced in both the muscular layer and myenteric plexus. SY and NF-immunorective nerve fibers were distributed throughout the myenteric plexus of the normal esophagus, but the scarcity of these immunoreactive nerve fibers in the atretic esophagus was apparent. In contrast, the density of immunorective nerve fibers for S100 in the myenteric plexus and muscular layer was increased in the distal end of the atretic esophagus.
We concluded that the distribution of ganglion cells and some nerve fibers in the distal end of the atretic esophageal segment is deficient. Inadequate and abnormal neuronal innervation of the esophagus could be related to the esophageal dysmotility seen in EA. Because GDNF is a survival factor for central and peripheral neurons, defective expression of GDNF could have an important role in the defective and/or abnormal neuronal innervation of atretic esophageal segment.
It has been demonstrated that a spectrum of esophageal motility abnormalities exist in patients after surgical correction of their esophageal atresia (EA) and distal tracheoesophageal fistula (TEF) [1, 2, 3, 4, 5]. These abnormalities include dysphagia, gastroesophageal reflux, aperistaltis, and low-amplitude simultaneous contractions in the affected esophagus as documented by radiologic findings, manometric observations, and intraesophageal pH recordings [2, 5, 6]. In the past this was attributed to a neurologic defect, probably as a consequence of partial denervation of the esophagus during operative dissection. The observation of disorders of the esophageal motor activity in patients who have EA-TEF before surgical intervention make it unreasonable to attribute the dysfunction entirely to esophageal denervation resulting from surgical mobilization and dissection. It is possible that patients with EA-TEF have preexisting congenital neuromotor abnormality in the atretic esophagus.
There is a paucity of information about the intrinsic nevre supply to the human esophagus in the presence of EA-TEF, mainly because of the difficulty of obtaining specimens to examine. We obtained the specimens of atretic esophagus trimmed during the primary anastomosis for EA-TEF. The aim of the study was to document whether there is any demonstrable primary intrinsic neuronal abnormality in the distal end of the proximal atretic esophageal pouch in neonates with EA-TEF that may cause dysmotility.
Patients and methods
Nine neonates (four boys and five girls, with the average gestational age 36.66 ± 2.23 days and postnatal age of 4.55 ± 2.96 days) with EA and TEF have been enrolled in the study to date. All patients have the proximal esophageal atresia and distal tracheoesophageal fistula form of EA. The tissue of the distal end of proximal atretic segment trimmed at the time of surgery for preparation of a primary anastomosis was used. Nine control esophagi were obtained from children 10 to 60 day old (five boys and four girls, with the average gestational age 37.77 ± 2.33 days, postnatal age 25.33 ± 14.89 days) undergoing autopsy, and the tissues were confirmed histologically to be unaffected. The distal end of the proximal atretic esophagus from the controls was 1 to 2 mm above the bifurcation of the trachea because this level is the counterpart of that in the esophageal atresia group.
Histology and immunohistochemistry
Histological and immunohistochemical analyses were performed in all cases on formalin-fixed, paraffin-embedded tissues. Serial sections (4–5 μm) from the distal end of the proximal atretic segment and its counterpart from the controls were cut and used for hematoxylin and eosin (H&E) and immunohistochemistry (IHC) preparations. Each glass slide was mounted with three histologic sections, one from each group, all stained for the same marker. One set consisted of 15 glass slides, three for the each of the five markers (including H&E) giving a total of 45 histologic sections in each set.
Immunohistochemical staining was performed with GDNF D-20 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), an affinity-purified rabbit polyclonal antibody. GDNF D-20 recognized specifically GDNF of human cell origin. The primary antibody at a dilution of 1:100 was used. The immunohistochemical techniques of S-100 protein Ab-1 (clone 4C4.9; NeoMarkers, Fremont, CA, USA), neurofilament (Clone 2F11; Dako, Carpinteria, CA, USA), and synaptophysin (polyclonal rabbit antibody; Dako) were used to define the neuronal and ganglion cell components of the esophageal segments more precisely. The avidin-biotin-peroxidase complex methods were used for each antibody.
