- Email: [email protected]
Kip1 expressions are potentially poor prognostic factors in lung adenocarcinoma patients
Lung Cancer 34 (2001) 59 – 65 www.elsevier.com/locate/lungcan
High cyclin E and low p27/Kip1 expressions are potentially poor prognostic factors in lung adenocarcinoma patients Hiroyuki Hayashi a,*, Nobuo Ogawa b, Naoki Ishiwa b, Takuya Yazawa a, Yoshiaki Inayama a, Takaaki Ito a, Hitoshi Kitamura a a
Department of Pathology, Yokohama City Uni6ersity School of Medicine, 3 -9 Fukuura, Kanazawa-ku, Yokohama, 236 -0004, Japan b Di6ision of Pulmonary Surgery, Kanagawa Cardio6ascular and Respiratory Centre, 6 -16 -1 Tomiokahigashi, Kanazawa-ku, Yokohama, 236 -0051, Japan Received 21 November 2000; received in revised form 15 February 2001; accepted 27 February 2001
Abstract Cyclin E is an important regulator of entry into the S phase of the cell cycle. p27/Kip1 (p27) binds to cyclin E/Cdk2 complex and negatively regulates cell proliferation. We immunohistochemically examined the expression of cyclin E and p27 in 98 cases of resected lung adenocarcinoma to evaluate the prognostic significance of cyclin E and p27. Cyclin E was expressed in 16 cases (16%), and p27 was expressed in 41 cases (42%). Using Kaplan– Meier survival analysis, patients with cyclin E positive (P= 0.0017) and p27 negative (P= 0.011), both individually and in combination (P B0.0001), had a worse prognosis. We also analyzed the relationship of these findings to clinicopathological parameters, which revealed that cyclin E-positive, p27-negative cases had a higher Ki67 expression (P = 0.012) and a higher rate of lymph node metastasis (P =0.0078) than other groups. Our results suggested that cyclin E over expression, in association with p27 reduction in particular, may potentially be a poor prognostic factor in lung adenocarcinoma patients. However, to verify the prognostic significance of these factors, a multivariate analysis of a larger number of patients should be undertaken. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cyclin E; p27/Kip1; Ki67; Lung; Adenocarcinoma; Survival
1. Introduction Lung cancer is one of the leading causes of cancer death throughout the world. Among the heterogeneous histological subtypes, the incidence * Corresponding author. Tel.: + 81-45-787-2583; fax: + 8145-789-0588. E-mail address: [email protected] (H. Hayashi).
of adenocarcinoma (AC) is increasing [1]. Recent studies of cell-cycle regulatory proteins have contributed to elucidating the development and extension of cancer. Cyclins, cyclin-dependent kinases (Cdks), and Cdk inhibitors play crucial roles in the passage of checkpoints in the cell cycle. Cyclin E is associated with Cdk2 in an active complex in the late G1 phase and regulates progression from G1 into the S phase [2]. Overexpression of cyclin E has been reported in lung
0169-5002/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 1 ) 0 0 2 1 1 - 2
60
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
cancer and other various organs [3– 11]. To date, there are three reports in which the expression of cyclin E and patients’ survival were analyzed in non-small cell lung cancer (NSCLC). Dobashi et al. reported those patients whose tumors were cyclin E positive showed significantly higher survival rate compared with those without cyclin E expression [3]. To the contrary, Fukuse et al. and Mishina et al. reported that high expression of cyclin E in lung cancer tissue correlates with a poor prognosis [4,9]. NSCLC consists of heterogeneous subtypes. Squamous cell carcinoma (SQCC) and AC have different aspects in their histogenesis, gender trend, p53 mutation spectrum, and etiology particularly the relationship to smoking [12]. In studies of earlier survival analysis on NSCLC, cyclin E expression was significantly lower in AC than in SQCC [3,9]. Moreover, cyclin E expression has been reported to be observed in premalignant lesions of SQCC [13], however, such observation was not found in precancerous lesion of AC [5]. We consider that it is important to analyze the prognostic significance of cyclin E in the respective histological subtypes because of those distinct differences. p27/Kip1 (p27), one of the Cdk inhibitors, binds to the cyclin E/Cdk2 complex and negatively regulates cell proliferation [14]. Recently, several studies have indicated that low levels of p27 expression are associated with a poor prognosis and an advanced grade in human neoplasms of the lung as well as those of various other organs [15–18]. However, the prognostic significance of p27 and cyclin E in combination has not been well studied to date. Porter et al. reported that in breast cancer patients, the combination of high p27 and low cyclin E showed the most favorable prognosis, and the converse was associated with the highest mortality [19]. We earlier reported that high p27 expressers showed lower Ki67 labeling index values (LI) than low p27 expressers among cyclin E-positive lung AC patients [5]. In the present study, we analyzed the expression of cyclin E and p27 immunohistochemically to evaluate whether their expression correlated with 5-year survival in lung AC, either individually or in combination. We also compared expression of
these proteins with other clinicopathological parameters.
