THE AMERICAN JOURNAL OF GASTROENTEROLOGY
1, 2000
© 2000 by Am. Coll. of Gastroenterology
Published by Elsevier Science Inc.
ORIGINAL CONTRIBUTIONS
Congenital Esophageal Stenosis in Adults
David A. Katzka, M.D., Marc S. Levine, M.D., Gregory G. Ginsberg, M.D., Riyadh Hammod, M.D.,
Philip O. Katz, M.D., Erik K. Insko, M.D., Edward C. Raffensperger, M.D., and David C. Metz, M.D.
Departments of Medicine and Radiology, Hospital of the University of Pennsylvania,
and Department of Medicine, Graduate Hospital, Philadelphia, Pennsylvania
OBJECTIVE: Congenital esophageal stenosis is thought to be
a rare disease confined to infancy and childhood with only
a few case reports in adults described.
METHODS: We report five patients between the ages of 19
and 46 yr who presented with this disorder over a 2-yr
period.
RESULTS: Patients had been labeled with reflux strictures,
webs, or as idiopathic in the past. All patients had chronic
solid food dysphagia, some since early childhood. The
location of the stricture varied, occurring in the mid or
proximal esophagus in four, but throughout the esopha-
gus in one. Radiographic and endoscopic appearance was
a smooth concentric stricture or multiple rings, some-
times tracheal in appearance. Endosonography was per-
formed in two patients, both of whom had focal circum-
ferential hypoechoic wall thickening with disruption of
the normal layer pattern corresponding to the areas of
luminal narrowing. All patients dilated had good symp-
tomatic response, with resolution of symptoms up to 6
months in follow-up.
CONCLUSIONS: We suggest that congenital esophageal ste-
nosis does occur in adults and may be underrecognized. Its
endosonographic appearance is described. (Am J Gastroen-
terol 2000;95:32–36. © 2000 by Am. Coll. of Gastroenter-
ology)
INTRODUCTION
Congenital esophageal stenosis is thought to be a rare dis-
ease generally confined to childhood and infancy. It is
estimated to occur in 1 in 25,000 –50,000 live births (1).
Presentation may occur in infancy when the stenosis is
severe, but less severe forms of stenosis may present later in
childhood as these individuals start to ingest solids. Its
occurrence in adults has only been documented by an
occasional case report (2– 6). During a 2-yr period, we
have had the opportunity to diagnose and treat five pa-
tients with congenital esophageal stenosis who presented
during adulthood. Furthermore, we describe the clinical,
radiographic, and endoscopic characteristics of this pa-
tient group.
Vol. 95, No.
ISSN 0002-9270/00/$20.00
PII S0002-9270(99)00727-3
MATERIALS AND METHODS
After identification of a single case, other patients with
suspected congenital esophageal stenosis in adults from two
institutions were identified. All patients underwent double-
contrast esophagography. All radiological studies were re-
viewed by one of the authors (M.S.L.). Endoscopy and
Savary dilation were performed in patients by three of the
authors (D.A.K., P.O.K., or D.C.M.) using diagnostic stan-
dard gastroscopes, G1F100 and 140 series (Olympus Amer-
ica, Mellville, NY). Endoscopic ultrasound using Olympus
GF-UM-20 (Olympus America) was performed by one of
the authors (G.G.) in two patients. Ambulatory pH moni-
toring was performed using a three-lead Synectics portable
pH-measuring unit with a model 93-6200 Digitrapper, re-
corded and analyzed using the Synectics Esophagram pro-
gram (Esophagram Software Package, Gastrosoft Inc., Chi-
cago, IL). Esophageal manometry was performed in two
patients using a solid-state intraluminal transducer assembly
(Konigsberg Instruments, Pasadena, CA).
Clinical Characteristics
Five cases were assembled from the two institutions over a
2-yr period (Table 1). These five patients ranged in age from
19 to 46 yr. All presented with long-standing dysphagia to
solids, with the average duration of dysphagia 16 yr (range,
5–25 yr). One patient had superimposed episodes of food
impaction. Two others had episodes of regurgitation of solid
food. The two youngest patients (ages 19 and 21 yr), with
clear symptoms since childhood, had adapted by either
cutting food into small pieces and eating slowly (patient 3)
or by restricting the diet to soft foods (patient 2). The latter
patient also remembers storing food in her cheek as a child
before swallowing to prevent choking. Prior diagnoses in-
cluded peptic stricture, treated with omeprazole in two pa-
tients, and so-called “webs” in one. No patient had any
history of skin disease associated with esophageal strictures
or webs, history of caustic or foreign body ingestion, pill-
induced esophagitis, iron deficiency anemia, tylosis, or ra-
diation exposure.
