Changing Perceptions of Early-Stage Breast Cancer: Contributions from Histopathology and Biology A New Look at Node-Negative Breast Cancer

The Oncologist Changing Perceptions of Early-Stage Breast Cancer: Contributions from Histopathology and Biology A New Look at Node-Negative Breast Cancer NADIA HARBECK, a CHRISTOPH THOMSSEN b a Breast Center, University of Cologne, Cologne, Germany; b Frauenklinik der Universitaet Halle, Saale, Germany Key Words. Node-negative breast cancer Tumor grade Adjuvant chemotherapy Urokinase-type plasminogen activator upa Plasminogen activator inhibitor type 1 PAI-1 Disclosures: Nadia Harbeck: Honoraria: Sanofi-Aventis, Genomic Health; Christoph Thomssen: Consultant/advisory role: Sanofi- Aventis; Honoraria: Sanofi-Aventis; Research funding/contracted research: Sanofi-Aventis. The content of this article has been reviewed by independent peer reviewers to ensure that it is balanced, objective, and free from commercial bias. No financial relationships relevant to the content of this article have been disclosed by the independent peer reviewers. ABSTRACT Node-negative breast cancer is a prevalent form of the disease worldwide, particularly in regions with rigorous screening and disease awareness efforts. Although there is a common biology between node-negative and nodepositive breast tumors, it is still important to specifically address risk assessment and predictive factors in nodenegative breast cancer. The relative risks and benefits are more pronounced in these patients, but there is no single prognostic factor available for deciding whether to administer chemotherapy and selecting the best adjuvant chemotherapy regimen. In the absence of universal predictive factors, the trend is to give chemotherapy to all patients to ensure the highest possibility for cure. Tumor grade is important in that it is predictive of risk over time, but lacks standardization. Adjuvant! Online, a web-based algorithm, is also used to guide treatment decisions. Recently, the urokinase-type plasminogen activator (upa) and plasminogen activator inhibitor type 1 (PAI-1) biomarkers have been used to determine disease risk and, consequently, determine whether or not chemotherapy is needed. However, the assay requires a fresh-frozen tissue sample, which is not always feasible. The oncotype DX genomic assay has also been used to help prognosis estimation and treatment decisions. It is currently under evaluation in conjunction with the upa/pai-1 assay in the Plan B trial. The question that remains in adjuvant chemotherapy today for patients with node-negative disease is proper patient selection. Node-negative breast cancer does not automatically suggest a good prognosis, or the preclusion of chemotherapy benefits, and additional biomarkers are needed to help identify patients who do benefit from chemotherapy. The Oncologist 2010;15(suppl 5):29 38 INTRODUCTION Although there is a common biology between node-negative and node-positive tumors, it is still important to use node-negative disease as an example in addressing risk assessment and predictive factors in breast cancer as a whole. A considerable amount of research has been performed in node-negative breast cancer patients, but it is still valid to take a new look at this disease, because the risk-to-benefit analysis is more pronounced in these patients. Correspondence: Nadia Harbeck, M.D., Breast Center, University of Cologne, Cologne, Germany. Telephone: 49-221-478-97303; Fax: 49-221-478-97304; e-mail: nadia.harbeck@uk-koeln.de AlphaMed Press 1083-7159/2010/$30.00/0 doi: 10.1634/theoncologist.2010-S5-29 The Oncologist 2010;15(suppl 5):29 38 www.theoncologist.com

