|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 7 (March 1), 2006: pp. 1152-1160 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.6631 Differential Cell Cycle–Regulatory Protein Expression in Biliary Tract Adenocarcinoma: Correlation With Anatomic Site, Pathologic Variables, and Clinical OutcomeFrom the Departments of Surgery, Pathology, and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY Address reprint requests to William R. Jarnagin, MD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021; e-mail: jarnagiw{at}mskcc.org
PURPOSE: Biliary tract adenocarcinomas (BTAs), although anatomically related, arise through ill-defined and possibly different location-related pathogenetic pathways. This clinicopathologic study characterizes differences in cell cycle–regulatory protein expression across the spectrum of BTA. METHODS: Tissue microarrays were prepared from paraffin-embedded surgical specimens with triplicate cores of BTA and benign tissue. Immunohistochemical expression of p53, cyclin D1, p21, Bcl2, p27, Mdm2, and Ki-67 was assessed, and the results were correlated with pathologic variables and survival. Hierarchical clustering was used to partition the data based on protein expression, and then the data were analyzed according to anatomic location. RESULTS: Tissue from 128 surgical patients (1992 to 2002) was obtained. Tumor sites of origin were intrahepatic cholangiocarcinoma (IH; n = 23), hilar cholangiocarcinoma (Hilar; n = 54), gallbladder (GB; n = 32), and distal bile duct (Distal; n = 19). p27 expression decreased progressively from proximal to distal in the biliary tree and correlated with location-related differences in outcome; cyclin D1 and Bcl2 overexpression also varied according to anatomic site. Aberrant p53 staining and cyclin D1 overexpression were lower in papillary tumors compared with the more common sclerosing tumors. The expression profiles of GB and Hilar were more similar to each other than either was to IH or Distal (86% clustering in the first partition). After an R0 resection, overexpression of Mdm2 (P = .0062) and absent p27 expression (P = .0165) independently predicted poor outcome. CONCLUSION: BTAs differentially express cell cycle–regulatory proteins based on tumor location and morphology. Prognostic roles were identified for Mdm2 and p27. Overlap in the pathogenesis of GB and Hilar tumors was suggested.
Biliary tract adenocarcinoma (BTA) can arise anywhere in the biliary tree, resulting in a wide spectrum of diseases (Fig 1). 1-4 Adenocarcinoma of the gallbladder and the extrahepatic bile ducts (cholangiocarcinoma) are the most common BTA, although recent data has shown an increasing incidence of intrahepatic cholangiocarcinoma.5 Despite greater risk associated with certain conditions (ie, sclerosing cholangitis6 and hepatolithiasis7,8), most cases are sporadic. The molecular changes associated with BTA development and progression are ill defined, and it is uncertain whether suggested location-related differences in pathogenesis9-12 influence clinical behavior.
Pertubation of check points governing progression through G1-S phase of the cell cycle is common among neoplasms.13 Loss of control at this site leads to unregulated entry into S phase, a hallmark of neoplastic growth. The p53 gene is pivotal in controlling cellular proliferation, largely by regulating proteins that control cell cycle progression, such as p21WAF1/Cip1, or apoptosis, such as Bax, which antagonizes the antiapoptotic activity of Bcl2. Loss of p53 activity is common in human cancers, resulting from mutational inactivation or functional inactivation through upregulation of Mdm2.14 Another frequent site of loss of regulatory control is cyclin D1. An upstream inhibitor of the Rb gene product, cyclin D1 overexpression has been identified in several malignancies, including BTA.15 p27Kip1 is also an important cyclin-dependent kinase inhibitor, promoting G1 arrest,16 and its loss of expression is associated with worse prognosis in some cancers. p27Kip1 is also proapoptotic.17 Variable rates of abnormal cell cycle–mediator expression have been shown in subgroups of BTA,15,18-25 and although some reports suggest anatomic site-specific alterations, none has comprehensively compared differential gene expression across the full spectrum of these tumors. This study aims to define differential expression of cell cycle–regulatory proteins based on anatomic site of origin and tumor morphology and to determine whether such differences have clinical relevance.
Patients This study was approved by the Memorial Sloan-Kettering Cancer Center Institutional Review Board and is Health Insurance Portability and Accountability Act compliant. Patients with BTA and adequate tissue for analysis, who were treated from 1992 to 2002, were identified from a database. The study group consisted of patients with intrahepatic (IH), hilar (Hilar), and distal (Distal) cholangiocarcinoma and gallbladder carcinoma (GB); histologies other than adenocarcinoma were excluded. Demographic, operative, recurrence, and survival data were obtained from the database and supplemented with review of the medical record. The authors' approach to evaluation and resection of these tumors has been described previously.1,4,26,27 Routine histologic analysis was performed on all specimens (resection margin status, tumor size and differentiation [highest grade], lymph node status, and vascular invasion). Tumor morphology was also recorded (sclerosing type v papillary type), as previously described.27 Histologic data was confirmed by rereview of all slides by one pathologist (D.S.K.). Tumors were restaged according to the American Joint Committee on Cancer (AJCC) Staging Manual (sixth edition). Recurrence and survival data were recorded; the initial site of recurrence (distant v local/regional, as previously defined)2,4 was also recorded.
Tissue Array
Staining
Immunostaining Analysis Stained sections from tissue array were evaluated by a pathologist (D.S.K.) and graded for intensity on a scale of 0 to 4 (0 = none, 4 = most intense) and percent staining (p53, p21, p27, Bcl2, Mdm2, and cyclin D1); Ki-67 was graded only by the percentage of stained cells. Non-neoplastic tissue served as the internal control for each Ab to ensure the absence of nonspecific staining. Only staining of neoplastic biliary adenocarcinoma cells was scored; nonspecific staining or staining of other cell types was discounted. After compilation of the raw data, the results were dichotomized into positive or negative for each Ab, except Ki-67, for which an average percentage of positive-stained cells was calculated. For p53, p21, Bcl2, Mdm2, and cyclin D1, the criteria for positive staining were an intensity of  2 and more than 20% positive staining; p27 was considered positive if intensity was 1 and more than 10% of cells were positive. These reference values are consistent with those used in several prior studies.15,19,20,23,25,29 When the triplicate tissue cores were lost during sectioning/staining, the original full-face paraffin block was sectioned, stained, and evaluated as described earlier.
Statistics Hierarchical clustering was performed to partition the data based on protein expression and then compared with anatomic location, excluding patients with incomplete immunostaining data. To assess the clustering stability, an R index was calculated by perturbing the data through the introduction of random noise and then reclustering the perturbed data; this process is repeated multiple times, and the results are compared with the original cluster of unperturbed data.30 The R index is the proportion of times that a patient pair clusters the same way in the perturbed data sets as in the original data set. In the present study, the perturbed data sets were calculated by introducing a small probability (5%) of flipping the binary marker values rather than by the introduction of Gaussian noise.30
Demographics and Histopathology One hundred twenty-eight patients submitted to resection were analyzed (Distal, n = 18, 14%; GB, n = 32, 25%; Hilar, n = 55, 43%; and IH, n = 23, 18%). Fourteen GB patients had undergone a prior noncurative cholecystectomy before referral for definitive resection; in this group, tissue was obtained either from the original submitted cholecystectomy specimen or from residual cancer in the subsequently resected liver. Overall, 120 patients (93.8%) underwent a complete gross resection; eight resections were palliative, and 101 (78.9%) were R0. Ten patients had stage IV disease determined intraoperatively (n = 8) or on histologic review (n = 2). Papillary morphology was identified in 14 patients (Hilar, n = 8; GB, n = 5; and Distal, n = 1). Most tumors (n = 117, 91.4%) were AJCC stages I to III. Clinicopathologic details stratified by tumor location are listed in Table 1.
