Real-Time Dose-Guided Radiation Therapy

Abstract

Dramatic strides have been made in real-time adaptive radiation therapy, where treating single tumors as dynamic but rigid bodies has demonstrated a halving of toxicities for prostate cancer. However, the human body is much more complex than a rigid body. This review explores the ongoing development and future potential of dose-guided radiation therapy, where the three core process steps of volumetric imaging of the patient, dose accumulation, and dose-guided treatment adaptation occur quasi-continuously during treatment, fully accounting for the complexity of the dynamic human body. The clinical evidence supporting real-time adaptive radiation therapy was reviewed. The foundational studies, status, and potential of real-time volumetric imaging using both x-ray and magnetic resonance imaging technology were described. The development of real-time dose accumulation to the dynamic patient was evaluated, and a method to measure real-time dose delivery was assessed. The growth of real-time treatment adaptation was examined. Literature demonstrates continued improvements in patient outcomes because the treatment becomes more conformal to the dynamic patient. Real-time volumetric imaging using both x-ray and magnetic resonance imaging technology is poised for broader implementation. Real-time dose accumulation has demonstrated clinical feasibility, with approximations made to achieve real-time operation. Real-time treatment adaptation to deforming targets and multiple targets has been experimentally demonstrated. Tying together the inputs of the real-time volumetric anatomy and dose accumulation is real-time treatment adaptation that uses the available degrees of freedom to optimize the dose delivered to the patient, maximizing the treatment intent. Opportunities exist for artificial intelligence to accelerate the application of dose-guided radiation therapy to broader patient use. In summary, the emerging field of real-time dose-guided radiation therapy has the potential to significantly improve patient outcomes. The advances are primarily software-driven and therefore could be widely available and cost-effective upgrades to improve imaging and targeting cancer.