This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by van de Vijver, M. Search for Related Content PubMed PubMed Citation Articles by van de Vijver, M.

Gene-Expression Profiling and the Future of Adjuvant Therapy

Department of Pathology, Netherlands Cancer Institute, Amsterdam, The Netherlands

Key Words. Gene-expression profiling • Prognostic tests • Predictive tests • Adjuvant • Neoadjuvant

Correspondence: Marc van de Vijver, M.D., Department of Pathology, Netherlands Cancer Institute, Plesmanlaan 121, 1066CX Amsterdam, The Netherlands. Telephone: 31-20-5122750; Fax: 31-20-5122759; e-mail: m.vd.vijver{at}nki.nl

Received September 5, 2005; accepted for publication September 5, 2005.

	LEARNING OBJECTIVES

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
After completing this course, the reader will be able to:

1. Explain how clinical experience influences treatment selection and how this is likely to be refined by molecular diagnosis in the future.
   
2. Describe the potential of gene profiling in personalized cancer medicine.
   
3. Discuss some of the limitations of clinical medicine as it relates to the rapid application of gene-expression profiling.
   

	ABSTRACT

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
Gene-expression profiling can help distinguish between patients at high risk and those at low risk for developing distant metastases, and so identify patients for adjuvant therapy. For several years, the Netherlands Cancer Institute has been working on gene-expression profiling of breast cancer using a microarray platform containing 25,000 genes. Using supervised classification, a prognostic classifier consisting of 70 genes could be identified. In addition to providing prognostic information, gene profiling should also enable us to detect patients who are likely to respond to particular adjuvant interventions. Well-known predictors for response to systemic therapy include estrogen receptor status HER-2 status, c-kit mutation, and epidermal growth factor receptor mutation. Because of the long periods required for predicting responsiveness in the adjuvant setting, neoadjuvant trials promise far quicker results. Several neoadjuvant studies are under way or planned to investigate gene-expression profiling as a means of predicting the therapeutic response to doce taxel (Taxotere^®; Aventis Pharmaceuticals Inc., Bridgewater, NJ, http://www.aventispharma-us.com), paclitaxel (Taxol^®; Bristol-Myers Squibb, Princeton, NJ, http://www.bms.com), cyclophosphamide, and doxorubicin (Adriamycin^®; Bedford Laboratories, Bedford, OH, http://www.bedfordlabs.com) in breast cancer patients. It is expected that in the coming years an increasing number of novel prognostic and predictive tests will help in guiding the adjuvant systemic treatment of breast cancer and other malignancies.

	INTRODUCTION

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
In breast cancer in the adjuvant setting, as with other tumors, the first step toward individualizing therapy is to distinguish those patients who need treatment from those who do not. Bear in mind, for example, that even poor- prognosis patients, as defined by clinicopathological factors, with no adjuvant therapy have a 50%–60% risk for developing distant metastases, indicating that 40%–50% of these patients will remain free of disease without any adjuvant systemic therapy [1–3]. The second step is to tailor treatment to the characteristics of the specific tumor, given that certain cancers respond better to one form of adjuvant therapy and others to a different form.

Although still at an early stage of development, gene-expression profiling is now contributing both to the identification of patients at greatest risk for relapse and to the selection of treatment regimens according to the likely sensitivity of the tumor.

Several years ago, the Netherlands Cancer Institute (NKI) in collaboration with Roseatta Inpharmatics (Seattle, WA, http://www.rii.com) began working with a 25,000-gene microarray. This array requires fresh-frozen tumor rather than paraffin-embedded specimens. The array is hybridized with fluorescently labeled cRNA isolated from the tumor and reference cDNA, giving a readout of the expression of 25,000 genes [4,5].

While it is theoretically possible that a handful of hitherto undiscovered genes are highly prognostic, or predictive of response to a specific therapy, this has not been the case to date. We find that 50–100 genes need to be assayed and that the differences in expression between tumors with good prognosis and those with poor prognosis is relatively subtle.

