meds capea

8/22/2019 Meds Capea

1/19

www.medscape.com

Histology of Symptomatic GastroesophagealReflux Disease

Is It Predictive of Response to Proton Pump Inhibitors?

Hiroto Miwa, Kaiyo Takubo, Tomohiko Shimatani, Takahisa Furuta, Tadayuki Oshima, Junji

Tanaka, Junko Aida, Masanori Ito, Susumu Kurosawa, Takashi Joh, Tsuneya Wada, Yasuki

Habu, Yusuke Watanabe, Michio Hongo, Tsutomu Chiba, Yoshikazu Kinoshita

J Gastroenterol Hepatol. 2013;28(3):479-487.

Abstract and Introduction

Abstract

Background and Aim: To examine the differences in esophageal histopathology between

non-erosive reflux disease (NERD) and reflux esophagitis (RE), and to investigate whether

baseline esophageal histopathology can predict the therapeutic response to proton pump

inhibitors (PPIs).

Method: The subjects comprised 94 patients with NERD (n = 71) or mild RE (n = 23). Tissuewas biopsied from 5 cm above the squamo-columnar junction (SCJ), and the degree or

presence of nine histopathological markers was assessed. The patients were treated with

rabeprazole (RPZ) 10 mg once daily for 4 weeks. If complete heartburn relief was not

achieved, RPZ was increased to 10 mg twice daily for another 2 weeks, and then to 20 mg

twice daily for another 2 weeks if heartburn remained.

Results: Features of esophageal histopathology 5 cm above the SCJ differed between NERD

and RE patients. The esophageal histopathology in patients unresponsive to RPZ was

characterized by Protein Gene Product (PGP) 9.5 negativity in those with NERD, and

intraepithelial bleeding in those with RE. In addition, the combination of dilated intercellular

spaces (DIS) (+)/PGP 9.5 () was indicative of strong resistance to PPI therapy in NERDpatients.

Conclusion: The therapeutic efficacy of PPI can be predicted from the features of biopsied

esophageal tissue. Factors predictive of resistance to treatment with PPI are negativity for

PGP 9.5 in NERD patients and intraepithelial bleeding in RE patients.

Introduction

Patients with gastroesophageal reflux disease (GERD) have been increasing in number in

recent years.[1] GERD is commonly treated with proton pump inhibitors (PPIs), but a high

proportion of patients show no response.[2,3] Cases of GERD in which the symptoms cannotbe relieved by PPI therapy are an important medical concern. Because about 20% of patients
http://www.medscape.com/http://www.medscape.com/http://www.medscape.com/


2/19

with reflux esophagitis (RE) and about 50% of patients with non-erosive reflux disease

(NERD) do not respond to PPI therapy,[4] it is desirable to obtain information on whether PPI

therapy would be effective or ineffective in individual patients before the start of therapy,

especially for those with NERD. In other words, it would be desirable to identify non-

responders to PPI before starting the therapy. Nocturnal gastric acid breakthrough (NAB),[5]

gene polymorphism of CYP2C19 (extensive metabolizer),[6]

esophageal reflux of duodenalfluid,[2] female gender, and poor treatment compliance have been noted as factors associated

with resistance to PPI therapy. However, no previous study has investigated whether

histopathological findings in the esophagus are predictive of resistance to PPI treatment.

It has been reported that GERD patients show changes in various esophageal

histopathological parameters,[79] and in recent years, consensus guidelines for histological

recognition of microscopically evident esophagitis in patients with GERD have been

formulated.[10] These parameters include an increased eosinophil count, intrapapillary blood

vessel dilatation, intraepithelial bleeding, papillary extension, basal cell hyperplasia

(thickening), dilated intercellular spaces (DIS), enhanced cellular proliferation, appearance of

Langerhans cells, and hyperplasia of nerve fibers. However, no study has compareddifferences in esophageal histopathological findings between patients with NERD and those

with RE. We therefore examined differences in esophageal histopathology between these

patient groups, and investigated whether the pathological features of esophageal biopsy

samples obtained before the start of treatment could predict the therapeutic efficacy of PPI.

This investigation was conducted as part of the TORNADO (Treatment with high dose Of

Rabeprazole for NERD patients conducted by AciD-related symptOm research group)

study.[11]

Methods

Patient Enrollment

This was a multi-center, cooperative open study in Japan belonging to the Acid-Related

Symptom (ARS) Research Group. The study was started after approval from the Ethics

Committee of each medical institution, and conducted in compliance with the Declaration of

Helsinki.

Outpatients with GERD who satisfied the following conditions were included in this study:

those who were 20 years of age or older, those who gave written informed consent to

participate in this study of their own free will, those who agreed to undergo esophagealbiopsy, and those who had suffered heartburn repeatedly at least once a week during a period

of one month before the study. Heartburn was defined as a "burning sensation" welling up

from the stomach or lower thoracic region that tended to occur after meals, or when in a bent-

over position or upon abdominal compression. The Modified Los Angeles Classification[12,13]

was used for diagnosis by esophageal endoscopic examination. Only patients with grade N or

M NERD and only patients with grade A or B RE were included in the study.

Since the following conditions may affect evaluation of the therapeutic effect of PPIs, patients

to whom these conditions applied were excluded: those who had been definitively diagnosed

as having RE by endoscopic examination in the past, those who had used a PPI within one

month previously, those who had a history of upper digestive tract surgery, those who had anactive gastric ulcer or duodenal ulcer, those who had Barrett's esophagus (long-segment


3/19

Barrett's esophagus: LSBE), those who had a history or presence of angina pectoris, those

who had malignant disease, serious hepatic disease, or serious renal disease, and other

patients judged inappropriate for the study by the investigators.

Study Protocol

After acquisition of informed consent, biopsy samples were taken from the esophageal

mucosa using an endoscope. In the "Treatment Period" at the beginning of the study, a 10-mg

rabeprazole (RPZ) tablet was administered orally once daily for 4 weeks. Changes in the

symptoms of heartburn over time were recorded using the following four grades in a self-

administered patient diary every day. No symptoms, bothersome but tolerable (mild),

bothersome (moderate), very bothersome (severe).

