Use of Maximum Intensity Projections (MIPs) for target outlining in 4DCT radiotherapy planning.

INTRODUCTION: Four-dimensional computed tomography (4DCT) is currently being introduced to radiotherapy centers worldwide, for use in radical radiotherapy planning for non-small cell lung cancer (NSCLC). A significant drawback is the time required to delineate 10 individual CT scans for each patient. Every department will hence ask the question if the single Maximum Intensity Projection (MIP) scan can be used as an alternative. Although the problems regarding the use of the MIP in node-positive disease have been discussed in the literature, a comprehensive study assessing its use has not been published. We compared an internal target volume (ITV) created using the MIP to an ITV created from the composite volume of 10 clinical target volumes (CTVs) delineated on the 10 phases of the 4DCT. METHODS: 4DCT data was collected from 14 patients with NSCLC. In each patient, the ITV was delineated on the MIP image (ITV_MIP) and a composite ITV created from the 10 CTVs delineated on each of the 10 scans in the dataset. The structures were compared by assessment of volumes of overlap and exclusion. RESULTS: There was a median of 19.0% (range, 5.5-35.4%) of the volume of ITV_10phase not enclosed by the ITV_MIP, demonstrating that the use of the MIP could result in under-treatment of disease. In contrast only a very small amount of the ITV_MIP was not enclosed by the ITV_10phase (median of 2.3%, range, 0.4-9.8%), indicating the ITV_10phase covers almost all of the tumor tissue as identified by MIP. Although there were only two Stage I patients, both demonstrated very similar ITV_10phase and ITV_MIP volumes. These findings suggest that Stage I NSCLC tumors could be outlined on the MIP alone. In Stage II and III tumors the ITV_10phase would be more reliable. CONCLUSIONS: To prevent under-treatment of disease, the MIP image can only be used for delineation in Stage I tumors.

Clinical implications of the implementation of advanced treatment planning algorithms for thoracic treatments.

BACKGROUND AND PURPOSE: Radiotherapy treatment planning algorithms continue to develop and current planning systems typically offer simpler, but faster, algorithms, which may be 2, 2.5 or 3D in modelling scatter, but which do not model electron transport (type a) and more accurate algorithms which aim to be fully 3D, i.e. which model 3D scatter and also model electron transport (type b). A range of comparative planning studies and experiments indicate that the main situation where the changes are significant between the two types of algorithm is where lung tissue is involved. However, more generally, interface areas between materials of different electron density and composition are expected to show differences between the two types of algorithms. These are likely to pose potentially significant clinical consequences when a centre changes from using older simpler algorithms to more accurate fully 3D ones and require careful consideration. MATERIALS AND METHODS: Some modelling is presented using the different type algorithms for a recently available novel design of linear accelerator treatment head, as part of the commissioning of that machine and in preparing for a change in TPS algorithm. The TPS data are compared to measurements and to Monte Carlo calculations. RESULTS AND DISCUSSION: The results add to the evidence of other studies that 3D planning techniques and type b dose calculation algorithms lead to systematic changes in computation and delivery of radiotherapy dose and in dose distributions, as compared to simpler methods, and that these changes are more pronounced in treatments involving lung tissue. The type b algorithms agree well with Monte Carlo modelling. CONCLUSIONS: Careful analysis of the changes is required before adopting new algorithms into clinical treatment planning practice. Discussion is needed between physicists and oncologists to fully understand the effects and potential consequences. These include changes in delivered dose to the reference point, to coverage of the PTV and to the dose distribution and also to dosimetric parameters used to constrain toxicity for lung, e.g. V20, and other tissues. There are consequences for assessment of dose-effect relationships and of parameters used in treatment planning decisions and this is an opportune time to re-evaluate this information.