Abstract PR07: The strength of drug selection determines the maximum fitness cost of resistance mutations in culture and xenografts

Abstract

Abstract Cancer cells often acquire resistance to therapy due to genetic mutations. However, without drug selection, these mutations would likely have a fitness cost to cancer cells and we hypothesize that the severity of this fitness cost is correlated with the effectiveness of drug treatment, with more effective treatments leading to more drastic genetic lesions. The size of a genetic lesion often correlates with its fitness cost - point mutation having little cost, aneuploidy having a large cost. However, aneuploidy is fast because the search space is only 46 gains or losses, while point mutations are slow with billions of base pairs to search before finding the right mutations. We designed experimental and computational systems to understand the selection pressures that would lead to resistance by these different genetic mechanisms. Mice were implanted with xenograft tumors from the HT1080 fibrosarcoma cell line that were heterozygous for DR5 (TNFRSF10B) and extremely sensitive to treatment with a DR5 specific monoclonal antibody, which induced extrinsic apoptosis. Treatment with three doses of anti-DR5 cured 3/11 mice, while 8/11 tumors relapsed. Isolation and culture of the relapse tumors showed they were completely resistant to DR5-mediated killing and had lost the functional copy of DR5. Assaying copy number variation using next generation sequencing (CNVSeq) showed that the relapse lines had lost one copy of chromosome 8p, where the functional DR5 gene was located. Treatment with two doses of anti-DR5 lead to relapses in all mice with loss of 8p. Finally, treatment with a single dose resulted in 3/4 tumors with loss of 8p. However 1/4 tumors became more resistant to anti-DR5 but did not lose the DR5 gene, presumably becoming resistant by smaller genetic changes or by non-genetic means that altered DR5 signaling. Thus, in most cases, treatment of xenograft tumors led to resistance by the loss of a chromosome arm, while resistance was rarely induced by other mechanisms. In contrast, cultured HT1080 cells treated with three doses of anti-DR5 became resistant but never lost 8p. In culture, treatment with anti-DR5 only kills a maximum of 80% of the cells and was therefore weaker than in mice where anti-DR5 could be curative. Thus, we observed an association between the strength of drug selection and the genetic mechanism by which cells became resistant. Experiments to further decrease selective pressure in xenografts and increase it in culture are ongoing. Next, we developed a computational model of tumor evolution to further explore the interplay between strength of selection and type of adaptation. In this simulation, cells were under two selective pressures, drug selection and competition for nutrients. Drug sensitive tumor cells could become resistant by two mechanisms analogous to point mutations or loss of 8p. Point mutations in this model caused resistance incrementally thereby taking several generations to acquire resistance, but they had no fitness cost in terms of competing for nutrients. Loss of 8p in this model caused complete resistance but had a fitness cost. Simulations with strong selective pressure led to the emergence of tumors that had completely lost 8p, even when the fitness cost was high. This result occurred because cells that did not lose 8p were eliminated by the drug. In contrast, simulations with selective pressure similar to culture (80% kill) led to tumors that became resistant without loss of 8p. In this case, sensitive cells incrementally acquired resistance mutations and then out competed cells that lost 8p for nutrients. Thus, simulation showed that resistance mutations with a large fitness cost will only arise under strong drug selection. In conclusion, our data suggest that the closer a therapy is to curing a tumor, the more extreme and costly resistance mutations will be. Citation Format: Lee A. Albacker, Rainer Kohler, Ralph Weissleder, Peter K. Sorger. The strength of drug selection determines the maximum fitness cost of resistance mutations in culture and xenografts. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Drug Sensitivity and Resistance: Improving Cancer Therapy; Jun 18-21, 2014; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(4 Suppl): Abstract nr PR07.