Functional precision oncology is the practice of assessing the phenotype of biopsied tumor cells upon perturbation, e.g. treatment with candidate therapies, to yield actionable information fast enough to influence clinical decision making. For a functional precision approach to provide actionable information on tumors’ response to candidate therapeutics, it must: retain specimen heterogeneity, monitor all biologically important phenotypes, make longitudinal observations, and - since clonal expansion of individual cells is sufficient to drive tumor progression or resistance – have single cell resolution. One example where functional precision oncology could be valuable is the choice of therapy for the approximately 50% of melanoma patients harboring the BRAFV600E mutation. Such patients have two options – immune checkpoint inhibition (ICI) or targeted therapy (TT). Either strategy is capable of curing patients in many cases but neither option works for all patients. Poor response to either is caused by the pre-existence and/or emergence of phenotypes resistant to each therapeutic option. A clinical test that could predict, on a personalized level, which patients are likely to respond or acquire resistance to either of these therapies is among the most pressing clinical needs in melanoma care. On a population level, more patients present with durable response to ICI than to TT, and so ICI is the default standard of care and a predictive test must achieve high accuracy to influence clinical decision making. We have pioneered the use of quantitative phase imaging (QPI) for rapid and label-free phenotype assessment of melanoma cells, including monitoring for therapeutic resistance. Our area under the receiver operator characteristic curve (AUC) for predicting resistance under 48 h is 0.84-0.90 – which is promising, but insufficient and needs validation. We propose to construct a new technological and analytical platform with two modifications. First, we will augment QPI with a second imaging module to measure light scatter via a new method we have developed based on darkfield microscopy. Light scatter is traditionally measured using flow cytometry and is predictive of relevant cell phenotypes in a myriad of cancer types, including, our preliminary data show, therapeutic resistance in melanoma. Second, we will establish an analytical pipeline for assessing cell state dynamics which we anticipate will yield a classifier that is more accurate across heterogeneous biopsies as compared to current approaches. In this proposal, we describe a series of engineering and analytical steps, coupled with technical milestones and target quantitative goals benchmarked against existing approaches for developing an approach we call Quantitative Imaging of Cell state Kinetics (QuICK), as well as a proof of principle study using clinical biopsies. If successful, we will have built a prototype platform with high potential to improve the care of melanoma patients through accurate personalized matching to the best therapy and we will be well situated for prospective clinical trials. In addition, QuICK has the potential to inform functional precision medicine approaches for other cancers that would benefit from rapid classification of live cell phenotypes with single cell resolution.
Functional precision oncology has the potential to revolutionize treatment options for cancer patients by providing much needed new approaches to identify the correct therapy for individual patients. Most patients with melanoma, a deadly skin cancer, are treated with one of two very different therapeutic options – either of which cure many patients, but neither of which work on all patients, making melanoma an ideal model system in which to develop and demonstrate functional approaches. We propose to engineer a prototype technology that analyzes small biopsies of living tumors to predict which melanoma patients will respond to which therapy, to help address the burden of melanoma and to provide a novel, functional approach that can be extended to other cancers in the future.