ESTRO 33, 2014 interfraction geometrical uncertainty of a new patient. The prediction power of the model was patient specific. Future work will investigate the use of this model for the development of automated segmentation techniques and for high precision RT. 1 Budiarto et al, PMB, 2011, 56:1045-1061
POSTER DISCUSSION: YOUNG SCIENTISTS 5: RTT PD-0310 Can teaching patients with pelvic tumours about correct positioning improve treatment set-up in radiotherapy? H. Hansen1, A. Boejen1, M. Bjørklund Ellegaard1, A. Vestergaard2 1 Aarhus University Hospital, Department of Oncology, Aarhus C, Denmark 2 Aarhus University Hospital, Department of Medical Physics, Aarhus C, Denmark Purpose/Objective: The use of daily online IGRT has revealed deviations in roll, pitch and yaw from the planning CT position. In clinical practise we experience that some patients are good at positioning and some have difficulties. Teaching patients about positioning before start of treatment and staff use of a common language when assisting patients to position themselves might assist patients to adopt a more accurate position and thus reduce the number of repositionings. The purpose of this randomised trial was to test whether teaching patients about positioning would A) improve the precision of the position B) reduce the number of repositionings. Materials and Methods: The randomised study included 25 female pelvic cancer patients (6 patients with anal cancer, 4 patients with rectal cancer and 15 patients with gynaecological cancer) treated with RT of more than 10 fractions immobilised in supine position using a heel and knee fixation system. In the intervention group (12 patients) treatment and positioning before initiating treatment were visually explained using a 3D virtual reality environment (VERT 2.9, VERTUAL). Two practical exercises about movement of the pelvis and the lumbar were also introduced. The staff used specific and common language when assisting the patients concerning positioning before treatment. The control group (13 patients) received the standard treatment. Daily shifts in target at the most cranial, caudal and lateral borders were calculated using a simple triangular calculation after a rigid registration between the first CBCT of each treatment and planning CT allowing for all rotations. Target shifts above 5 mm were considered equal to a required repositioning of the patient. An unpaired t-test was used to compare rotational errors in the two groups. Results: Target displacements of more than 5 mm were found in 11 fractions in the intervention group, and in 10 fractions in the control group. Offline CBCT image registration resulted in pitch (mean: 0.23°; SD: 1.36°), roll (mean: -0.07°; SD: 0.78°) and yaw (mean: -0.17°; SD: 0.66°) for the intervention group and pitch (mean: 0.21°; SD: 1.32°), roll (mean: -0.25°; SD: 0.81°) and yaw (mean -0.30°; SD: 0.77°) for the control group. A significant difference was found between the control and intervention group when looking at roll (p = 0.005) and yaw (p = 0.021) but we found no significant difference when looking at pitch (p = 0.834). Conclusions: The results showed a tendency towards patients being better positioned in the intervention group compared to the control group when looking at roll and yaw. There were an almost equal number of target displacements above 5mm in the two groups. PD-0311 Reproducibility of bladder volume in prostate RapidArc radiotherapy treatment J. Bento1, R. Silva1, A. Espadinha1, F. Banha1, D. Silva1, I. Lobato1, A. Videira1, M.I. Antunes1, P. Chinita1 1 Hospital Espírito Santo, Radiotherapy, Évora, Portugal Purpose/Objective: There is evidence that dose escalation improves local control in prostate tumors, thereby increasing toxicity risk in normal tissues, such as the bladder. In order to minimize this risk, filling bladder protocols (consisting of water ingestion) can be adopted to increase its volume, resulting in a smaller irradiated bladder volume, also keeping the small intestine way from the irradiation field.Thus, the reproducibility of bladder's volume during the treatment period is crucial. In this work, a comparative study was implemented in order to compare the effectiveness of three different filling bladder protocols, assessing the respective variations in bladder filling during the treatment period.