Four patients had previously undergone previous Savary
dilation in the past. One patient who had undergone four
dilations had good symptomatic response lasting generally
2–3 yr per dilation and was able to eat most solids with little
restriction. The three other patients had a less favorable
AJG – January, 2000
Table 1. Dysphagia
Patient Age (yr) Age of Onset (yr) Duration (yr)
1 46 29 17
2 21 ? Birth 21
3 19 7 12
4 23 18 5 0
5 30 5 25
* Distance from incisors or radiographic location.
EUS ⫽ endoscopic ultrasound.
response, with either little response or rapid recurrence of
dysphagia, still requiring restriction of some solids or care-
ful chewing and cutting of food.
Ambulatory pH monitoring was performed in patient 3
and showed an elevation in recumbent pH in the distal
esophagus (6.1%; normal, ⬍1.2%) but upright and total
measurements were within normal limits. There was no
proximal acid exposure in this patient in the area of the
stricture. In this patient and one other (patient 5), esophageal
manometry was performed with normal lower esophageal
sphincter (LES) pressure and esophageal peristalsis. Of
note, however, is that both of these patients had high prox-
imal strictures so peristaltic function may not have been
adequately assessed in these areas.
Radiographic Findings
Double-contrast esophagrams (Fig. 1) revealed strictures in
the midthoracic esophagus in two patients, the upper and
midthoracic esophagus in one, the upper, mid, and distal
thoracic esophagus in one, and the cervical esophagus in
one. Not accounting for magnification, the strictures had an
averaged width of 1.0 cm (range, 0.8 –1.7 cm) and an aver-
age length of 9.5 cm (range, 1.5–20 cm). In two cases, there
was prolonged retention of a 1.2-cm– diameter barium tablet
above the proximal end of the stricture. The strictures had
relatively tapered proximal and distal borders in all five
patients. The narrowing was concentric in four patients and
asymmetric in one. In four patients, characteristic ring-like
indentations were detected in various portions of the stric-
tures.
Endoscopy and Endoscopic Ultrasound (EUS)
Three patients (patients 1, 4, and 5) underwent endoscopy
(one patient was pregnant at the time of presentation to us;
another had undergone endoscopy at another institution).
One patient (patient 5) had the appearance of so-called
“tracheal”-type rings 23–25 cm from the incisors with a
luminal diameter of approximately 8 mm, causing moder-
ately difficult passage of the endoscope (Fig. 2A). The
overlying mucosa appeared completely normal. The distal
esophagus and stomach also appeared grossly normal. Pa-
tient 1 had four to five thickened ring-like structures dis-
tributed throughout the body of the esophagus, starting
approximately 33 cm from the incisors and extending down
to just proximal to the gastroesophageal mucosa (Fig. 2B).
The endoscope was passed with only mild resistance
through these structures with the small ring diameter ap-
Congenital Esophageal Stenosis in Adults 33
Previous Dilations Stricture Location* EUS
4 33–30 cm Yes
2 Proximal esophagus No
1 Proximal esophagus No
29–33 cm Yes
1 23–25 cm No
proximately 9 mm. All overlying mucosa appeared normal.
Endoscopy of patient 4 demonstrated a smooth regular stric-
ture from 29 to 33 from the incisors, with normal overlying
mucosa with a luminal diameter of approximately 7 mm.
The rest of the esophagus appeared normal. Endoscopic
ultrasound was performed in these latter two patients (Fig.
3). Both these patients showed focal circumferential hypo-
echoic wall thickening with disruption of the normal layer
pattern corresponding to the areas of luminal narrowing.
Mural thickening increased up to 8 mm in involved areas.
Uninvolved segments of the esophagus had normal EUS
mural characteristics with preservation of the wall layer
pattern and normal thickness.
Biospsies were available for four patients. Only two pa-
tients had biopsies performed within the area of the stenosis.
