Clinical and molecular evaluation of anti-estrogen therapy resistance in breast cancer. Towards tailored use of anti-estrogens and targeted agents

The efficacy of anti-estrogen therapy in breast cancer is confined by the occurrence of intrinsic or acquired resistance. In cell biological assays for predictive biomarkers identification one should consider caveats like, like geno- and phenotypic drift, artefacts and the “out of tissue” context. Also many clinical studies suffer from methodological flaws. Importantly, the clinical validity (using a statistical test for interaction), clinical utility and technical validity of a potential predictive biomarker should be analyzed. Furthermore, the biomarker should be analyzed in pre- and postmenopausal patients separately. Cell biological studies of the ER and other key components of the ER signalling pathway have provided interesting leads, but very few of these have shown clinical validity to predict anti-estrogen therapy resistance. Anti-estrogen therapy resistance can be achieved through activated growth factor receptors that share common downstream kinase activation pathways; the MAPK and PI3K pathway. In postmenopausal patients, lack of clinical validity to predict tamoxifen resistance was shown for the canonical pathway drivers (PIK3CA mutations in exon 9 and 20, loss of expression of PTEN and over-expression of HER2 and/or IGF-1R). In contrast, interactions were found between tamoxifen and several downstream activated proteins in the PI3K and/or MAPK pathways, like p-p70S6K. Patients with a p-p70S6K negative tumor derived significant benefit from tamoxifen, while patients whose tumor did express p-p70S6K did not. PIK3CA mutations were not specifically enriched in lobular breast cancer, but associated with favorable prognostic factors like low grade and positive progesterone receptor status. Phosphorylation of the ER can be mediated – but not exclusively – by activated kinases in the PI3K/MAPK pathways. The ultimate effect of ER phosphorylation on anti-estrogen sensitivity might be dependent on the upstream pathway. In postmenopausal patients, phosphorylation of the ER at either Serine 118 or Serine 167 was not associated with a differential benefit from tamoxifen, but an association with better prognosis was observed. Furthermore, the predictive value for different proliferation markers (including mitotic count, Ki67, Cyclin D1 expression and CCND1 and PAK1 gene amplification) was tested. High mitotic count and high CCND1 log 2 copy number ratio were associated with reduced tamoxifen benefit. CYP2C19 affects both metabolism of tamoxifen to 4-hydroxytamoxifen as well as catabolism of estrone and estradiol. Analysis of different CYP2C19 variants indicated that patients with at least one (non-functional) CYP2C19*2 allele derived significantly more benefit from tamoxifen compared to patients without a CYP2C19*2 allele. It is hypothesized that patients carrying a CYP2C19*2 allele are susceptible to tumors that are highly dependent on estrogen signaling due to a life-long higher exposure to estrogens. The clinical relevance of PI3K/AKT/mTOR pathway activation as mechanism of acquired resistance was tested in a series of breast cancer patients from whom both primary tumor tissue as well as biopsies from metastatic tumor lesions were available. Adjuvant anti-estrogen therapy was significantly associated with an increase in p-mTOR, p-p70S6K and p-4EBP1 in metastatic tumor biopsies compared to primary tumor tissue, indicating that compensatory activation of the PI3K/AKT/mTOR pathway might indeed be a clinically relevant mechanism resulting in acquired resistance.