Radiation oncology: what can we achieve by optimized dose delivery?

Spectacular technical advances have marked the last twenty years in radiation oncology. This revolution began with CT-based planning which was followed by 3D conformal therapy. The latter approach produced two important capabilities. The most obvious was that tumors could be viewed in their true location with respect to normal tissues and treated with beams that were not in the axial plane. A second equally important advance was the development of 3D planning tools such as dose volume histograms. These tools permitted quantitative comparison of treatment plans and have supported the development of models relating normal tissue irradiation to the risk of complication. The '3D hypothesis' - that 3D treatment planning would permit higher doses of radiation to be safely delivered-has been proven. Dose escalation studies have been successfully conducted in the lung (= 100 Gy), liver (= 90 Gy), brain (= 90 Gy), and prostate (= 78 Gy). Prospective phase II and phase III trials suggest improved outcome using these higher doses for tumors in the liver and prostate compared to doses considered acceptable in the 2D era. The next technical revolution is underway, with advances in '4D' radiotherapy (accounting fully for organ motion) and in intensity-modulated radiation therapy (IMRT) to further improve the conformality and accuracy of treatment. Proton therapy will improve dose distributions still further. These improved dose distributions can be combined with more accurate tumor delineation provided by functional imaging to offer the potential for additional dose escalation without toxicity and for improved tumor control. These developments permit us to ask if we are approaching the limits of dose optimization and how (if?) research in radiation delivery should proceed