Precision Cardio-Oncology: A Multimodal Approach to Cancer Therapy-Related Cardiovascular Toxicity Prediction and Assessment

Abstract

Cancer therapy-related cardiovascular toxicity (CTR-CVT) is associated with morbidity, mortality, and interruptions to cancer treatment. Current cardio-oncology clinical assessment tools generally rely on a limited range of clinical data; however, in the era of precision medicine, it may be possible to leverage multimodal data, including imaging and multi-omic inputs, to personalise CTR-CVT prediction and assessment. In this review, we examine the genotypic and phenotypic predictors of CTR-CVT and the evidence supporting their incorporation into multimodal risk models. Most research focuses on cancer therapy-related cardiac dysfunction (CTRCD). We identify a strong evidence base for multimodal imaging-guided prediction, particularly cardiovascular magnetic resonance (CMR) and transthoracic echocardiography (TTE)-derived global longitudinal strain measurements. Emerging evidence also supports the potential value of novel imaging parameters, including computed tomography (CT) and TTE-derived radiomic features. From the genomic and proteomic perspective, truncating titin variants and microRNA-126 may be informative predictors of cancer therapy-related cardiac dysfunction (CTRCD). At the same time, the literature also highlights potential epigenomic markers of CTR-CVT. Finally, we consider how these diverse data can be integrated into multimodal risk models and how digital innovation and advanced machine learning methods may further improve model predictive performance. However, a large implementation gap remains, with challenges such as data privacy and a lack of external validation studies preventing routine clinical implementation.