This Article Abstract Full Text (PDF) Supplementary Data Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Thuerigen, O. Articles by Lichter, P. Search for Related Content PubMed PubMed Citation Articles by Thuerigen, O. Articles by Lichter, P. Social Bookmarking                            What's this?

Gene Expression Signature Predicting Pathologic Complete Response With Gemcitabine, Epirubicin, and Docetaxel in Primary Breast Cancer

Olaf Thuerigen, Andreas Schneeweiss, Grischa Toedt, Patrick Warnat, Meinhard Hahn, Heidi Kramer, Benedikt Brors, Christian Rudlowski, Axel Benner, Florian Schuetz, Bjoern Tews, Roland Eils, Hans-Peter Sinn, Christof Sohn, Peter Lichter

From the Divisions of Molecular Genetics, Theoretical Bioinformatics, and Biostatistics, Deutsches Krebsforschungszentrum; and Departments of Gynecology and Obstetrics and Pathology, University of Heidelberg, Heidelberg, Germany

Address reprint requests to Andreas Schneeweiss, MD, Department of Gynecology and Obstetrics, University of Heidelberg, Hospitalstrasse, D-69115 Heidelberg, Germany; e-mail: andreas.schneeweiss{at}med.uni-heidelberg.de

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
PURPOSE: Primary systemic therapy (PST) with gemcitabine (G), epirubicin (E), and docetaxel (Doc) has resulted in a pathologic complete response (pCR) in 26% of primary breast cancer patients. This study was aimed at the identification of a gene expression signature in diagnostic core biopsy tissue samples that predicts pCR.

PATIENTS AND METHODS: Core biopsy samples from patients with operable primary breast cancer, T2-4N0-2M0, enrolled onto two phase I and II trials evaluating GEDoc (n = 48) and GE sequentially followed by Doc (GEsDoc; n = 52) as PST were snap frozen and subjected to RNA expression profiling. A signature predicting pCR was discovered in the training set (GEsDoc) applying a support vector machine algorithm, and performance of this classifier was validated on the independent test set (GEDoc) by receiver operator characteristics analysis.

RESULTS: We identified a signature consisting of 512 genes, which was enriched in genes involved in transforming growth factor beta and RAS-mediated signaling pathways, that predicts pCR with a sensitivity of 78%, a specificity of 90%, and an overall accuracy of 88% (95% CI, 75% to 95%). Apart from our signature, only HER2 overexpression was an independent predictor of pCR in multivariate analysis.

CONCLUSION: In conclusion, our gene expression signature allows prediction of pCR to PST containing G, E, and Doc with unprecedented high overall accuracy and robustness.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
In early breast cancer, preoperative (primary systemic) and postoperative (adjuvant) chemotherapy are equally effective in terms of disease-free and overall survival rates.¹ However, primary systemic therapy (PST) permits observation of tumor response to treatment. This is of particular importance because pathologic complete response (pCR), which is defined as the disappearance of all invasive cancer cells in the breast after PST, strongly correlates with improved long-term survival.^2-4 Conversely, patients not achieving pCR might not generally benefit because they experience the same long-term outcome as patients who do not respond.⁵

Several strategies have been successful in significantly increasing the pCR rate in patients with primary breast cancer, one of which is the sequential addition of docetaxel to doxorubicin-based combination PST regimens, resulting in pCR rates of approximately 25% to 30%.^6-8 Triple-combination therapy and sequential, dose-dense PST with gemcitabine (G), epirubicin (E), and docetaxel (Doc) also yielded promising pCR rates of 26%.^9,10

