Overexpression of Cyclin B1 in Early-Stage Non-Small Cell Lung Cancer and ItsClinical Implication 1
Jean-Charles Soria, Se Jin Jang, Fadlo R. Khuri, Khaled Hassan, Diane Liu, Waun Ki Hong, and Li Mao2
Molecular Biology Laboratory, Department of Thoracic/Head and Neck Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030
Abstract
Cyclin B1 is a key molecule for G2-M-phase transition during the cellcycle and is overexpressed in various tumor types. However, the expres-sion status of cyclin B1 in lung cancer and its clinical significance remainunknown. We used immunohistochemistry studies to examine the expres-sion of cyclin B1 in 77 non-small cell lung cancer specimens from patientswith histological stage I disease. All of the patients underwent curativesurgical treatment. The median length of follow-up care is 8.2 years.High-level cyclin B1 expression (a cyclin B1 labeling index> 15%) wasobserved in 17 of the 77 (22%) tumors. Patients whose tumors expresseda high level of cyclin B1 had a significantly shorter survival time thanpatients whose tumors expressed a low level of cyclin B1 (P5 0.02,log-rank test). Interestingly, overexpression of cyclin B1 was more fre-quently observed in tumors with squamous cell histology than in tumorswith other histological cell types (P5 0.01, Fisher’s exact test). A sub-group analysis revealed that cyclin B1 overexpression seems to be anadverse prognostic factor only in patients with squamous cell carcinoma(SCC) of the lung (P5 0.02, log-rank test). Our data indicate that cyclinB1 may be dysregulated in non-small cell lung cancer, particularly in theSCC subtype, and that a high level of cyclin B1 expression may be aprognostic marker for patients with early-stage SCC of the lung.
Introduction
Lung cancer is a major cause of mortality worldwide. In the UnitedStates alone, it was estimated that 164,100 new cases of lung cancerwould be diagnosed in 2000, with an estimated 156,900 deaths (1).NSCLC3 represents about 80% of all lung cancers, and its dismalsurvival rate has not improved significantly in the past two decades.Even patients with pathological stage I NSCLC have only a 60%survival rate at 5 years (2). To further improve the survival rate in thisgroup of patients, their prognostic classification based on tumorbiology will be crucial. Such classification might help clinicians tomake the right management decisions for each subset of patients.
Altered regulation of the cell cycle is a hallmark of human cancers(3). Cell cycle progression is governed by a series of cyclins and cdks.Individual cyclins act at different phases of the cell cycle by bindingand activating corresponding cdks. Of the various cyclin/cdkcomplexes involved in cell cycle regulation, cyclin D1/cdk4/6 andcyclin B1/Cdc2 are of particular interest because the former directs
G1-S-phase transition and the latter controls G2-M-phase checkpointsurveillance, which are in turn essential for DNA synthesis and cellproliferation. Dysregulated expression of these cyclins, cdks, or bothmay lead to uncontrolled cell growth and malignant transformation.Overexpression and/or amplification of cyclin D1 has been reported ina large variety of human cancers, including those of the esophagus,head and neck, lung, liver, and breast (3) and is reported to be ofprognostic importance in patients with most of these tumor types(4–6). Overexpression of cyclin B1 has been reported more recentlyin breast, colon, prostate, oral, and esophageal carcinomas (7–11). Itsprognostic value has been suggested in patients with SCC of theesophagus (11). However, little is known about cyclin B1 expressionstatus in lung cancer and its potential clinical application in this tumortype.
To determine whether cyclin B1 expression is dysregulated inearly-stage NSCLC and whether it can be used as a prognostic markerin patients with NSCLC, we examined immunohistochemically theexpression pattern of cyclin B1 in 77 patients with pathological stageI NSCLC. We found that cyclin B1 was overexpressed in about 20%of the tumors analyzed (34% of SCC and 12% of non-SCC tumors)and that this overexpression is an adverse prognostic factor of survivaltime for patients with stage I NSCLC.
