8. Estimation and Presentation of Uncertainty in the Delivered Dose

The practice for most patients in radiotherapy is still to generate a treatment plan, which is consequently used over the entire course of therapy, typically for 25–30 fractions. This leads to a number of systematic uncertainties associated with the generation of a treatment plan (imaging for treatment planning, contouring, dose calculation). Even as this practice gradually changes with the introduction of adaptive concepts, the mentioned systematic uncertainties apply to the number of fractions for which the plan is used. A second category of uncertainties is of random nature and can be separated into interfraction and intrafraction uncertainties. Most of these random uncertainties will show a tendency to average out over the number of fractions. All uncertainties taken together will inevitably lead to deviations between the delivered dose and the prescribed absorbed dose. For example, a random setup uncertainty will lead to a smearing out of the delivered dose but not to a shift of the dose distribution. A systematic error in the definition of the target point, however, will lead to a systematic shift of the delivered dose distribution in all subsequent fractions if not corrected. In ion-beam therapy uncertainties play a more important role for two reasons. First, a lower number of fractions is delivered, which may lead to a larger impact of random uncertainties. Second, the calculation of RBE-weighted dose, which is widely used, necessitates the use of biological models with inherently larger uncertainties in the input parameters. Consequently, the uncertainties involved in ion-beam therapy will be discussed below in more detail. Due to the nature of the underlying biological processes and their variations, the response of an organ or tumor to a radiation dose is not completely predictable. Only from a group of patients, meaningful parameters, that affect the outcome, can be derived statistically. In addition to these inherently probabilistic effects, there are also uncertainties connected to the underlying physical and clinical parameters on which the biological effects of radiation are dependent. To derive meaningful clinical results from the application of ion beams, these parameters should be specified, controlled, and if possible, kept to a reasonable accuracy and uncertainty level. Some of these parameters can, in principle, be determined by measurements while others are more difficult to determine precisely. Among the latter, there are all the treatment parameters, which are defined through the experience and expertise of the radiation oncologist. These are: