Summary

Recently, various cell cycle regulators have been investigated as biological markers of malignant potential. These regulators might influence the survival rate and the effect of adjuvant therapies. In this study, we analyzed p53, p21(Waf1/Cip1) and cyclin D1 expression in 64 esophageal cancer patients and the relationship between clinicopathologic parameters and patient survival. The positive expression rate was 48.4%, 42.2% and 43.8% in the p53, p21 and cyclin D1 groups respectively. Multivariant analysis revealed that tumor depth, chemotherapy, p53, p21 and cyclin D1 expression showed significant values. p53- and cyclin D1-negative patients had a worse prognosis. p21-positive patients had a better prognosis. In stage 0, I and II patients, there was a significant difference between p53-positive and -negative, p21-positive and -negative, and cyclin D1-positive and -negative groups. In stage III and IV patients, there was no significant difference between any two groups. However, a significant difference was seen in the p21 group: among patients who received adjuvant chemotherapy, the p21-positive group had a 5-year survival rate of 50% compared with 13.4% in the p21-negative group (not significant).

Introduction

Many trials of adjuvant therapies and various operative procedures have been performed to improve the survival rate of esophageal cancer patients, however clear successes have not been obtained to date. Recently, various cell cycle regulators have been investigated as biological markers of potential malignancy. These cell regulators might influence the survival rate and the effect of the adjuvant therapies. Therefore, the strategy for esophageal cancer treatments must be reconsidered from the viewpoint of these new biological markers.

In this study, we investigated p53, p21 (Waf1/Cip1) and cyclin D1 expression in 88 esophageal cancer patients, and in 64 patients the relationship between clinicopathologic parameters and patient survival was analyzed. Moreover, the effects of chemo- and radiotherapy were re-estimated.

Patients and Methods

Eighty-eight esophageal cancer patients who had undergone resection from 1986 to 1996 were investigated. Of these, 64 patients, excluding palliatively resected cases and mucosal cancer cases, were analyzed. They were described according to the guidelines for the clinical and pathological studies on carcinoma of the esophagus.1

For immunohistochemical investigations, the avidin–biotin–peroxidase complex (ABC) method was performed. In p53, a monoclonal mouse anti-human p53 protein antibody (Do-7, Dako Denmark) was used; in p21 (Waf1/Cip1), a monoclonal mouse anti-human p21 protein antibody (purified mouse anti-human Sdi1 (p21) monoclonal antibody, Pharmingen, San Diego, CA, USA), and in cyclin D1 a monoclonal mouse anti-human cyclin D1 protein antibody (purified mouse anti-human cyclin D1 monoclonal antibody, Pharmingen). All slides were examined by a single pathologist who was unaware of the patients’ clinical status. At ×40 magnification, 1000 random cells per slide were examined for nuclear staining. Their overexpression was arbitrarily defined as positive when 10% of cells examined were stained.

Statistical analysis

Survival rates were calculated by Kaplan–Meier analysis. Differences between patient groups were tested by the Cox–Mantel test. Characteristic distribution in each group was compared using the chi-squared test. The multivariate relationship was analyzed according to Cox’s proportional hazards general linear model. Differences with p-values <0.05 were considered significant.

Results

Positive rates of expression in the p53, p21 and cyclin D1 groups were 48.4%, 42.2% and 43.8% respectively. Characteristics of positive and negative groups in each group showed no significant difference. There was no increase in positive rate by stage.

Table 1 shows the multivariate analysis. Tumor depth, chemotherapy, p53, p21 and cyclin D1 were significant. p53- and cyclin D1-negative patients had a worse prognosis. p21-positive patients showed better prognosis.

Table 1

. Multivariate analysis of esophageal cancer patients who underwent resection

Variable Hazard ratio 95% Confidence interval p-Value 
Age 1.24190 0.6807–2.2657 0.48332 
Sex 0.40981 0.0820–2.0465 0.28217 
Approach 0.53742 0.2479–1.1649 0.12197 
Histology 0.89288 0.5386–1.4802 0.66133 
Depth 2.85405 1.5154–5.3751 0.00209 
2.87311 0.8029–10.2801 0.11096 
Curability 1.06467 0.4430–2.5586 0.88916 
Complication 1.29864 0.6044–2.7902 0.50614 
Chemotherapy 5.59328 2.2151–14.1230 0.00064 
Radiotherapy 1.23027 0.5617–2.6943 0.60664 
p53 1.73097 0.7677–3.9024 0.19188 
p21 3.40878 1.3878–8.3725 0.01008* 
Cyclin D 2.46941 1.0053–6.0652 0.04972** 
Variable Hazard ratio 95% Confidence interval p-Value 
Age 1.24190 0.6807–2.2657 0.48332 
Sex 0.40981 0.0820–2.0465 0.28217 
Approach 0.53742 0.2479–1.1649 0.12197 
Histology 0.89288 0.5386–1.4802 0.66133 
Depth 2.85405 1.5154–5.3751 0.00209 
2.87311 0.8029–10.2801 0.11096 
Curability 1.06467 0.4430–2.5586 0.88916 
Complication 1.29864 0.6044–2.7902 0.50614 
Chemotherapy 5.59328 2.2151–14.1230 0.00064 
Radiotherapy 1.23027 0.5617–2.6943 0.60664 
p53 1.73097 0.7677–3.9024 0.19188 
p21 3.40878 1.3878–8.3725 0.01008* 
Cyclin D 2.46941 1.0053–6.0652 0.04972** 

n = 64, Excluding pallatively resected and mucosal cancer cases.

