This Article Full Text (PDF) 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 Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Brenton, J. D. Articles by Caldas, C. Search for Related Content PubMed Articles by Brenton, J. D. Articles by Caldas, C. Related Articles Related Correspondence Social Bookmarking                            What's this?

In Reply:

Cancer Genomics Program, Department of Oncology, University of Cambridge, Hutchison/Medical Research Council Research Centre, Cambridge, United Kingdom

Dr Loi et al raise two scientific questions in their letter commenting on our recent review article. Firstly, they are correct in stating that the molecular taxonomy of breast cancer is still a work in progress. It is however reassuring that novel molecular subtypes such as basal and luminal reflect what we as clinicians already know to be true—hormone receptors, HER2, and only a few additional markers identify tumors that differ in behavior. We disagree with their comments that describe the results of unsupervised molecular classifications of breast cancer as unstable. These classifications have been consistent across two external datasets and across platforms, and have similar differences in outcomes.¹ In addition, the examples cited by Dr Loi et al to show inconsistencies in the classifications are explained by noting that the luminal B and luminal C categories were clustered together and considered a single group in Sorlie et al 2003¹ article. The clustering of the luminal B group with the estrogen-receptor negative samples is not evidence of a biologic association as the correlation values of the upper dendrogram branches were negative. It is much more relevant to note the consistent finding of groups that reflect basal-like, HER2+/estrogen-receptor negative and luminal tumors. However as we stated in our review, "unsupervised methods do not derive statistically valid information... [and] are poorly suited for identifying prognostic variables." The important point we are making is that it is now difficult to argue for the use of unselected breast cancer cases to derive or to test novel prognostic signatures when clear distinctions such as HER2 classification have important prognostic and therapeutic implications. It is now becoming clear that even the recent Early Breast Cancer Trialists’ Collaborative Group overview² which has not considered HER2 status might not be able to inform future management.³

Secondly, there are statistical issues concerning bias,⁴ the uniqueness of the signature,⁵ and overfitting⁶ in the Amsterdam^7,8 signature. This should not be interpreted to mean that the Amsterdam signature does not contain genes important in prognosis—rather that there are many prognostic signatures in the data set, and it remains unclear which are the best. Further validation to identify the most important genes is urgently needed. The Program for Assessment of Clinical Cancer Tests has produced draft guidelines for marker evaluation and this specifically highlights the importance of assay standardization and clinical utility assessment.⁹ Dr Loi et al state that level 3 evidence is now available for the Amsterdam signature and highlight work published as a meeting abstract¹⁰ that tested the signature in 301 available samples from six centers and compared hazard ratios (HR) for time to distant metastasis and overall survival against the Nottingham prognostic index and the St Gallen criteria. The differences between the three prognostic tests are modest and the HRs for the Amsterdam signature are smaller than previously published (HR 1.85; 95% CI 1.14 to 3.0 v 5.1; 95% CI 2.9 to 9.0).⁸ We also note that the Nottingham prognostic index performed poorly as a prognostic tool in this cohort of cases (time to distant metastasis HR 1.46; CI 0.87 to 2.47)¹⁰ raising questions about the representativeness of the samples. It therefore remains unclear whether significant additional prognostic information was provided by the Amsterdam signature and a fuller assessment of these issues will have to await the peer-reviewed publication. Dr Loi et al also cite the recent article by Espinosa et al¹¹ as further supporting the prognostic value of the Amsterdam signature. However, this work was performed with only 60 genes from the original 70-gene signature, using a different expression assay and prediction rule and the 60-gene signature was not prognostic for node-negative breast cancers.

Authors' Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.

REFERENCES

1. Sorlie T, Tibshirani R, Parker J, et al: Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A 100:8418-8423, 2003[Abstract/Free Full Text]

2. Early Breast Cancer Trialists' Collaborative Group: Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomised trials. Lancet 365:1687-1717, 2005[CrossRef][Medline]

3. Burstein HJ, Winer EP: HER2 or not HER2: That is the question. J Clin Oncol 23:3656-3659, 2005[Free Full Text]

4. Ransohoff DF: Gene-expression signatures in breast cancer. N Engl J Med 348:1715-1717, 2003[Free Full Text]

5. 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]

6. Simon R, Radmacher MD, Dobbin K, et al: Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification. J Natl Cancer Inst 95:14-18, 2003[Free Full Text]

7. van 't Veer LJ, Dai H, van de Vijver MJ, et al: Expression profiling predicts outcome in breast cancer. Breast Cancer Res 5:57-58, 2003[Medline]

8. 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]

9. National Cancer Institute: Guidelines for Marker Development. http://www.cancerdiagnosis.nci.nih.gov/assessment/progress/markerdev.html

10. Piccart MJ, Loi S, Van't Veer L, et al: Multi-center external validation study of the Amsterdam 70-gene prognostic signature in node negative untreated breast cancer: Are the results still outperforming the clinical-pathological criteria? San Antonio Breast Cancer Symposium, San Antonio, TX, December 8-11, 2004

11. Espinosa E, Vara JA, Redondo A, et al: Breast Cancer Prognosis Determined by Gene Expression Profiling: A Quantitative Reverse Transcriptase Polymerase Chain Reaction Study. J Clin Oncol 23:7278-7285, 2005[Abstract/Free Full Text]

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

Related Correspondence

• Molecular Forecasting of Breast Cancer: Time to Move Forward With Clinical Testing
  Sherene Loi, Christos Sotiriou, Marc Buyse, Emiel Rutgers, Laura Van't Veer, Martine Piccart, and Fatima Cardoso
  JCO 2006 24: 721-722 [Full Text]

This Article Full Text (PDF) 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 Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Brenton, J. D. Articles by Caldas, C. Search for Related Content PubMed Articles by Brenton, J. D. Articles by Caldas, C. Related Articles Related Correspondence Social Bookmarking                            What's this?