Journal of the Formosan Medical Association (2017) 116, 4e9

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.jfma-online.com

ORIGINAL ARTICLE

Antigastric parietal cell and antithyroid

autoantibodies in patients with recurrent
aphthous stomatitis
Yang-Che Wu a,b, Yu-Hsueh Wu a,b, Yi-Ping Wang a,b,c,
Julia Yu-Fong Chang a,b,c, Hsin-Ming Chen a,b,c, Andy Sun a,b,*

a
Graduate Institute of Clinical Dentistry, School of Dentistry, National Taiwan University,
Taipei, Taiwan
b
Department of Dentistry, National Taiwan University Hospital, College of Medicine, National Taiwan
University, Taipei, Taiwan
c
Graduate Institute of Oral Biology, School of Dentistry, National Taiwan University, Taipei, Taiwan

Received 22 September 2016; accepted 30 September 2016

KEYWORDS Background/Purpose: Anti-gastric parietal cell antibody (GPCA), anti-thyroglobulin antibody

antigastric parietal (TGA), and anti-thyroid microsomal antibody (TMA) have not yet been reported in patients
cell antibody; with recurrent aphthous stomatitis (RAS). This study mainly assessed the frequencies of the
antithyroglobulin presence of serum GPCA, TGA, and TMA in different types of RAS patients.
antibody; Methods: Serum GPCA, TGA, and TMA levels were measured in 355 RAS patients of different
antithyroid subtypes and in 355 age- and sex-matched healthy control individuals.
microsomal Results: We found that 13.0%, 19.4%, and 19.7% of 355 RAS patients, 16.7%, 23.3%, and 21.7% of
antibody; 60 major-typed RAS patients, 12.2%, 18.6%, and 19.3% of 295 minor-typed RAS patients, 18.1%,
atrophic glossitis; 20.0%, and 21.9% of 160 atrophic glossitis-positive RAS (AGþ/RAS) patients, and 8.7%, 19.0%,
recurrent aphthous and 17.9% of 195 AG-negative RAS (AGe/RAS) patients had the presence of GPCA, TGA, and
stomatitis TMA in their sera, respectively. RAS, major-typed RAS, minor-typed RAS, AGþ/RAS, and AG
e/RAS patients all had a significantly higher frequency of GPCA, TGA, or TMA positivity than
healthy control individuals (all p < 0.001). Of 65 TGA/TMA-positive RAS patients whose serum
thyroid-stimulating hormone (TSH) levels were measured, 76.9%, 12.3%, and 10.8% of these
TGA/TMA-positive RAS patients had normal, lower, and higher serum TSH levels, respectively.
Conclusion: We conclude that approximately one-third RAS patients may have GPCA/TGA/TMA
positivity in their sera. Because some GPCA-positive patients may develop pernicious anemia,

Conflicts of interest: The authors have no conflicts of interest relevant to this article.
* Corresponding author. Department of Dentistry, National Taiwan University Hospital, Number 1, Chang-Te Street, Taipei 10048, Taiwan.
E-mail address: andysun7702@yahoo.com.tw (A. Sun).

