Cancer Letters 247 (2007) 243–252

www.elsevier.com/locate/canlet

p53 protein accumulation, cancer multiplicity, and

aldehyde dehydrogenase-2 genotype in Japanese alcoholic
men with early esophageal squamous cell carcinoma
Akira Yokoyama a,*, Tai Omori b, Yoichi Tanaka c, Tetsuji Yokoyama d,
Hitoshi Sugiura e, Takeshi Mizukami a, Sachio Matsushita a, Susumu Higuchi a,
Katsuya Maruyama a, Hiromasa Ishii f, Toshifumi Hibi f
a
National Hospital Organization Kurihama Alcoholism Center, 5-3-1 Nobi, Yokosuka, Kanagawa 239-0841, Japan
b
Departments of Gastroenterology and Surgery, Kawasaki Municipal Hospital, Kanagawa 210-0013, Japan
c
Department of Pathology, Tokyo Dental College, Ichikawa General Hospital, Chiba 272-0824, Japan
d
Department of Technology Assessment and Biostatistics, National Institute of Public Health, Saitama 351-0197, Japan
e
Department of Pathology, Kawasaki Municipal Hospital, Kanagawa 210-0013, Japan
f
Department of Internal Medicine, School of Medicine, Keio University, Tokyo 160-8582, Japan

Received 22 February 2006; received in revised form 2 May 2006; accepted 2 May 2006

Abstract

Synchronous multiple intra-esophageal squamous cell carcinomas (SCCs) or oropharyngolaryngeal SCCs are common
in alcoholics with esophageal SCC, and more frequently found in those with inactive heterozygous aldehyde dehydroge-
nase-2 (ALDH2). p53 alterations have been suspected as key molecular events in such multifocal esophageal carcinogen-
esis. We studied 95 Japanese alcoholic men with Tis and mucosal invasive esophageal SCC and found very high levels of
p53 protein accumulation occurring in early esophageal SCC. Synchronous cancer multiplicity in the upper aerodigestive
tract was found in 40 patients. p53 expression was not correlated with either cancer multiplicity or ALDH2 genotype. The
risk for cancer multiplicity was associated with inactive heterozygous ALDH2 alone (OR = 4.22) among the risk factors
investigated, which also included smoking, less-active alcohol dehydrogenase-1B, and macrocytosis, enhancing the validity
of the link between acetaldehyde exposure and cancer multiplicity.
 2006 Elsevier Ireland Ltd. All rights reserved.

Keywords: Esophageal cancer; p53; Multiple cancer; Aldehyde dehydrogenase-2; Alcoholism

1. Introduction

Multiple primary intra-esophageal squamous cell

*
carcinomas (SCCs) and oropharyngolaryngeal
Corresponding author. Tel.: +81 46 848 1550; fax: +81 46 849
SCCs concurrent with esophageal SCC are fre-
7743.
E-mail address: a_yokoyama@kurihama1.hosp.go.jp (A. Yo- quently found in Japanese alcoholics [1,2]. This can-
koyama). cer multiplicity is often explained by the concept of