Materials and Methods
Study Population. The primary NSCLC specimens were archived tissuesamples of surgically resected pathological stage I tumors from 77 patientstreated at The University of Texas M. D. Anderson Cancer Center (Houston,TX). All patients were treated by surgery alone and received a median of 8.2years of follow-up care after surgical treatment. Surgical specimens werecollected between 1975 and 1990. Survival data were available for all patientsand with a minimum length of follow-up care of 5 years. The study populationconsisted of 54 men and 23 women. The mean age of patients was 65 years.Histological subtypes included 35 SCCs, 33 adenocarcinomas, 3 broncho-alveolar carcinomas, 3 large cell carcinomas, 1 adenosquamous carcinoma, and2 unclassified tumors (Table 1).
Immunohistochemical Staining for Cyclin B1 and Ki-67 Protein. Par-affin-embedded, 4-mm-thick tissue sections from all 77 primary tumors werestained for the cyclin B1 protein using a primary mouse monoclonal antibody(NCL-Cyclin B1; Novocastra, Newcastle, United Kingdom) and for Ki-67 (allbut one case) using a primary rabbit polyclonal antibody (Dako, Carpinteria,CA). Slides were baked at 60°C for 1 h and then deparaffinized through aseries of xylene baths. Rehydration was performed through graded alcohols.To retrieve the antigenicity, tissue sections were then treated with microwavesin 10 mM citrate buffer (pH 6.0) for 3 min (three times for cyclin B1, and sixtimes for Ki-67). The sections were then immersed in methanol containing0.3% hydrogen peroxidase for 20 min to block the endogenous peroxidaseactivity and incubated in 2.5% blocking serum to reduce nonspecific binding.Sections were incubated overnight at 4°C with primary anti-cyclin B1 oranti-Ki-67 antibody at dilutions of 1:15 and 1:100, respectively. The sectionswere then processed using standard avidin-biotin immnuohistochemistry ac-cording to the manufacturer’s recommendations (Vector Laboratories, Burl-ingame, CA). Diaminobenzidine was used as a chromogen, and commercialhematoxylin was used for counterstaining. Routinely processed tissue sectionsof normal lymph node and tonsil were used as positive staining controls and
Received 3/14/00; accepted 6/7/00.The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.
1 Supported in part by American Cancer Society Grant RPG-98-054 and NationalCancer Institute Grant P01 CA74173 (to L. M.), Fondation de France, AP-HP and LillyFondation Grant (to J-C. S.), American Cancer Society-Clinical Oncology Career Devel-opment Award 96-41 (to F. R. K.), National Cancer Institute Grant U19 CA 68437 (toW. K. H.), Cancer Center Grant P30 CA 16620 (to the M. D. Anderson Cancer Center),and Tobacco Research Fund from the State of Texas (M. D. Anderson Cancer Center).W. K. H. is an American Cancer Society Clinical Research Professor.
2 To whom requests for reprints should be addressed, at Molecular Biology Labora-tory, Department of Thoracic/Head and Neck Medical Oncology, The University of TexasM. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. Phone:(713) 792-6363; Fax: (713) 796-8655; E-mail: [email protected].
3 The abbreviations used are: NSCLC, non-small cell lung cancer; SCC, squamous cellcarcinoma; cdk, cyclin-dependent kinase.
were also stained with the primary antibody omitted to confirm stainingspecificity.
The cyclin B1 labeling index was defined as the percentage of tumor cellsdisplaying cytoplasmic or nuclear immunoreactivity and calculated by count-ing the number of cyclin B1-stained tumor cells among at least 1000 tumorcells for each section. Similarly, the Ki-67 proliferative index was defined asthe percentage of nuclear-stained cells among 1000 or more tumor cells.Representative areas of each tissue section were selected, and cells werecounted in at least four fields (at3400) in these areas. All slides were scoredconcomitantly by a pathologist (S. J. J.) and another investigator (J-C. S.).
Statistical Analysis. Survival curves were estimated by the Kaplan-Meiermethod, and the resulting curves were compared using the log-rank test.Fisher’s exact test and thex2 test were used to analyze the association betweentwo categorical variables.P , 0.05 was considered to be statistically signif-icant. Immunohistochemical analysis was performed in a blinded manner withrespect to the clinical information of the subjects. The Pearson correlationcoefficient was used to test the strength of association between continuousvariables (i.e., cyclin B1 and Ki-67).
