Customization of tamoxifen therapy : not only a dream

Abstract

Breast cancer (BC) is the most common form of cancer in western women. The grand majority of the afflicted women are eligible for endocrine treatment. Tamoxifen has been the golden standard treatment for more than three decades. Unfortunately, resistance towards this drug is a major concern in the clinic. The aim of this thesis is to increase our knowledge of the resistance mechanisms by merging information from cell model systems and human tumors.

One of the main inducers of angiogenesis, vascular endothelial growth factor (VEGF), is overexpressed in many types of cancers and has also been associated to poor outcome in estrogen receptor (ER) positive BC. When analyzing VEGF in 404 ER positive tumors from patients treated with tamoxifen as the sole adjuvant treatment, we found a significant negative effect of high VEGF on survival. Interestingly, the effect was overcome by a prolonged tamoxifen regimen. Moreover, studies on MCF7 cells and its 4-hydroxytamoxifen (4-OHT)-resistant counterpart, MCF/LCC2 (LCC2) revealed the existence of an autocrine signaling loop involving VEGF, its receptor VEGFR2 and p38 in LCC2 cells.

As the resistance mechanism likely involves alterations in multiple parts of the signaling machinery we used a high throughput methodology to study this further. We performed quantitative mass spectrometry (MS)-based proteomics on these cell lines to explore key proteomic differences between them and in response to 4-OHT treatment. Pathway analysis on significantly deregulated proteins revealed a connection to the retinoic acid receptor alpha (RARA), a receptor in close interaction with ER signaling. The 4-OHT-resistant cells exhibited attenuated anti-growth response in response to RARA activation in contrast to MCF7 cells. LCC2 cells were also dependent on RARA for maintenance of viability. Primary validation of the impact of RARA on survival, in the previously described cohort, revealed a correlation between high RARA expression and reduced relapse-free survival specifically in those treated with two years of tamoxifen. Moreover, VEGF was highly correlated to RARA expression. We conclude that RARA may be of potential prognostic or predictive value for patients eligible for tamoxifen therapy.

We used a similar MS-based proteomics approach in order to discover potential biomarkers predictive for tamoxifen response by comparing patients with early relapse (<2years) versus non-relapse (>7years). This pilot study of 24 patients yielded in the identification of 3101 proteins out of which 13 were classified as characteristic for tamoxifen-resistant tumors. CAPS and MX1 are part of this 13-protein signature and have shown to be overexpressed in patients with early relapses.

Altogether, our findings suggest that it is possible to apply information from cell lines to find patient-based associations with potential clinical utility. The response to tamoxifen may be dependent on the involvement of VEGF and RARA. The predictive value of CAPS and MX1 remains to be elucidated. The validation of these factors in larger patient populations is expected in the near future.