University of Wollongong Research Online University of Wollongong esis Collection University of Wollongong esis Collections 2012 Commissioning of a new total body irradiation protocol Zoe Baldwin University of Wollongong Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: [email protected]Recommended Citation Baldwin, Zoe, Commissioning of a new total body irradiation protocol, Master of Science - Research thesis, Faculty of Engineering, University of Wollongong, 2012. hp://ro.uow.edu.au/theses/3756
122
Embed
2012 Commissioning of a new total body irradiation protocol
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
University of WollongongResearch Online
University of Wollongong Thesis Collection University of Wollongong Thesis Collections
2012
Commissioning of a new total body irradiationprotocolZoe BaldwinUniversity of Wollongong
Research Online is the open access institutional repository for theUniversity of Wollongong. For further information contact the UOWLibrary: [email protected]
Recommended CitationBaldwin, Zoe, Commissioning of a new total body irradiation protocol, Master of Science - Research thesis, Faculty of Engineering,University of Wollongong, 2012. http://ro.uow.edu.au/theses/3756
A thesis submitted in fulfilment of the requirements for the Master of
Science – Research degree
University of Wollongong
Faculty of Engineering
2012
ABSTRACT
COMMISSIONING OF A NEW TOTAL BODY IRRADIATION
TREATMENT PROTOCOL
by Zoë Baldwin
Centre for Medical Radiation Physics
The current planning process for adult Total Body Irradiation (TBI) using the
PLATO 2D treatment planning system (TPS) at the Royal Brisbane and
Women’s Hospital is cumbersome; it does not adequately simulate the
treatment technique and a basic manual calculation is used to account for the
effects of heterogeneities. A new treatment delivery technique is proposed;
this thesis addresses the commissioning of this new regimen, including the
acquisition of dosimetric data under treatment conditions to quantify the
accuracy achievable by performing dose calculations using Oncentra
MasterPlan TPS for this new treatment technique.
Two treatment planning calculation algorithms (a pencil beam and a collapsed
cone) and a total of five separate beam models (four of which were
formulated specifically for TBI) were examined. Comparisons were made
between calculated and measured data, which include percentage depth dose
(PDD) values, profiles, an output factor, doses within homogeneous water-
equivalent phantoms and doses within anthropomorphic phantoms.
ii
TABLE OF CONTENTS
Introduction ......................................................................................................................... 1 1.1 Total Body Irradiation (TBI) ....................................................................... 1 1.2 Treatment planning ....................................................................................... 2 1.3 Calculation algorithms .................................................................................. 5 1.3.1 Pencil Beam (PB) ........................................................................................... 5 1.3.2 Collapsed Cone (CC)..................................................................................... 6 1.4 Dosimetric verification with phantoms..................................................... 7 1.5 Considerations in the design of a TBI programme................................. 9 1.6 Complications and Organs At Risk (OAR) ............................................10 1.7 Compensation ..............................................................................................11 1.8 Technique and patient set-up ....................................................................14 1.8.1 Patient set-up ................................................................................................16 1.8.2 Complex patient set-up...............................................................................19 1.9 Beam energy..................................................................................................20 1.9.1 Dose build-up...............................................................................................21 1.10 Treatment Planning System (TPS) data...................................................21 1.10.1 Commissioning of TBI TPS Calculations........................................23
Confirmation of beam remodelling and data requirments PHASE I .....................25 2 Initial investigation..............................................................................................25 2.1 Patient positioning.......................................................................................25 2.2 Dosimetric verification ...............................................................................26 2.3 Small cubic phantom...................................................................................27 2.4 PDD acquisition ..........................................................................................29 2.4.1 PDD for MLC-delineated Collimator 45° beam...................................31 2.4.2 PDD comparison for MLC-delineated Collimator 45° beam.............32 2.4.3 PDD acquisition for Collimator 0°, 40 cm x 25 cm field ....................34 2.4.4 PDD comparison for Collimator 0°, 40 cm x 25 cm field ..................35 2.4.4.1 Absence of build-up screen ................................................................35 2.4.4.2 Bolus density settings...........................................................................37 2.5 Acquisition of beam profile free in air.....................................................39 2.6 Comparison of in air-profiles ....................................................................40 2.7 Cable signal ...................................................................................................42 2.8 Simple solid water phantom ......................................................................42 2.9 Simple solid water phantom comparison................................................43 2.10 Thorax phantom data acquisition.............................................................45 2.11 Thorax phantom comparison....................................................................47 2.12 Anthropomorphic phantom data acquisition.........................................48 2.13 Anthropomorphic phantom comparison................................................52 2.14 Outcomes......................................................................................................55 2.14.1 Nucletron data submission .................................................................55
Data model validation PHASE II ..................................................................................57
iii
3 Dosimetric verification ......................................................................................57 3.1 Output Factor...............................................................................................57 3.2 PDD...............................................................................................................59 3.2.1 Model 1 ..........................................................................................................62 3.2.2 Model 2 ..........................................................................................................64 3.2.3 Model 3 ..........................................................................................................65 3.2.4 Model 4 ..........................................................................................................67 3.3 Profiles ...........................................................................................................69 3.3.1 Model 1 ..........................................................................................................70 3.3.2 Model 2 ..........................................................................................................71 3.3.3 Model 3 ..........................................................................................................72 3.3.4 Model 4 ..........................................................................................................72 3.4 Simple phantom comparison.....................................................................73 3.5 Measurement consistency investigation ..................................................74 3.6 Thorax Phantom Comparison ..................................................................76 3.7 Anthropomorphic Phantom......................................................................78 3.7.1 TLD variation investigation.......................................................................80 3.7.1.1 TLD calibration at ESSD....................................................................81
Discussion and Conclusion.............................................................................................86 4 Project Overview ................................................................................................86 4.1 Discussion .....................................................................................................87 4.1.1 PDDs .............................................................................................................87 4.1.2 Beam profiles measured in air ...................................................................87 4.1.3 Output factor................................................................................................87 4.1.4 Thorax phantom..........................................................................................88 4.1.5 Simple phantom...........................................................................................88 4.1.6 Anthropomorphic phantom......................................................................88 4.1.7 General Discussion......................................................................................89 4.2 Conclusion and Future Work....................................................................90
ICRU International Commission on Radiation Units and Measurements
IMRT Intensity Modulated Radiation Therapy
MU Monitor Units
OAR Organ At Risk
OMP Oncentra MasterPlan
PBE Pencil Beam Enhanced (calculation algorithm)
PDD Percentage Depth Dose
PTV Planning Target Volume
vi
RBWH Royal Brisbane and Women’s Hospital
RT Radiation Therapy
SSD Source to Surface Distance
SCD Source to Chamber Distance
TBI Total Body Irradiation
TLD Thermoluminescent Dosimeter
TPS Treatment Planning System
uA Type A standard uncertainty
vii
LIST OF FIGURES
Figure 1: An image of a digitised axial CT slice from the Plato 2D TPS, in which the lungs and spine have been outlined, however the true relative electron density (or physical density) from the CT data are not used in the Plato 2D dose calculation; default nominal values are used instead. ............................................................................................................................................3
Figure 2: The simulation position for the current treatment planning protocol, which will be used as the simulation and treatment position for the proposed new treatment technique. ..........................................................................................4
Figure 3: Lateral view of the current treatment set-up at the RBWH. The yellow lines indicate the 6 slice positions for the current treatment technique. .......................................................................................................................................4
Figure 4: An axial view of a treatment plan as displayed in Plato 2D. Note the bulk density corrections have been applied. ..........................................................................11
Figure 5: A jig used to hold the lead sheets employed in head shielding in TBI. .........................12 Figure 6: A simulation film taken of a patient for the lung compensator
design; this image is compared with a film taken at each treatment fraction for confirmation of treatment position. ..................................................................13
Figure 7: Chest plane from TBI plan, indicating the position of the bolus and its use in the treatment plan at RBWH...................................................................................14
Figure 8: Bolus used in a TBI treatment for both contour compensation and as a patient positioning aid. .......................................................................................................14
Figure 9: Lateral view of the current semi-reclining treatment position.........................................16 Figure 10: The proposed new supine treatment position..................................................................18 Figure 11: Demonstration of how the new technique can be utilised to treat
tall adult patients (pictured is a 194 cm “patient”), the treatment field is indicated by the white line and the dimensions of the BUS indicated by the green line. .........................................................................................................................19
Figure 12: The simple small cubic solid water phantom consisting of Gammex RMI 457 solid water slabs. ......................................................................................27
Figure 13: Dose plotted as a function of depth (corrected for effective point of measurement) in a 20 cm cubic solid water phantom, with and without the BUS..........................................................................................................................29
Figure 14: The PTW MP3 water tank used in the acquisition of the PDD data.................................................................................................................................................30
Figure 15: The phantom arrangement used to acquire the build-up region measurements for the PDD......................................................................................................31
Figure 16: The PDDs acquired at the patient plane in a large water tank at 380 cm SSD using the field delineated by the MLCs with a collimator angle of 45º. The plot labelled CAX was acquired 7 cm directly below the beam central axis (CAX); the other plot was acquired 43 cm away from the CAX plot in the gun-target direction. ....................................................................32
