Full text of DOI:10.1097/00005537-200203000-00011

The Laryngoscope
Lippincott Williams & Wilkins, Inc., Philadelphia
© 2002 The American Laryngological,
Rhinological and Otological Society, Inc.
The Association Between Glutathione
S-Transferase P1 Genotype and Plasma
Level in Head and Neck Cancer
Willem P. A. Kelders, MD; Michael B. Oude Ophuis, MD; Hennie M. J. Roelofs, MSc;
Wilbert H. M. Peters, PhD; Johannes J. Manni, MD, PhD
Objectives/Hypothesis: Glutathione S-transferase
P1–1 (GSTP1–1) is an important enzyme because it plays
a major role in many detoxification reactions, including
tobacco-related metabolic products. GSTP1–1 is the
most abundant of all glutathione S-transferase (GST)
enzymes in normal human head and neck epithelium,
whereas it is overexpressed in head and neck malignan-
cies. At least three different GSTP1 genotypes exist, AA,
AB, and BB, which have been correlated with reduced
enzyme activity. Many authors have studied the GSTP1
genotypes in relation to the risk for human head and
neck squamous cell carcinoma (HNSCC). A correlation
between GSTP1–1 genotype and GSTP1–1 plasma levels
has not been made before. We investigated the correla-
tion between GSTP1 genotype and GSTP1–1 plasma lev-
els. Study Design: To evaluate the possible association
between the genetic polymorphisms in GSTP1 and the
phenotypic expression (GSTP1–1 plasma levels) in pa-
tients with HNSCC. Methods: GSTP1 genotype and
GSTP1–1 plasma level were established in 87 patients
with HNSCC and 51 patients with benign head and neck
lesions who served as control subjects. Results: For all
GSTP1 genotypes (AA, AB, and BB) in patients with
HNSCC, the mean plasma GSTP1–1 values were signif-
icantly higher compared with the control subjects.
There was no significant difference in the plasma
GSTP1–1 levels between the different genotypes in pa-
tients with HNSCC. Conclusion: There is no association
between GSTP1 genotype and GSTP1–1 plasma levels in
patients with head and neck cancer. Key Words: Gluta-
thione S-transferase P1 genotype, glutathione
S-transferase P1–1 plasma level, head and neck squa-
mous cell carcinoma, detoxification enzymes.
INTRODUCTION
Squamous cell carcinoma of the head and neck region
(HNSCC) is the fifth most common malignancy in the
world. Most cases of HNSCC are associated with alcohol
abuse or smoking cigarettes, or both, but only a small
number of smokers do develop HNSCC. This suggests the
influence of genetic components in the susceptibility to
HNSCC. Some of the factors that could explain this dif-
ference in susceptibility may be the presence of polymor-
phisms in genes coding for enzymes involved in carcinogen
metabolism, such as the glutathione S-transferase (GST)
enzymes. GST enzymes remain a focal point for research
because of their role as defense mechanism against xeno-
biotics such as polycyclic aromatic hydrocarbons, which
are an important class of carcinogens present in tobacco
smoke.1 Most carcinogens have to be activated by phase 1
enzymes such as cytochrome P450 1A1 (CYP1A1) to be-
come a carcinogen in the form of reactive epoxide inter-
mediates such as benzo[␣]pyrene diol epoxide.1 These car-
cinogens subsequently may be detoxified by various phase
2 enzymes such as GST enzymes.
Human GST enzymes can be subdivided into five
classes, alpha (A), mu (M), pi (P), theta (T), and zeta (Z).
Each class includes one or more isoenzymes. Glutathione
S-transferase P1–1 (GSTP1–1) is a major enzyme in the
detoxification of tobacco-related metabolic products,2 and
because it is quantitatively the most abundant GST en-
zyme in the human head and neck area,3 many research-
ers have tried to link its genotype to the risk for
HNSCC.4–7
Several genotypes of the GSTP1 gene have been de-
scribed, one resulting in isoleucine versus valine amino
acid substitution at position 104 in the corresponding
enzyme.8 This 104 valine or GSTP1 B variant is homozy-
gously present in 10% of the Caucasian population. Ap-
proximately 55% of the Caucasian population is homozy-
gous for the 104 isoleucine or GSTP1 A allele, whereas
approximately 35% of the Caucasian population is het-
erozygous, containing both alleles (Ile and Val).
