Abstract 2709: Mathematical modeling for prediction of secondary BRCA1 and BRCA2 mutations in ovarian cancers with deleterious germline BRCA1 and BRCA2 mutations

Abstract

Abstract Background: DNA damaging agents such as cisplatin or carboplatin and poly(ADP-ribose) polymerase inhibitors (PARPi) have clinical activity in germline BRCA1 and BRCA2 mutation-associated ovarian cancers. However, BRCA1 or BRCA2-mutated tumors often develop resistance to these drugs. Restoration of BRCA function due to secondary BRCA mutations has been recognized as one of the clinical resistance mechanisms. The aim of this mathematical analysis is to study the emergence and frequency of secondary BRCA mutations in a cohort of BRCA-deficient in silico ovarian cancer-positive cases. Methods: We develop a 2D stochastic cell-cycle specific mathematical model of BRCA-deficient ovarian cancer growth and progression prior to and post-treatment, which includes debulking surgery, platinum-based chemotherapy or PARPi. This is a continuum model based on a combination of ordinary differential equations using Gompertzian growth kinetics subject to random discrete-time jump processes. This approach enables us to reflect the inter-individual heterogeneity in disease progression and drug administration reported in published data. Calculations are performed for in silico cancer-positive growth curves in which the inception of the first ovarian malignant cell is assumed to have already occurred. Results: Our modeling approach generates an empirical cumulative distribution of disease-free survival times prior to recurrence or disease progression, given the heterogeneous cancer progression illustrated by our in silico cohort. Our preliminary computational results indicate that an initially rare resistant cell clone carrying a secondary BRCA mutation pre-exists to treatment in circa 90% of the simulated progression curves. For the remaining 10% of the simulated curves, it is more likely that the secondary mutations are acquired during chemotherapy or PARPi via increased mutation rates, possibly conferred by the increasing amount of damaged DNA lesions caused by the cancer treatment. Conclusions: Our mathematical model provides some insight on the dynamics of secondary BRCA mutations that cannot be accessed through existing genomic analysis methods, as current DNA sequencing depth thresholds cannot detect the presence of rare tumor cell clones carrying such mutations. We predict that a large proportion of initially resistant cell clones with restored BRCA function more likely originate prior to rather than post-treatment. If such clones get selected for by subsequent platinum-based chemotherapy or PARPi in a Darwinian fashion, our model generates an empirical distribution of disease-free survival times prior to recurrence. This may have direct implications on the possible duration of subsequent treatment response in our simulated ovarian cancer-positive population. Citation Format: Dana-Adriana Botesteanu, Doron Levy, Jung-Min Lee. Mathematical modeling for prediction of secondary BRCA1 and BRCA2 mutations in ovarian cancers with deleterious germline BRCA1 and BRCA2 mutations. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2709.