504 Background: Gene expression profiling has identified four main intrinsic subtypes of breast cancer which have distinct clinical behaviours and responses to therapy. A simplified immunohistochemical (IHC) panel can provide meaningful discrimination and help guide patient management. Methods: Using tissue microarrays (TMA) created from two randomized trials of adjuvant chemoendocrine therapy in node negative early breast cancer (IBCSG VIII and IX,n= 1220) we assessed intrinsic subtype as follows: “luminal A” (LA): ER and/or PR present, Ki-67 low (<median: 19%), HER2-; “Luminal B” (LB): ER and/or PR present, Ki-67 high(>=19% median) and/or HER2+ (IHC 3+ or FISH amplified); HER2 enriched: ER and PR negative, HER2 positive (IHC 3+ or FISH amplified), Triple Negative (TNP): ER, PR and HER2 negative. p53+ and high cyclin D1 expression were then assessed to determine if they provided any further additive discriminatory value. Results: The standard definitions of intrinsic subtype demonstrated a significant difference in disease-free survival between the subtypes in both trials VIII and IX (p= 0.02), but this was not significant in multivariate analysis. The discrimination between luminal A and B tumors could not be improved upon by the addition of p53 or cyclin D1 expression. The standard definitions demonstrated that TNP and HER2-enriched tumors benefited from the addition of chemotherapy to endocrine therapy, but LA and LB did not. p53+ was associated with a poorer prognosis in ER+ luminal tumors but improved prognosis in ER- non-luminal tumors (Interaction p = 0.002). Conclusions: Intrinsic subtyping using a panel of ER, PR, HER2 and Ki67 provides meaningful prognostic information to help guide patient management in early breast cancer. Whilst it predicts benefit from the addition of chemotherapy in TNP and HER2-enriched tumors it did not in this series do so for LB. Although p53 status did not improve the definition of luminal tumors, p53+ appeared to have a divergent effect on outcome dependent on ER status.
Read full abstract