0304-3835/$ - see front matter  2006 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.canlet.2006.05.004
244 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
multicentric or field cancerization, which suggests
that common intrinsic and/or extrinsic factors are
involved in their carcinogenesis [3]. Alcohol drink-
ing and tobacco smoking predispose to multiple
cancers in the upper aerodigestive tract (UADT)
[4,5]. Aldehyde dehydrogenase-2 (ALDH2) encoded
by ALDH2*1/*1 eliminates acetaldehyde [6], but
about half of East Asians have a mutant inactive
ALDH2*2 allele, and their blood acetaldehyde lev-
els increase dramatically after drinking. Homozy-
gotes for inactive ALDH2 cannot drink heavily
because they develop severe acetaldehydemia and
unpleasant flushing responses [6]. However, the
aversive effect of heterozygous ALDH2*1/*2 on
drinking is incomplete, and 26% of heavy drinkers
[7] and 12% of alcoholics in Japan [8] have been
found to be heterozygous for inactive ALDH2.
ALDH2*1/*2 is a strong risk factor for esophageal
SCC in East Asian drinkers [2,7,9,10], and for syn-
chronous or metachronous multiple cancers associ-
ated with esophageal SCC in both Japanese
alcoholic men [1,2,11,12] and Japanese men in the
general population [13,14]. Acetaldehyde plays a
critical role in field cancerization throughout the
entire mucosal surface of the UADT.
p53 tumor suppressor protein plays a major role
in cell cycle control, and its functional inactivation
by mutations is associated with the development
of human cancers [15]. The mutant p53 proteins
are more stable than the wild-type protein and thus
immunohistochemical detection of p53 protein may
indicate the presence of p53 mutations [15]. p53 pro-
tein accumulation occurs at the Tis and T1 stages of
esophageal SCC [16,17], and Japanese studies have
shown a correlation between p53 expression and a
combination of heavy drinking and heavy smoking
in esophageal SCC [18–21]. The rate of p53 expres-
sion in cancerous lesions has been found to be high-
er in multiple esophageal SCC patients with the
high-risk combination than in those without it
[21]. One study showed a higher prevalence of p53
accumulation in multiple esophageal SCC than in
solitary esophageal SCC [20], and p53 expression
has been more frequently detected in esophageal
dysplasia with SCC than in esophageal dysplasia
without SCC [22]. The frequency of p53-positive
staining of the squamous epithelium surrounding
multicentric SCCs in the UADT has been shown
to be higher than in the squamous epithelium sur-
rounding unicentric SCC [23]. The results of these
Japanese studies may mean that p53 alterations
are key molecular events in multifocal esophageal
carcinogenesis. Whether the ALDH2-associated
susceptibility to cancer multiplicity is linked to p53
expression is an interesting question, because the
direct molecular mechanism linking acetaldehyde
to cancer is poorly understood.
Ki-67 protein, which is closely associated with
cell proliferation, has been reported to accumulate
in the early stages of esophageal carcinogenesis
[16,24]. However, little is known about whether
the presence of inactive ALDH2 or cancer multiplic-
ity in the UADT affects the proliferative activity of
early esophageal SCC.
Smoking [25], less-active alcohol dehydrogenase
1B (previously called ADH2) encoded by
ADH1B*1/*1 [2], macrocytosis (mean corpuscular
volume: MCV P 106 fl) within 7 days of their last
drink [26] have also been demonstrated to increase
the susceptibility to esophageal SCC of Japanese
alcoholic men. This study was designed to find out
if correlations exist between p53 accumulation, Ki-
67 proliferative activity, and multiple field cancer-
ization of early esophageal SCC and various risk
factors in Japanese alcoholic men.
2. Materials and methods
2.1. Subjects
At the Kurihama Alcoholism Center, we routinely use
a regimented cancer-screening program consisting of
endoscopy combined with oropharyngolaryngeal inspec-
tion and esophageal iodine staining in male alcoholic
patients aged 40 or older who come to the Center for
the treatment of alcoholism [25]. Of the 3204 patients
examined in the initial screening between January 1993
and June 2000 and between November 2002 and August
2004, 131 were histologically diagnosed with esophageal
SCC, and the study subjects consisted of the 95 cancer
patients whose cancer was Tis or mucosal invasive esoph-
ageal SCC and treated by endoscopic mucosal resection.
The diagnosis of alcohol dependence was made based
on The Diagnostic and Statistical Manual of Mental Disor-
ders, third edition revised [27]. The Ethics Committee of
each collaborating institution reviewed and approved
the proposed study, and each of the participants gave
informed consent.
The entire resected esophageal mucosa was fixed in
buffered formalin solution and cut into 1.5 mm slices for
complete histologic examination. A 4-lm-thick section
was obtained from each slice embedded in paraffin for
staining with haematoxylin and eosin (H&E). The criteria
for histological diagnosis and assessment of the depth of
the cancer invasion were based on the Guidelines for the
Clinical and Pathologic Studies on Carcinoma of the
 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
Esophagus adopted by the Japanese Society for Esopha-
geal Diseases [28]. SCC was classified into four stages:
(i) stage m1, carcinoma confined within the epithelium,
diagnosed in 49 of the alcoholic patients; (ii) stage m2,
carcinoma invading the lamina propria mucosae, diag-
nosed in 27 patients; and (iii) stage m3, carcinoma invad-
ing the muscularis mucosae but not beyond, diagnosed in
19 patients.
2.2. Cancer multiplicity
The diagnosis of multiple primary carcinomas of the
esophagus was based on the results of iodine staining
examination during endoscopy [1] and histological exam-
ination. We used the following criteria proposed by Kuw-
ano et al. [29]: (i) each cancerous lesion showed definite
features of malignancy and was localized without continu-
ity with the other; and (ii) each carcinoma was accompa-
nied by areas of intraepithelial carcinoma. Synchronous
multiple intra-esophageal primary SCCs were found in
33 of the 95 alcoholic patients. At the start of the endo-
scopic screening examination, we asked the patients to
open their mouth wide in the supine position, and we
examined the palate, oral floor, tongue, and gingiva with
the endoscope. After inserting of the endoscope into the
pharynx with the patient in the left lateral position, the
oropharynx, hypopharynx, and epilarynx were carefully
examined while removing secretions by suction. Synchro-
nous oropharyngolaryngeal SCCs were detected in 11
alcoholic patients (6 hypopharyngeal; 2 oropharyngeal;
1 oropharyngeal and hypopharyngeal; 1 oral floor; and
1 gingiva). Overall, cancer multiplicity in the UADT by
cancer at one or both sites was observed in 40 (42.1%)
patients.
2.3. Immunostaining for p53 and Ki-67
In each case, we selected the paraffin block containing
the largest dimension of m1 carcinoma or the deepest front
of m2–m3 cancer invasion of all the esophageal cancer
lesions. Three serial sections, 4-lm-thick, cut from the par-
affin blocks were deparaffinized for staining with H&E and
immunostaining for p53 and Ki-67. After rehydration
through a graded series of ethanol and washing in Tris-buf-
fered saline (TBS), when antibody against p53 was used, the
sections were heated in a citrate buffer (pH 6.0) for 40 min at
95 C, and when antibody against Ki-67 was used, they
were heated in a Tris–EDTA buffer (pH 9.0) for 40 min at
95 C. The sections were then cooled for 20 min at room
temperature, and after washing in TBS, they were exposed
to 5% normal bovine serum for 5 min to block nonspecific
binding, and then incubated with primary antibody for
30 min at room temperature. The primary monoclonal anti-
bodies raised against p53 (DO7, DACO, Glostrup, Den-
mark) and Ki-67 (MIB-1, DACO) were diluted at 1:100.
Endogenous peroxidase activity was blocked by incubation
245
in 5% aqueous hydrogen peroxide for 5 min. The sections
were incubated with biotinylated secondary antibody
(ChemMate ENVISION, DACO) for 30 min at room tem-
perature, and after washing in TBS, the color was devel-
oped with diaminobenzidine (DACO DAB, DACO).
Finally, the nuclei were counterstained with Mayer’s hae-
matoxylin. Negative control staining was performed by
omitting the primary antibodies. The positive rates for
p53 and Ki-67 were calculated as the number of positive
cells divided by the total number of cancer cells examined
and expressed as a percentage. There was some heterogene-
ity in p53 as well as Ki-67 immunoreactivity among sites
observed in the cancerous lesion. To diminish the selection
bias, fields were selected from the area containing the thick-
est layer of m1 carcinoma or the deepest front of cancer
invasion of m2 and m3 carcinoma based on the examina-
tion of the H&E stained sections by one experienced
pathologist (Y.T.) blinded to the results of immunohisto-
chemistry, and then approximately 500–1000 nuclei of can-
cer cells in the corresponding fields of the p53 and Ki-67
stained sections were independently examined by two inves-
tigators (Y.T. and A.Y), one of whom (Y.T.) was blinded to
all the clinical data of the patients. The degree of the p53
and Ki-67 positivity was classified into five categories:
( ), 0–0.9% of cancer cells positive; (±), 1–9.9% of cancer
cells positive; (1+), 10–49.9% of cancer cells positive;
(2+), 50–79.9% of cancer cell positive; (3+), 80–100% of
cancer cells positive (Fig. 1).
2.4. Drinking and smoking habits
The information on drinking and smoking habits was
obtained from the patients and, when available, from
their significant others. Daily alcohol consumption during
the preceding year was expressed in grams of ethanol per
day by using a standard conversion table for alcohol bev-
erages. Beer was assumed to be 5% ethanol (v/v), wine
12%, sake 16%, shochu 25%, and whiskey 40%.
2.5. ALDH2 and ADH1B genotyping
Samples of lymphocyte DNA from the patients’ blood
were genotyped for ALDH2 and ADH1B by slight mod-
ifications [2] of the method described by Harada and
Zhang [30] and the method described by Xu et al. [31]
using polymerase chain reaction-restriction fragment
length polymorphism (PCR-RFLP).
2.6. MCV measurement
MCV was measured the first day the patients came the
Center for treatment of alcoholism. Blood was drawn
from the antecubital vein into a tube containing EDTA.
MCV was measured by the electrical impedance method
with an autoanalyzer (CELL-DYN 3500, Abbott, North
Chicago, IL) within 4 h of collection.
246 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
Fig. 1. Immunohistochemistry for p53 and Ki-67 in mucosal esophageal squamous cell carcinoma. Three serial sections showing staining
with H&E (a) and immunostaining for p53 (b) and Ki-67 (c) (40·).
3. Results
Age, alcohol consumption, cigarette consumption,
ALDH2/ADH1B genotypes, and MCV within 7 days of
their last drink did not differ among the alcoholics with
m1, m2, and m3 carcinoma (Table 1).
A very high prevalence of p53 overexpression was
found throughout all depth categories of esophageal
SCC (Table 2). There were no difference in Ki-67 prolifer-
ative activity or frequency of cancer multiplicity among
those with m1, m2, and m3 carcinoma.
Table 3 shows the relationships between p53 immuno-
reactivity, Ki-67 labeling index, and selected variables.
Age, alcohol, and cigarette consumption, ALDH2/
ADH1B genotype, MCV, and cancer multiplicity in the
UADT had no effect on p53 or Ki-67 expression. Nor
was there any correlation between p53 expression and
Ki-67 expression.
The risk for cancer multiplicity in the UADT in the
form of either intra-esophageal multiple SCCs or concur-
rent oropharyngolaryngeal SCC was higher in the
ALDH2*1/*2 heterozygotes (OR = 4.22) than in the
ALDH2*1/*1 homozygotes. p53 expression, Ki-67
expression, smoking, ADH1B genotype, and MCV were
unassociated with cancer multiplicity in the UADT
(Table 4).
4. Discussion
High rates of intense p53 expression were found
in Tis and mucosal invasive esophageal SCCs in
the Japanese alcoholic men. The rate of p53 immu-
noreactivity in the alcoholics was 84% for (1+),
(2+), or (3+), 74% for (2+) or (3+), and 62% for
(3+). When monoclonal antibody DO7 was used,
the rates in Japanese patients with Tis and T1
esophageal SCC in general hospitals were reported
to be 31–61% for (1+), (2+), or (3+) [16,18,20]
and 12% for (2+) or (3+) [16]. The first Japanese
study of Tis and T1 esophageal SCC patients
showed a lower frequency of p53 accumulation in
Tis carcinoma than in T1 carcinoma [16], but a later
Japanese study showed that p53 protein expression
was not correlated with histological stage, including
Tis and T1 [17]. In our alcoholic population, the
rate of p53 overexpression was very high even in
Tis carcinoma. The high levels of p53 accumulation
occurring in very early stages of esophageal carcino-
genesis in the alcoholics are not surprising, because
a combination of heavy drinking and heavy smok-
ing is correlated with p53 expression [18–21].
 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252 247