Nevertheless, chemotherapy is applied empirically, and not all patients benefit from this approach. Currently, there is no clinically useful molecular predictor of response to any cytotoxic drug used in the treatment of breast cancer.¹¹ Furthermore, single clinical or molecular parameters, such as tumor size, histology, hormone receptor or human epidermal growth factor receptor 2 (HER2) expression, and tumor grade, among others, show only weak association with response (summarized in Hortobagyi et al¹²). Partially, this might be a result of the heterogeneous definition of response. Recent technologic advances, however, have enabled researchers to scan the expression pattern of thousands of genes (ie, thousands of possible predictive factors) in individual tumors at once and to identify gene expression patterns that are predictive of response and outcome in breast cancer patients.^13-16

This study was designed to discover a gene expression profile that predicts pCR to PST with G, E, and Doc administered either as triple therapy (GEDoc; every 3 weeks) or as a dose-dense sequential therapy (GE Doc [GEsDoc] every 2 weeks) for a total of 18 weeks in primary breast cancer (T2-4N0-2M0).

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
Eligibility Criteria
Women with newly diagnosed nonmetastatic breast cancer presenting at the University of Heidelberg (Heidelberg, Germany) were enrolled onto this study if they had a biopsy-proven T2-4N0-2M0 primary breast cancer with sufficient amount of snap-frozen tissue collected for microarray analysis, were 18 to 65 years old, had an Eastern Cooperative Oncology Group performance status of 0 to 2, had not received any prior therapy for breast cancer, had adequate organ function, and gave written informed consent. No evidence of distant metastatic disease was assessed by physical examination, chest x-ray, liver ultrasound, bone scan, and full blood count. The study protocol was approved by the Joint Ethical Committee of the University of Heidelberg.

Treatment and Response
Within two sequential phase I and II studies, patients received one of the following two regimens: (1) five cycles of E (90 to 100 mg/m²) in combination with G (1,250 mg/m²) repeated every 2 weeks, sequentially followed by four cycles Doc (80 to 100 mg/m²) every 2 weeks (GEsDoc) with prophylactic pegfilgrastim; or (2) six cycles of G (800 mg/m²), E (60 to 90 mg/m²), and Doc (60 to 75 mg/m²) every 3 weeks (GEDoc) with prophylactic filgrastim. G (800 mg/m²) was repeated on day 8 of each cycle. Detailed drug schedules have been published elsewhere.^9,10 Clinical response after 6 weeks of PST was assessed by ultrasound according to Response Evaluation Criteria in Solid Tumors.¹⁷

Patients proceeded to surgery within 4 weeks after the last dose of chemotherapy. If breast-conserving surgery was not possible, a modified radical mastectomy was recommended.^9,10 Pathologic response was defined as no invasive tumor residuals in the removed breast tissue. All patients undergoing a breast-conserving procedure received standard radiotherapy to the remaining breast. Radiotherapy to the chest wall or regional lymph nodes was performed according to national standards.

Immunohistochemistry
Tumor tissue from core biopsy was embedded in paraffin. Tissue sections (approximately 2 µm thick) were stained using the system BioTek TechMate (BioTek Solutions, Newport Beach, CA). The primary antibodies were directed against (clones/lots in parentheses) HER2 (A0485), KI-67 (MIB-1), p53 (DO7), BCL2 (124), estrogen receptor (ESR1) (1D5), and progesterone receptor (PGR) (PR88) (reagents from DakoCytomation Ltd, Ely, United Kingdom). Nuclear staining was recorded for KI-67 and p53. BCL2- and HER2-specific immunoreactions were scored from 0 to 3, with respect to cell membrane staining. Hormone receptors were scored from 0 to 12.

Tissue Samples
At time of diagnosis, core biopsies were taken before PST using a 14-gauge needle, snap frozen in liquid nitrogen (–196°C) immediately after extraction, and stored at –80°C. One randomly selected biopsy for each patient was used for microarray analysis.

RNA Extraction
RNA was extracted from biopsy samples after milling of the frozen sample in a Mikro-Dismembrator S (B. Braun, Melsungen, Germany). Frozen cell powder was immediately disintegrated in 5 mL of Trizol (Invitrogen, Carlsbad, CA) solution. RNA purification was performed as recommended by the manufacturer. The aqueous phase containing RNA was mixed 1:2 with ethanol, applied to RNeasy mini columns (Qiagen, Hilden, Germany), and eluted as recommended.