Patients having complete heartburn relief (defined as no heartburn during the previous 7 days)

were defined as "Responders to 4-week treatment," and completed the study drug

administration at Week 4. Patients not having complete heartburn relief at the end of the

Treatment Period were defined as "Non responders to 4-week treatment" and the dose of RPZwas increased. Such patients received oral administration of a RPZ 10-mg tablet twice daily

for 2 weeks in "Continuous Treatment Period I." Patients who had complete heartburn relief

at the end of Continuous Treatment Period I (after successive 6-week administration)

completed the study drug administration. Patients who did not have complete heartburn relief

at the end of Continuous Treatment Period I received a further increased dose of RPZ orally

with a RPZ 20-mg tablet twice daily for another 2 weeks (successive 8-week administration)

in "Continuous Treatment Period II." Patients who were completely relieved of heartburn

after the 4-week, 6-week, or 8-week treatment were defined as "Cumulative responders to 8-

week treatment," while patients who were not completely relieved after the 8-week treatment

were defined as "Cumulative non-responders to 8-week treatment."

During the period of RPZ administration, concomitant use of the following drugs and therapy

was prohibited: other PPIs, H2 receptor antagonists, M3 receptor antagonists, prokinetics, anti-

cholinergic agents, antacids, anti-gastrin agents, prostaglandin preparations, mucosal

protective agents, sodium alginate, andHelicobacter pylori eradication therapy.

In investigation 1, one endoscopic biopsy specimen was obtained from each patient before

treatment. The features of nine histopathological and immunohistopathologically stained

sections () from each NERD patient were compared with those from RE patients. PGP 9.5 is a

protein belonging to the ubiquitin C-terminal hydrolase proteins and highly expressed in

vertebral neurons and nerve fibers.

[14]

Free ending nerves have been described in interepithelial spaces and probably mediated esophageal pain and dysphagia in patients GERD.[15]

To estimate increased number of nerve fibers in patients with GERD, we stained

immunohistochemical biopsy specimens with antibody for PGP 9.5. In Investigation 2, the

therapeutic effects of RPZ (rate of complete heartburn relief after the 4-week treatment and

rate of cumulative relief after the 8-week treatment) in NERD patients and RE patients were

calculated and aggregated. Then, the aggregated data and findings for the nine different

histopathological and immunohistopathological sections were compared between

"Responders to 4-week treatment" and "Non-responders to 4-week treatment," and

"Cumulative responders to 8-week treatment" and "Cumulative non-responders to 8-week

treatment."

Table 1. Histopathological parameters investigated


4/19

Item Evaluation

EosinophilsTotal number of intraepithelial

eosinophils in biopsy sample

Intrapapillary vessel dilatation

Presence of intrapapillary vessel

dilatation

Intraepithelial bleeding

Presence of red blood cell

diapedesis from intrapapillary

vessels into epithelium

Dilated intercellular spaces (DIS) Positive (+) or negative ()

Papillary extension(Length of papilla)/(thickness of

whole epithelium) 100 = %

Basal cell hyperplasia

(Thickness of basal cell

layer)/(thickness of whole

epithelium) 100 = %

Ki-67 immunoreactivity (MIB-1; anti-human

monoclonal antibody, Dako Cytomation, Glostrup,

Denmark)

Number of cell layers positive for

Ki-67 (cellular marker of

proliferation = marker of DNA

synthesis)

Langerhans cells (positivity on CD1a immunostaining;

CD1a101, Dako Cytomation, Glostrup, Denmark)

Number of positive cells (from

most concentrated area, cell nuclei

must be recognized in a 400

view)

PGP 9.5 immunoreactivity (a soluble protein isolatedfrom brain, used as a general marker for neuronal and

neuroendocrine tissue;14 anti-PGP 9.5 polyclonal

antibody, Dako Cytomation, Glostrup, Denmark)

Number of positively stained nerve

fiber sections per squamous

epithelial area (2 3 mm)

It has been reported that, in gastroesophageal reflux disease (GERD) patients, the basal cell

layer is thickened (basal cell hyperplasia) to 1550% of the thickness of stratified squamousepithelium and that the thickness of the lamina propria mucosal papilla exceeds 1/22/3 of the

thickness of the epithelium (papillary extension).16 In addition, it has been reported that if Ki-

67 (MIB-1) immunoreactivity is positive in three or more layers of esophageal epithelium

cells, some type of abnormality consistent with GERD is present.

17

Histopathological Features of the Esophageal Mucosa

One biopsy specimen was taken from 5 cm above the squamo-columnar junction (SCJ) using

an endoscope. The specimen was taken from part of the anterior wall of the esophagus and at

the top of the esophageal mucosal plicae; any erosion or ulcer was excluded from biopsy. The

biopsy specimen was fixed in 10% formalin solution, embedded in paraffin, subjected to

hematoxylin-eosin staining or immunostaining, and histopathologically assessed at the Tokyo

Metropolitan Institute of Gerontology using light microscopy. Two pathologists (KT, JA)

blind to the patient backgrounds assessed the degree or presence of nine items[79] listed in .