Table 1
The basic characteristics of Japanese alcoholic men with esophageal squamous cell carcinoma
Total (n = 95) Depth of the main cancerous lesion Pa
m1 (n = 49) m2 (n = 27) m3 (n = 19)
Age (years) 57.2 ± 8.6 56.9 ± 8.0 56.9 ± 8.4 58.5 ± 10.4 0.78
Alcohol drinking (units)b 4.5 [3.0, 6.0] 4.5 [3.3, 6.0] 5.0 [3.5, 6.5] 4.0 [3.0, 5.0] 0.41
Cigarette smoking (cigs/day) 20 [15, 30] 20 [15, 30] 25 [15, 40] 20 [0, 20] 0.12
ALDH2 genotype
*1/*1 44.2% 32.7% 59.3% 52.6%
*1/*2 55.8% 67.4% 40.7% 47.4% 0.055
ADH1B genotype
*1/*1 49.5% 40.8% 66.7% 47.4%
*1/*2 23.2% 32.7% 7.4% 21.1%
*2/*2 27.4% 26.5% 25.9% 31.6% 0.11
MCV, flc 104.2 ± 9.1 105.3 ± 9.5 101.0 ± 8.0 105.4 ± 8.9 0.23
Values are means ± standard deviation, median [25, 75 percentiles], or column percent.
a
P for homogeneity among m1 through m3 groups by analysis of variance, Kruskall–Wallis test, and Fisher’s exact test for comparing
mean, median, and percent values, respectively.
b
1 U = 22 g ethanol.
c
MCV within 7 days of taking their last drink was obtained from 67 alcoholic men.

