Tumor Heterogeneity

Resistance to treatment in cancer evolution is caused by tumor heterogeneity, making gaining greater comprehension of underlying dynamics which drive cancer cell variation fundamental to development of new more efficient therapies. Treatment resistance can be caused by growth of preexisting subclonal populations or evolution of cells which are resistant to drugs.

Heterogeneous tumors are classed as spatial and temporal types, dependent on whether nonuniform distribution of cancer cells is dispersed across and within disease sites; or whether there is cell variation over time.

Tumor heterogeneity is currently analyzed by examining bulk specimens, but it is limited due to admixture of diverse cancer cell types and nonmalignant cells. Advent of single cell sequencing provides ability to characterized individual cells within a diverse population to define complex clonal relationships.

Cancer is by far a stagnant disease and genomic instability with tumor cells provides genetic diversity underpinning tumor heterogeneity. Instability ranges from single based substitutions to doubling of whole genomes caused by exposure to mutagens including medications, UV radiation, or faults in internal regulation of process such as DNA repair and replication.

Chemotherapy may also create genomic instability by increasing mutational spectrum of tumors. Studies have found that some cancers can integrate endogenous homostatic processes to increase overall burden of mutation, indicating tumorigenesis is linked to higher spontaneous mutation rates.

Genomic instability within brain tumors are produced from chromosomal changes where the whole genome segment is removed or doubled due to segregation of errors that occur during cell division.

Clonal evolution selection framework was developed to explain how diversity is maintained and describes two patterns of evolution: Linear tumor evolution can have successive acquisition of mutations which may provide a survival advantage or growth promotion. Genetic instability produces a new clone type with advantage that is outcompeted by the next subclones. Cooperation between subpopulations is required for tumor propagation where there is nonautonomous initiating events. Second pattern that can emerge is branching evolution when multiple genetically distinct populations form from common ancestral clones, which produces an environment that is more likely to create a heterogeneous tumor. Linear evolution is most common in haematological malignancies, while branched often forms solid tumors.

Increased levels of tumor heterogeneity can cause decreased effectiveness of therapies, utilizing mathematical modeling can help to combat this and help to design combinational anticancer treatments. Simulations can help to determine optimal dosage schedule for withdrawal of targeted therapies to prevent drug resistant cells and provide time for vulnerable cells to repopulate.

Studies have shown that optimal drug combination for tumor cell death with minimal outgrowth of clonal subpopulations can be defined with RNA inference models. Short hairpin RNA knockdowns used as model of individual loss of function events with combined subpopulations reflect heterogeneous tumor populations. Models then use data sets of known hairpin RNA knockdown responses to agents, which simulations can then supply optimal drug combinations for treatment of heterogeneous tumor populations.