|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 22, No 11 (June 1), 2004: pp. 2053-2060 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.11.046 Signal Transducer and Activator of Transcription-5 Activation and Breast Cancer PrognosisFrom the Lombardi Comprehensive Cancer Center, Department of Oncology, Georgetown University; Institute for Pathology, University of Basel, Basel, Switzerland; and Diomeda Life Sciences Inc, Sun Prairie, WI. Address reprint requests to Hallgeir Rui, Lombardi Comprehensive Cancer Center NRB E504, Georgetown University, 3970 Reservoir Rd NW, Washington, DC 20057-1469; e-mail: ruih{at}georgetown.edu
PURPOSE: Transcription factor signal transducer and activator of transcription-5 (Stat5) promotes breast epithelial cell differentiation. We retrospectively analyzed whether levels of active Stat5 in breast cancer were linked to clinical outcome. MATERIALS AND METHODS: Immunohistochemistry was used to detect active, tyrosine-phosphorylated Stat5 in paraffin-embedded breast cancer specimens from three archival tissue microarray materials A, B, and C. Material A included 19 healthy human breast tissues and a progression series of primary lymph node-negative, primary lymph node-positive, and metastatic breast cancer (n = 400). Materials B (n = 785) and C (n = 570) represented two independent arrays of unselected primary breast cancer specimens with clinical follow-up data. RESULTS: Material A demonstrated that Stat5 activation, but not Stat5 protein expression, was gradually lost during cancer progression, with detectable activation in 100% of healthy breast specimens compared with less than 20% of node-positive breast cancers and metastases. Stat5 activation in tumors of material B was associated with favorable prognosis. This observation was confirmed and extended in material C to include both breast cancer-specific survival and disease-free survival. Stat5 activation remained an independent prognostic marker after adjusting for patient age, tumor size, histological grade, estrogen receptor, progesterone receptor, and Her2/neu status by Cox multivariate analysis (hazard ratio, 2.0; P = .029). Stat5 activation was a particularly favorable marker in the lymph node-negative breast cancer subpopulation (hazard ratio, 7.5; P = .003). CONCLUSION: In our study, active Stat5 distinguishes breast cancer patients with favorable prognosis, and may be a useful marker for selection of more individualized treatment, especially in localized disease. These findings require confirmation in a large prospective study.
Signal transducer and activator of transcription-5 (Stat5) functions as a regulatory switch, maintaining cellular differentiation of breast epithelium.1–3 Stat5 becomes highly activated during pregnancy and lactation in response to elevated circulating prolactin, but Stat5 is also continuously activated at a basal level outside of pregnancy and lactation in healthy mouse and human breast epithelia.4 When activated, cytoplasmic Stat5 is phosphorylated on a conserved tyrosine residue, translocates to the cell nucleus, and binds to target DNA.5 In addition to promoting cellular differentiation, Stat5 may also stimulate survival and proliferation of breast epithelial cells.6 However, the involvement of Stat5 in breast cancer development and progression remains unresolved based on a limited few initial studies. In transgenic mouse mammary tumor models driven by transforming growth factor-alpha or SV40 large T antigen, cellular apoptosis was increased, and tumor formation delayed when the mice were crossed with Stat5a-deficient mice,7,8 suggesting a tumor promoting role of Stat5. Furthermore, in vitro overexpression of a dominant-negative Stat5 mutant in the human breast cancer cell line T47D, but not MCF7, was associated with increased apoptosis.9 In human breast cancer, one study reported low or undetectable levels of Stat5 DNA binding activity in extracts of 63 tumors,10 while another study detected nuclear localized Stat5a in 49% to 76% of 78 tumors.11 Using in situ detection of tyrosine phosphorylated Stat5 in paraffin-embedded tissue, we evaluated levels of active Stat5 in three independent human breast cancer materials, and correlated the data with metastatic tumor progression and clinical outcome in a retrospective study.
