Gynecologic Oncology 93 (2004) 34 – 40 www.elsevier.com/locate/ygyno
A significance of immunohistochemical determination of steroid receptors, cell proliferation factor Ki-67 and protein p53 in endometrial carcinoma S. Oreskovic, a,* D. Babic, b D. Kalafatic, a D. Barisic, a and L. Beketic-Oreskovic c a
Department of Obstetrics and Gynecology, School of Medicine, University of Zagreb, Zagreb, Croatia b Department of Pathology, School of Medicine, University of Zagreb, Zagreb, Croatia c University Hospital for Tumors, Zagreb, Croatia Received 28 April 2003
Abstract Objective. The aim of the study was to preoperatively predict the biologic behavior of the endometrial carcinoma using immunohistochemical analysis of the p53 protein and Ki-67 expression, and estrogen receptor (ER) and progesterone receptor (PR) status, in the material obtained by fractional curettage. Methods. One hundred and thirty-six patients with primary endometrial carcinoma were included in the study. In all 136 patients, the fractional curettage was performed before the hysterectomy, and the diagnosis of endometrial carcinoma was confirmed pathohistologically after the surgical procedure on the hysterectomy specimens. The significance of the prognostic factors was assessed using univariate and multivariate analyses. The cutoff values of the percentage of ER, PR, p53, and Ki-67 positive cells in terms of survival probability determination were obtained as the values of the highest chi-square test, using proportional-risk regression method. A multivariate Cox regression analysis was performed to estimate the influence of several clinical, pathohistologic, and immunohistochemical covariates to patients’ survival. Survival curves were determined by the Kaplan – Meier product-limit method based on the most recent clinical status. Results. According to the histologic type of the tumor, fractional curettage specimens revealed 111 histologically favorable types (81.6%) and 25 unfavorable types (18.4%). The data indicate that ER, PR, Ki-67, and p53 levels of the hysterectomy specimens and those of the preoperative specimens were in fairly good agreement. The patients with the most favorable tumor grade (G I) had significantly better prognosis when the percentage of p53 positive cells was less than 15%. In the group of patients with histologic grade II, the survival was affected by ER expression (more than 30% of positive cells) and p53 levels (less than 15% of positive cells). None of the parameters was predictive in the group of patients with histologic grade III. Conclusion. We found that determination of immunohistochemical parameters (ER, PR, and p53) on well-differentiated and moderately differentiated endometrial carcinoma of favorable histologic type obtained by curettage enables the recognition of the patients with favorable prognosis, who should not be treated by radical surgery. D 2004 Elsevier Inc. All rights reserved. Keywords: Estrogen receptor; Progesterone receptor; p53; Ki-67; Endometrial carcinoma
Introduction Endometrial carcinoma is the most common malignant disease of the female reproductive tract in the developed countries, affecting about 2 – 3% of all women [1]. The standard diagnostic workup consists several clinical and laboratory procedures. Fractional curettage is mandatory during the diagnostic follow-up of the endometrial carcino-
* Corresponding author. Department of Obstetrics and Gynecology, School of Medicine, University of Zagreb, Petrova 13, Zagreb, Croatia. E-mail address:
[email protected] (S. Oreskovic). 0090-8258/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2003.12.038
ma [2]. Besides the pathohistologic verification and classification of the disease, the procedure theoretically offers the determination of other important characteristics of the tumor, such as histologic grade and cervical involvement. Correlations between steroid hormone receptor status of endometrial carcinoma and known prognostic parameters such as tumor stage, tumor grade, and depth of myometrial infiltration have been documented [3 – 11]. The significance of the progesterone receptor status on the overall survival rate has been particularly demonstrated in several papers [8,9,11 –15]. Recently, some other factors have been proven to offer additional data about the biologic behavior of the endometrial carcinoma. The mutation of the p53 tumor
S. Oreskovic et al. / Gynecologic Oncology 93 (2004) 34–40
suppressor gene, the most frequently observed genetic alteration in human cancer, has also been documented in endometrial cancer, although the prognostic significance of its overexpression is still conflicting [16 –19]. The Ki-67 nuclear antigen expression, an indicator of cell proliferation, has also been evaluated as prognostic factor on survival in many human malignant tumors, including endometrial carcinoma, although the published data are generally sparse and difficult to compare [20 – 25]. The immunohistochemical determination of the above tumor characteristics has been commonly performed on the hysterectomy specimens after the surgical procedure [26 – 28]. As far as we know, none of the studies published so far has dealt with the diagnostic and prognostic significance of the fractional curettage specimens obtained preoperatively. Furthermore, a correlation between risk factors obtained pre- and postoperatively has never been extensively studied. The purpose of this study was to preoperatively evaluate the biologic behavior of the endometrial carcinoma using immunohistochemical analysis of the p53 and Ki-67 expression, and estrogen receptor (ER) and progesterone receptor (PR) status, in the material obtained by fractional curettage. If knowledge of tumor receptor status or p53 or Ki-67 expression could preoperatively estimate the biologic aggressiveness of the tumor, this information could be helpful in making
35
the decisions of the extensivity of the surgical procedure, as well as in estimating the survival rate of the disease.