Transcriptome Amplification and Labeling
Two micrograms of total RNA from each patient and Human Universal Reference RNA (Stratagene, La Jolla, CA) were used in a T7-polymerase–based transcriptome amplification method (TAcKLE procedure¹⁸). The performance of this protocol has been validated by comparison to an Affymetrix expression platform (Affymetrix, Santa Clara, CA).¹⁹

Microarrays
A set of 21,329 gene-specific 70mere oligonucleotides (Human Oligo set 2.1; Operon, Cologne, Germany) was printed in duplicate on glass slides coated with epoxy-silane (Schott Nexterion, Jena, Germany). Microarray production, prehybridization treatment, hybridization, and posthybridization washes, and data acquisition were performed as described previously.^18-20 All hybridization experiments were repeated with inversely labeled sample and reference.

Data Preprocessing
Result files containing all relevant raw data were processed using the in-house–developed ChipYard microarray analysis software (http://www.dkfz.de/genetics/ChipYard/), the statistical programming language R, and packages of the Bioconductor project.^21,22 Raw fluorescence intensity values were normalized applying variance stabilization.²³ Quality criteria were calculated as outlined in the Supplementary Methods. We assessed the quality of all hybridizations by generating scatter plots, gradient plots, and M/A plots using the Bioconductor package "limma." The GEsDoc data set showed a better hybridization quality compared with the GEDoc data set and was, therefore, chosen as the training set. The raw and normalized data are deposited in the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/; accession No. GSE4056).

Data Analysis
Preprocessed microarray data from patients receiving GEsDoc (n = 52) and GEDoc therapy (n = 48) were used as training and test sets, respectively, to discover and independently validate a gene expression profile that predicts pCR. The algorithms used, namely Support Vector Machine and Recursive Feature Elimination,^24,25 are described in detail in Supplementary Methods.

A penalized logistic regression analysis²⁶ was performed to compare the predictive value of the gene expression signature (based on the previously described binary classification into pCR and non-pCR patients) with conventional parameters predicting pCR (tumor grade, partial response after 6 weeks of chemotherapy, hormone receptor status, HER2 status, and tumor size) for the patients in the independent test set. As a result of the sample size, a wide CI was expected.

Validation of Microarray Results
The mRNA levels of selected candidate genes were validated by quantitative real-time reverse transcriptase polymerase chain reaction (RT-PCR). Templates were generated from 32 ng of tumor sample and reference RNA using SuperScript II (Invitrogen) and oligo-d(T)₂₁VN primer, respectively. For quantification, SYBR Green ROX Mix (ABgene, Epsom, United Kingdom) was used. To prevent amplification of genomic DNA, all primer pairs were designed to span exon-exon boundaries and tested with genomic DNA. Primer sequences are listed in Supplementary Table S1. RT-PCR was performed in a 7900RT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) with the following settings: initial denaturation at 95°C for 15 minutes, amplification, and quantification (95°C, 15 seconds; 60°C, 10 seconds, 72°C, 60 seconds; single fluorescence measurement; 50 cycles). Products were analyzed by gel electrophoresis and melting-curve analysis (60°C to 95°C with 0.1 K/sec continuous fluorescence measurements). Calculation of efficacy, normalization to reference RNA, and relative quantification versus nonregulated standard genes (DCTN2 and GALNAC4S-6ST) were performed according to published algorithms.^27,28

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
A total of 100 patients were enrolled between January 2002 and November 2004. Of these, 52 and 48 patients received chemotherapy with GEsDoc and GEDoc, respectively. Patient characteristics are listed in Table 1. From the core biopsies of these patients, expression profiles were assessed using DNA microarrays, which test for 21,139 human genes. Expression data obtained by microarray analysis were validated for a small subset of differentially expressed genes by RT-PCR (Fig 1). Differential expression was confirmed for the known marker genes ESR1 and HER2 as well as for BAMBI, DAPK2, LMO4, and SMAD3, but not for SRC.

View larger version (7K): [in this window] [in a new window]   Fig 1. Comparison of real-time quantitative polymerase chain reaction (RQ-PCR) and microarray data for selected genes (gene expression ratios of pathologic complete response [pCR] v non-pCR patients). ESR1 and HER2 genes as known markers are used as positive controls. BAMBI, LMO4, SMAD3, and SRC are members of the TGF-ß pathway; DAPK2 is the top-ranked gene within the signature.

View larger version (8K): [in this window] [in a new window]   Fig 2. Validation of gene expression predictor on the test set (patients who received triple therapy with gemcitabine, epirubicin, and docetaxel). (A) Receiver operator characteristics. Arrow marks selected ratio of sensitivity and specificity resulting in a maximal Youden's Index. (B) Performance of the gene expression signature shown for both patient groups and in total. pCR, pathologic complete response; PPV, positive predictive value; NPV, negative predictive value.

View larger version (21K): [in this window] [in a new window]   Fig 3. Functional annotation of the genes from the predictive signature by Gene Ontology (GO). Only 447 of the 512 genes were eligible for GO, and for 304 of these genes (68%), an annotation was obtained. Several genes may have multiple entries into different categories (therefore, the sum extends 100%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

Most information about the usefulness of gene expression arrays in human breast cancer is available for questions of predicting whether a primary breast tumor has metastasized or will metastasize to lymph nodes or distant organs (for an overview, see Weigelt et al²⁹); only a few studies have investigated gene signatures with respect to prediction of response to a PST.^16,30-32 A recent study by Rouzier et al³² investigated a series of 82 patients treated with paclitaxel followed by fluorouracil, doxorubicin, and cyclophosphamide, who were subdivided into four molecular subtypes by different means according to the work of Perou et al³³ and Sørlie et al.³⁴ Two of these molecular subgroups displayed high rates of pCR; these subgroups were the basal-like and the erbB2-positive (HER2-positive) subgroups. However, none of the genes associated with pCR in the basal-like subgroup was associated with pCR in the HER2-positive subgroup. Because there was a clear preselection of the sample sets based on a published breast cancer intrinsic gene set, the basal-like and HER2-positive subgroups in this study cannot serve as training and test sets. Future validation of the gene set identified in the basal-like subgroup, using an independent set of basal-like breast tumors, will allow the assessment of the predictive potential of this gene set with respect to pCR.

To date, three other published studies have aimed at the identification of gene signatures predicting response to PST. Notably, two of these trials did not use the end point of pCR,^30,31 which is the only one that had been shown to prospectively correlate with patient survival.^4,5 Because Chang et al³⁰ scored clinical response by a 50% tumor reduction, these data are not comparable to the data in our study, and it cannot be expected that their gene signature predicts pCR. Hannemann et al³¹ pooled pCR and near pCR for the end point of the trial. Because near pCR likely does not correlate with patient survival,⁵ this might explain why the authors were unable to identify a gene signature within their training set correlating with therapy response. Clearly, prediction of pCR as a valuable surrogate for survival is highly dependent on strict application of the criteria for pCR, as performed in the present study and in the trial by Ayers et al.¹⁶ In the latter study, a training set of 24 patients, including six pCR patients, was used to discover a predictive gene signature, which was validated in a test set of 18 patients including seven pCR patients. Although the overall accuracy amounted to 78% (95% CI, 52% to 94%), three of the seven pCR patients (sensitivity, 43%) and 11 of the 11 residual disease patients (specificity, 100%) were correctly predicted. Here, we present a study with a considerably larger series of tumors, allowing the definition of a predictive gene signature with improved robustness (88% overall accuracy; 95% CI, 75% to 95%). The main focus of our model to predict pCR was to achieve a high overall accuracy, resulting from a particularly high sensitivity of 78% and a good specificity of 90% compared with the results of Ayers et al.¹⁶ Optimization towards the sensitivity of the marker set was performed because our aim was the identification of the maximal number of patients benefiting from the PST. The current development of drug prescription based on improved molecular diagnostics (as exemplified in the case of trastuzumab administration depending on amplification and overexpression of HER2) emphasizes the necessity of identifying those patients who benefit from the therapy in a robust manner. Clearly, this development also creates a strong pressure for the development of further treatment options for those patients who do not benefit from current therapy regimens.