Table 2
p53 immunoreactivity and Ki-67 labeling index and cancer multiplicity in Japanese alcoholic men with esophageal squamous cell
carcinoma
Total (n = 95) Depth of the main cancerous lesion Pa
m1 (n = 49) m2 (n = 27) m3 (n = 19)
p53 immunoreactivity (%) 85.7 [34.0, 95.7] 84.8 [72.2, 93.6] 89.5 [29.7, 96.9] 84.6 [1.4, 96.4] 0.75
: 0–0.9 9.5% 6.1% 7.4% 21.1%
±: 1–9.9 6.3% 4.1% 7.4% 10.5%
+: 10–49.9 10.5% 10.2% 11.1% 10.5%
2+: 50–79.9 11.6% 16.3% 11.1% 0.0%
3+: 80–100 62.1% 63.3% 63.0% 57.9% 0.45
Ki-67 labeling index (%) 29.8 [19.5, 49.9] 38.0 [20.3, 54.3] 30.2 [19.5, 48.6] 24.0 [12.9, 45.6] 0.33
: 0–0.9 0.0% 0.0% 0.0% 0.0%
±: 1–9.9 10.5% 4.1% 18.5% 15.8%
+: 10–49.9 65.3% 67.4% 59.3% 68.4%
2+: 50–79.9 23.2% 28.6% 22.2% 10.5%
3+: 80–100 1.1% 0.0% 0.0% 5.3% 0.10
Intra-esophageal SCC
Solitary 65.3% 63.3% 70.4% 63.2%
Multiple 34.7% 36.7% 29.6% 36.8% 0.84
Concurrent oropharyngolaryngeal SCC
Absent 88.4% 93.9% 88.9% 73.7% 0.076
Present 11.6% 6.1% 11.1% 26.3%
Cancer multiplicity in the UADTb
Absent 57.9% 59.2% 66.7% 42.1% 0.24
Present 42.1% 40.8% 33.3% 57.9%
Values are median [25, 75 percentiles] or column percent.
a
P for homogeneity among m1 through m3 groups by analysis of variance, Kruskall–Wallis test, and Fisher’s exact test for comparing
mean, median, and percent values, respectively.
b
Intra-esophageal multiple SCCs and/or concurrent oropharyngolaryngeal SCC. UADT, upper aerodigestive tract.

The most common p53 mutations are missense mutations in esophageal SCC do not result in
mutations, and they generally result in accumula- detectable accumulation of p53 protein [32]. Since
tion of altered p53 protein with an extended the proportion of nonsense or frameshift mutations
half-life, but nonsense mutations and frameshift among all p53 mutations in Japanese esophageal
248 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252

Table 3
Correlation of p53 immunoreactivity and Ki-67 labeling index with selected factors
p53 Ki-67
r P r P
Age 0.15 0.14 0.17 0.095
Alcohol drinkinga 0.04 0.68 0.03 0.74
Cigarette smokinga 0.06 0.60 0.04 0.72
ALDH2b 0.11 0.28 0.01 0.96
ADH1Bb 0.03 0.79 0.06 0.54
Mean corpuscular volumea 0.14 0.25 0.11 0.37
Cancer multiplicity in the UADTa 0.08 0.42 0.04 0.70
p53a – 0.03 0.74
Ki-67a 0.03 0.74 –
r, P: Spearman’s rank correlation coefficient and P value. ALDH2 genotype was coded as 1 (*1/*2) or 0 (*1/*1); ADH1B genotype, as 1
(*1/*1) or 0 (others); cancer multiplicity, as 1 (present) or 0 (absent).
a
Adjusted for age.
b
Adjusted for age and drinking.

