The changing landscape of cardio-oncology research: usual suspects, novel exposures, and the dawn of artificial intelligence

Abstract

The role of a cardio-oncologist has expanded significantly due to the broadened definition of cardiovascular toxicity, now encompassing a wide range of cardiovascular issues beyond the traditional focus on left ventricular dysfunction and heart failure. This complexity is further compounded by the continuous discovery of new associations between modern cancer therapies and cardiovascular complications. While ideally, all cancer patients would be thoroughly screened and monitored for cardiovascular toxicity, this approach is impractical, resource-intensive, and often lacks strong evidence-based guidelines. Therefore, identifying patients at increased risk for cardiovascular toxicity is essential to optimize monitoring and maximize patient benefit without unnecessary strain. The aim of this thesis is twofold. First, it seeks to identify novel predictors for the spectrum of cardiovascular toxicity resulting from both conventional and modern anticancer therapeutics. Second, it aims to explore innovative methods for harnessing the vast amount of data within the electronic health records of cancer patients, utilizing natural language processing, artificial intelligence, and automations.