OBSTn

European Journal of Obstetrics & Gynecology

GYNECOU
ELSEVIER and Reproductive Biology 63 (1995) 55 59

p53 tumor suppressor gene protein expression in cervical cancer:

relationship to prognosis

Gerard L. B r e m e r *~', A n t o n T.M.G. T i e b o s c h b ", H a n s W . H . M . v a n d e r P u t t e n ~, J e l t e d e

H a a n a,
Jan-Willem Arends b
~'Department (~['Obstetrics and Gynecoh)gy, Unit'ersiO, of Limburg, and UniversiO, Ho~q)ital Maastricht, The Netherlands
bDepartment o/ Pathology, UniversiO' (?]'L#77burg and b~li~'ersio' Hospital, Maastricht, The Netherlands

Received 13 April 1995: revision received 12 June 1995; accepted 24 July 1995

Abstract

Objective: Mutation of the p53 gene can be found in several human tumors. We tested the hypothesis whether overexpression
of p53 protein is a parameter of more aggressive disease in patients with cervical cancer. Study design: In this study, we describe
the effects of p53 overexpression in 156 patients with cervical cancer (Figo stage IB-IV) by assessing expression patterns of the
p53 gene product using a monoclonal anti-p53 antibody (DO7). Results: Overexpression of p53 tumor suppressor gene protein
was observed in 30.2'7,, of the tumors, low expression in 30.7'7,, and 39.1% of the tumors showed no p53 immunoreactivity. With
increase in stage, p53 overexpression raised from 20.1% in stage IB to 60% in stage IV. A significant correlation between p53
overexpression and disease-free survival of patients was observed, however, after stratification for stage, this effect disappeared.
Conclusions: The p53 mutation expressed as p53 tumor suppressor gene protein overexpression is a late event in cervical cancer
genesis and does not appear to be of prognostic significance in cervical cancer.

Keywords: Cervical Cancer; p53; Tumor suppressor gene; Immunohistochemistry; Survival

1. Introduction
T u m o r suppressor genes are able to suppress the
growth and metastasis of cells driven to uncontrolled
proliferation by oncogenes. The p53 tumor suppressor
gene is located on the short arm of chromosome 17 and
codes for a 375-amino-acid nuclear phosphoprotein
with a short half life time of several minutes [1}. This
wild type p53 protein shows growth suppression activi-
ties, particularly leading to a reversible arrest in either
the G1 phase or in the S phase [2]. This cell-cycle arrest
may result in apoptosis, an irreversible process culmi-
nating in individualized cell death [3], or allow for
repair of genomic aberrations [4]. Mutation of the p53
gene has been found in several human tumors [5] and
can be shown by immunohistochemical detection of the
mutated gene product, which demonstrates an in-
* Corresponding author, Department of Obstetrics and Gynecol-
ogy, Saint Joseph Hospital, P.O. Box 7777, 5500 MB Veldhoven, The
Netherlands. Tel.: + 31 40 588384; Fax: + 31 40 589564.
creased stability and a longer half live than the wild-
type protein.
In the cervix, infection with human papillomavirus
(HPV) is associated with tumor development and HPV
D N A sequences of the high risk types 16 and 18 can be
detected in a high percentage of these malignancies [6].
The oncoprotein E6 encoded by HPV type 16 and 18
forms a complex with and promotes the degradation of
the p53 protein [7]. The effect of neutralisation of p53
wild-type protein through E6 is more extensively stud-
ied in cervical cancer tumorigenesis than the effects of
p53 mutation per se.
There is evidence that HPV-negative tumors are
more aggressive and show a greater metastatic potential
and worse prognosis than HPV positive cancers [8,9].
In HPV positive cancers, loss of p53 function results
from loss of p53 protein in the cell. In contrast, muta-
tions of p53 occur preferentially in tumors without
HPV infection [10]
This suggests that p53 mutations may result in a
more severe deregulation of normal growth than p53
elimination as a result of E6 production.