Laryngoscope, 112:462–466, 2002
From the Department of Otorhinolaryngology—Head and Neck Sur-
gery (W.P.A.K., M.B.O.O., J.J.M.), University Hospital Maastricht, Maastricht,
The Netherlands, and the Department of Gastroenterology (H.M.J.R.,
W.H.M.P.), University Hospital St. Radboud, Nijmegen, The Netherlands.
Editor’s Note: This Manuscript was accepted for publication May 22,
2001.
Send Correspondence to Johannes J. Manni, MD, PhD, Department
of Otorhinolaryngology—Head and Neck Surgery, University Hospital
Maastricht, Post Office Box 5800, 6202 AZ Maastricht, The Netherlands.
Homozygosity for the wild-type A allele has been
found in a lower incidence in patients with oral and pha-
Laryngoscope 112: March 2002
462
Kelders et al.: Glutathione S-Transferase P1 Genotype

ryngeal carcinoma in a Caucasian population,4,5 whereas
a higher incidence of the A allele in patients with pharyn-
geal and laryngeal cancer in a Japanese study has been
described.6 Another study revealed no significant differ-
ences in GSTP1 genotypes in patients with laryngeal can-
cer as compared with healthy control subjects.7 Our group
found that genetic polymorphisms of GSTP1 are not asso-
ciated with altered susceptibility to HNSCC (M.B.O.O. et
al., unpublished data, 2001).
tients with HNSCC (21 female patients [24%] and 66 male pa-
tients [76%]), all with no previous history of cancer. In the group
with benign head and neck lesions, the mean age was 49 Ϯ 18
years (range, 16–86 y). In the HNSCC group, the mean age was
58 Ϯ 12 years (range, 23–82 y). All sites of tumor expression are
given in Table I.
Informed consent was obtained, and medical and other an-
amnestic data were obtained by interview. The Medical Ethical
Review Committee approved the investigations.
GSTP1 genotypes and their possible influence on the
risk of other forms of malignant and benign tumors have
been evaluated in patients with lung cancer; one study
showed a significantly higher incidence of the GSTP1 BB
genotype in the cancer group,1 whereas another study
revealed no difference.9 In urogenital malignancies signif-
icant higher incidence of the BB genotype was found,10
whereas in gastrointestinal malignancies no higher inci-
dence was found.11,12 However, in premalignant Barrett’s
esophagus a higher incidence of the GSTP1 B genotype
was demonstrated.13 In breast carcinoma a higher inci-
dence of the GSTP1 BB genotype was found as well.14
Because GSTP1–1 was shown to be overexpressed in
a wide variety of tumors,15 another field of interest is the
relevance of plasma GSTP1–1 levels as tumor marker in
several categories of patients with cancer.16–20 However,
arguments pleading against a role of GSTP1–1 as a
plasma tumor marker in patients with HNSCC and gas-
trointestinal cancer were recently given by us.21,22 We
found significantly higher GSTP1–1 plasma levels in pa-
tients with HNSCC patients as compared with a group of
patients with benign head and neck lesions as control
subjects. However, only 14% of patients with HNSCC had
plasma levels greater than the upper normal reference
limit of 21.8 ␮g/l, which was estimated after analyzing
samples of 230 healthy blood donors. In the past, several
investigators have correlated GSTP1 genotype and
GSTP1–1 tissue levels (e.g., in lung malignancies23 in
human placenta24 and in Barrett’s esophagus13). They all
found significant differences between the individual geno-
types GSTP1 AA, AB, and BB. However, an association
between the GSTP1 genotype and GSTP1–1 plasma levels
in patients with head and neck malignancies has never
been studied.
Blood Sampling
Approximately 3 mL of blood was collected by venapuncture
in vacuumed, silicone-coated, K3 ethylenediamine tetra-acetic
acid (EDTA–containing tubes (Vacutainer, Order No. 606601,
Becton Dickinson, Grenoble, France). Immediately after collec-
tion, blood was gently mixed with the EDTA solution. Approxi-
mately 0.5 mL. of whole blood was stored at Ϫ20°C for genotyp-
ing. The remaining amount was centrifuged for 10 minutes at
3000g at room temperature within 2 hours. The upper two-thirds
of the plasma was collected, and care was taken not to aspirate
the platelets on top of the cell layer. Plasma samples were stored
at Ϫ20°C until use.