http://dx.doi.org/10.1016/j.jfma.2016.09.008
0929-6646/Copyright ª 2016, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Autoantibodies in recurrent aphthous stomatitis
autoimmune atrophic gastritis, and gastric carcinoma, and some TGA/TMA-positive patients
may have thyroid dysfunction such as hyperthyroidism and hypothyroidism, these patients
should be referred to doctors for further management.
Copyright ª 2016, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Introduction
Recurrent aphthous stomatitis (RAS) is a common oral
mucosal disease characterized by recurrent and painful
ulcerations on the nonkeratinized oral mucosa. The prev-
alence of RAS varies from 5% to 20% depending on the
population evaluated.1 In Taiwan, the prevalence of RAS is
10.5% in the general population.2
Although several etiological factors have been pro-
posed, the exact cause underlying RAS remains unclear.3
The study results on tissue infiltrated mononuclear
cells favor the role of cell-mediated cytotoxicity in the
immunopathogenesis of RAS.4 In addition to immune
dysregulation, multiple nutritional deficiencies including
deficiencies of vitamins B1, B2, B6, and B12, folate, iron,
and ferritin are found to be the possible etiologies of RAS.4
Although several different types of antibody or autoanti-
body have already been reported in RAS patients,5e9 the
organ-specific autoantibodies such as anti-gastric parietal
cell antibody (GPCA), anti-thyroglobulin antibody (TGA),
and anti-thyroid microsomal antibody (TMA) have not yet
been reported in RAS patients.
In our oral mucosal disease clinic, patients with atro-
phic glossitis (AG), burning mouth syndrome (BMS), oral
lichen planus (OLP), RAS, oral submucous fibrosis (OSF),
and other oral mucosal diseases are frequently encoun-
tered.4,10e24 For AG, BMS, OLP, RAS, and OSF patients,
complete blood count, serum iron, vitamin B12, folic acid,
homocysteine, GPCA, TGA, and TMA levels are frequently
examined to assess whether these patients have anemia,
hematinic deficiencies, and serum GPCA, TGA, or TMA
positivity.4,10e21,25 The serum GPCA, TGA, and TMA levels
were evaluated because patients with GPCA are more
likely to have pernicious anemia and to develop autoim-
mune atrophic gastritis, which may subsequently progress
to gastric carcinoma,26,27 and patients with TGA or TMA
may develop autoimmune thyroid disease and finally
result in thyroid dysfunction.10,28 For early diagnosis and
treatment of subsequent diseases, it is very important to
evaluate whether RAS patients have GPCA, TGA, and TMA
in their sera.
In this study, the serum autoantibodies including GPCA,
TGA, and TMA were measured in 355 RAS patients and 100
healthy control individuals. The purposes of this study were
to assess whether a certain percentage of RAS patients
might have GPCA, TGA, and TMA in their sera; to evaluate
whether RAS patients might have a significantly higher
frequency of GPCA, TGA, or TMA positivity than healthy
control individuals; and to find whether TGA-positive and/
or TMA-positive (TGA/TMA-positive) patients might have
thyroid dysfunction.
5
Methods
Participants
The study group consisted of 355 RAS patients (106 men and
249 women; age range, 18e90 years; mean age,
52.8  15.9 years). The normal control group consisted of
355 age- ( 2 years of each patient’s age) and sex-matched
healthy control individuals (106 men and 249 women; age
range, 20e89 years; mean age, 53.1  14.7 years). All RAS
patients and control individuals were seen consecutively,
diagnosed, and treated in the Department of Dentistry,
National Taiwan University Hospital (NTUH; Taipei, Taiwan)
from July 2007 to July 2016. Patients were diagnosed as
having RAS when they had at least one episode of oral ul-
cerations on movable oral mucosa per month since child-
hood.4 RAS were further divided into major-typed RAS
(n Z 60) when patients had recurrent oral ulcerations with
a diameter more than 1 cm and minor-typed RAS (n Z 295)
when patients had recurrent oral ulcerations with a diam-
eter less than 1 cm.4 In this study, 160 RAS patients had
concomitant partial or complete atrophic glossitis (AG),
which was defined as partial or complete absence or flat-
tening of filiform papillae on the dorsal surface of the
tongue, respectively.12 Thus, RAS patients could be further
divided into AG-positive RAS (AGþ/RAS) patients (n Z 160)
and AG-negative RAS (AGe/RAS) patients (n Z 195). RAS
patients with betel quid chewing habit or autoimmune
diseases (such as systemic lupus erythematosus, rheuma-
toid arthritis, Sjogren’s syndrome, pemphigus vulgaris, and
cicatricial pemphigoid) were excluded. Moreover, patients
with traumatic ulcers or with aphthous-like ulcers associ-
ated with systemic disorders including Behcet’s syndrome,
celiac disease, gluten-sensitive enteropathy, inflammatory
bowel diseases, human immunodeficiency virus infection,
and cyclic neutropenia were also excluded.29 In addition,
RAS patients with serum creatinine concentrations indica-
tive of renal dysfunction (i.e., men, >131 mmol/L; women,
>115 mmol/L), and who reported a history of stroke, heavy
alcohol use, or diseases of the liver, kidney, or coronary
arteries were also excluded.30 Healthy control individuals
had either dental caries, pulpal disease, malocclusion, or
missing of teeth but did not have any oral mucosal or sys-
temic diseases. None of the RAS patients had taken any
prescription medication for RAU or AG at least 3 months
prior to entering the study.
The blood samples were drawn from all RAS patients
and healthy control volunteers for measurement of
serum GPCA, TGA, and TMA levels. All RAS patients and
healthy control individuals signed the informed consent
6 Y.-C. Wu et al.