30 Node-Negative Breast Cancer Figure 1. Approximate percentages of patients with node-negative disease at time of diagnosis in different parts of the world. CHEMOTHERAPY IN NODE-NEGATIVE DISEASE Node-negative disease is a prevalent form of breast cancer worldwide (Fig. 1). In regions with widespread breast cancer screening and disease awareness among women, the rates of node-negative disease are likely to be in the range of 65% 70% of breast cancer patients. There is no single factor for choosing the type of adjuvant chemotherapy in node-negative disease. It is important to address the relative risk-to-benefit ratio of administering chemotherapy. If the risk in terms of disease progression is greater than the potential side effects of chemotherapy, the trend is to give chemotherapy to all patients to ensure the highest possibility for cure. This is in the absence of adequate, universal predictive factors. Breast cancer has been better characterized with Adjuvant! Online, a data-based tool that quantifies clinical instincts and guides decision making with respect to a patient s risk assessment. Data from Adjuvant! Online have shown that even patients with grade 1, estrogen receptor negative, node-negative tumors may have a high relapse rate of almost 20% over 10 years. The relapse rate is even higher in patients with grade 2 and grade 3 tumors [1]. Consequently, it is clear that node-negative breast cancer is by no means a low-risk disease that will only respond to and thus require endocrine therapy alone. International guidelines, including the German Arbeitsgemeinschaft Gynaekologische Onkologie guidelines, recommend the best available chemotherapy in nodepositive disease, which means anthracycline- and taxane-based regimens. In node-positive patients with a low degree of nodal involvement, the possibility of dying from their disease depends on established risk factors and tumor biology [2]. Tumor grade is certainly important in that it is predictive of risk over time, but it lacks standardization. Node-positive disease is associated with an overall mortality rate of approximately 20%, and oncologists do not hesitate to prescribe chemotherapy for these patients. In patients with node-negative disease, tumor size is a crude factor and does not reliably predict chemotherapy response in 2 cm versus 3 cm tumors, for instance. It is important to note that Adjuvant! Online is an epidemiologically based database. Data from this database show that the mortality risk may even be higher in patients with node-negative grade 3 tumors than the risk demonstrated in some patients with node-positive disease, suggesting that these patients do have a high enough risk to indicate adjuvant chemotherapy. However, there is still a good degree of uncertainty in determining whether patients with node-negative disease actually benefit from chemotherapy, leading many clinicians to hesitate before indicating chemotherapy in many node-negative patients. The Grupo Español de Investigación del Cáncer de Mama (GEICAM) study was the first and possibly the only completed trial to date with results that really focused on high-risk, node-negative breast cancer. The GEICAM trial assessed an optimal taxane- and anthracycline-based option (docetaxel doxorubicin cyclophosphamide) compared with the best available anthracycline-based option (fluorouracil doxorubicin cyclophosphamide) in high-risk, node-negative breast cancer patients. Patient risk was defined by grade, tumor size, age, and hormone receptor status. The investigators demonstrated that patients with node-negative disease did benefit significantly from the addition of a taxane to an anthracycline-based combination chemotherapy (Fig. 2) [3]. In addition, the U.S. Oncology (USON) 9735 trial in-

Harbeck 31 cluded both node-negative and node-positive patients and demonstrated the clinical value of a docetaxel cyclophosphamide (TC [DocC]) regimen, establishing TC as an anthracycline-free chemotherapy option in breast cancer management. The node-negative cohort experienced a similar degree of benefit to that shown in nodepositive patients, but the trial was not powered with a patient number adequate to reach significance for such subgroup analyses (Fig. 3) [4]. There is no hint that nodenegative patients did not respond as well or had different outcomes in the USON 9735 trial, which makes sense because nodal status per se is not a predictive factor for response to chemotherapy. If the estimated patient risk is high enough, there will be a benefit with the best available chemotherapy. A