Table 1. Histopathological parameters investigated


5/19

Item Evaluation

EosinophilsTotal number of intraepithelial

eosinophils in biopsy sample

Intrapapillary vessel dilatation

Presence of intrapapillary vessel

dilatation

Intraepithelial bleeding

Presence of red blood cell

diapedesis from intrapapillary

vessels into epithelium

Dilated intercellular spaces (DIS) Positive (+) ornegative ()

Papillary extension(Length of papilla)/(thickness of

whole epithelium) 100 = %

Basal cell hyperplasia

(Thickness of basal cell

layer)/(thickness of whole

epithelium) 100 = %

Ki-67 immunoreactivity (MIB-1; anti-human

monoclonal antibody, Dako Cytomation, Glostrup,

Denmark)

Number of cell layers positive for

Ki-67 (cellular marker of

proliferation = marker of DNA

synthesis)

Langerhans cells (positivity on CD1a immunostaining;

CD1a101, Dako Cytomation, Glostrup, Denmark)

Number of positive cells (from

most concentrated area, cell nuclei

must be recognized in a 400

view)

PGP 9.5 immunoreactivity (a soluble protein isolatedfrom brain, used as a general marker for neuronal and

neuroendocrine tissue;14 anti-PGP 9.5 polyclonal

antibody, Dako Cytomation, Glostrup, Denmark)

Number of positively stained nerve

fiber sections per squamous

epithelial area (2 3 mm)

It has been reported that, in gastroesophageal reflux disease (GERD) patients, the basal cell

layer is thickened (basal cell hyperplasia) to 1550% of the thickness of stratified squamousepithelium and that the thickness of the lamina propria mucosal papilla exceeds 1/22/3 of the

thickness of the epithelium (papillary extension).16 In addition, it has been reported that if Ki-

67 (MIB-1) immunoreactivity is positive in three or more layers of esophageal epithelium

cells, some type of abnormality consistent with GERD is present.

17

Immunostaining with the two antibodies was carried out on 3-m-thick tissue sections using

ChemMate Envision/HRP (Polymer-Immuno Complex method, Dako Cytomation, Glostrup,

Denmark). Micrographs of intraepithelial bleeding, dilated intercellular spaces (DIS), Ki-67

immunoreactivity, and PGP 9.5 immunoreactivity are shown in Figure 1.


6/19

Figure 1.

Histopathological findings. (a) Intraepithelial bleeding in a 41-year-old man with m-LA grade

M, who achieved complete heartburn relief after 4 weeks; (b) Dilated intercellular spaces

(DIS) circled (dotted line) in a 54-year-old woman with m-LA grade M, who achieved

complete heartburn relief after 4 weeks; (c) Ki-67 immunoreactivity (34 cell layers) in a 55-

year-old man with m-LA grade B, who achieved complete heartburn relief after 4 weeks; (d)

Protein Gene Product (PGP) 9.5 immunoreactivity revealed as dots or dotted lines (arrows) in

intercellular spaces of the epithelium in a 23-year-old woman with m-LA grade N, who

achieved complete heartburn relief after 6 weeks.

m-LA, modified Los Angeles Classification.


7/19

Statistical Analysis

Data distributions were determined by summary statistics for continuous variables and by

frequency tables for classified variables. The significance level for statistical tests was set at

5% on both sides. Intergroup comparisons of demographic data for the study subjects and

esophageal histopathological findings were performed between NERD and RE patients usingthe two-sample Wilcoxon test or 2 test. In addition, intragroup comparisons of each

esophageal histopathological feature were also performed between the two patient groups

using the two-sample Wilcoxon test, 2 test, or Fisher's exact test.

For pathologic parameters deemed likely to exhibit significant differences between responders

and non-responders, subgroup analyses were performed for complete heartburn relief rates

after the 4-week treatment, and for cumulative complete heartburn relief rates after the 8-

week treatment.

Results

Subject Disposition and Demographics

Ninety-four (94) subjects who gave consent to participate in the study, and for whom

esophageal histopathological findings were evaluable, were included in the analyses. These

subjects comprised 71 patients with NERD and 23 patients with RE. Disease severity graded

by the modified Los Angeles Classification was Grade N in 24 subjects, Grade M in 47

subjects, Grade A in 17 subjects, and Grade B in six subjects. Demographics of the subjects

are shown for NERD and for RE separately in .

Table 2. Patient demographics (non-erosive reflux disease [NERD]vs

reflux esophagitis[RE])

Items NERD (n= 71) RE (n= 23) P-value

Gender (male/female) 32/39 17/6 0.016*a

Age, years (mean SD) 44.6 17.8 41.7 18.5 0.428

BMI, kg/m (mean SD) 22.5 3.4 23.9 5.9 0.402

Frequency of heartburn, days/week (mean SD) 4.3 2.4 3.7 2.1 0.308

Severity of heartburn (%)

Mild 59.2 43.5 0.351a

Moderate 35.2 52.2

Severe 5.6 4.3

Hiatal hernia + (%) 40 52.2 0.306a

Atrophy: Kimura-Takemoto Classification (%)

C-O 42.9 52.2 0.738a

C-I-III 41.4 34.8

O-I-III 15.7 13.0


8/19

Helicobacter pylori + (%) 31 17.4 0.205a

CYP2C19 genotype (%)

homo EM 38.2 34.8 0.679a

hetero EM 44.1 39.1PM 17.6 26.1

Smoking+: sometimes +every day (%) 21.4 43.5 0.103a

Alcohol+: sometimes + every day (%) 47.1 56.5 0.220a

Caffeine +: sometimes +every day (%) 71.6 69.6 0.864a

*P< 0.05,a)2 test, b)Wilcoxon two-sample test, SD = standard deviation.n = 70.n = 68.n

= 67.

Comparison of Esophageal Histopathological Findings Between NERD and RE

Nine (9) esophageal histopathological markers in biopsy specimens taken before treatment

were compared between NERD and RE, and the results are shown in . There was no

statistically significant difference in any marker between NERD and RE. Though there was

no significant difference in the proportions of patients showing Ki-67 immunoreactivity in 3

cell layers, the proportion in the RE group (61.9%) was higher than that in the NERD group

(39.4%) (P= 0.071). In addition, the mean number of nerve fibers showing staining for PGP

9.5 (per squamous epithelial area) was higher in the NERD group than in the RE group (4.4 vs

1.9 respectively;P= 0.078).