viii
Figure 17: Treatment plan configuration for the case of the PDD taken 7 cm directly below the beam axis in a 40 cm x 25 cm field.........................................................33
Figure 18: Treatment plan configuration for the case of the PDD taken 7 cm below the beam axis and 43 cm towards the target in a 40 cm x 25 cm field. ...............................................................................................................................................33
Figure 19: The PDDs acquired at the patient plane in a large water tank at 380 cm SSD using the field delineated by the jaws only with a collimator angle of 0º. The plot labelled CAX was acquired 7 cm directly below the beam central axis (CAX); the other plot was acquired 43 cm away from the CAX plot in the gun-target direction. ....................................................................34
Figure 20: The comparison of the PDDs measured 7 cm below the central axis and calculated by the treatment planning system in the absence of the BUS using both the pencil beam and collapsed cone algorithms, normalised to 10 cm depth. ......................................................................................................35
Figure 21: The comparison of the PDDs measured 43 cm away from the central axis and calculated by the treatment planning system in the absence of the BUS using both the pencil beam and collapsed cone algorithms, normalised to 10 cm depth. .................................................................................36
Figure 22: The comparison of the PDDs measured on the central axis and those produced by the treatment planning system using the collapsed cone algorithm, normalised to a depth of 10 cm, with three different density values applied. ................................................................................................................38
Figure 23: The comparison of the PDDs measured on the central axis and those produced by the treatment planning system using the pencil beam algorithm, normalised to a depth of 10 cm, with three different density values applied.................................................................................................................38
Figure 24: The y-axis profile obtained in air with a build-up cap with 50% of the central dose occurring at ±80 cm. ....................................................................................40
Figure 25: The y-axis profile normalised at the central axis position, calculated by the Collapsed Cone and Pencil Beam algorithms as compared with that measured...............................................................................................................................41
Figure 26: The simple solid water phantom replicated in the TPS for the 14 cm separation arrangement. ......................................................................................................44
Figure 27: Transverse view of the thorax phantom as set-up on the treatment bed. The mid-plane of the phantom is positioned 400 cm from the source (as indicated by the laser). The chamber is in the chamber cavity in the lung on the beam entrance side.........................................................................46
Figure 28: A beam’s eye view of the thorax phantom behind the BUS on the treatment couch...........................................................................................................................46
Figure 29: The treatment plan produced by OMP from a CT scan of the thorax phantom...........................................................................................................................47
Figure 30: The anthropomorphic phantom arrangement used in this investigation on the treatment bed. Lasers are used to align the phantom in the same manner as for a patient undergoing treatment...............................49
ix
Figures 31. a), b) c): Photos taken of a subset of the sections within the anthropomorphic phantom in which TLDs were placed to acquire dosimetric data.............................................................................................................................50
Figure 32.1 – 32.10: CT slices of the anthropomorphic phantom, showing the individual TLD positions within the phantom. The beam direction as it appears in the above figures is from the right. The percentage difference between doses at the TLD positions calculated by the collapsed cone and pencil beam algorithms and the corresponding measured doses are shown in blue and red respectively......................................................54
Figure 33: The water phantom used to acquire the output factor. Solid water, Perspex and plastic water are used to provide lateral scatter..............................................56
Figure 34: PDD as calculated at 7 cm below the CAX by the pencil beam algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the maximum dose, in a large volume water phantom..............................60
Figure 35: PDD as calculated at 7 cm below the CAX by the collapsed cone algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the maximum dose, in a large volume water phantom..............................60
Figure 36: PDD as calculated at 7 cm below and 43 cm lateral to the CAX by the pencil beam algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the maximum dose ............................................................61
Figure 37: PDD as calculated at 7 cm below and 43 cm lateral to the CAX by the collapsed cone algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the maximum dose ............................................................62
Figure 38: Treatment beam model PDD as calculated by the pencil beam algorithm for a 10 cm x 10 cm field normalised at dmax rather than the 40 cm x 25 cm field, showing better agreement with the measured PDD for the 40 cm x 25 cm field for models 2 and 3.........................................................68
Figure 39: Treatment beam model PDD as calculated by the collapsed cone algorithm for a 10 cm x 10 cm field normalised at dmax rather than the 40 cm x 25 cm field, showing better agreement with the measured PDD for the 40 cm x 25 m field for models 2 and 3...........................................................69
Figure 40: Profiles as calculated at 7 cm below the CAX by the pencil beam algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the central position ...........................................................................................69
Figure 41: Profiles as calculated at 7 cm below the CAX by the collapsed cone algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the central position. ..........................................................................................70
Figure 42: PDD as calculated by the pencil beam algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the depth of 10 cm...................................................................................................................................................77
Figure 43: PDD as calculated by the collapsed cone algorithm for each treatment beam model for a 40 cm x 25 cm field normalised to the depth of 10 cm. ...........................................................................................................................78
Figure 44: The graph displaying the average, minimum and maximum percentage deviations of the TPS calculated doses from the TLD doses for Pencil Beam calculations..........................................................................................79
x
Figure 45: The graph displaying the average percentage deviations of the TPS calculated doses from the TLD doses for Collapsed Cone calculations. ..................................................................................................................................80
Figure 46: The anthropomorphic/ solid water phantom arrangement used in the TLD accuracy and linearity investigation. .......................................................................81
Figures 47.1 – 47.10: CT slices of the anthropomorphic phantom, showing the individual TLD positions within the phantom. The percentage differences from the TLD of doses calculated by collapsed cone and pencil beam algorithms are shown in blue and red respectively for model 0. The TLDs were calibrated in a large phantom at ESSD. ...................................84
Figure 48: The graph displaying the average, minimum and maximum percentage deviations of the TPS calculated doses from the recalibrated TLD doses for PBE calculations. ......................................................................85
Figure 49: The graph displaying the average, minimum and maximum percentage deviations of the TPS calculated doses from the recalibrated TLD doses for CCE calculations.......................................................................85
xi
LIST OF TABLES
Table 1: The gamma analysis results comparing the measured PDD and the Pencil Beam algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm)...............................................35
Table 2: The gamma analysis results comparing the measured PDD and the Collapsed Cone algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm)...............................................36
Table 3: The gamma analysis results comparing the 43 cm off axis measured PDD and Pencil Beam algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................37
Table 4: The gamma analysis results comparing the 43 cm off axis measured PDD and the Collapsed Cone algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build- up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm)..................................37
Table 5: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and the measured data.......................................................41
Table 6: The results of gamma analysis performed to compare the Collapsed Cone calculation of the profile and the measured data .......................................................42
Table 7: The doses as measured by converting the charge collected in the simple solid water phantom to dose, where the standard uncertainty (uA) is derived according to TRS398[72] as the standard deviation of the mean. ......................................................................................................................................43
Table 8: The comparison of doses calculated by OMP using the Model 0 data with the doses measured by the Farmer chamber in the simple phantom. The tabulated values are the percentages by which the calculated dose exceeds the measured dose. ..........................................................................44
Table 9: The doses to water in cGy derived from the charge collected from the chamber in the central tissue plug, the spinal bone plug and the central lung plug on the entrance side of the beam. Each measurement was corrected for extra signal from cable irradiation..................................47
Table 10: The percentage difference between doses measured by the ionisation chamber in the tissue, lung and bone cavities and those calculated by the treatment planning system for 500 MU. A negative difference indicates a lower calculated dose than the measured dose...............................48
xii
Table 11: The average doses measured by the TLDs at each point of interest in the anthropomorphic phantom, including the mean average dose and the mean standard deviation of doses measured...........................................................51
Table 12: The dose at the reference point in cGy/MU calculated by the treatment planning system for each beam model and calculation algorithm for a 40 cm x 25 cm field defined at the linac isocentre. ..................................58
Table 13: The dose at the calibration point in cGy/MU calculated by the treatment planning system for each beam model and calculation algorithm for a 10 cm x 10 cm field defined at the linac isocentre. The measured calibration value was 0.06661 cGy/MU for a 40 cm x 25 cm field. ........................................................................................................................................58
Table 14: Percentage difference between dose at the calculation point in cGy/MU calculated by the treatment planning system for each beam model and calculation algorithm and the measured value of 0.06661 cGy/MU at the calibration point. A negative difference indicates a lower calculated dose than the measured dose......................................................................59
Table 15: The gamma analysis results comparing the measured PDD below the central axis and the PBE Model 1 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm) ...............................................................................................................................................62
Table 16: The gamma analysis results comparing the measured PDD below the central axis and the CCE Model 1 PDD with 10% or 2% and 1 mm distance to agreement, regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................63
Table 17: The gamma analysis results comparing the measured PDD 43 cm off axis and the PBE Model 1 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................63
Table 18: The gamma analysis results comparing the measured PDD 43 cm off axis and the CCE Model 1 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................63
Table 19: The gamma analysis results comparing the measured PDD below the central axis and the PBE Model 2 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm). ..............................................................................................................................................64
Table 20: The gamma analysis results comparing the measured PDD below the central axis and the CCE Model 2 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-
xiii
Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm). ..............................................................................................................................................64
Table 21: The gamma analysis results comparing the measured PDD 43 cm off axis and the PBE Model 2 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................64
Table 22: The gamma analysis results comparing the measured PDD 43 cm off axis and the CCE Model 2 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................65
Table 23: The gamma analysis results comparing the measured PDD below the central axis and the PBE Model 3 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm). ..............................................................................................................................................65
Table 24: The gamma analysis results comparing the measured PDD below the central axis and the CCE Model 3 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm). ..............................................................................................................................................66
Table 25: The gamma analysis results comparing the measured PDD 43 cm off axis and the PBE Model 3 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................66
Table 26: The gamma analysis results comparing the measured PDD 43 cm off axis and the CCE Model 3 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................66
Table 27: The gamma analysis results comparing the measured PDD below the central axis and the PBE Model 4 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm). ..............................................................................................................................................67
Table 28: The gamma analysis results comparing the measured PDD 43 cm off axis and the PBE Model 4 PDD with 10% or 2% and 1 mm distance to agreement. Regions of interest include the In Build-Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1 mm to 500 mm).......................................67
xiv
Table 29: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and the measured data for Model 1. ...............................70
Table 30: The results of gamma analysis performed to compare the Collapsed Cone calculation of the profile and the measured data for Model 1.................................71
Table 31: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and the measured data for Model 2. ...............................71
Table 32: The results of gamma analysis performed to compare the Collapsed Cone calculation of the profile and the measured data for Model 2.................................71
Table 33: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and the measured data for Model 3. ...............................72
Table 34: The results of gamma analysis performed to compare the Collapsed Cone calculation of the profile and the measured data for Model 3.................................72
Table 35: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and the measured data for Model 4. ...............................72
Table 36: The percentage differences between the doses measured in the simple solid water phantom and those calculated using the PBE and CCE algorithms for all beam models investigated for the same phantom geometry and number of MU in OMP. The positive differences indicate higher calculated doses than the measured doses.............................73
Table 37: The percentage difference between doses measured by the ionisation chamber in the tissue, lung and bone and those calculated by the TPS for 500 MU, where the chamber cavity has been replicated in the TPS as being filled with water. The ionisation chamber readings have been corrected for cable signal. A negative difference indicates lower calculated dose than the measured dose......................................................................76
Table 38: The results of the comparison of TLD dose measurements in the anthropomorphic/ solid water phantom with doses measured by the ionisation chamber......................................................................................................................82
Table 39: The doses in cGy measured with TLDs within the anthropomorphic phantom and comparison with the existing beam model (model 0) doses calculated by the Pencil Beam algorithm in OMP. The “% diff TLD” is the percentage by which the calculated dose exceeds the measured dose..............................................................................................93
Table 40: The doses in cGy measured with TLDs within the anthropomorphic phantom and comparison with the existing beam model (model 0) doses calculated by the Collapsed Cone algorithm in OMP. The “% diff TLD” is the percentage by which the calculated dose exceeds the measured dose..............................................................................................94
Table 41: The doses calculated using the PBE algorithm for each beam model, at each TLD position, and a comparison with the doses measured by the TLDs. For each model, the first column shows the calculated dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the TLD-measured dose at that point for the same number of MU..................................................................................95
Table 42: The doses calculated using the CCE algorithm for each beam model, at each TLD position and a comparison with the doses
xv
measured by the TLDs. For each model, the first column shows the calculated dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the TLD-measured dose at that point for the same number of MU..................................................................................96
Table 43: The doses calculated using the pencil beam algorithm for all models and the percentage differences between the measured and calculated values. For each model, the first column shows the calculated dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the revised TLD-measured dose at that point for the same number of MU..........................................................................................97
Table 44: The doses calculated using the Collapsed Cone algorithm for all models and the percentage differences between the measured and calculated values. For each model, the first column shows the calculated dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the revised TLD-measured dose at that point for the same number of MU....................................................................98
C h a p t e r 1
INTRODUCTION
1.1 Total Body Irradiation (TBI)
Total Body Irradiation is a treatment technique that involves the irradiation of the whole
body. Several disorders where the entire body requires treatment have been proven to
respond to TBI, including lymphomas[1-3], leukemias[4-6] and aplastic anaemia[7]. In
situations where an allogeneic bone marrow transplant (BMT) or stem cell transplant is
prescribed it is common for such patients to undergo TBI in combination with chemical
therapy as part of a pre-transplant cytoreductive conditioning process[8, 9]. TBI yields two
beneficial outcomes in this process: firstly in the destruction of cancerous cells (also an
outcome of the chemotherapy) and secondly in the suppression of the immune system to
combat any rejection of transplanted bone marrow[10]. Although this preparatory
regimen for BMT has evolved from TBI alone, the combination with chemical therapy
agents has proven much more effective.
Advantages of TBI are that no organ is unintentionally spared and there are no sanctuary
sites, as the dose is homogeneous throughout and independent of blood supply[11]. The
radiation dose is not affected by any detoxification or excretion and if desired the dose
distribution may be adjusted through the use of partial compensators for the
radiosensitive normal tissues, and boosts may be prescribed for the radio-resistant regions.
As in all forms of patient therapy, in TBI it is desirable to achieve the maximum possible
therapeutic ratio[12] providing greater disease control and minimal normal tissue toxicity.
The radiation induced toxicity of normal tissue is a major limiting factor in the
effectiveness of total body irradiation. Clift et al[5] demonstrated the fine line between an
improvement in tumour control/ immuno-suppression and the transplant-related
complications for different irradiation regimens. A lower relapse rate for one regimen
translates to an improved long term survival, but with an increased risk of early mortality
resulting from increased acute toxicity.
2
1.2 Treatment planning
The use of Computed Tomography (CT)-based treatment planning systems in
conventional external beam radiation therapy is well established[13-16]. The tomographic
data provides the necessary information for the treatment planning system to determine
the patient’s external contour, to calculate the transmission and absorption of the
radiation in external beams, and hence to display the distribution of absorbed dose within
the patient. The CT data set which provides details of external contour, geometry of the
internal organs and their composition, constitutes a suitable base from which to calculate
the effect of tissue heterogeneities. Conventional external beam radiotherapy is conducted
at a nominal isocentric distance of 100 cm (often referred to as 100 cm Source-Axis-
Distance); treatment techniques are described as extended SSD when the source to
surface distance (SSD) is greater than 100 cm.
CT-based TBI treatment planning is less common. Because of the complex nature of an
extended SSD total body irradiation, Hui [17] investigated the accuracy of a CT-based
treatment planning system (Pinnacle™), finding that the use of CT data provides the
necessary anatomical and density information that ultimately results in a more accurate
dose calculation for TBI. This is especially important for the lung dose (an organ at risk),
where compensation is required to achieve the desired uniformity, and the CT data can
assist in compensator design. A Dose Volume Histogram (DVH) generated using the CT
data can provide homogeneity indexes for organs at risk (OAR) as well as the Planning
Target Volume (PTV).
Plato 2D (Nucletron, Veenendaal, The Netherlands) - the treatment planning system
currently used for planning TBI treatments at the Royal Brisbane and Women’s Hospital
(RBWH) - utilises 6 slices of CT data for the purpose of determining the location of the
kidneys and lungs (organs at risk) (see figure 3). The lung is outlined and a bulk density
correction (of 0.33 for adults and 0.41 for children) is applied. All other anatomical
structures are assigned a relative density of 1 (water). The external patient contours
obtained from the CT slices are used to account for the contour variation of the patient
(see figure 1). The technique is time consuming and does not reflect the dose to all
3
pertinent organs, nor the treatment position as the patient is simulated in a more supine
position (see figure 2) than the semi-reclining position used for treatment.
Figure 1: An image of a digitised axial CT slice from the Plato 2D TPS, in which the lungs and spine have
been outlined, however the true relative electron density (or physical density) from the CT data are not used
in the Plato 2D dose calculation; default nominal values are used instead.
The Plato 2D software is no longer supported by the manufacturer and therefore the use
of a new treatment planning system was proposed as the basis of a new TBI protocol.
Oncentra MasterPlan (OMP) (Nucletron, Veenendaal, The Netherlands) is also a CT-
based planning system that utilises helical CT data and is currently used for treatment
planning of all other three-dimensional conformal radiotherapy (3D-CRT) treatments at
the RBWH. In OMP, a 3D density matrix derived from pixel CT number values from CT
images or from user-specified density values assigned to regions of interest is generated,
and this is used by one of two well established[13, 18-22] calculation algorithms to
calculate doses to the patient,.
4
Figure 2: The simulation position for the current treatment planning protocol, which will be used as the
simulation and treatment position for the proposed new treatment technique.
Figure 3: Lateral view of the current treatment set-up at the RBWH. The yellow lines indicate the 6 slice
positions for the current treatment technique.
The prescribed doses range from 5 to 14 Gy delivered in up to 8 fractions, generally twice
daily. At RBWH doses for both paediatric and adult TBI are prescribed at the midpoint
5
behind the umbilicus. Dose limits for the critical organs are often included in the
prescription.
1.3 Calculation algorithms
Each commercial treatment planning system provides a selection of photon beam
treatment planning algorithms. Two algorithms available in Oncentra Masterplan for
photon beam treatment planning are Pencil Beam and Collapsed Cone. The same EGS4
Monte Carlo code proposed by Mackie et al [23] is used as the basis of the point dose
kernel for both algorithms. This point kernel is later used to obtain a pencil kernel (for the
pencil beam algorithm) through discrete integrations over the depth of calculations. The
accuracy and limitations of both calculation models have been investigated[19, 20, 22, 24-
32]. Oncentra Masterplan - unlike other commercially available treatment planning
systems - does not allow the user (physicist) to “tweak” the Monte Carlo generated energy
spectrum derived from the measured data. Instead this tweaking or beam characterisation
is conducted by the manufacturer off-site, based on the measured data supplied by the
user.
The dose calculation engine in OMP calculates the dose per incident energy fluence, based
on energy deposition kernels and combines this with calculated incident energy fluence
for the specified beam. The energy deposition kernel describes the dose response to an
incident radiation beam in water. For dose calculations the beams are treated separately
and the energy fluence exiting the treatment machine per beam is calculated as an initial
step of the dose calculation. The transport and dose deposition of the secondary particles
is taken into account by way of a pencil kernel for the pencil beam algorithm or a point
kernel for the collapsed cone algorithm.
1.3.1 Pencil Beam (PB)
This algorithm as implemented in OMP was originally proposed by Ahnesjö[32] and
utilises a particular technique to calculate the transport of photons and electrons and
deposition of dose. As the name suggests it is based on a pencil beam. For volumetric
integrations, the pencil beam is parameterised every 2.5 mm along the propagation
direction to a maximum of 502.5 mm. The actual beam is divided into a discrete grid of
6
infinitesimally narrow beams (pencils) which originate at the beam’s virtual source.