meta-analysis of docetaxel-containing adjuvant chemotherapy trials was recently published and further characterizes chemotherapy response in node-negative disease. The first message from that analysis is that large trials comparing two chemotherapy regimens in unselected patient groups are no longer of value. For instance, the results of these trials in node-negative patients were sometimes strikingly negative. It is possible that this is not because the taxane does not work in node-negative disease, but that patients were not selected appropriately. If only low-risk patients are selected, an effect with the better chemotherapy would be difficult to demonstrate with the patient numbers analyzed (Fig. 4) [5]. The meta-analysis also demonstrated that node-negative patients do derive a benefit, but the benefit is not significant for overall survival. This is not necessarily because docetaxel does not work in these patients, but because, again, the patient number was far too small, with not enough risk selection, to adequately demonstrate a clinical benefit. Figure 2. Grupo Español de Investigación del Cáncer de Mama (GEICAM) 9805: Disease-free survival; median follow-up, 67 months. Abbreviations: CI, confidence interval; FAC, fluorouracil doxorubicin cyclophosphamide; HR, hazard ratio; TAC, docetaxel doxorubicin cyclophosphamide (DocAC). From Martin M, Iluch A, Segui M et al. Multicenter, randomized phase III study of adjuvant chemotherapy for high-risk, node-negative breast cancer comparing TAC with FAC: 5-year efficacy analysis of the GEICAM 9805 trial. J Clin Oncol 2008; 26(15 suppl):abstract 542. Reprinted with permission. 2008 American Society of Clinical Oncology. All rights reserved. www.theoncologist.com NODE STATUS AND RELAPSE RATE Nodal status is an important risk factor for disease relapse. In fact, approximately 70% of node-negative patients respond sufficiently to surgery, radiotherapy, and endocrine therapy, and do not require additional chemotherapy. The dilemma that clinicians now face is that approximately 30% of node-negative patients will need chemotherapy because of their risk for recurrence, but there are no tools currently available to identify this subset of patients. Consequently, depending on the country, chemotherapy is still given to the majority of these patients (Fig. 5) [6]. RISK STRATIFICATION IN NODE-NEGATIVE DISEASE Recent advances have improved patient risk assessment and selection for chemotherapy in node-negative disease, but many questions remain. The prospective, randomized multicenter Chemo N0 trial was conducted in the 1990s and was innovative because it included only patients with node-negative disease, with no other risk stratification, evaluating urokinase-type plasminogen activator (upa) and plasminogen activator inhibitor type 1 (PAI-1) biomarkers in primary tumor tissue (Fig. 6). The upa/pai-1 biomarker is a protein-based enzyme-linked immunosorbent assay and is very well standardized. Researchers were able to stratify patients into groups with low and high upa/pai-1 values; those with low values received observation only, whereas those with high upa and/or PAI-1 values were randomized to receive either cyclophosphamide methotrexate fluorouracil (CMF) chemotherapy or observation. Another patient arm included those who refused randomization and chose observation or CMF chemotherapy [7]. The 10-year results of that trial revealed that half of the node-negative patients with low upa/pai-1 values and no systemic therapy whatsoever, not even endocrine therapy, had a 10-year survival rate of almost 90% [8]. Most likely, the administration of adjuvant endocrine therapy in these patients could have further increased the 10-year survival rate. Although adjuvant chemotherapy does not seem indicated in patients with low upa and low PAI-1 levels, clinicians would be more likely to recommend chemotherapy in higher risk patients with high upa and/or PAI-1 levels, and thus a high risk for relapse. Although researchers were not entirely sure whether these biomarkers would still be predictive over 10 years, these findings are reassuring.