2. Materials and methods
2.1. Patients and tissues Of the consecutive 149 lung AC patients who were surgically treated between 1987 and 1995 in Kanagawa Cardiovascular and Respiratory Centre, 128 patients underwent complete tumor resection with mediastinal lymph node dissection. Those who died within 3 months after surgery, those who died of causes other than lung cancer within 5 years, and those who were followed for less than 5 years (for surviving patients) were excluded from this study. Patients with tumors on whom we could not perform immunohistochemical evaluation were also excluded. Thus, the subjects were 98 patients, consisting of 42 females and 56 males whose ages at surgery ranged from 31 to 81 years. They had not undergone chemotherapy or radiotherapy preoperatively. Six cases presented in our earlier report [5] were included.
2.2. Immunohistochemistry For immunohistochemistry, we used mouse primary monoclonal antibodies against p27 (57; Transduction Lab., Lexington, KY) at a 1:300 dilution, Ki67 (MIB-1; Immunotech, Marseilles, France) at a 1:200 dilution, and cyclin E (13A3; Novocastra Lab., Newcastle, UK) at a 1:25 dilution. After deparaffinization, microwave pre-treatment was performed in 10 mmol/l citrate buffer (pH 6.0) for 15 min at 95°C. The sections were incubated overnight at 4°C with the primary antibodies, then processed with a streptavidin– biotin–peroxidase detection kit (LSAB2 kit; Dako) in which diaminobenzidine was used as the chromogen. Cells were considered positive when the nucleus was distinctly stained. LI was determined in each lesion for each of the antibodies by counting 2000 cells. Cases showing \10% of tumor cells positive for p27, or showing \ 5% for cyclin E were defined as positive for each according to
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
the criteria used in the earlier studies [3,5]. Lung cancer and breast cancer specimens known to express these antigens were used as positive controls.
2.3. Statistical analysis The associations between the expression of those antigens and the clinicopathological parameters were determined by Fisher’s exact test. Kaplan–Meier survival analysis with log-rank test was used to evaluate the relationships between expressions of those antigens and survival distributions. Cox’s proportional hazard analysis was used for multivariate analysis. P values less than 0.05 were considered significant.
3. Results
3.1. Clinicopathological findings The patients consisted of 42 females and 56 males whose ages at surgery ranged from 31 to 81
61
years with a mean 9 S.D. of 62.79 10.4 years (Table 1). There were 49 smokers and 41 nonsmokers, and the smoking history of eight patients was unknown. Forty-six cases were well differentiated AC, 33 were moderately differentiated, and 19 were poorly differentiated. pT and pN classification for each case was based on the TNM classifications of American Joint Committee on Cancer and Union Internationale Contre le Cancer [20].
3.2. Immunohistochemistry The results were summarized in Table 1 and Fig. 1. There were no cells positive for cyclin E in non-neoplastic lung tissues. Cyclin E expression was found in the nuclei of tumor cells in 16 cases (16%, Fig. 1a and b). Among positive cases, 6– 68% of the cells were positive for cyclin E mostly with moderate to strong intensity. p27 was expressed in the nuclei of lymphocytes, fibroblasts, and normal bronchiolar epithelial cells, which served as internal positive control. In neoplastic lesions, p27 immunostaining was
Table 1 Clinicopathological characteristics of patients according to the results of immunohistochemistrya n
Cyclin E Positive (%)
p27
Ki67
P
Positive (%)
P
High expressor (%)
P
Gender Female Male
42 56
7 (16) 9 (17)
ns
17 (42) 24 (43)
ns
14 (33) 22 (39)
ns
Smoking history Smokers Non-smokers Unknown
49 41 8
9 (18) 5 (12) 2 (25)
ns
20 (41) 20 (49) 1 (13)
ns
21 (43) 11 (27) 4 (50)
ns
Differentiation Well Moderately Poorly
46 33 19
2 (4) 8 (24) 6 (31)
0.011* 0.006**
26 (57) 13 (39) 2 (11)
0.0005** 0.026***
9 (20) 14 (42) 13 (68)
0.026* 0.0003**
pT classification 1 and 2 3 and 4
81 17
12 (15) 4 (24)
ns
37 (46) 4 (24)
ns
28 (35) 8 (47)
ns
pN classification 0 1B
64 34
6 (9) 10 (30)
0.013
29 (45) 12 (35)
ns
18 (28) 18 (53)
0.014
a
Abbreviations: ns, not significant; *, well versus moderately; **, well versus poorly; ***, moderately versus poorly.