Although it is currently not possible to determine to what degree our predictive gene signature is discovering response to PST in general or specific effects of the administered drugs, the fact that the study by Ayers et al,¹⁶ which applied a different treatment (paclitaxel followed by fluorouracil, doxorubicin, and cyclophosphamide), identified a nonoverlapping signature might indicate that gene signatures predictive for PST response also contain drug-specific features. In contrast, the differences between our predictive signature and the signatures to predict metastases (eg, see van't Veer et al¹³) might be well explained by the different scientific questions aimed at considerably different biologic properties of breast tumors.

Moreover, for complex diseases like breast cancer, several sets of outcome-predictive genes may exist as an answer to the same scientific question because many genes can show a tight correlation to patient survival.³⁵ The biologic function of the genes contained in our signature, as revealed by the Gene Ontology database, is summarized in Figure 3 and shows a predominance of genes encoding enzymes and proteins binding to nucleic acids, many of which are transcriptional regulators. Consideration of related gene functions revealed the following four prominent groups of genes known to be often affected in tumorigenesis: genes coding for members of the TGF-ß pathway, an even larger number of RAS-related genes, and genes involved in DNA damage response and apoptosis (Table 3). An interesting feature of the TGF-ß pathway is its downstream effect on BRCA1 (Fig 4) and its association with DNA damage response.³⁶ The signature suggests that BRCA1 is inhibited because BAMBI and LMO4 are upregulated, whereas the possible BRCA1 inductor SRC is not, according to quantitative PCR-based expression analysis (Fig 1). It is tempting to speculate that tumors with reduced DNA repair capacity are more sensitive to the applied PST, which contains a purin analog (gemcitabine) to be incorporated into cellular DNA that is not effectively repaired. The second prominent group of genes encoding RAS-related proteins also provides a connection to DNA damage–associated cell growth regulation. However, the genes of a predictive signature are not necessarily the key players involved in the etiology of the investigated disease.³⁷ The task to exclusively find all genes with biologic importance for a given disease is not only a task that is different than the task of finding a set of genes with discriminative power, but it is also a task for which no mature methodology exists.

View larger version (6K): [in this window] [in a new window]   Fig 4. Schematic drawing illustrating the relationship between main members of the TGF-ß signaling pathway. Black and gray text represents proteins whose genes are contained or not contained in the predictive signature, respectively. Dashed lines represent that molecular binding is known but nature of interaction is unknown.

	Authors' Disclosures of Potential Conflicts of Interest

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
Although all authors completed the disclosure declaration, the following authors or their immediate family members 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. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Authors Employment Leadership Consultant Stock Honoraria Research Funds Testimony Other Andreas Schneeweiss Sanofi-Aventis (A); Lilly (B) Sanofi-Aventis (A); Lilly (B) Sanofi-Aventis (A); Lilly (B) Christian Rudlowski Sanofi-Aventis (A) Sanofi-Aventis (A) Florian Schuetz Sanofi-Aventis (A); Lilly (B)

Dollar Amount Codes (A) < $10,000 (B) $10,000-99,999 (C) $100,000 (N/R) Not Required