Equivalent path length corrections are applied to manage any regions of heterogeneity for
the primary dose contribution; however this does not take account of lateral transport of
secondary particles in heterogeneous media. A 1-dimensional convolution is applied along
fan lines with an exponential function to manage the scattered radiation[32-34]. The dose
at any point of interest is the result of the summation of the dose contributions from each
pencil beam. Heterogeneity corrections are performed by simply scaling the dose spread
function using the physical density in the path to save calculation time rather than
including density scaling for lateral heterogeneities. A decrease in the calculation accuracy
in regions where lateral electronic equilibrium is not present (where lung and air regions
are present) is the result of this simplification of the lateral transport of secondary particles
in the dose calculation. The penumbral widening observed in regions of low density (such
as lung) is not well predicted when using the pencil beam algorithm (the implementation
of which in OMP is known as Pencil Beam Enhanced PBE).
1.3.2 Collapsed Cone (CC)
Collapsed cone algorithms are gaining in popularity; as they provide a superior dose
calculation for regions where lateral charged particle equilibrium does not hold, such as in
lung and where the beam overshoots the patient’s body. The collapsed cone convolution
algorithm implemented in Oncentra Masterplan™ was first proposed by Ahnesjö[35]. It
involves apportioning all the energy released into coaxial cones of equal solid angles about
the point of interaction. The direct energy fluence and the head scatter energy fluence are
portioned into point kernels which describe the primary and scattered absorbed doses.
The dose at each point is obtained from interpolation in the 3D calculation grid. For
heterogeneities the radiological path length (as calculated from the density matrix) along
each cone axis replaces the geometrical path length. A limitation of the implementation of
the collapsed cone algorithm (known as Collapse Cone Enhance CCE) in OMP is in the
voxelisation of the calculation. In the case of the pencil beam algorithm the voxel
calculation size is determined by the user, however in the interest of calculation efficiency
the collapsed cone algorithm employs a finite number of voxels and as a result where the
volume is large (such as in the case of full body CT) the voxel size may be too large to
produce a dose distribution with high spatial resolution. This would limit the extent to
7
which small variations in the patient density or external contour would be handled by the
TPS calculation. A too coarse voxel size would be one that gives rise to significant errors
in reconstructing the positions of or dose distributions within OARs or targets. For TBI
voxel sizes of 5 mm or less should be satisfactory given the size of the OARs and the
PTV (the whole body).
1.4 Dosimetric verification with phantoms
The dose distribution of a clinical treatment may be inferred from measurements made in
a phantom. The term phantom refers to a material that is used to simulate the scattering
and absorption of radiation in tissue. Phantoms commonly consist of water, which has
similar absorption properties to muscle and soft tissue, however solid substances, such as
Plastic Water™[36] have been manufactured to also replicate the same properties.
Measurements of dose are made with dosimeters, the most commonly used type of which
in radiotherapy physics is the ionisation chamber, with various designs available[37-39]. A
common type of ionisation chamber (or ion chamber) consists of a conductive wall
surrounding a cylindrical gas-filled cavity within which is a centrally placed collecting
electrode. When a voltage is applied between this electrode and the wall, ions created in
the active volume by incident ionising radiation are attracted to the electrode and the
collected charge recorded by an electrometer. The primary measurement quantity is mass
ionisation (C/kg), the ionisation Q (C) per unit mass m (kg) of gas in the chamber, which
is related to dose (J/kg) to the gas in the chamber Dgas by the mean energy required to
produce one ion pair W (J). W/e (J/C) is then the mean energy expended per unit charge
produced, where e is the electronic charge (C), leading to the following equation:
=e
W
m
QDGas
The gas is most commonly air; more dense gases are sometimes used to increase
sensitivity by increasing the ionisation density within the volume[38]. One primary
requirement under which most radiotherapy dosimetry measurements are made is that in a
phantom the chamber is assumed to be small enough so as not to significantly perturb the
radiation field within the phantom. In such situations the charge produced and collected
8
in the chamber volume is proportional to the dose delivered (at the same point in the
absence of the chamber itself) to the phantom medium, from which the energetic
electrons producing the gas ionisation come in a megavoltage beam. Thus the dose to the
surrounding medium is given by the Bragg-Gray equation[39]:
GasMedGasMed SDD ,=
Where Smed,gas is the ratio of the mean mass stopping power of the energetic electrons in
the phantom medium and in the gas. For indirectly ionising radiation such as high energy
photons the contribution of any primary radiation to the ionisation events in the chamber
may be ignored. Since the theory assumes that all the measured ionisation is produced by
energetic electrons liberated by the photons in the surrounding phantom material, the wall
and other solid material of the chamber need to be similar to the phantom material and
the wall of chamber needs to be sufficiently thin, to ensure that almost all of the electrons
producing the charge measured have effectively been generated in the phantom.
Deviations from strict Bragg-Gray cavity conditions are dealt with by the use of
perturbation factors. In practice ion chambers are calibrated to provide chamber – specific
conversion coefficients which enable a direct conversion from measured charge to dose to
water.
Thermoluminescent Dosimeters (TLDs) have been used for in-vivo dosimetry[12, 14, 40].
TLDs provide a means of acquiring quick in-vivo measurements for radiotherapy and are
commonly used to verify the dose delivered to the eye when treating the head. Many
radiotherapy centres utilise TLDs for in-vivo measurement of superficial doses in total
body irradiation cases. There is a variety of TLDs available today, the most common
compositions being lithium fluoride, lithium borate and calcium fluoride, of which lithium
fluoride is the most widely used[41]. The most commonly used TLD in a radiotherapy
department is the Harshaw Chemical Company LiF TLD-100 made in the form of a small
disk. A TLD’s dose response may be described as the thermoluminescent output per unit
of absorbed dose in the phosphor and under certain conditions the dose response curve
of the LiF (TLD- 100) is linear up to 103 cGy and supralinear beyond this point[12]. The
TLDs can be calibrated so that their response is related to absorbed dose to water under
9
reference conditions (i.e. calibrated against an ion chamber). Metcalfe[12] outlines a cheap
and easy method for packaging TLD chips for clinical in-vivo measurements. Each centre
usually develops its own procedures for acquiring and processing TLD measurements
(which may include pre-irradiation annealing, post-irradiation annealing and a particular
heating program during the readout process). If these procedures are adhered to correctly,
they lead to accurate and consistent dose measurements and an uncertainty equal to 1
standard deviation of the mean value. The Australasian College of Physical Scientists and
Engineers in Medicine (ACPSEM)[15] recommends a calibration of the TLDs to be
undertaken with every measurement and a linearity check to be completed following the
initial use of each mode used or after a malfunction or repair of the TLD equipment.
Ionisation chambers and TLDs were used extensively in the work presented in this thesis.
1.5 Considerations in the design of a TBI programme
Historically TBI has been given as a single fraction, however the radiation induced side
effects can be better managed under a fractionated regimen, based on radiobiological
principles of preferential normal tissue repair[12]. Fractionation allows the prescription of
a higher total dose and improvement in the long term survival. Research by Deeg[42],
Cosset[43] and Cosset[44] has shown fractionated TBI can produce a better therapeutic
ratio than a non-fractionated scheme. According to Cox[45] the most common regimens
are 12 Gy given in six fractions over 6 days or twice a day over three days.
The aim of TBI is the ablation of bone marrow and the destruction of circulating
leukemic cells; since these are widely distributed throughout the body, the target for TBI
treatments is the entire body.
The achievement of uniform dose throughout the patient in TBI is a challenge; an
overdose to any critical organs will increase the toxicity while an underdose to the target
will increase the risk of relapse. A region of low density results in higher dose within and
beyond the low density region, as a result of the reduced attenuation of the beam as it
travels through the low density medium. The works of Keane et al[46] and Van Dyk et
al[47] have shown that a change of 5% in the dose to lung could result in a 20% change in
the incidence of pneumonitis, a complication potentially fatal for a whole lung irradiation.
10
Khan[36] discusses the clinical achievement of a high dose uniformity in TBI, accepting
dose uniformity along the body axis of ±10%, with extremities and certain non-critical
structures receiving higher doses, in agreement with AAPM report 17[8].
As survival rates have increased for patients treated with total body irradiation, the
investigation of long term side effects and the quality of life has become more significant
clinically. In TBI, treatment parameters prescribed such as the total dose, the fractionation
regimen[43, 44] and dose rate[48], are influenced by the possible toxicity to sensitive
organs and normal tissues such as lungs and kidneys. Consequently the accuracy attainable
in TBI treatment planning and dose delivery is a factor limiting the application of certain
techniques. The International Commission on Radiation Units and Measurements
(ICRU)[49, 50], recommends that the overall accuracy in the dose delivery of external
beam radiation therapy be ±5% and that the dose range within the target volume not
exceed -5% to +7% of the reference dose. However the American Association of
Physicists in Medicine (AAPM) Task Group 29[8] suggests that a 5% accuracy could be
unachievable in TBI, but if the prescribed dose falls far below the onset of normal tissue
toxicity levels then a 5% accuracy may be relaxed.
1.6 Complications and Organs At Risk (OAR)
In radiation therapy, to achieve the desired result, the destruction of cancerous cells needs
to occur without excessive destruction of normal tissues. Certain organs and tissues are
more radiosensitive than others and become a limitation when treating patients with
radiation therapy. It is important to limit the radiation dose to such organs so as to avoid
undesirable outcomes such as pulmonary toxicity.
In TBI the entire body is irradiated and as a result a number of critical structures can be
considered organs at risk. Limitations to the effectiveness of TBI in the conditioning
process exist, the most notable limitation being the normal tissue tolerance of certain
organs, such as the lungs, kidneys and eyes[8, 13, 14, 18]. Each of these is associated with
a higher tissue complication factor which results in a greater likelihood of cumulative
toxicity than other organs with lower complication factors[14]. In the prescribing and
planning of TBI, it is important to develop the appropriate procedures to protect these
11
organs at risk. The most concerning OAR is the lung, the associated complications for
which are one of the major causes of mortality following total body irradiation[18].