Immunohistochemistry Complete staining data with all Abs were obtained in 121 patients, and partial data were obtained in four patients. In three patients, no tumor was identified in any of the array cut sections, and additional tissue could not be obtained. p53, cyclin D1, Mdm2, and Bcl2 staining was absent in nonmalignant biliary epithelium; by contrast, p27 and p21 staining was seen in normal biliary cells. Significant differences in p27, cyclin D1, and Bcl2 expression were observed according to anatomic location, all of which were proportionately higher in IH (Table 2). p27 expression progressively decreased as tumor location moved from proximal to distal in the biliary tree. Cyclin D1 and Bcl2 did not manifest such a pattern but positive staining with these Abs was more common in IH compared with extrahepatic biliary tumors (Hilar, Distal, and GB). Ki-67 labeling index also varied according to tumor location; it was significantly lower in IH and Hilar compared with Distal and GB. Detectable p53 staining was somewhat higher in Distal (n = 9, 50%) compared with tumors arising more proximally in the biliary tract (n = 24, P = .083). Mdm2 overexpression was uniformly high in tumors at all sites; p21 expression also did not vary by location.
Of the 33 tumors with detectable p53, 27 (81.8%) were also Mdm2 positive compared with only six (18.2%) that were p53 positive and Mdm2 negative. By contrast, 65 (70.6%) of the 92 Mdm2-positive tumors were p53 negative.
Correlation With Pathologic Variables
Cyclin D1 overexpression was seen in 36 patients (32.4%) but was absent from all papillary tumors (P = .019). Bcl2-positive staining correlated inversely with both tumor stage (51.2% tumor stage  2 v 21.8% tumor stage > 2; P = .001) and TNM stage (51.5% TNM stage 1 v 25.0% TNM stage 2; P = .005). There was no difference in the proportion of Bcl2-positive tumors with (34.1%) and without (28.2%) nodal metastases (P = .51), but Bcl2 staining was absent from all tumors associated with distant metastases (0% with metastases v 37.1% without metastases; P = .046). p27 expression was also more common in tumors without nodal or distant metastases, but this difference was not significant.
Hierarchical Clustering
The R index of the cluster (measure of stability) was 0.86, suggesting a high level of reproducibility of the partitions. Additionally, when the clustering analysis was repeated without the papillary tumors, similar results were obtained, with only eight patients reclassified with this partition (Distal, n = 2; GB, n = 1; Hilar, n = 3; and IH, n = 2).
Recurrence and Survival Survival after an R0 resection was greatest for IH (46.3 months) and Hilar (42.7 months) compared with GB (19.3 months) and Distal (18.9 months; P = .007). Significant variables on univariate analysis included anatomic location, lymph node involvement, tumor stage, vascular invasion, and grade (Table 5). There was a trend toward improved survival in p53-negative, Mdm2-negative, and p27-positive tumors. The association between Mdm2 and survival was strongest for Hilar (median survival time, 75.7 months for Mdm2 negative v 33.5 months for Mdm2 positive; P = .087). Vascular invasion, lymph node metastases, absence of p27 expression, and Mdm2 overexpression independently predicted poor outcome on multivariate analysis (Table 5). Some caution should be exercised in interpreting the survival data because this was not a consecutive series. However, patients were selected only on the basis of tissue availability, and the analyses were restricted to R0 resections.
