Discordance between PAM50 intrinsic subtyping and immunohistochemistry in South African women with breast cancer

Abstract

PurposeBreast cancer is a heterogeneous disease with different gene expression profiles, treatment options and outcomes. In South Africa, tumors are classified using immunohistochemistry. In high-income countries multiparameter genomic assays are being utilized with implications for tumor classification and treatment.MethodsIn a cohort of 378 breast cancer patients from the SABCHO study, we investigated the concordance between tumor samples classified by IHC and the PAM50 gene assay.ResultsIHC classified patients as ER-positive (77.5%), PR-positive (70.6%), and HER2-positive (32.3%). These results, together with Ki67, were used as surrogates for intrinsic subtyping, and showed 6.9% IHC-A-clinical, 72.7% IHC-B-clinical, 5.3% IHC-HER2-clinical and 15.1% triple negative cancer (TNC). Typing using the PAM50 gave 19.3% luminal-A, 32.5% luminal-B, 23.5% HER2-enriched and 24.6% basal-like. The basal-like and TNC had the highest concordance, while the luminal-A and IHC-A group had the lowest concordance. By altering the cutoff for Ki67, and realigning the HER2/ER/PR-positive patients to IHC-HER2, we improved concordance with the intrinsic subtypes.ConclusionWe suggest that the Ki67 be changed to a cutoff of 20–25% in our population to better reflect the luminal subtype classifications. This change would inform treatment options for breast cancer patients in settings where genomic assays are unaffordable.