Figure 4: An axial view of a treatment plan as displayed in Plato 2D. Note the bulk density corrections have
been applied.
1.7 Compensation
Compensation is used for tissue density heterogeneities, surface irregularities and overall
separation variations. Because the large variations in body thickness and the presence of
tissue heterogeneities in TBI would lead to considerable non-uniformity of the dose along
the body, bolus and/or compensators are usually used to provide a much more uniform
dose distribution throughout the body.
Bolus material is often placed around the patient for support as well as contour
compensation. In regions where the patient thickness is significantly reduced, tissue
equivalent bolus may be added to achieve the required thickness. To account for
variations in patient shape either tissue-equivalent bolus or tissue compensators should be
employed. Although using the AP/PA treatment technique would reduce the amount of
12
the required lung tissue compensation, positioning of the arms to shadow the lungs in the
bilateral treatment technique partially compensates for lack of tissue homogeneity,
reducing the additional shielding and bolus needed for compensation[17, 51]. Bolus
materials generally employed in clinical use are pliable, soft tissue equivalent and non-
toxic. The equivalence refers to the volumetric electron density. A number of materials
have been used[8, 52]; bolus material if placed under the legs in TBI can provide both
added scatter and support during the treatment.
Figure 5: A jig used to hold the lead sheets employed in head shielding in TBI.
High density material such as lead is also utilised in the reduction of the intensity of
incident radiation to regions of OAR (see figure 5). For the regions of low electron
density such as the lung (a highly radiosensitive organ) the dose absorbed in the absence
of compensation would be higher than the intended dose, due to the lower attenuation of
the beam in the lung. Additional compensation to protect the lungs leads to an underdose
of the bone marrow in the sternum. At RBWH a megavoltage photon isocentric boost is
prescribed and delivered to the sternum at conventional SSD to address this issue.
13
Figure 6: A simulation film taken of a patient for the lung compensator design; this image is compared with a
film taken at each treatment fraction for confirmation of treatment position.
The current technique at the RBWH uses lead sheets (the thickness of which is manually
calculated based on measured transmission factors) to reduce the dose to the head and
lungs, positioned on a build-up screen (BUS) with their exact placement determined and
verified by film prior to each treatment. In addition bolus positioned adjacent to the
regions of the abdomen under the arms is used to provide additional attenuation to
reduce the dose to both the lungs and kidneys whilst also maintaining a relatively uniform
external surface. Although the bolus materials have densities a little different from water, a
bulk density of 1 g/cm2 is used when outlining the bolus material.
The majority of bolus material at used in TBI at RBWH is beeswax (see figure 8) with a
measured relative electron density of 0.96 (compared to water). Thicknesses of this bolus
used (up to 12 cm) are therefore unlikely to differ from water in their attenuation by more
than the attenuation produced by 0.5 cm of water. This equates to ~1.3% dose error
beyond the wax depth. For the current technique this could be an issue, however for the
proposed technique the patient will be CT scanned with the bolus material in place and
therefore taken into account in the TPS calculations.
14
Figure 7: Chest plane from TBI plan, indicating the position of the bolus and its use in the treatment plan at
RBWH.
Figure 8: Bolus used in a TBI treatment for both contour compensation and as a patient positioning aid.
1.8 Technique and patient set-up
There are a number of published techniques[9, 53-59] for the delivery of total body
irradiation. Each technique attempts to achieve uniform dose throughout the body while
not exceeding the tolerance dose to any OAR[14]. These techniques vary in the number
and type of the radiation sources used. Where most techniques use beams provided by
linear accelerators, use beams coming from multiple sources from several directions[36,
54]. A common and rather simple technique involves patients treated with opposing
bolus
15
fields at a fixed extended source to patient midline distance[36, 60], thereby employing
large, uniform photon beams to ensure the entire body receives a uniform dose.
Although it is possible to use multiple adjacent fields delivered at standard treatment
distances (100 cm SAD) to achieve total body irradiation, the use of such a technique
would result in complicated patient treatment set-ups and field junctions which introduce
non-uniformity and uncertainty in the dose distribution[36]. A further problem in using
this technique for TBI is that while the aim is to deliver radiation to the entire body, cells
circulating through the body may receive a reduced dose as a result of delivering smaller
fields separately. Thus an extended SSD treatment using opposed fields, which can easily
be implemented, is usually preferred if the treatment room is sufficiently large.
To achieve a field size large enough to encompass an adult human, the geometric
divergence of the beam from the source is utilised, which results in a required source to
patient mid-sagittal distance of 3 to 5 meters. Current conventional linear accelerators are
often used in large bunkers and as a result can apply a broad lateral horizontal beam (with
a gantry angle of 90º or 270º), with the maximum field size typically 40 cm x 40 cm (as
defined at 100 cm from the x-ray source), which diverges to 160 cm x 160 cm at 400 cm
from the source. The patient is set up at this distance with their mid-sagittal plane
perpendicular to the beam axis. The correct location of the mid-sagittal plane (patient
midline) is indicated by a ceiling-mounted laser. A collimator angle of 45º may be used to
take advantage of the longest available field dimension, which is 200 cm along the
diagonal at 400 cm from the source (the linac’s primary collimator limits the maximum
field diameter to 50 cm at 100 cm form the source). The opposing lateral field is achieved
by rotating the patient couch through 180º about a vertical axis. The technique currently
employed at RBWH utilises a conventional linear accelerator (Elekta Precise™) delivering
opposed lateral beams of 10 MV, with a BUS to avoid skin sparing.
16
Figure 9: Lateral view of the current semi-reclining treatment position.
Another advantage of performing the TBI treatment at an extended SSD is the
consequent reduction in dose rate to the patient which should be below 10 cGy/min to
prevent any dose rate related toxicity[48, 61]. This reduction occurs as a result of the
inverse square law. However if the dose rate is still not sufficiently low at the position of
the patient, a further reduction is required (achieved by decreasing the pulse repetition rate
of the linear accelerator). With the application of a decreased dose rate, an increase in the
treatment time results and the use of immobilisation devices may be necessary for certain
treatment protocols.
1.8.1 Patient set-up
In the case of extended SSD single source treatments, patients may be treated lying supine
with lateral beams; or standing[9, 62] or seated or lying on their side treated with an
anterior and a posterior beam[57]; in all instances it is necessary to rotate the patient in
order to obtain dose uniformity via an effective parallel opposed pair of beams.
In the cases where the maximum achievable field size is not sufficiently large, the patient’s
legs may be bent to reduce the overall length of the patient. Some techniques achieve this
by treating the patient seated in a custom chair; however such a treatment position may
17
not be replicated in the CT simulation, which can result in a reduction in the calculation
accuracy of the treatment planning process. In departments that have a sufficiently large
bunker it is common for patients to be treated lying supine on a bed. This has the
advantage of the patient being imaged for treatment planning and treated in the same
position. If the treatment requires the patient to have bent legs, with the use of a wide
bore CT the patient may still be imaged and treated in the same position, maintaining the
same pelvic tilt and anatomical geometry, potentially producing more accurate dose
estimates.
The Anterior-Posterior/ Posterior-Anterior (AP/ PA) technique requires appropriate lung
blocking to reduce the dose to the lung, concurrently allowing sufficient chest wall dose
coverage. Boost fields to regions of unacceptably low dose are often employed at the
RBWH as a part of the treatment process. In an opposed lateral supine treatment, the
patient’s arms are used as partial lung compensators. However with the AP/ PA
technique, if the patient is supported in an upright position, the overall separation of the
effective fields on the patient surface is decreased and less variable, which is advantageous
for dose uniformity, particularly for lower beam energies such as 60Co.
For reproducibility of set-up the supine position is superior[55]. However in using a
supine position in conjunction with a lateral beam (as is most commonly found) the dose
to the head would generally be too high, as a result of the greatly reduced lateral
separation. To reduce this high dose to the clinical volume a compensation filter (or
shield) is normally applied.
After consideration of the issues discussed above, the technique involving treatment of
supine patients with lateral beams, at a distance from the x-ray source to the patient’s mid-
sagittal plane of 400 cm, was proposed for the new TBI protocol at RBWH. However for
patients with a sufficiently large separation, consideration may be given to treating such
patients using an anterior-posterior technique to yield the required dose uniformity. In
order to maintain patients in such positions patient immobilisation and support systems as
complex as those used in the current treatment technique will be required.
18
The proposed technique (see figure 10) will enable the treatment of tall patients by
treating with the collimator rotated to 45º (see figure 11). In order to minimise dosimetry
complications resulting from scattered radiation from the direct irradiation of the steel
frame of the BUS and metal frame of the treatment bed, MLCs will be used to collimate
the treatment field to reduce the effect of this.
Figure 10: The proposed new supine treatment position.
19
Figure 11: Demonstration of how the new technique can be utilised to treat tall adult patients (pictured is a
194 cm “patient”), the treatment field is indicated by the white line and the dimensions of the BUS indicated
by the green line.
1.8.2 Complex patient set-up
Certain institutions have adopted more complex set-ups to address some of the issues
associated with TBI, to improve patient comfort and treatment reproducibility. Papiez[63]
utilised an extended SSD with the patient on the floor. The technique comprised 4 fields
in Anterior-Posterior (AP) and Posterior-Anterior (PA) treatments with the overlapped
regions modulated by multileaf collimators (MLCs).
Similarly to cater for a small bunker design, La Macchia[64], utilised an in-house designed
floor couch and segmented MLC fields, producing fields that were adjusted to deliver the
prescribed dose to the midline. An early technique utilised on 60Co units was the
translation of the couch during treatment, where the patient was moved continuously
through the treatment beam[59, 65].
