Dissecting the predictive value of MAPK/AKT/estrogen-receptor phosphorylation axis in primary breast cancer to treatment response for tamoxifen over exemestane: a Translational Report of the Intergroup Exemestane Study (IES)—PathIES

Abstract

Purpose The prognostic and predictive values of the MAPK/AKT/ERα phosphorylation axis (pT202/T204MAPK, pT308AKT, pS473AKT, pS118ERα and pS167ERα) in primary tumours were assessed to determine whether these markers can differentiate between patient responses for switching adjuvant endocrine therapy after 2–3 years from tamoxifen to exemestane and continued tamoxifen monotherapy in the Intergroup Exemestane Study (IES). Methods Of the 4724 patients in IES, 1506 were managed in a subset of centres ( N = 89) participating in PathIES. These centres recruited 1282 (85%, 1282/1506) women into PathIES of whom 1036 had phospho-marker data. All phospho-markers were analysed by immunohistochemistry staining. Multivariable Cox proportional hazards models of the phospho-markers for disease-free survival (DFS) and overall survival (OS) were adjusted for clinicopathological factors. Treatment effects on the biomarker expression were determined by interaction tests. Benjamini–Hochberg adjustment for multiple testing with a false discovery rate of 10% was applied ( p BH ). Results Phospho-T202/T204MAPK, pS118ERα and pS167ERα were all found to be correlated ( p BH = 0.0002). These markers were not associated with either DFS or OS when controlling for the established clinicopathological factors. Interaction terms between the phospho-markers and treatment strategies for either DFS or OS were not statistically significant ( p BH > 0.05 for all). Conclusions This PathIES study confirmed previously described associations between the phosphorylation site markers of AKT, MAPK and ERα activity in postmenopausal breast cancer patients. No prognostic correlations between the phosphorylation markers and clinical outcome were found, nor were they predictive for clinical outcomes among patients who switched therapy over those treated with tamoxifen alone. Electronic supplementary material The online version of this article (10.1007/s10549-018-05110-x) contains supplementary material, which is available to authorized users.

A scoping review of statistical methods in studies of biomarker-related treatment heterogeneity for breast cancer

“…While 8 of the 31 articles reported results for only one analysis with regard to a predictive biomarker, many studies performed multiple evaluations, mostly for several biomarkers, but also for several outcomes and subpopulations (Additional File 2: Supplementary Table 3 ). Five studies performed 10 or more analyses on the same patients [ 21 , 30 , 35 , 40 , 41 ]. The maximum number of analyses performed was 20 in a study with 18 different biomarkers [ 21 ].…”

“…The maximum number of analyses performed was 20 in a study with 18 different biomarkers [ 21 ]. Only 3 of the studies with more than one analysis adjusted their results for multiple testing [ 40 , 41 , 46 ]. The adjustment in all three studies was done by controlling the family-wise type 1 error rate via a Benjamini-Hochberg correction.…”

“…Half of the reviewed studies analyzed more than one biomarker and/or endpoint, but only 3 studies corrected for multiple testing [ 40 , 41 , 46 ] in order to reduce the probability of false positive findings. This is important since the chance of false positive conclusions is already increased for underpowered studies [ 59 ].…”

New insights into acquired endocrine resistance of breast cancer

“…In another aspect, ER α can be phosphorylated by activated MAPK and Akt, resulting in ligand-independent transcriptional activity [36,37] . Particularly, aberrant activation of PI3K/Akt by PI3KCA mutations has been implicated in endocrine resistance of breast cancer [38,39] . Additionally, more investigations disclose that activity of non-receptor tyrosine kinase c-Src is increased [40,41] and plays a key role in the mediation of interaction between ERα and growth factor receptor in endocrine resistant breast cancer cells [30] .…”