Table 3. Comparison of baseline histopathological parameters in the non-erosive reflux

disease (NERD) vsreflux esophagitis (RE) groups

Items NERD (n) RE (n)P-

value

Number of eosinophils (mean SD)0.1 0.3

(68)

0.0 0.2

(23)0.990b)

Proportion of patients with intrapapillary vessel

dilatation (+)30.3% (66) 36.4% (22) 0.597a)

Proportion of patients with intraepithelial bleeding (+) 47.8% (67) 52.2% (23) 0.715a

Proportion of patients with dilated intercellular spaces

(DIS) (+)70.4% (71) 73.9% (23) 0.448a)

Papillary extension % (mean SD)47.2 17.2

(49)

52.8 12.6

(17)0.192b)

Basal cell hyperplasia % (mean SD)10.6 4.8

(55)

11.0 7.3

(17)0.651b)

Proportion of patients with Ki-67 (+) immunoreactivity

3 cell layers

39.4% (66) 61.9% (21) 0.071a)

Number of Langerhans cells (+) (mean SD) 6.8 2.6 6.9 3.4 0.557


9/19

(66) (23)

Number of PGP 9.5 (+) nerve fibers (/unit area) (mean

SD)

4.4 5.1

(51)

1.9 2.5

(15)0.078b)

*P< 0.05,a)

2

test,b)

Wilcoxon two-sample test. Note: The number of patients is differentaccording to the histopathological parameters because of presence of biopsy specimens unable

to assess by each parameter.

Comparison of Esophageal Histopathology Between Responders and Non-responders in

the NERD Group

Histopathological features of the esophagus in NERD patients were compared between

responders and non-responders to RPZ for the 4-week treatment and between responders and

non-responders (cumulative) for the 8-week treatment, and the results are shown in .

Table 4. Comparison of histopathological parameters in the esophagus after 4 or 8weeks of rabeprazole (RPZ) treatment between responders and non-responders

Items Time NERD group RE group

Responde

rs (n)

Non-

responde

rs (n)

P-

value

Responde

rs (n)

Non-

responde

rs (n)

P-

value

Number of

eosinophils

(mean SD)

4 weeks0.0 0.0

(27)

0.1 0.4

(34)0.121b)

0.0 0.0

(17)

0.2 0.4

(6)0.113b)

8 weeks(Cumulativ

e)

0.1 0.4

(37)

0.0 0.2

(24)0.828b)

0.0 0.0

(19)

0.3 0.5

(4)

0.039*b)

Proportion of

patients with

intrapapillary

vessel

dilatation (+)

4 weeks23.1%

(26)

33.3%

(33)0.388a)

25.0%

(16)66.7% (6) 0.070a)

8 weeks

(Cumulativ

e)

25.7%

(35)

33.3%

(24)0.526a)

33.3%

(18)50.0% (4) 0.531a)

Proportion of

patients withintraepithelial

bleeding (+)

4 weeks57.7%

(26)

44.1%

(34)0.297a)

41.2%

(17)83.3% (6) 0.076a)

8 weeks

(Cumulativ

e)

52.8%

(36)

45.8%

(24)0.598a)

42.1%

(19)100% (4)

0.035*a)

Proportion of

patients with

dilated

intercellular

spaces (DIS)

(+)

4 weeks56.7%

(30)

79.4%

(34)0.050a)

76.5%

(17)66.7% (6) 0.638a)

8 weeks

(Cumulativ

e)

62.5%

(40)

79.2%

(24)0.164a)

73.7%

(19)75.0% (4) 0.957a)

Papillary 4 weeks 43.8 49.2 0.273 53.3 51.6 9.9 0.597


10/19

extension %

(mean SD)

18.8 (18) 17.0 (25) 13.9 (12) (5)

8 weeks

(Cumulativ

e)

46.6

18.4 (26)

47.6

17.2 (17)0.931b)

53.8

13.5 (13)

49.5

10.1 (4)0.427b)

Basal cellhyperplasia %

(mean SD)

4 weeks 10.8 5.6(22)

10.7 4.4(27)

0.739b) 11.4 8.4(12)

10.0 4.3(5)

0.958b)

8 weeks

(Cumulativ

e)

10.9 5.5

(30)

10.6 4.1

(19)0.829b)

10.8 8.3

(13)

11.5 3.8

(4)0.393b)

Proportion of

patients with

Ki-67 (+)

immunoreactiv

ity 3 cell

layers

4 weeks42.3%

(26)

42.4%

(33)0.993a)

73.3%

(15)33.3% (6) 0.088a)

8 weeks

(Cumulativ

e)

40.0%

(35)

45.8%

(24)

0.656a)64.7%

(17)

50.0% (4) 0.586a)

Number of

Langerhans

cells (+) (mean

SD)

4 weeks6.7 2.9

(26)

6.9 2.4

(33)0.859b)

6.3 3.0

(17)

8.5 4.1

(6)0.256b)

8 weeks

(Cumulativ

e)

6.7 2.8

(35)

6.9 2.3

(24)0.779b)

6.6 3.4

(19)

8.3 3.3

(4)0.285b)

Ratio of

patients with

PGP 9.5 (+)

immunoreactivity

4 weeks88.2%

(17)

60.7%

(28)

0.048*a)

50.0%

(10)60.0% (5) 0.714a)

8 weeks

(Cumulativ

e)

84.6%(26)

52.6%(19)

0.019*a)

50.0%(12)

66.7% (3) 0.605a)

*P< 0.05,a)2 test,b)Wilcoxon two-sample test. Note: Responder or non-responder status was

defined as achieving or failing to exhibit lack of heartburn onset during the 7-day period

preceding evaluation, respectively. Patients who were completely relieved of heartburn after

the 4-week, 6-week, or 8-week treatment were defined as "Cumulative responders to 8-week

treatment." The number of patients is different according to the histopathological parameters

because of presence of biopsy specimens unable to assess by each parameter.

The proportion of patients whose biopsy samples showed staining for PGP 9.5 wassignificantly higher in responders to the 4-week treatment and in cumulative responders after

the 8-week treatment (responders vs non-responders: 88.2% [15/17] vs 60.7% [17/28],P=

0.048, for 4-week treatment and 84.6% [22/26] vs 52.6% [10/19],P= 0.019, for cumulative

8-week treatment). (See Fig. 2a "Rate of complete heartburn relief in the NERD group in

relation to presence or absence of PGP 9.5 immunoreactivity").


