Acta Oto-Laryngologica

ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20

Epidermal Growth Factor Receptor, c-erbB-2 and

p53 Overexpressions in Epithelial Hyperplastic
Lesions of the Larynx

Nina Gale, Nina Zidar, Vinko Kambič, Mario Poljak & Andrej Cör

To cite this article: Nina Gale, Nina Zidar, Vinko Kambič, Mario Poljak & Andrej Cör (1997)
Epidermal Growth Factor Receptor, c-erbB-2 and p53 Overexpressions in Epithelial
Hyperplastic Lesions of the Larynx, Acta Oto-Laryngologica, 117:sup527, 105-110, DOI:
10.3109/00016489709124048

To link to this article: http://dx.doi.org/10.3109/00016489709124048

Published online: 08 Jul 2009.

Submit your article to this journal

Article views: 11

View related articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ioto20

Download by: [University of Newcastle] Date: 16 March 2016, At: 07:08

 Acta Otolaryngol (Stockh) 1997; Suppl 527: 105- 110

Epidermal Growth Factor Receptor, c-erbB-2 and p53 Overexpressions in

Epithelial Hyperplastic Lesions of the Larynx
NINA GALE,' NINA ZIDAR,' VINKO KAMBIC,2 MARIO POLJAK3 and ANDREJ C 0 R 4
From ilie ' Iirstiiuic, of Ptrtliology. Mediccil F u c u l t ~ ~
'Sloueriicrri
, o f Sciences LIIIN' Arts, 'Institute of' Microbiology and
Actideni~~ Immunology.
Medical Fuculiy. "hstirute of' Histology ant1 Enihrvology, Medictrl Fuciilty, Uniuersiry of. ~j~rhljcrnu,
Slouenitr

Gale N, Zidar N, Karnbit V, Poljak M, Cor A. Epideri?inl groit'ili /kctor receptor. c-erhB-2 om/ p5.1 ocercspre.s.si~ins in
epiilirlicrl liyperplusiic lesions of' the I a r y i ~ . Acta
~. Otolaryngol (Stockh) 1997; Suppl 527: 105- 110.
An immunohistochemical analysis of overexpression of epidermal growth factor receptor (EGFR), c-erhB-2, and p53
proteins was performed on 43 biopsies of laryngeal epithelial hyperplastic lesions (EHLL), classified according to the
KambiE-Lenart classification, and in I 1 cases of laryngeal carcinoma (SCCL). The aim of the present study was to
determine whether there is a correlation between the staining patterns of these proteins and different grades of EHLL, and
to reveal their possible prognostic value. We compared the staining patterns of atypical hyperplasia adjacent to cancer
Downloaded by [University of Newcastle] at 07:08 16 March 2016

with the same type of lesions which have not turned malignant. p53 and EGFR overexpressions were detected in 28/54
(52%~)and 33/54 cases (6I'X)) respectively, and tend to increase with the degree of epithelial changes. The intensity of
staining in various grades of EHLL adjacent to cancer was more pronounced than the same type or lesions which have
not progressed to cancer. c-erhB-2 was weakly positive in the majority of cases, and changed from predominantly
membranous in simple hyperplasia to cytoplasmic staining in abnormal and atypical hyperplasias. There was no
significant statistic correlation between the amount of positive cells for all proteins and the grade of epithelial
abnormalities. We conclude that the overexpression of each biomarker itself adds little predictive value over routine
histomorphology, and cannot be regarded as a reliable prognostic factor for EH LL. However. the histologic characteris-
tics of atypical hyperplasia together with the immunostaining patterns of EGFR and p53 up to two-thirds or more of the
epithelial thickness could be considered a reliable pattern which correlates with the progression to cancer. Kev words:
epitheliul hyperpicistic lesions, Icirynr, bnniiinohistotlien~i.~t~~,
EGFR. c-erhB-2 protein, p53 protrin.

