ORIGINAL ARTICLE

Inlet patch: Associations with endoscopic findings in theupper gastrointestinal system

ILHAMI YUKSEL1, OGUZ USKUDAR1, SEYFETTIN KOKLU1, OMER BASAR1,

SELCAN GULTUNA1, SELMAN UNVERDI2, ZEYNEL A. OZTURK2, DEMET SENGUL3,

ATA TURKER ARIKOK4, OSMAN YUKSEL5 & SAHIN COBAN6

1Department of Gastroenterology, Ankara Dıskapı Yıldırım Beyazıt Education and Research Hospital, Ankara, Turkey,2Department of Internal Medicine, 3Department of Pathology, Ankara Education and Research Hospital, Ankara, Turkey,4Department of Pathology, Ankara Dıskapı Yıldırım Beyazıt Education and Research Hospital, Ankara, Turkey,5Department of Gastroenterology, Ankara Numune Education and Research Hospital, Ankara, Turkey, and 6Department of

Gastroenterology, Kırıkkale Yuksek Ihtısas Hospital, Kırıkkale, Turkey

AbstractObjective. Ectopic gastric tissue in the esophagus (inlet patch) mostly presents in the upper part of the esophagus and isusually under-diagnosed because of its localization. Little is known about its pathogenesis and significance. The aim of thisstudy was to investigate whether there is an association between ectopic gastric tissue development and endoscopic featuresof the upper gastrointestinal tract, especially in the esophagus. Material and methods. A total of 9437 endoscopicexaminations were analyzed prospectively. Endoscopic features and histological examinations of inlet patch and stomachspecimens were documented. Endoscopic findings in patients with inlet patch were compared with those in patients withoutinlet patch. Results. Inlet patch was present in 171 (1.8%) of all patients. Forty-three (25.1%) patients with inlet patch and519 (5.6%) patients without inlet patch had esophagitis (p�0.000). Histologically proven Barrett’s esophagus was morefrequent among patients with inlet patch than among patients without inlet patch (3.5% versus 0.5%, p�0.000).Prevalences of hiatal hernia in the two groups were similar. Open cardia was diagnosed more frequently in the inlet patchgroup than in the other group (24.5% versus 10.0%, p�0.000). Helicobacter pylori colonization was detected in only 11% ofinlet patch specimens, whereas 58% of stomach specimens from the same patients contained H. pylori colonies.Conclusions. Patients with inlet patch seem to have predisposing factors for gastroesophageal reflux, and Barrett’sesophagus is found more frequently in those patients. H. pylori colonization is involved in ectopic gastric tissue lessfrequently than in gastric tissue.

Key Words: Barrett’s esophagus, ectopic gastric tissue, esophagus, heterotopic gastric tissue, inlet patch, open cardia

Introduction

Ectopic (heterotopic) gastric tissue in the cervical

esophagus is known as inlet patch, and consists of

antral or fundic type gastric mucosa. Schmidt first

described inlet patch (IP) in 1805 [1]. IP can occur in

a single site or in multiple sites, and can range from a

few millimeters to 5 cm in size [2]. It can be

overlooked if specific attention is not given to the

proximal esophagus. Although IP is generally asymp-

tomatic, it can cause dysphagia, chest pain, cough or

dyspnea. Ulcer, fistula, intestinal metaplasia, high-

grade dysplasia and adenocarcinoma are rare com-

plications of IP [3�6].

There is some speculation about the pathogenesis

of IP, which is generally considered as a congenital

condition, resulting from an incomplete embryolo-

gic esophageal epithelization process. The columnar

epithelium of the embryo’s esophagus is gradually

replaced by squamous cell epithelium. This process

starts in the mid-esophagus and extends vertically

in both directions, the cervical esophagus being the

last region to become stratified. If the squamous

Correspondence: Seyfettin Koklu, MD, Karargahtepe mahallesi, Kumrulu sokak, 18/1, Kecioren, Ankara, Turkey. Tel: �90 312 3612 568. Fax: �90 312

3124 120. E-mail: gskoklu@yahoo.com

Scandinavian Journal of Gastroenterology, 2008; 43: 910�914

(Received 24 November 2007; accepted 12 February 2008)

ISSN 0036-5521 print/ISSN 1502-7708 online # 2008 Informa UK Ltd.