One patient had mild chronic inflammation; another had
inflammatory change with eosinophils. This second patient
(patient 1) had presented with the multiple rings rather than
an isolated stricture. Of the biopsies performed in the distal
esophagus, one patient had focal inflammatory cell infiltra-
tion of the lamina propria. One patient also had moderate
chronic inflammation with eosinophils, basal cell hyperpla-
sia, and elongation of the Rete pegs. This patient (patient 3)
had mildly elevated recumbent reflux on ambulatory pH
study.
All three patients endoscoped underwent Savary dilation
successfully, without complication. Three patients were di-
lated up to 15 mm. One of these patients had severe chest
pain after dilation, which resolved spontaneously without
need for further study. All three patients had resolution of
their dysphagia postdilation. Clinical follow-up has not re-
vealed any relapse with follow-up ranging from 2 to 6
months.
DISCUSSION
In this report, we describe five patients who have a clinical
picture consistent with congenital esophageal stenosis. A
long history of dysphagia is typical, with two of these
patients having lifelong symptoms. Presentation is most
commonly with solid food dysphagia but episodes of super-
imposed food impaction may also occur. These patients
typically have had previous dilations with variable results
and have often been labeled with several diagnoses, includ-
ing peptic stricture and esophageal webs. Congenital esoph-
ageal stenosis is manifest on barium studies by relatively
34 Katzka et al. AJG – Vol. 95, No. 1, 2000

Figure 1. Barium esophagography of congenital esophageal stenosis. A. Double-contrast esophagram shows a tapered area of concentric
narrowing (arrows) in the midesophagus. Transverse indentations are faintly seen in the region of the stricture (patient 4). B. Double-
contrast esophagram shows a focal area of mild narrowing in the midesophagus with several discrete ring-like indentations (arrows) in the
narrowed segment (patient 1).

long strictures in the upper, mid, or distal esophagus at a
discrete distance from the gastroesophageal junction, with a
normal overlying and intervening mucosal pattern. These
strictures usually appear to be tapered areas of concentric
narrowing and are often associated with multiple ringlike
indentations.
Three pieces of evidence support the conclusion that
these adults have congenital esophageal stenosis. Firstly, the
descriptions of congenital esophageal stenosis as it exists in
children are very similar to what is described in our patients.
These studies (7, 8) describe endoscopically and radiograph-
ically regular stenoses with normal appearing overlying
mucosa or stenoses with appearances very similar to tra-
chea. Furthermore, both the proximal and distal esophagus
may be involved. These studies also describe stenosis as
fibromuscular based on pathological resection, with the
finding of localized circumferential proliferation of smooth
muscle embedded in connective tissue. Also described in the
children’s series are fragments of cartilage in ring-like struc-
tures similar to trachea. Some of the radiographic and en-
AJG – January, 2000
Figure 2. Endoscopic appearance of congenital esophageal stenosis. A. “Tracheal”-type appearance to stricture in proximal esophagus
(patient 5). B. Multiple rings throughout the esophagus (patient 1).
doscopic images in the present series may have been com-
patible with this process, although a tissue diagnosis is
lacking. Finally, the endoscopic ultrasound images in two of
our patients with generalized hypoechoic mural thickening
Figure 3. Endoscopic ultrasound image of congenital esophageal
stenosis. There is focal disruption of the normal five-layer wall
pattern with circumferential, hypoechoic (dark black) thickening of
the esophageal wall up to 8 mm (normal, 3– 4 mm) at locations
corresponding to the luminal stenosis (large arrows). The hypere-
choic (bright white) interface between the esophageal wall and the
surrounding adventitia is intact (small arrows). The echoendoscope
transducer (TR) is seen centrally.
Congenital Esophageal Stenosis in Adults 35
with obscuring of the normal layer pattern were consistent
with these descriptions, although tissue was not obtained to
confirm this. Of note, the EUS findings observed in these
cases are more extensive than those previously described in
the single case report by Mc Nally et al. (9). In their case
report, focal thickening was limited to the muscularis pro-
pria layer and there was preservation of the normal five-
layer echo pattern of the esophageal wall. In our patients,
there is transmural hypoechoic thickening of the esophageal
wall with elimination of the normal wall layer pattern.
Secondly, other reported cases of congenital esophageal
stenosis have had similar presentations (2– 6). These include
a history of longstanding and often lifelong compensated
dysphagia and presentation with food impaction. Patients
tend to be young, but reports of this lesion in patients as old
as 63 yr have been reported (10). Interestingly, two of these
patients who underwent dilation experienced significant
chest pain after the procedure, similar to one of our patients
and to others described (9).