Results
In normal ciliated bronchial epithelial cells, peribronchial glandcells, or alveolar pneumocytes, cyclin B1 expression was not detect-able. Interstitial or lymphoid cells of tumor areas were occasionallyimmunostained. In tumor cells that stained positively for cyclin B1,the immunostaining was detected predominantly in the cytoplasm(Fig. 1, a–c), although in a few cases, nuclear staining was alsoobserved. In the positive control tissues (tonsil), the normal stratifiedsquamous epithelium displayed cyclin B1-positive cells in the basaland parabasal layers. In the adjacent lymph nodes, cyclin B1-positivecells were localized predominantly in the germinal centers. PositiveKi-67 immunostaining was restricted to the cell nuclei (Fig. 1,d–f).For Ki-67, as well as for cyclin B1, the immunohistochemical stainingshowed a wide heterogeneity from rare scattered cells to a homoge-neous pattern for the vast majority of cells examined, suggesting thatphenotypic heterogeneity is a major feature in NSCLC (Fig. 1). If$15% of the cells were positive for cyclin B1, the case was consid-ered to have cyclin B1 overexpression. For Ki-67 staining, cases with$25% positive cells were considered to have high expression ofKi-67 based on previous reports (12–14). As shown in Fig. 2, a weakbut statistically significant correlation (correlation coefficient5 0.30;P 5 0.009) was found between cyclin B1 index and Ki-67 score.
Cyclin B1 overexpression was observed in 17 of the 77 (22%) stageI NSCLC specimens. Low cyclin B1 expression was observed in 60tumors. Table 1 shows the relationships between expression of cyclinB1 and clinicopathological factors. The frequency of cyclin B1 over-expression did not differ significantly in different age groups or bysex. Interestingly, overexpression of cyclin B1 was observed morefrequently in the SCC subtype than it was in other histologicalsubtypes. Twelve of the 35 (34%) SCCs exhibited high levels ofcyclin B1 expression, whereas only 5 of 42 (12%) patients withtumors with non-SCC subtypes (mainly adenocarcinoma) showedoverexpression of cyclin B1 (P5 0.01, Fisher’s exact test).
We subsequently analyzed the relationship between cyclin B1expression and length of survival. Fig. 3Ashows a comparison of theKaplan-Meier survival curves between patients whose tumors ex-pressed a high level of cyclin B1 and those whose tumors had lowcyclin B1 expression. Patients with tumors that overexpressed cyclinB1 had significantly shorter survival times than patients with tumorsthat displayed low levels of cyclin B1 (P 5 0.02, log-rank test). About60% of the patients whose tumors had a low cyclin B1 expressionwere alive at 5 years compared with only 30% of the patients whosetumors had high cyclin B1 expression (Table 1). The prognosticsignificance of cyclin B1 expression was further explored in terms ofthe major histological subtypes. Cyclin B1 overexpression was asignificant adverse prognostic factor among patients with tumors ofthe SCC subtype (P5 0.02, log-rank test; Fig. 3B) as opposed topatients with adenocarcinoma (Fig. 3C). Because Ki-67 and cyclin B1expression had a weak correlation, the prognostic significance ofKi-67 was also explored, but no statistically significant effect wasobserved (Fig. 3D).
Discussion
Cyclin B1 is an important mitotic cyclin in the G2 and M phases ofthe cell cycle. Its association with the active form of cdc2 initiateschromosome condensation, destruction of the nuclear membrane, andassembly of the mitotic spindle. An increasing body of data suggeststhat altered expression of cyclin B1 is a frequent event in tumor cells.Overexpression of cyclin B1 has been demonstrated in colorectal,prostate, breast, esophagus, and head and neck cancers as well asHodgkin and MALT lymphomas (7–11, 15, 16). Significant differ-ences in the labeling index of cyclin B1 between benign/premalignantlesions and breast carcinomas suggest a consequential role of cyclinB1 overexpression in the malignant transformation of breast epithelialcells (7). Furthermore, cyclin B1 overexpression was statisticallyassociated with depth of tumor invasion and the presence of venousinvasion in esophageal SCC, which was thought to affect length ofsurvival as established in a multivariate analysis (11).
