Trial Information

800 breast cancer patients who fulfill the eligibiltiy criteria will be selected from the
NUH breast cancer registry (400 for training, 400 for validation). Eligible patients should
have at least 5 years' follow-up at NUH and have an available archival paraffin-embedded
tumor block stored at the Department of Pathology, NUH. 8-10 ten-micron sections from each
tumor block will be cut and RNA extracted from the sections. RNA will then be profiled using
high-throughput gene expression platforms (Illumina Veracode assay, Affymetrix Quantigene
assay). Raw gene expression data will be applied in the following signatures previously
generated by the Genome Institute of Singapore:

1. 32-gene p53 Pathway Signature

2. 5-gene Genetic Grade Signature

3. 33-gene TuM1 Signature

4. 10-gene ER Signature (Miller & Tan, et. al., unpublished)

5. 6-gene 3-Ratio Predictor (Miller & Karuturi, et. al., unpublished)

6. 5-gene HER2 Amplicon Predictor (Miller & Karuturi, et. al., unpublished)

7. 7-gene Basal-Luminal Discriminator (Miller, unpublished)

The prognostic and/or predictive abilities of these signatures will be compared with
conventional clinical prognosticators and predictors with the goal of developing archival
tumor-derived genomic tests for breast cancer management in the future.

The archival paraffin-embedded tumor blocks are left-over samples after clinical use. They
are not samples that have been consented for research purpose. We are requesting for waiver
of consent as this is a minimal risk study. We will ensure that sufficient tissue be left
behind for future routine diagnostic purposes. Sections taken from each tumor will be coded
with no patient identifiers to protect the privacy and confidentiality of the participants.

The results generated from the tumor samples in this study will not impact on the clinical
management of the patients.