Early Detection and Diagnosis
Prognosis of adenocarcinoma of the uterine cervix: p53 expression correlates with higher incidence of mortality
Version of Record online: 12 MAR 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 121, Issue 1, pages 106–110, 1 July 2007
How to Cite
Baalbergen, A., Ewing-Graham, P. C., Eijkemans, M. J. and Helmerhorst, T. J.M. (2007), Prognosis of adenocarcinoma of the uterine cervix: p53 expression correlates with higher incidence of mortality. Int. J. Cancer, 121: 106–110. doi: 10.1002/ijc.22678
- Issue online: 24 APR 2007
- Version of Record online: 12 MAR 2007
- Manuscript Accepted: 30 JAN 2007
- Manuscript Received: 3 SEP 2006
- cervical adenocarcinoma;
We investigated the significance of prognostic markers-estrogen receptor, progesterone receptor, p53, MIB-1 and bcl-2 - in adenocarcinoma of the uterine cervix. In 101 patients with primary cervical adenocarcinoma, treated from 1989 to 2000, we evaluated clinical parameters in relation to these prognostic markers. Mean age of patients was 45 years. Seventy eight percent of the patients were in FIGO stage I, 16% stage II, 7% stage III and IV. estrogen receptor, progesterone receptor, p53 and bcl-2 immunoreactivity was scored as 0 (up to 5% positive cells), 1+ (5–25% of cells positive), 2+ (26–50% of cells positive), 3+ (51–75% of cells positive) or 4+ (>76% of cells positive). MIB-1 was scored in 10 categories: 0–10, 11–20, 21–30, 31–40, 41–50, 51–60, 61–70, 71–80, 81–90, 91–100. The overall survival rate was 67%. Survival was not influenced by estrogen receptor, progesterone receptor, MIB-1, or bcl-2 strongly positive staining. Only p53 showed significant influence on survival, even when adjusted for stage or tumor grade. In conclusion, it does not seems useful to determine estrogen receptor, progesterone receptor, MIB-1 or bcl-2 in cervical adenocarcinomas as an indication of prognosis: survival is not influenced by presence or absence. However, if p53 staining is strongly positive survival is significantly worse than in tumors scored as negative or weak positive. © 2007 Wiley-Liss, Inc.
The prevalence of cervical adenocarcinoma has risen worldwide, in recent studies accounting for 25% of all cervical carcinomas.1 In The Netherlands there are ∼700 new cases of cervical cancer per year, of which 19% are adenocarcinomas.2 The overall survival rate for cervical adenocarcinoma is worse than for the more common squamous cell carcinoma of the uterine cervix.3, 4, 5 However, for early stage adenocarcinoma survival is equivalent to squamous cell carcinoma.6, 7 Stage is one of the most important factors for survival in cervical adenocarcinoma.8, 9, 10
In breast and uterine corpus carcinoma expression of estrogen receptor and progesterone receptor are associated with better prognosis.11, 12 The prognostic significance of hormone receptor status in adenocarcinoma of the uterine cervix is still unclear.
The p53 tumor suppressor gene plays a major role in cell cycle control and growth arrest following DNA damage. Mutations of the p53 tumor suppressor gene are the most common genetic alterations in human cancers. P53 protein overexpression has been found to be a associated with poor prognosis in several malignancies13, 14, 15; however results of studies in cervical cancer mostly show no association.16
It has been shown that the aberrant expression of cell cycle regulatory proteins is a potential prognostic indicator within different tumor groups. Ki-67 nuclear antigen is found in proliferating cells. It identifies the growth fraction of normal and neoplastic cells, and is expressed solely in cells in the G1, S, G2 and mitotic phases. It has been reported that the percentage Ki-67 index correlates inversely with the prognosis of various tumors. The MIB-1 murine monoclonal antibody, which reacts with the Ki-67 nuclear antigen, is a marker for proliferating cells and in breast and ovarian cancer it is a good prognostic marker.17, 18
Bcl-2 is the protein product of a proto-oncogene that inhibits apoptosis, and it has been shown to prevent apoptosis instead of promoting cell proliferation. Together with the proapoptotic protein Bax it forms a dimeer to control progression to apoptosis. Expression of bcl-2 is associated with better survival in patients with solid tumors.19 The present study evaluates the prognostic significance of estrogen and progesterone receptor, p53, MIB-1, and bcl-2 in adenocarcinoma of the uterine cervix.