62
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
Fig. 1. Immunohistochemistry of cyclin E (a and b), p27 (c and d), and Ki67 (e and f). Cyclin E-positive tumor (a) without p27 expression (c) showed high Ki67 LI (e), while cyclin E-negative (b), p27-positive (d) tumor showed low Ki67 LI (f). Scale bar =20 mm.
mostly confined to the cell nuclei and was usually strong (Fig. 1c and d), but a few cases showed cytoplasmic staining. Forty-one cases (42%) were positive for p27, whose LI varied from 11 to 87%. Very few cells were positive for Ki67 in nonneoplastic lung tissues. In tumors, distinct nuclear staining was observed, if present. The mean LI of Ki67 in tumors used in this study was 12.69 14.1% (mean9S.D.), which we defined as the cut-off level on statistical analysis. According to this criterion, 36 cases (37%) showed high Ki67 expression (Fig. 1e and f). Cyclin E-positive cases showed higher Ki67 LI than negative cases (P =
0.0001). There was no definite correlation between p27 and Ki67 expression. Comparing the expression of those proteins to clinicopathological parameters, Ki67 and cyclin E expression increased and p27 expression decreased parallel to the grade of AC (Table 1).
3.3. Prognostic significance of cyclin E and p27 Patients with cyclin E-positive tumors had a significantly worse prognosis than those with cyclin E-negative tumors (P= 0.0017, Fig. 2a). Patients with p27-positive tumors had a significantly
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
63
better prognosis than those with p27-negative tumors (P=0.011, Fig. 2b). Since these proteins interact in the G1/S phase, we analyzed the prognosis by combination of both factors. Tumors positive for cyclin E and negative for p27 (n = 11) had a worse prognosis than the cyclin-E negative, p27-positive group (n =36, P B0.0001, Fig. 2c), both negative group (n =46, P =0.016) and both positive group (n= 5, P = 0.12). Moreover, comparing the cyclin E-positive, p27-negative group with the other three groups combined, the former had a significantly worse prognosis (P B 0.0001).
3.4. Multi6ariate prognostic factor analysis The results of 98 patients were summarized in Table 2. pT and pN classifications and Ki67 expression were significant factors. However, neither cyclin E and p27 expressions nor their combination were significant factors. Therefore, we analyzed the correlation between their combination status and other factors. This analysis revealed that cyclin E and p27 status correlate with Ki67 expression (P =0.012) and pN classification (P =0.0078, Table 3). Similar comparison was not possible in sufficient number of patients with node-negative group alone, because tumors of most node-negative patients were cylcin E negative. Nevertheless, among node-negative patients, those with cyclin E-positive, p27-negative tumors (three cases) died within 3 years, whereas those with tumors positive for both antibodies (three cases) survived for more than 5 years.
4. Discussion A body of evidences has been accumulated showing that overexpression of cyclin E contributes to the development of many types of human cancers and their precancerous lesions [3– 11,13]. Two earlier reports demonstrated that overexpression of cyclin E was an unfavorable prognostic marker in lung NSCLC [4,9], while one report was contradictory [3]. This discrepancy between the results may be due to a difference in the histological subtypes of lung cancers that were
Fig. 2. Kaplan – Meier survival curves of lung adenocarcinoma patients. In (c), ‘cyclin E status/p27 status’ was indicated, and ‘( +/ −)’ group had worse prognosis than ‘( −/ +)’ group (PB 0.0001), ‘( −/ −)’ group (P= 0.016) and ‘( +/ +)’ group (P =0.12).
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
64
Table 2 Cox’s proportional hazards analysis of prognostic factors in AC patientsa Factors
Hazard ratio
95% CI
P
Model 1 Cyclin E p27 Ki67 Differentiation pT pN
1.3 1.7 1.9 1.1 2.3 3.5
0.6–2.8 0.9–3.3 1.0–3.8 0.5–2.5 1.2–4.4 1.7–7.2
0.48 0.13 0.048 0.81 0.015 0.0008
1.6
0.7–3.4
0.28
2.0 1.0 2.4 3.2
1.1–3.8 0.4–2.4 1.3–4.7 1.5–6.9
0.031 0.95 0.0079 0.0021
Model 2 Cyclin E and p27 in combination Ki67 Differentiation pT pN a
variate analysis. However, the results showed that neither cyclin E nor p27 nor the combination was significant factor. Instead, it was reconfirmed that high Ki67 and positive N factor were very important. It would be speculated that overexpression of cyclin E in lung AC might be associated with high proliferation activity of tumor cells as well as lymph node metastasis as noted in gastric cancer [21]. It may be noteworthy that cyclin E-positive, p27-negative tumors seemed to have unfavorable prognosis even in node-negative cases because patients with such tumor died within 3 years. Larger scale studies using cases limited to stage I are required to clarify the effect of the combination of cyclin E and p27 expressions on prognosis in lung AC.
Abbreviation: CI, confidence interval.
5. Conclusions analyzed. In terms of cancer development, the effect of cyclin E expression seems to differ in degree between lung AC and SQCC, because its expression is lower in AC than in SQCC, including premalignant lesions. This is the first study which analyzed prognostic significance of cyclin E expression in lung AC, and it suggested that overexpression of cyclin E might be correlated with unfavorable prognosis in lung AC. It was also suggested that a low p27 expression was another poor prognostic factor. In the present study, patients with p27-negative and cyclin E-positive tumors had the most unfavorable prognosis, as reported in breast cancer [19]. In order to confirm this, we performed multi-
Our results suggested that cyclin E overexpression in lung AC, particularly with low p27 expression, is potentially correlated to poor prognosis. Larger scale analysis is required to verify the prognostic significance of these factors.