Assays
Glutathione S-transferase P1 genotyping. Genomic
DNA was isolated from whole blood using the Wizard genomic
DNA purification kit (Promega, Madison, WI) according to the
instructions of the manufacturer.
Genetic polymorphisms in GSTP1 were detected by poly-
merase chain reaction–restriction fragment length polymorphism
analysis, using a primer set (105F and 105R) that was designed
so that the presence of the rare B allele (valine) resulted in the
appearance of an ALw261 restriction enzyme site according to
Harries et al.10 All primers were synthesized by Pharmacia Bio-
tech (Roosendaal, The Netherlands). All chemicals needed for
polymerase chain reaction were purchased from Promega.
Plasma glutathione S-transferase P1–1 measurement.
Plasma GSTP1–1 levels were measured using a recently devel-
oped enzyme-linked immunosorbent assay.22 Briefly, microtiter
plates were coated overnight with purified anti–GSTP1–1 mono-
clonal antibody. One hundred microliters of standard (0.4–100
␮g/l GSTP1–1) or 1:1 diluted plasma samples were added to the
wells and incubated overnight. The next day, the plates were
washed and incubated with rabbit anti–GSTP1–1 serum, washed
again, and incubated with horseradish peroxidase–labeled swine
anti-rabbit antibody; after a final wash, plates were stained with
O-phenylenediamine-H202. Standards and samples were mea-
sured in duplicate. A four-parameter weighted logistic regression
model was used to calculate standard curves and unknown vari-
ables. The assay had no cross-reactivity with GST enzymes of the
alpha or mu classes; the detection limit was 0.4 ␮g/l GSTP1–1,
and the intra-assay and interassay coefficients of variation were
5.8% and 10.9%, respectively.
Because GST enzyme activity may be associated with
GSTP1 genotype,13,23,24 we wanted to investigate whether
patients with elevated GSTP1–1 plasma levels all belong
to a certain GSTP1 genotype. Therefore, we investigated
whether one of the genotype subcategories differed signif-
icantly in mean GSTP1–1 plasma level. We also compared
the mean plasma levels in 51 control subjects and 87
patients with HNSCC in each of the three genotype sub-
categories (GSTP1 AA, AB, and BB).
TABLE I.
Site of Tumor Origin in HNSCC.
PATIENTS AND METHODS
Site of Tumor Origin
No. of Cases
In total, 138 patients were included preoperatively during
their scheduled admission at the Department of Otorhinolaryn-
gology—Head and Neck Surgery of the Academic Hospital, Uni-
versity of Maastricht (Maastricht, The Netherlands), between
1995 and 1998 (Table I). The group consisted of 51 patients with
benign head and neck lesions who served as control subjects (23
female patients [45%] and 28 male patients [55%]) and 87 pa-
Oral cavity
Oropharynx
Hypopharynx
Larynx
25
23
12
27
HNSCC ϭ squamous cell carcinoma of the head and neck region.
Laryngoscope 112: March 2002
Kelders et al.: Glutathione S-Transferase P1 Genotype
463

TABLE II.
Plasma GSTP1-1 Levels.
Group
Genotype
Mean*
N
SD*
Minimum*
Maximum*
Median*
Control
AA
AB
9.1
9.9
26
18
7
3.2
6.1
5.1
4.2
3.3
3.3
4.6
3.7
9.3
3.7
4.6
3.7
3.3
3.3
17.9
32.3
8.3
8.2
BB
9.3
4.5
15.9
7.4
Total
AA
9.4
51
36
38
13
87
62
56
20
138
4.5
32.3
8.2
HNSCC
Total
12.0
19.8
15.0
15.9
10.8
16.6
13.0
13.5
5.1
26.8
10.7
11.8
14.5
11.2
9.5
AB
26.1
5.3
141.5
26.8
BB
Total
AA
17.9
4.6
141.5
26.8
AB
22.1
5.6
141.5
26.8
10.0
12.8
10.0
BB
Total
14.8
141.5
*Values are given as ␮g/L.
N ϭ number of subjects in each category; SD ϭ standard deviation.
Statistical Analysis
Table II shows the mean GSTP1–1 plasma levels
according to GSTP1 genotype in the two study groups
(patients with HNSCC and control subjects). For all
GSTP1 genotypes (AA, AB, and BB) in HNSCC, the mean
and median values were higher as compared with corre-
sponding control subjects. The highest mean values were
observed in the heterozygotic GSTP1 AB groups in both
study groups.