Material and methods
All primary invasive cervical adenocarcinomas diagnosed between 1989 and 2000 in the Rotterdam area were retrieved from local cancer registries: IKR regional cancer registration body, Palga pathology registration and oncology registration at the Daniel den Hoed Cancer Clinic. Case notes were retrieved from the Erasmus MC University Hospital Rotterdam, the Daniel den Hoed Cancer Clinic and the affiliated hospitals in the region. The patients were staged according to the FIGO system. From 103 patients formalin-fixed and paraffin-embedded tissue was available for immunohistochemical testing. The haematoxylin and eosin slides (n = 103 cases) were reviewed by an experienced gynaecopathologist (P.C.E-G). Any cases in which there was clinical or pathological doubt concerning the primary site were excluded from the study, as were patients for whom clinical follow-up was inadequate or unavailable. All cases of usual/mucinous type cervix adenocarcinoma were included, along with the variants of mucinous carcinoma.20 Where there was a suggestion of endometrioid morphology the case was not included unless clearly located in the cervix only. If a squamous component was present the case was classified separately as adenosquamous. A number of rare tumors were excluded: serous, clearcell and adenoid basal.
The tumors were graded architecturally into well, moderately or poorly differentiated depending on the proportion of solid growth.21 The project was approved by the Medical Ethical Committee of the ErasmusMC University Hospital Rotterdam (nr.211.651/2002/48).
Marker analysis immunohistochemistry
All immunohistochemical analyzes were carried out by applying the avidin-biotin complex (ABC) method, on 4 μm sections cut from formalin-fixed, paraffin embedded tissues using commercially available mouse- monoclonal antibodies and DAB as color technique. Sections were deparaffinized and then manually incubated in the primary antibodies for 30 min at 20°C Before incubation the sections were placed in a microwave oven with citric acid buffer solution (pH 6.0, 0.1 M) for 15 min to retrieve the antigens. In all cases a microwave pretreatment of the slides was used.
For estrogen and progesterone receptors the antibodies and dilutions used were as follows: oestrogen receptor DAKO M7047, ID5, 1:160, progesterone receptor DAO M3569 PgR 636, 1:40. P53 was analyzed with the mouse monoclonal antibody p53, clone DO-7, isotype IgG2b, which recognizes wildtype and mutant forms of the p53 protein (1:50; DAKO, Glostrup, Denmark). For p53 only cells with a distinct brown stain confined to the nucleus were regarded as immunoreactive. In slides stained for MIB-1 the antibody used was Immunotech 0505 MIB-1, 1:100. Immunohistochemical staining for bcl-2 was with DAKO M0887 124, 1:80 and a semi-quantitative assessment was made of the percentage of cells showing cytoplasmatic staining.
Scoring of immunostaining
Immunostaining was scored semiquantitatively. For estrogen receptor, progesterone receptor, p53 and bcl-2 the percentage of positive cells was graded as follows: 0 (up to 5% positive cells), 1+ (5–25% of cells positive), 2+ (26–50% of cells positive), 3+ (51–75% of cells positive) or 4+ (>76% of cells positive). MIB-1 was scored in 10 categories: 0–10, 11–20, 21–30, 31–40, 41–50, 51–60, 61–70, 71–80, 81–90, 91–100. Estrogen receptor, progesterone receptor, p53 and MIB-1 were scored for nuclear staining, whereas bcl-2 was scored for cytoplasmic staining. In the analysis, a cut-off value at ≤50% versus >50% was used as it gave the best survival curve separation.
Patients with early disease underwent radical hysterectomy and pelvic lymph node dissection, unless their clinical condition did not permit this procedure. The criteria for postoperative radiotherapy were: positive lymph node involvement, compromised surgical margin or extension of tumor into the parametrium. Patients with stage II-B disease or higher were treated with external irradiation and brachytherapy. In 1999 hyperthermia was added to this regimen to improve survival.22
Patients were followed-up every three months during the first 2 years, and thereafter every 6 months until the fifth year.
Patient disease-specific survival distribution was calculated using the Kaplan-Meier method. Patients who died of intercurrent disease or who were lost to follow-up, were censored at the time of last known follow-up. The significance of the survival was tested by log-rank test. To obtain independent prognostic significance of the variables, Cox's proportional hazards regression analysis was used in multivariate analyzes. A value of p < 0.05 was considered statistically significant in all analyzes.
Tumor characteristics such as FIGO stage and tumor grade are presented in Table I. The mean age of the patients (n = 101) was 45 years (26–81 years). Seventy nine percent of the patients were in FIGO stage I, 14% stage II, 7% stage III and IV. Eight percent had primary surgery, whereas 13% had primary radiotherapy and 3 patients received palliative therapy. The overall disease-specific survival was 67%.
|Number of patients||Percentage|
The results of immunohistochemical staining are shown in Table II. In 7 patients there was insufficient material for all immunohistochemical analyzes. The estrogen receptor was analyzed in 101 patients. In 83% of the patients less than 5% of the tumor cells stained positively for estrogen receptor, leading to a classification as ‘negative’ for the receptor. In 88% of the 101 patients we found a negative staining for progesterone receptor.
|Marker||No. of tumors with negative staining (≤50%)||No. of tumor with positive staining by extent of distribution||No of positive tumors with strong positive staining (>50%)|
|1+ (5–25%)||2+ (26–30%)||3+ (51–75%)||4+ (>76%)|
p53 was analyzed in 94 patients. In 62% of the patients less than 5% of the tumor cells stained positively for p53, leading to a classification as ‘negative’ 80% of the 96 patients had less than 5% staining for the apoptosis control gene, bcl-2.