Acknowledgements This study was partly supported by a grant from the Ministry of Health and Welfare and by Kanagawa Prefectural Hospital Cancer Research Fund, Japan. The authors thank Masaichi Ikeda for excellent technical assistance.
Table 3 Relationship between combination of cyclin E and p27 status (cyclin E/p27) and clinicopathological factors (significant factors selected from Table 1) Ki67
(−/+) group Other groups combined
pN
n
High (%)
Low (%)
P
N0 (%)
N1B (%)
P
11 87
8 (73) 28 (32)
3 (27) 59 (68)
0.012
3 (27) 61 (70)
8 (73) 26 (30)
0.0078
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
References [1] Barsky SH, Cameron R, Osann KE, Tomita D, Holmes EC. Rising incidence of bronchioloalveolar carcinoma and its clinicopathological features. Cancer 1994;73:1163 – 70. [2] Koff A, Giordano A, Desai D, et al. Formation and activation of cyclin E –cdk2 complex during the G1 phase of the human cell cycle. Science 1992;257:1689 –94. [3] Dobashi Y, Shoji M, Jiang S-X, Kobayashi M, Kawakubo Y, Kameya T. Active cyclin A –CDK2 complex, a possible critical factor for cell proliferation in human primary lung carcinomas. Am J Pathol 1998;153:963 – 72. [4] Fukuse T, Hirata T, Naiki H, Hitomi S, Wada H. Prognostic significance of cyclin E overexpression in resected non-small cell lung cancer. Cancer Res 2000;60:242 –4. [5] Hayashi H, Ito T, Yazawa T, et al. Reduced expression of p27/Kip1 is associated with pulmonary adenocarcinoma development. J Pathol 2000;192:26 –31. [6] Ito Y, Matsuura N, Sakon M, et al. Expression and prognostic roles of the G1-S modulators in hepatocellular carcinoma: p27 independently predicts the recurrence. Hepatology 1999;30:90 –9. [7] Makiyama K, Masuda M, Takano Y, et al. Cyclin E overexpression in transitional cell carcinoma of the bladder. Cancer Lett 2000;151:193 –8. [8] Muller W, Noguchi T, Wirtz HC, Hommel G, Gabbert HE. Expression of cell-cycle regulatory proteins cyclin D1, cyclin E, and their inhibitor p21WAF1/CIP1 in gastric cancer. J Pathol 1999;189:186 –93. [9] Mishina T, Dosaka-Akita H, Hommura F, et al. Cyclin E expression, a potential prognostic marker for non-small cell lung cancers. Clin Cancer Res 2000;6:11 –6. [10] Nielsen NH, Arnerlo¨ v C, Emdin SO, Landberg G. Cyclin E overexpression, a negative prognostic factor in breast cancer with strong correlation to oestrogen receptor status. Br J Cancer 1996;74:874 –80. [11] Shimizu M, Nikaido T, Toki T, Shiozawa T, Fujii S.
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
65
Clear cell carcinoma has an expression pattern of cell cycle regulatory molecules that is unique among ovarian adenocarcinomas. Cancer 1999;85:669 – 77. Greenblatt MS, Bennett WP, Hollstein M, Harris CC. Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res 1994;54:4855 – 78. Lonardo F, Rusch V, Langenfeld J, Dmitrovsky E, Klimstra DS. Overexpression of cyclins D1 and E is frequent in bronchial preneoplasia and precedes squamous cell carcinoma development. Cancer Res 1999;59:2470 – 6. Polyak K, Lee M-H, Erdjument-Bromage H, et al. Cloning of p27Kip1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell 1994;78:59 – 66. Catzavelos C, Tsao M-S, DeBoer G, Bhattacharya N, Shepherd FA, Slingerland JM. Reduced expression of the cell cycle inhibitor p27Kip1 in non-small cell lung carcinoma: a prognostic factor independent of Ras. Cancer Res 1999;59:684 – 8. Esposito V, Baldi A, DeLuca A, et al. Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res 1997;57:3381 – 5. Yatabe Y, Masuda A, Koshikawa T, et al. p27KIP1 in human lung cancers: differential changes in small cell and non-small cell carcinomas. Cancer Res 1998;58:1042 – 7. Lloyd RV, Erickson LA, Jin L, et al. p27kip1: a multifunctional cyclin-dependent kinase inhibitor with prognostic significance in human cancers. Am J Pathol 1999;154:313 – 23. Porter PL, Malone KE, Heagerty PJ, et al. Expression of cell cycle regulators p27Kip1 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nature Med 1997;3:222 – 5. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710 – 7. Sakaguchi T, Watanabe A, Sawada H. Prognostic value of cyclin E and p53 expression in gastric carcinoma. Cancer 1998;82:1238 – 43.