form prior to entering the study. This study was reviewed
and approved by the Institutional Review Board at the
NTUH.
Determination of serum anti-GPCA level
The serum GPCA level was detected using the indirect
immunofluorescence technique with rat stomach as a sub-
strate as described previously.12,15e17 Sera were scored as
positive when they produced fluorescence at a dilution of
10-fold or more.
Determination of serum anti-TGA or anti-TMA level
TGA and TMA titers were measured by a semiquantitative
microtiter particle agglutination test using Serodia-AMC
kits (Fujirebio Inc., Tokyo, Japan) as described previ-
ously.10,11 TGA or TMA titers  1:40 were considered
positive.
Determination of thyroid-stimulating hormone
concentrations
The blood thyroid-stimulating hormone (TSH) concentra-
tions of the TGA/TMA-positive RAS patients and healthy
control individuals were determined by the routine tests
performed in the Department of Laboratory Medicine of
NTUH.10,11
Statistical analysis
Comparisons of frequencies of presence of serum autoan-
tibodies such as GPCA, TGA, and TMA between 355 RAS
patients, 60 major-typed RAS patients, 295 minor-typed
RAS patients, 160 AGþ/RAS patients, or 195 AGe/RAS pa-
tients and 100 healthy control individuals, between 60
major-typed RAS patients and 295 minor-typed RAS pa-
tients, and between 160 AGþ/RAS patients and 195 AG/
RAS patients were performed using the Chi-square test.
Comparison of frequency of patients with different serum
TSH levels between 65 TGA/TMA-positive patients and 100
healthy control individuals was also done using the Chi-
square test. The result was considered to be significant if
the p value was < 0.05.
Results
Comparisons of frequencies of the presence of serum au-
toantibodies such as GPCA, TGA, and TMA between 355 RAS
patients or patients in each of four different RAS subgroups
and 100 healthy control individuals, between 60 major-
typed RAS patients and 295 minor-typed RAS patients, and
between 160 AGþ/RAS patients and 195 AG/RAS patients
are shown in Table 1. We found that 13.0%, 19.4%, and
19.7% of 355 RAS patients, 16.7%, 23.3%, and 21.7% of 60
major-typed RAS patients, 12.2%, 18.6%, and 19.3% of 295
minor-typed RAS patients, 18.1%, 20.0%, and 21.9% of 160

Table 1 The patient number and frequencies of presence of serum autoantibodies such as anti-gastric parietal cell antibody
(GPCA), anti-thyroglobulin antibody (TGA), or anti-thyroid microsomal antibody (TMA) in 355 recurrent aphthous stomatitis
(RAS), 60 major-typed RAS (major RAS), 295 minor-typed RAS (minor RAS), 160 atrophic glossitis-positive RAS (AGþ/RAS), and
195 AG-negative RAS (AG/RAS) patients and in 355 age- and sex-matched healthy control individuals.
Autoantibodies
Positive patient number (%)
GPCA TGA TMA
RAS (n Z 355) 46 (13.0) 69 (19.4) 70 (19.7)
pa <0.001 <0.001 <0.001
Major RAS (n Z 60) 10 (16.7) 14 (23.3) 13 (21.7)
pa <0.001 <0.001 <0.001
pb 0.467 0.511 0.812
Minor RAS (n Z 295) 36 (12.2) 55 (18.6) 57 (19.3)
pa <0.001 <0.001 <0.001
AGþ/RAS (n Z 160) 29 (18.1) 32 (20.0) 35 (21.9)
pa <0.001 <0.001 <0.001
pc 0.014 0.914 0.429
AGe/RAS (n Z 195) 17 (8.7) 37 (19.0) 35 (17.9)
pa <0.001 <0.001 <0.001
Healthy control 7 (2.0) 8 (2.3) 8 (2.3)
individuals (n Z 355)
a
Comparisons of different autoantibody frequencies between each of different groups of RAS patients and healthy control individuals
by the Chi-square test.
b
Comparisons of different autoantibody frequencies between 60 major RAS patients and 295 minor RAS patients by the Chi-square
test.
c
Comparisons of different autoantibody frequencies between 160 AGþ/RAS patients and 195 AGe/RAS patients by the Chi-square
test.
Autoantibodies in recurrent aphthous stomatitis 7