	GENE-EXPRESSION PROFILING AND PROGNOSIS

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
In an initial study, tumor tissue was obtained from 78 breast tumors [6]. All patients were aged under 55 years. Tumors were <5 cm diameter, and no patient had lymph node involvement. Patients were all treated at the NKI, and outcome over 5 or more years was known.

From the initial pool of 25,000 genes, a system of supervised classification (Fig. 1) identified a group of 70 genes that were associated with poor prognosis; that is, different levels of expression tended to be found in the 34 patients who developed distant metastases within 5 years compared with the 44 patients with no recurrence in this period. The threshold was set at 10% for false negatives (Fig. 2). A subsequent validation study investigated the prognostic power of these 70 genes in 295 patients, some of whom were lymph node positive (but all aged under 53 years) [7]. Figure 3 shows that patients with a good prognosis signature had a <15% risk for developing distant metastases over 10 years and a <10% risk for dying. Patients with tumor genes associated with a poor prognosis had a 50% risk for distant metastases and a 50% mortality rate. Further work, conducted in collaboration with Stanford University, showed that the prognostic value of the 70-gene assay could be refined [8]. This was achieved by the inclusion of genes associated with a wound response in fibroblasts, which the Stanford group had previously shown to be predictive of outcome in breast cancer [9].

View larger version (28K): [in this window] [in a new window]   Figure 1. Supervised classification for prognosis. From [6], with permission.

View larger version (52K): [in this window] [in a new window]   Figure 2. Supervised classification prognosis: "Leave-one-out" crossvalidation. From [6], with permission.

View larger version (28K): [in this window] [in a new window]   Figure 3. Validation study investigating the prognostic power of the 70 genes in stage I and II patients (n = 295). From 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 2002; 347:1999–2009[Abstract/Free Full Text]. ©2002 Massachusetts Medical Society. All rights reserved. Adapted with permission, 2005.

View larger version (14K): [in this window] [in a new window]   Figure 4. Combining gene-expression signatures. From [8], with permission.

Interestingly, genes identified by that group, who used an Affymetrix platform, showed very little overlap with those identified at the NKI in Amsterdam; only three were common to the two expression profiles. However, survival data showed the Rotterdam gene profile to have much the same prognostic value as the profile we had developed. A greater degree of overlap would have enhanced confidence in our identification of the exact set of genes needed to predict outcome. Nevertheless, we are confident that the two datasets, ultimately, will prove compatible.

	GENE PROFILES AS PREDICTORS OF RESPONSE TO THERAPY

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
Clearly, several factors predicting response to therapy of various kinds have already been established. Thus, estrogen receptor positivity predicts response to hormonal therapy, HER-2 positivity predicts response to trastuzumab (Herceptin^®; Genentech, Inc., South San Francisco, CA, http://www.gene.com), presence of c-kit mutation predicts response to imatinib (Gleevec^®; Novartis Pharmaceuticals Corporation, East Hanover, NJ, http://www.pharma.us.novartis.com), and presence of epidermal growth factor receptor mutation predicts response to gefitinib (Iressa^®; AstraZeneca Pharmaceuticals, Wilmington, DE, http://www.astrazeneca-us.com). However, we are still far from being able to individualize therapy with any great degree of confidence.

Two groups have attempted to predict responsiveness to tamoxifen (Nolvadex^®; AstraZeneca Pharmaceuticals). Jansen et al. [11] produced a molecular classification of tamoxifen-resistant breast carcinomas by gene-expression profiling, and Ma et al. [12] reported that a two-gene expression ratio predicted clinical outcome in breast cancer patients treated with tamoxifen.

However, attempts to predict responsiveness in the adjuvant setting inevitably require very long periods of follow-up to generate outcome data. Neoadjuvant trials promise far quicker results, but data published so far relate only to small series of patients.