*

Positive cases showed better prognosis.

**

Negative cases showed worse prognosis.

The combined analysis with p53 and p21 expression in the same esophageal cancer specimens revealed that 5-year survival rates were 53%, 33.2%, 28.5% and 27.5% in patients with p53−/p21+, p53+/p21+, p53+/p21− p53−/p21− tumors (Fig. 1). There was a significant difference between p53−/p21+ and p53−/p21−.

Fig. 1

—Survival curves according to p53 and p21 relationship in all cases: N-P, p53 negative/p21 positive (n=11); P-P, p53 positive/p21 positive (n=16); P-N, p53 positive/p21 negative (n=15); N-N, p53 negative/p21 negative (n=22)

Fig. 1

—Survival curves according to p53 and p21 relationship in all cases: N-P, p53 negative/p21 positive (n=11); P-P, p53 positive/p21 positive (n=16); P-N, p53 positive/p21 negative (n=15); N-N, p53 negative/p21 negative (n=22)

Table 2 shows 5-year survival rates in all cases. A significant difference was revealed between the p53-positive and -negative groups, and cyclin D1-positive and-negative groups, with negative groups showing better survival. Table 3 shows 5-year survival rates in stage 0, I and II patients. A significant difference was revealed between p53-positive and -negative groups, p21-positive and -negative groups, and cyclin D1-positive and -negative groups. Table 4 shows 5-year survival rates in stage III and IV cases, without a significant difference between any two groups. However a great difference was seen in the p21 group: among patients who received adjuvant chemotherapies using cisplatin or bleomycin, the p21-positive group had a 5-year survival rate of 50% compared with 13.4% in p21-negative group (not significant) (Fig. 2).

Table 2

. Five-year survival rates in all cases

 Immunoreactivity Case number Five-year survival rate (%) p-Value 
p53 31 31.9 <0.01 
 – 33 35.6  
p21 27 42.4 NS 
 – 37 25.9  
Cyclin D 28 28.1 <0.01 
 – 36 35.4  
 Immunoreactivity Case number Five-year survival rate (%) p-Value 
p53 31 31.9 <0.01 
 – 33 35.6  
p21 27 42.4 NS 
 – 37 25.9  
Cyclin D 28 28.1 <0.01 
 – 36 35.4  
Table 3

. Five-year survival rates in stages 0, I, II

 Immunoreactivity Case number 5-Year survival rate (%) p-Value 
p53 16 31.7 <0.01 
 – 10 67.5  
p21 10 54.0 <0.05 
 – 16 46.3  
Cyclin D 11 34.1 <0.01 
 – 15 56.6  
 Immunoreactivity Case number 5-Year survival rate (%) p-Value 
p53 16 31.7 <0.01 
 – 10 67.5  
p21 10 54.0 <0.05 
 – 16 46.3  
Cyclin D 11 34.1 <0.01 
 – 15 56.6  
Table 4

. Five-year survival rates in stages III, IV

 Immunoreactivity Case number 5-Year survival rate (%) p-Value 
p53 15 29.1 NS 
 – 23 19.9  
p21 17 37.6 NS 
 – 21 9.2  
Cyclin D 17 24.4 NS 
 – 21 21.2  
 Immunoreactivity Case number 5-Year survival rate (%) p-Value 
p53 15 29.1 NS 
 – 23 19.9  
p21 17 37.6 NS 
 – 21 9.2  
Cyclin D 17 24.4 NS 
 – 21 21.2  
Fig. 2

—Survival curves of chemotherapy group according to p21 in stages III and IV.

Fig. 2

—Survival curves of chemotherapy group according to p21 in stages III and IV.