SCC has been reported to be 29% [32], it is plausible
that a substantial fraction of the p53 ( /±) alcohol-
ics had nonsense or frameshift mutations rather
than no mutations.
Contrary to expectation, we failed to find any
correlations between p53 expression by esophageal
SCC and cancer multiplicity in the UADT or
ALDH2 genotype in the alcoholic patients. A previ-
ous Japanese study that included 9 patients with
multiple esophageal SCCs and 80 patients with sol-
itary SCC linked p53 expression to cancer multiplic-
ity [20]. That study also demonstrated associations
between both heavy drinking and heavy smoking
and the rate of p53 expression. The difference
between the results of the two studies can be
explained as follows. p53 expression and cancer
multiplicity in the UADT developed much more fre-
quently as a result of chronic heavy exposure to
alcohol and tobacco in the alcoholics, and the
resulting homogeneity of the tissues with regard to
their extremely high rate of p53 alterations may
have eclipsed possible associations between p53
expression and cancer multiplicity in the UADT, if
they existed. Or, the contribution of p53 alterations
per se to multicentric cancerization may be small,
even though both p53 alterations and multicentric
cancerization are frequently present due to the com-
bination of heavy drinking and heavy smoking. A
study that directly analyzes p53 mutations is needed
to clarify these issues.
Multiple factors associated with alcoholism may
contribute to the overexpression of p53. Alcoholics’
poor dietary habits lead to deficiency of several
nutrients, including vitamins and minerals. Chronic
alcohol consumption and alcohol metabolism result
in the generation of several classes of DNA-damag-
ing molecules, including reactive oxygen species and
lipid peroxidation products [33]. Chronic alcohol
consumption leads to induction of cytochrome
P-4502E1, which metabolizes ethanol to acetalde-
hyde and increases activation of the nitrosamines
present in tobacco smoke to their ultimate carcino-
gens [34]. Ethanol-induced P4502E1 has been
demonstrated in the UADT of humans [35].
The proliferative activity of cancerous lesions
assessed by immunostaining for Ki-67 has been
reported to be correlated with the outcome of a
variety of neoplasias [36], including esophageal
SCC [37]. A previous Japanese study showed that
Ki-67 expression was more prominent in mucosal
invasive carcinoma than in Tis carcinoma [16]. In
the present study, the Ki-67 labeling index in Tis
carcinoma was as high as in mucosal invasive carci-
noma, indicating high proliferative activity by very
early carcinoma in this population. The Ki-67
expression was not associated with ALDH2 geno-
type or cancer multiplicity in the UADT, and these
findings were consistent with the fact that the depth
of the main cancerous lesion was also unassociated
with either of them.
Among the risk factors, investigated smoking,
inactive heterozygous ALDH2, less-active ADH1B,
and macrocytosis, only inactive heterozygous
ALDH2 was linked to multiplicity of the SCC,
enhancing the validity of the link between acetal-
dehyde exposure and cancer multiplicity. The high
level exposure of the UADT to acetaldehyde may
occur through several other pathways besides
systemic exposure. Alcoholic beverages themselves
and tobacco smoke contain high levels of
 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252 249

Table 4
Associations of cancer multiplicity in the upper aerodigestive tract with p53 immunoreactivity, Ki-67 labeling index, and selected variables
Cancer multiplicity
Present (n = 40) Absent (n = 55) ORa 95% CI
Age (years)
<50 13% 18% 1 (referent)
50–59 53% 45% 1.67 0.43–7.24
60–69 28% 27% 1.45 0.33–7.06
70+ 8% 9% 1.19 0.13–9.69
P = 0.90 for trend
p53b
10% 9% 1 (referent)
± 5% 7% 0.64 0.04–7.83
+ 15% 7% 1.81 0.22–16.45
2+ 5% 16% 0.30 0.02–2.98
3+ 65% 60% 0.98 0.19–5.46
P = 1.0 for trend
Ki-67 (%)b
± 10% 11% 1 (referent)
+ 70% 62% 1.21 0.26–6.42
2+ 20% 25% 0.85 0.14–5.37
3+ 0% 2% 1.65 0.00–64.41
P = 0.60 for trend
Alcohol drinking (units/day)b
63.5 45% 33% 1 (referent)
3.6–5.0 38% 27% 1.02 0.33–3.18
5.1+ 18% 40% 0.33 0.09–1.06
P = 0.057 for trend
Cigarette smokingb
<20 cigs/day 30% 25% 1 (referent)
20–29 cigs/day 30% 35% 0.74 0.23–2.39
30+ cigs/day 30% 36% 0.70 0.21–2.33
Ex-smoker 10% 4% 2.25 0.27–28.77
P = 0.59 for homogeneity
ALDH2 genotypec
*1/*1 25% 58% 1 (referent)
*1/*2 75% 42% 4.22 1.58–12.16
P = 0.0015
ADH1B genotypec
*1/*1 43% 55% 0.57 0.22–1.44
*1/*2 + *2/*2 58% 45% 1 (referent)
P = 0.21
MCV, fl (n = 30) (n = 37)
<106 60% 81% 1 (referent)
106+ 40% 19% 2.75 0.82–9.86
P = 0.10
a
ORs (odds ratios), 95% CIs (confidence intervals), and P values were calculated by the exact logistic regression model; ORs with ‘0%’
were the median unbiased estimates.
b
ORs were adjusted for age.
c
ORs were adjusted for age and alcohol drinking.

