Abstract The polyploid giant cancer cell (PGCC) state is a common response of cancer cells to various stressors including chemotherapy, irradiation, hypoxia and viral infection. Upon stress, PGCCs adopt an endoreplication in which the genome replicates, mitosis is omitted, and cells grow in size, causing drug resistance. How accessible endoreplication is to a cell, therefore, directly translates to its resistance to therapy. In this study, we hypothesized that endoreplication and PGCC state are more accessible to cancer cells that already have a higher ploidy content prior to therapy. To test this hypothesis, we developed a comprehensive three-tier framework consisting of i) computational, ii) in-vitro and iii) in silico components. First, we designed a set of ordinary differential equations (ODEs) to mathematically model how cells enter and exit the PGCC state in response to a given stressor. We engineered how this in silico approach interacts with in vitro experiments into a broadly applicable software solution called CLONEID. The software uses computer vision to monitor phenotypic changes in cell and nuclear size from standard bright-field microscopy and classify cells into PGCC and non-PGCC states. We used CLONEID to test various therapeutic agents for their ability to select for a stable near-tetraploid (4N) population in a set of near-diploid (2N) cell lines. Through spontaneous cell fusions, we also obtained tetraploid breast cancer cells matched with their parental lines. Altogether, this framework enabled us to i) monitor the PGCC state experimentally in ii) two sets of matched isogenic cell lines with 2N and 4N DNA content to iii) model the successful entry and exit rates to and from the endoreplication state, respectively. As the first application of this framework, we tested the ability of our 2N and 4N TNBC lines (SUM159 and MDA-MB-231) to access the PGCC state upon treatment with 18 chemotherapy agents. Among those drugs, we observed that only gemcitabine caused continued cell growth without cell division in both tetraploid SUM159 and MDA-MB-231 cells whereas near-diploid parental lines were hypersensitive to the treatment. Consequently, tetraploid cancer cells continued to safely grow in the presence of gemcitabine. Furthermore, these PGCCs re-entered the proliferative cell cycle and grew in cell number when treatment is terminated. Gemcitabine-based chemotherapy is a standard treatment for patients with TNBC although its efficacy is limited mainly due to drug resistance. We expect our findings and three-component framework strategy to help stratify the TNBC patient population by their response to gemcitabine. In addition, our mathematical modelling approach has the promising potential to inform personalized dose optimization and to effectively decrease administered gemcitabine dose for a subset of patients, which would alleviate severe therapy-associated side effects and co-morbidities. Citation Format: Vural Tagal, Jackson Cole, Daria Miroshnychenko, Andriy Marusyk, Noemi Andor. Ploidy as a predictive biomarker for gemcitabine sensitivity in triple-negative breast cancers [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr A024.
Read full abstract