High cyclin E and low p27/Kip1 expressions are potentially poor prognostic factors in lung adenocarcinoma patients Hiroyuki Hayashi a,*, Nobuo Ogawa b, Naoki Ishiwa b, Takuya Yazawa a, Yoshiaki Inayama a, Takaaki Ito a, Hitoshi Kitamura a a
Department of Pathology, Yokohama City Uni6ersity School of Medicine, 3 -9 Fukuura, Kanazawa-ku, Yokohama, 236 -0004, Japan b Di6ision of Pulmonary Surgery, Kanagawa Cardio6ascular and Respiratory Centre, 6 -16 -1 Tomiokahigashi, Kanazawa-ku, Yokohama, 236 -0051, Japan Received 21 November 2000; received in revised form 15 February 2001; accepted 27 February 2001
Abstract Cyclin E is an important regulator of entry into the S phase of the cell cycle. p27/Kip1 (p27) binds to cyclin E/Cdk2 complex and negatively regulates cell proliferation. We immunohistochemically examined the expression of cyclin E and p27 in 98 cases of resected lung adenocarcinoma to evaluate the prognostic significance of cyclin E and p27. Cyclin E was expressed in 16 cases (16%), and p27 was expressed in 41 cases (42%). Using Kaplan– Meier survival analysis, patients with cyclin E positive (P= 0.0017) and p27 negative (P= 0.011), both individually and in combination (P B0.0001), had a worse prognosis. We also analyzed the relationship of these findings to clinicopathological parameters, which revealed that cyclin E-positive, p27-negative cases had a higher Ki67 expression (P = 0.012) and a higher rate of lymph node metastasis (P =0.0078) than other groups. Our results suggested that cyclin E over expression, in association with p27 reduction in particular, may potentially be a poor prognostic factor in lung adenocarcinoma patients. However, to verify the prognostic significance of these factors, a multivariate analysis of a larger number of patients should be undertaken. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cyclin E; p27/Kip1; Ki67; Lung; Adenocarcinoma; Survival
1. Introduction Lung cancer is one of the leading causes of cancer death throughout the world. Among the heterogeneous histological subtypes, the incidence * Corresponding author. Tel.: + 81-45-787-2583; fax: + 8145-789-0588. E-mail address: [email protected] (H. Hayashi).
of adenocarcinoma (AC) is increasing [1]. Recent studies of cell-cycle regulatory proteins have contributed to elucidating the development and extension of cancer. Cyclins, cyclin-dependent kinases (Cdks), and Cdk inhibitors play crucial roles in the passage of checkpoints in the cell cycle. Cyclin E is associated with Cdk2 in an active complex in the late G1 phase and regulates progression from G1 into the S phase [2]. Overexpression of cyclin E has been reported in lung
0169-5002/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 1 ) 0 0 2 1 1 - 2
60
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
cancer and other various organs [3– 11]. To date, there are three reports in which the expression of cyclin E and patients’ survival were analyzed in non-small cell lung cancer (NSCLC). Dobashi et al. reported those patients whose tumors were cyclin E positive showed significantly higher survival rate compared with those without cyclin E expression [3]. To the contrary, Fukuse et al. and Mishina et al. reported that high expression of cyclin E in lung cancer tissue correlates with a poor prognosis [4,9]. NSCLC consists of heterogeneous subtypes. Squamous cell carcinoma (SQCC) and AC have different aspects in their histogenesis, gender trend, p53 mutation spectrum, and etiology particularly the relationship to smoking [12]. In studies of earlier survival analysis on NSCLC, cyclin E expression was significantly lower in AC than in SQCC [3,9]. Moreover, cyclin E expression has been reported to be observed in premalignant lesions of SQCC [13], however, such observation was not found in precancerous lesion of AC [5]. We consider that it is important to analyze the prognostic significance of cyclin E in the respective histological subtypes because of those distinct differences. p27/Kip1 (p27), one of the Cdk inhibitors, binds to the cyclin E/Cdk2 complex and negatively regulates cell proliferation [14]. Recently, several studies have indicated that low levels of p27 expression are associated with a poor prognosis and an advanced grade in human neoplasms of the lung as well as those of various other organs [15–18]. However, the prognostic significance of p27 and cyclin E in combination has not been well studied to date. Porter et al. reported that in breast cancer patients, the combination of high p27 and low cyclin E showed the most favorable prognosis, and the converse was associated with the highest mortality [19]. We earlier reported that high p27 expressers showed lower Ki67 labeling index values (LI) than low p27 expressers among cyclin E-positive lung AC patients [5]. In the present study, we analyzed the expression of cyclin E and p27 immunohistochemically to evaluate whether their expression correlated with 5-year survival in lung AC, either individually or in combination. We also compared expression of
these proteins with other clinicopathological parameters.
