|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 27 (September 20), 2007: pp. 4317-4318 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.12.8694
In ReplyJules Bordet Institute, Université Libre de Buxelles, Brussels, Belgium; and the Peter MacCallum Cancer Center, Melbourne, Australia
Jules Bordet Institute, and the Machine Learning Group, Université Libre de Bruxelles, Brussels, Belgium
Jules Bordet Institute, Université Libre de Buxelles, Brussels, Belgium We thank Debled and colleagues for their thoughtful comments. However, we would like to highlight that our article1 is not proposing the gene expression grade index (GGI) as a new prognostic tool, rather that the heterogeneity of clinical outcome that is associated with estrogen receptor–positive breast cancers is related to the extent of expression of proliferation and cell cycle genes (ie, the proliferation status of a breast cancer is far more important in determining prognosis than the quantitative expression of hormone receptors). Using the GGI, we concluded that proliferation status can easily and accurately identify two estrogen receptor–positive subgroups resembling the previously described luminal A and B classification proposed by Perou and colleagues2 and importantly provides an easy to understand biologic basis for this distinction. Debled et al note that it is not surprising that the GGI has such strong prognostic value as it consists of genes associated with cell cycle and progression. This is to be expected as the GGI was developed to be an objective quantification of histologic grade.3 While we agree that demonstration of a prognostic ability of a gene signature in uni- and multivariate analysis with other traditional clinicopathologic factors is less adequate than comparing its performance with classifications such as the Nottingham Prognostic Index (NPI), St Gallen, or Adjuvant! Online, we wish to emphasize that the GGI was developed to enhance the prognostic ability of histologic grade. However, the GGI is significantly correlated with a 70-gene signature4 and a 76-gene signature5 both of which have been extensively validated against these tools.6,7 In addition, replacing histologic grade with the GGI in our previously reported data set3 significantly improved the prognostic performance of the NPI (data not shown). Ivshina et al8 have reported similar results, suggesting that the use of genomic grade can improve significantly the prognostication of breast cancer patients. Debled et al are interested in the prognostic ability of the GGI and progesterone receptor (PgR) status. The authors used PgR levels determined by immunohistochemistry to better characterize grade 2 tumors, resulting in a high risk subgroup representing only 14% of their data set compared with 44% in our data set based on GGI levels. Unfortunately, we do not have PgR levels determined by immunohistochemistry nor the full components of histologic grade (nuclear grade, tubule formation, and mitotic rate) or other proliferation factors to make a direct comparison between the genomic and clinicopathologic assessment. However, for those patients with estrogen receptor–positive breast cancers who had not received systemic adjuvant treatment and classified as histological grade 2 (n = 111), those with poor PgR expression and high GGI have a worse outcome than those with rich PgR expression and low GGI, as expected (hazard ratio, 5.43; 95% CI, 2.19 to 15.12; P = .0038), but not rich PgR and high GGI levels (P = .1). This highlights the additional and important prognostic information provided by GGI. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. REFERENCES
1. Loi S, Haibe-Kains B, Desmedt C, et al: Definition of clinically distinct molecular subtypes in estrogen receptor-positive breast carcinomas through genomic grade. J Clin Oncol 25:1239-1246, 2007 2. Perou CM, Sorlie T, Elsen MB, et al: Molecular portraits of human breast tumours. Nature 406:747-752, 2000[CrossRef][Medline] 3. Sotiriou C, Wirapati P, Loi S, et al: Gene expression profiling in breast cancer: Understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst 98:262-272, 2006 4. van 't Veer LJ, Dai H, van de Vrijver MJ, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415:530-536, 2002[CrossRef][Medline] 5. Wang Y, Klijn JG, Zhang Y, et al: Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet 365:671-679, 2005[Medline] 6. Buyse M, Loi S, van't Veer L, et al: Validation and clinical utility of a 70-gene prognostic signature for women with node-negative breast cancer. J Natl Cancer Inst 98:1183-1192, 2006 7. Desmedt C, Piette F, Loi S, et al: Strong time dependence of the 76-gene prognostic signature for node-negative breast cancer patients in the TRANSBIG multicenter independent validation series. Clin Cancer Res 13:3207-3214, 2007 8. Ivshina AV, George J, Senko O, et al: Genetic reclassification of histologic grade delineates new clinical subtypes of breast cancer. Cancer Res 66:10292-10301, 2006
Related Correspondence
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