Materials and Methods: The bladder volume was measured in 60 patients undergoing prostate RapidArc radiotherapy treatment (38-40 fractions). Before performing the planning CT scan and daily treatment, the patients were instructed to fill the bladder according to one of three protocols: protocol 1 (drink 500 ml, wait 30 minutes), protocol 2 (drink 667 ml, wait 30 minutes) or protocol 3 (drink 667 ml, wait 40 minutes). The bladder was contoured and its volume was measured in the planning CT scan and in seven Cone Beam CT (CBCT) scans weekly acquired during treatment. The relative deviation between bladder volume at the CT planning scan and the CBCT scans and their mean values were determined. Pearson correlation test was applied to evaluate the correlation between bladder volume at CT planning scan and its variation during treatment. Results: There was a decrease between mean bladder volume at the planning CT (199.5±112.4 ml) and at treatment period (150.5±69.7 ml). The relative deviation between planning CT and weekly CBCT scans suggests that protocol 1 is more reproducible; the mean value for squared relative deviation (0.35 for protocol 1, 0.53 for protocol 2 and 0.45 for protocol 3) also indicates that protocol 1 has lowest bladder volume variation during treatment. There is a weak correlation (r=0.47) between the bladder volume in the planning CT and its variation during treatment, indicating that patients with higher bladder volumes at the planning CT tend to have greater bladder filling variation. Conclusions: Generally, the reproducibility of bladder filling during treatment is low, since the estimated mean bladder volume is smaller when compared with the value at the planning CT scan. Still, protocol 1 seems to be more reproducible; as this protocol involves less water amount and less waiting time, these results are in agreement with the correlation test that indicates the higher bladder volume at the planning CT, the larger variations during treatment. PD-0312 Flattening Filter Free for head and neck and prostate irradiation L.W. Kool1, A. Bruggeman1, F. Van Hofwegen1, M. Jeulink1, W. Verbakel1 1 VUMC, Radiation Oncology, Amsterdam, The Netherlands Purpose/Objective: Earlier studies have shown benefits of using flattening filter free (FFF) irradiation for faster delivery of stereotactic body radiotherapy and reduction of peripheral doses. If FFF beams could be used for all treatment sites, in combination with an intensity modulated radiotherapy technique, flattened (FF) beams would not be needed anymore. The purpose of this planning study is to evaluate if FFF beams can be used for prostate and head and neck (h&n) radiotherapy. Materials and Methods: The CT datasets of three prostate, three larynx, three oropharynx and two nasopharynx patients were used for replanning. All treatment plans used two full arcs RapidArc (Varian Medical Systems). For each prostate patient three plans were created with FF 6MV, FFF 6MV and FFF 10MV and for each h&n patient two plans were created (FF 6MV, FFF 6MV). All plans per patient used the same set of optimization objectives. OAR of h&n patients contained both the salivary and the swallowing structures. Planning Target Volume (PTV) coverage, dose homogeneity (standard deviation of PTV dose), mean/maximum dose in organs at risk (OAR), integral dose, peripheral dose, monitor units (MU's),length and volume of the PTV were assessed. Results: The prostate treatment plans using FFF 10MV resulted in a significantly better PTV coverage and peripheral and integral dose compared to FFF 6MV and FF 6MV, with same OAR doses. The larynx plans using FFF 6MV resulted in a slightly better PTV coverage, homogeneity, dose max to spinal cord, integral dose and peripheraldose compared to the FF 6MV. However, MU's increased by 16%. For nasopharynxand oropharynx, FF 6MV plans were better for the PTV coverage, dose homogeneity, mean/maximum dose in OAR and MU's, but not significant for all the parameters. In this study, FFF resulted in better plans for patients with a smaller PTV.
ESTRO 33, 2014 PD-0314 Online evaluation of tumor match and need for adaptation for lung cancer patients - Can radiotherapists do the job? A.B. Rasmussen1, D.S. Moeller2, M.H. Andersen1, L. Hoffmann2 1 Aarhus University Hospital, Department of Oncology, Aarhus C, Denmark 2 Aarhus University Hospital, Department of Medical Physics, Aarhus C, Denmark Purpose/Objective: Treating large tumor volumes in the lung and avoiding high risk of radiation pneumonitis demands minimal margins. This is obtained by daily soft tissue match on the tumor combined with adaptive radiotherapy (ART) for patients with systematic anatomical changes during RT. The aim was to evaluate Radiotherapists (RTT´s) assessment and evaluation of soft tissue match and the need for adaptation.