Breast Tumor Materials Paraffin-embedded breast cancer specimens from three archival tissue microarray materials A, B, and C were used in this retrospective study. The specimens were randomly selected from a consecutive series of more than 3,000 breast cancer specimens that were routinely examined in the Institute of Pathology at the Kantonsspital Basel (Basel, Switzerland) from 1985 through 1995. The use of the deidentified archival tissue specimens in research was approved by the ethics committee of the Basel University Hospital, and by the Georgetown University institutional review board. Material A. In addition to 19 healthy breast tissue specimens, material A included a breast cancer progression array containing 80 lymph node-negative breast cancer specimens, 80 lymph node-positive breast cancer specimens, and three matching lymph node metastases from the same lymph node-positive patients—a total of 240 metastases.12 Immunohistochemical analysis of levels of active Stat5 was uninformative for 20 specimens because of missing or unrepresentative tumor samples in the array sections studied, and assessable levels of active Stat5 were determined in 380 specimens. Among lymph node-negative tumors, approximately 68% were ductal carcinomas, 19% were lobular carcinomas, and 13% of tumors were of other histologic subtypes. The node-positive tumors consisted of primary breast cancers that each had three or more positive axillary lymph nodes with metastases larger than 0.5 cm in diameter, and for which tissue blocks were available from both the primary cancer and the axillary lymph node metastases. Among the lymph node-positive primary tumors, approximately 73% were ductal carcinomas, 13% were lobular carcinomas, and 13% were tumors of other histologic subtypes. Material B. Material B represented an array of 785 unselected primary breast cancer specimens for which clinical follow-up data in the form of raw survival data were available.13 Assessable levels of active Stat5 were determined in 683 specimens. Of the 683 tumors for which levels of active Stat5 were determined, histological subtypes were 65% ductal, 15% lobular, 1.6% medullary, 3.1% mucinous, 5.3% cribriform, 2.6% tubular, and 7.3% of other rare histological subtypes. Nodal status was not known for 87 patients, and of the remaining patients, 54% had lymph node-negative disease and 46% had lymph node-positive disease. The Elston-Ellis grade14 distribution was 31% grade 1, 40% grade 2, and 29% grade 3. The postoperative tumor-node-metastasis (pT) stage was 30% pT1, 53% pT2, 8% pT3, and 10% pT4, according to classification by the International Union Against Cancer.15 Material C. Material C represented an independent tumor tissue microarray of unselected primary breast cancer samples of variable stage for which breast cancer-specific survival data and disease-free (relapse-free) survival data were available for 553 and 570 patients, respectively.16 Immunohistochemical analysis of levels of active Stat5 was uninformative for 125 specimens because of missing or unrepresentative tumor samples in the array sections studied. Of the 443 tumors for which levels of active Stat5 were determined, histological subtypes were 76% ductal, 13%, lobular, 2.3% medullary, 1.9% mucinous, 1.9% cribriform, 1.4% tubular, 1.4% papillary, and 2.5% of other rare histological subtypes. pT stage was 39% pT1, 46% pT2, 4% pT3, and 11% pT4. Additional demographic and clinical characteristics of material C are presented in Table 1.
Immunohistochemistry of Active Stat5 Mouse monoclonal antibody AX1 was provided by Advantex BioReagents (Conroe, TX), and its specificity to tyrosine phosphorylated Stat5 in immunohistochemistry and immunoblotting has been validated by peptide competition, inducible phosphorylation studies, site-directed mutagenesis, and Stat5 knockout model analyses.4 Sections of paraffin-embedded, formalin-fixed tissues from healthy or malignant human breast were deparaffinized in xylene, rehydrated in graded ethanol, and heated in a pressure-cooker with antigen-retrieval solution AXAR1 (Advantex BioReagents). Following antigen retrieval, endogenous peroxidase activity was blocked by incubating slides in 0.3% H2O2 for 10 minutes at room temperature, and nonspecific binding of immunoglobulin was minimized by preincubation in normal goat serum for 2 hours at room temperature. The primary antibody AX1 recognizing phosphorylated tyrosine 694/699 of active Stat5a/b was diluted in 1% bovine serum albumin in phosphate-buffered saline and incubated with the samples at a final concentration of 0.6 µg/mL for 16 hours. Antigen-antibody complexes were detected using biotinylated goat antimouse secondary antibody (Biogenex, San Ramon, CA) followed by streptavidin-horseradish-peroxidase complex, using 3,3' diaminobenzidine as chromogen, and Mayer's hematoxylin as counterstain. Individual breast tumor samples were scored for active Stat5 levels as defined by nuclear localized and tyrosine phosphorylated Stat5 on a scale from 0 to 3, where 0 was undetectable and 1 to 3 represented detectable staining at three steps of increasing staining intensity and proportion of stained tumor cells (low, intermediate, and high). This scoring method corresponded to a simplified version of a general immunohistochemical scoring method.17 Scoring was performed in a blinded manner by two investigators (M.T.N. and H.R.), and cases of divergence were resolved by re-examination and consensus. Immunostaining with AX1 in formalin-fixed, paraffin-embedded archival samples verified that Stat5 was active in epithelial cells of normal, nonpregnant human breast tissue (Fig 1A), and was further activated in lactating human breast tissue (Fig 1B), consistent with previous observations.4 Also presented are examples of human primary breast carcinomas displaying either high or undetectable staining for active Stat5, including both infiltrating ductal (Figs 1C and 1D) and lobular breast carcinoma specimens (Figs 1E and 1F).
