Abstract Book

S1240

ESTRO 37

Purpose or Objective The aim of this study was to compare the use of kV planar imaging with combined kV and CBCT imaging in patients receiving radical Intensity Modulated Radiotherapy (IMRT) for head and neck cancer Material and Methods The verification images of 30 patients receiving radical IMRT for head and neck cancer were retrospectively reviewed by a single, experienced observer. Standard imaging protocol was kV planar imaging on days 2, 3 and then weekly with a CBCT acquired on day 1 and week 4. Additional kV images were taken in addition to the weekly schedule if corrections or gross displacements in shifts occurred for either kV or CBCT. Displacements were recorded in all 3 planes with the addition of the Pitch, Roll and Rotation (PRR) from the CBCT images. PRR were evaluated in the matching but could not be corrected as we do not have a 6 degrees of freedom treatment couch in our department. PRR values <3 degrees were accepted, if greater than this the patient was re-setup. Online shifts for both kV and CBCT images were performed for displacements of >0.3 cm. The displacement values in the 3 planes were used to produce population random and systematic errors using the method as outlined in On Target(1) Results The addition of 3-dimensional information of patient position provided by CBCT gave a more precise match to bone. Population random and systematic errors for kV (219 AP, 238 Lateral images) and CBCT (72 images) are shown below (Figure 1). Errors were similar for kV and CBCT matching except systematic error in the Superior- Inferior which was larger for CBCT. Figure 1. Population random and systematic errors Margin recipes (Van Herk2 and Stroom3) were calculated to determine the clinical target volume (CTV) to planning target volume (PTV) and are shown in Table 1. Uncorrected data uses raw displacement measurements assuming patient position was not corrected and gives worst case scenario. For corrected, displacements greater >0.3cm were assumed to be corrected to ‘0’ at treatment. Table 1. Planning target volume margins from margin recipes The results indicate that if all shifts were corrected for then the minimum CTV-PTV margin would be 0.3cm as is current departmental protocol. However, this would require the introduction of daily imaging which may have training and resource implications. For current imaging protocol, CTV-PTV margins need to be increased. Conclusion The use of kV imaging alone during treatment can underestimate necessary shifts. Introduction of limited CBCT imaging provides additional displacement data without a major cost or resource increase and may be an alternative to daily CBCT imaging if CTV-PTV margins are adjusted accordingly. References 1. On Target: Ensuring geometric accuracy in radiotherapy. The Royal College of Radiologists, 2008.

The aim of this study was to evaluate the dosimetric impact of computed tomography (CT) versus MRI based contouring variability for cervical cancer RT. Material and Methods 5 definitive cervical cancer patients had clinical target volumes (CTV) independently contoured by three radiation oncologists on CT and MRI scans as part of a previous study). Planning target volumes (PTV) were generated by applying 1cm margin around the CTV. Gold standard PTVs for each image modality were generated using the Simultaneous Truth and Performance Level Estimation (STAPLE) algorithm. All organs at risks (OAR) were contoured by one of two radiation therapy technologists (RTT) on CT scan and checked by a radiation oncologist. Two volumetric modulated arc therapy plans were generated for all patients; one using CT-STAPLE volume, and another using MRI-STAPLE volume. A prescription of 50Gy in 25 fractions was used for all cases. All plans were generated by one RTT and checked by another senior RTT, both experienced in cervical cancer radiotherapy planning. Dose metrics for planning target volumes (PTV) and OARs were compared to analyse the difference based on imaging modality (CT vs MRI). Paired t-tests were used to analyse the differences between the two modalities. Results Table 1 shows the average target and OARs doses comparing CT and MRI. Paired t-tests showed no statistically significant differences between the two modalities for all dose metrics assessed.

CT

based

MRI

based

Dose metric

Structure

plans

plans

PTV

D95 (Gy) 49.1

49.1

PTV

D50 (Gy) 51

51.1

PTV

D98 (Gy) 48.9

48.2

PTV

D2 (Gy)

52.2

52.5

Small bowel

V40Gy (%) 40.8

42.2

Rectum

V40Gy (%) 77.5

76.7

Bladder

V45Gy (%) 57.2

54

Left

femoral

V30Gy (%) 18.1

15.6

head

Right femoral head

V30Gy (%) 13.9

14.3

Table 1: Dose metrics averages (Dx - dose to x% of the PTV; Vx - the percentage of the volume receiving more than x% of the prescribed dose) Conclusion This preliminary study showed no dosimetric differences between CT and MRI based target volumes. Further work will include planning on a larger sample size and analysing contouring similarity metrics to correlate with dosimetric outcomes for cervical cancer radiotherapy.

Electronic Poster: RTT track: Image guided radiotherapy and verification protocols

EP-2370 Cone Beam Computed Tomography for radical Head and Neck Radiotherapy D. Thornberry 1 , R. McLauchlam 1 , D.M. Gujral 1 1 Imperial College NHS Trust, Radiotherapy, London, United Kingdom

2.

Van Herk et al. Int. J. Radiation Oncology Biol. Phys., Vol. 47, pp. 1121-1135, 2000 Stroom et al. Int. J. Radiation Oncology Biol. Phys., Vol. 43, pp. 905-919, 1999

3.