Integrated Informatic and Experimental Evaluations of Cancer Chronotherapy

点击翻译按钮获取中文摘要

ABSTRACT Years of clinical experience and a growing body of basic research suggest that chemotherapeutic activity can change with time-of-day. But when should our patients take their medicines? Must we test each new agent for circadian modulation in both efficacy and toxicity? Which tumors are most sensitive to chemotherapy administration time? Can we tailor our recommendations for individual patients? Temporal variation in the abundance of drug targets, transporters, and metabolizing enzymes, in both tumors and normal tissues, underlies circadian variation in drug activity. Until recently almost all we knew about tissue specific circadian rhythms came from normal mice. Without human data, a mechanistic, hypothesis-driven transition to medical practice has been slow. Recently we developed CYCLOPS (CYCLic Ordering by Periodic Structure) a machine-learning algorithm to uncover human transcriptional oscillations using existing, unordered biopsy samples. We used CYCLOPS to explore circadian rhythms in human lung and liver, identify disrupted rhythms in hepatocellular carcinoma, and predict circadian changes in drug effectiveness. This proposal will greatly expand that work and accelerate its translation to clinical oncology. Using public data, we will describe the molecular rhythms in an array of normal human tissues and thus the times of day when these tissues are least sensitive to specific drug toxicities. We will also describe rhythms in select tumors, identifying circadian times and cell cycle phases when cancers are most distinct from surrounding tissue and thus uniquely sensitive to various treatments. We will explore the influence of specific mutations and tumor markers on the rhythms observed in patients. Mapping these data onto pharmacogenomics databases we can make testable prediction as to the drugs and side effects most influenced by circadian time. Finally using both experimental mouse data and our preliminary human results, we have compiled a list of some of the most promising chronotheraputic candidates. We will expand and refine this list over the course of the study, testing several of these predictions in established animal models, and exploring the promise and practical principles of cancer chronotherapy. Taken together these aims will help catalyze chronotheraputic translation to clinical oncology and help delineate the role of time in precision cancer therapy.