BTA remains a difficult clinical problem; the origin of BTAs from diverse anatomic sites and their rarity limit clinical and biologic investigation. Resection is the most effective treatment but is applicable in a minority of patients, whereas the remainder present with advanced disease and have few effective therapeutic options.26,27 Compounding the therapeutic challenge is a lack of understanding of the etiology and molecular pathogenesis of BTAs, most of which are sporadic. Although linked anatomically and histopathologically, it is unclear whether BTAs from different sites share common pathogenetic features. The primary aim of this study was to examine the entire spectrum of BTA to define anatomic site-related similarities and differences in cell cycle–regulatory protein expression. A major finding in this regard was the differential expression of the cycle-dependent kinase inhibitor p27, which was more common in proximal biliary tumors (IH and Hilar) compared with GB and Distal tumors. This observation is of particular interest given the association between low p27 expression and poor outcome, as documented in the present study and other reports.15,21-24 The findings suggest that the survival differences according to tumor location are at least partly related to differences in p27 expression. Other anatomic site–related differences included overexpression of Bcl2 and cyclin D1, both of which were significantly greater in the IH group. By contrast, no obvious differences or patterns were observed for p53, p21, and Mdm2, although aberrant p53 staining tended to be more common among Distal tumors. Mdm2 overexpression was uniformly high among all groups, proportionately higher than in some prior studies.18,31 It is evident from the expression profiles that BTAs differ in their genetic composition. Hierarchical clustering, which grouped tumors based on the combined differences, showed that GB and Hilar had similar and robust expression profiles that were unrelated to tumor stage and morphology and also different from those of Distal and IH. The findings suggest that, at least with respect to the proteins examined in the present study, GB and Hilar are more similar to each other than either is to IH or Distal and further suggest the possibility of common underlying pathogenetic features. Some overlap in the pathogenesis of these two tumors is plausible, given their anatomic proximity. However, despite the similarities in the global expression pattern of the proteins analyzed, there were notable individual differences between GB and Hilar, particularly in p27 and cyclin D1 expression. Additionally, the results do not account for differences in the clinical behavior between GB and Hilar, which are well described and may reflect differences in other molecular pathways.2 Several correlations were observed between the immunohistochemical staining results and histopathologic variables, most notably the absence of cyclin D1 overexpression in tumors with papillary morphology; aberrant p53 expression was also less common in this group. Consistent with these results, a recent study showed that cyclin D1 overexpression and aberrant p53 staining were infrequent in early intrahepatic papillary tumors but increased progressively with the degree of cellular atypia, although lower rates of Mdm2 overexpression and Ki-67 indices were observed, even among invasive tumors.32 These differences may simply reflect dissimilarities in the patient populations, which, in the latter study, focused on patients with hepatolithiasis. Shimonishi et al33 reported similar findings for aberrant p53 staining, and Abraham et al34 found no aberrant p53 staining in 12 patients with intrahepatic papillary tumors. On the basis of these results, it has been suggested that mutational inactivation of p53 plays little role in the pathogenesis of papillary tumors. However, functional p53 inactivation can occur as a result of Mdm2 binding, and such a mechanism is clearly possible because Mdm2 overexpression in papillary tumors was high in the present study. Indeed, the data revealed an important role for Mdm2 in all BTAs, regardless of morphology or anatomic site of origin. After an R0 resection, Mdm2 overexpression was associated with a higher rate of distant recurrence and independently predicted survival. This is consistent with several prior studies showing Mdm2 overexpression to be a common and important molecular event in the pathogenesis of several tumors, including cholangiocarcinoma.18,20,31 Della Torre et al20 reported an Mdm2 overexpression rate of 70%, and most Mdm2-positive tumors were also p53 positive; however, the rate of p53 mutations was only 15%, which is consistent with data showing that the coexistence of p53 mutations and Mdm2 overexpression is rare.35,36 Because Mdm2-bound p53 is largely targeted for proteosomal degradation,37 absent p53 staining would be expected in the setting of Mdm2 overexpression. On the contrary, p53-positive/Mdm2-positive tumors were relatively common in the present study and also in the study by Della Torre et al.20 Whether this is a result of Mdm2 splice variant(s) that retain p53 binding capacity but lack the ability to signal degradation is unclear.38,39 Additionally, Mdm2 has pro-oncogenic, p53-independent activity that may also be involved.40 Overall, however, the data from this and other studies strongly suggest that mutational inactivation of p53 is less common than Mdm2-mediated functional inactivation in BTA. In summary, BTAs exhibit differential expression of cell cycle–regulatory proteins according to tumor site of origin and morphology. The expression profiles of GB and Hilar were similar, suggesting the possibility of overlap in pathogenesis. Prognostic roles were identified for Mdm2 and p27, and the latter seemed to be a determinant of anatomic site-related differences in survival.
