Evolutionary Double-Bind Treatment Using Radiation Therapy and Natural Killer Cell-Based Immunotherapy in Prostate Cancer.

PURPOSE: Evolution-informed therapies exploit evolutionary consequences of drug resistance to inhibit treatment resistance and prolong time to progression. One strategy, termed an evolutionary double-bind, uses an initial therapy to elicit a specific adaptive response by cancer cells, which is then selectively targeted by a follow-on therapy. Although the concept of an evolutionary double-bind has long been hypothesized in cancer, it has not been measured. Here, to our knowledge, we present the first example of a quantifiable double-bind: radiation therapy (RT) with natural killer (NK) cells. RT induces lethal double-strand DNA breaks, but cancer cells adapt. Although this increases resistance to DNA-damaging agents, it also enhances expression of NK cell ligands creating an obvious choice for a double-bind strategy.

METHODS AND MATERIALS: We investigated this potential evolutionary double-bind through in vitro studies and evolution-based mathematical models. Using multiple prostate cancer cell lines, we evaluated surface and soluble NK ligand expression following RT. In vitro competition experiments were performed with an isogenic radiation-resistant cell line model. We introduced a two-population Lotka-Volterra competition model, consisting of radiation-sensitive and radiation-resistant populations modeling intrinsic growth rates with fixed carrying capacity and inter-specific competition terms.

RESULTS: Alterations in NK cell ligands resulted in a twofold increase in sensitivity to NK cell-mediated killing with selective targeting of RT-resistant cells. These dynamics were framed mathematically to quantify the double bind. RT alone slowed overall growth but strongly selected for RT-resistant cells. NK cell therapy alone suppressed the RT-resistant population, but with a surviving population of radiation-sensitive cells. Model simulation predicted that optimal tumor control would be achieved through initial RT followed by NK cells. Subsequent experiments confirmed the model prediction.

CONCLUSIONS: We conclude that RT and NK cell-based immunotherapy produce an evolutionary double-bind. This multidimensional approach addresses the immediate challenge of treatment resistance and lays the groundwork for the development of personalized treatment regimens tailored to the evolving dynamics of individual tumors.

SIGNIFICANCE: Clinical experience demonstrates that prostate cancer has a remarkable capacity to evolve resistance to all currently available treatments resulting in progression and, ultimately, patient death. Resistance mechanisms often come at a fitness cost placing cells in a bind when competing with surrounding cells. A carefully chosen secondary drug can introduce a double-bind targeting the adaptive resistance mechanism. This manuscript provides the first direct experimental evidence quantifying an "evolutionary double-bind' in prostate cancer supporting the combination of DNA-damaging agents and NK cell-based immunotherapy in evolutionarily guided treatment designs. Our work is mathematically novel in that it extends Evolutionary Game Theory models and establishes an experimental-mathematical framework to quantify genuine evolutionary double binds applicable across cancer types and treatment modalities.