In the present study, we demonstrated that cyclin B1 is overex-pressed in a significant fraction of NSCLCs. Overall, 17 of the 77(22%) tumors expressed cyclin B1 in$15% of tumor cells. Interest-ingly, cyclin B1 expression was different when SCC tumors werecompared with non-SCC samples. High cyclin B1 expression wasobserved in 34% of SCC tumors but in only 12% of tumors with otherhistological subtypes (Table 1). Furthermore, our data show that highcyclin B1 expression is a significant unfavorable prognostic factor inpatients with stage I NSCLC. The fact that all of the patients in thestudy were treated at a single institute and received lengthy follow-upcare after surgery makes survival analysis reliable. In the histologicalsubgroup analysis, we found that the prognostic value of cyclin B1seems to be limited to the squamous cell subtype and has no predictivevalue among patients with adenocarcinoma (Fig. 3,B and C). Thisparticular observation, together with the fact that cyclin B1 expressionitself differs among histological subtypes, highlights the biologicaldifferences among different subtypes of NSCLC. Different abnormal-ities in oncogenes and tumor suppressor genes among histologicalsubtypes of NSCLC are well-known. Indeed,K-ras mutations aremuch more common in adenocarcinomas than in SCCs, whereas thep53 mutant immunophenotype is more frequent in squamous carci-nomas than in adenocarcinoma (17, 18).
To determine whether overexpression of cyclin B1 in NSCLCsimply reflects the increased cell proliferation, we analyzed Ki-67, awell-established marker of cell proliferation that is not phase specificin the cell cycle, in contiguous sections of the tumors. Consistent withprevious reports, a weak correlation (0.30;P 5 0.009) was found
Table 1 Cyclin B1 status in stage I NSCLC tumors according to clinicopathologicalfeatures of patients
between the cyclin B1 index and Ki-67 score (9, 10). Although someof the previous studies have suggested that Ki-67 is a possible indi-cator of length of survival (12, 19, 20), we failed to show anysignificant prognostic value for Ki-67 in our study population. Theabsence of prognostic value for Ki-67 remained even when survivalanalysis was performed using different cutoff levels published previ-ously [i.e., 5%, 10%, or 20%; data not shown (9, 20, 21)]. Neverthe-less, the prognostic significance of Ki-67 in lung cancer is not yetfirmly established because some studies suggest that a high Ki-67expression level is predictive of poor survival (12, 19, 20), whereasothers did not demonstrate any effect (13, 14, 21). Collectively, theseobservations suggest that high cyclin B1 expression in NSCLCs is nota mere consequence of cell proliferation, but rather an indicator ofaberrant cell cycle progression at the G2-M-phase transition in cancercells. Cyclin B1 also seems to play a pivotal role in the biologicalbehavior of NSCLCs, particularly in SCC, and, in doing so, representsa potential new prognostic marker as well as a therapeutic target inNSCLC.
Fig. 1. Immunohistochemical staining patterns of cyclin B1 and Ki-67 in NSCLC.a, an adenocarcinoma with most cancer cells expressing cyclin B1 in the cytoplasm.b, a squamouscell carcinoma with most carcinoma cells positive for cyclin B1.c, an adenocarcinoma with rare scattered cells expressing cyclin B1 (original magnification,3400). d, anadenocarcinoma showing a strong positive reaction in the nuclei of carcinoma cells.e, a squamous cell carcinoma with most carcinoma cells positive for Ki-67.f, a bronchoalveolarform with few cells positive for Ki-67 (original magnification,3400).
Fig. 2. Distribution of the cyclin B1 labeling index and Ki-67 proliferative index for thestudy population.
4002
CYCLIN B1 EXPRESSION AS A PROGNOSTIC MARKER FOR NSCLC
How overexpression of cyclin B1 participates in tumor progressionremains to be established. Many authors suggest that the oncogenicrole of cyclins is probably related to their unscheduled expression,namely, their expression throughout the cell cycle (22, 23). Theunscheduled expression of cyclins in tumor cells could be the result ofeither impaired degradation or continued synthesis during the cellcycle. Regardless of the cause, the continuous presence of cyclins maycause activation of their respective partner kinases (cdks), whichremain essentially invariable throughout the cell cycle. This, in turn,might lead to unscheduled phosphorylation of a variety of proteins,thus driving the cell through the cycle and bypassing the respectivecheckpoints. Recent findings that p53 controls a G2 checkpointthrough down-regulation of cyclin B1 and that constitutive activationof cyclin B1 and associated cdc2 kinase can override this p53-mediated G2-M-phase arrest support this notion (24, 25).