	Author Contributions

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 

Conception and design: Andreas Schneeweiss, Meinhard Hahn, Christian Rudlowski, Peter Lichter Financial support: Peter Lichter Administrative support: Andreas Schneeweiss, Meinhard Hahn, Roland Eils, Christof Sohn, Peter Lichter Provision of study materials or patients: Andreas Schneeweiss, Florian Schuetz, Hans-Peter Sinn, Christof Sohn Collection and assembly of data: Olaf Thuerigen, Andreas Schneeweiss, Grischa Toedt, Heidi Kramer, Christian Rudlowski, Florian Schuetz, Hans-Peter Sinn, Christof Sohn Data analysis and interpretation: Olaf Thuerigen, Andreas Schneeweiss, Grischa Toedt, Patrick Warnat, Benedikt Brors, Axel Benner, Florian Schuetz, Bjoern Tews, Hans-Peter Sinn, Christof Sohn, Peter Lichter Manuscript writing: Olaf Thuerigen, Andreas Schneeweiss, Grischa Toedt, Patrick Warnat, Meinhard Hahn, Bjoern Tews, Peter Lichter Final approval of manuscript: Olaf Thuerigen, Andreas Schneeweiss, Grischa Toedt, Patrick Warnat, Meinhard Hahn, Heidi Kramer, Benedikt Brors, Christian Rudlowski, Axel Benner, Florian Schuetz, Bjoern Tews, Roland Eils, Hans-Peter Sinn, Christof Sohn, Peter Lichter

 

	ACKNOWLEDGMENTS

 
We thank Johannes Blatter, Sabine Greulich, and Thomas Bauknecht (Lilly Deutschland), Allen Melemed (Eli Lilly & Company, Indianapolis, IN) and Gunther Bastert (Heidelberg) for support.

	NOTES

 
Supported by Research Fund from Lilly Deutschland GmbH.

Both O.T. and A.S. contributed equally to this study.

Presented in part at the 41st Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 13-17, 2005.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
1. Kaufmann M, von Minckwitz G, Smith R, et al: International expert panel on the use of primary (preoperative) systemic treatment of operable breast cancer: Review and recommendations. J Clin Oncol 21:2600-2608, 2003[Abstract/Free Full Text]

2. Fisher B, Bryant J, Wolmark N, et al: Effect of preoperative chemotherapy on the outcome of women with operable breast cancer. J Clin Oncol 16:2672-2685, 1998[Abstract]

3. Kuerer HM, Newman LA, Smith TL, et al: Clinical course of breast cancer patients with complete pathologic primary tumor and axillary lymph node response to doxorubicin-based neoadjuvant chemotherapy. J Clin Oncol 17:460-469, 1999[Abstract/Free Full Text]

4. Bear HD, Anderson S, Smith RE, et al: A randomized trial comparing preoperative (preop) doxorubicin/cyclophosphamide (AC) to preop AC followed by preop docetaxel (T) and to preop AC followed by postoperative (postop) T in patients (pts) with operable carcinoma of the breast: Results of NSABP B-27. Breast Cancer Res Treat 88:S16, 2004 (suppl, abstr)

5. Wolmark N, Wang J, Mamounas E, et al: Preoperative chemotherapy in patients with operable breast cancer: Nine-year results from National Surgical Adjuvant Breast and Bowel Project B-18. J Natl Cancer Inst Monogr 30:96-102, 2001[Medline]

6. Smith IC, Heys SD, Hutcheon AW, et al: Neoadjuvant chemotherapy in breast cancer: Significantly enhanced response with docetaxel. J Clin Oncol 20:1456-1466, 2002[Abstract/Free Full Text]

7. Bear HD, Anderson S, Brown A, et al: The effect on tumor response of adding sequential preoperative docetaxel to preoperative doxorubicin and cyclophosphamide: Preliminary results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 21:4165-4174, 2003[Abstract/Free Full Text]