20
Harden[66] describes a standing technique used at Addenbrooke’s Hospital, in which
patients are treated on a non-dedicated offset (relative to the centre of the bunker)
machine with anterior and posterior fields using large horizontal fields with customised
lung compensators.
1.9 Beam energy
Historically TBI was initially delivered with cobalt-60 low energy megavoltage (1.25 MeV)
beams[8] with protocol dose rates around 5 to 10 cGy/min. However 60Co machines have
largely been superseded by linear accelerators and total body irradiation techniques have
been adopted for linear accelerator beams*.
For lateral treatment techniques the photon beam energy required is dependent upon the
thickness of the patient and the dose homogeneity specification. The variation in
thickness of the patient along their sagittal axis will affect the homogeneity of dose
distribution; in addition the patient thickness along the central ray of the beam will also
impact upon the overall homogeneity. For the parallel lateral opposed pair technique, as
the patient thickness increases (or the beam energy decreases) the central axis maximum
dose near the surface increases relative to the midpoint dose, an effect referred to by
Khan[36] as the tissue lateral effect. In general for the sparing of subcutaneous tissues for
overall thickness exceeding 20 cm, Khan’s analysis indicates that 10 MV or higher energy
beams should be used.
The Task Group 29 of the AAPM[8] investigated the ratio of the peak dose to the
midline dose on the central ray versus the patient thickness, for 6- and 25-MV beams;
concluding that the greater the energy, the lower the dose variation throughout the
patient, the greater the treatment distance the lower the dose variation; the greater the
patient’s diameter the greater the dose variation. For a lateral opposed beam set-up where
the patients’ thicknesses range from approximately 38 cm to 50 cm only 25-MV x-rays at a
distance of 300 cm could result in a dose uniformity within 15% for the largest patients.
* in Australia there are no Cobalt- 60 units suitable for TBI
21
In summary 10 MV or higher energies are preferred, producing a more uniform dose
distribution for large separations in the case of the parallel lateral opposed pair technique.
They do however require a spoiler or bolus to increase the skin dose as discussed below.
The RBWH currently uses 10-MV beams and intends to continue the use of 10 MV (the
highest energy available on its current linacs) for future TBI treatments with the proposed
new protocol.
1.9.1 Dose build-up
It is well known that due to an initial electron fluence build-up with depth, the maximum
dose on the central axis of a megavoltage beam occurs not at the surface but at a certain
depth below it, with a progressive increase in dose from the surface to the depth of dose
maximum. The surface dose is typically about 20% of the maximum dose. Although this
skin sparing effect depends on a number of factors including SSD, field size and
configuration of any blocking tray present, in general the effect becomes more
pronounced as the photon energy increases. For use of higher energy x-rays, consideration
must be given to the effects of any low dose in the build-up region.
Because most TBI protocols do not require skin sparing, a “beam spoiler” is introduced
to bring the surface dose to at least 90% of the maximum dose; most of the electron
fluence build-up then takes place in the beam spoiler (or build-up screen), a sheet of low
atomic number material placed between the patient and the linac[8]. The thickness and
location of this BUS will be dependent upon the desired clinical dose criteria. In general
the screen is placed as close to the patient as possible, in order to achieve the greatest
increase in the skin dose over the whole surface of the patient facing the beam. Sanchez-
Nieto et al[67] state that a 1 cm thick PerspexTM sheet for a 10-MV photon beam will
increase the skin dose up to 97% of the maximum dose. A Perspex BUS of this thickness
is currently in use for TBI treatments at RBWH, and it is proposed to continue using it in
the new protocol.
1.10 Treatment Planning System (TPS) data
Treatment planning systems are used to calculate the dose to the patient, and need to do
this accurately if the desired outcome is to be achieved. They require data collected under
22
specified conditions to perform these calculations, and this data is normally collected
during the commissioning of the treatment unit and the TPS. Whether this data is
adequate for accurate calculations at an extended treatment distance is a matter of
conjecture.
The change in percentage depth dose (PDD) with increasing SSD is a consequence of the
inverse square law and the Mayneord Factor can be used to estimate this change[36].
However if this method is solely relied on for the dose calculations for patient dosimetry it
can lead to errors of up to 6% and therefore measurements of the PDD under treatment
conditions are recommended[36]. It should be noted for extended SSD TBI, Houdek et
al[68] compared the accuracy of dose calculations using beam data measured at the
extended SSD to that using data recalculated (from 100 cm SSD measured data) for a 10-
MV beam. Their limited investigation yielded comparable results for the two methods
which can be explained by considering the scattering conditions surrounding the patient
and water tank. A primary motivation or rationale for the investigation presented in this
thesis is the potential to improve upon the accuracy with which the current treatment
planning system calculates the dose distribution in TBI of a patient during treatment.
Phantom Scatter Factors (SP) can be applied to correct for the variation in the amount and
distribution of scatter resulting from a change in phantom size[36]. These factors quantify
the change in dose at a fixed point as the volume of phantom irradiated with a fixed
collimator opening is varied.
These factors used in dose calculation are normally interpolated from data acquired at 100
cm SSD, from measurements acquired using a phantom whose dimensions exceed the
field size; however for TBI this is not easily achievable and the TBI field is usually larger
than the phantom area presented to the beam. This difference in geometry would be
expected to impact any Tissue Maximum Ratios (TMRs) and field output factors for the
TBI calculations, however Curran et al[69] show that for TMRs measured in three
different phantom sizes at 100 cm SAD and 400 cm SAD there was at most a change of
1.5% in TMRs.
23
Collimator Scatter Factors (SC) are used to correct for the variation in the amount of
scatter in the beam incident on the phantom resulting from change in field size from the
calibration reference field[36]. Such factors are used in dose calculations and may be
derived empirically from interpolation of data acquired at 100 cm SSD. The variation
between SC measured at extended distance and 100 cm SSD should be limited[69] (for
field sizes larger than 25 cm x 25 cm of the order 0.3%[69]), however the SC measurement
at ESSD is potentially complicated by backscatter from adjacent surfaces and objects not
present in 100 cm SSD fields. This investigation aims to assess the suitability of utilising
the data available in the OMP measured at SSD 100 cm for accurate dose calculations for
400 cm source to mid-line distance.
In any case, measurements are needed at the extended distance to confirm the TPS
calculation accuracy under the treatment conditions.
1.10.1 Commissioning of TBI TPS Calculations
Following the selection of a treatment technique, the commissioning work begins.
According to Van Dyk[14] the beam data collected for conventional radiotherapy has little
application in the TBI setting; a separate set of dosimetric data is required for TBI
geometry. For a better prediction of dose by the treatment planning system, absolute
beam output calibration, percentage depth doses and beam profiles should be acquired, at
the extended SSD.
There are a number of dosimetric problems to consider in the commissioning of a TBI
protocol, specific to large field dosimetry which do not arise in conventional radiotherapy.
These issues are associated with the phantoms and ionisation chambers used in the
acquisition of the dosimetric data. Dosimetric inaccuracies can arise from the irradiation
of the ionisation chamber cable as a result of the relatively large radiation fields and
chamber leakage currents become more significant at the low dose rates typical of TBI.
Charge collection efficiency of the ionisation chamber will be greater at the low dose rate
and perturbation corrections used in dose calculations can depend on the source-chamber
distance (SCD)[14].
24
The selection of the phantom sizes for which this data is acquired is often considered to
be critical however for large distances Podgorsak[70] found that for output factors the size
of the phantom made a limited difference, although for the depth dose distribution there
was a significant variation in percentage doses at depths of 15 cm or greater. This effect
decreases slightly as the beam energy increases, attributed to the increase in the more
forward directed scattering of photons and a reduction in the amount of scatter produced
for higher energies.
25
C h a p t e r 2
CONFIRMATION OF BEAM REMODELLING AND DATA REQUIRMENTS PHASE I
2 Initial investigation
2.1 Patient positioning
The current treatment position involves the patient in a semi-reclining position proposed
by Khan[36] with knees elevated to reduce the overall length of the patient in the beam’s
eye view. The new position proposed in this thesis is designed to be more comfortable for
the patients during the lengthy treatment.
A survey of a number of RBWH TBI patients indicated a strong preference for the supine
position for comfort, confirming one of the critical benefits of the change in the
treatment protocol.
The overall height of the patient is an obvious limiting factor with regard to treatment
position. Although supine may be more comfortable, it makes it more difficult to fit a
very tall patient within the treatment field. In the event that a patient does not fit within
the size of the field, they will be treated with their knees bent. This position will be
maintained during both simulation and treatment. A low density patient support device (a
pillow) can be used to assist the patient in maintaining the desired position during the scan
and long treatment time, which will ultimately assist in the reproducibility of the treatment
setup. Because of the air gap between the BUS and the patient surface, the beam and the
screen need to extend some distance beyond the head and feet of the patient to produce
adequate build-up of electron fluence in these regions.
Although 40 cm x 40 cm is the largest field possible, the MLC system is used to reduce
the size of the field at the extended SSD to approximately the same dimensions as the
BUS. The collimator angle is set to 45º to take advantage of the maximum horizontal field
dimension of 50 cm (at 100 cm SSD) along a diagonal. The high density frame of the BUS
26
is kept outside the field, as OMP does not allow the simulation of materials with a density
exceeding 2.8 g/cm3.
2.2 Dosimetric verification
The first step in assessing the ability of a planning system to accurately model the
complexities of a treatment plan is to assess its ability to accurately model a simple case.