2. Materials and methods
2.1. Patients and tissues Of the consecutive 149 lung AC patients who were surgically treated between 1987 and 1995 in Kanagawa Cardiovascular and Respiratory Centre, 128 patients underwent complete tumor resection with mediastinal lymph node dissection. Those who died within 3 months after surgery, those who died of causes other than lung cancer within 5 years, and those who were followed for less than 5 years (for surviving patients) were excluded from this study. Patients with tumors on whom we could not perform immunohistochemical evaluation were also excluded. Thus, the subjects were 98 patients, consisting of 42 females and 56 males whose ages at surgery ranged from 31 to 81 years. They had not undergone chemotherapy or radiotherapy preoperatively. Six cases presented in our earlier report [5] were included.
2.2. Immunohistochemistry For immunohistochemistry, we used mouse primary monoclonal antibodies against p27 (57; Transduction Lab., Lexington, KY) at a 1:300 dilution, Ki67 (MIB-1; Immunotech, Marseilles, France) at a 1:200 dilution, and cyclin E (13A3; Novocastra Lab., Newcastle, UK) at a 1:25 dilution. After deparaffinization, microwave pre-treatment was performed in 10 mmol/l citrate buffer (pH 6.0) for 15 min at 95°C. The sections were incubated overnight at 4°C with the primary antibodies, then processed with a streptavidin– biotin–peroxidase detection kit (LSAB2 kit; Dako) in which diaminobenzidine was used as the chromogen. Cells were considered positive when the nucleus was distinctly stained. LI was determined in each lesion for each of the antibodies by counting 2000 cells. Cases showing \10% of tumor cells positive for p27, or showing \ 5% for cyclin E were defined as positive for each according to
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
the criteria used in the earlier studies [3,5]. Lung cancer and breast cancer specimens known to express these antigens were used as positive controls.
2.3. Statistical analysis The associations between the expression of those antigens and the clinicopathological parameters were determined by Fisher’s exact test. Kaplan–Meier survival analysis with log-rank test was used to evaluate the relationships between expressions of those antigens and survival distributions. Cox’s proportional hazard analysis was used for multivariate analysis. P values less than 0.05 were considered significant.
3. Results
3.1. Clinicopathological findings The patients consisted of 42 females and 56 males whose ages at surgery ranged from 31 to 81
61
years with a mean 9 S.D. of 62.79 10.4 years (Table 1). There were 49 smokers and 41 nonsmokers, and the smoking history of eight patients was unknown. Forty-six cases were well differentiated AC, 33 were moderately differentiated, and 19 were poorly differentiated. pT and pN classification for each case was based on the TNM classifications of American Joint Committee on Cancer and Union Internationale Contre le Cancer [20].
3.2. Immunohistochemistry The results were summarized in Table 1 and Fig. 1. There were no cells positive for cyclin E in non-neoplastic lung tissues. Cyclin E expression was found in the nuclei of tumor cells in 16 cases (16%, Fig. 1a and b). Among positive cases, 6– 68% of the cells were positive for cyclin E mostly with moderate to strong intensity. p27 was expressed in the nuclei of lymphocytes, fibroblasts, and normal bronchiolar epithelial cells, which served as internal positive control. In neoplastic lesions, p27 immunostaining was
Table 1 Clinicopathological characteristics of patients according to the results of immunohistochemistrya n
Cyclin E Positive (%)
p27
Ki67
P
Positive (%)
P
High expressor (%)
P
Gender Female Male
42 56
7 (16) 9 (17)
ns
17 (42) 24 (43)
ns
14 (33) 22 (39)
ns
Smoking history Smokers Non-smokers Unknown
49 41 8
9 (18) 5 (12) 2 (25)
ns
20 (41) 20 (49) 1 (13)
ns
21 (43) 11 (27) 4 (50)
ns
Differentiation Well Moderately Poorly
46 33 19
2 (4) 8 (24) 6 (31)
0.011* 0.006**
26 (57) 13 (39) 2 (11)
0.0005** 0.026***
9 (20) 14 (42) 13 (68)
0.026* 0.0003**
pT classification 1 and 2 3 and 4
81 17
12 (15) 4 (24)
ns
37 (46) 4 (24)
ns
28 (35) 8 (47)
ns
pN classification 0 1B
64 34
6 (9) 10 (30)
0.013
29 (45) 12 (35)
ns
18 (28) 18 (53)
0.014
a
Abbreviations: ns, not significant; *, well versus moderately; **, well versus poorly; ***, moderately versus poorly.