Statistical Methods Statistical analyses were performed using SPSS software version 10.1 (SPSS Inc, Chicago, IL). For material A, differences in ratios between multiple groups were initially analyzed by Kruskal-Wallis test, and post hoc by the Mann-Whitney U test with Bonferroni correction for multiple comparisons. For material B, overall survival analysis was performed by constructing Kaplan-Meier curves,18 and differences between curves were evaluated by log-rank test. For material C, both breast cancer-specific survival data and disease-free (relapse-free) survival data were available, and the hazard ratio [HR] for disease relapse or breast cancer-specific death was estimated using the Cox proportional hazards regression model.19 For the data presented in Table 2, the Cox model was used for both univariate and multivariate analyses. The assumption of proportional hazards was verified graphically. In the multivariate analyses, the Cox regression models were adjusted for Stat5 activation status (negative v positive), tumor size (continuous variable in millimeters), patient age at diagnosis (continuous variable in years), number of affected lymph nodes (continuous variable), tumor histological grade (III v I-II), estrogen receptor (ER) status (negative v positive), progesterone receptor status (negative v positive), and Her2/neu status (positive v negative). Because of incomplete information on adjuvant therapy (Table 1), adjuvant therapy was not included in the model. Variables with P values  .20 in univariate analysis were included in the initial multivariate regression models. Reduced models were obtained by forward selection with exclusion level of 0.10. Considered significant was P < .05, and 95% CI for the HR was calculated.
Stat5 is activated in healthy human breast epithelium and is gradually inactivated during metastatic progression. Immunohistochemical analysis of material A identified a gradual reduction in detectable activation of Stat5 during breast cancer progression. Specifically, while Stat5 was active in epithelial cells in all of the healthy human breast tissue specimens examined (100%; 19 of 19), Stat5 was active in 32% of node-negative breast cancers (24 of 76), in 19% (13 of 70) of node-positive breast cancers, and in 16% (38 of 234) of lymph node metastases (Fig 2). In contrast, immunohistochemistry of replicate samples with an antipan Stat5 antibody4 showed that Stat5 protein expression was less affected during breast cancer progression, being detectable in 80% of node-negative breast cancers (58 of 73), 94% (64 of 68) of node-positive breast cancers, and 84% (194/231) of lymph node metastases (Fig 2). These data suggest that gradual loss of Stat5 activation during breast cancer progression involves changes in tyrosine phosphorylation pathways rather than loss of Stat5 protein expression.
Survival analysis of material B. Kaplan-Meier analysis of overall survival of material B indicated that levels of active Stat5 corresponding to scores 1, 2, and 3 were associated with reduced risk of death relative to that associated with a score of 0 (P = .0264 by log-rank analysis; Fig 3). This analysis presented the first evidence of a favorable prognostic significance of Stat5 activation status of primary breast cancer. However, the patient survival data for material B did not discriminate between death due to breast cancer and death from other causes. To verify and extend the association between Stat5 activation and overall patient survival to breast cancer-specific clinical outcome, we analyzed Stat5 activation in an independent material C, which represented 570 well-characterized invasive breast malignancies of various stages and with clinical follow-up data, both in the form of breast cancer-specific, and relapse-free survival. For analysis of material C, undetectable staining for Stat5 activation was categorized as negative, and detectable staining levels of 1, 2, or 3 were classified as positive.
