Current Clinical Practice Versus New Developments in Target Volume and Dose Specification Procedures: A Contradiction?

Modern radiotherapy requires a geometric accuracy of the order of a few millimetres and a dosimetric accuracy of the order of a few percent. The exact delineation of the target volume as well as the choice of the corresponding irradiation fields is one of the most difficult tasks in radiotherapy. Too large field sizes will cause unnecessary irradiation of healthy normal tissues, while too small fields will result in geographic misses as reflected in several studies ( I . 2). Accurate knowledge of the dose is required because many dose-effect curves for both tumour control ( 3 ) and especially normal tissue damage (4) are very steep. The need for a high dosimetric accuracy has resulted in the world wide implementation of dosimetry protocols which prescribe in detail how to assess the dose under reference conditions. More recently additional recommendations have been provided to determine the dose under non-reference conditions; i.e. at the dose specification point along the central beam axis (5 ) . The next step to facilitate communication and consistency in radiotherapy is to define unambiguous procedures for the specification of both target volumes and dose delivery. Although such a statement is obvious and for a long time now recommendations have been available for volume and dose specification (6). there is still wide diversity among the procedures applied in the various radiotherapy institutions or even within the same clinic. This has been reported after an inventory of dose specification in the Nordic radiotherapy centres (7) as well as by a careful review of publications of radiotherapy results (8). In order to use common procedures more widely, the International Commission on Radiation Units and Measurements, ICRU, updated its recommendations for target volume and dose specification given in 1978 (9). In ICRU Report 50 the previous definitions have been largely repeated, but some definitions and recommendations have been clarified or modified. For instance, the Planning Target Volume, PTV, is identical to the previous (Report 29) definition of 'Target Volume' but newly defined volumes such as the Gross Tumour Volume, GTV, and the Clinical Target Volume, CTV, have been introduced in ICRU Report 50. The report can therefore be considered as a logical extension of the earlier Report 29 and most of its recommendations are relatively easy to implement. Furthermore, these recommendations are already in use in a large number of radiotherapy institutions and have been implemented in various protocols of clinical trials (e.g. (10)). Although difficult to quantify, it can be stated that the ideas and definitions provided by ICRU Report 50 have had a great impact on the radiotherapy community and have improved the communication between the various people involved in radiation oncology considerably. Despite its widespread use, the concepts applied in ICRU Report 50 have limitations. First, according to ICRU, the dose should be specified at a (reference) point centrally located in the PTV. By applying biological models, Brahme predicted that the mean dose to the cancer cell population might be the best parameter which is correlated with tumour response (1 I ) . Although the ICRU also mentions, as additional information. the specification of the mean dose to the PTV, such a procedure is conceptually not correct because the PTV contains margins for setup uncertainties which d o not contain tumour cells. The mean dose to the PTV might therefore differ from the mean dose to the cancer cells and from the dose at the specification point, although the differences are small for a number of common treatment techniques (12). Furthermore, no distinction is made between uncertainties in the position of the CTV as a result of internal movements or inaccuracies