acetaldehyde, and the acetaldehyde levels in saliva oral microflora also form acetaldehyde from etha-
are markedly higher than in blood after ethanol nol and contribute to acetaldehyde levels in saliva
ingestion, especially in inactive-ALDH2 heterozyg- [38,39]. Another pathway is related to topical
otes [38] and during active smoking [39]. Normal mucosal enzymes in the UADT. High-Km
250 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
ADH4 (previously called ADH7) is strongly
expressed in the UADT and is active in producing
acetaldehyde upon exposure to a locally high dose
of ethanol, whereas ALDH2 activity in the
UADT is extremely weak, if present at all. Thus,
a high level of exposure and inefficient degrada-
tion of acetaldehyde may occur in the UADT [40].
The carcinogenicity of acetaldehyde has been
adequately demonstrated in experimental animals
[41], and acetaldehyde has been shown to interact
with DNA to form DNA adducts in humans [42].
The high level of exposure of the UADT to ethanol
and acetaldehyde may affect the local metabolism of
folate and retinoic acid, local deficiencies of which
may be involved in alcohol-induced carcinogenesis.
Folate deficiency leads to alterations in DNA meth-
ylation and disruption of DNA integrity and repair
[43]. Acetaldehyde increases folate catabolism in vi-
tro [44], and serum folate levels are more likely to be
lowered by alcohol drinking in men with inactive
ALDH2 than in men with active ALDH2 [45]. Ret-
inoic acid is an important factor in regulating cell
differentiation and proliferation and is synthesized
from retinol via various enzymatic steps involving
ADH and ALDH. High concentrations of ethanol
and acetaldehyde inhibit retinoic acid formation in
rat esophagus [46] and human prenatal tissues
[47]. Retinoic acid exerts its biological effect by
binding to specific nuclear receptors, including reti-
noic acid receptor-b, loss of whose expression is an
early event during esophageal carcinogenesis
[24,48]. These findings provide evidence of several
possible mechanisms of acetaldehyde-related field
cancerization in the UADT.
ALDH2*1/*2 greatly increases the risk of
metachronous cancer multiplicity as well as syn-
chronous cancer multiplicity in Japanese alcoholics
[11,12]. The proportion of patients with metachro-
nous development of intra-esophageal SCC and
hypopharyngeal/epilaryngeal SCC had reached
70% and 25%, respectively, 5 years after the diagno-
sis of the first esophageal SCC in ALDH2*1/*2
alcoholics, as opposed to less than 10% and 0%,
respectively, in ALDH2*1/*1 alcoholics [12]. p53
protein accumulation is frequently detected in mul-
tifocal noncancerous lesions of the esophagus with
SCC [22,23,49]. Some of the metachronous cancers
may have been present in the form of p53-positive
precursor lesions at the time of diagnosis of the first
cancer and subsequently become cancerous lesions
that were distinct enough to be identifiable. In the
present study we investigated only cancerous lesions
resected by endoscopic mucosectomy. However,
esophageal iodine staining shows numerous minute
areas that do not take up iodine in most Japanese
alcoholics with esophageal SCC. Investigation of
the condition of the background mucosa, which
might have premalignant potential, may clarify the
relationship between p53 expression and cancer
multiplicity, and ALDH2-associated p53 alterations
also may be demonstrated in studies of the associa-
tion between p53 expression and ALDH2 genotype
in such precursor lesions.
Acknowledgements
This work was supported by a Grant-in-Aid for
Cancer Research from the Ministry of Health, La-
bor, and Welfare and by Medical School Faculty
and Alumni Grants from Keio University Medical
Science Fund.
References
[1] A. Yokoyama, T. Muramatsu, T. Ohmori, H. Makuuchi, S.
Higuchi, S. Matsushita, K. Yoshino, K. Maruyama, M.
Nakano, H. Ishii, Multiple primary esophageal and concur-
rent upper aerodigestive tract cancer and the aldehyde
dehydrogenase-2 genotype of Japanese alcoholics, Cancer 77
(1996) 1986–1990.
[2] A. Yokoyama, T. Muramatsu, T. Omori, T. Yokoyama, S.
Matsushita, S. Higuchi, K. Maruyama, H. Ishii, Alcohol and
aldehyde dehydrogenase gene polymorphisms and orophar-
yngolaryngeal, esophageal and stomach cancers in Japanese
alcoholics, Carcinogenesis 22 (2001) 433–439.
[3] D.P. Slaughter, H.W. Southwick, W. Smejkal, ‘‘Field canc-
erization’’ in oral stratified squamous epithelium. Clinical
implications of multicentric origin, Cancer 6 (1953) 963–968.
[4] M. Morita, H. Kuwano, S. Ohno, K. Sugimachi, Y. Seo, H.
Tomoda, M. Furusawa, T. Nakashima, Multiple occurrence
of carcinoma in the upper aerodigestive tract associated with
esophageal cancer: reference to smoking, drinking, and
family history, Int. J. Cancer 58 (1994) 207–210.
[5] L.H. Schwartz, M. Ozsahin, G. Ni Zhang, E. Touboul, F. De
Vataire, P. Andlenko, J. Lacau-Saint Guily, A. Laugier, M.
Schlienger, Synchronous and metachronous head and neck
carcinomas, Cancer 74 (1994) 1933–1938.
[6] W.F. Bosron, T.K. Li, Genetic polymorphism of human