0301-211505/$09.50 ~ 1995 Elsevier Science Ireland Ltd. All rights reserved

SSDI 0301-21 I 5(95)02225-V
56 G.L. Bremer et al. ,; European Journal ol Obstetrics & Gynecology and Reproduclit:e Biology 63 (1995) 55 59
Against this background, the purpose of this study
was to determine whether overexpression of p53 protein
is indeed a parameter of more aggressive disease in
patients with cervical cancer.
2. Materials and methods
2.1. Clinical &aa
From 1978 to 1993, a total of 205 patients with
invasive cervical cancer were admitted to the University
Hospital Maastricht, including 67 referrals from four
different regional community hospitals. All patients
were staged according to the classification of the Inter-
national Federation of Gynecology and Obstetrics.
The number of patients assessed as having early
cervical cancer was 145:21 (10.2%) patients, stage IA;
124 (60,5%) patients, stage IB and IIA. Advanced
disease (stage lIB, IlI and IV) was diagnosed in 60
(29.3%) patients.
This study focuses on patients with stage IB-IV cervi-
cal cancer. In the group with stage IB/IIA disease,
paraffin blocks were no longer tracable for 8 patients.
In the group with advanced disease, paraffin blocks
were not available for 9 patients, insufficient tumor
tissue remained in the paraffin blocks for 5 patients for
study and follow up data were incomplete for 6 pa-
tients. Thus, 28 patients were excluded and the remain-
ing 156 patients form the basis of the study: 116
patients with stage IB and IIA and 40 patients with
advanced disease.
The primary treatment for stage IB and I1A patients
was radical surgery according to Wertheim-Meigs. Dur-
ing the surgical procedure, pelvic lymph nodes were
examined by frozen sections and when they were nega-
tive for tumor the procedure was completed.
When pelvic lymph nodes were involved by cancer,
the surgical procedure was abandoned and full pelvic
and intracavitary radiation therapy was started after 2
weeks. Of the 116 stage IB and IIA patients, 75 patients
were treated by complete radical hysterectomy and
pelvic lymphadenectomy.
Fourteen of these patients received post-operative
radiation therapy because of positive surgical margins
(n = 5), near positive (less than 2 mm) surgical margins
(n = 6) or microscopical involvement of the
parametrium (n = 3). Survival without evidence of dis-
ease for this group of 75 patients was 88%. In 27 of the
116 patients, the procedure was abandoned because of
pelvic lymph node involvement found by frozen section
during operation. All of these patients underwent radia-
tion therapy and survival without evidence of disease
was 65%.
Fourteen out of the 116 patients received primary
radiation therapy because of relative or absolute con-
traindications for surgery such as age, obesity or poor
medical condition. Survival without evidence of disease
was 65%.
A detailed report of the results of this treatment
policy is described elsewhere [11].
All patients with advanced disease were treated with
primary radiation therapy. Survival in stage liB was
38%, in stage III 22"/,, and no patient survived for at
least 5 years in stage IV. Fig. 1 shows the survival
curves for the study group classified by stage.
Patient follow up ranged from 6 months to 15 years
(median 6.2 years).
2.2. H&tological analysis
Diagnostic cervical biopsies and representative sam-
ples of tumors found in the hysterectomy specimens
were routinely processed, paraffin-embedded and
stained by standard histochemical techniques. Histo-
logic examination showed squamous cell carcinoma in
139 (89.2%), adenocarcinoma in 14 (8.9%) and
adenosquamous carcinoma in 3 (1.9%) patients.
Sections for immunobistochemistry were stained us-
ing the monoclonal mouse anti-human p53 protein,
DO7 (Dako Corporation Denmark), recognizing an
epitope in the NHe-terminus between amino acids 35
45 of the wild and mutant type of the p53 protein.
Tissue sections (3/lm) of the tumor were deparaffinized
and endogenous peroxidase was blocked with 0,6%
H 2 0 2 in methanol. The slides were pretreated in 10 mM
citrate buffer (pH, 6.0) and boiled in a microwave oven
for 10 min. After cooling slowly, the sections were
incubated with the primary p53 monoclonal antibody
(dilution 1:500) for 45 min at room temperature. After
washing with PBS (3 x 5 min), the sections were
incubated with biotinylated rabbit anti mouse im-
munoglobulin (DAKO 1:500) for 30 min, washed and
incubated with strept-avidin-biotin complex HRP la-
belled (DAKO 1:1000) for 30 min.
1.00
113
0.75
r~ HA
0.50
HB
0.25 111
0.00 IV
0 2 4 6 8 10 12 14
T i m e (year)
Fig. 1. Kaplan-Meier analysis of survival of patients with cervical
cancer, according to stage. IB: stage IB (n : 85); IlA: stage IIA
(n = 31); liB: stage l i b ( n - 2 1 ) ; Ill: stage 111 (it = 14); IV: stage IV
( n - 5 ) , (Log Rank Z ~ 45.1; D F - 4 ; P < 0.0001).
 G.L. Bremer et al. / European Journal o! Obstetrics & Gynecoh~gy and Reproductive Biology 63 (19951 55 59 57

