A reduced qRT-PCR-based gene expression signature with prognostic value in early breast cancer

Abstract #2030

Background: Gene profiling may improve prognostic accuracy in patients with early breast cancer, but this technology is not widely available. We used commercial assays for qRT-PCR to measure the expression of 83 genes previously found to determine breast cancer prognosis. The objective of our study was to develop a simple molecular signature predicting relapse in early breast cancer with positive hormonal receptors.
 Methods: 153 patients (age 29-82 years-old) with early breast cancer and a minimum follow-up of 5 years were included. All tumours were positive for hormonal receptors and 38% had positive lymph nodes; 64% of patients received adjuvant chemotherapy (with anthracycline if N+ or high-risk N0). RNA was isolated from tissue slices of formalin-fixed paraffin-embedded samples using the MasterPure RNA Purification Kit (Epicentre Biotech). qRT-PCR amplifications were performed with TaqMan Gene Expression Assays products in an ABI PRISM 7900 HT Sequence Detection System (Applied Biosystems). The reactions were carried out using the TaqMan Low Density Arrays (Applied Biosystems). A supervised analysis was done to identify a prognostic expression signature using BRB Array Tools software. Survival curves were derived from Kaplan-Meier estimates and compared by log-rank test. The association of gene expression and clinical variables with distant metastasis-free survival (DMFS) was analyzed by Cox regression models. SAS 9.1 software package was used for statistical analyses.
 Results: After a median follow-up of 92 months, 28% of patients relapsed. We defined an 8-gene prognostic signature. The distant metastasis-free survival at 5 years was calculated for this profile. DMFS was 97% -good risk- versus 60% -poor risk, HR 20.3 (95% CI 6.17-67.2), p<0.001. This gene expression profile was highly informative in identifying patients with distant metastasis, even when corrected for traditional prognostic factors in Cox multivariate analysis (HR 20, 95% CI 5.9-67.4). Performance was similar to that of three validated gene profiles. The validity of this signature was verified in an independent cohort obtained at the GEO database.
 Conclusions: This study identifies a new molecular signature with prognostic utility that outperforms traditional histopathologic prognostic factors. Our 8-gene qRT-PCR assay is suitable for analyzing routine formalin-fixed paraffin-embedded samples in the clinical setting.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 2030.