liver alcohol and aldehyde dehydrogenases, and their rela-
tionship to alcohol metabolism and alcoholism, Hepatology
6 (1986) 502–510.
[7] A. Yokoyama, H. Kato, T. Yokoyama, T. Tsujinaka, M.
Muto, T. Omori, T. Haneda, Y. Kumagai, H. Igaki, M.
Yokoyama, H. Watanabe, H. Fukuda, H. Yoshimizu,
Genetic polymorphisms of alcohol and aldehyde dehydro-
genases and glutathione S-transferase M1 and drinking,
smoking, and diet in Japanese men with esophageal squa-
mous cell carcinoma, Carcinogenesis 23 (2002) 1851–1859.
[8] S. Higuchi, S. Matsushita, M. Murayama, S. Takagi, M.
Hayashida, Alcohol and aldehyde dehydrogenase polymor-
 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
phisms and the risk for alcoholism, Am. J. Psychiatry 152
(1995) 1219–1221.
[9] A. Yokoyama, T. Omori, Genetic polymorphisms of alcohol
and aldehyde dehydrogenases and risk for esophageal and
head and neck cancers, Jpn. J. Clin. Oncol. 33 (2003) 111–
121.
[10] C.F. Wu, D.C. Wu, H.K. Hsu, E.L. Kao, J.M. Lee, C.C.
Lin, M.T. Wu, Relationship between genetic polymorphisms
of alcohol and aldehyde dehydrogenases and esophageal
squamous cell carcinoma risk in males, World J. Gastroen-
terol. 11 (2005) 5103–5108.
[11] A. Yokoyama, T. Ohmori, T. Muramatsu, T. Yokoyama, K.
Okuyama, H. Makuuchi, H. Takahashi, S. Higuchi, M.
Hayashida, K. Maruyama, H. Ishii, Short-term follow-up
after endoscopic mucosectomy of early esophageal cancer
and aldehyde dehydrogenase-2 genotype in Japanese alco-
holics, Cancer Epidemiol. Biomark. Prev. 7 (1998) 473–476.
[12] A. Yokoyama, T. Omori, T. Yokoyama, Alcohol and
aldehyde dehydrogenase gene polymorphisms and multiple
cancers associated with esophageal squamous cell carcinoma
in Japanese alcoholics. I to Cho 38 (2003) 339–348 (in
Japanese with English abstract).
[13] A. Yokoyama, H. Watanabe, H. Fukuda, T. Haneda, H.
Kato, T. Yokoyama, T. Muramatsu, H. Igaki, Y. Tachimori,
Multiple cancers associated with esophageal and orophar-
yngolaryngeal squamous cell carcinoma and the aldehyde
dehydrogenase-2 genotype in male Japanese drinkers, Can-
cer Epidemiol. Biomark. Prev. 11 (2002) 895–900.
[14] M. Muto, M. Takahashi, A. Ohtsu, S. Ebihara, S. Yoshida,
H. Esumi, Risk of multiple squamous cell carcinomas both
in the esophagus and the head and neck region, Carcino-
genesis 26 (2005) 1008–1012.
[15] L.J. Hofseth, S.P. Hussain, C.C. Harris, p53. 25 years after
discovery, Trends Pharmacol. Sci. 25 (2004) 177–181.
[16] K. Ohashi, T. Nemoto, Y. Eishi, A. Matsuno, K. Nakamura,
K. Hirokawa, Proliferative activity and p53 protein accu-
mulation correlate with early invasive trend, and apoptosis
correlates with differentiation grade in oesophageal squa-
mous cell carcinomas, Virchows Arch. 430 (1997) 107–115.
[17] O. Chino, H. Kijima, H. Shimada, T. Nishi, H. Tanaka, Y.
Kise, T. Kenmochi, S. Himeno, T. Machimura, M. Tanaka,
S. Inokuchi, T. Tajima, R.Y. Osamura, H. Makuuchi,
Accumulation of p53 in esophageal squamous cell carci-
noma, Int. J. Mol. Med. 8 (2001) 359–363.
[18] M. Miyazaki, S. Ohno, M. Futatsugi, H. Saeki, T. Ohga, M.
Watanabe, The relation of alcohol consumption and ciga-
rette smoking to the multiple occurrence of esophageal
dysplasia and squamous cell carcinoma, Surgery 131 (2002)
S7–S13.
[19] H. Saeki, S. Ohno, K. Araki, A. Egashira, H. Kawaguchi, Y.
Ikeda, M. Morita, K. Kitamura, K. Sugimachi, Alcohol
consumption and cigarette smoking in relationship to high
frequency of p53 protein accumulation in oesophageal
squamous cell carcinoma in the Japanese, Br. J. Cancer 82
(2000) 1892–1894.
[20] H. Kato, M. Yoshikawa, T. Miyazaki, M. Nakajima, Y.
Fukai, K. Tajima, N. Masuda, K. Tsukada, T. Fukuda, T.
Nakajima, H. Kuwano, Expression of p53 protein related to
smoking and alcohol beverage drinking habits in patients
with esophageal cancers, Cancer Lett. 167 (2001) 65–72.
[21] H. Saeki, S. Ohno, M. Miyazaki, K. Araki, A. Egashira, H.
Kawaguchi, M. Watanabe, M. Morita, K. Sugimachi, p53
251
protein accumulation in multiple oesophageal squamous cell
carcinoma: relationship to risk factors, Oncology 62 (2000)
175–179.
[22] H. Kawakubo, S. Ozawa, N. Ando, Y. Kitagawa, M. Mukai,
M. Ueda, M. Kitajima, Alterations of p53, cyclin D1 and
pRB expression in the carcinogenesis of esophageal squa-
mous cell carcinoma, Oncol. Rep. 14 (2005) 1453–1459.
[23] Y. Nakanishi, M. Noguchi, Y. Matsuno, M. Saikawa, K.
Mukai, Y. Shimosato, S. Hirohashi, p53 expression in
multicentric squamous cell carcinoma and surrounding
squamous epithelium of the upper aerodigestive tract.
Immunohistochemical analysis of 95 lesions, Cancer 75
(1995) 1657–1662.
[24] M. Xu, Y.L. Jin, J.J. Fu, H. Huang, S.Z. Chen, P. Qu, H.M.
Tian, Y.Z. Lui, W. Zhang, The abnormal expression of
retinoic acid receptor-b, P53 and Ki67 protein in normal,
premalignant and malignant esophageal tissues, World J.
Gastroenterol. 8 (2002) 200–202.
[25] A. Yokoyama, T. Ohmori, H. Makuuchi, K. Maruyama, K.
Okuyama, H. Takahashi, T. Yokoyama, K. Yoshino, M.
Hayshida, H. Ishii, Successful screening for early esophageal
cancer in alcoholics using endoscopy and mucosal iodine