The Baylor group of Chang et al. [13] investigated gene-expression profiling as a means of predicting the therapeutic response to docetaxel (Taxotere^®; Aventis Pharmaceuticals Inc., Bridgewater, NJ, http://www.aventispharma-us.com) in breast cancer patients. There were 24 cases in the training set and six for validation. At MD Anderson Cancer Center, Ayers et al. [14] looked at gene-expression profiles in a similar number of patients as a means of predicting pathological complete response to neoadjuvant paclitaxel (Taxol^®; Bristol-Myers Squibb, Princeton, NJ, http://www.bms.com) and fluorouracil, doxorubicin (Adriamycin^®; Bedford Laboratories, Bedford, OH, http://www.bedfordlabs.com), and cyclophosphamide. Many centers are now investigating the predictive value of the gene profiling of biopsied tumor in larger groups of patients.

At the NKI in Amsterdam, we have concluded a small study in 49 patients with locally advanced breast cancer randomized to neoadjuvant chemotherapy with either doxorubicin plus cyclophosphamide or docetaxel plus cyclophosphamide (each for six cycles) prior to surgery and/or radiotherapy and tamoxifen (Fig. 5) [15]. To date, it has not been possible to identify a gene profile in initial biopsy material that predicts responsiveness to one or the other of the alternative regimens. This will very likely require the study of a large number of patients. However, there is little doubt that genes predictive of response will eventually be found.

View larger version (16K): [in this window] [in a new window]   Figure 5. Predicting responsiveness to neoadjuvant chemotherapy: study design. Abbreviations: AC, doxorubicin/cyclophosphamide; AD, docetaxel/cyclophosphamide; MRI, magnetic resonance imaging. From [15], with permission.

	GENE PROFILING IN ROUTINE CLINICAL PRACTICE

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

The aim was to look at T1 and T2, node-negative patients aged less than 60 years, with a view to establishing a 70-gene prognostic profile for these patients. The protocol required standardized acquisition of tissue using punch biopsy, with the specimen being placed in a commercially available reagent (RNAlater^®;Qiagen, Venlo, The Netherlands, http://www1.qiagen.com/SelectCountry.aspx, conveyed by conventional mail, and frozen by the receiving center within 7 days of the specimen being taken. Extracted RNA is analyzed using a custom-made 70-gene microarray test, and patients are classified into good and poor prognosis groups.

In 60 weeks, 251 patients have been entered into the RASTER study, and 133 microarray tests have been conducted. Since the study is confined to lymph node–negative patients, and lymph node status is established only at subsequent surgery, the main reason for tissue not being analyzed (in 24% of cases) is that the patients were identified as node positive. Technical problems, such as the tumor being <50% of the sample, poor RNA quality, and poor tissue preservation, were minor considerations. Results of the study are awaited, but the initial phase should give us confidence that procedures for acquiring material for genetic profiling are robust.

	PROSPECTIVE VERSUS RETROSPECTIVE VALIDATION

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
Recently, a company called Genomic Health (Redwood City, CA, http://www.genomichealth.com) has developed a system that can establish gene profiles in material that has been paraffin-embedded [16]. Using this system, they have collaborated with the National Surgical Adjuvant Breast and Bowel Project (NSABP) to identify and validate a 21-gene prognostic classifier using retrospectively collected paraffin-embedded tumors from patients treated in NSABP randomized clinical trials. Such randomized controlled trials provide the ideal environment in which to assess the predictive value of gene profiles.

The fact that the 70-gene assay requires frozen material may limit its use on archival samples. The fact that the TRANSBIG validation study, using retrospective material, found that the 70-gene profile was not a strong prognostic factor needs to be taken into account. Prospective clinical studies are under way to further test microarray-based prognostic tests. In the Microarray for Node-Negative Disease May Avoid Chemotherapy (MINDACT) trial, more than 5,000 patients will be randomized to treatment according to their 70-gene profile as well as clinical prognostic factors, and results are awaited with interest.

Gene-expression profiling and other high-throughput techniques are helping to discover novel prognostic tests and novel therapy response–predicting tests. These tests will help in guiding adjuvant systemic treatment.

	DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 
The author indicates no potential conflicts of interest.