Table I The immunostaining pattern of p53 protein in the

Correlation between p53 tumor suppressor gene protein expression
and stage in cervical cancer (n = 156)
different stages is presented in Table 1. In stage IB and
I1A overexpression was observed in, respectively, 20.1
Stage p53 ( ) Low expression Overexpression and 32.2% of the tumors; no p53 reactivity was found
I% 5% p 5 3 ( + ) >5% p53(+) in 51.7% of the tumors. In stage IIB, low and overex-
pression were both observed in 47.6% and only one
I1 %, n %~ n %~
tumor was negative for p53 immunostaining. In stage
IB 44 51.7 24 28.2 17 20.1 III and IV, all tumors showed p53 expression.
IIA 16 51.7 5 16.1 l0 32.2 Since no major differences in p53 expression could be
liB 1 4.8 10 47.6 10 47.6 observed between stage IB and IIA, and stage IIB, III
111 0 0 7 50 7 50 and IV statistical analysis was performed after
IV 0 0 2 40 3 60
combining stage IB and IIA patients in early cervical
Total 61 39.1 48 30.7 47 30.2 cancer, and combining patients in stage IIB, III and IV
in advanced cervical cancer. The difference in p53 ex-
pression pattern between early and advanced stage
After final washing with PBS, a diaminobenzidine
cervical cancer was statistically significant (chi square
H202 complex with 0.01 M imidazol (Merck) was used
Z 2=30.37; D F = 2 ; P < 0.00001).
to visualize the immunoreaction. Positive controls con-
sisted of simultaneously immuno-stained sections of a
3.2. p53 expression, histoh)gic cell type and peh, ic
carcinoma that was previously shown to be immunore- lymph node status
active for p53 by this technique.
Table 2 shows p53 protein accumulation according to
histological cell type and pelvic lymph node involve-
2.3. Analysis oJp53 protein accumulation ment.
Many representative sections of each tumor were The percentage of p53 protein expression (low and
investigated for the presence of tumor cells immunore- overexpression) was significantly higher in squamous
active for p53 protein. A semiquantitative estimate of cell carcinoma compared to tumors with an adenocom-
the percentage of cells positive was performed by ponent (chi square Z 2 = 9.65; D F = 2: P = 0.008).
counting a minimum of 1,000 cells in representative One hundred and two stage IB and IIA patients were
high-power fields. All samples were evaluated in a treated surgically by Wertheim-Meigs procedure and
blinded manner, without knowledge of the clinical out- pelvic lymph node status of these patients was known.
come of the patients. Arbitrarily, low expression was In both, pelvic lymph node positive and negative can-
defined as the presence of 1 5% p53 positive tumor cers, approximately 50% of the tumors were negative
cells, and overexpression was detected when at least 5% for p53 accumulation. Overexpression of p53 protein
of the cells were p53 positive. was detected in 29.6% of the cancers with positive
pelvic lymph nodes compared to 18.7% with negative
2.4. Statistical analysis pelvic lymph nodes. Statistical analysis showed no sig-
The objective of this study was to assess the relation- nificant association.
ship between p53 expression, pelvic lymph node metas-
tasis and disease free survival. In a univariate analysis,
disease free survival rates were calculated based upon
the method of Kaplan and Meier and survival curves Table 2
were compared using the log-rank test. Difference in Correlation between p53 t u m o r suppressor gene protein expression,
histological type and pelvic lymph node metastasis in cervical cancer
p53 expression between patients with and without
pelvic lymph node involvement was tested using chi p53 (-) Low expression Overexpression
square statistics. Data were analyzed with B M D P
Statistical Software version 1990 (University of Califor- 1% 5 % p 5 3 ( + ) >5% p53(+)
nia Press, Berkeley) on an IBM personal computer. II I~j n %, ti I~/o