11/19

Figure 2.


12/19

Rate of complete heartburn relief in the non-erosive reflux disease (NERD) group in relation

to presence or absence of Protein Gene Product (PGP) 9.5 immunoreactivity (a) and dilated

intercellular spaces (DIS) (b). Note: Complete heartburn relief was defined as no heartburn

onset during the 7-day period preceding evaluation.

The proportion of patients whose biopsy samples showed DIS tended to be higher in non-responders to the 4-week treatment and in cumulative responders to the 8-week treatment

(responders vs non-responders: 56.7% [17/30] vs 79.4% [27/34],P= 0.050, for 4-week

treatment and 62.5% [25/40] vs 79.2% [19/24],P= 0.164, for 8-week treatment), although

this difference was not significant. (See Fig. 2b "Rate of complete heartburn relief in relation

to presence or absence of DIS in the NERD group").

There were no significant differences in the other pathological parameters between responders

and non-responders for either the 4-week treatment or the 8-week treatment.

Comparison of Esophageal Histopathology Between Responders and Non-responders in

the RE Group

Histopathological features of the esophagus in RE patients were compared between

responders and non-responders to RPZ for the 4-week treatment and between responders and

non-responders (cumulative) to the 8-week treatment, and the results are shown in .

Table 4. Comparison of histopathological parameters in the esophagus after 4 or 8

weeks of rabeprazole (RPZ) treatment between responders and non-responders

Items Time NERD group RE group

Responde

rs (n)

Non-responde

rs (n)

P-

value

Responde

rs (n)

Non-responde

rs (n)

P-

value

Number of

eosinophils

(mean SD)

4 weeks0.0 0.0

(27)

0.1 0.4

(34)0.121b)

0.0 0.0

(17)

0.2 0.4

(6)0.113b)

8 weeks

(Cumulativ

e)

0.1 0.4

(37)

0.0 0.2

(24)0.828b)

0.0 0.0

(19)

0.3 0.5

(4)

0.039*b)

Proportion of

patients withintrapapillary

vessel

dilatation (+)

4 weeks23.1%

(26)

33.3%

(33)0.388a)

25.0%

(16)66.7% (6) 0.070a)

8 weeks

(Cumulativ

e)

25.7%

(35)

33.3%

(24)0.526a)

33.3%

(18)50.0% (4) 0.531a)

Proportion of

patients with

intraepithelial

bleeding (+)

4 weeks57.7%

(26)

44.1%

(34)0.297a)

41.2%

(17)83.3% (6) 0.076a)

8 weeks

(Cumulativ

e)

52.8%

(36)

45.8%

(24)0.598a)

42.1%

(19)100% (4)

0.035*a)

Proportion ofpatients with

4 weeks 56.7%(30)

79.4%(34)

0.050a) 76.5%(17)

66.7% (6) 0.638a)


13/19

dilated

intercellular

spaces (DIS)

(+)

8 weeks

(Cumulativ

e)

62.5%

(40)

79.2%

(24)0.164a)

73.7%

(19)75.0% (4) 0.957a)

Papillaryextension %

(mean SD)

4 weeks 43.8 18.8 (18) 49.2 17.0 (25) 0.273b) 53.3 13.9 (12) 51.6 9.9(5) 0.597

b)

8 weeks

(Cumulativ

e)

46.6

18.4 (26)

47.6

17.2 (17)0.931b)

53.8

13.5 (13)

49.5

10.1 (4)0.427b)

Basal cell

hyperplasia %

(mean SD)

4 weeks10.8 5.6

(22)

10.7 4.4

(27)0.739b)

11.4 8.4

(12)

10.0 4.3

(5)0.958b)

8 weeks

(Cumulativ

e)

10.9 5.5

(30)

10.6 4.1

(19)0.829b)

10.8 8.3

(13)

11.5 3.8

(4)0.393b)

Proportion of

patients with

Ki-67 (+)

immunoreactiv

ity 3 cell

layers

4 weeks42.3%

(26)

42.4%

(33)0.993a)

73.3%

(15)33.3% (6) 0.088a)

8 weeks

(Cumulativ

e)

40.0%

(35)

45.8%

(24)0.656a)

64.7%

(17)50.0% (4) 0.586a)

Number of

Langerhans

cells (+) (mean

SD)

4 weeks6.7 2.9

(26)

6.9 2.4

(33)0.859b)

6.3 3.0

(17)

8.5 4.1

(6)0.256b)

8 weeks

(Cumulative)

6.7 2.8

(35)

6.9 2.3

(24) 0.779b)6.6 3.4

(19)

8.3 3.3

(4) 0.285b)

Ratio of

patients with

PGP 9.5 (+)

immunoreactiv

ity

4 weeks88.2%

(17)

60.7%

(28)

0.048*a)

50.0%

(10)60.0% (5) 0.714a)

8 weeks

(Cumulativ

e)

84.6%

(26)

52.6%

(19)

0.019*a)

50.0%

(12)66.7% (3) 0.605a)

*P< 0.05,a)2 test,b)Wilcoxon two-sample test. Note: Responder or non-responder status was

defined as achieving or failing to exhibit lack of heartburn onset during the 7-day periodpreceding evaluation, respectively. Patients who were completely relieved of heartburn after

the 4-week, 6-week, or 8-week treatment were defined as "Cumulative responders to 8-week

treatment." The number of patients is different according to the histopathological parameters

because of presence of biopsy specimens unable to assess by each parameter.

The mean eosinophil count was 0 in cumulative responders to the 8-week treatment and 0.3 in

cumulative non-responders to the 8-week treatment, being significantly higher (P= 0.039) in

the latter. The proportion of patients with intraepithelial bleeding was 42.1% in cumulative

responders to the 8-week treatment and 100% in cumulative non-responders to the 8-week

treatment, being also significantly higher (P= 0.035) in non-responders.


14/19

There were no significant differences in other pathological parameters between responders

and non-responders to the 4-week or 8-week treatment.

