The objective of the study was to evaluate the pathogenetic and prognostic value of p53 protein expression in squamous cell carcinoma of the vulva.
The objective of the study was to evaluate the pathogenetic and prognostic value of p53 protein expression in squamous cell carcinoma of the vulva.
The clinical data in charts of 167 patients with International Federation of Gynecology and Obstetrics (FIGO) Stages I-III primary tumors who were treated by surgery were reviewed. Samples from the primary tumor were immunostained for p53 protein. p53 overexpression was defined as immunoreactivity in > 5% of nuclei.
p53 overexpression was observed in 92 tumors (55%). p53 overexpression did not correlate with age at diagnosis, FIGO stage, histologic grade, vessel invasion, tumor thickness, tumor greatest dimension, DNA ploidy, or inguinal lymph node metastasis. In the whole group a significantly reduced 5-year corrected survival was observed in patients with p53 overexpression compared with p53 negative patients (P = 0.04). In the different FIGO stages, disease-related survival was not influenced by p53 overexpression in 37 patients with Stage I disease (P = 0.60) or in 86 patients with Stage II disease (P = 0.96). In 44 patients with Stage III disease, p53 overexpression was significantly associated with poorer prognosis (P = 0.004). Independent prognostic factors for corrected survival in the entire group of 167 patients were: vascular invasion, groin metastasis, tumor greatest dimension, and p53 overexpression. In patients with FIGO Stage III disease p53 overexpression was not an independent prognostic factor.
p53 protein overexpression appears to be involved in the pathogenesis of vulvar squamous cell carcinoma. p53 protein overexpression was significantly associated with disease-related survival. p53 prognostic impact was observed only in patients with advanced disease. Cancer 1999;85:1133–8. © 1999 American Cancer Society.
To our knowledge, mutations in the p53 gene are the most frequent molecular events described in human tumors thus far.1 p53 is a tumor suppressor gene located on the short arm of chromosome 17. It is involved in the regulation of the cell cycle and the triggering of apoptosis. The mutant p53 protein has a longer half-life than the wild-type p53, and overexpression of the mutant p53 protein, associated with mutation of the p53 gene, can be detected by immunohistochemistry.2 p53 abnormalities may have diagnostic, prognostic, and therapeutic implications. In squamous cell carcinoma of the genital tract, p53 mutation appears to play a role in the pathogenesis, but most likely not in the progression, of cervical, vaginal, and vulvar carcinomas.3–10 Some authors also have found p53 protein expression to be of prognostic value.8, 10 However, the number of patients in the majority of published series is too small to draw definite conclusions. The objective of the current study was to investigate the possible pathogenetic and prognostic value of p53 protein expression and its relation to other clinical, histopathologic, and flow cytometric (FCM) parameters in squamous cell carcinoma of the vulva.
During the 10-year period between 1982-1991, 215 patients with squamous cell carcinoma of the vulva were admitted to the Norwegian Radium Hospital for primary treatment. The hospital records were reviewed for clinical data. All lesions were staged retrospectively according to the 1971 International Federation of Gynecology and Obstetrics (FIGO) system.11 Included in the current study were patients with Stage I-III disease who primarily were treated by surgery (N = 192 patients).
Formalin fixed, paraffin embedded sections from the primary tumors were stained with hematoxylin and eosin, alcian blue, and periodic acid-Schiff stain. Paraffin embedded material was prepared by the avidin-biotin-peroxidase complex method for immunohistochemical studies with antibodies against p53 protein. Two different antisera against p53 protein were used, one polyclonal (NCL-CM1; Novocastra Laboratory Ltd., UK) and one monoclonal (DO-1; Scandinavian Diagnostic Services, Denmark). Reactions were performed with and without pretreatments with microwaves in citrate buffer (0.01 M, pH 6.0, twice for 5 minutes each).The method for p53 analysis was described by Skomedal et al.3 All series included positive and negative controls. All controls gave satisfactory results.
Histologic specimens were reviewed by one of the authors (J.M.N) who had no access to clinical information. The tumors were classified and graded according to World Health Organization recommendations. Depth of tumor invasion was measured by an ocular micrometer from the top of the nearest dermal papilla to the deepest point of stromal penetration. Vessel invasion was recorded as positive only when tumor cells were observed within unquestionably endothelium-lined spaces.
FCM DNA measurements were performed on paraffin embedded samples from the primary tumor. The tissue was dewaxed with xylol and rehydrated according to the method described by Hedley et al.12 with minor modifications. The method and DNA ploidy definitions are described by Kaern et al.13 Follow-up information was retrieved from the hospital records, from the Cancer Registry of Norway, and from the death certificates. All patients were followed until death or until January 1, 1996.