Dosimetric verification of a planning system can be achieved in many ways, the simplest
of which is the point measurement within a simple homogeneous phantom. Ionisation
chambers are used to collect charge deposited at a particular depth and the charge
collected can be converted to absorbed dose in a specified medium at that depth. The
charge collected is corrected for the influence quantities temperature and pressure,
polarity and recombination. An assessment was made to ascertain the need to modify the
parameters of the beam model, most notably the energy fluence kernels in OMP, to better
predict the doses at the extended distance: this was done by comparing a series of results
produced by OMP to experimental data, using the beam model already used clinically for
conventional (non-TBI) treatments in the TPS. This model will be referred to as Model 0
in this thesis.
Analysis performed using the Gamma Index as discussed by Low[71] is commonly
employed as a measure of the relevant difference between two dose profiles. For regions
of low dose gradients the difference between profiles is best assessed as the dose value
difference, in contrast to regions of high dose gradients (such as penumbra) where the
difference is best assessed by the distance-to-agreement for a particular dose value. As a
measurement of agreement, at each point along a profile the gamma index is given as a
combination of the dose and positional differences. When the gamma value exceeds 1 the
deviation between the two profiles is no longer within the tolerance limit set. The Gamma
analysis software employed in this thesis was developed in-house (using Matlab,
Mathworks, Natick, Ma, USA) based on the Low[71] evaluation method.
For the chambers used, the recombination effects are small (of the order of 1% of the
charge collected) at standard treatment distances and dose rates. The predominant
recombination effect component is approximately proportional to the dose (to the air in
27
the chamber) per pulse of radiation. At the extended (~400 cm) TBI treatment distance,
the recombination effects are negligible as the dose per pulse is lower by more than an
order of magnitude due to the inverse square law. The distance from the source to the
back wall is 495 cm.
2.3 Small cubic phantom
A small 20 cm cubic solid water phantom consisting of Gammex RMI 457 (Gammex
RMI, Madison, WI) slabs was used for a preliminary set of measurements due to the
geometrical simplicity and ease of set-up it provided. It was placed in the centre of the 40
x 40 cm MLC modified field at 389.75 cm SSD (400 cm to the centre of the phantom,
congruent with the clinical set-up).
The phantom comprised slabs of solid water, and a Farmer Type 2571 (Nuclear
Enterprises, Reading, England) ion chamber (used because of its compatibility with the
phantom – see Figure 12) connected to a PTW Unidos Webline (PTW-Frieburg, Frieburg,
Germany) electrometer was placed with its axis at 1 cm depth in the solid water. The slices
were re-arranged to allow the placement of the chamber at other depths while keeping the
SSD constant.
Figure 12: The simple small cubic solid water phantom consisting of Gammex RMI 457 solid water slabs.
To check the effect of the BUS (to be used during treatment), the charge was collected for
a series of depths with and without the BUS in the beam. The distance from the source to
the BUS is maintained at 380 cm.
28
The charge was converted to dose to water by applying a calibration factor. This factor
with units of cGy/nC was determined by acquiring the charge collected for a known dose
in the solid water phantom at dmax for a 10 cm x 10 cm field, at 100 cm SSD. This
process assumes that the spectrum of the large beam at the extended SSD is sufficiently
similar to that of the reference beam at 100 cm SSD that the mass stopping power ratio of
water or solid water to air is essentially the same for both measurements. The justification
for this assumption is as follows: (i) The effect of the differential attenuation of the beam
spectral components, by the additional 3 m of air is approximately equivalent to that of 3
mm of water, which would be negligible. (ii) The large area beam would have slightly
lower average energy than the 10 cm x 10 cm reference beam due to the increased
proportion of collimator scatter, but the effect of this would be no greater than for
conventional treatments at ~ 1 m SSD, where it is considered sufficiently small that no
correction is made for it in ion chamber dosimetry. (iii) The build-up with depth of
phantom scatter (Compton-scattered photons which have lower average energy than the
primary photons) reduces the average energy of a beam, and this reduction will increase as
the phantom becomes larger and generates more scatter. However the small cubic
phantom described in this section is of a size where again no correction is made in
conventional ion chamber dosimetry. (iv) The stopping power ratio of water to air varies
slowly with the variation in average energy in a 10-MV beam, so even a significant change
in average photon energy may produce an insignificant change in the response of the
chamber. The graphs in appendix 3 of the IAEA TRS-398[72] show that a 10% change in
the beam quality specifies TPR20,10, centred on the value 0.7 (the approximate value for a
10-MV beam), produces only a 1.5% change in Sw,air which is the predominant factor
affecting chamber response.
The results of these dose measurements within the solid water phantom are indicated in
figure 13. For all graphs presented in this thesis the error bars are insignificantly greater
than the thickness of the lines and are omitted for clarity.
29
10 MV, 389.75 SSD, 400 MUVarying depth in solid water at centre of 40 cm x 40 cm MLC modified field
0
5
10
15
20
25
30
0 2 4 6 8 10 12 14 16 18
Depth in solid water
Do
se (
cGy)
No screen
With screen
Figure 13: Dose plotted as a function of depth (corrected for effective point of measurement) in a 20 cm
cubic solid water phantom, with and without the BUS.
The results shown in figure 13 confirm the suitability of the BUS for increasing the
entrance dose and demonstrate that for a small volume phantom an average reduction in
local dose of 2.8% beyond the depth of maximum dose can be expected due to this BUS.
This is equivalent to shifting the PDD by ~ 1 cm towards the source for points beyond
the build-up region.
In a full patient-sized phantom the entrance dose would be a little larger than in this small
phantom due to increased backscatter. The build-up effect of the screen in the same sized
large field would be the same with a large phantom, so these preliminary measurements
indicated that the screen performed as anticipated and was likely to be satisfactory. This
was verified by later measurements in the build-up region of a large phantom with the
screen present - see section 2.4.
2.4 PDD acquisition
In order to acquire the pertinent data to generate a PDD, two phantom systems had to be
employed, one a large volume MP3 (PTW-Frieburg, Frieburg, Germany) computer
controlled water phantom system shown in figure 14 for the majority of the
30
measurements and the other a composite phantom consisting of Plastic Water and solid
water shown in figure 15 (CIRS (Computerized Imaging Reference Systems Inc.,
Norfolk, VA, USA) Plastic Water™ for build up and Gammex RMI 457 solid water for
backscatter) was used to acquire data for depths less than 52 mm. The second phantom
was required because of the physical limitations of the first phantom, for which the
minimum depth of measurement in a lateral beam was equivalent to 52 mm of water. The
attenuation coefficient of the Plastic Water when compared with real water in a 10-MV
beam was measured as 0.995 and applied in the calculation of water-equivalent depths in
this phantom.
Figure 14: The PTW MP3 water tank used in the acquisition of the PDD data.
The PDD data was acquired in both phantoms (regions of overlap between the two were
used to confirm no variation in response as a function of scatter and relative transmission)
using the same parallel-plate ionisation chamber PPC 40 (IBA Dosimetry,
Schwarzenbruck, Germany) so as to exclude any individual chamber response variations.
The effective measuring point of the ionisation chamber (1 mm below the front surface of
the chamber) was used for the PDD measurements.
31
Figure 15: The phantom arrangement used to acquire the build-up region measurements for the PDD.
The distance between the BUS and the phantom surface (17.8 cm) was maintained in this
PDD acquisition and the effect of varying this distance was not investigated.
The entrance dose which can be a significant parameter in TBI treatment techniques
would only decrease quite slowly by increasing this distance because of the very large field
size. Additionally in the treatment technique proposed the patient will be positioned on
the bed as close as practicable to the proximal edge of the bed and hence to the screen
which abuts the bed and will be less than the distance of 17.8 cm to the phantom. ICRU
[50] recommends that the dose within the target volume (in TBI this includes the skin)
should not vary more than +7% and -5% of the prescribed dose, so an acceptable value
for the entrance dose is considered to be at least 95% of the target dose. The entrance
dose measured in this phantom is greater than 96%.
2.4.1 PDD for MLC-delineated Collimator 45°°°° beam
Two separate PDDs were acquired, both at the extended SSD, one 7 cm below the central
axis of the beam and a second displaced from the first 43 cm parallel to the gun-target
direction (towards the target). A PDD is recommended by Andreo et al[73] as a part of
the commissioning process to verify the correct modelling of the energy spectra in a
treatment planning system.
32
PDD LA3, 400 MU, 7 cm below of CAX, 380 cm source-tank distance, 17.8 cm screen-tank separation
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
110%
0 50 100 150 200 250 300 350 400 450 500 550
Depth (mm)
Per
cen
tage
Do
se
CAX
43 cm off axis
Figure 16: The PDDs acquired at the patient plane in a large water tank at 380 cm SSD using the field
delineated by the MLCs with a collimator angle of 45º. The plot labelled CAX was acquired 7 cm directly
below the beam central axis (CAX); the other plot was acquired 43 cm away from the CAX plot in the gun-
target direction.
As expected the beam is slightly less penetrating off axis as shown in figure 16. This
results from the lower initial energy and reduced beam hardening of the radiation further
off axis, (the latter due to reduction in the thickness of the flattening filter the off-axis
photons pass through), the obliquity of off-axis rays in the phantom and the decrease with
distance from the axis of the rate of build up of scatter with depth.
2.4.2 PDD comparison for MLC-delineated Collimator 45°°°° beam
In order to replicate the treatment geometry accurately an investigation was conducted to
assess the possibility of including the BUS as a region of interest to improve calculation
accuracy. An artificial CT image series was created to be sufficiently large to enable the
replication of the water tank within the image series in OMP. The BUS which has a
physical density of 1.19 g/cm3 had to be replicated in the CT data set as a bolus structure
(or region of interest).
Calculations were attempted however the treatment planning system was found to be
incapable of performing a calculation for a beam where the field size is set at 40 cm x 40
cm when the collimator is rotated away from 0º. This limitation of the planning system
33
was confirmed via correspondence with the manufacturer. Consequently any further
investigation of the large MLC-delineated collimator 45º beam was abandoned. Instead a
simple second treatment field was investigated, where the collimator is not rotated and the
field size is set to 40 cm x 25 cm.