Table III shows the results of the independent-
samples t tests comparing the mean GSTP1–1 plasma
values in each genotype subgroup. When comparing the
HNSCC to the control group for the AA genotype, plasma
GSTP1–1 values were significantly higher in the HNSCC
group (12.0 and 9.1 ␮g/L, respectively; P ϭ .008). A similar
result was obtained for the AB genotype, even when cor-
rected for one abnormal, high value that appeared unre-
liable according to Cook’s test (19.8 vs. 9.9 ␮g/L, P ϭ .031).
The GSTP1 BB genotype also showed a difference that
was significant (15.0 vs. 9.3 ␮g/L, P ϭ .028). The difference
in GSTP1–1 plasma values for the combined control and
HNSCC groups was significant as well (9.4 vs. 15.9 ␮g/L,
P ϭ .002).
For statistical analysis we used the statistical package
SPSS for Windows 8.0. The statistical significance of differences
between the individual means for the control and HNSCC groups
for each genotype was tested with an independent t test for
equality of means. Possible unreliable variables were detected by
Cook’s test.
The significance of the difference of means for the different
genotypes in each category of disease (benign or malignant) and
the total group (control patients and HNSCC patients) was tested
with a one-way analysis of variances, combined with a Levene
test for equality of variances and completed with Bonferroni and
Tamhane post hoc tests.
RESULTS
The results of the assays and different comparative
tests are shown in Tables II–IV.
The distribution of genotypes in the entire study
group (patients with HNSCC and control subjects) was 62
individuals (44.9%) with the GSTP1 AA genotype, 56
(40.6%) with the AB genotype, and 20 (14.5%) with the BB
genotype. For the control group the distribution was 26
(51.0%) for GSTP1 AA, 18 (35.3%) for AB, and 7 BB
(13.7%). For the HNSCC group the distribution was 36
(41.4%) for GSTP1 AA, 38 (43.7%) for AB, and 13 (14.9%)
for the BB genotype.
Table IV shows the results of one-way analyses of
variance comparing the mean GSTP1–1 plasma concen-
trations for the different genotypes in each study group.
The difference between the mean plasma GSTP1–1 levels
for each GSTP1 genotype in the control group was not
TABLE III.
T-tests.
TABLE IV.
One-way Analysis of Variances.
t-test Control vs.
HNSCC (P values)
GSTP1 Genotype
AA
.008
.031
.035
.028
.002
ANOVA Significance
Group
(P values)
AB
AB, exclusion of highest value
Control
HNSCC
Total
.862
.170
.100
BB
Total
Difference in GSTP 1-1 plasma level between control and HNSCC for
the different genotypes.
Differences in mean GSTP1-1 plasma level between the genotypes in
control, HNSCC and both.
Laryngoscope 112: March 2002
464
Kelders et al.: Glutathione S-Transferase P1 Genotype

significant (9.1 [AA], 9.9 [AB], and 9.3 [BB] ␮g/l, respec-
tively; P ϭ .862), nor was it significant for the HNSCC
group (12.0 [AA], 19.8 [AB], and 15.0 [BB] ␮g/L; P ϭ .170)
or for the total of both groups (both control subjects and
patients with HNSCC) (10.8, 16.6, and 13.0 ␮g/L; P ϭ
.100).
Dutch patient groups that were investigated have a dif-
ferent distribution of GSTP1 genotypes. The GSTP1 fre-
quency distribution in the Dutch population is 55% for
GSTP1 AA, 35% for GSTP1 AB, and 10% for GSTP1 BB
genotypes. Hirata et al.16 gave no data on the genotype
distribution, but another Japanese study by Morita et al.6
indicated a frequency distribution of 69% for GSTP1 AA,
29% for the AB, and 2% for BB genotypes, which, indeed,
differs considerably as compared with corresponding data
in our Dutch population.
DISCUSSION
The purpose of the current study was to evaluate the
possible association between the genetic polymorphisms
in GSTP1 and the phenotypic expression (GSTP1–1
plasma levels) in patients with malignant head and neck
tumors.
CONCLUSION
There is no clear association between GSTP1 geno-
type and GSTP1–1 phenotypic expression in the form of
plasma levels in patients with head and neck lesions as
was found previously for GST enzyme activity in tissues of
patients with lung cancer23 or Barrett’s esophagus.13 Per-
haps, further investigation of the corresponding GSTP1–1
tissue enzyme activity in patients with head and neck
disease will reveal such a correlation.