The proliferation marker MIB-1 was analyzed in 100 patients. See Figure 1. The mean MIB-1 index was 60%. In univariate analysis using the Kaplan-Meier method only marker P53 was significant for survival; patients with a higher p53 staining had a worse survival; in the p53 negative group (staining less than 5%) the 5-year survival was 68% which declined to 40% when staining was 4+ (p < 0.05). Accordingly, patients were grouped into those with strongly positive tumors (>50% staining) and those with negative or weak immunostaining (≤50%). Ten percent of our patients had a strongly positive staining for ER and 8% of the patients for PR. Only 5% of our patients had a strongly positive staining for bcl-2. Fifty nine percent of patients had a strongly positive staining for MIB-1. Survival was not significantly influenced by strongly positive staining for ER, PR, bcl-2 or MIB-1. Patients with strongly p53 immunoreactive tumors had a significantly worse outcome than patients whose tumors were p53-negative or weakly positive (p < 0.05). See Table III and Figure 2.
|Weak-negative||5-yr survival||Strong positive||5-yr survival||p|
|ER||91/101 (90%)||68%||10/101 (10%)||64%||0.32|
|PR||93/101 (92%)||68%||8/101 (8%)||67%||0.26|
|p53||83/94 (88%)||71%||11/94 (12%)||26%||0.0004|
|Bcl-2||91/96 (95%)||67%||5/96 (5%)||60%||0.71|
|MIB-1||41/100 (41%)||61%||59/100 (59%)||72%||0.32|
Strongly positive staining for p53 also was significant in multivariate analysis using the Cox's proportional regression mode, when survival was adjusted for stage and grade (p < 0.05). These results are shown in Table IV.
|Variables||Beta (SE)||Relative risk (95% CI)||p-value|
|P53 expression||1.55 (0.49)||4.7 (1.81–12.34)||0.002|
|Stage||1.30 (0.31)||3.66 (1.98–6.76)||0.000|
|Grade||0.54 (0.29)||1.7 (0.96–3.06)||0.068|
Comparison of different studies in this area is difficult owing to lack of standardized fixation procedures, protocols, heterogeneous tumor stage, differing antibodies and fixation procedures, determination of tumor cell positivity qualitatively or quantitatively, and differing statistical analyses. Another reason for the difference between the immunohistochemical studies are the misclassified cases, for example endometrial cancers that could have been misclassified as being of endocervical origin. There are a number of immunohistochemical reports about adenocarcinoma of the uterine cervix, but there is no consistency among these studies in the definition of positive and negative staining. Used cut-off values are 1%,23, 24, 25 5%,26 10% or 50%.16, 27, 28 We showed staining results semi-quantitatively from 0–4+ and for analysis used cut-off value off strong positive, which is more that 50% staining. Unlike breast cancer and cancer of the uterine corpus where hormone receptor status is of prognostic significance12, 29, 30 and can determine response to endocrine therapy, the significance of hormone receptor status in adenocarcinoma of the uterine cervix still remains unclear. The expression of estrogen receptor in adenocarcinoma of the uterine cervix has been found to vary between 4%31 and 81%.32 Ten percent of our patients were classified as having tumors strongly positive for the estrogen receptor. This is lower than in many studies, possibly because of differences in tumor stage distribution or different cut-off values for determining hormone receptor positivity like others32, 33, 34 we found no significant difference in survival between those with positive or negative estrogen receptor status. Some groups have reported better disease free survival where estrogen receptor is positive35, 36 although in some studies this was marginal.33
The prevalence of progesterone receptor in adenocarcinomas of the uterine cervix varies between 4%31 and 54%.32 In our study 8% of cases were strongly positive for progesterone receptor. As in other studies32, 33, 36 we found that survival was not influenced by presence or absence of the progesterone receptor, although Masood et al.35 and Suzuki et al.34 found a better survival in patients with positive progesterone receptor. In endometrial cancer high estrogen/progesterone receptor levels are more frequent in well-differentiated adenocarcinomas.37 We did not detect any relation between tumor grade and immunohistochemical staining for hormone receptors.