Histological type
3. Results Squamouscell 37 35.0 46 32.1 45 32.9
carcinoma
3. I. p53 expression and stage Adeno + adeno- 12 75.0 2 12.5 2 12.5
squamousca
p53 expression was localized in the nuclear and peri-
nuclear areas. No reactivity was observed in normal Pelvic lymph
nodes
cervical epithelium. Overexpression of p53 was detected
Negative 37 49.3 24 32.0 14 18.7
in 30.2% of the tumors, low expression was seen in Positive 14 51.9 5 18.5 8 29.6
30.7% and no reactivity was found in 39.1%.
 The prognostic value of ~53 protein overexpression
was assessed comparing the results with pelvic lymph
node metastasis and disease free survival. A higher
percentage of tumors with pelvic lymph node involve-
ment showed ~53 protein overexpression but this was
not statistically significant.
Survival was significantly reduced for the entire
group of patients in correlation with p53 protein over-
expression however. these results were highly dependent
on stage. In each FIG0 stage. no correlation between
~53 protein ovcrexpression and disease free survival
was observed.
In cervical carcinoma. deregulation of ~53 function
through direct neutralisation of the gene product by E6,
a \:iral protein produced by HPV types 16 and 18, has
received much attention [7,14]. Elimination of ~53
protein is thought to be an important step in tumorige-
nesis of HPV-associated cervical cancer [I 51. Little data,
however. are available about the significance of muta-
tion of the pS3 gene per SC in the development of
cervical cancer. Studies by Crook and coworkers have
shown that ~53 mutation is associated with HPV nega-
tive status and that loss of pS3 wild type function, due
Fig. 2 shows the survival curves in relation to ~53 either to neutralisation of the gene product by E6 or to
protein expression. Considering the entire study group. pS3 gene mutation, is involved in the etiology of cervi-
a significantly lower percentage of patients survived cal cancer [16]. It is also suggested that HPV-negative
without evidence of disease when overexpression was tumors show ;I more aggressive behavior in terms of
observed (Logrank test z’ = 9.14; P = 0.01). This was lymph node metastasis. disease free interval and sur-
caused by the high incidence of p53 overexpression in vival compared to HPV-positive tumors [8,9].
the stages IIB. 111 and IV because no correlation be- As to the HPV status. no data are available in our
twcen ~53 protein overexpression and survival was retrospective patient series. Overexpression of the ~53
observed in the different stages. protein would not be expected to occur in HPV 16- 1%
associated tumors where E6 binds and degradates the
4. Discussion pS3 protein.
It is. therefore, conceivable that ~53 mutation could
In the present retrospective study of 156 cervical only play ;I role in the tumorigenesis of HPV 16 1%
cancers, we demonstrated that low and overexpression negative cervical cancers.
of pS3 protein by immunohistochemistry was found in. Our data suggest that this indeed could be the case.
respcctii,ely. 30.7 and 30.2%. We considered only tu- be it that in cervical cancer ~53 mutations occur in a
mors showing overexpression to be ~53 protein positive relatively late stage of disease.
because thcrc is evidence that in tumors showing an Several studies have shown the prognostic signifi-
occasional focus of p53 immunoreactive cells in an cance of ~5.3 expression in breast, gastric and ovarian
otherwise negative neoplasm (below 5% p53 positive). cancer [17,18].
not only mutant but also wild type p53 protein is being Our results imply that the influence of p53 protein
stained particularly when sensitive techniques like mi- overexpression on the prognosis of patients with cervi-
crowave irradiation are used [12]. cal cancer is limited. Th,s observation confirms data
Therefore, overcxpression of ~53 protein is more from an investigation of 192 stage III cervical cancers
likely to be associated with somatic p53 gene mutation. treated by radiation therapy showing ~53 immunoreac-
This assumption is in agreement with the results of ticity in 26% of the tumors. In this study, no correla-
Baas and coworkers. who found X0% concordance be- tion \v;ts observed between patient survival and ~53
tween ~53 gene mutation and pS3 protein overexprcs- protein expression 1191. Another investigation showed
sion, using mouse monoclonal antibody DO7 and the similar results also in a group stage III cervical cancers
detection of mutations by DNA sequencing 1131. [Xl.