The authors indicated no potential conflicts of interest.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Fong Y, Blumgart LH, Lin E, et al: Outcome of treatment for distal bile duct cancer. Br J Surg 83:1712-1715, 1996[Medline] 2. Jarnagin WR, Ruo L, Little SA, et al: Patterns of initial disease recurrence after resection of gallbladder carcinoma and hilar cholangiocarcinoma: Implications for adjuvant therapeutic strategies. Cancer 98:1689-1700, 2003[CrossRef][Medline] 3. Lazaridis KN, Gores GJ: Cholangiocarcinoma. Gastroenterology 128:1655-1667, 2005[CrossRef][Medline] 4. Weber SM, Jarnagin WR, Klimstra D, et al: Intrahepatic cholangiocarcinoma: Resectability, recurrence pattern, and outcomes. J Am Coll Surg 193:384-391, 2001[CrossRef][Medline] 5. Patel T: Increasing incidence and mortality of primary intrahepatic cholangiocarcinoma in the United States. Hepatology 33:1353-1357, 2001[CrossRef][Medline] 6. MacFaul GR, Chapman RW: Sclerosing cholangitis. Curr Opin Gastroenterol 21:348-353, 2005[Medline] 7. Shaib Y, El-Serag HB: The epidemiology of cholangiocarcinoma. Semin Liver Dis 24:115-125, 2004[CrossRef][Medline] 8. Chen DW, Tung-Ping PR, Liu CL, et al: Immediate and long-term outcomes of hepatectomy for hepatolithiasis. Surgery 135:386-393, 2004[Medline] 9. Rashid A: Cellular and molecular biology of biliary tract cancers. Surg Oncol Clin N Am 11:995-1009, 2002[CrossRef][Medline] 10. Yang B, House MG, Guo M, et al: Promoter methylation profiles of tumor suppressor genes in intrahepatic and extrahepatic cholangiocarcinoma. Mod Pathol 18:412-420, 2005[CrossRef] 11. Kang YK, Kim WH, Lee HW, et al: Mutation of p53 and K-ras, and loss of heterozygosity of APC in intrahepatic cholangiocarcinoma. Lab Invest 79:477-483, 1999 12. Rullier A, Le BB, Fawaz R, et al: Cytokeratin 7 and 20 expression in cholangiocarcinomas varies along the biliary tract but still differs from that in colorectal carcinoma metastasis. Am J Surg Pathol 24:870-876, 2000[CrossRef][Medline] 13. Sherr CJ: Cancer cell cycles. Science 274:1672-1677, 1996 14. Momand J, Zambetti GP, Olson DC, et al: The mdm-2 oncogene product forms a complex with the p53 protein and inhibits p53-mediated transactivation. Cell 69:1237-1245, 1992[CrossRef][Medline] 15. Ito Y, Takeda T, Sasaki Y, et al: Expression and clinical significance of the G1-S modulators in intrahepatic cholangiocellular carcinoma. Oncology 60:242-251, 2001[CrossRef][Medline] 16. Polyak K, Lee MH, Erdjument-Bromage H, et al: Cloning of p27Kip1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell 78:59-66, 1994[CrossRef][Medline] 17. Katayose Y, Kim M, Rakkar AN, et al: Promoting apoptosis: A novel activity associated with the cyclin-dependent kinase inhibitor p27. Cancer Res 57:5441-5445, 1997 18. Horie S, Endo K, Kawasaki H, et al: Overexpression of MDM2 protein in intrahepatic cholangiocarcinoma: Relationship with p53 overexpression, Ki-67 labeling, and clinicopathological features. Virchows Arch 437:25-30, 2000[CrossRef][Medline] 19. Furubo S, Harada K, Shimonishi T, et al: Protein expression and genetic alterations of p53 and ras in intrahepatic cholangiocarcinoma. Histopathology 35:230-240, 1999[CrossRef][Medline] 20. Della Torre G, Pasquini G, Pilotti S, et al: TP53 mutations and mdm2 protein overexpression in cholangiocarcinomas. Diagn Mol Pathol 9:41-46, 2000[Medline] 21. Filipits M, Puhalla H, Wrba F: Low p27Kip1 expression is an independent prognostic factor in gallbladder carcinoma. Anticancer Res 23:675-679, 2003[Medline] 22. Taguchi K, Aishima S, Asayama Y, et al: The role of p27kip1 protein expression on the biological behavior of intrahepatic cholangiocarcinoma. Hepatology 