	REFERENCES

Top Learning objectives Abstract Introduction Gene-expression profiling and... Gene profiles as predictors... Gene profiling in routine... Prospective versus retrospective... Disclosure of potential... References

 

1. Early Breast Cancer Trialists’ Collaborative Group. Polychemotherapy for early breast cancer: an overview of the randomised trials. Lancet 1998;352:930–942.[CrossRef][Medline]
2. Robain M, Pierga JY, Jouve M et al. Predictive factors of response to first-line chemotherapy in 1426 women with metastatic breast cancer. Eur J Cancer 2000;36:2301–2312.
3. Chang J, Clark GM, Allred DC et al. Survival of patients with metastatic breast carcinoma: importance of prognostic markers of the primary tumor. Cancer 2003;97:545–553.[CrossRef][Medline]
4. ‘t Hoen PA, de Kort F, van Ommen GJ et al. Fluorescent labelling of cRNA for microarray applications. Nucleic Acids Res 2003;31:e20.[Abstract/Free Full Text]
5. Sterrenburg E, Turk R, Boer JM et al. A common reference for cDNA microarray hybridizations. Nucleic Acids Res 2002;30:e116.[Abstract/Free Full Text]
6. van’t Veer LJ, Dai H, van de Vijver MJ et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002;415:530–536.[CrossRef][Medline]
7. 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 2002;347:1999–2009.
8. Chang HY, Nuyten DS, Sneddon JB et al. Robustness, scalability, and integration of a wound-response gene expression signature in predicting breast cancer survival. Proc Natl Acad Sci U S A 2005;102:3738–3743.[Abstract/Free Full Text]
9. Chang HY, Sneddon JB, Alizadeh AA et al. Gene expression signature of fibroblast serum response predicts human cancer progression: similarities between tumors and wounds. PLoS Biol 2004;2:E7. Epub 2004 Jan 13.[CrossRef][Medline]
10. Wang Y, Klijn JG, Zhang Y et al. Gene-expression profiles to predict distant metastasis of lymphnode-negative primary breast cancer. Lancet 2005;365:671–679.[Medline]
11. Jansen MP, Foekens JA, van Staveren IL et al. Molecular classification of tamoxifen-resistant breast carcinomas by gene expression profiling. J Clin Oncol 2005;23:732–740.[Abstract/Free Full Text]
12. Ma XJ, Wang Z, Ryan PD et al. A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell 2004;5:607–616.[CrossRef][Medline]
13. 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 2003;362:362–369.[CrossRef][Medline]
14. 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 2004;22:2284–2293.[Abstract/Free Full Text]
15. Hannemann J, Oosterkamp HM, Bosch CA et al. Changes in gene expression associated with response to neoadjuvant chemotherapy in breast cancer. J Clin Oncol 2005;23:3331–3342.[Abstract/Free Full Text]
16. Paik S, Shak S, Tang G et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med

This article has been cited by other articles:

				M. Mimeault, P. P. Mehta, R. Hauke, and S. K. Batra Functions of Normal and Malignant Prostatic Stem/Progenitor Cells in Tissue Regeneration and Cancer Progression and Novel Targeting Therapies Endocr. Rev., April 1, 2008; 29(2): 234 - 252. [Abstract] [Full Text] [PDF]

				L. Pusztai Current Status of Prognostic Profiling in Breast Cancer Oncologist, April 1, 2008; 13(4): 350 - 360. [Abstract] [Full Text] [PDF]

				M. Mimeault, S. L. Johansson, G. Vankatraman, E. Moore, J.-P. Henichart, P. Depreux, M.-F. Lin, and S. K. Batra Combined targeting of epidermal growth factor receptor and hedgehog signaling by gefitinib and cyclopamine cooperatively improves the cytotoxic effects of docetaxel on metastatic prostate cancer cells Mol. Cancer Ther., March 1, 2007; 6(3): 967 - 978. [Abstract] [Full Text] [PDF]

				N. Wolmark Provocative Optimism in the Treatment of Early Stage Disease Oncologist, October 1, 2005; 10(suppl_2): 1 - 3. [Full Text] [PDF]

This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by van de Vijver, M. Search for Related Content PubMed PubMed Citation Articles by van de Vijver, M.