As survival improves with advances in cancer care, cardiovascular (CV) complications associated with treatment have become more prevalent. Effects of traditional chemotherapeutics are generally well known, but incorporation of small molecule inhibitors and immunotherapeutics has led to the emergence of new and unexpected toxicities. The mechanistic drivers underlying many of these have not been well characterized, undermining both appropriate monitoring and effective intervention. This is further complicated by reliance upon models of CV toxicity that do not fully recapitulate the complicated landscape of human cancer. While in vitro studies permit dissection of cellular and molecular alterations in response to drug exposure, they lack context of the whole organism that contributes to pathogenesis. Rodent models have been instrumental in defining fundamental characteristics of treatment induced CV complications, but, there are significant differences in duration of exposure to therapeutics and an absence of co-morbidities that likely influence outcome. Moreover, their small size and short lifespan limit instrumentation, longitudinal analysis, and repeated sampling. Pet dogs with spontaneous cancer are routinely treated with anti-cancer agents known to produce CV toxicity including doxorubicin, tyrosine kinase inhibitors, and more recently immune checkpoint inhibitors and may thus provide an opportunity for mechanistic interrogation in a more clinically relevant context to bridge the gap from cells and mice to humans. Their larger size and longer lifespan permit the use of prospective study designs in the setting of standard cancer treatment that more closely represent the human experience, thereby overcoming some limitations of rodent models. As such, the fundamental premise underlying this proposal is that no single model system of cancer treatment-induced CV toxicity is sufficient to effectively interrogate mechanistic drivers and assess approaches to therapeutic intervention. Instead, a coordinated, integrated effort across the landscape of multiple in vitro and in vivo model systems is required to efficiently identify and validate biomarkers for early intervention, evaluate novel treatments to address complications, and ultimately develop algorithms for predicting potential CV toxicity in the setting of combination therapy. We therefore propose that inclusion of data generated from dogs with spontaneous cancer treated with agents known to induce CV toxicity will permit a more accurate characterization and confirmation of key mechanistic drivers and therapeutic intervention strategies critical for advancing human outcomes. To accomplish this, we created a non-reductionist, multi-species framework for analyzing data generated in the laboratory, mouse models, dogs with spontaneous cancer, and human patients. The studies in this proposal will credential and optimize this novel platform using two established yet unique CV toxicities that constrain...