Precision oncology involves tailoring diagnostics and selecting treatments based on the unique molecular profile and driver alterations of an individual patient’s tumor; in other words, delivering the right treatment at the right time for every patient. A plethora of antitumor agents and companion diagnostic assays have been evaluated in clinical trials to best match patients, leading to unprecedented extended survival and improved quality of life compared to non-matched antitumor therapies. Genomically-driven precision treatment plans are now being integrated into mainstream clinical oncology and are revolutionizing treatment and healthcare management for cancer patients. For example, next-generation sequencing, which allows for the rapid and accurate sequencing of many genes at once, has become the cornerstone of precision oncology. However, due to the increasing number of cancer diagnoses and the significant amount of data generated during cancer treatment, data-driven identification of disease states and treatment options remains a challenge. A recent editorial piece written by Pedro J. Ballester and Javier Carmona and published in Nature (2021) highlighted the fact that with advancements in artificial intelligence (AI), particularly machine learning (ML), the data limitation aspects of precision oncology could be transformed. Approaches such as using ML models to predict driver mutations based on histopathological images, for example, could be incorporated into clinical pathways to improve histological subtyping. Although still in its infancy, AI ML has the potential to identify the most effective therapeutic approach and maximize therapeutic efficacy for each cancer patient, whether through drug repurposing, identifying synergistic drug combinations or optimizing dosing schedules. The application of AI to precision oncology will inevitably encounter many challenges. Ballester and Carmona emphasized the need for “best practice” guidelines to ensure AI methods are developed and integrated in a manner that maximizes the benefit for all patients. They noted that it would become increasingly important to define and test the interfaces that enable human-computer collaboration. Furthermore, to ensure the successful deployment of AI tools in routine practice, educating users on the working principle and degree of interpretability will also be paramount.
In this issue of healthbook TIMES Oncology Hematology, Sara de Dosso from the Oncology Institute of Southern Switzerland (IOSI) and her colleagues discuss the evolving systemic treatment for advanced hepatocellular carcinoma (HCC). In particular, they describe several potential precision biomarkers that are currently being evaluated to guide the clinical development of immune checkpoint inhibitors in HCC. Next, Alexander Siebenhüner from the Cantonal Hospital Schaffhausen in Switzerland provides an overview of the evolving immune-chemotherapeutic paradigms in esophageal and gastric cancers. He emphasizes the fact that although targeted immune therapy combinations in the field of HER2-positive adenocarcinomas look promising, better identification and classification of predefined biomarkers remains an unmet clinical need. With an increased emergence of novel combinations for the treatment of various cancers in recent years, it will be interesting to see how the clinical utility of different biomarkers, together with precision oncology and AI ML, will evolve towards personalized treatment of patients with HCC and adenocarcinomas, among other cancer types. In a third article, Evelyne Bischof from the Shanghai University of Medicine and Health Sciences, China, and colleagues provide insight into the added value of performing comprehensive geriatric assessments to manage elderly and old cancer patients. Since there is exponential growth in both the size and the proportion of older persons in the population worldwide, it will be fascinating to see in the future how precision oncology will extend beyond tumor-specific markers to incorporate host factors that are evaluated as part of a routine geriatric assessment to significantly improve outcomes for older patients.
We are confident you will enjoy the latest content of healthbook TIMES Oncology Hematology. Please do not hesitate to give us your feedback on this issue and upcoming issues that will most certainly contain information about various advances in precision oncology and AI technologies.
Dr Ellen Heitlinger