DOI: 10.1080/00365520801986619

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y B

row

n U

nive

rsity

on

10/2

9/14

For

pers

onal

use

onl

y.

epithelization remains incomplete, the persisting

columnar-lined area differentiates to IP [7].

Although IP shares similar morphological charac-

teristics with Barrett’s esophagus, there are no

adequate data on whether they have a common

etiopathogenesis [3,7]. Moreover, an association

between IP and gastroesophageal reflux has also

been reported [7]. Overall, the etiopathogenesis and

clinical associations of IP need to be clarified.

The pertinent literature about IP includes a small

number of case series [8�10]. To our knowledge, our

report contains the largest series of IP in the

literature. In this study the aim was to analyze the

probable associations between IP and endoscopic

features of the upper gastrointestinal tract, especially

the esophagus. We also determined the prevalence in

a large endoscopic series.

Material and methods

Design

All patients undergoing an upper gastrointestinal

endoscopy (UGE) at the Dıskapı Yıldırım Beyazıt

Hospital and the Ankara Education and Research

Hospital during the period July 2006�July 2007 were

included in this prospective study. We surveyed 9437

consecutive adult patients (�16 years old) who were

referred for endoscopies from several departments

and outpatient clinics for upper gastrointestinal

system symptoms. Patients requiring urgent endos-

copies for upper gastrointestinal bleeding and in-

complete or duplicate procedures were excluded.

Data collection was carried out based on the endo-

scopy reports.

Endoscopic procedure

Written informed consent was obtained from all

patients before commencement of the endoscopic

procedures. UGE was done by four endoscopists

using a videogastroscope with forward viewing

(Fujinon EVE S400). All examinations were per-

formed with the patient in the left lateral decubitus

position. Premedication consisted mainly of topical

anesthesia only, and in combination with midazolam

2.5�5 mg and/or meperidine 25�50 mg i.v. when

needed. Special attention was paid to the proximal

part of the esophagus, especially while the endo-

scope was being slowly withdrawn.

A yellowish pink color with a velvet-like soft

appearance in the upper esophagus was defined as

IP (Figure 1). In all cases, we also documented the

endoscopic findings of the upper gastrointestinal

system including esophagitis, Barrett’s esophagus,

hiatal hernia, open cardia, gastritis, bulbitis and

ulcer. We compared the prevalences with those of

patients that did not have IP.

Patients were classified as having esophagitis if the

endoscopic examination of the esophagus revealed

ulcers, strictures, or erosive or exudative lesions.

Diagnoses of Barrett’s esophagus were based on

histologically proven intestinal metaplasia within the

distal esophagus, and were analyzed for presence or

absence. Hiatal hernia was diagnosed if the gastric

folds extended at least 3 cm proximal to the

diaphragmatic hiatus during normal respiration.

Open cardia was diagnosed when the cardia did

not encompass the endoscope snugly despite deep

inspiration within a poor insufflate stomach at

inversion of the endoscope.

Histopathological assessment

Biopsies were taken from the 69 IP patients. Biopsies

were also obtained from the antrum and/or corpus of

these patients. The Sydney classification method was

employed in histopathological evaluation of the

biopsy samples. In histological analysis, cell type of

ectopic tissue was determined depending on the

presence of parietal and chief cells (fundic or antral

type). The presence of Helicobacter pylori colonies in

both the IP and gastric body was documented.

Statistical analysis

All data were analyzed using the SPSS 12 software

program. Results were expressed as mean values

9standard deviation. For statistical comparison,

Student’s t-test and the x2 test were used whenever

appropriate. A p-value of less than 0.05 was taken as

the level of significant difference.

Figure 1. Endoscopic appearance of inlet patch in the cervical

esophagus.

Inlet patch 911

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y B

row

n U

nive

rsity

on

10/2

9/14

For

pers

onal

use

onl

y.

Results

There were 171 and 9266 cases with and without IP,

respectively. The prevalence of IP among all patients

who underwent endoscopy was 1.8%. Gender dis-

tribution and mean age were similar for both groups

(Table I).

Forty-three (25.1%) patients with IP and

519 (5.6%) patients without IP had esophagitis

(pB0.05). Patients with IP had higher frequencies

of histologically proven Barrett’s esophagus than

those without IP (3.5% versus 0.5%) (pB0.05).

