Authors:
Marc Hillairet de Boisferon, Ismahene Benzaid, Elodie Marie Dit Chatel, Nicolas Hoffmann, and Bruce A Littlefield
Abstract:
Worldwide, numerous preclinical agents that enter oncology clinical trials fail to demonstrate sufficient efficacy in patients to gain regulatory approval. This failure rate reflects both a poor understanding of the complexity of human cancers as well as the weak predictive value of existing preclinical models. PDXs are now widely embraced as better preserving characteristics of the original tumor and thus providing better tools for translational research. However, tumor collection for PDX generation usually coincides with initial surgery or biopsy, meaning that PDXs typically reflect the state of patient tumors at the times of collection. Unfortunately, this may not reflect changes that occur after treatment, particularly those associated with acquisition of resistance. Although analysis of changes associated with acquisition of resistance would be of great value for translational purposes, opportunities to obtain paired initial primary and later resistant samples are limited in current practice.
In this study we describe establishment of a luminal breast cancer PDX model with acquired resistance to hormonotherapy. Starting from a luminal A breast cancer PDX that was both estrogen-dependent and fulvestrant-sensitive, we derived delayed outgrowth tumors under in vivo selection conditions of estrogen deprivation and/or fulvestrant treatment, thus corresponding to acquisition of resistance following hormonotherapy in the clinical setting. To understand phenotypic changes associated with resistance acquisition, resistant PDXs were analyzed for estrogen and progesterone receptor expression and subjected to whole exome sequencing, RNA sequencing, and DNA methylation analysis comparatively with the parental PDX. Intriguingly, all derived resistant PDXs showed complete loss of PR expression by immunohistochemistry regardless of selection method, whereas ER expression was only decreased in resistant PDXs that had been selected by fulvestrant treatment. Genetic analyses showed that resistant PDX tumors fell into two different transcriptomic signatures, depending on whether resistance was driven by estrogen deprivation or fulvestrant administration. Compared to the naïve parental tumor, the cluster corresponding to fulvestrant-selected PDX showed up-regulated and down-regulated pathways associated with genomic alterations related to endocrine therapy and KRAS pathways. Interestingly, the other cluster in which PDX were selected without fulvestrant showed only a partial genomic alterations. Furthermore, whole exome sequencing of the PDX revealed similar driver genes mutations existed in both derived resistant and sensitive parental PDX. Finally, we confirmed pharmacological resistance to fulvestrant in the resistant PDX. New experiments investigating drug combinations in the context of endocrine resistance are ongoing.