Relationship Between Presence or Absence of DIS and PGP 9.5 Immunoreactivity in

NERD and RE Patients The above findings suggested that both the presence or absence of

DIS or PGP 9.5 immunoreactivity might be related to the efficacy of RPZ in NERD patients.The efficacy of RPZ was therefore compared among patients classified into the following four

groups: DIS (+)/PGP 9.5 (+), DIS (+)/PGP 9.5 (), DIS ()/PGP 9.5 (+), and DIS ()/PGP 9.5

().

With regard to the efficacy of RPZ in 45 NERD patients, the rate of complete heartburn relief

after 4 weeks was 0% for DIS (+)/PGP 9.5 (), representing the lowest value, and there were

significant differences among the four histopathology groups (P= 0.046) ( ). Similarly, the

rate of cumulative complete heartburn relief after 8 weeks was lowest (22.2%) in the DIS

(+)/PGP 9.5 () group (P= 0.093), although the difference relative to the other groups was

not significant. In 15 RE patients, there were no significant differences in the rates of

complete heartburn relief among the four groups after the 4-week treatment, or in thecumulative rate after the 8-week treatment.

Table 5. Response to rabeprazole (RPZ) therapy based on presence or absence of

dilated intercellular spaces (DIS) and Protein Gene Product (PGP) 9.5 immunoreactivity

in non-erosive reflux disease (NERD) and reflux esophagitis (RE) patients (%, rates of

complete heartburn relief)

Time

Presence or

absence

NERD (n = 45) RE (n = 15)

DIS (+) DIS ()

P-

value DIS (+) DIS ()

P-

value

After 4 weeksPGP 9.5 (+)

47.8%

(11/23)

44.4%

(4/9)0.046*

60.0%

(3/5)

66.7%

(2/3)1.000

PGP 9.5 () 0% (0/9)50.0%

(2/4)

80.0%

(4/5)

50.0%

(1/2)

After 8 weeks

(cumulative)PGP 9.5 (+)

69.6%

(16/23)

66.7%

(6/9)0.093

80.0%

(4/5)

66.7%

(2/3)1.000

PGP 9.5 ()22.2%

(2/9)

50.0%

(2/4)

80.0%

(4/5)

100%

(2/2)

*P< 0.05, Fisher's exact test.

Discussion

The present study is the first to have examined differences in esophageal histopathology

between NERD and RE. Although several reports have described histopathological findings

in the esophagus of patients with GERD,[79] it has been unclear whether they are

physiological characteristics of gastroesophageal reflux and not specific to GERD.[18] One

reason for this lack of clarity is probably the diverse nature of the sites from which biopsy

samples have been obtained.[9,19]


15/19

In the present investigation, biopsy specimens were taken from 5 cm above the SCJ near the

site of an indwelling sensor used for pH monitoring to detect morbid gastric acid reflux. This

biopsy site was selected because it might enable clarification of the relationship between the

duration of exposure to gastric acid and tissue histopathology. Although there was no

significant correlation between any of the histopathological features of esophageal biopsy

specimens and pH parameters (% period during which pH < 4, frequency of reflux) (data notshown), one difference in pathology between NERD and RE was of interest: a greater

proportion of nerve fibers were immunopositive for PGP 9.5 in NERD patients than in RE

patients, though the difference was not statistically significant. This finding may explain the

greater sensory hypersensitivity of NERD patients.[20] On the other hand, a greater proportion

of RE patients had three or more Ki 67-positive cell layers, an indicator of cell proliferation,

than was the case in NERD patients, though the difference was also not statistically

significant. Enhancement of cellular proliferative potential may reflect inflammation due to

mucosal damage by gastric acid. This is consistent with the reported finding that the

expression of Ki 67 increases in proportion to worsening of esophageal histopathology.[21,22]

Although it has been speculated that NERD may be a mild form of RE, the present results

suggest that the spectra of these diseases may differ.

Here we also investigated the histopathological features of the esophagus that might predict

the therapeutic efficacy of PPIs in patients with GERD. In this study, we compared the

esophageal histopathology between responders and non-responders. The results indicated that

among patients with NERD, the proportion of responders to PPI was significantly higher in

those showing PGP 9.5 immunoreactivity in nerve fibers, and that the proportion of non-

responders to PPI was significantly higher in patients lacking such PGP 9.5 immunoreactivity

even after an increase in the PPI dose. As mentioned above, positive PGP 9.5

immunoreactivity in nerve fibers may be related to hypersensitivity of NERD patients.

Therefore, the patients with positive PGP9.5 will respond to the increased dose of PPI, which

somehow decreases the amount of acid or weakly acid reflux. On the other hand, the

negativity to PGP9.5 suggests that their symptoms are not related to acid reflux event.

Therefore such patients are unlikely to respond to the treatment with PPI even if its dosage is

increased. In fact, it is reported that the total esophageal acid exposure time correlates

significantly with density of PGP 9.5 immunoreactivity in NERD patients.[23] Thus, lack of

PGP 9.5 immunoreactivity appeared to be a useful predictive factor of resistance to treatment

with PPI.

Because NERD patients with DIS tended to be more resistant to standard therapy with PPI

than NERD patients without DIS (although not to a significant degree;P= 0.050), DIS was

also considered to be a potentially important factor predicting the efficacy of PPI. DIS hasbeen studied mainly in the context of GERD symptoms.[24] Since it has been reported that DIS

disappears with improvement of heartburn,[25] and that it is evident microscopically in 90% of

RE patients, 68% of NERD patients, and 8% of control subjects, being correlated with

mucosal damage to some extent,[26] it is not surprising that DIS might be a factor predictive of

the therapeutic effectiveness of PPI. In the present study, the difference in the symptom relief

rate between patients with and without DIS decreased (P= 0.164) after PPI dose escalation.

This suggested that strong and long-term suppression of gastric acid might ameliorate DIS.