Active Stat5 is an independent marker of favorable prognosis in breast cancer. To assess the ability of Stat5 activation status to predict breast cancer-specific survival and disease progression, breast cancer specimens of material C were analyzed separately by univariate and multivariate survival analysis. By univariate analysis, positive Stat5 activation status was associated with significantly improved breast cancer-specific survival (P < .001) and disease-free survival (P < .001; Table 1). Consistent with previous observations, number of positive lymph nodes, tumor size, histological grade, ER, progesterone receptor, and Her2/neu status were also highly significant prognostic markers based on both breast cancer-specific survival and disease-free survival by univariate analysis (Table 1). Multivariate analysis using Cox's proportional hazard regression method19 to take all prognostic factors into account, showed that Stat5 activation status remained an independent prognostic factor for breast cancer-specific survival (HR, 2.0; P = .029) and disease-free survival (HR 1.9; P = .014; Table 1 and Fig 4A). In addition, number of positive lymph nodes, tumor size, tumor grade, and Her2/neu and ER status were independent prognostic factors.
Active Stat5 is associated with particularly favorable prognosis in node-negative breast cancer. Because lymph node status is regarded as the single most critical prognostic factor for breast cancer, and there is a lack of simple immunohistochemical markers to predict prognosis of node-negative breast cancer patients,20 we specifically analyzed the ability of Stat5 activation status to predict prognosis of node-negative patients. In the subpopulation of patients with localized, invasive breast cancer, univariate analysis identified Stat5 activation status, tumor size, and tumor grade as prognostic factors by breast cancer-specific survival and disease-free survival (Table 1B). In addition, Her2/neu status was a statistically significant marker in breast cancer-specific survival but not in disease-free survival by univariate analysis (Table 1B). Importantly, in patients with node-negative disease, Stat5 continued to be a significant and independent prognostic factor for breast cancer-specific survival (HR, 7.5; P = .003) and disease-free survival (HR, 4.5; P = .016; Table 1). The risk for disease progression and death due to breast cancer was therefore highly elevated in node-negative patients whose primary tumors were negative for active Stat5. This corresponded to breast cancer-specific 10-year survival rates as high as 97% (SE, 1.9) in node-negative patients with positive Stat5 activation status by Kaplan-Meier estimates, compared with approximately 72% (SE, 7.0) in patients with Stat5-negative tumors (Fig 4B). Thus, there was an approximately 25% survival benefit at 10 years associated with active Stat5 status among node-negative breast cancer patients. Other than Stat5 activation status, tumor size also remained an independent prognostic marker for patients with localized disease among prognostic factors in the present analysis. In addition, Her2/neu was an independent marker in breast cancer-specific survival, but not in disease-free survival by multivariate analysis (Table 1). In patients with established lymph node metastases, Stat5 activation status of the primary tumor was a prognostic factor of more marginal statistical significance based on univariate analysis of breast cancer-specific survival (P = .068) and disease-free survival (P = .022; Table 1). Correspondingly, multivariate analysis also indicated that Stat5 activation status may be less important as an independent prognostic factor in node-positive breast cancer, as reflected both by breast cancer-specific survival (HR, 1.8; P = .070) and disease-free survival (HR, 1.8; P = .044). Instead, number of affected lymph nodes, histologic grade, ER, and possibly Her2/neu status, were markers that had stronger independent prognostic value on survivorship in node-positive cases (Table 1). Clinically, this suggests that once detectable metastases have been established, Stat5 activity of the primary tumor seems to be of lesser prognostic value.