AGþ/RAS patients, and 8.7%, 19.0%, and 17.9% of 195 AG/
RAS patients had the presence of GPCA, TGA, and TMA in
their sera, respectively. RAS, major-typed RAS, minor-
typed RAS, AGþ/RAS, and AGe/RAS patients all had a
significantly higher frequency of GPCA, TGA, or TMA posi-
tivity than healthy control individuals (all p < 0.001; Table
1). Moreover, 160 AGþ/RAS patients had a significantly
higher frequency of GPCA positivity than 195 AGe/RAS
patients (p Z 0.014). However, there were no significant
differences in the frequencies of GPCA, TGA, and TMA
positivities between major-typed RAS and minor-typed RAS
patients as well as in the frequencies of TGA and TMA
positivities between AGþ/RAS and AGe/RAS patients
(Table 1).
We also found that some RAS patients might have the
presence of one, two, or three organ-specific autoanti-
bodies such as GPCA, TGA, and TMA in their sera. The pa-
tient number and frequencies of presence of one, two, or
three organ-specific autoantibodies such as GPCA, TGA, and
TMA in 355 RAS patients and patients in four different RAS
subgroups are shown in Table 2. We found that 14 (3.9%)
RAS patients, 42 (11.8%) RAS patients, and 59 (16.6%) RAS
patients, five (8.3%) major-typed RAS patients, seven
(11.7%) major-typed RAS patients, and eight (13.3%) major-
typed RAS patients, nine (3.1%) minor-typed RAS patients,
35 (11.9%) minor-typed RAS patients, and 51 (17.3%) minor-
typed RAS patients, eight (5.0%) AGþ/RAS patients, 20
(12.5%) AGþ/RAS patients, and 31 (19.4%) AGþ/RAS
patients, and six (3.1%) AG/RAS patients, 22 (11.3%) AG/
RAS patients, and 28 (14.4%) AG/RAS patients had the
presence of three (GPCA þ TGA þ TMA), two (GPCA þ TGA,
GPCA þ TMA, or TGA þ TMA), or one (GPCA only, TGA only,
or TMA only) organ-specific autoantibody in their sera,
respectively (Table 2).
In this study, the serum TSH levels were measured in 65 out
of 84 TGA/TMA-positive RAS patients and in 100 healthy
control individuals (Table 3). All healthy control individuals
and 76.9% of 65 TGA/TMA-positive RAS patients had their
serum TSH levels within the normal range (0.4e4.0 mIU/mL).
However, 12.3% and 10.8% of 65 TGA/TMA-positive RAS pa-
tients had a lower serum TSH level (< 0.4 mIU/mL) and a
higher serum TSH level (> 4.0 mIU/mL), respectively (Table 3).
Discussion
The serum organ-specific autoantibodies such as GPCA,
TGA, and TMA have not yet bone reported in RAS patients.
This study found GPCA in 13.0%, TGA in 19.4%, and TMA in
19.7% of 355 RAS patients. Although the positive rates of
GPCA, TGA, and TMA were slightly higher in major-typed
RAS patients than in minor-typed RAS patients, the differ-
ences were not significant. However, the positive rate of
GPCA was significantly higher in AGþ/RAS patients (18.1%)
than in AGe/RAS patients (8.7%). This finding suggests that
RAS patients with concomitant AG do have a relatively high

Table 2 The patient number and frequencies of presence of one, two, or three organ-specific autoantibodies such as anti-
gastric parietal cell antibody (GPCA), anti-thyroglobulin antibody (TGA), and anti-thyroid microsomal antibody (TMA) in 355
recurrent aphthous stomatitis (RAS), 60 major-typed RAS (major RAS), 295 minor-typed RAS (minor RAS), 160 atrophic glossitis-
positive RAS (AGþ/RAS), and 195 AG-negative RAS (AGe/RAS) patients.
Autoantibodies Patient number (%)
RAS Major RAS Minor RAS AGþ/RAS AGe/RAS
(n Z 355) (n Z 60) (n Z 295) (n Z 160) (n Z 195)
GPCA þ TGA þ TMA 14 (3.9) 5 (8.3) 9 (3.1) 8 (5.0) 6 (3.1)
GPCA þ TGA 1 (0.3) 0 (0.0) 1 (0.3) 0 (0.0) 1 (0.5)
GPCA þ TMA 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
TGA þ TMA 41 (11.5) 7 (11.7) 34 (11.5) 20 (12.5) 21 (10.8)
GPCA only 31 (8.7) 5 (8.3) 26 (8.8) 20 (12.5) 11 (5.6)
TGA only 13 (3.7) 2 (3.3) 11 (3.7) 4 (2.5) 9 (4.6)
TMA only 15 (4.2) 1 (1.7) 14 (4.7) 7 (4.4) 8 (4.1)
None 240 (67.6) 40 (66.7) 200 (67.8) 101 (63.1) 139 (71.3)