No previous studies have been performed on the cor-
relation between GSTP1 genotype and GSTP1–1 plasma
level. However, a correlation between enzyme function
and GSTP1 genotype has been made.8,23,24 Ali-Osman et
al.8 and Zimniak et al.24 have provided evidence of poly-
morphisms in the human GSTP1 gene locus, resulting in
functionally different GSTP1–1 proteins.8,24 Watson et
al.23 and Van Lieshout et al.13 found significant differ-
ences in GST conjugating activity between the AA wild-
type genotype and the AB and BB genotypes in lung and
esophageal tissue, respectively.
We found no significant correlation between the dif-
ferent GSTP1 genotypes and the corresponding GSTP1–1
plasma levels. This was the case when the control and
HNSCC groups were investigated separately and also
when they were combined. This means that both in pa-
tients with HNSCC and in control subjects, no direct re-
lationship was present between GSTP1 genotype and phe-
notype, as was found earlier in lung or esophageal tissue
of patients with corresponding lung or esophageal
lesions.13,23
In a previous study on HNSCC, we found that
GSTP1–1 is not a plasma tumor marker for individual
patients, despite the fact that median plasma levels in
HNSCC were significantly higher as compared with con-
trol subjects, but that only a small number of patients
with HNSCC had plasma levels elevated above the normal
reference level of 21.8 ␮g/l.21
When comparing patients with HNSCC and control
subjects, significantly different plasma levels were found
for all GSTP1 genotype categories. These findings show
that there is no difference in the influence of the genotypes
on GSTP1–1 plasma levels between patients with HNSCC
and control subjects. The control group had a slightly
higher percentage of GSTP1 AA genotypes than the
HNSCC group (51.0% vs. 41.4%) and a lower percentage of
GSTP1 AB genotypes (35.3% vs. 43.7%). The clinical sig-
nificance of this difference must be investigated in a larger
patient group.
BIBLIOGRAPHY
1. Ryberg D, Skaug V, Hewer A, et al. Genotypes of glutathione
transferase M1 and P1 and their significance for lung DNA
adduct levels and cancer risk. Carcinogenesis 1997;18:
1285–1289.
2. Hayes JD, Pulford DJ. The glutathione S-transferase super-
gene family: regulation of GST and the contribution of the
isoenzymes to cancer chemoprotection and drug resistance.
Crit Rev Biochem Mol Biol 1995;30:445–600.
3. Mulder TP, Manni JJ, Roelofs HM, Peters WHM, Wiersma A.
Glutathione S-transferases and glutathione in human
head and neck cancer. Carcinogenesis 1995;16:619–624.
4. Matthias C, Bockmuhl U, Jahnke V, et al. The glutathione
S-transferase GSTP1 polymorphism: effects on susceptibil-
ity to oral/pharyngeal and laryngeal carcinomas. Pharma-
cogenetics 1998;8:1–6.
5. Jahnke V, Matthias C, Bockmühl U, Strange RC. Genetische
Pra¨disposition fu¨r die Entstehung von Kopf- und Hals-
karzinomen. Laryngorhinootologie 1999;78:24–27.
6. Morita S, Yano M, Tsujinaka T, et al. Genetic polymorphisms
of drug-metabolizing enzymes and susceptibility to head-
and-neck squamous-cell carcinoma. Int J Cancer 1999;80:
685–688.
7. Jourenkova-Mironova N, Voho A, Bouchardy C, et al. Gluta-
thione S-transferase GSTM3 and GSTP1 genotypes and
larynx cancer risk. Cancer Epidemiol Biomarkers Prev
1999;8:185–188.
8. Ali-Osman F, Akande O, Antoun G, Mao JX, Buolamwini J.
Molecular cloning, characterization, and expression in
Escherichia coli of full-length cDNAs of three human glu-
tathione S-transferase Pi gene variants: evidence for dif-
ferential catalytic activity of the encoded proteins. J Biol
Chem 1997;272:1004–1012.
9. Saarikoski ST, Voho A, Reinikainen M, et al. Combined effect
of polymorphic GST genes on individual susceptibility to
lung cancer. Int J Cancer 1998;77:516–521.
