Abstract Book

S1244

ESTRO 37

Results The overall mean and standard deviation of the cumulative IFM (translations and rotations) are summarized in Table 1. Figure 1 shows the correlation between the session time and the IFM vector. The mean (M) session time was 27.1 min (range 19.8 - 41.3). The correlation between the session time and IFM vector is not significant, p=0.11. For 98% of the patients the IFM vector is ≤ 1.5mm. Two outliers of 3.0 mm and 2.5 mm were found.

rotational setup errors; (2) Simulation of residual rotational errors. Volume of PTVs receiving the prescription dose (V100), Coverage index (CI) and Homogeneity index (HI) and Maximum dose (D max ) were the parameters to evaluate the impact of setup errors. Friedman test was used and statistical significance was considered when p < 0.05. Results 3mm errors in translational and 3 o error in rotational positioning errors in any direction could result in V100 loss to <95% and D max > 5400cGy and 4500cGy for brainstem and spinal cord respectively. SI shifts and pitch rotations caused larger deviation of dose than other axis of shifts and rotations for PTVs; AP shifts and pitch rotations caused larger deviation of dose than other axis of shifts and rotations for OARs. Translational errors caused higher numbers of PTVs losing coverage to less than 95% for V100 than that with rotational errors. Pilot data showed that with translational correction, V100 of all PTVs remained greater than 98% and showed slightly increase in D max for brainstem and spinal cord. Conclusion For the best dosimetric results, with consideration of balancing treatment efficiency in busy clinical settings, NPC treatment using VMAT should have setup translational errors of <2mm and rotational errors of <2 o in all directions. The setup errors tolerance should be evaluated for individual cases if critical structures are in close proximity to the targets. EP-2377 Image guidance using paired, non-cardinal kV imaging for whole breast RT A. Leong 1 , C. Cowles 1 , J. Ransley 1 , R. Louwe 2 1 Wellington Blood & Cancer Centre, Radiation Therapy, Wellington, New Zealand 2 Wellington Blood & Cancer Centre, Medical Physics, Wellington, New Zealand Purpose or Objective There is no clear consensus on the best approach for image-guidance for whole breast radiation therapy (WBRT). Further challenges arise for WBRT involving regional lymph nodes (WBRT-LN) due to the independent movement of anatomical regions. Transitioning from reliance on clinical set up using skin marks and offline imaging, our department implemented an online imaging protocol that utilises paired, non-cardinal kV images. This study reports on set up accuracy and residual errors obtained with this protocol. Material and Methods A customised online imaging protocol was applied for two sequential cohorts of patients during WBRT (n=10) and WBRT-LN (n=12). Briefly, a tangential kV imaging field was paired with an orthogonal anterior-oblique field. Matching was guided by patient-specific match regions highlighted in the DRR that defined the ribs immediately underlying the treated volume. For WBRT-LN treatments, additional imaging included an AP or PA kV field prior to treatment of the nodal volume. Analysis included assessment of initial set up error, the position of the anterior breast/chest wall derived from tangential kV images, and the residual error of surgical clips. For WBRT-LN, the residual error of the nodal treatment region was determined using AP or PA kV images, as well as rotation of the chest wall and position of the humeral head. Group mean (M), systematic (Σ) and random (σ) errors were calculated for all measurements.

Figure 1. Correlation between session time and cumulative IFM vector Conclusion 98% of the patients had an IFM vector length smaller than 1.5 mm (Mean 0.65 mm). Furthermore, there appears to be no correlation between the session time length and cumulative IFM. This study can therefore be used as an initial estimate for the IFM and corresponding margins that can be expected for Gamma Knife ICON treatments. EP-2376 Evaluation of dosimetric consequences of setup errors on NPC using VMAT C.H.A. Liu 1 , D.L.W. Kwong 2 , W.Y.V. Lee 1 1 Tuen Mun Hospital, Department of Clinical Oncology, New Territories, Hong Kong SAR China 2 The University of Hong Kong, Department of Clinical Oncology, Pokfulam, Hong Kong SAR China Purpose or Objective (1) To investigate the impacts of setup errors on dose distribution to planning target volumes (PTVs) and organs at risk (OARs) in nasopharyngeal carcinoma (NPC) treatment using Volumetric Modulated Arc Therapy (VMAT); (2) To provide pilot data for evaluating the adequacy of current clinical practice by only correcting translational setup errors; (3) To give recommendation of the maximum setup errors allowed according to the dosimetric impact on PTVs and OARs. Material and Methods The data from 12 patients with NPC treated with VMAT were investigated. A total of 432 plans, including 6 translational-shifts new plans (+1mm, +2mm, +3mm) in Supero-Inferior (SI), Antero-Posterior (AP), Left-Right (LR) directions respectively; and 6 rotational-error-simulated new plans (+1 o , +2 o , +3 o ) for Row, Yaw and Pitch directions respectively, were created for each isocenter treatment. 1 patient, who suffered the greatest dosimetric effect by setup errors, was selected for further simulation. The weekly cone beam computed tomography (CBCT) images were reviewed and 2 sets of simulation were performed according to the recorded setup errors: (1) Simulation of combined translational and