The present study determined that there was a gradual loss of activation of Stat5 in human breast cancer as the tumors became more advanced. The reduced frequency of Stat5 activation within metastases indicated that Stat5 signaling does not promote breast cancer metastasis. Instead, our data identified active Stat5 as an independent favorable prognostic marker for human breast cancer, based on statistically significant reduced risk of death associated with active Stat5 in two independent tumor materials. Importantly, active Stat5 predicted especially favorable prognosis in node-negative breast cancer. The biologic mechanisms that may explain Stat5 activation as a marker of favorable prognosis in breast cancer remain to be determined. However, loss of normal differentiation in tumor cells is particularly prominent at the transition from localized and surgically curable cancer to metastatic disease.21–23 Furthermore, epithelial-to-mesenchymal dedifferentiation is required for migration of breast carcinoma cells.24 As a recognized breast epithelial cell differentiation factor, Stat5 may directly suppress epithelial-to-mesenchymal dedifferentiation. Based on the present observations, loss of activation of Stat5 in primary breast cancer could represent a tumor progression event that contributes to increased risk of peritumoral or extravascular invasion. The data also indicate that once independent lymph node metastases have been established, Stat5 activity within the primary breast tumor is of lesser prognostic value. Mechanistic testing of Stat5 as a suppressor of breast cancer invasion and metastasis is now warranted. Despite the consistency of our clinical findings with the biology of Stat5 as a breast epithelial differentiation factor, several limitations of this work need to be addressed in follow-up investigations. First, the reported studies are retrospective. Second, the key patient material C was heterogenous in terms of adjuvant treatment and treatment information. It will therefore be important to validate the conclusions of this report in prospective studies with well-defined adjuvant therapies. Third, the particularly strong prognostic value of active Stat5 in node-negative breast cancer patients needs to be verified, because it was identified subsequent to initial statistical analysis of Material C as a whole. Fourth, while preliminary analyses of whole tumor sections indicate that Stat5 phosphotyrosine staining is homogenous within individual breast tumors (data not shown), the present tumor microarray studies need to be repeated on full-size tumor sections. Finally, the higher frequency of active Stat5 in primary breast cancer observed in materials A and C compared to material B is probably due to improved sensitivity, and additional studies will yield more accurate information. Provided that the results of the present work are verified in prospective studies, the most immediate utility of Stat5 would be to identify a subgroup of node-negative breast cancer patients with excellent prognosis. To this end, Stat5 activation may be useful alone or in conjunction with other new biochemical or expression signatures that reportedly forecast prognosis in node-negative breast cancer, including urokinase-type plasminogen activator and its inhibitor PAI-1,25 Cyclin E variants,26 and gene expression profiles.27,28 Stat5 activation status may improve individualized medicine for patients with node-negative breast cancer, especially when deciding on whether to choose potentially toxic and expensive adjuvant therapies.29 It will therefore be important to validate our initial observations in a large prospective cohort.
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Owns stock (not including shares held through a public mutual fund): Hallgeir Rui, Advantex BioReagents. Acted as a consultant within the last 2 years: Hallgeir Rui, Advantex BioReagents; Marja Nevalainen, Advantex BioReagents. Served as an officer or member of the Board of a company: Hallgeir Rui, Advantex BioReagents.
We thank Robert A. Kirken, Olli-P. Kallioniemi, and Richard G. Pestell for helpful suggestions and critical reading of the manuscript. We also thank Markus Zuber, Department of Surgery, University of Basel, Basel, Switzerland, Ossi R. Köchli, Universitäts-Frauenklinik, University of Basel, Basel, Switzerland, Frank Mross, Kreiskrankenhaus, Lörrach, Germany, and Holger Dieterich, Frauenklinik, Rheinfelden, Germany, for extensive support with clinical data analysis. Material support for this study has been provided by Advantex BioReagents, Conroe, Texas, and Advantex BioReagents has patents pending on the primary observations.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Liu X, Robinson GW, Wagner KU, et al: Stat5a is mandatory for adult mammary gland development and lactogenesis. Genes Dev 11:179–186, 1997 2. Teglund S, McKay C, Schuetz E, et al: Stat5a and Stat5b proteins have essential and nonessential, or redundant, roles in cytokine responses. Cell 93:841–850, 1998[CrossRef][Medline]
