P1-07-21: Analysis of Molecular Markers by Immunohistochemistry (IHC) Method on Formalin Fixed Paraffin Embedded (FFPE) Tissues Could Predict Shorter Recurrence Free Survival (RFS) and Overall Survival (OS) among Patients Who Have Received Adjuvant Chemotherapy for Early Breast Cancer.

Abstract

Abstract Background: Various molecular markers assessed by IHC (ER, PR, HER2) and gene expression profiling (e.g. Oncotype Dx) have been developed as prognostic and predictive tools for breast cancer. Gene profiling is said to be superior to IHC but at a considerable cost with limited availability. IHC is relatively inexpensive and more readily available. If early breast cancer patients who are going to relapse within 5 years of curative surgery despite adjuvant chemotherapy could be identified by IHC on FFPE tissue alternative adjuvant therapies could be explored. In this context, here we evaluate IHC for expression of a panel of molecular markers implicated in: growth signalling pathways (ER, PR, HER2, EGFR, CD71, Ki67, MCM2), cell survival (Bcl-2, Bag 1), angiogenesis (PDGFRa) and cell cycle progression (Aurora A, MCM2). Of note, this study includes markers of breast cancer molecular subtype (ER, PR, HER2, Ki67, EGFR, also CK5/6) and several proteins encoded by genes in the Oncotype Dx test (ER, PR, HER2, Ki 67, Bcl2, Bag1 and CD68). Materials and Method: 72 cases (R) relapsing within 5 years of curative surgery, 72 controls (C), relapse free > 5 years were identified from the hospital records. All patients had adjuvant chemotherapy. Controls were matched to cases by Adjuvant! recurrence risk (ARR). Optimised IHC was performed on FFPE TMA slides using a Ventana autostainer. Protein expression was evaluated on digitalised images (Mirax scanner). Survival analysis by molecular markers expression and also 5 molecular subtypes, Luminal A (LA = ER/PR+, HER2−, Ki67-), Luminal B (LB = ER/PR+, HER2/Ki67+), HER2 enriched (H = ER-, PR-, HER2+), Core Basal (CB = ER-, PR-, HER2−, CK5/6/EGFR+) and 5-negative (5N = negative for ER,PR,HER2,EGFR,CK5/6)], were performed. SPSS 16v. was used for statistical analysis. Findings: All but four cases had died at the time of analysis. Four controls developed relapse at 83.8, 90.6, 107.7, 127.6 months respectively. Two controls died from non-breast cancer causes. Median (m) follow-up for the controls group (ie. mOS)was 104.9 mo (72.8 - 164.4). For cases, mRFS and mOS were 23.2 (4.5 - 59.9) and 39.7(8.1 - 129). mRFS and mOS for IHC molecular subtypes were: Subtype LA = not yet & not yet; LB = 58.1 & 86.1; CB = 15.4 & 30.4; H = 28 & 55.9; 5N = 19.9 & 26 (p < 0.0001 & <0.0001 by Log rank test). Better RFS and OS were found for positive Bcl2 (p = 0.036 & 0.058) and MCM2 (p = 0.022 & 0.048), negative Aurora A (p = 0.01 & 0.001) and PDGFRa (p = 0.07 & 0.086) expressions. For this study cohort there was no correlation between ARR and survival outcome or molecular subtypes. Result of ongoing multivariate analysis and correlation between survival and CD68, CD71 and Bag 1 expressions will be presented in the conference. Discussion: Subtypes CB & 5N, negative Bcl-2 & MCM2, positive Aurora A & PDGFRa expression as measured by IHC were predictive of poor RFS and OS. While these findings need to be verified in an independent cohort, IHC profiles nevertheless have potential to stratify different risk groups for clinical trials and effective adjuvant treatments. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-07-21.