Conclusions: We expect that the length of the PTV is of influence on the suitability of FFF beams for RapidArc plans consisting of 2 arcs. FFF 10MV is a good alternative for the current FF 6MV prostate treatment planning. For head and neck patients, 2 arc FFF plans could only be used for Larynx treatments. Possibly, good FFF plans for large PTV lengths require more leaf modulation than what can be generated in plans of 2 arcs. PD-0313 A novel approach to implement dose-guided ART for HN cases treated with tomotherapy W. Fung1, G. Chiu1, A. Mui1 1 Hong Kong Sanatorium & Hospital, Department of Radiotherapy, Happy Valley, Hong Kong (SAR) China Purpose/Objective: Dose-guided (DG) ART requires systematic evaluation of the dose progress during treatment course for replan consideration, but its clinical implementation is usually hindered by the time and labor consuming process of massive data generation and by the lack of effective reporting tool and mechanism. Here we use a novel approach for implementing true DG-ART in clinical practice and investigate its effectiveness for the process. Materials and Methods: Three HN cases treated with tomotherapy completed the DG-ART process by far. An automated data processing and reporting tool, ART AssistTM (MIM Software Inc., Cleveland, OH, USA), was used to monitor patient changes during the treatment course. On each day, the system performed contour propagation to MVCTs and delivered dose accumulation using deformable registration. Daily reports were generated to display the dose and anatomic changes in terms of tables, graphs, figures and DVHs. An in-house ART protocol with patient-specific thresholds was used for assessment. Whenever the report revealed changes that exceeded tolerance for 3 consecutive days, senior dosimetrist would perform in-depth investigation e.g. checking of isodose distribution and contour changes on MVCT. Oncologist would decide whether a replan should be initiated based on the findings. Results: All 3 cases suffered from NPC and the total number of fractions was 30-35. Ninety-eight MVCTs were involved in total. Major triggers were found at 13th and 29th fractions for Case 1, 7th fraction for Case 2, and 13thfraction for Case 3, in which the observed dose and/or anatomic changes were severely deviated from the original plan that the senior dosimetrist decided to consult oncologist for replan consideration. At the time of first major trigger, Case 1 and Case 3 had their lymph node target dose decreased 8-13% and their parotid mean dose increased 1428%, while Case 1 and Case 2 had their brainstem maximum dose increased 11-18%. For all three cases, the oncologist decided not to replan but loosened the preset tolerance according to the actual reported changes and dose distribution. For Case 1 at 29th fraction, the dose changes had exceeded the newly set tolerance that a replan was originally initiated. However due to patient poor condition, replan was ceased. The actual total dose delivered throughout the whole treatment was estimated for these cases. Only Case 1 showed dissatisfactory outcome with underdose in the lymph node targets, in which further boost treatment might be needed. Conclusions: The combined use of ART Assist and in-house ART protocol allows easy and effective monitoring of treatment progress and guarantees prompt action to significant patient changes. The estimated actual total dose further helps the oncologist in deciding future treatment to patients. All these make DG-ART clinically feasible. DG-ART using this novel approach is ongoing for more HN cases and results will be presented shortly.
Materials and Methods: Soft tissue match and ART were implemented for all curatively intended lung cancer patients in April 2013 in our clinic. Before clinical start 32 RTTs underwent an education programme including e-learning and hands on training using CBCTs from 22 patients. An exam consisting of an evaluation of soft tissue match from 6 patients had to be passed. 50 consecutive patients treated with ART were included in this study. The RTT’s performed a daily online evaluation of the primary tumor volume (GTV-T), the lymph nodes (GTV-N), the spine, mediastinum, patient contour (body) and lung density changes (atelectasis, pneumonia or pleural effusion). The GTV-T should be encompassed by an evaluation structure (GTV T+2mm) generated by adding an isotropic margin of 2 mm to GTV T, hence allowing for 2 mm deviations. Lymph nodes are difficult to observe on CBCT, therefore GTV-N was evaluated using a surrogate structure called Match-N, which include surrounding normal tissue visible on CBCT (trachea, hilum, aortic arch, aortic decendens or the spine). A deviation of 5 mm was accepted for Match-N. For the spine 5/10 mm deviation was allowed depending on the treatment plan, while for mediastinum, body and pleural effusion the accepted deviations were 10, 15 and 10 mm, respectively. Changes in atelectasis or pneumonia were always scored as a deviation. 3 consecutive deviations prompted an evaluation of the demands for ART performed offline by a medical physicist. In this study, an experienced RTT retrospectively evaluated the deviations offline for 13 patients. It was noted how many times the online evaluations scored three consecutive deviations correct, incorrect (i.e. deviations below the limit) or missed. For each patient, deviations in tumor, lymph nodes and normal tissue were scored separately. Results: 78% of the RTTs immediately passed the exam. The remaining RTTs had additional hands on training and 19/3% passed a second/third test. The experienced RTT fully agreed with the evaluation made by two medical physicists on the 6 patients in the exam. The offline evaluations showed that in 20 cases correct deviations were found online, in 3 cases incorrect deviations were found and in 6 cases deviations were missed. Correct, incorrect and missed deviations were divided into causes of deviation as seen in the figure.
Conclusions: ART driven by online evaluation of daily CBCT’s is feasible and the majority of the deviations are found by the RTT’s. The deviations missed at the online evaluation are due to tumor deformation and deviations in patient positioning.