Background: The negative association between the oestrogen receptor (ER) and the human epidermal growth factor receptor 2 (HER-2) in breast cancer travels in both directions. ER+ tumours are less likely HER-2+ and HER-2+ tumours are less likely ER+.
Methods: We studied the age-related immunohistochemical (IHC) expression of ER, progesterone receptor (PR) and HER-2 in 2,227 tumours using age as a continuous variable. Steroid receptors were considered positive for any nuclear staining of invasive cancer cells and for HER-2, either for strong expression by IHC (score 3+) or gene amplification by fluorescence in situ hybridisation (FISH). Based on nonparametric regression, the age-related association between steroid receptors and HER-2 was presented as likelihood curves.
Results: The association between ER or PR and HER-2 is age-related. The age-related expression of ER and PR is HER-2 dependent. In HER-2(-) cases, the odds ratio (OR) for being ER+ was 2.594 (95% CI = 1.874-3.591) up to age 50 and age-independent thereafter; for PR-expression the OR was 2.687 (95% CI = 1.780-4.057) up to age 45 and 0.847 (95% CI = 0.761-0.942) thereafter. In HER-2+ cases, the OR was 0.806 (95% CI = 0.656-0.991) to be ER+ and 0.722 (95% CI = 0.589-0.886) to be PR+. The age-related OR for breast cancers to be HER-2+ is steroid receptor dependent. Taking together, ER+PR+HER-2+ breast cancers appear on average 5.4 years earlier than breast cancers of any other ER/PR/HER-2 phenotype (95% CI = 3.3-7.5; P < 0.0001).
Conclusion: There is a qualitative interaction between age and expression of steroid and HER-2 receptors. Our findings suggest a strong age-related selective growth advantage for breast tumour cells belonging to the ER+PR+HER-2+ subgroup.