Agentic Artificial Intelligence for Precision Oncology: Autonomous Clinical Decision Systems from Diagnosis to Therapy

Abstract

Cancer remains one of the leading causes of morbidity and mortality worldwide despite substantial advances in molecular diagnostics, targeted therapeutics, immunotherapy, and precision medicine. The extraordinary biological complexity of malignant diseases, together with rapidly expanding volumes of clinical, radiological, pathological, genomic, and molecular data, has created increasing demand for intelligent computational systems capable of supporting individualized clinical decision-making. Recent developments in agentic artificial intelligence (AI) have introduced a new generation of autonomous clinical systems that extend beyond conventional predictive models by continuously perceiving, reasoning, planning, acting, and learning from dynamic healthcare environments. Unlike traditional AI systems that perform isolated prediction tasks, agentic AI integrates multimodal biomedical information, electronic health records, medical imaging, genomic sequencing, digital pathology, laboratory biomarkers, wearable technologies, and clinical guidelines to autonomously coordinate diagnostic reasoning, therapeutic planning, longitudinal monitoring, and adaptive decision support under physician supervision. Advances in large language models, foundation models, multimodal learning, graph neural networks, reinforcement learning, retrieval-augmented generation, and autonomous AI agents have accelerated the development of intelligent oncology ecosystems capable of supporting cancer diagnosis, molecular characterization, treatment optimization, immunotherapy prediction, adaptive radiation therapy, clinical trial matching, survivorship management, and precision therapeutics. Despite these advances, important challenges remain regarding explainability, algorithmic reliability, interoperability, ethical governance, cybersecurity, regulatory approval, accountability, and safe clinical implementation. This review provides a comprehensive overview of agentic artificial intelligence in precision oncology, highlighting computational foundations, current clinical applications, emerging innovations, and future perspectives for autonomous clinical decision systems across the cancer care continuum.[1]