8. von Minckwitz G, Raab G, Caputo A, et al: Doxorubicin with cyclophosphamide followed by docetaxel every 21 days compared with doxorubicin and docetaxel every 14 days as preoperative treatment in operable breast cancer: The GEPARDUO study of the German Breast Group. J Clin Oncol 23:2676-2685, 2005[Abstract/Free Full Text]

9. Schneeweiss A, Huober J, Sinn HP, et al: Gemcitabine, epirubicin and docetaxel as primary systemic therapy in patients with early breast cancer: Results of a multicentre phase I/II study. Eur J Cancer 40:2432-2438, 2004[Medline]

10. Schneeweiss A, Schuetz F, Rudlowski C, et al: Dose-dense primary systemic chemotherapy with gemcitabine plus epirubicin sequentially followed by docetaxel for early breast cancer: Final results of a phase I/II trial. Anticancer Drugs 16:1023-1028, 2005[Medline]

11. Bast RC Jr, Ravdin P, Hayes DF, et al: 2000 update of recommendations for the use of tumor markers in breast and colorectal cancer: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19:1865-1878, 2001[Abstract/Free Full Text]

12. Hortobagyi GN, Hayes DF, Pusztai L: Integrating newer science into breast cancer prognosis and treatment: A review of current molecular predictors and profiles. 38th Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 18-21, 2002, pp 191-202

13. 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]

14. 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[Abstract/Free Full Text]

15. Ahr A, Karn T, Solbach C, et al: Identification of high risk breast-cancer patients by gene expression profiling. Lancet 359:131-132, 2002[CrossRef][Medline]

16. Ayers M, Symmans WF, Stec J, et al: Gene expression profiles predict complete pathologic response to neoadjuvant paclitaxel and fluorouracil, doxorubicin, and cyclophosphamide chemotherapy in breast cancer. J Clin Oncol 22:2284-2293, 2004[Abstract/Free Full Text]

17. Therasse P, Arbuck SG, Eisenhauer EA, et al: New guidelines to evaluate the response to treatment in solid tumors: European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 92:205-216, 2000[Abstract/Free Full Text]

18. Schlingemann J, Thuerigen O, Ittrich C, et al: Effective transcriptome amplification for expression profiling on sense-oriented oligonucleotide microarrays. Nucleic Acids Res 33:e29, 2005

19. Schlingemann J, Habtemichael N, Ittrich C, et al: Patient-based cross-platform comparison of oligonucleotide microarray expression profiles. Lab Invest 85:1024-1039, 2005[CrossRef][Medline]

20. Wrobel G, Schlingemann J, Hummerich L, et al: Optimization of high-density cDNA-microarray protocols by ‘design of experiments.’ Nucleic Acids Res 31:e67, 2003

21. R Development Core Team: R: A Language and Environment for Statistical Computing. Vienna, Austria, R Foundation for Statistical Computing, 2004

22. Gentleman RC, Carey VJ, Bates DM, et al: Bioconductor: Open software development for computational biology and bioinformatics. Genome Biol 5:R80, 2004

23. Huber W, von Heydebreck A, Sultmann H, et al: Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 18:S96-S104, 2002 (suppl 1)

24. Vapnik V: The support vector method of function estimation, in Suykens JAK, Vandewalle J (eds): Nonlinear Modeling: Advanced Black-Box Techniques. Boston, MA, Kluwer Academic Publishers, 1998, pp 55-85

25. Guyon I, Weston J, Barnhill S, et al: Gene selection for cancer classification using support vector machines. Machine Learning 46:389-422, 2002[CrossRef]

26. Heinze G, Ploner M: Fixing the nonconvergence bug in logistic regression with SPLUS and SAS. Comput Methods Programs Biomed 71:181-187, 2003[Medline]

27. Pfaffl MW: A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29:e45, 2001

28. Korshunov A, Neben K, Wrobel G, et al: Gene expression patterns in ependymomas correlate with tumor location, grade, and patient age. Am J Pathol 163:1721-1727, 2003[Abstract/Free Full Text]