To evaluate the ganglion cells in each case, five visual fields that consisted of myenteric and submucous plexuses were examined under microscope with a 20× objective lens and 10× oculars. The ganglion cells were counted in each field. The total ganglion numbers were noted for each plexus separately. These data were compared using the Mann-Whitney U-test, taking p < 0.05 as statistically significant.
The images of an H&E-stained and an IHC-stained specimen were captured with a camera mounted on a light microscope, transmitted to a desktop computer, and projected onto the monitor. The thickness of the mucosa, submucosa, and muscle were measured.
The immunoreactivity (IR) was assessed in two ways as previously described by Cheng et al.  using the image analyzer system (Leica; Leica Micros Imaging Solutions, Cambridge, UK): (1) Quantity: The net area of immunostaining per each cross section of esophagus was measured. (2) Density: The IR was calculated by expressing the immunostained area as a percentage of the cross-sectional area of the esophagus. Three to five high power fields (×400) in each histological section were assessed to quantitate the results. Descriptive statistics are given as the mean ± SEM. These data were compared using the Mann-Whitney U-test, taking p < 0.05 as statistically significant.
Pearson correlations were used to assess the relation between thickening of the esophageal wall, ganglion numbers, and neuronal density. A value of p < 0.05 was considered statistically significant.
Number of ganglia and thickness of the mucosal, submucosal, and muscular layers of control and atretic esophagus
723.260 ± 32.543
774.693 ± 85.005
1245.586 ± 50.218
1283.266 ± 60.486
3656.813 ± 35.757
3807.52 ± 92.363
9101.246 ± 3167.27
6889.60 ± 76.165
No. of ganglia
4.22 ± 4.99
6.55 ± 3.74
IR (quantity) and IR per unit area of the control and atretic esophagus (density) for four neuronal markers
IR of each section (quantity)
IR as a percentage of the cross-sectional area (density)
Synaptophysin and S100 immunoreactivity
All neuronal IRs have negative and positive correlations with the esophageal wall thickness in the two groups, but they were not statistically significant (p > 0.05).
Esophageal dysmotility is frequent in children suffering from esophageal atresia and TEF [6, 8, 9, 10, 11]. The cause of esophageal dysmotility may involve the extrinsic nerves (e.g., recurrent laryngeal and vagus nerves), intrinsic intramural nervous components, and/or muscular elements. Previous experimental studies exploring the anatomy of the extrinsic nerves of fetal rats with EA-TEF demonstrated an abnormality in the vagus nerves, which may explain the occurrence of esophageal dysmotility . However, bilateral high vagatomy or complete vagal blockade regularly produced a period of smooth muscle aperistalsis. Within a few hours to 4 months, there is a return of peristaltic activity to this portion of the esophagus [13, 14]. This finding suggests that the complex intrinsic nervous system in the esophagus, as seen in the other parts of the gastrointestinal tract, is essential for normal esophageal motility. It has been demonstrated that both contraction and relaxation of the normal esophageal smooth muscle are regulated directly by intramural excitatory and inhibitory motor neurons in the myenteric plexus [15, 16]. Using an adriamycin fetal rat model, some workers have reported that there were intrinsic neuronal abnormalities in the myenteric plexus of the atretic esophagus [7, 17, 18].
With improving survival rates, there has been increasing interest in esophageal function after successful repair, but few studies have addressed the histologic structure of the human esophagus with EA, mainly because of difficulty obtaining specimens to examine. For our study we were able to obtain specimens of the distal end of the proximal atretic segment trimmed at the time of surgery from nine neonates. We have prepared specimens with the H&E stain and four different IHC stains to look for intrinsic neuronal abnormalities in the distal end of the human atretic esophageal pouch.