Difference in proportions were evaluated by the chi-square test or Fisher's exact test, whichever was appropriate. Disease-related survival was calculated from the start of treatment until death from disease or January 1, 1996, using the method of Kaplan and Meier.14 The log rank test15 was used to compare survival rates. Cox proportional hazards analysis was used to identify the independent prognostic factors.16 A backward selection procedure was performed. The analysis was verified by computing the log cumulative hazard against time. Statistical significance was considered at P < 0.05.
Successful p53 protein analyses were obtained in 167 of 192 cases (87%): 37 of 53 Stage I cases, 86 of 90 Stage II cases, and 44 of 49 Stage III cases. The relative high frequency of unsuccessful analyses in Stage I cases reflects an inability to obtain enough material from the very small tumors. p53 protein overexpression, defined as immunoreactivity in > 5% of nuclei, was observed in 92 of 167 tumors (55%), 29 with 5-50% of cells with immunoreactive nuclei and 63 with >50% of cells with immunoreactive nuclei. Seventy-five tumors (45%) were p53 negative. Correlations between p53 overexpression and the patients' clinical, histopathologic, and FCM characteristics are presented in Table 1. No correlation was observed between p53 protein overexpression and the different parameters analyzed in this study.
|Factor||No. of patients||No. of cases (%) with positive p53 staining||P value|
|< 60||29||14 (48)|
|≥ 80||38||24 (63)|
|Tumor dimension (cm)||0.508|
|≤ 2||37||20 (54)|
|> 2–≤ 5||95||55 (58)|
|> 5||31||14 (45)|
|Tumor thickness (mm)||0.172|
|≤ 1||17||9 (53)|
|> 1–≤ 3||24||7 (29)|
|> 3–≤ 5||26||17 (65)|
|> 5–≤ 7||31||19 (61)|
|> 7||68||39 (57)|
|Not performed||32||18 (56)|
|Inguinal lymph node metastases||0.386|
|Not assessed||31||18 (58)|
Surgical treatment was local excision and/or vulvectomy in 32 patients, and radical vulvectomy with unilateral or bilateral inguinal lymph node dissection in 135 patients (Table 1). Groin metastases were demonstrated histologically in 49 of the 135 patients who underwent inguinal lymph node dissection (36%). Thirty-two patients had unilateral and 17 patients had bilateral groin metastases. In addition, metastatic disease to one groin was verified cytologically in one patient who underwent vulvectomy. Postoperative external irradiation was given to 31 of the patients with proven groin metastases and to 2 patients with narrow surgical margins in the vulva. In five patients external irradiation to the inguinal lymph nodes was preferred to inguinal lymph node dissection due to the patient's compromised medical condition.
Three patients died of postoperative complications and 7 patients (4%) never achieved disease free status. Of the remaining 157 patients, 46 (28%) developed recurrences: 26 of the 84 p53 positive cases (31%) and 20 of the 73 p53 negative cases (27%) (P = 0.7). The mean recurrence free survival was 47 months and 42 months, respectively, in the p53 positive and p53 negative patients (P = 0.1615). When analyzing each FIGO stage separately, no difference in recurrence free survival was found between patients with p53 positive tumors and those with p53 negative tumors in patients with Stage I and Stage II disease. Patients with Stage III disease with p53 positive tumors had a significantly poorer recurrence free survival compared with p53 negative patients (P = 0.0210).
Treatment in the patients with recurrent disease was comprised of surgery in 7 patients, external irradiation in 13 patients, chemotherapy in 5 patients, and a combination of all 3 treatment modalities in 9 patients. Twelve patients received no treatment. Eight of the patients with recurrent disease were salvaged by secondary therapy.
At the end of follow-up 86 patients were alive without evidence of disease. Three patients had died of postoperative complications and 33 patients had died of unrelated causes, 2 of whom had been treated successfully with surgery for recurrent disease. Forty-five patients had died of vulvar squamous cell carcinoma. For the 86 patients who were alive at the end of follow-up, the median observation time was 92 months (range, 51-168 months).
Figure 1 shows the 5-year corrected survival according to p53 expression in the whole group of 167 patients. The patients with p53 overexpression had a significantly worse outcome compared with the patients with p53 negative tumors (P = 0.039). The 5-year corrected survival according to p53 expression in patients with Stage I disease is shown in Figure 2 , and the 5-year corrected survival in patients with Stage II disease is shown in Figure 3. p53 overexpression had no impact on disease-related survival in these two groups of patients. However, patients with Stage III disease with p53 overexpression had a significantly worse prognosis than the p53 negative patients (P = 0.0035) (Fig. 4).