29. Weigelt B, Peterse JL, van't Veer LJ: Breast cancer metastasis: Markers and models. Nat Rev Cancer 5:591-602, 2005[CrossRef][Medline]

30. Chang JC, Wooten EC, Tsimelzon A, et al: Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer. Lancet 362:362-369, 2003[CrossRef][Medline]

31. Hannemann J, Oosterkamp HM, Bosch CA, et al: Changes in gene expression associated with response to neoadjuvant chemotherapy in breast cancer. J Clin Oncol 23:3331-3342, 2005[Abstract/Free Full Text]

32. Rouzier R, Perou CM, Symmans WF, et al: Breast cancer molecular subtypes respond differently to preoperative chemotherapy. Clin Cancer Res 11:5678-5685, 2005[Abstract/Free Full Text]

33. Perou CM, Sorlie T, Eisen MB, et al: Molecular portraits of human breast tumours. Nature 406:747-752, 2000[CrossRef][Medline]

34. Sørlie T, Perou CM, Tibshirani R, et al: Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 98:10869-10874, 2001[Abstract/Free Full Text]

35. Ein-Dor L, Kela I, Getz G, et al: Outcome signature genes in breast cancer: Is there a unique set? Bioinformatics 21:171-178, 2005[Abstract/Free Full Text]

36. Elliott RL, Blobe GC: Role of transforming growth factor Beta in human cancer. J Clin Oncol 23:2078-2093, 2005[Abstract/Free Full Text]

37. Markowetz F, Spang R: Molecular diagnosis: Classification, model selection and performance evaluation. Methods Inf Med 44:438-443, 2005[Medline]

Submitted October 21, 2005; accepted February 13, 2006.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				P. D. Williams, S. Cheon, D. M. Havaleshko, H. Jeong, F. Cheng, D. Theodorescu, and J. K. Lee Concordant Gene Expression Signatures Predict Clinical Outcomes of Cancer Patients Undergoing Systemic Therapy Cancer Res., November 1, 2009; 69(21): 8302 - 8309. [Abstract] [Full Text] [PDF]

				P. VICI, F. GIOTTA, L. DI LAURO, M. BRANDI, V. GEBBIA, P. FOGGI, V. LORUSSO, C. VITUCCI, D. SERGI, S. I. FATTORUSO, et al. Multicenter Phase II Trial of First-line Docetaxel/Gemcitabine in Advanced Breast Cancer Pretreated with Adjuvant Anthracyclines Anticancer Res, May 1, 2009; 29(5): 1841 - 1845. [Abstract] [Full Text] [PDF]

				L. Li, B. Fridley, K. Kalari, G. Jenkins, A. Batzler, S. Safgren, M. Hildebrandt, M. Ames, D. Schaid, and L. Wang Gemcitabine and Cytosine Arabinoside Cytotoxicity: Association with Lymphoblastoid Cell Expression Cancer Res., September 1, 2008; 68(17): 7050 - 7058. [Abstract] [Full Text] [PDF]

				P. Lonning, S Knappskog, V Staalesen, R Chrisanthar, and J. Lillehaug Breast cancer prognostication and prediction in the postgenomic era Ann. Onc., August 1, 2007; 18(8): 1293 - 1306. [Abstract] [Full Text] [PDF]

				F. Peintinger, K. Anderson, C. Mazouni, H. M. Kuerer, C. Hatzis, F. Lin, G. N. Hortobagyi, W. F. Symmans, and L. Pusztai Thirty-Gene Pharmacogenomic Test Correlates with Residual Cancer Burden after Preoperative Chemotherapy for Breast Cancer Clin. Cancer Res., July 15, 2007; 13(14): 4078 - 4082. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplementary Data Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Thuerigen, O. Articles by Lichter, P. Search for Related Content PubMed PubMed Citation Articles by Thuerigen, O. Articles by Lichter, P. Social Bookmarking                            What's this?