When DNA damage occurs, the cell cycle is stopped in G1 and DNA repair is carried out prior to cell division. If DNA damage is irreversible, apoptosis is induced. P53 is involved in the regulation of cell proliferation by stimulating the transcription of other specific cell cycle control genes. Cells with inactivated or mutant p53 cannot delay progression from the G1 to the S phase of the cell cycle and thereby cannot prevent the replication of abnormal DNA. Mutations result in a conformational change of the protein, which becomes stabilized, thus allowing for immunohistochemical detection.38 Usually when p53 mutation is present, there is a diffuse intense nuclear positivity. However, lower levels of p53 positivity may occur without mutation, as a result of stabilization of wild-type p53 by nonmutational events.39 It has been suggested that over-expression of wild type p53 gene products may indicate a poorer prognosis. There are conflicting reports about p53 as a prognostic factor in cervical cancer. Some found no relationship between p53 and survival in cervical adenocarcinoma.19, 40, 41, 42, 43, 44 However, in accordance with some other groups,45, 46, 47, 48 we found a significantly worse 5-year disease specific survival if p53 staining was strongly positive. Multivariate analysis showed, adjusted for FIGO stage and grade, a significantly worse survival in patients with strong positive p53 staining.
Ki-67 antibody is found in proliferative cells and is observed during the late G1, S, G2 and M phases of the cell cycle; cells in G0 and early G1 consistently lack reactivity.17 MIB-1 is a monoclonal antibody which reacts with the Ki-67 antigen. MIB-1 is a prognostic factor in breast cancer,49 ovarian cancer,18 squamous cell cervical carcinomas50 and endocervical adenocarcinomas.51 In general a high MIB-1 index reflects a poorer prognosis. Higher proliferative activity of cancer cells is associated with a more aggressive behavior and results in a higher frequency of recurrence or metastases. However, we found no difference in survival with higher MIB-1 index. Suzuki et al.34 and Nakano et al.52 showed a better survival in patients with squamous cell carcinoma of the uterine cervix with higher levels of MIB-1. This study group studied a large homogeneous group of patients with long follow-up. The primary treatment was radiotherapy, and the authors suggest that this treatment may explain their results. Tumors containing a large numbers of proliferating cells, as indicated by a high Ki-67 growth fraction or MIB-1 index, may be more sensitive to radiation therapy, possibly conferring a better prognosis. In our group most patients (84%) were treated by primary surgery. It may be that growth fraction of the tumor does not correlate directly with its biological behavior. Van der Putte et al. found in a large group of early SCC, no differences in survival, but an inverse relation between Ki-67 and both tumor size and stromal invasion.28 It is our interpretation that MIB-1 seems not to be of clinical significance for survival in adenocarcinomas of the cervix.
The prevalence of bcl-2 positivity in cervical adenocarcinomas varies between 27%53 and 61%.42 Using a cutoff value of 50% we found a much lower level of bcl-2 positivity: only 5% of cases. Even when the cutoff was lowered to 1%, the level used by other groups,15, 23 the percentage of bcl-2 positivity remained low, 23%. Unlike other studies where a better19, 44, 45 or worse16, 41, 54 survival was demonstrated for bcl-2 positive tumors, we found no significant difference in survival.
In summary, our study suggests that it is not of clinical significance to determine estrogen receptor, progesterone receptor, MIB-1 or bcl-2 in cervical adenocarcinomas as an adjunct to determine survival. However, determination of p53 seems useful since p53 staining is a marker for survival. p53 positivity appears to be linked to poorer survival in cervical adenocarcinoma, and adjuvant therapy may need to be adjusted.
- 2Dijck vanJAAM,CoeberghJWW,SieslingS,VisserO, eds. Trends of Cancer in the Netherlands 1989–1998. Utrecht: Vereniging van Integrale Kankercentra, 2002.
- 20Pathology and genetics tumours of the breast and female organs. In: TavassoliFA,DevileeP, eds. World Health Organisation classification of tumours. Lyon: IARC Press, 2003. 272–7., .
- 21Tumors of the cervix, vagina and vulva. In: RosaiJ, ed. Atlas of tumor pathology. Washinton, DC: Armed Forces Institute of Pathology, 1992. p 87., , .
- 24Expression of HER-2/neu, epidermal growth factor receptor, vascular endothelial growth factor, cyclooxygenase-2, estrogen receptor, and progesterone receptor in small cell and large cell neuroendocrine carcinoma of the uterine cervix: a clinicopathologic and prognostic study. Int J Gynecol Cancer 2005; 15: 646–56., , , , , , .
- 30Multivariate independent prognostic factors in endometrial carcinoma: a clinicopathologic study in 181 patients: 10 years experience at the Department of Obstetrics and Gynecology of the Mainz University. Int J Gynecol Cancer 2003; 13: 197–203., , , , , , , , .