Materials and methods Between 1986 and 1990, a series of 136 patients with primary endometrial carcinoma were included in the study. In all 136 patients, the fractional curettage was performed before the hysterectomy, and the diagnosis of the endometrial carcinoma was confirmed pathohistologically after the surgical procedure on the hysterectomy specimens. The clinical stage, histologic type, and tumor grade were assessed using the Federation of Gynecology and Obstetrics (FIGO; 1988) system of classification. According to the histologic type of the tumor, the patients were divided in two groups: group I, histologically favorable types (endometrioid carcinoma, mucinous carcinoma, and adenoacanthoma); and group II, histologically unfavorable types (serous, clear-cell, and adenosquamous carcinomas). According to the level of tumor differentiation, the patients were divided into three groups: grade I (G I), well-differentiated tumors; grade II (G II), moderately differentiated tumors; and grade III (G III), poorly differentiated tumors.
Fig. 1. Immunohistochemical staining of Ki-67, p53, estrogen, and progesterone receptors on curettage specimens (white arrows pointing immunohistochemically positive cells).
36
S. Oreskovic et al. / Gynecologic Oncology 93 (2004) 34–40
Formalin-fixed, paraffin-embedded blocks from obtained materials were retrieved. The paraffin blocks were cut in 3to 5-Am sections after the 24-h fixation in 10% buffered formalin. The hemalaun– eosin staining, as well as PAS (periodic acid-Schiff), PAS – diasthase, Mallory and Gomory staining, where needed, were performed. For immunohistochemical analysis, the sections were incubated in methanol with 0.3% hydrogen peroxide to block endogenous peroxidase activity. The samples were incubated with the primary antibody for 60 min at room temperature. The primary antibodies used in the study were anti-ER mouse antihuman monoclonal antibodies 1D5, antiPR mouse antihuman monoclonal antibodies 1AG, mouse antihuman p53 protein monoclonal antibodies DO-7, and rabbit antihuman Ki-67 antigen monoclonal antibodies, using original DAKO LSAB 2 kit (DAKO Corporation, Carpinteria, CA). Methyl green was used as a counterstain for nuclei. The color index was graded from 0 to 3 (0, negative finding; 1, weak intensity; 2, moderate intensity; and 3, strong intensity). The most densely positive area of tumor cells was selected, and more than 500 tumor cells were scored at 400 magnification (Fig. 1). The index of positivity of ER, PR, p53, and Ki-67 was scored from 0 to 4 (index 0, no immunohistochemically positive cells; index 1, 1 – 10% positive cells; index 2, 11 – 50% positive cells; index 3, 51 –80% positive cells; and index 4, >80% positive cells). The color index was then multiplicated with the percentage of positive cells, and index of immunoreactivity (IR index) was obtained. The tumor was considered as highly immunoreactive if the product was between 7 and 12 (IR index = 3). The moderately immunoreactive tumors (IR index = 2) had the product between 4 and 6. The IR index 1 was used for weakly immunoreactive tumors (product < 4). The significance of the prognostic factors was assessed using univariate and multivariate analysis. The univariate analysis involved performing a chi-square analysis for categorical variables and a nonparametrical Mann –Whitney test for continuous variables. The cutoff values of the percentage of ER, PR, p53, and Ki-67 positive cells in terms of survival probability determination were obtained as the values of the highest chi-square test, using proportionalrisk regression method. A multivariate Cox regression analysis was performed to estimate the influence of several clinical, pathohistologic, and immunohistochemical covariTable 1 Estrogen receptor (ER), progesterone receptor (PR), Ki-67 and p53 protein levels on the fractional curettage and hysterectomy specimens
ER PR Ki-67 p53
Fractional curettage (% of positive cells)
Hysterectomy specimens (% of positive cells)
Mean (FSD)
Range
Mean (FSD)
Range
53.4 59.0 24.7 10.6
0 – 95 0 – 97 6 – 68 1 – 89
54.1 (34.6) 54.9 (33.3) 26.9 (13.3) 11.2 (17.2)
0 – 95 0 – 95 6 – 70 1 – 89
(34.6) (33.8) (13.1) (16.8)
Table 2 The distribution of patients according to the cutoff levels of ER, PR, Ki-67, and p53 in curettage and hysterectomy specimens Fractional curettage specimens
ER z 30%