In conclusion, this study has shown, using a homogeneous popu-lation of 77 patients with stage I NSCLC, that there is overexpressionof cyclin B1 in lung tumors, particularly in SCC, and that suchoverexpression was associated with patients’ prognosis. More com-prehensive studies involving greater numbers of patients are necessaryto confirm these findings.
Acknowledgments
We thank Susan Cweren, Lakshmi Kakarala, Shyla Kalapurakai, Julie Izzo,and Mathilde Sibony for helpful discussions and technical assistance withimmunohistochemistry.
References
1. Greenlee, R. T., Murray, T., Bolden, S., and Wingo, P. A. Cancer statistics, 2000. CACancer J. Clin.,50: 27–33, 2000.
2. Williams, D. E., Pairolero, P. C., Davis, C. S., Bernatz, P. E., Payne, W. S., Taylor,W. F., Uhlenhopp, M. A., and Fontana, R. S. Survival of patients surgically treatedfor stage I lung cancer. J. Thorac. Cardiovasc. Surg.,82: 70–76, 1981.
3. Sherr, C. J. Cancer cell cycles. Science (Washington DC),274: 1672–1677, 1996.4. Akervall, J. A., Michalides, R. J. A. M., Mineta, H., Balm, A., Borg, A., Dictor, M.,
Jin, Y., Loftus, B., Mertens, F., and Wennerberg, J. P. Amplification of cyclin D1 insquamous cell carcinoma of the head and neck and the prognostic value of chromo-somal abnormalities and cyclin D1 overexpression. Cancer (Phila.),79: 380–389,1997.
5. McIntosh, G. G., Anderson, J. J., Milton, I., Steward, M., Parr, A. H., Thomas, M. D.,Henry, J. A., Angus, B., Lennard, T. W. J., and Horne, C. H. W. Determination of theprognostic value of cyclin D1 overexpression in breast cancer. Oncogene,11:885–891, 1995.
6. Nishida, N., Fukuda, Y., Komeda, T., Kita, R., Sando, T., Furukowa, M., Amenomori,M., Shibagaki, I., Nakao, K., Ikenaga, M., and Ishizaki, K. Amplification andoverexpression of the cyclin D1 gene in aggressive human hepatocellular carcinoma.Cancer Res.,54: 3107–3110, 1994.
7. Kawamoto, H., Koizumi, H., and Uchikoshi, T. Expression of the G2-M checkpointregulators cyclin B1 and cdc2 in nonmalignant and malignant breast lesions. Am. J.Pathol.,150: 15–23, 1997.
8. Wang, A., Yoshimi, N., Ino, N., Tanaka, T., and Mori, H. Overexpression of cyclinB1 in human colorectal cancers. J. Cancer Res. Clin. Oncol.,123: 124–127, 1997.
9. Mashal, R. D., Lester, S., Corless, C., Richie, J. P., Chandra, R., Propert, K. J., andDutta, A. Expression of cell cycle-regulated proteins in prostate cancer. Cancer Res.,56: 4159–4163, 1996.
10. Kushner, J., Bradley, G., Young, B., and Jordan, R. C. K. Aberrant expression ofcyclin A and cyclin B1 proteins in oral carcinoma. J. Oral Pathol. Med.,28: 77–81,1999.
11. Murakami, H., Furihata, M., Ohtusi, Y., and Ogoshi, S. Determination of the prog-nostic significance of cyclin B1 overexpression in patients with oesophageal squa-mous cell carcinoma. Virchows Arch.,434: 153–158, 1999.
Fig. 3.A, survival curves of patients with stage I NSCLC according to cyclin B1 expression;B, survival curves of patients with stage I squamous cell carcinoma according to cyclinB1; C, survival curves of patients with stage I adenocarcinoma according to cyclin B1;D, survival curves of patients with stage I NSCLC according to Ki-67.