3. Xie J, LeBaron MJ, Nevalainen MT, et al: Role of tyrosine kinase Jak2 in prolactin-induced differentiation and growth of mammary epithelial cells. J Biol Chem 277:14020–14030, 2002
4. Nevalainen MT, Xie J, Bubendorf L, et al: Basal activation of transcription factor signal transducer and activator of transcription (stat5) in nonpregnant mouse and human breast epithelium. Mol Endocrinol 16:1108–1124, 2002 5. Gouilleux F, Wakao H, Mundt M, et al: Prolactin induces phosphorylation of Tyr694 of Stat5 (MGF), a prerequisite for DNA binding and induction of transcription. Embo J 13:4361–4369, 1994[Medline]
6. Miyoshi K, Shillingford JM, Smith GH, et al: Signal transducer and activator of transcrip-tion (Stat) 5 controls the proliferation and differentiation of mammary alveolar epithelium. J Cell Biol 155:531–542, 2001 7. Ren S, Cai HR, Li M, et al: Loss of Stat5a delays mammary cancer progression in a mouse model. Oncogene 21:4335–4339, 2002[CrossRef][Medline]
8. Humphreys RC, Hennighausen L: Signal transducer and activator of transcription 5a influences mammary epithelial cell survival and tumorigenesis. Cell Growth Differ 10:685–694, 1999 9. Yamashita H, Iwase H: The role of Stat5 in estrogen receptor-positive breast cancer. Breast Cancer 9:312–318, 2002[Medline]
10. Widschwendter A, Tonko-Geymayer S, Welte T, et al: Prognostic significance of signal transducer and activator of transcription 1 activation in breast cancer. Clin Cancer Res 8:3065–3074, 2002 11. Cotarla I, Ren S, Zhang Y, et al: Stat5a is tyrosine phosphorylated and nuclear localized in a high proportion of human breast cancers. Int J Cancer 108:665–671, 2004[CrossRef][Medline]
12. Monni O, Barlund M, Mousses S, et al: Comprehensive copy number and gene expression profiling of the 17q23 amplicon in human breast cancer. Proc Natl Acad Sci U S A 98:5711–5716, 2001 13. Simon R, Panussis S, Maurer R, et al: KIT (CD117)-positive breast cancers are infrequent and lack KIT gene mutations. Clin Cancer Res 10:178–183, 2004[CrossRef][Medline] 14. Elston CW, Ellis IO: Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: Experience from a large study with long-term follow-up. Histopathology 19:403–410, 1991[Medline] 15. International Union Against Cancer: TNM Classification of Malignant Tumors. Berlin, Germany, Springer, 1992 16. Kononen J, Bubendorf L, Kallioniemi A, et al: Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 4:844–847, 1998[CrossRef][Medline] 17. Allred DC, Harvey JM, Berardo M, et al: Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol 11:155–168, 1998[Medline] 18. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481, 1958[CrossRef] 19. Cox DR, Hinkley DV, Reid N, et al: Statistical theory and modelling, in Honour of Sir David Cox, FRS (ed 1). London, UK, Chapman and Hall, 1991 20. Hayes DF, Isaacs C, Stearns V: Prognostic factors in breast cancer: Current and new predictors of metastasis. J Mammary Gland Biol Neoplasia 6:375–392, 2001[CrossRef][Medline] 21. Freije JM, MacDonald NJ, Steeg PS: Nm23 and tumour metastasis: Basic and translational advances. Biochem Soc Symp 63:261–271, 1998[Medline] 22. Hart IR, Easty D: Tumor cell progression and differentiation in metastasis. Semin Cancer Biol 2:87–95, 1991[Medline] 23. Rivadeneira DE, Simmons RM, Christos PJ, et al: Predictive factors associated with axillary lymph node metastases in T1a and T1b breast carcinomas: Analysis in more than 900 patients. J Am Coll Surg 191:1–8, 2000[CrossRef][Medline] 24. Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2:442–454, 2002[CrossRef][Medline]
25. Look MP, van Putten WL, Duffy MJ, et al: Pooled analysis of prognostic impact of urokinase-type plasminogen activator and its inhibitor PAI-1 in 8377 breast cancer patients. J Natl Cancer Inst 94:116–128, 2002
26. Keyomarsi K, Tucker SL, Buchholz TA, et al: Cyclin E and survival in patients with breast cancer. N Engl J Med 347:1566–1575, 2002
27. van de Vijver MJ, He YD, van't Veer LJ, et al: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 347:1999–2009, 2002 28. van 't Veer LJ, Dai H, van de Vijver MJ, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415:530–536, 2002[CrossRef][Medline] 29. Thomssen C, Janicke F: Do we need better prognostic factors in node-negative breast cancer? Eur J Cancer 36:293–398, 2000[Medline] Submitted November 12, 2002; accepted March 15, 2004.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