INTRODUCTION tional data are available concerning overexpression of

Epithelial hyperplastic lesions of the larynx (EHLL)
with the clinical appearance of chronic laryngitis
represent a broad spectrum of histomorphologic
changes. These range from entirely benign simple hy-
perplasia (squamous cell hyperplasia) to atypical hy-
perplasia or risky epithelium (severe dysplasia), which
is a condition associated with an increased probabil-
ity for subsequent cancer (1). Multiple genetic abnor-
malities are believed to occur during the tumor-
igenesis of the head and neck squamous cell cancer
(SCCHN). Recognition of these early events might be
considered an aid to predicting the risk of tumor
development ( 2 ) . Determining which of the genetic
changes appearing in EHLL, particularly in the risky
epithelium, might correlate with particular clinical
and histomorphologic abnormalities, could explain
more precisely the critical events in the multistep
carcinogenesis in the respiratory tract (3).
The epidermal growth factor receptor (EGFR)
gene, located on chromosome 7~12-13,codes for a
170 kDa transmembrane growth-regulating receptor
glycoprotein, which has an intrinsic tyrosine-specific
kinase activity that regulates cell growth in cell lines
and in a variety of cancers (4, 5). Overexpression of
EGFR is associated with alterations on chromosome
7, and amplification or rearrangement of the EGFR
gene (4, 6). These genetic changes have been reported
in several human tumors, including SCCHN. Addi-
EGFR in premalignant lesions in the head and neck
region, mainly in the larynx (2, 4, 7). These authors
share the same opinion that overexpression of EGFR
could be a useful biomarker for predicting the risk of
cancer development.
The oncogene c-erbB-2, located on chromosome
17, encodes a 185 kDa transmembrane protein which
expresses a significant sequence homology with
EGFR (8). The oncogene potential of c-erbB-2 can
be released by point mutations, truncation of extra-
cellular and cytoplasmic domains, and amplification
and/or overexpression of the normal gene (9). Over-
expression and amplification of c-erbB-2 has been-
mainly observed in different adenocarcinomas ( I 0)
and less frequently in SCCHN. (6). Some authors,
in contrast, have observed overexpression of c-erbB-2
in 46-60'%) SCCHN (1 1, 12). It is, therefore, still
controversial whether c-erbB-2 amplification is a
common event in precancerous lesions and SCCHN
( 1 3).
The prevalence of p53 protein overexpression con-
cerning benign and premalignant lesions of the SC-
CHN, particularly of the laryngeal epithelium, has
been extensively studied in the last few years (14- 17).
It might be assumed that the appearance of the
mutant p53 protein in different laryngeal epithelial
lesions could help in understanding cell transforma-
tion in the process of carcinogenesis.

