ORIGINAL ARTICLE Gene Expression Profiling for Breast Cancer Prognosis in Chinese Populations Bing Sun, MD,* ,1 Feng Zhang, PhD,  ,1 Shi-kai Wu, MD,* Xiaohong Guo, M.M.,   Li-li Zhang, PhD,   Ze-fei Jiang, MD,* Duen-mei Wang, PhD,   and San-tai Song, MD* *Department of Breast Cancer, Hospital Affiliated Academy of Military Medical Science, Beijing 100071, China; and   Key Laboratory of Genome Sciences and Information, Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing 100029, China n Abstract: To investigate a quantitative reverse transcription polymerase chain reaction (QRT-PCR) assay different from 21-gene assay which can be used to prognosticate the risk of recurrence in patients with estrogen receptor (ER) posi- tive, lymph node (LN) negative breast cancer. To accurately determine the relationship between the Recurrence Score (RS) derived from our assay and the risk of distant recurrence in Chinese patients with LN negative and positive breast cancer through the analysis of paraffin tissues. We obtained archival paraffin-embedded tissues from patients with invasive breast cancer and varying axillary lymph node involvement. QRT-PCR reaction was performed by using the method of SYBR Green I dye with primers. Expression of the 21-genes was converted to RS by a prespecified algorithm. We then assessed the probability of the test to accurately predict distant recurrence-free survival in this retrospective cohort. Ninety-three patients were eligible based on gene expression profiles. In our population, most breast cancer patients were premenopausal (82.6%), at early stage (93.6%) and ER positive (91.4%). Median follow-up was 65.9 months. The 5-year recurrence-free survival rate for the group was 58.8%. The concordance between the reverse transcription-PCR and immunohistochemical (IHC) measurement for ER, progesterone receptor (PgR), and HER-2 determinations was high and comparable. High RS was predictive of an elevated risk of relapse (p < 0.001). In subgroups of patients, RS had significantly predictive perfor- mance both in node-negative (p = 0.009) and node-positive patients (p = 0.038). Multivariable analysis showed that nodal status, adjuvant hormonal therapy and RS were significantly related to prognosis. RS category is a better predictor than the other risk assessment criteria or clinicopatholic features, with which we can determine more accurately the risks for recurrence of various patients. We have established an easy and economical QRT-PCR assay and validated in concor- dance with IHC measurements for ER, PgR, and HER-2. RS was associated with distant recurrence among Chinese patients with hormone receptor (HR) positive breast cancer. This study may promote the use of RS estimated from the expression of the 21-gene set for prognostication and routine clinical diagnostic application in Chinese populations. n Key Words: breast cancer, gene expression, paraffin-embedded tissues, real-time reverse transcription polymerase chain reaction T he 21-gene assay is the representative result of combining high-throughput quantitative reverse transcription polymerase chain reaction (QRT-PCR), optimization of the normalization method, and the development of RS algorithm based on mRNA expres- sion levels (1). RS was a more accurate predictor of relapse than standard clinical characteristics for patients with hormone receptor (HR) positive, lymph node (LN)-negative operable breast cancer from origi- nal papers (1,2). Studies indicated that high-risk patients identified with RS had elevated risk of distant recurrence and were most likely to benefit from chemo- therapy (3). 21-gene also provided significant prognos- tic information even among patients with positive axillary lymph nodes (4,5). The neo-adjuvant studies further developed RS and suggested a correlation between high RS and response to chemotherapy (6,7). Therefore, increasing researches of 21-gene assay and a large prospective study in North America (TAILORx) will test the value of its application in clinical practice. Currently, RS assay (oncotype DX) is commercially available with Genomic Health (Redwood City, CA) (8). There have a few different characteristic of breast Address correspondence and reprint requests to: Duen-mei Wang, PhD, Key Laboratory of Genome Sciences and Information, Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing 100029, China, or e-mail: wangdm@big.ac.cn; Song San-tai, MD, Department of Breast Cancer, Hospital Affiliated Academy of Military Medical Science, Beijing 100071, China, or e-mail: songsantai2006@163.com. 1 Joint first authors. DOI: 10.1111/j.1524-4741.2010.01049.x Ó 2011 Wiley Periodicals, Inc., 1075-122X/11 The Breast Journal, Volume 17 Number 2, 2011 172–179