Prevalence of hiatal hernia was comparable among

both groups (13.5% and 17.0% in patients with and

those without IP, respectively). Open cardia was

diagnosed more frequently in the IP group (24.5%

versus 10.0%) (pB0.05) (Table I).

In the IP group, 117 and 24 patients had gastritis

and gastritis plus bulbitis, respectively. Gastric or

duodenal ulcer was diagnosed in 10 patients. Ten

patients had normal endoscopic findings and the

remaining 10 had other findings such as telangiecta-

sia, polyps, etc. Pylor dysfunction was present in 13

(7.6%) cases.

Diameter of IP was 10.297.3 mm (range from 3

to 40 mm). Biopsies from the IP and antrum and/or

corpus were taken in 40.4% of IP patients. Ectopic

gastric tissue was demonstrated histologically in 55

(80.9%) of 69 patients. Biopsy was unsuccessful in

the remaining patients because of technical difficul-

ties or because the samples were too small for

histological analysis. There was no atrophy, dyspla-

sia or adenocarcinoma in the IP biopsy specimens.

Cell types of IP were fundic, antral and transitional

in 30%, 60% and 10% of the cases, respectively.

H. pylori colonization was detected in 6 of 55

(10.9%) ectopic gastric tissues. Those cases also

had H. pylori in their gastric biopsies. Fifty-eight

percent of antrum and/or corpus biopsies were H.

pylori positive. All of the H. pylori-positive ectopic

tissues were of fundic type, with the exception of

one sample.

Discussion

In this prospective study, the prevalence of IP in the

cervical esophagus was 1.8% in sequential upper

endoscopies. The prevalence of H. pylori in IP is

significantly low when compared with gastric

H. pylori. We also demonstrated that patients with

IP have a predisposition for gastroesophageal reflux

disease. Moreover Barrett’s esophagus was more

frequently present in patients with IP than in

patients without IP.

The largest autopsy series revealed that IP pre-

valence was 4.5% in 1000 autopsies in children [11].

In an endoscopic study, IP was present in 0.1�10%

of cases [12,13]. The difference depends on the

heterogeneity of populations and, of course, sample

size. Within our ethnic origin, IP has been reported

in 1.6�3.6% of cases in two studies [10,14]. The

exact prevalence of IP is vague, since the localization

of IP in the upper esophagus is not readily accessible

by endoscope and the region is quickly passed when

the endoscope protrudes over the sphincter. There-

fore, qualified endoscopists and large series are

necessary to detect a clear prevalence of IP. Our

large esophagogastroduodenoscopy (EGD) series,

which is set up by our experienced endoscopists,

revealed an IP prevalence of 1.8% in the Turkish

population. This result is not as high as those

mentioned in the literature.

H. pylori colonization in heterotrophic gastric

mucosa has been found to be 5.3% to 73% in the

reported series [12,15�17]. H. pylori colonization of

IP closely relates to H. pylori density in the stomach

[17]. In this study, H. pylori colonization was present

in about one-tenth of the IP mucosa, and those

patients also had H. pylori in their gastric biopsies.

This finding is lower than the previous prevalences

[12,15�17]. Lower H. pylori colonies in IP may be

due to contact with saliva, being small in diameter

and containing few parietal cells and thus low acid

secretion by IP. Because of difficulties in sampling IP

tissue, inaccurate targeting can result in a higher

ratio of biopsy failure (about one-fifth in the present

study). Hence, biopsy samples that are too small and

superficial may also affect determination of H. pylori

in IP. Moreover, specific stains have been used in

order not to miss detecting H. pylori histologically

[17].

There is some speculation about IP and Barrett’s

esophagus having a shared pathogenesis [3,18,19]. It

has been suggested that there might be a common

embryological origin [3] and acid reflux injury may

contribute to the development of both conditions

[8,18]. Several studies have reported a higher

frequency (20% to 34%) of such an association

[3,9]. Our findings also demonstrated such an

Table I. Demographics and endoscopic findings in patients and

controls.