With regard to the efficacy of PPI, as shown in , NERD patients who showed an especially

low response to the therapy were DIS (+)/PGP 9.5 (), and the difference in the efficacy rate

was significantly lower than in the other three groups. Since these patients continued to showa low efficacy rate even after the dose of RPZ had been increased, these two parameters


16/19

appeared to be potentially predictive of therapeutic effectiveness of PPI. The precise

mechanism why NERD patients with DIS (+)/PGP 9.5 () is still unclear. However, it is not

surprising that the combination of these is likely to serve as a predictor for poor responder to

PPI because positive DIS is a marker of symptoms and negative PGP 9.5 suggests the

symptoms are less related to acid reflux.

Table 5. Response to rabeprazole (RPZ) therapy based on presence or absence of

dilated intercellular spaces (DIS) and Protein Gene Product (PGP) 9.5 immunoreactivity

in non-erosive reflux disease (NERD) and reflux esophagitis (RE) patients (%, rates of

complete heartburn relief)

Time

Presence or

absence

NERD (n = 45) RE (n = 15)

DIS (+) DIS ()P-

valueDIS (+) DIS ()

P-

value

After 4 weeksPGP 9.5 (+)

47.8%

(11/23)

44.4%

(4/9) 0.046*60.0%

(3/5)

66.7%

(2/3) 1.000

PGP 9.5 () 0% (0/9)50.0%

(2/4)

80.0%

(4/5)

50.0%

(1/2)

After 8 weeks

(cumulative)PGP 9.5 (+)

69.6%

(16/23)

66.7%

(6/9)0.093

80.0%

(4/5)

66.7%

(2/3)1.000

PGP 9.5 ()22.2%

(2/9)

50.0%

(2/4)

80.0%

(4/5)

100%

(2/2)

*P< 0.05, Fisher's exact test.

Interestingly, the present results suggested that the pathologic factors possibly reflecting the

efficacy of RPZ differed between NERD and RE groups. That is, DIS (+)/PGP 9.5 ()

indicated resistance to PPI therapy in NERD patients, while the presence of both eosinophil

infiltration and intraepithelial bleeding were predictive of resistance to PPI therapy in RE

patients. RE patients whose biopsy samples had eosinophil infiltration exhibited a poor

response to PPI therapy. Based on the reported finding that the eosinophil count reflects the

severity of gastric acid reflux,[19] it can be inferred that the greater the degree of eosinophil

infiltration, the longer the symptoms have continued, despite treatment with PPI. However,

since only one of the four patients in the non-responders of RE group showed eosinophil

positivity, no statistically firm conclusion can yet be reached. Similarly, patients who hadintraepithelial bleeding also exhibited a poor response to PPI therapy. Intraepithelial bleeding

may be caused by dilation or hyperplasia of intrapapillary blood vessels,[27,28] and since

bleeding is an erythrogenically severe condition clinically, it may prolong the symptoms.

The advantages of the study were that it was a multi-center study, biopsy specimens were

assessed objectively at a completely independent center, and that the site of biopsy was

carefully specified. However, since biopsy specimens were taken from only one site in each

subject at each study center, and assessed on this basis, it was unclear whether the data

obtained were actually representative of each subject. In addition, the number of subjects may

have been too small to allow detection of statistically significant differences among the

variables investigated. Although the sample size reflected the actual epidemiologic ratio ofNERD and RE patients (71 vs 23), it would have been better if the number of RE subjects had


17/19

approximated that of NERD patients. A further study including healthy adults, patients with

Grade C and D reflux esophagitis, and patients with asymptomatic reflux esophagitis would

allow a more comprehensive assessment of the characteristics of biopsy specimens from

GERD patients and their clinical significance.

In conclusion, our study demonstrated the histological assessment of biopsied esophagealtissue is likely to predict therapeutic efficacy of PPI, and the current study suggests factors

predictive of resistance to treatment with PPI are negativity for PGP 9.5 in NERD patients

and intraepithelial bleeding in RE patients.

Appendix

Members of the ARS (Acid-Related Symptom) Research Group

Shuichi Ohara, Department of Gastroenterology, Tohoku Rosai Hospital, Sendai, Japan;

Tomoyuki Koike, Division of Gastroenterology, Tohoku University Graduate School of

Medicine, Sendai, Japan; Motoyasu Kusano and Yasuyuki Shimoyama, Department ofEndoscopy and Endoscopic Surgery, Gunma University Hospital, Maebashi, Japan; Osamu

Kawamura, Department of Medicine and Molecular Science, Gunma University Graduate

School of Medicine, Maebashi, Japan; Yoshio Hoshihara, Clinic of the Ministry of Economy,

Trade, and Industry, Tokyo, Japan; Kouji Yakabi, Department of Gastroenterology and

Hepatology, Saitama Medical Center, Saitama Medical University, Kawagoe, Japan;

Mitsushige Sugimoto and Masafumi Nishino, First Department of Medicine, Hamamatsu

University School of Medicine, Hamamatsu, Japan; Makoto Sasaki, Department of

Gastroenterology and Metabolism, Nagoya City University Graduate School of Medical

Sciences, Nagoya, Japan; Yasuhiro Fujiwara, Department of Gastroenterology, Osaka City

University Graduate School of Medicine, Osaka, Japan; Kazuhide Higuchi, Second

Department of Internal Medicine, Osaka Medical College, Takatsuki, Japan; Ken Haruma,

Division of Gastroenterology, Department of Internal Medicine, Kawasaki Medical School,

Kurashiki, Japan; Noriaki Manabe, Division of Endoscopy and Ultrasonography, Kawasaki

Medical School, Kurashiki, Japan; Kyoichi Adachi, Department of Clinical Nursing, Shimane

University Faculty of Medicine, Izumo, Japan; Kenji Furuta, Department of Internal Medicine

II, Shimane University Faculty of Medicine, Izumo, Japan; Kazuma Fujimoto, Department of

Internal Medicine, Saga Medical School, Saga, Japan.