62
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
Fig. 1. Immunohistochemistry of cyclin E (a and b), p27 (c and d), and Ki67 (e and f). Cyclin E-positive tumor (a) without p27 expression (c) showed high Ki67 LI (e), while cyclin E-negative (b), p27-positive (d) tumor showed low Ki67 LI (f). Scale bar =20 mm.
mostly confined to the cell nuclei and was usually strong (Fig. 1c and d), but a few cases showed cytoplasmic staining. Forty-one cases (42%) were positive for p27, whose LI varied from 11 to 87%. Very few cells were positive for Ki67 in nonneoplastic lung tissues. In tumors, distinct nuclear staining was observed, if present. The mean LI of Ki67 in tumors used in this study was 12.69 14.1% (mean9S.D.), which we defined as the cut-off level on statistical analysis. According to this criterion, 36 cases (37%) showed high Ki67 expression (Fig. 1e and f). Cyclin E-positive cases showed higher Ki67 LI than negative cases (P =
0.0001). There was no definite correlation between p27 and Ki67 expression. Comparing the expression of those proteins to clinicopathological parameters, Ki67 and cyclin E expression increased and p27 expression decreased parallel to the grade of AC (Table 1).
3.3. Prognostic significance of cyclin E and p27 Patients with cyclin E-positive tumors had a significantly worse prognosis than those with cyclin E-negative tumors (P= 0.0017, Fig. 2a). Patients with p27-positive tumors had a significantly
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
63
better prognosis than those with p27-negative tumors (P=0.011, Fig. 2b). Since these proteins interact in the G1/S phase, we analyzed the prognosis by combination of both factors. Tumors positive for cyclin E and negative for p27 (n = 11) had a worse prognosis than the cyclin-E negative, p27-positive group (n =36, P B0.0001, Fig. 2c), both negative group (n =46, P =0.016) and both positive group (n= 5, P = 0.12). Moreover, comparing the cyclin E-positive, p27-negative group with the other three groups combined, the former had a significantly worse prognosis (P B 0.0001).
3.4. Multi6ariate prognostic factor analysis The results of 98 patients were summarized in Table 2. pT and pN classifications and Ki67 expression were significant factors. However, neither cyclin E and p27 expressions nor their combination were significant factors. Therefore, we analyzed the correlation between their combination status and other factors. This analysis revealed that cyclin E and p27 status correlate with Ki67 expression (P =0.012) and pN classification (P =0.0078, Table 3). Similar comparison was not possible in sufficient number of patients with node-negative group alone, because tumors of most node-negative patients were cylcin E negative. Nevertheless, among node-negative patients, those with cyclin E-positive, p27-negative tumors (three cases) died within 3 years, whereas those with tumors positive for both antibodies (three cases) survived for more than 5 years.
4. Discussion A body of evidences has been accumulated showing that overexpression of cyclin E contributes to the development of many types of human cancers and their precancerous lesions [3– 11,13]. Two earlier reports demonstrated that overexpression of cyclin E was an unfavorable prognostic marker in lung NSCLC [4,9], while one report was contradictory [3]. This discrepancy between the results may be due to a difference in the histological subtypes of lung cancers that were
Fig. 2. Kaplan – Meier survival curves of lung adenocarcinoma patients. In (c), ‘cyclin E status/p27 status’ was indicated, and ‘( +/ −)’ group had worse prognosis than ‘( −/ +)’ group (PB 0.0001), ‘( −/ −)’ group (P= 0.016) and ‘( +/ +)’ group (P =0.12).
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
64
Table 2 Cox’s proportional hazards analysis of prognostic factors in AC patientsa Factors
Hazard ratio
95% CI
P
Model 1 Cyclin E p27 Ki67 Differentiation pT pN
1.3 1.7 1.9 1.1 2.3 3.5
0.6–2.8 0.9–3.3 1.0–3.8 0.5–2.5 1.2–4.4 1.7–7.2
0.48 0.13 0.048 0.81 0.015 0.0008
1.6
0.7–3.4
0.28
2.0 1.0 2.4 3.2
1.1–3.8 0.4–2.4 1.3–4.7 1.5–6.9
0.031 0.95 0.0079 0.0021
Model 2 Cyclin E and p27 in combination Ki67 Differentiation pT pN a
variate analysis. However, the results showed that neither cyclin E nor p27 nor the combination was significant factor. Instead, it was reconfirmed that high Ki67 and positive N factor were very important. It would be speculated that overexpression of cyclin E in lung AC might be associated with high proliferation activity of tumor cells as well as lymph node metastasis as noted in gastric cancer [21]. It may be noteworthy that cyclin E-positive, p27-negative tumors seemed to have unfavorable prognosis even in node-negative cases because patients with such tumor died within 3 years. Larger scale studies using cases limited to stage I are required to clarify the effect of the combination of cyclin E and p27 expressions on prognosis in lung AC.
Abbreviation: CI, confidence interval.
5. Conclusions analyzed. In terms of cancer development, the effect of cyclin E expression seems to differ in degree between lung AC and SQCC, because its expression is lower in AC than in SQCC, including premalignant lesions. This is the first study which analyzed prognostic significance of cyclin E expression in lung AC, and it suggested that overexpression of cyclin E might be correlated with unfavorable prognosis in lung AC. It was also suggested that a low p27 expression was another poor prognostic factor. In the present study, patients with p27-negative and cyclin E-positive tumors had the most unfavorable prognosis, as reported in breast cancer [19]. In order to confirm this, we performed multi-
Our results suggested that cyclin E overexpression in lung AC, particularly with low p27 expression, is potentially correlated to poor prognosis. Larger scale analysis is required to verify the prognostic significance of these factors.