Inlet patch

group n: 171

Control

group n: 9266 p-value

Age 40.19913.65 42.54917.24 0.076

Gender

Female/Male 97/74 5299/3967 0.900

Barrett’s esophagus 6 (3.5%) 46 (0.5%) 0.000

Esophagitis 43 (25.1%) 519 (5.6%) 0.000

Hiatal hernia 23 (13.5%) 1575 (17.0%) 0.220

Open cardia 42 (24.5%) 927 (10.0%) 0.000

912 I. Yuksel et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y B

row

n U

nive

rsity

on

10/2

9/14

For

pers

onal

use

onl

y.

association, but the prevalence was much lower. The

difference may, on the one hand, be due to a low

number of cases in the previous reports. On the

other hand, Barrett’s esophagus includes metaplasia

and is considered to be a premalignant lesion,

whereas metaplasia is not common in IP and IP

rarely results in carcinoma. The degree of acid

exposure may determine such a differentiation of

both of these embryologically similar entities.

More than one-third of our IP cases had predis-

posing factors (hiatal hernia and open cardia) for

gastroesophageal reflux disease. Moreover, one-

quarter of patients with IP had evidence of reflux

(esophagitis). Previous reports have also described

such associations [3,10,20]. In the present study, we

found a significant relationship between IP and

Barrett’s esophagus and esophagitis but not with

hiatal hernia. This may be due to the considerable

difference between the number of cases of IP and

control groups (171 versus 9266). Furthermore, not

all patients with hiatal hernia have gastric acid reflux.

Although acid secretion has not been demonstrated

in all IPs, there is evidence of acid secretion by IP, at

least by those of greater size [20]. However, it is not

clear whether esophageal symptoms and metaplasia

or carcinoma of IP are attributable to gastric acid or

acid secreted by IP, or both. To clarify these

questions, we need to have esophageal manometric

and pH investigations of both the proximal and distal

esophagus in further studies.

The pertinent literature does not clarify many of

the issues surrounding IP. It is not clear whether

endoscopic surveillance is necessary for all cases.

Since metaplasia and even carcinoma can develop

in IP, it seems that endoscopic follow-up is neces-

sary, especially in symptomatic cases. We should

also clarify whether it is mandatory to take biopsy

samples from all IPs. Generally, it has been

suggested that asymptomatic patients with inciden-

tally diagnosed IP do not require any treatment. A

biopsy to establish an exact diagnosis and to rule

out metaplasia and further differentiation would be

ideal. Patients with symptoms and without mor-

phological changes may benefit from acid suppres-

sion. Finally, cases with morphological (strictures,

fistulas and stenosis) and/or neoplastic changes

should be treated accordingly [7]. Endoscopic

mucosectomy, argon plasma coagulation and surgi-

cal resection are among the invasive treatment

options in those cases.

In conclusion, IP was found in 1.8% of a large

study population. Endoscopic examination of the

upper gastrointestinal tract revealed that many

patients with IP have endoscopic signs of gastro-

esophageal reflux. Barrett’s esophagus is more com-

mon in those patients. H. pylori colonization is found

less frequently in ectopic gastric tissue than in gastric

tissue.

References

[1] Schmidt FA. De Mammalian Esophago atque Ventriculo

[inaugural dissertation]. Halle, in off, Batheana; 1805. Cited

by: Truong LD, Stroehlein JR, McKechnie JC. Gastric

heterotopia of the proximal esophagus: a report of four cases

detected by endoscopy and a review of the literature. Am J

Gastroenterol 1986;/81:/1162�6.

[2] Jacobs E, Dehou MF. Heterotopic gastric mucosa in the

upper esophagus: a prospective study of 33 cases and a

review of the literature. Endoscopy 1997;/29:/710�5.

[3] Avidan B, Sonnenberg A, Chejfec G, Schnell TG, Sontag SJ.

Is there a link between cervical inlet patch and Barrett’s

esophagus? Gastrointest Endosc 2001;/53:/717�21.

[4] Goodwin WJ Jr, Larson DL, Sajjad SM. Adenocarcinoma of

the cervical esophagus in a patient with extensive columnar

cell-lined (Barrett’s) esophagus. Otolaryngol Head Neck

Surg 1983;/91:/446�9.

[5] Klaase JM, Lemaire LC, Rauws EA, Offerhaus GJ, van

Lanschot JJ. Heterotopic gastric mucosa of the cervical

esophagus: a case of high-grade dysplasia treated with argon

plasma coagulation and a case of adenocarcinoma. Gastro-

intest Endosc 2001;/53:/101�4.