The total immunoreactivities of NF, SY, and GDNF in the esophagus in each group showed significant differences in our study. The expression of these neuronal markers was significantly reduced in the distal end of the proximal atretic esophagus. This possibly translated into a reduction in the intrinsic neuronal innervation of the atretic esophagus. It can also be speculated that the nerve tissue that normally encircles most of the esophagus became thin and short in the distal end of the proximal atretic esophagus, and the nerve distribution was markedly reduced. In addition, atretic segment specimens showed that there were marked hypoganglionosis and immature ganglion cells in the myenteric plexus.
It seems that with the development of EA-TEF concurrent changes may occur in intrinsic nerve development. Nakazato et al.  described a similar finding in five biopsies from atretic human esophagus; that is, the interganglionic fibers were thinner and looser than normal. In our study, we also find significantly elevated expression of S100 in the muscular layer and the myenteric plexus, which stains for glial cells. The hypertrophied glial tissue may signify compensation for defective neuronal tissue.
The most novel and relevant finding of the present work is the observation that GDNF expression is significantly reduced in the atretic esophageal specimens. The GDNF is vital to development and maintenance of neural tissues. It is distant member of the transforming growth factor-β) (TGFβ) superfamily, which comprises an expanding list of multifunctional proteins serving as regulators of cell proliferation and differentiation. Formation of cranial parasympathetic and enteric nervous system by neuronal precursors migrating from the neuronal crest requires GDNF signaling via GDNF family receptor α-1 and Ret receptor tyrosine kinase [20, 21]. Because GDNF is a survival factor for central and peripheral neurons, to explain why the nerve fibers become thin and sparse it can be speculated that defective GDNF expression is responsible (as a trophic or induction factor) for the underdevelopment of these intrinsic nerve fibers. However, it cannot be excluded that poor GDNF expression and/or poor neuronal development leads to atresia and requires further investigation.
In our study we observed that, compared with the normal esophagus, the total cross-sectional area in the atresia group was slightly increased; and this finding was different from the adriamycin-induced esophageal atresia model. This finding suggests a defect in the adriamycin-induced esophageal atresia model. The total immunoreactivities of NF, SY, S100, and GDNF in the esophagus in each group showed no statistically significant correlation with the esophageal wall thickness. For this reason, it is not possible to say that these changes in neuronal density are related to the esophageal wall changes in our study.
We conclude that the distribution of neuronal tissue in the atretic esophageal segment is deficient. The alteration of GDNF expression may be involved in the pathogenesis, and it may be responsible for the deficient neuronal innervation of the esophageal atretic segment. The results may be helpful for explaining the esophageal motility disorders seen in patients with EA-TEF before surgical correction. Further research is needed to determine more precisely the exact nature of innervation and neuromotor abnormalities in patients who have EA-TEF.
- 1.Kirkpatrick JA, Cresson SL, Pilling GP (1961) The motor activity of the esophagus in association with esophageal atresia and tracheoesophageal fistula. AJR Am J Roentgenol 86:884–887Google Scholar
- 3.Romeo G, Zuccarello B, Proietto F, et al. (1987) Disorders of the esophageal motor activity in atresia of the esophagus. J Pediatr 22:120–124Google Scholar
- 5.Duranceau A, Fisher SR, Flye MW, et al. (1997) Motor function of the esophagus after repair of esophageal atresia and tracheo-esophageal fistula. Surgery 82:116–123Google Scholar
- 8.Beasley SW (1998) Esophageal atresia: surgical aspects. In: Stringer MD, Oldham KT, Mouriquand PDE, et al., editors. Pediatric Surgery and Urology: Long Term Outcomes. London, Saunders, pp 166–180Google Scholar
- 15.Crist J, Gidda JS, goayal RK (1984) Intramural mechanism of esophageal peristalsis: role of cholinergic and noncholinergic nerves. Physiol Sci 81:3595–3599Google Scholar