Thirdly, although strictures caused by mediastinal irradi-
ation, caustic ingestion, or rare skin diseases such as epi-
dermolysis bullosa dystrophica and benign mucous mem-
brane pemphigoid might have a similar appearance, there
was no evidence for these secondary diseases in these pa-
tients. No patients had history of caustic ingestion or evi-
dence of gastroesophageal reflux or Barrett’s esophagus
with completely normal overlying mucosa, nor evidence of
severe reflux on barium study or ambulatory pH study.
Barrett’s strictures also tend to be shorter and are not typ-
ically associated with ringlike indentations. Thus, in the
absence of a history of reflux disease, caustic ingestion, or
36 Katzka et al.
mediastinal irradiation, a relatively long, tapered stricture
involving the upper, mid, and/or distal esophagus should be
highly suggestive of congenital esophageal stenosis, partic-
ularly if associated with ring-like indentations.
The etiology of these strictures is unclear. In the case of
stenosis and web, atresia has been hypothesized in infants
on the basis of pathological analysis and by association with
other anomalies that may be associated with atresia, such as
tracheoesophageal fistula, duodenal atresia, and anorectal or
cardiac abnormalities (8). None of these were present in the
patients in this series.
One of the essential questions is determining how com-
mon this entity is in adults. Initial determination would
suggest that it is rare, as only isolated case reports have been
described (2– 6). In contrast, preliminary data were recently
presented in abstract form on seven adult patients with
congenital esophageal stenosis (10). We reported five pa-
tients in 2 yr. We suspect that congenital esophageal steno-
sis is a more common condition that previously recognized
and that many prior cases may have been mislabled as
idiopathic or with other diagnoses.
In summary, we describe the clinical, radiographic, en-
doscopic, endosonographic, and therapeutic aspects of con-
genital esophageal stenosis in adults, an entity that may be
more common than previously appreciated. We hypothesize
that this lesion is congenital because of its characteristics,
which are similar to those described in infancy and child-
hood and because of the exclusion of other diseases known
to be associated with esophageal stricture formation in
young adults. Effective therapy requires vigorous dilation
but with the expectation that it may be accompanied by
chest pain. Finally, endoscopic ultrasound may be useful
diagnostically in confirming the diagnosis, particularly be-
cause biopsy of the esophageal wall may not be practical.
AJG – Vol. 95, No. 1, 2000
Reprint requests and correspondence: David A. Katzka, M.D.,
Division of Gastroenterology, 3 Ravdin Building, Hospital of the
University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA
19104.
Received Mar. 10, 1999; accepted July 6, 1999.
REFERENCES
1. Wright VM. In: Walker WA, Durie PR, Hamilton JR, et al.,
eds. Pediatric gastrointestinal disease. Philadelpha: Decker,
1991:369 –70.
2. Pokieser P, Schiima W, Schober E, et al. Congenital esopha-
geal stenosis in a 21-year old man. Am J Roentgenol 1998;
170:147– 8.
3. Case records of the Massachusetts General Hospital. N Engl
J Med 1956;255:707–10.
4. McNally PR, Collier EH, Lopiano MC, et al. Congenital
esophageal stenosis. A rare cause of food impaction in the
adult. Dig Dis Sci 1990;35:263– 6.
5. Harrison CA, Katon RM. Familial multiple congenital esoph-
ageal rings: Report of an affected father and son. Am J
Gastroenterol 1992;87:1813–5.
6. Singaram C, Sweet MA, Gaumnitz EA, et al. Peptidergic and
nitrinergic denervation in congenital esophageal stenosis. Gas-
troenterol 1995;109:275– 81.
7. Nishima T, Tsuchida Y, Saito S. Congenital esophageal ste-
nosis is due to tracheobronchial remants and its associated
anomolies. J Pediatr Surg 1981;16:190 –3.
8. Nihoul-Fekete C, DeBacker A, Lortat-Jacob S, et al. Congen-
ital esophageal stenosis. A review of 20 cases. Pediatr Surg Int
1987;2:86 –92.
9. McNally PR, Lemon JC, Goff JS, et al. Congenital esophageal
stenosis presenting as noncardiac, esophageal chest pain. Dig
Dis Sci 1993;38:369 –73.
10. Younes Z, Johnson DA. Congenital esophageal stenosis:
Clinical and endoscopic features in adults. Am J Gastroen-
terol 1996;91:1901 (abstract).