4003
CYCLIN B1 EXPRESSION AS A PROGNOSTIC MARKER FOR NSCLC
12. Scagliotti, V. G., Micela, M., Gubetta, L., Leonardo, E., Cappia, S., Borasio, P., andPozzi, E. Prognostic significance of Ki-67 labelling in resected non-small cell lungcancer. Eur. J. Cancer,29A: 363–365, 1993.
13. Pujol, J. L., Simony, J., Jolimoy, G., Jaffule, D., Demoly, P., Quantin, X., Marty-Ane,C., Boher, J. M., Charpentier, R., and Michel, F. B. Hypodiploidy, Ki-67 growthfraction and prognosis of surgically resected lung cancers. Br. J. Cancer,74: 964–970, 1996.
14. Mehdi, S. A., Etzell, J. E., Newman, N. B., Weidner, N., Kohman, L. J., and Graziano,S. L. Prognostic significance of Ki-67 immunostaining and symptoms in resectedstage I and II non-small cell lung cancer. Lung Cancer,20: 99–108, 1998.
15. Ohshima, K., Haraoka, S., Fujiki, T., Yoshioka, S., Suzumiya, J., Kanda, M., andKikuchi, M. Expression of cyclin E, A and B1 in Hodgkin and Reed-Stenberg cells:not suppressed by cyclin-dependent kinase inhibitor p21 expression. Pathol. Int.,49:506–512, 1999.
16. Banerjee, S. K., Weston, A. P., Zoubine, M. N., Campbell, D. R., and Cherian, R.Expression of cdc2 and cyclin B1 inHelicobacter pylori-associated gastric MALTand MALT lymphoma. Am. J. Pathol.,156: 217–225, 2000.
17. Rossell, R., Li, S., Skacel, Z., Mate, J. L., Maestre, J., Canela, M., Tolosa, M., Tolosa,E., Armengol, P., Barnadas, A., and Ariza, A. Prognostic impact of mutatedK-rasgene in surgically resected non-small cell lung cancer patients. Oncogene,8: 2407–2412, 1993.
18. Kishimoto, Y., Murakami, Y., Shiraishi, M., Hayashi, K., and Sekiya, T. Aberrationsof the p53 tumor suppressor gene in human non-small cell carcinomas of the lung.Cancer Res.,52: 4799–4804, 1992.
19. Pence, J. C., Kerns, B. J. M., Dodge, R., and Iglehart, D. Prognostic significance ofthe proliferation index in surgically resected non-small-cell lung cancer. Arch. Surg.,128: 1382–1390, 1993.
20. Viberti, L., Papotti, M., Abbona, G. C., Celano, A., Filosso, P. L., and Bussolati, G.Value of Ki-67 immunostaining in preoperative biopsies of carcinomas of the lung.Hum. Pathol.,28: 189–192, 1997.
21. Tungekar, M. F., Gatter, K. C., Dunnill, M. S., and Mason, D. Y. Ki-67 immuno-staining and survival in operable lung cancer. Histopathology,19: 545–550, 1991.
22. Gong, J., Ardelt, B., Traganos, F., and Darzynkiewicz, Z. Unscheduled expression ofcyclin B1 and cyclin E in several leukemic and solid tumor cell lines. Cancer Res.,54:4285–4288, 1994.
23. Kakino, S., Saski, K., Kurose, A., and Ito, H. Intracellular localization of cyclin B1during the cell cycle in glioma cells. Cytometry,24: 49–54, 1996.
24. Innocente, S. A., Abrahamson, J. L. A., Gogswell, J. P., and Lee, J. M. p53 regulatesa G2 checkpoint through cyclin B1. Proc. Natl. Acad. Sci. USA,96: 2147–2152,1999.
25. Park, M., Chae, H. D., Yun, J., Jung, M., Kim, Y. S., Kim, S. H., Han, M. H., andShin, D. Y. Constitutive activation of cyclin B1-associated cdc2 kinase overridesp53-mediated G2-M arrest. Cancer Res.,60: 542–545, 2000.
4004
CYCLIN B1 EXPRESSION AS A PROGNOSTIC MARKER FOR NSCLC
2000;60:4000-4004. Cancer Res Jean-Charles Soria, Se Jin Jang, Fadlo R. Khuri, et al. Lung Cancer and Its Clinical ImplicationOverexpression of Cyclin B1 in Early-Stage Non-Small Cell