0 1997 Scandinavian University Press. ISSN 0365-5237

 106 N . Gule et al.
Considering the frequent overexpression of EGFR
and p53 proteins and the controversial role of c-
erbB-2 in the tumorigenesis of SCCHN, we intended
to analyze the frequency of overexpression of these
genes in EHLL and to determine their relationship to
the histologic grading of the lesions, as well as to
explore the potential clinical benefit of the particular
immunohistochemical staining patterns. We particu-
larly wished to compare the staining pattern of atypi-
cal hyperplasia adjacent to cancer with the same type
of lesions which have not turned malignant.
MATERIAL AND METHODS
A retrospective, non-random study included 54
Downloaded by [University of Newcastle] at 07:08 16 March 2016
biopsy samples of EHLL from 34 patients. There
were 6 women and 28 men, with mean age 50 years,
ranging from 15 to 80 years. All patients had at least
4 years clinical follow-up. Benign EHLL were graded
according to the KambiE-Lenart classification into 3
groups (1, 18): 5 cases of simple (SH), 20 of abnormal
(AbH), and 18 of atypical hyperplasias (AtH); 11
samples of the squamous cell carcinoma of the larynx
(SCCL) were additionally selected because they
demonstrated all types of EHLL adjacent to invasive
cancer in the same section.
The control group consisted of autopsy samples of
the laryngeal mucous membrane of 5 non-smokers,
aged 20-30 years, who died accidentally and had no
clinical or histologic evidence of laryngeal disease.
Immediately after excision, the specimens were
fixed in 10% neutral formalin, embedded in paraffin,
cut at 4-5 p m and stained with hematoxylin and
eosin for traditional light microscopy. For immuno-
histochemistry, tissue slides were deparaffinized in
xylene, rehydrated through graded alcohols, and sub-
jected to proteinase K digestion, except for EGFR.
Subsequently, the sections for c-erhB-2 and p53
staining were pretreated in a microwave oven twice
for 5 min in citrate buffer.
All samples were immunostained using the follow-
ing primary antibodies: c-erbB-2 (rabbit polyclonal,
1 : 50, Dako, Glostrup, Denmark), EGFR (mouse
monoclonal, 1 : 5, Zymed Laboratories Inc, San Fran-
cisco, USA), p53 PAb 1801 (mouse monoclonal,
1 : 50, Oncogene Science, Uniondale, USA). Immuno-
peroxidase detection was employed using the standard
avidin-streptavidin method (ABC) (Dako, Glostrup,
Denmark) and diaminobenzidin as substrate.
The specificity of reaction was determined using
nonimmune serum instead of the primary antibody as
negative controls, SCCL exhibiting overexpression of
EGFR and p53, and a breast cancer showing c-erhB-
2 as positive staining.
The immunostaining signals were evaluated semi-
quantitatively. The EGFR immunoreactivity was
assessed on the cell membranes while c-erbB-2 im-
munoreactivity was interpreted as membranous or
cytoplasmic staining. Both staining patterns were
scored as follows: 0, no staining; 1, staining of basal
cells and suprabasal region up to one third of the
epithelial thickness; 2, staining of two thirds of the
epithelial thickness; 3, staining of the whole epithelial
thickness. The intensity of nuclear staining (weak,
moderate, strong) with p53 antibody, and the distri-
bution within the epithelium was scored as follows: 0,
no staining; 1, focal staining of basal cells, < 10% of
positive cells cells; 2, staining of basal cells and
suprabasal region up to the midportion of the epithe-
lial thickness, 11 to 50% of cells; 3, of the whole
epithelial thickness, > 50% of cells. The same scoring
relating to the percentage of the positive cells was
used for all carcinomas.
Stutistics
Quantitative data, such as histomorphologic parame-
ters of EHLL and immunostaining appearance were
analysed by the Pearson x -squared statistical test.
Values of p < 0.05 were considered to be statistically
significant.
RESULTS
Immunohistochemistry showed that 33/54 (61%)
cases were positive for EGFR (Fig. lu), among them
27/43 (63%) lesions of EHLL and 6/11 (55'Yn) of
SCCL. c-erbB-2 staining was seen in 24 (44%) le-
sions, 18/43 (42%) of EHLL and 6/11 (55'Yn) of SCCL
(Fig. 16). Overexpression of p53 protein was found in
28 of 54 (52%) cases, among them in 21/43 (49%)
EHLL and 5/11 (45%))SCCL (Fig. Ic). The intensity
of c-erbB-2 and EGFR signals, membranous and
cytoplasmic, appeared to increase with the grade of
epithelial abnormalities, as well as the p53 immunos-
taining, which was strictly confined to the epithelial
nuclei. In addition to the positivity of the basal layer
and the lower third of the epithelium in simple and
abnormal hyperplasia, staining in atypical hyper-
plasia was notable, particularly in the lower two
thirds of the epithelium or in its whole thickness.
c-erbB-2 immunoreactivity, which was mainly
weakly positive and not very strictly confined to
basal or augmented basal type cells, changed from
predominantly membranous in simple hyperplasia
to cytoplasmic staining with the increasing grade of
epithelial abnormalities. Thus, in SCCL positive
cases c-erhB-2 showed predominantly cytoplasmic
staining. Although a trend of increasing positivity
relating to the degree of abnormalities was found,