33:1118-1123, 2001[CrossRef][Medline] 23. Tenjo T, Toyoda M, Okuda J, et al: Prognostic significance of p27(kip1) protein expression and spontaneous apoptosis in patients with colorectal adenocarcinomas. Oncology 58:45-51, 2000[CrossRef][Medline] 24. Fiorentino M, Altimari A, D'Errico A, et al: Low p27 expression is an independent predictor of survival for patients with either hilar or peripheral intrahepatic cholangiocarcinoma. Clin Cancer Res 7:3994-3999, 2001 25. Ito Y, Takeda T, Sasaki Y, et al: Bcl-2 expression in cholangiocellular carcinoma is inversely correlated with biologically aggressive phenotypes. Oncology 59:63-67, 2000[Medline] 26. Fong Y, Jarnagin W, Blumgart LH: Gallbladder cancer: Comparison of patients presenting initially for definitive operation with those presenting after prior noncurative intervention. Ann Surg 232:557-569, 2000[CrossRef][Medline] 27. Jarnagin WR, Bowne W, Klimstra DS, et al: Papillary phenotype confers improved survival after resection of hilar cholangiocarcinoma. Ann Surg 241:703-712, 2005[CrossRef][Medline] 28. Wreesmann VB, Sieczka EM, Socci ND, et al: Genome-wide profiling of papillary thyroid cancer identifies MUC1 as an independent prognostic marker. Cancer Res 64:3780-3789, 2004 29. Wiksten JP, Lundin J, Nordling S, et al: The prognostic value of p27 in gastric cancer. Oncology 63:180-184, 2002[Medline] 30. McShane LM, Radmacher MD, Freidlin B, et al: Methods for assessing reproducibility of clustering patterns observed in analyses of microarray data. Bioinformatics 18:1462-1469, 2002 31. Rayburn E, Zhang R, He J, et al: MDM2 and human malignancies: Expression, clinical pathology, prognostic markers, and implications for chemotherapy. Curr Cancer Drug Targets 5:27-41, 2005[CrossRef][Medline] 32. Ishikawa A, Sasaki M, Sato Y, et al: Frequent p16ink4a inactivation is an early and frequent event of intraductal papillary neoplasm of the liver arising in hepatolithiasis. Hum Pathol 35:1505-1514, 2004[Medline] 33. Shimonishi T, Zen Y, Chen TC, et al: Increasing expression of gastrointestinal phenotypes and p53 along with histologic progression of intraductal papillary neoplasia of the liver. Hum Pathol 33:503-511, 2002[Medline] 34. Abraham SC, Lee JH, Hruban RH, et al: Molecular and immunohistochemical analysis of intraductal papillary neoplasms of the biliary tract. Hum Pathol 34:902-910, 2003[Medline] 35. Leach FS, Tokino T, Meltzer P, et al: p53 Mutation and MDM2 amplification in human soft tissue sarcomas. Cancer Res 53:2231-2234, 1993 36. Momand J, Jung D, Wilczynski S, et al: The MDM2 gene amplification database. Nucleic Acids Res 26:3453-3459, 1998 37. Kubbutat MH, Jones SN, Vousden KH: Regulation of p53 stability by Mdm2. Nature 387:299-303, 1997[CrossRef][Medline] 38. Pinkas J, Naber SP, Butel JS, et al: Expression of MDM2 during mammary tumorigenesis. Int J Cancer 81:292-298, 1999[CrossRef][Medline] 39. Bartel F, Taubert H, Harris LC: Alternative and aberrant splicing of MDM2 mRNA in human cancer. Cancer Cell 2:9-15, 2002[CrossRef][Medline] 40. Zhang Z, Zhang R: p53-independent activities of MDM2 and their relevance to cancer therapy. Curr Cancer Drug Targets 5:9-20, 2005[CrossRef][Medline] Submitted October 28, 2005; accepted December 20, 2005.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test, respectively. Survival probabilities were computed using the Kaplan-Meier method and compared using the log-rank test; a multivariable model was performed using proportional hazards regression, including only variables with univariate P < .2. Survival analysis was limited to completely resected tumors (margin negative, R0).
lightest, gallbladder 