10. Harries LW, Stubbins MJ, Forman D, Howard GC, Wolf CR.
Identification of genetic polymorphisms at the glutathione
S-transferase Pi locus and association with susceptibility
to bladder, testicular and prostate cancer. Carcinogenesis
1997;18:641–644.
Hirata et al.16 considered estimation of GSTP1–1
plasma levels a useful aid for early diagnosis, prediction of
tumor extent, and determination of treatment efficacy and
prognosis for oral cancer. We did not find arguments for
such a role of GSTP1–1 for our patients with HNSCC.21
The question is whether this difference can be explained
by either racial or demographic factors or by the fact that
Hirata et al.16 limited their investigation to oral cancers.
Another explanation could be that the Japanese and
11. Morita S, Yano M, Shiozaki H, et al. CYP1A1, CYP2E1 and
GSTM1 polymorphisms are not associated with suscepti-
bility to squamous-cell carcinoma of the esophagus. Int J
Cancer 1997;71:192–195.
12. Lin DX, Tang YM, Peng Q, Lu SX, Ambrosone CB, Kadlubar
FF. Susceptibility to esophageal cancer and genetic poly-
morphisms in glutathione S-transferases T1, P1, and M1
Laryngoscope 112: March 2002
Kelders et al.: Glutathione S-Transferase P1 Genotype
465

and cytochrome P450 2E1. Cancer Epidemiol Biomarkers
Prev 1998;7:1013–1018.
13. van Lieshout EM, Roelofs HM, Dekker S, et al. Polymorphic
expression of the glutathione S-transferase P1 gene and its
susceptibility to Barrett’s esophagus and esophageal car-
cinoma. Cancer Res 1999;59:586–589.
14. Helzlsouer KJ, Selmin O, Huang HY, et al. Association be-
tween glutathione S-transferase M1, P1, and T1 genetic
polymorphisms and development of breast cancer. J Natl
Cancer Inst 1998;90:512–518.
15. Tsuchida S, Sato K. Glutahione transferases and cancer. Crit
Rev Biochem Mol Biol 1992;27:337–384.
of regulatory sequences near the pi-class glutathione
S-transferase gene accompanies human prostatic carcino-
genesis. Proc Natl Acad Sci U S A 1994;91:11733–11737.
20. Veneroni S, Silvestrini R, Costa A, Salvatori P, Faranda A,
Molinari R. Biological indicators of survival in patients
treated by surgery for squamous cell carcinoma of the oral
cavity and oropharynx. Oral Oncol 1997;33:408–413.
21. Oude Ophuis MB, Mulder TP, Peters WHM, Manni JJ.
Plasma glutathione S-transferase P1–1 levels in patients
with head and neck squamous cell carcinoma. Cancer 1998;
82:2434–2438.
22. Mulder TP, Peters WHM, Wobbes T, Witteman BJ, Jansen
JB. Measurement of glutathione S-transferase P1–1 in
plasma: pitfalls and significance of screening and follow-up
of patients with gastrointestinal carcinoma. Cancer 1997;
80:873–880.
16. Hirata S, Odajima T, Kohama G, Ishigaki S, Niitsu Y. Sig-
nificance of glutathione S-transferase-pi as a tumor
marker in patients with oral cancer. Cancer 1992;70:
2381–2387.
17. Niitsu Y, Takahashi Y, Saito T, et al. Serum glutathione-S-
transferase-pi as a tumor marker for gastrointestinal ma-
lignancies. Cancer 1989;63:317–323.
23. Watson MA, Stewart RK, Smith GB, Massey TE, Bell DA.
Human glutathione S-transferase P1 polymorphisms: re-
lationship to lung tissue enzyme activity and population
frequency distribution. Carcinogenesis 1998;19:275–280.
24. Zimniak P, Nanduri B, Pikula S, et al. Naturally occurring
human glutathione S-transferase GSTP1–1 isoforms with
isoleucine and valine in position 104 differ in enzymic
properties. Eur J Biochem 1994;224:893–899.
18. Hida T, Kuwabara M, Ariyoshi Y, et al. Serum glutathione
S-transferase-pi level as a tumor marker for non-small cell
lung cancer. Potential predictive value in chemotherapeu-
tic response. Cancer 1994;73:1377–1382.
19. Lee WH, Morton RA, Epstein JI, et al. Cytidine methylation
Laryngoscope 112: March 2002
466
Kelders et al.: Glutathione S-Transferase P1 Genotype