When stage. according to the FIG0 classification. In summary. the data of this and other studies sug-
increased ~53 protein overexpression raised from 20.1% gest that apart from pS3 function and regulation in
in stage IB to 60% in stage IV. HPV I6 I K-associated cervical cancers, deregulation of
 G.L. Bremer et al. / European Journal o[ Obstetrics & Gynecology and R~TwO&tctive Biology 63 (1995) 55 59
p53 per se could be another possible step in cervical
cancer tumorigenesis.
p53 protein overexpression, however, seems to occur
in late stages of cervical cancer development and does
not appear to be of prognostic significance in cervical
cancer.
Acknowledgements
This investigation was supported in part by the
Dutch National Cancer Society 'Koningin Wilhelmina
Fonds'.
We are grateful to the gynecologists of the St Lauren-
tius Hospital Roermond, the Maasland Hospital Sit-
tard/Geleen and the St Joseph Hospital Kerkrade for
referring their patients.
References
[1] Finlay CA, Hinds PW, Levine AJ. The p53 proto-oncogen can
act as a suppressor of transformation. Cell 1989; 57:1083 1093.
[2] Casey G, Lo-Hsueh M, Lopez ME, Vogelstein B, Stanbridge JB.
Growth suppression of h u m a n breast cancer cells by the intro-
duction of a wild type p53 gene. Oncogene 1991; 6:1791 1797.
[3] Ulrich SJ, Anderson CW, Mercer WE, Appella E. The p53
tumor suppressor protein, a modulator of cell proliferation. J
Biol Chem 1992: 267:15259 15262.
[4] Bartek J, Iggo R, G a n n o n J, Lane DP. Genetic and i m m u n o
chemical analysis of mutant p53 in h u m a n cancer cell lines.
Oncogene 1990: 5:893 899.
[5] Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 muta-
tions in h u m a n cancers. Science 1991; 253:49 53.
[6] Vousden KH. H u m a n papillomaviruses and cervical carcinoma.
Cancer Cells 1989; 1:43 50.
[7] Werness BA, Levine A J, Howley PM. Association of h u m a n
papillomavirus 16 and 18 E6 proteins with p53. Science 1990;
2 4 8 : 7 6 79.
[8] Riou G, Eavre M, Jeannel D, Bourhis J. Le Doussal V Orth G.
Association between poor prognosis in early stage invasive cervi-
cal carcinomas and non-detection of HPV DNA. Lancet I990:
335:1171 1174.
59
[9] Higgins GD, Davy M, Roder D, Uzelin DM, Phillips GE Burrell
CJ. Increased age and mortality associated with cervical car-
cinomas negative for h u m a n papillomavirus RNA. Lancet 1991;
338:910 913.
[10] Paquette RL, Lee YY, Wilczynki SP, K a r m a k a r A, Kizaki M,
Miller CW, Koeffler HP. Mutations of p53 and h u m a n papillo-
mavirus infection in cervical cancer. Cancer 1993; 7 2 : 1 2 7 2
1280.
[11] Bremer GL, van der Putten H W H M , Dunselman GAJ, de Haan
J. Early stage cervical cancer: Aborted versus completed radical
hysterectomy. Eur J Obstet Gynecol Reprod Biol 1992; 47:
147 151.
[12] Gown AM, De Wever N, Battifora H. Micro-wave based anti-
genic unmasking: A revolutionary new technique for routine
immunohistochemistry. AP l m m u n o h i s t o c h e m 1993: 1:256 266.
[13] Baas IO, Mulder J-WR, Offerhaus JA, Vogelstein B Hamilton
SR. An evaluation of six antibodies for immunohistochemistry
of mutant p53 gene product in archival colorectal neoplasms. J
Pathol 1994; 172:5 12.
[14] Scheffner M, Munger K, Byrne JC, Howley PM. The state of the
p53 and retinoblastoma genes in h u m a n cervical carcinoma cell
lines. Proc Natl Acad Sci USA 1991: 88:5523 55227.
[15] Scheffner M, Werness BA, Huibregtse JM, Levine AJ Howley
PM. The E6 oncoprotein encoded by h u m a n papillomavirus
types 16 and 18 promotes the degradation of p53. Cell 1990: 63:
1129 1136.
[16] Crook T, Wrede D, Tidy JA, Mason WP, Evans DJ, Vousden
KH. Clonal p53 mutation in primary cervical cancer: association
with human-papilloma-negative tumours. Lancet 1992; 39:
1070 1073.
[17] Thor AD, Moore 1I DH, Edgerton SM, Kawasaki ES Reihsaus
E, Lynch HT, Marcus JN, Schwartz L, Chert L-C, Mayall BH,
Smith HS. Accumulation of p53 tumor suppressor gene protein:
an independent marker of prognosis in breast cancers. J Natl
Cancer lnst 1992; 84:845 855.
[18] Klemi PJ, Takahashi S, Joensuu H, Kiilholma P Narimatsu E,
Mori M. Immunohistochemical detection of p53 protein in bor-
derline and malignant serous ovarian tumors, lnt J Gynecol
Pathol 1994; 13:228 233.
[19] Oka K, N a k a n o T, Arai T. p53CMI expression is not associated
with prognosis in uterine cervical carcinoma. Cancer 1993; 72:
160 164.
[20] Gitsch G, Kainz C, Joura E, Breitenecker G. Mutant p53
product in patients with stage Ill cervical cancer. Anticancer Res
1992; 12:2241 2242