Table 3 Number and percentage of individuals with different serum levels of thyroid-stimulating hormone (TSH) in 65 anti-
thyroglobulin antibody (TGA)-positive and/or anti-thyroid microsomal antibody (TMA)-positive (TGA/TMA-positive) recurrent
aphthous stomatitis (RAS) patients and in 100 healthy control individuals.
TSH level TGA/TMA-positive RAS patients Healthy control individuals p
(n Z 65) (n Z 100) Chi-square test
<0.4 mIU/mL 8 (12.3) 0 (0) 0.001*
0.4e4.0 mIU/mL 50 (76.9) 100 (100) <0.001*
>4.0 mIU/mL 7 (10.8) 0 (0) 0.003*
Data are presented as n (%).
* Comparison of frequency of patients with different levels of TSH between 65 TGA/TMA-positive RAS patients and 100 healthy control
individuals by the Chi-square test with p < 0.05.
8 Y.-C. Wu et al.
frequency of the presence of GPCA in their sera. Our pre-
vious study showed serum GPCA positivity in 47 (26.7%) of
176 AG patients.12 The GPCA-positive rate was greater in
AG patients than in AGþ/RAS patients (p Z 0.081, marginal
significance).
It is still not clear why some of RAS patients have GPCA,
TGA, and TMA in their sera. Our previous studies showed a
strong association of HLA-DRw9 with RAS in Chinese patients
and a strong association of antiepithelial cell antibodies with
HLA-DR3 or DR7 phenotype in RAS patients.31,32 The HLA DR9
is known to be associated with increased prevalence of
autoimmune diseases such as Graves’ disease, myasthenia
gravis, insulin-dependent diabetes, and Hashimoto’s
thyroiditis in Southern Chinese.33e35 Moreover, HLA-DR3 has
been reported to be related to systemic lupus erythemato-
sus, Graves’ disease, myasthenia gravis, insulin-dependent
diabetes, and erosive OLP in Caucasians.35,36 Thus, the
intimate association of HLA-DRw9 with RAS in Chinese pa-
tients and of HLA-DR3 or DR7 with production of anti-
epithelial cell antibodies in RAS patients may partially
explain why a small percentage of RAS patients may have
GPCA, TGA, and TMA in their sera. Further studies are
needed to elucidate the mechanisms that may induce part of
RAS patients to generate GPCA, TGA, and TMA in their sera.
Different types of antibodies or autoantibodies other
than GPCA, TGA, and TMA have been reported in RAS
patients.5e9 Antiendomysial (or antitransglutaminase) IgA
and IgG antibodies were found in two RAS patients with
concomitant celiac disease.5 Of 87 minor-typed RAS pa-
tients, antireticulin IgG, antireticulin IgA, and anti-
endomysial IgA antibodies were reported in three (3.4%)
patients, one (1.1%) patient, and one (1.1%) patient,
respectively.6 Healy et al7 studied the antiendothelial cell
autoantibody and antineutrophil cytoplasmic autoantibody
levels in 20 RAS patients and 20 control individuals. IgG
antiendothelial cell autoantibody was detected in 19 (95%)
of 20 RAS patients and four control individuals. However,
antineutrophil cytoplasmic autoantibody was detected in
only one RAS patient and none of the control individuals.
Sun and Wu8 found the antimucosal antibody in 15 (71%) of
21 RAS patients and anti-intercellular substance antibody in
14 (67%) of 21 RAS patients. The positive rate of anti-
intercellular substance antibody increased to 67% in the
active phase but decreased to 25% in the 1-week remission
phase and to 10% in the 2-week remission phase of RAS.8
The serum anti-nuclear antibody (ANA) was detected in
six (12%) of 50 RAS patients and in three (5%) of 57 healthy
control individuals; no significant difference in the inci-
dence of serum ANA positivity was found between RAS pa-
tients and healthy control individuals.9 The above-
mentioned findings indicate the low frequencies of pres-
ence of antiendomysial and antireticulin antibodies in RAS
patients.5,6 The high positive rate of antiendothelial cell