Discussion

The altered p53 expression detected by immunohistochemistry is considered to reflect the nuclear accumulation of mutant p53 protein, which is encoded by the p53 gene with missense mutations and has a prolonged half-life.2 However, despite the large number of published investigations, there are no unequivocal examples of tumors in which p53 gene mutations and/or p53 protein accumulation provide prognostic information. The reason for the discrepancies in the literature depends, at least in part, on the different techniques used to detect p53 aberrations in human tumors. Most of the investigations that attempted to assess the prognostic significance of p53 aberrations relied on the demonstration of p53 protein accumulation using immunohistochemistry techniques. Unfortunately, it is almost impossible to compare the results of many studies because of the use of different antibodies and interpretation criteria for the evaluation of tumors.3 Greenblatt et al4 reviewed 84 studies in which both immunohistochemical techniques and sequence analysis were performed on the same tumors. Although 44% of cases showed p53 immunoreactivity, only 36% displayed p53 gene mutations.4 Coggi et al5 reported that p53 protein accumulation was demonstrated in 41 (57%) tumors, but p53 gene mutations were detected in 39 (53%) tumors. These results supports the theory that the detection of p53 protein expression might not reflect the p53 gene mutation exactly.

In our study, p53 protein accumulation did not correlate with age, gender, location, histology, stage, curability, lymphatic invasion or blood vessel invasion. Multivariate analysis did not reveal any prognostic value, however in early stages of 0, I and II, p53 was the most import prognostic indicator.

Ogawa et al6 also reported that gastric cancer patients with p21-positive tumors had a significantly better prognosis than those with p21-negative tumors. Aoyagi et al7 reported that the 3-year survival rate of the p53+/p21− group was significantly lower than either that of the p21+/p53− or the p53−/p21− groups. By analyzing both p53 and p21 (Waf1/Cip1) expression in the same tumor specimens, we revealed that the p53−/p21+ group showed the best prognosis compared with other patients with esophageal cancer.

In multivariate analysis of survival, p21(Waf1/Cip1) expression was significantly correlated with prognosis. In advanced stages of III and IV, p21- positive patients showed much better prognosis than p21-negative patients, but without significant difference because of the small number of patients. Moreover, in p21-positive patients, chemotherapy was more effective than in p21-negative patients. These cell cycle regulators might be useful as a chemosensitivity test in the future.8

Callender et al9 reported that cyclin D1 amplification correlated significantly with elevated proliferative activity and noted a tendency for amplification to occur in poorly differentiated, high-stage and aneuploid tumors. Our results showed that there was poor correlation with cyclin D1 expression and prognosis in esophageal cancer patients.

In conclusion, we noted that (1), in multivariate analyses, p21 (Waf1) expression in esophageal cancer was significantly associated with the prognosis; (2) in the early stages (stages 0, I and II), p53, p21 (Waf1/Cip1) and cyclin D1 protein expression was significantly correlated with prognosis; (3) advanced-stage (stages III and IV), p21-positive patients showed much better prognosis than p21-negative patients (not significant); (4) by analyzing both p53 and p21 expression in the same tumor specimens, the p53−/ p21+ group showed the best prognosis compared with other patients; (5) chemotherapy was more effective in p21-positive patients than in p21-negative patients.

References

1
Japanese Society for Esophageal Diseases.
Guidelines for the clinical and pathological studies on carcinoma of the esophagus, 8th edn. Tokyo: Kanehara Shuppan,
1992
.
2
Levine
A J
,
Momand
J
,
Finlay
C A
.
The p53 tumor suppressor gene
.
Nature
 
1991
;
351
:
453
456
.
3
Bosari
S
&
Viale
G
.
The clinical significance of p53 aberrations in human tumors
.
Virchow Arch
 
1995
;
427
:
229
241
.
4
Greenblatt
M S
,
Bennett
W P
,
Hollstein
M
,
Harris
C C
.
Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis
.
Cancer Res
 
1994
;
54
:
4855
4878
.
5
Coggi
G
,
Bosari
S
,
Roncalli
M
, et al.
p53 protein accumulation and p53 gene mutation in Esophageal carcinoma – a molecular and immunohistochemical study with clinicopathologic correlations
.
Cancer
 
1997
;
79
:
425
432
.
6
Ogawa
M
,
Maeda
K
,
Onoda
N
, et al.
Immunohistochemical study of p21 expression in gastric carcinoma
.
Gan To Kagaku Ryoho
 
1997
;
24
:
292
295
.
7
Aoyagi
K
,
Koufuji
K
,
Kodama
I
, et al.
The significance of p21 expression in gastric cancer
.
Gan To Kagaku Ryoho
 
1997
;
24
:
285
291
.
8
Hickey
K
,
Grehan
D
,
Reid
I M
,
O’briain
S
,
Walsh
T N
,
Hennesy
T P J
.
Expression of epidermal growth factor receptor and proliferating cell nuclear antigen predicts response of esophageal squamous cell carcinoma to chemoradiotherapy
.
Cancer
 
1994
;
74
:
1693
1698
.
9
Gaffey
M J
,
Frierson
H F
,
Williams
M E
.
Chromosome 11q13, c-erbB-2, and c-myc amplification in invasive breast carcinoma: Clinicopathologic correlations
.
Mod Pathol
 
1993
;
6
:
654
659
.

Author notes

*
Presented at the VIIth World Congress, International Society for Diseases of the Esophagus, Montreal, September 1998.