Metastatic disease to the lymph nodes was demonstrated in 49 of 135 patients who underwent inguinal lymph node dissection. The prognosis was significantly worse in lymph node positive patients compared with lymph node negative patients (P < 0.0001), but p53 expression did not influence outcome in either lymph node positive or lymph node negative patients (Fig. 5).
The probability of 5-year corrected survival related to the prognostic factors and the corresponding P values from univariate analyses in the entire group of 167 patients are shown in Table 2.
|Factor||No. of patients||5-year survival||P value|
|Tumor dimension (cm)||0.0573|
|> 2–≤ 5||95||0.625|
|Tumor thickness (mm)||0.0137|
|> 1–≤ 3||24||0.738|
|> 3–≤ 5||26||0.782|
|> 5–≤ 7||31||0.582|
|Vessel invasion||< 0.0001|
|Inguinal lymph node metastases||< 0.0001|
When tested in Cox multivariate analysis independent prognostic factors in the entire group of 167 patients were vessel invasion, inguinal lymph node metastases, maximum tumor dimension, and p53 expression. One hundred fifty-five patients were available for analysis (Table 3 ). Stage was not included in the multivariate analysis because the classification is based on maximum tumor dimension and metastatic disease. In patients with Stage III disease only vessel invasion was found to be an independent prognostic factor (Table 3). In this group 42 patients were available for the multivariate analysis.
|Factor||P value||95% confidence interval||Relative risk|
|167 patients with FIGO Stage I–III|
|Inguinal lymph node metastases||0.003||1.136–1.828||1.439|
|44 patients with FIGO Stage III|
p53 mutations and p53 protein overexpression have been reported in 24-86% of patients with squamous cell carcinoma of the vulva.5–10 In the current study p53 overexpression was demonstrated in 55% of the cases, supporting the suggestion that p53 mutation is involved in the neoplastic process. We found no correlation between p53 overexpression and stage of disease, indicating that p53 mutations are oncogenic events early in carcinogenesis. This is in accordance with Milde-Langosch et al.,6 who found p53 mutations occurring frequently in Stage I disease; they, like McConnell et al.,9 demonstrated a high proportion of p53 mutations, even in precancerous lesions.
Some authors have claimed that p53 aberrations are associated with inguinal lymph node metastases,5, 7, 8 a finding not confirmed by others.17 In our series no correlation was found between p53 overexpression and metastatic disease. However, because the sample size was relatively small, a significant difference may exist that was not detected. In the report by McConnell et al.9 well and moderate tumor differentiation was significantly more likely to be associated with positive p53 immunostaining than poor tumor differentiation. Like Kohlberger et al.,8 we found no correlation between tumor grade and p53 overexpression. To our knowledge, with the exception of inguinal lymph node metastases and tumor differentiation, a possible association between p53 protein expression and other well known clinical and histopathologic parameters has not been evaluated in previous reports. We tested for age at diagnosis, maximum tumor dimension, tumor thickness, vessel invasion, and DNA ploidy and found no significant correlation between p53 protein expression and any of these parameters.
To our knowledge, few series exist concerning the possible prognostic value of p53 protein expression in vulvar squamous cell carcinoma. In a series of 38 patients Sliutz et al.10 reported a significantly shorter recurrence free survival and overall survival in patients with tumors bearing p53 mutations compared with patients without p53 mutations, and Kohlberger et al.8 found p53 overexpression was correlated significantly with a reduced overall survival rate in 25 patients with Stage I-II disease. However, in their study of 115 patients McConnell et al.9 found no prognostic impact of tumor p53 expression. In our study p53 expression influenced outcome significantly; patients whose tumors overexpressed p53 had the poorest prognosis. However, the prognostic impact was linked to advanced disease. In early disease (i.e., Stages I and II) p53 protein expression did not influence recurrence free survival or disease-related survival.
As the treatment of vulvar carcinoma has evolved during the last 20 years toward more conservative and individualized surgery, there is a need for prognostic factors to identify patients suited for less than radical procedures. According to the results of the current study p53 expression adds no information with regard to treatment selection in patients with early disease. Patients with clinical Stage III disease have a relatively poor prognosis even if radical treatment is used. We found that patients with p53 negative tumors had a significantly better prognosis compared with patients whose tumors showed p53 overexpression in Stage III disease. However, the number of patients in our series is too small to recommend less than radical treatment in any group of patients with advanced vulvar squamous cell carcinoma.