autoantibody in RAS patients suggests that vasculitis may
play a role in etiology of RAS.7 Moreover, the raise in the
frequency of anti-intercellular substance antibody in sera
of active RAS patients is only a transient phenomenon.8
The two major criteria for the diagnosis of chronic
autoimmune thyroiditis are high TSH concentration and the
presence of TGA/TMA in patients’ sera.28 This study
measured the serum TSH levels in 65 TGA/TMA-positive RAS
patients; we found that 50 (76.9%) RAS patients, eight
(12.3%) RAS patients, and seven (10.8%) RAS patients had
normal (suggestive of euthyroid), lower (suggestive of hy-
perthyroidism), and higher serum TSH levels (suggestive of
hypothyroidism), respectively. Our previous study quantified
the serum TSH levels in 190 TGA/TMA-positive patients
(including 83 AG patients and 107 BMS patients) and found
that 163 (85.8%) TGA/TMA-positive patients, eight (4.2%)
TGA/TMA-positive patients, and 19 (10.0%) TGA/TMA-
positive patients had normal, lower, and higher serum TSH
levels, respectively.10 Previous community survey studies
discovered that 50e75% of individuals with TGA/TMA posi-
tivity are euthyroid, 25e50% have subclinical hypothyroid-
ism, and only 5e10% have overt hypothyroidism.28 The
results of the above-mentioned studies are comparable and
indicate that the majority of the TGA/TMA-positive pa-
tients actually have euthyroid, and only a small portion of
TGA/TMA-positive patients have overt hypothyroidism.10,28
It is interesting to know what percentage of GPCA-
positive oral mucosal disease patients may have pernicious
anemia. Our previous study showed that 12.9% of 124 GPCA-
positive oral mucosal disease patients (including 75 AG
patients and 49 BMS patients) have pernicious anemia.17 In
our oral mucosal disease clinic, we also found that 14.1% of
92 GPCA-positive erosive OLP patients with desquamative
gingivitis have pernicious anemia15 and 8.7% of 46 GPCA-
positive RAS patients had pernicious anemia (unpublished
data). These findings indicate that approximately 9e14% of
GPCA-positive oral mucosal disease patients do have per-
nicious anemia.15,17
In conclusion, we found that 13.0%, 19.4%, and 19.7% of
355 RAS patients had the presence of GPCA, TGA, and TMA
in their sera, respectively. Moreover, 3.9%, 11.8%, or 16.6%
RAS patients had the presence of three, two, or one organ-
specific autoantibody in their sera, respectively. Of 65
TGA/TMA-positive RAS patients whose serum TSH levels
were measured, 12.3% and 10.8% RAS patients had lower
(suggestive of hyperthyroidism) and higher (suggestive of
hypothyroidism) serum TSH levels, respectively. As stated
before, without proper early diagnosis and treatment,
GPCA-positive patients are more likely to have pernicious
anemia and to develop autoimmune atrophic gastritis,
which may subsequently progress to gastric carcinoma,26,27
and TGA/TMA-positive patients may develop autoimmune
thyroid disease and finally result in thyroid dysfunction.10,28
Those TGA/TMA-positive RAS patients with either hyper-
thyroidism or hypothyroidism should be referred to the
endocrinology department for further treatment. More-
over, those GPCA-positive patients should be referred to
the department of gastroenterology for endoscopic exam-
ination of the stomach to check for the presence of auto-
immune atrophic gastritis that can be further treated by
doctors in that department. In addition, a long-term follow-
up study is warranted to assess whether GPCA-positive RAS
patients with or without treatment may develop gastric
carcinoma.
Acknowledgments
This study was supported by the grant of Ministry of Science
and Technology, ROC (No. 102-2314-B-002-125-MY3 and No.
105-2314-B-002-075-MY2).
Autoantibodies in recurrent aphthous stomatitis 9