32 Node-Negative Breast Cancer Figure 3. U.S. Oncology 9735 trial: TC regimen. Abbreviations: AC, doxorubicin cyclophosphamide; DFS, disease-free survival; q3w, every 3 weeks; ER, estrogen receptor; HR, hazard ratio; N, node; OS, overall survival; PR, progesterone receptor; R, randomize; TC, docetaxel cyclophosphamide. From Jones S, Holmes FA, O Shaughnessy J et al. Docetaxel with cyclophosphamide is associated with an overall survival benefit compared with doxorubicin and cyclophosphamide: 7-Year follow-up of US Oncology research trial 9735. J Clin Oncol 2009;27:1177 1183. Reprinted with permission. 2009 American Society of Clinical Oncology. All rights reserved. The Chemo N0 10-year analysis also looked at the impact of upa/pai-1 levels on disease-free survival in subgroups according to tumor grade in an as-treated subset of 406 patients who had not received any adjuvant systemic therapy. Decentralized tumor grade evaluations from 12 institutions with different pathologists showed that patients with grade 3 tumors had a higher risk for recurrence, suggesting that these patients do require chemotherapy. Meanwhile, among patients with grade 1 tumors, only two recurrences were observed over 10 years, confirming that these patients do not need chemotherapy. Grade 2 tumors still pose a clinical problem in assessing overall risk, but the upa/ PAI-1 information helped to distinguish those patients who had a low risk for recurrence from those who had a high risk of recurrence. Therefore, the Chemo N0 results are not just statistically significant, but are also clinically meaningful [8]. The Chemo N0 investigators also sought to determine whether high-risk patients according to upa/pai-1 levels benefited from chemotherapy, and were able to show that these patients did benefit from the addition of chemotherapy [8]. The risk for relapse was lower by almost half in patients treated with CMF chemotherapy, and other pooled analyses looking at upa/pai-1 levels with regard to chemotherapy response confirm that this is a real predictive effect. Patients with high upa/pai-1 values had a three times higher benefit from adjuvant chemotherapy, indicating that the aggressive process that these markers represent is very susceptible to early chemotherapy. If these patients receive chemotherapy in the metastatic setting, this seems to be too late and they will not respond well, according to published data. It is important to note that the 10-year survival curves from the Chemo N0 trial could be even higher with more modern chemotherapy regimens. The Chemo N0 group also compared actual patient outcomes in their trial, based on risk assessment according to upa/pai-1 levels, with Adjuvant! Online estimates, finding that the 10-year Adjuvant! Online survival data corresponded well to that observed in the Chemo N0 study. Meanwhile, Adjuvant! Online overestimates risk in patients with low upa/pai-1 levels and underestimates risk in those with high upa/pai-1 levels [9]. Adjuvant! Online is a good foundation tool, but knowing more about tumor biology can further refine patient evaluation and therapeutic selection. The Node-Negative Breast Cancer (NNBC)-3 study closed enrollment earlier this year with 4,000 patients (Figs. 7 and 8). It aimed to further characterize node-negative risk assessment and optimize adjuvant chemotherapy for high-risk node-negative breast cancer patients. Partici-

Harbeck 33 Figure 4. Docetaxel meta-analysis: DFS and OS according to nodal status. Abbreviations: CI, confidence interval; DFS, disease-free survival; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; OS, overall survival. From Laporte S, Jones S, Chapelle C et al. Consistency of effect of docetaxel-containing adjuvant chemotherapy in patients with early stage breast cancer independent of nodal status: Meta-analysis of 12 randomized clinical trials [abstract 605]. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Texas, December 9 13, 2009. Figure 5. Lymph node status and relapse rate in primary breast cancer. Abbreviations: LN, lymph node; RFS, relapse-free survival. From Harbeck N, Dettmar P, Thomssen C et al. Prognostic impact of tumor biological factors on survival in node-negative breast cancer. Anticancer Res 1998;18:2187 2197, with permission. Modified from Harbeck N. M.D. Ph.D. thesis, 1998. pating institutions could either evaluate upa/pai-1 or use a clinical pathological algorithm. Based on these findings, patients were stratified into low-risk and high-risk groups. Those classified as low risk did not receive chemotherapy, whereas those classified as high risk received either six cycles of fluorouracil epirubicin cyclophosphamide (FEC) chemotherapy or three cycles of FEC chemotherapy and three cycles of docetaxel [10]. It is important to note that 10% of the patients enrolled were from France, which as a country did not im- www.theoncologist.com