70 ,
6o
" SH AbH AtH Ca
100 ~~ ~~~
I
~
Downloaded by [University of Newcastle] at 07:08 16 March 2016
" SH AbH AtH Ca
Fig. l ( u ) Frequency ((%I) of EGFR positivity in different
grades of EHLL. ( b ) Frequency ('%I) of c-erbB-2 positivity
in different grades of EHLL. Legend for ( u ) and ( b ) :0, no
staining; I , staining of basal cells and suprabasal region up
to one third of the epithelial thickness; 2, staining of the
two thirds of the epithelial thickness; 3, staining of the
whole epithelial thickness. Legend for Ca: 0, no staining, I ,
focal staining of < 10% of cells; 2, staining of 11-50'%1of
cells; 3, staining of > 50'%,of cells. (c) Frequency (%) of p53
positivity in different grades of EHLL and Ca. Legend for
(c): 0, no staining, 1, focal staining of basal cells, < 10'%1of
cells; 2, staining of basal cells and suprabasal region up to
the midportion of the epithelial thickness, 11 to 50% of
cells; 3, staining of the whole epithelial thickness, > 50% of
cells.
the levels of EGFR, c-erbB-2 and p53 immunostain-
ing did not statistically correlate with the grade of
epithelial changes. In all cases of SCCL, an adjacent
EGFR, c-erbB-2 a n d p 5 3 in the lurynx 107
Fig. 2. Immunohistochemical staining of EGFR in atypical
hyperplasia without progression to cancer (up) and the
same type of lesion adjacent to invasive cancer (down).
Note the increased intensity of staining (down). StreptABC,
immunoperoxidase, hematoxylin counterstain, x 120.
transitional area from SH to AtH and carcinoma
in situ was found. In these cases, the strongly positive
foci of p53 and EGFR in atypical hyperplasia
adjacent to cancer were interspersed with inconspicu-
ous areas of simple hyperplasia. The intensity of
staining in various grades of EHLL adjacent to can-
cers was more pronounced than the same type of
lesions which have not progressed to cancer (Figs. 2
and 3). Such a difference was not observed in c-erbB-
2 staining.
DISCUSSION
The results of the present study are in accordance
with the findings of other recent studies showing that
p53 protein overexpression, one of the most frequent
genetic events in tumorigenesis of SCCL occurs in
early EHLL in the head and neck region even before
histomorphologic abnormalities are seen, and that
their intensity and frequency increase according to
the severity of the epithelial changes (1, 2, 15-17, 19,
20). The current findings are similar to those pre-
sented in our previous work (20). Namely, the p53
 108 N . Gale rt a1
Downloaded by [University of Newcastle] at 07:08 16 March 2016
immunostaining pattern was partially focal and to a
limited extent diffuse and was confined to a basal and
proliferative basal type in SH and AbH. In all posi-
tive cases (39°K) of AtH, a focal expression of p53
protein was seen up to two third or the whole epithe-
lial thickness. This finding might suggest that only a
subclone of epithelial cells exhibits the p53 gene
Fig. 3. Immunohistochemical staining
of p53 in atypical hyperplasia without
progression to cancer (up) and the
same type of lesion adjacent to inva-
sive cancer (down). Note the increased
intensity of staining (down). Strept-
ABC, immunoperoxidase, hematoxylin
counterstain, x 49.5, x 31.5, respec-
tively.
alteration in precancerous lesions. On the other hand,
overexpression of the p53 protein in areas of car-
cinoma in situ and atypical hyperplasia adjacent to
invasive cancer was identified as strong, diffuse
within two thirds or the whole epithelial thickness.
These results support the concept of field canceriza-
tion proposed by Shin et al. (15), whereby the whole