References 18. Sun A, Chang JYF, Wang YP, Cheng SJ, Chen HM, Chiang CP. Do
all the patients with vitamin B12 deficiency have pernicious
anemia? J Oral Pathol Med 2016;45:23e7.
1. Lopez-Jornet P, Camacho-Alonso F, Martos N. Hematological
19. Chang JYF, Wang YP, Wu YC, Cheng SJ, Chen HM, Sun A. He-
study of patients with aphthous stomatitis. Int J Dermatol
matinic deficiencies and pernicious anemia in oral mucosal
2014;53:159e63.
disease patients with macrocytosis. J Formos Med Assoc 2015;
2. Chiang CP, Chueh LH, Lin SK, Chen MY. Oral manifestations in
114:736e41.
human immunodeficiency virus-infected patients in Taiwan. J
20. Chang JYF, Wang YP, Wu YC, Cheng SJ, Chen HM, Sun A. He-
Formos Med Assoc 1998;97:600e5.
matinic deficiencies and anemia statuses in oral mucosal dis-
3. Scully C, Gorsky M, Lozada-Nur F. The diagnosis and manage-
ease patients with folic acid deficiency. J Formos Med Assoc
ment of recurrent aphthous stomatitis: a consensus approach.
2015;114:806e12.
J Am Dent Assoc 2003;134:200e7.
21. Chang JYF, Wang YP, Wu YC, Cheng SJ, Chen HM, Sun A. Blood
4. Sun A, Chen HM, Cheng SJ, Wang YP, Chang JYF, Wu YC, et al.
profile of oral mucosal disease patients with both vitamin B12
Significant association of deficiency of hemoglobin, iron,
and iron deficiencies. J Formos Med Assoc 2015;114:532e8.
vitamin B12, and folic acid and high homocysteine level with
22. Lee JJ, Yang FY, Wu YC, Chen HM. Metastatic lung carcinoma to
recurrent aphthous stomatitis. J Oral Pathol Med 2015;44:
the lower anterior gingiva. J Formos Med Assoc 2014;113:
300e5.
978e80.
5. Olszewska M, Sulej J, Kotowski B. Frequency and prognostic
23. Lin HP, Liu CJ, Chiang CP. Spindle cell lipoma of the tongue. J
value of IgA and IgG endomysial antibodies in recurrent aph-
Formos Med Assoc 2015;114:477e9.
thous stomatitis. Acta Derm Venereol 2006;86:332e4.
24. Shen WR, Chang JYF, Wu YC, Cheng SJ, Chen HM, Wang YP. Oral
6. Robinson NA, Porter SR. Low frequency of anti-endomysial
traumatic ulcerative granuloma with stromal eosinophilia: a
antibodies in recurrent aphthous stomatitis. Ann Acad Med
clinicopatholoical study of 34 cases. J Formos Med Assoc 2015;
Singapore 2004;33:43e7.
114:881e5.
7. Healy CM, Carvalho D, Pearson JD, Thornhill MH. Raised anti-
25. Sun A, Chang JYF, Chiang CP. Blood examination is necessary
endothelial cell autoantibodies (AECA), but not anti-
for oral mucosal disease patients. J Formos Med Assoc 2016;
neutrophil cytoplasmic autoantibodies (ANCA), in recurrent
115:1e2.
oral ulceration: modulation of AECA binding by tumour necrosis
26. Lo CC, Hsu PI, Lo GH, et al. Implications of anti-parietal cell
factor-alpha (TNF-alpha) and interferon-gamma (IFN-gamma).
antibodies and anti-Helicobacter pylori antibodies in histo-
Clin Exp Immunol 1996;106:523e8.
logical gastritis and patient outcome. World J Gastroenterol
8. Sun A, Wu YC. Anti-mucosal antibodies in recurrent aphthous
2005;11:4715e20.
ulcers. J Formos Med Assoc 1989;88:122e7.
27. Lee HW, Hahm KB, Lee JS, Ju YS, Lee KM, Lee KW. Association
9. Sun A, Chiang CP, Chiou PS, Wang JT, Liu BY, Wu YC. Immu-
of the human leukocyte antigen class II alleles with chronic
nomodulation by levamisole in patients with recurrent aph-
atrophic gastritis and gastric carcinoma in Koreans. J Dig Dis
thous ulcers or oral lichen planus. J Oral Pathol Med 1994;23:
2009;10:265e71.
172e7.
28. Dayan CM, Daniels GH. Chronic autoimmune thyroiditis. N Engl
10. Wang YP, Lin HP, Chen HM, Kuo YS, Lang MJ, Sun A. Hemo-
J Med 1996;335:99e107.
globin, iron, and vitamin B12 deficiencies and high blood ho-
29. Scully C. Aphthous ulceration. N Engl J Med 2006;355:
mocysteine levels in patients with anti-thyroid autoantibodies.
165e72.
J Formos Med Assoc 2014;113:155e60.
30. Morris MS, Jacques PF, Rosenberg IH, Selhub J. Folate and
11. Chang JYF, Chiang CP, Wang YP, Wu YC, Chen HM, Sun A. Anti-
vitamin B-12 status in relation to anemia, macrocytosis, and
gastric parietal cell and anti-thyroid autoantibodies in patients
cognitive impairment in older Americans in the age of folic acid
with desquamative gingivitis. J Oral Pathol Med 2016 Sep 7.
fortification. Am J Clin Nutr 2007;85:193e200.
http://dx.doi.org/10.1111/jop.12490 [Epub ahead of print].
31. Sun A, Hsieh RP, Chu CT, Wu YC. Strong association of HLA-
12. Sun A, Lin HP, Wang YP, Chiang CP. Significant association of
DRw9 in Chinese with recurrent oral ulcers. J Am Acad Der-
deficiency of hemoglobin, iron and vitamin B12, high homo-
matol 1991;24:195e8.
cysteine level, and gastric parietal cell antibody positivity with
32. Sun A, Hsieh RP, Liu BY, Wang JT, Leu JS, Wu YC, et al. Strong
atrophic glossitis. J Oral Pathol Med 2012;41:500e4.
association of antiepithelial cell antibodies with HLA-DR3 or
13. Lin HP, Wang YP, Chen HM, Kuo YS, Lang MJ, Sun A. Significant
DR7 phenotype in patients with recurrent oral ulcers. J Formos
association of hematinic deficiencies and high blood homo-
Med Assoc 2000;99:290e4.
cysteine levels with burning mouth syndrome. J Formos Med
33. Hawkins BR, Ma JT, Lam KS, Wang CC, Yeung RT. Association of
Assoc 2013;112:319e25.
HLA antigens with thyrotoxic Graves’ disease and periodic
14. Chen HM, Wang YP, Chang JYF, Wu YC, Cheng SJ, Sun A. Sig-
paralysis in Hong Kong Chinese. Clin Endocrinol (Oxf) 1985;23:
nificant association of deficiency of hemoglobin, iron and
245e52.
vitamin B12 and high homocysteine level with oral lichen pla-
34. Hawkins BR, Ip MS, Lam KS, Ma JT, Wy CL, Yeung RT, et al. HLA
nus. J Formos Med Assoc 2015;114:124e9.
antigens and acetylcholine receptor antibody in the subclas-
15. Chang JYF, Wang YP, Wu YC, Wu YH, Tseng CH, Sun A. Hema-
sification of myasthenia gravis in Hong Kong Chinese. J Neurol
tinic deficiencies and anemia statuses in anti-gastric parietal
Neurosurg Psychiatry 1986;49:316e9.
cell antibody-positive erosive oral lichen planus patients with
35. Hawkins BR, Lam KS, Ma JT, Wang C, Yeung RT. Strong as-
desqamative gingivitis. J Formos Med Assoc 2016;115:860e6.
sociation between HLA DRw9 and Hashimoto’s thyroiditis
16. Wang YP, Wu YC, Cheng SJ, Chen HM, Sun A, Chang JYF. High
in southern Chinese. Acta Endocrinol (Copenh) 1987;114:
frequencies of vitamin B12 and folic acid deficiencies and
543e6.
gastric parietal cell antibody positivity in oral submucous
36. Jontell M, Ståhlblad PA, Rosdahl I, Lindblom B. HLA-DR3
fibrosis patients. J Formos Med Assoc 2015;114:813e9.
antigens in erosive oral lichen planus, cutaneous lichen
17. Sun A, Wang YP, Lin HP, Jia JS, Chiang CP. Do all the patients
planus, and lichenoid reactions. Acta Odontol Scand 1987;
with gastric parietal cell antibodies have pernicious anemia?
45:309e12.
Oral Dis 2013;19:381e6.