[6] Sperling RM, Grendell JH. Adenocarcinoma arising in an

inlet patch of the esophagus. Am J Gastroenterol 1995;/90:/

150�2.

[7] von Rahden BH, Stein HJ, Becker K, Liebermann-Meffert

D, Siewert JR. Heterotopic gastric mucosa of the esophagus:

literature-review and proposal of a clinicopathologic classi-

fication. Am J Gastroenterol 2004;/99:/543�1.

[8] Azar C, Jamali F, Tamim H, Abdul-Baki H, Soweid A.

Prevalence of endoscopically identified heterotopic gastric

mucosa in the proximal esophagus: endoscopist dependent? J

Clin Gastroenterol 2007;/41:/468�71.

[9] Tang P, McKinley MJ, Sporrer M, Kahn E. Inlet patch:

prevalence, histologic type, and association with esophagitis,

Barrett esophagus, and antritis. Arch Pathol Lab Med 2004;/

128:/444�7.

[10] Poyrazoglu OK, Bahcecioglu IH, Dagli AF, Ataseven H,

Celebi S, Yalniz M. Heterotopic gastric mucosa (inlet patch):

endoscopic prevalence, histopathological, demographical

and clinical characteristics. Int J Clin Pract 2007; Epub

ahead of print.

[11] Rector LE, Connerly ML. Aberrant mucosa in the esopha-

gus in infants and children. Arch Pathol Lab Med 1941;/31:/

285�94.

[12] Borhan-Manesh F, Farnum JB. Incidence of heterotopic

gastric mucosa in the upper esophagus. Gut 1991;/32:/968�9.

[13] Maconi G, Pace F, Vago L, Carsana L, Bargiggia S, Bianchi

Porro G. Prevalence and clinical features of heterotopic

gastric mucosa in the upper oesophagus (inlet patch). Eur J

Gastroenterol Hepatol 2000;/12:/745�9.

[14] Akbayir N, Alkim C, Erdem L, Sokmen HM, Sungun A,

Basak T, et al. Heterotopic gastric mucosa in the cervical

esophagus (inlet patch): endoscopic prevalence, histological

and clinical characteristics. J Gastroenterol Hepatol 2004;/

19:/891�6.

[15] Flejou JF, Potet F, Molas G, Bogomoletz WV, Nasca S,

Rigaud C, et al. Campylobacter-like organisms in heterotopic

gastric mucosa of the upper oesophagus. J Clin Pathol 1990;/

43:/961.

[16] Song ZY, Huang X, Qian KD, Peng JP, Sun AW, Zhang YY,

et al. Clinical analysis of 39 cases of heterotopic gastric

Inlet patch 913

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y B

row

n U

nive

rsity

on

10/2

9/14

For

pers

onal

use

onl

y.

mucosa in the upper esophagus. Zhonghua Yi Xue Za Zhi

2005;/85:/244�7.

[17] Gutierrez O, Akamatsu T, Cardona H, Graham DY, El-

Zimaity HM. Helicobacter pylori and hetertopic gastric

mucosa in the upper esophagus (the inlet patch). Am J

Gastroenterol 2003;/98:/1266�70.

[18] Malhi-Chowla N, Ringley RK, Wolfsen HC. Gastric meta-

plasia of the proximal esophagus associated with esophageal

adenocarcinoma and Barrett’s esophagus: what is the con-

nection? Inlet patch revisited. Dig Dis 2000;/18:/183�5.

[19] Feurle GE, Helmstaedter V, Buehring A, Bettendorf U,

Eckardt VF. Distinct immunohistochemical findings in

columnar epithelium of esophageal inlet patch and of

Barrett’s esophagus. Dig Dis Sci 1990;/35:/86�92.

[20] Baudet JS, Alarcon-Fernandez O, Del Rio AS, Aguirre-Jaime

A, Leon-Gomez N. Heterotropic gastric mucosa: a signifi-

cant clinical entity. Scand J Gastroenterol 2006;/41:/1398�404.

914 I. Yuksel et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y B

row

n U

nive

rsity

on

10/2

9/14

For

pers

onal

use

onl

y.