-
epithelium, which is continuously exposed to mito-
gens, accumulates genetic injuries and is disposed at
high risk for developing multiple lesions which may
turn to overt cancer. However, when analyzing the
follow-up of the patients included in this non-random
study, we can conclude, as others did, that overex-
pression of p53 protein itself may not be considered a
reliable prognostic factor determining the risk of
cancer development.
The pattern of EGFR expression was analogous
with p53 immunostaining in all types of EHLL and
SCCL. It was ascertained in an even higher percent-
age (61%) than p53 protein expression (52%). The
present findings indicate that expression of EGFR
progressively increases with the grade of EHLL, as
described also by some other authors (2-4). Shin et
Downloaded by [University of Newcastle] at 07:08 16 March 2016
al. (4) presented 2 stages in EGFR expression in
carcinogenesis of SCCHN: the first increase occurred
with the transition from control epithelium of non-
smokers to normal epithelium adjacent to carcinoma,
and the second within the transition of dysplastic
lesions to SCCHN. EGFR increase adjacent to inva-
sive SCCL was also constantly seen in our series, as
well as a higher polysomies of chromosome 7 com-
pared to the results of the same type of lesions which
have not progressed to cancer. The polysomies were
demonstrated by nonisotopic in situ hybridization
(data not presented) using chromosome-specific cen-
tromeric DNA probes on the same tissue samples.
According to the results of our study, we share the
opinion of Uhlman et al. (2) that immunohistochemi-
cal EGFR expression is a rather common event,
which strongly correlates with atypias of EHLL. Al-
though dysregulation of the EGFR gene is clearly
involved in carcinogenesis of SCCL, its specificity
and predictive role as an independent biomarker is
still of unpredictable value.
In our series, the entire expression of c-erbB-2 was
weak in most cases. Although the c-erbB-2 protein
shares a significant sequence homology and similar
gross structural organization with EGFR, we have
not found a clear correlation between their overex-
pressions. In agreement with other studies, c-erbB-2
protein overexpression is seen in the process of can-
cerogenesis of SCCHN (1 1, 12); we found it in 42%
and 55% of EHLL and cancer, respectively. The
c-erbB-2 immunoreactivity pattern tended to change
from predominantly membranous in simple hyper-
plasia to cytoplasmic staining with an increasing
grade of epithelial abnormalities, although these
staining patterns were not quite constant. There is no
general agreement in the literature as to whether both
staining patterns should be considered positive. How-
ever, as immuno-electronmicroscopy studies have lo-
calized c-erbB-2 to the plasma membrane, many
EGFR, c-erbB-2 undp5-3 in the Iurynx 109
investigators believe that only the membranous stain-
ing pattern is associated with clear cut positivity and
could, therefore, be crucial for the interpretation of
the c-erbB-2 expression. Here, we also failed to
demonstrate either an exact correlation with the
severity of epithelial abnormalities or with the course
of the disease, since the immunostaining expression
was always weak.
We conclude that EGFR and p53 overexpressions
are fairly common in EHLL and SCCL and strongly
correlate with the grade of epithelial abnormalities.
The intensity of staining pattern, especially in atypi-
cal hyperplasia, is more constant and pronounced
adjacent to cancer than the same type of lesions
which have not turned malignant. The overexpression
of each biomarker itself, e.g. EGFR, p53 and c-erbB-
2 adds little clinical predictive value over routine
histomorphology and cannot be considered reliable
prognostic factors for EHLL. Our results strongly
suggest that the histologic characteristics of atypical
hyperplasia together with the immunostaining pat-
terns of EGFR and p53 up to the two thirds or more
of the epithelial thickness and detection of
polysomies of chromosomes 7 and 17 (data not pre-