34 Node-Negative Breast Cancer Figure 6. Prospective randomized multicenter Chemo N0 trial in node-negative breast cancer. Abbreviations: CMF, cyclophosphamide methotrexate fluorouracil; LN, lymph node; PAI-1, plasminogen activator inhibitor type 1; R, randomize; T, tumor; upa, urokinase-type plasminogen activator. From Harbeck N, Schmitt M, Meisner C et al. Final 10-year analysis of prospective multicenter Chemo N0 trial for validation of ASCO-recommended biomarkers upa/pai-1 for therapy decision making in node-negative breast cancer. J Clin Oncol2009; 27(suppl 15):511, reprinted with permission, 2009 American Society of Clinical Oncology, all rights reserved; and Jänicke F, Prechtl A, Thomssen C et al.; German N0 Study Group. Randomized adjuvant chemotherapy trial in high-risk, lymph nodenegative breast cancer patients identified by urokinase-type plasminogen activator and plasminogen activator inhibitor type 1. J Natl Cancer Inst 2001;93:913 920, with permission. Figure 7. Node-Negative Breast Cancer-3 study design. Abbreviations: FEC, fluorouracil epirubicin cyclophosphamide; PAI-1, plasminogen activator inhibitor type 1; upa, urokinase-type plasminogen activator. From Thomssen C, Vetter M, Geurts-Moespot A et al. Determination of ASCO recommended prognostic factors upa and PAI-1 in daily clinical routine and the node-negative NNBC 3-Europe trial [abstract 1092]. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Texas, December 10 14, 2008, with permission. mediately adopt the upa/pai-1 test, suggesting that committed research centers can effectively change postsurgical tissue handling procedures so that the freshfrozen tissue upa/pai-1 biomarker assay can be performed. In addition, as evidenced by NNBC-3 enrollment patterns, clinicians sometimes need time to buy into the idea that these biomarker-guided treatment decisions are worthwhile, and then there are challenges in determining how to supply the laboratory staff with the necessary materials for the assay. As such, recruitment began slowly but ramped up toward the end of enrollment. Forty percent of the patients enrolled in NNBC-3 trial with grade 2 node-negative tumors have already been spared from chemotherapy, using the upa/pai-1 assay (Harbeck, ASCO presentation, 2010). Clinical data will probably be available in 2011.

Harbeck 35 Figure 8. Node-Negative Breast Cancer-3 trial (Arbeitsgemeinschaft Gynaekologische Onkologie, European Organization for Research and Treatment of Cancer Pathobiology Group, German Breast Group). The upa/pai-1 biomarker is included in both the American Society of Clinical Oncology (ASCO) and the German guidelines as a factor for use in treatment decisions based on its high level of evidence [11]. However, as mentioned previously, the upa/pai-1 assay requires a fresh-frozen tissue sample, and there is some reluctance to adopt the test in countries in which fresh-frozen tissue is not readily available after surgery. On the other hand, the assay is still the only level I evidence prognostic marker in breast cancer, it is relatively inexpensive at approximately 150 euros, and there is good quality control, so substantial advantages to the assay do exist. ONCOTYPE DX AND UPA/PAI-1 The oncotype DX assay is also recommended by ASCO and, being a paraffin-based assay, it is more commonly used in the U.S. to characterize tumor biology. The U.S.-led Trial Assigning IndividuaLized Options for Treatment (Rx) trial is currently evaluating the benefit of adjuvant chemotherapy in node-negative breast cancer patients with intermediate recurrence score (RS) values. The Trans- Anastrozole or Tamoxifen, Alone or in Combination study retrospectively looked at the RS determined by the oncotype DX assay within different nodal groups as a predictor of risk. The assay is prognostic and works independently of nodal status, yet it seems only helpful in patients with three or fewer involved lymph nodes to determine the clinical indication for chemotherapy. Patients with more than three involved lymph nodes have a substantial relapse