References

1. Miwa H, Oshima T, Tomita T, Kim Y, Hori K, Matsumoto T. Gastro-esophagealreflux disease: the recent trend in Japan. Clin. J.Gastroenterol. 2008; 1: 1338.2. Fass R, Gasiorowska A. Refractory GERD: what is it? Curr.Gastroenterol. Rep. 2008;

10: 2527.

3. Hershcovici T, Fass R. Management of gastroesophageal reflux disease that does notrespond well to proton pump inhibitors. Curr.Opin. Gastroenterol. 2010; 26: 36778.

4. Miwa H, Sasaki M, Furuta T et al. Efficacy of rabeprazole on heartburn symptomresolution in patients with non-erosive and erosive gastro-oesophageal reflux disease:

a multicenter study from Japan.Aliment. Pharmacol. Ther. 2007; 26: 6977.

5. Peghini PL, Katz PO, Bracy NA, Castell DO. Nocturnal recovery of gastric acidsecretion with twice-daily dosing of proton pump inhibitors.Am. J. Gastroenterol.

1998; 93: 7637.


18/19

6. Furuta T, Shirai N, Sugimoto M, Ohashi K, Ishizaki T. Pharmacogenomics of protonpump inhibitors.Pharmacogenomics 2004; 5: 181202.

7. Frierson HF Jr. Histology in the diagnosis of reflux esophagitis. Gastroenterol. Clin.North Am. 1990; 19: 63144.

8. Riddell RH. The biopsy diagnosis of gastroesophageal reflux disease, "carditis," andBarrett's esophagus, and sequelae of therapy.Am. J.Surg. Pathol. 1996; 20 (Suppl. 1):S3150.

9. Takubo K, Honma N, Aryal G et al. Is there a set of histologic changes that areinvariably reflux-associated?Arch. Pathol. Lab.Med. 2005; 129: 15963.

10.Fiocca R, Mastracci L, Riddell Ret al. Development of consensus guidelines for thehistologic recognition of microscopic esophagitis in patients with gastroesophageal

reflux disease: the Esohisto project.Hum. Pathol. 2010; 41: 22331.

11.Furuta T, Shimatani T, Sugimoto M et al. Investigation of pretreatment prediction ofproton pump inhibitor (PPI)-resistant patients with gastroesophageal reflux disease

and the dose escalation challenge of PPIs-TORNADO study: a multicenter prospective

study by the Acid-related Symptom Research Group in Japan.J.Gastroenterol. 2011;

46: 127383.12.Hoshihara Y. Endoscopic findings of GERD.Nippon Rinsho 2004; 62: 145964.13.Hongo M. Minimal changes in reflux esophagitis: red ones and white ones.J.

Gastroenterol. 2006; 41: 959.

14.Piccinini M, Merighi A, Bruno Ret al. Affinity purification and characterization ofprotein gene product 9.5 (PGP9.5) from retina.Biochem. J. 1996; 318: 71116.

15.Takubo K. Free nerve endings. In: Takubo K, ed.Pathology of theEsophagus, 2ndedn, Tokyo, Berlin, Heidelberg: Springer, 2007; 12.

16.Ismail-Beigi F, Pope CE 2nd. Distribution of the histological changes ofgastroesophageal reflux in the distal esophagus of man. Gastroenterology 1974; 66:

110913.

17.Haber MM, Lu L, Modi A, Garcia FU. Use of MIB-1 in the assessment of esophagealbiopsy specimens from patients with gastroesophageal reflux disease in well- and

poorly oriented areas.Appl. Immunohistochem. Mol. Morphol. 2002; 10: 12833.

18.Vieth M. Structural abnormalities of endoscopy-negative reflux diseasereal orperceived?Digestion 2008; 78 (Suppl. 1): 2430.

19.Haggitt RC. Histopathology of reflux-induced esophageal and supraesophagealinjuries.Am. J. Med. 2000; 108 (Suppl. 4a): 109S11S.

20.Miwa H, Minoo T, Hojo M et al. Oesophageal hypersensitivity in Japanese patientswith non-erosive gastro-oesophageal reflux diseases.Aliment. Pharmacol. Ther. 2004;

20 (Suppl. 1): 11217.

21.Binato M, Gurski RR, Fagundes RB, Meurer L, Edelweiss MI. P53 and Ki-67overexpression in gastroesophageal reflux diseaseBarrett's esophagus and

adenocarcinoma sequence.Dis.Esophagus 2009; 22: 58895.

22.Mastracci L, Grillo F, Zentilin P et al. Cell proliferation of squamous epithelium ingastro-oesophageal reflux disease: correlations with clinical, endoscopic and

morphological data.Aliment. Pharmacol. Ther. 2007; 25: 63745.

23.Bhat YM, Bielefeldt K. Capsaicin receptor (TRPV1) and non-erosive reflux disease.Eur. J. Gastroenterol. Hepatol. 2006; 18: 26370.

24.van Malenstein H, Farr R, Sifrim D. Esophageal dilated intercellular spaces (DIS)and nonerosive reflux disease.Am. J.Gastroenterol. 2008; 103: 10218.

25.Orlando LA, Orlando RC. Dilated intercellular spaces as a marker of GERD. Curr.Gastroenterol. Rep. 2009; 11: 1904.


19/19

26.Solcia E, Villani L, Luinetti O et al. Altered intercellular glycoconjugates and dilatedintercellular spaces of esophageal epithelium in reflux disease. Virchows Arch. 2000;

436: 20716.

27.Geboes K, Desmet V, Vantrappen G, Mebis J. Vascular changes in the esophagealmucosa. An early histologic sign of esophagitis. Gastrointest. Endosc. 1980; 26: 29

32.28.Kobayashi S, Kasugai T. Endoscopic and biopsy criteria for the diagnosis of

esophagitis with a fiberoptic esophagoscope.Am. J.Dig. Dis. 1974; 19: 34552.

Financial support

Unrestricted donation was offered by Eisai Co., Ltd. for this study.

J Gastroenterol Hepatol. 2013;28(3):479-487. 2013 Blackwell Publishing

meds capea

Documents