sented) could be considered a reliable pattern which
correlates with the progression to cancer.
REFERENCES
1. Kambii: V, Gale N. Epithelial hyperplastic lesions of
the larynx. Amsterdam: Elsevier, 1995; 1-265.
2. Uhlman DL, Adams G, Knapp D, Aeppli DM, Nie-
hans G . Immunohistochemical staining for markers of
future neoplastic progression in the larynx. Cancer Res
1996; 56: 2199-205.
3. Rusch V. Klimstra D, Linkov I, Dmitrovsky E. Aber-
rant expression of p53 or the epidermal growth factor
receptor is frequent in early bronchial neoplasia, and
coexpression precedes squamous cell carcinoma devel-
opment. Cancer Res 1995; 55: 1365-72.
4. Shin DM, R o JY, Hong WK, Hittelman WN. Dys-
regulation of epidermal growth factor receptor expres-
sion in premalignant lesions during head and neck
tumorigenesis. Cancer Res 1994; 54: 3 153-9.
5 . Wagner U, Sauter G, Moch H , et al. Patterns of p53,
erbB-2, and EGF-r expression in premalignant lesions
of the urinary bladder. Hum Pathol 1995; 26: 970-8.
6. Kearsley JH, Leonard JH, Walsh MD, Wright GR. A
comparison of epidermal growth factor receptor
(EGFR) and c-erbB-2 oncogene expression in head and
neck squamous cell carcinomas. Pathology 199I ; 23:
189-94.
7. Miyaguchi M , Olofsson J, Hellquist HB. Expression of
epidermal growth factor receptor in laryngeal dysplasia
and carcinoma. Acta Otolaryngol (Stockh) 1990; 110:
309- 13.
8. Underwood M, Bartlett J , Reeves J, Gdrdiner DS,
Scott R, Cooke T. c-erbB-2 gene amplification: a
molecular marker in recurrent bladder tumors? Cancer
Res 1995; 55: 2422-30.
 9. Bacus S, Zelnick CR, Plowman G, Yarden Y. Expres-
sion of the erbB-2 family of growth factor receptors
and their ligands in breast cancers. Am J Clin Pathol
1994; 102 (Suppl I ) : S13-24.
10. Gorgoulis VG, Barbatis C, Poulias I. Karameris AM.
Molecular and immunohistochemical evaluation of epi-
dermal growth factor receptor and c-evh-B-2 gene
product in transitional cell carcinomas of the urinary
bladder: a study in Greek patients. Mod Pathol 1995;
8: 758-64.
11. Craven JM, Pavelic ZP, Stambrook PJ, et al. Expres-
sion of c-erbB-2 gene in human head and neck car-
cinoma. Anticancer Res 1992; 12: 2273-6.
12. Field JK, Spandidos DA, Yiagnisis M , Gosney JR,
Papadimitriou K, Stell PM. c-erbB-2 Expression in
squamous cell carcinoma of the head and neck. An-
ticancer Res 1992: 12: 613-20.
13. Bron J, Monnier P. Molecular alterations in head and
neck squamous cell carcinoma. Clin Otolaryngol 1995;
Downloaded by [University of Newcastle] at 07:08 16 March 2016
20: 291-8.
14. Dolcetti R, Doglioni C, Maestro R, et al. p53 over-ex-
pression is an early event in the development of human
squamous-cell carcinoma of the larynx: genetic and
prognostic implications. Int J Cancer 1992; 52: 178-82.
15. Shin DM. Kim J, R o JY, et al. Activation of p53 gene
expression in premalignant lesions during head and
neck tumorigenesis. Cancer Res 1994: 54: 321-6.
16. Munck-Wikland E, Edstrom S, Jungmark E, Kuy-
lenstierna R, Lindholm J, Auer G. Nuclear DNA con-
tent, proliferating-cell nuclear antigen (PCNA) and p53
immunostaining in predicting progression of laryngeal
cancer in situ lesions. Int J Cancer 1994; 56: 95-9.
17. Pavelic ZP, Li YQ, Stambrook PJ, et al. Overexpres-
sion of p53 protein is common in premalignant head
and neck lesions. Anticancer Res 1994; 14: 2259-66.
18. Kambii? V, Gale N, Ferluga D. Laryngeal hyperplastic
lesions, follow-up study and application of lectins and
anticytokeratins for their evaluation. Pathol Res Pract
1992; 188: 1067-77.
19. Lavieille JP. Brainbilla E, Riva-Lavieille C, Reyt E,
Charachon R , Brambilla C. Immunohistochemica~de-
tection of p53 protein in preneoplastic lesions and
squainous cell carcinoma of the head and neck. Actd
Otolaryngol (Stockh) 1995; 115: 334-9.
20. Poljak M, Gale N. Kambii? V, Ferluga D , Fischinger J.
Overexpression of p53 protein in benign and malignant
laryngeal epithelial lesions. Anticancer Res 1996; 16:
1947- 52.
Address for correspondence:
Nina Gale, M.D.
Institute of Pathology
Korytkova 2
SI-1105 Ljubljana
Slovenia
E-mail: gale@ibmi.mf.uni-lj.si