Acknowledgements This study was partly supported by a grant from the Ministry of Health and Welfare and by Kanagawa Prefectural Hospital Cancer Research Fund, Japan. The authors thank Masaichi Ikeda for excellent technical assistance.
Table 3 Relationship between combination of cyclin E and p27 status (cyclin E/p27) and clinicopathological factors (significant factors selected from Table 1) Ki67
(−/+) group Other groups combined
pN
n
High (%)
Low (%)
P
N0 (%)
N1B (%)
P
11 87
8 (73) 28 (32)
3 (27) 59 (68)
0.012
3 (27) 61 (70)
8 (73) 26 (30)
0.0078
H. Hayashi et al. / Lung Cancer 34 (2001) 59–65
References [1] Barsky SH, Cameron R, Osann KE, Tomita D, Holmes EC. Rising incidence of bronchioloalveolar carcinoma and its clinicopathological features. Cancer 1994;73:1163 – 70. [2] Koff A, Giordano A, Desai D, et al. Formation and activation of cyclin E –cdk2 complex during the G1 phase of the human cell cycle. Science 1992;257:1689 –94. [3] Dobashi Y, Shoji M, Jiang S-X, Kobayashi M, Kawakubo Y, Kameya T. Active cyclin A –CDK2 complex, a possible critical factor for cell proliferation in human primary lung carcinomas. Am J Pathol 1998;153:963 – 72. [4] Fukuse T, Hirata T, Naiki H, Hitomi S, Wada H. Prognostic significance of cyclin E overexpression in resected non-small cell lung cancer. Cancer Res 2000;60:242 –4. [5] Hayashi H, Ito T, Yazawa T, et al. Reduced expression of p27/Kip1 is associated with pulmonary adenocarcinoma development. J Pathol 2000;192:26 –31. [6] Ito Y, Matsuura N, Sakon M, et al. Expression and prognostic roles of the G1-S modulators in hepatocellular carcinoma: p27 independently predicts the recurrence. Hepatology 1999;30:90 –9. [7] Makiyama K, Masuda M, Takano Y, et al. Cyclin E overexpression in transitional cell carcinoma of the bladder. Cancer Lett 2000;151:193 –8. [8] Muller W, Noguchi T, Wirtz HC, Hommel G, Gabbert HE. Expression of cell-cycle regulatory proteins cyclin D1, cyclin E, and their inhibitor p21WAF1/CIP1 in gastric cancer. J Pathol 1999;189:186 –93. [9] Mishina T, Dosaka-Akita H, Hommura F, et al. Cyclin E expression, a potential prognostic marker for non-small cell lung cancers. Clin Cancer Res 2000;6:11 –6. [10] Nielsen NH, Arnerlo¨ v C, Emdin SO, Landberg G. Cyclin E overexpression, a negative prognostic factor in breast cancer with strong correlation to oestrogen receptor status. Br J Cancer 1996;74:874 –80. [11] Shimizu M, Nikaido T, Toki T, Shiozawa T, Fujii S.
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
65
Clear cell carcinoma has an expression pattern of cell cycle regulatory molecules that is unique among ovarian adenocarcinomas. Cancer 1999;85:669 – 77. Greenblatt MS, Bennett WP, Hollstein M, Harris CC. Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res 1994;54:4855 – 78. Lonardo F, Rusch V, Langenfeld J, Dmitrovsky E, Klimstra DS. Overexpression of cyclins D1 and E is frequent in bronchial preneoplasia and precedes squamous cell carcinoma development. Cancer Res 1999;59:2470 – 6. Polyak K, Lee M-H, Erdjument-Bromage H, et al. Cloning of p27Kip1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell 1994;78:59 – 66. Catzavelos C, Tsao M-S, DeBoer G, Bhattacharya N, Shepherd FA, Slingerland JM. Reduced expression of the cell cycle inhibitor p27Kip1 in non-small cell lung carcinoma: a prognostic factor independent of Ras. Cancer Res 1999;59:684 – 8. Esposito V, Baldi A, DeLuca A, et al. Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res 1997;57:3381 – 5. Yatabe Y, Masuda A, Koshikawa T, et al. p27KIP1 in human lung cancers: differential changes in small cell and non-small cell carcinomas. Cancer Res 1998;58:1042 – 7. Lloyd RV, Erickson LA, Jin L, et al. p27kip1: a multifunctional cyclin-dependent kinase inhibitor with prognostic significance in human cancers. Am J Pathol 1999;154:313 – 23. Porter PL, Malone KE, Heagerty PJ, et al. Expression of cell cycle regulators p27Kip1 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nature Med 1997;3:222 – 5. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710 – 7. Sakaguchi T, Watanabe A, Sawada H. Prognostic value of cyclin E and p53 expression in gastric carcinoma. Cancer 1998;82:1238 – 43.