risk even with low RS values so that adjuvant chemotherapy seems warranted (Fig. 9) [2]. Based on the available data, the West German Study Group Plan B trial (Fig. 10) was initiated in order to evaluate an anthracycline-free regimen in both node-negative and node-positive early breast cancer patients. Patients will be randomized to six cycles of TC or a standard regimen with four cycles of epirubicin cyclophosphamide followed by four cycles of docetaxel. That trial also aims to determine the efficacy of the risk group stratification used. The oncotype DX RS will be prospectively determined for all hormone receptor positive patients with zero to three positive nodes, and a very conservative threshold RS of 11 has been determined to forego chemotherapy. Patients categorized as lower risk will be recommended for endocrine therapy alone. UPA/PAI-1 DEVELOPMENT IN PERSPECTIVE The currently available trial data suggest that risk stratification in early-stage breast cancer patients has not improved considerably over the past 10 years. Clinicians still use tumor size, grade, and age, with the addition of biomarkers such as upa/pai-1, oncotype DX, or MammaPrint in some centers, to determine recurrence risk in patients with node-negative disease. Meanwhile, clinicians are eager to implement new techniques, including upa/pai-1 assessment, to better refine the decision to recommend chemotherapy in node-negative patients. The timeline for the development of upa/pai-1 as a reliable marker of disease risk in breast cancer demonstrates an almost 20-year span (Fig. 11). Patients and clinicians cannot wait this long for new biomarkers to be developed. However, now that the work has been completed, researchers are now moving quickly to incorpo- www.theoncologist.com

36 Node-Negative Breast Cancer Figure 9. Trans-Anastrozole or Tamoxifen, Alone or in Combination: For any recurrence score, the rate of distant recurrence increases with the number of positive nodes. Abbreviation: CI, confidence interval. From Dowsett M, Cuzick J, Wale C et al. Prediction of risk of distant recurrence using the 21-gene recurrence score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: A TransATAC study. J Clin Oncol 2010;28:1829 1834. Reprinted with permission. 2010 American Society of Clinical Oncology. All rights reserved. Figure 10. West German Study Group Plan B trial: Anthracycline free. Abbreviations: BC, breast cancer; Doc, docetaxel; EC, epirubicin cyclophosphamide; HER-2, human epidermal growth factor receptor; HR, hormone receptor; LN lymph node; PAI-1, plasminogen activator inhibitor type 1; RS, recurrence score; TC, docetaxel cyclophosphamide; upa, urokinase-type plasminogen activator. rate upa/pai-1 and other potential biomarkers into prospective trials to transition these tests into routine clinical practice. In discussing the upa/pai-1 assay, it is also important to address the fact that although tumor tissue was standard 20 years ago, core biopsy specimens are more common now. Researchers therefore need to demonstrate that assays such as the upa/pai-1 assay are reliable from core biopsies. Also, in tests on surgical specimens, clinicians need to be aware that some of these markers may have already changed as a result of wound-healing processes initiated by a prior core biopsy. Of note, one unicenter study has already determined that upa/pai-1 determination renders comparable results in core biopsies and sur-

Harbeck 37 Figure 11. Time line for development of upa/pai-1 assay. Abbreviations: AGO, Arbeitsgemeinschaft Gynaekologische Onkologie; BC, breast cancer; EORTC, PBG, European Organization for Research and Treatment of Cancer Pathobiology Group; GBG, German Breast Group; NN, node negative; NNBC-3, Node-Negative Breast Cancer-3; PAI-1, plasminogen activator inhibitor type 1; upa, urokinase-type plasminogen activator. Figure 12. upa and PAI-1 determination is feasible in clinical routine. Abbreviations: PAI-1, plasminogen activator inhibitor type 1; upa, urokinase-type plasminogen activator. From Thomssen C, Harbeck N, Dittmer J et al. Feasibility of measuring the prognostic factors upa and PAI-1 in core needle biopsy breast cancer specimens. J Natl Cancer Inst 2009;101:1028 1029, by permission of Oxford University Press. gical specimens, and is thus feasible in routine clinical practice (Fig. 12) [12]. A recent press release from Genomic Health, the manufacturer of the oncotype DX assay, found that nontumor www.theoncologist.com