staining, Cancer 76 (1995) 928–934.
[26] A. Yokoyama, T. Yokoyama, T. Muramatsu, T. Omori, S.
Matsushita, S. Higuchi, K. Maruyama, H. Ishii, Macrocy-
tosis, a new predictor for esophageal squamous cell carci-
noma in Japanese alcoholic men, Carcinogenesis 24 (2003)
1773–1778.
[27] American Psychiatric Association, Diagnostic and Statistical
Manual of Mental Disorders. third rev ed., APA, Washing-
ton, DC, 1987 pp. 173–175.
[28] Japanese Society for Esophageal Diseases, Guidelines for
Clinical and Pathologic Studies on Carcinoma of the
Esophagus. nineth ed., Kanehara & Co., Ltd., Tokyo, 1999
pp. 11.
[29] H. Kuwano, S. Ohno, H. Matsuda, M. Mori, K. Sugimachi,
Serial histologic evaluation of multiple primary squamous
cell carcinomas of the esophagus, Cancer 61 (1988) 1635–
1638.
[30] S. Harada, S. Zhang, New strategy for detection of ALDH2
mutant, Alcohol Alcohol. 28 (Suppl. 1A) (1993) 11–13.
[31] Y. Xu, L.G. Carr, W.F. Bosron, T.K. Li, Genotyping of
human alcohol dehydrogenases at the ADH2 and ADH3 loci
following DNA sequence amplification, Genomics 2 (1988)
209–214.
[32] T. Wagata, I. Shibagaki, M. Imamura, Y. Shimada, J.
Toguchida, D.W. Yandell, M. Ikenaga, T. Tobe, K. Ishizaki,
Loss of 17p, mutation of the p53 gene, and overexpression of
p53 protein in esophageal squamous cell carcinomas, Cancer
Res. 53 (1993) 846–850.
[33] P.J. Brooks, DNA damage, DNA repair, and alcohol
toxicity: a review, Alcohol. Clin. Exp. Res. 21 (1997) 1073–
1082.
[34] H.K. Seitz, G. Poschl, U.A. Simanowski, Alcohol and
cancer, in: M. Galanter (Ed.), Recent Development in
Alcoholism: The Consequences of Alcoholism, Plenum
Press, New York, London, 1988, pp. 67–96.
[35] G.. Baumgarten, R. Waldherr, F. Stickel, U.A. Simanowski,
M. Ingelmann-Sundverg, H.K. Seitz, Enhanced expression
of cytochrome p450 2E1 in the oropharyngeal mucosa in
alcoholics with cancer [Abstract]. Alcohol. Clin. Exp. Res. 20
(Suppl. 2) (1996) 80A (abstract no. 458).
252 A. Yokoyama et al. / Cancer Letters 247 (2007) 243–252
[36] J. Gerdes, Ki-67 and other proliferation markers useful for
immunological diagnostic evaluations in human malignan-
cies, Semin. Cancer Biol. 1 (1990) 199–206.
[37] E.M. Youssef, T. Matsuda, N. Takada, H. Osugi, M.
Higashino, H. Kinoshita, T. Watanabe, Y. Katsura, H.
Wanibuchi, S. Fukushima, Prognostic significance of the
MIB-1 proliferation index for patients with squamous cell
carcinoma of the esophagus, Cancer 76 (1995) 358–366.
[38] S. Väkeväinen, J. Tillonen, D.P. Agarwal, N. Srivastava, M.
Salaspuro, High salivary acetaldehyde after a moderate dose
of alcohol in ALDH2 deficient subjects: strong evidence for
local carcinogenetic action of acetaldehyde, Alcohol. Clin.
Exp. Res. 24 (2000) 873–877.
[39] V. Salaspuro, M. Salaspuro, Synergistic effect of alcohol
drinking and smoking on in vivo acetaldehyde concentration
in saliva, Int. J. Cancer 111 (2004) 480–483.
[40] S.J. Yin, C.S. Liao, C.W. Wu, T.T. Li, L.L. Chen, C.L.
Lai, T.Y. Tsao, Human stomach alcohol and aldehyde
dehydrogenases: comparison of expression pattern and
activities in alimentary tract, Gastroenterology 12 (1997)
766–775.
[41] International Agency for Research on Cancer, Allyl com-
pounds, aldehydes, epoxides and peroxides, in: IARC
Monographs on the Evaluation on the Carcinogenic Risk
of Chemicals to Humans, vol. 36, IARC, Lyon, 1985, pp.
101–132.
[42] J.L. Fang, C.E. Vaca, Detection of DNA adducts of
acetaldehyde in peripheral white blood cells of alcohol
abusers, Carcinogenesis 18 (1997) 627–632.
[43] S.W. Choi, J.B. Mason, Folate and carcinogenesis: an
integrated scheme, J. Nutr. 130 (2000) 129–132.
[44] S. Shaw, E. Jayatilleke, V. Herbert, N. Colman, Cleavage of
folates during ethanol metabolism. Role of acetaldehyde/
xanthine oxidase-generated superoxide, Biochem. J. 257
(1989) 277–280.
[45] T. Yokoyama, K. Saito, H. Lwin, N. Yoshiike, A. Yamam-
oto, Y. Matsushita, C. Date, H. Tanaka, Epidemiological
evidence that acetaldehyde plays a significant role in the
development of decreased serum folate concentration and
elevated mean corpuscular volume in alcohol drinkers,
Alcohol. Clin. Exp. Res. 29 (2005) 622–630.
[46] H. Shiraishi-Yokoyama, H. Yokoyama, M. Matsumoto, H.
Imaeda, T. Hibi, Acetaldehyde inhibits the formation of
retinoic acid from retinal in the rat esophagus, Scand. J.
Gastroenterol. 41 (2006) 80–86.
[47] M. Khalighi, M.R. Brezezinski, H. Chen, M.R. Juchau,
Inhibition of human prenatal biosynthesis of all-trans-
retinoic acid by ethanol, ethanol metabolites, and products
of lipid peroxidation reactions: a possible role for CYP2E1,
Biochem. Pharmacol. 57 (1999) 811–821.
[48] H. Qui, W. Zhang, A.K. El-Naggar, S.M. Lippman, P. Lin,
R. Lotan, X.C. Xu, Loss of retinoic acid receptor-beta
expression is an early event during esophageal carcinogen-
esis, Am. J. Pathol. 155 (1999) 1519–1523.
[49] D. Tian, Z. Feng, N.M. Hanley, R.W. Setzer, J.L. Mumford,
D.M. De Marini, Multifocal accumulation of p53 protein in
esophageal carcinoma: evidence for field cancerization, Int.
J. Cancer 78 (1998) 568–575.