38 Node-Negative Breast Cancer tissue taken too close to the core biopsy wound results in an elevated oncotype DX result [13]. This is similar to evidence found with the upa/pai-1 assay. Even though it is potentially important to quickly adopt these tests in clinical practice, clinicians should take into account that these modern tests are extremely sensitive. An understanding of tumor biology and precise tissue sampling methodology are both important, because these assays have the potential to determine whether a patient receives chemotherapy. CONCLUSIONS There is no substantial difference in the underlying tumor biology between node-negative and node-positive disease, REFERENCES 1 www.adjuvantonline.com, accessed November 1, 2010. 2 Dowsett M, Cuzick J, Wale C et al. Prediction of risk of distant recurrence using the 21-gene recurrence score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: A TransATAC study. J Clin Oncol 2010;28:1829 1834. 3 Martin M, Iluch A, Segui M et al. Multicenter, randomized phase III study of adjuvant chemotherapy for high-risk, node-negative breast cancer comparing TAC with FAC: 5-year efficacy analysis of the GEICAM 9805 trial. J Clin Oncol 2008;26(15 suppl):abstract 542. 4 Jones S, Holmes FA, O Shaughnessy J et al. Docetaxel with cyclophosphamide is associated with an overall survival benefit compared with doxorubicin and cyclophosphamide: 7-Year follow-up of US Oncology research trial 9735. J Clin Oncol 2009;27:1177 1183. 5 Laporte S, Jones S, Chapelle C et al. Consistency of effect of docetaxelcontaining adjuvant chemotherapy in patients with early stage breast cancer independent of nodal status: Meta-analysis of 12 randomized clinical trials [abstract 605]. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Texas, December 9 13, 2009. 6 Harbeck N, Dettmar P, Thomssen C et al. Prognostic impact of tumor biological factors on survival in node-negative breast cancer. Anticancer Res 1998;18:2187 2197. 7 Jänicke F, Prechtl A, Thomssen C et al.; German N0 Study Group. Randomized adjuvant chemotherapy trial in high-risk, lymph node-negative breast cancer patients identified by urokinase-type plasminogen activator and the question that remains in adjuvant chemotherapy today is in proper patient selection. Node-negative breast cancer does not automatically suggest a good prognosis, or the lack of a need for chemotherapy, and additional biomarkers are needed to help identify those node-negative patients who benefit from chemotherapy. AUTHOR CONTRIBUTIONS Conception/design: Nadia Harbeck Financial support: Nadia Harbeck Administrative support: Nadia Harbeck Provision of study material or patients: Nadia Harbeck, Christoph Thomssen Collection and/or assembly of data: Nadia Harbeck, Christoph Thomssen Data analysis and interpretation: Nadia Harbeck, Christoph Thomssen Manuscript writing: Nadia Harbeck, Christoph Thomssen Final approval of manuscript: Nadia Harbeck, Christoph Thomssen and plasminogen activator inhibitor type 1. J Natl Cancer Inst 2001;93: 913 920. 8 Harbeck N, Schmitt M, Meisner C et al. Final 10-year analysis of prospective multicenter Chemo N0 trial for validation of ASCO-recommended biomarkers upa/pai-1 for therapy decision making in node-negative breast cancer. J Clin Oncol 2009;27(suppl 15):511. 9 Euler U, Meisner C, Friedel C et al. Comparison of outcome prediction in node-negative breast cancer based on biomarkers upa/pai-1 or Adjuvant Online using the 10-year follow-up of the randomized multicenter Chemo N0 trial. J Clin Oncol 2006;24(suppl18):Abstract 534. 10 Thomssen C, Vetter M, Geurts-Moespot A et al. Determination of ASCO recommended prognostic factors upa and PAI-1 in daily clinical routine and the node-negative NNBC 3-Europe trial [abstract 1092]. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Texas, December 10 14, 2008. 11 Harris L, Fritsche H, Mennel R et al. American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J Clin Oncol 2007;25:5287 5312. 12 Thomssen C, Harbeck N, Dittmer J et al. Feasibility of measuring the prognostic factors upa and PAI-1 in core needle biopsy breast cancer specimens. J Natl Cancer Inst 2009;101:1028 1029. 13 Genomic Health. Genomic Health Announces Data Confirming Manual Microdissection of Biopsy Cavities Is Essential for Accurate Recurrence Risk Assessment in Early-Stage Breast Cancer Patients. Available at http:// investor.genomichealth.com/releasedetail.cfm?releaseid 414933, accessed April 5, 2010.