Figure 17: Treatment plan configuration for the case of the PDD taken 7 cm directly below the beam axis in
a 40 cm x 25 cm field.
Figure 18: Treatment plan configuration for the case of the PDD taken 7 cm below the beam axis and 43 cm
towards the target in a 40 cm x 25 cm field.
34
2.4.3 PDD acquisition for Collimator 0°°°°, 40 cm x 25 cm field
Two separate PDDs were acquired at the extended SSD (380 cm to the tank) with the 40
cm x 25 cm field, the first 7 cm below the central axis and a second 43 cm from the first
at the same height.
PDD LA3, 400 MU, 7 cm below of CAX, 380 cm source-tank distance, 17.8 cm screen-tank separation
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
110%
0 50 100 150 200 250 300 350 400 450 500 550
Depth (mm)
Per
cen
tage
Do
se
CAX
43 cm off axis
Figure 19: The PDDs acquired at the patient plane in a large water tank at 380 cm SSD using the field
delineated by the jaws only with a collimator angle of 0º. The plot labelled CAX was acquired 7 cm directly
below the beam central axis (CAX); the other plot was acquired 43 cm away from the CAX plot in the gun-
target direction.
The ratio of doses between the two curves (PDD on CAX and PDD 43 cm off CAX)
with the MLC delineated field and the jaw defined field differ by only 0.4 %. The
relationship between the CAX and 43 cm off axis PDD curves for the 40 cm x 25 cm
field is similar to that observed for the field delineated by the MLC system with the
collimator rotated to 45º.
35
2.4.4 PDD comparison for Collimator 0°°°°, 40 cm x 25 cm field
2.4.4.1 Absence of build-up screen
PDD Comparision, 400 MU, 7 cm below axis, 380 cm source-tank distance, 17.8 cm screen-tank separation
0%
20%
40%
60%
80%
100%
120%
140%
0 10 20 30 40 50 60
Depth (cm)
Per
cen
tage
Do
se
Measured PBE no build-up screen CCE no build-up screen
Figure 20: The comparison of the PDDs measured 7 cm below the central axis and calculated by the
treatment planning system in the absence of the BUS using both the pencil beam and collapsed cone
algorithms, normalised to 10 cm depth.
IBUR ABUR User defined
Dist tol (mm) 1 1 1
Dose tol (%) 10 2 2
G_max 28.73 4.47 143.67
G_mean 2.95 3.14 4.64
Pass rate (%) 52.81 0.04 1.36
Table 1: The gamma analysis results comparing the measured PDD and the Pencil Beam algorithm PDD
with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-Up Region
(IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1
mm to 500 mm).
The maximum gamma value (G_max) represents the worst agreement between the data
sets within the regions compared; the mean gamma value (G_mean) represents the mean
gamma value within the region compared. A value greater than 1 represents a failure to
fall within the limits set. The Pencil Beam model has an unacceptably low pass rate
(~53%) and will not be used for clinical use.
36
IBUR ABUR User defined
Dist tol (mm) 1 1 1
Dose tol (%) 10 2 2
G_max 12.02 2.41 60.08
G_mean 2.20 1.52 1.64
Pass rate (%) 59.43 27.60 50.43
Table 2: The gamma analysis results comparing the measured PDD and the Collapsed Cone algorithm PDD
with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-Up Region
(IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined region (range 1
mm to 500 mm).
Although an improvement from the Pencil Beam model, the Collapsed Cone model too
has an unacceptably low pass rate (~28%) and will not be used for clinical use
PDD Comparision, 400 MU, 43 cm off axis, 380 cm source-tank distance, 17.8 cm screen-tank separation
0%
20%
40%
60%
80%
100%
120%
140%
0 10 20 30 40 50 60
Depth (cm)
Per
cen
tage
Do
se
Measured PBE no build-up screen CCE no build-up screen
Figure 21: The comparison of the PDDs measured 43 cm away from the central axis and calculated by the
treatment planning system in the absence of the BUS using both the pencil beam and collapsed cone
algorithms, normalised to 10 cm depth.
37
IBUR ABUR User defined
Dist tol (mm) 1 1 1
Dose tol (%) 10 2 2
G_max 29.57 6.27 76.94
G_mean 3.01 4.22 5.89
Pass rate (%) 52.81 0.04 1.16
Table 3: The gamma analysis results comparing the 43 cm off axis measured PDD and Pencil Beam
algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-
Up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined
region (range 1 mm to 500 mm).
IBUR ABUR User defined
Dist tol (mm) 1 1 1
Dose tol (%) 10 2 2
G_max 12.0164 2.4052 60.0822
G_mean 2.1984 1.5214 1.6428
Pass rate (%) 59.4306 27.6001 50.4299
Table 4: The gamma analysis results comparing the 43 cm off axis measured PDD and the Collapsed Cone
algorithm PDD with 10 % or 2 % and 1 mm distance to agreement. Regions of interest include the In Build-
up Region (IBUR), After Build-Up Region (ABUR) (range from dmax to 300 mm) and the user defined
region (range 1 mm to 500 mm).
From the above comparisons it is apparent that the data in the TPS based on data
collected at 100 cm SSD will be insufficient to accurately predict the doses at the ESSD by
using either PBE or CCE.
2.4.4.2 Bolus density settings
It was hypothesised that the OMP, operating on a model derived from measurements
made on the 10-MV Elekta (Elekta, Stockholm, Sweden) Precise beam at conventional
treatment distances, would be sufficiently accurate for TBI provided a structure replicating
the BUS is included in the calculation. This structure was included in the TPS plan as a 1
cm thick bolus region of interest (ROI). Although the physical density of this screen was
known, an investigation was conducted in order to assess if this structure could be
38
tweaked to force the simulated beam PDD to match the measured data. This was
performed by altering the density of the screen.
Percentage Depth Dose 7 cm below CAX, Measured data and Collpased Cone data comparision
0%
20%
40%
60%
80%
100%
120%
140%
0 5 10 15 20 25 30 35 40 45 50
Depth (cm)
Per
cen
tage
Do
se
Measured
Bolus 1 g/cm3
Bolus 1.2 g/cm3
Bolus 1.5 g/cm3
Figure 22: The comparison of the PDDs measured on the central axis and those produced by the treatment
planning system using the collapsed cone algorithm, normalised to a depth of 10 cm, with three different
density values applied.
Percentage Depth Dose 7 cm below CAX, Measured data and Pencil Beam data comparision
0%
20%
40%
60%
80%
100%
120%
140%
0 5 10 15 20 25 30 35 40 45 50
Depth (cm)
Per
cen
tage
Do
se
Measured
Bolus 1 g/cm3
Bolus 1.2 g/cm3
Bolus 1.5 g/cm3
Figure 23: The comparison of the PDDs measured on the central axis and those produced by the treatment
planning system using the pencil beam algorithm, normalised to a depth of 10 cm, with three different
density values applied
39
From the results presented in figures 22 and 23 above the maximum deviation appears to
be ~6 % of dMax. at ≥ 35 cm depth where the PDD is ~ 40%. This would correspond to
%100*140
6~
which is equal to ~ 4% deviation for a bilateral treatment, which could be considered
clinically acceptable in TBI.
Although the inclusion of the ROI replicating the BUS in treatment plans significantly
improved agreement with measured data, it was deemed too time consuming and in
certain clinical cases would prove to be unachievable, as OMP restricts the calculation of
dose to within the CT data set; any structure outside of the CT data is ignored. For the
case of large patients the inclusion of such a ROI would not be possible. The aim of this
project is to enable accurate calculations for all TBI patients at RBWH, not just a subset.
For both of the investigations described in this section, the current 10-MV model does
not predict the dose variation with depth correctly within the tolerance values set for each
investigation respectively, leading to the conclusion that an attempt to refine the method
of dose calculation is warranted. It is expected that generally the dose calculation should
be accurate to within ± 5 % of the measured data.
2.5 Acquisition of beam profile free in air
In order to investigate the overall limitation of the TPS in the cross beam direction at a
depth beyond the build-up an in-air profile was measured. A Farmer type ion chamber
fitted with a small Perspex build-up cap (5 cm in diameter) to produce charged particle
equilibrium was clamped to a retort stand. Charge measurements were acquired at the
height of the treatment beam isocentre, at various points across the 40 cm x 25 cm field.
40
10 MV, 400 cm SCD, 40 cm x 25 cm field, Y- Profile
3
3.5
4
4.5
5
5.5
6
6.5
7
7.5
8
-110 -90 -70 -50 -30 -10 10 30 50 70 90 110
Position on Y axis, at height of the beam axis
Ch
arge
co
llect
ed (
nC
) as
a f
un
ctio
n o
f M
U
Figure 24: The y-axis profile obtained in air with a build-up cap with 50% of the central dose occurring at
±80 cm.
The above investigation was conducted to check the symmetry and flatness in air at the
extended distance. The resultant y-axis profile was also expected to be useful in
interpreting the measured and TPS–calculated doses at off-axis points covering the length
of the patient. In the x-axis direction only a very limited range of the beam area will be
used, covering the anterior-posterior dimension of the patient, therefore this data was not
collected separately.
2.6 Comparison of in air-profiles
The phantom geometry (the build-up cap) was replicated in the treatment planning
system; the profiles produced by each calculation algorithm were extracted.
41
Profile Collapsed Cone, Longitudinal profile400 cm Source to chamber distance, 40 x 25 cm field
25%
35%
45%
55%
65%
75%
85%
95%
105%
115%
-100 -80 -60 -40 -20 0 20 40 60 80 100
Position (cm)
Per
cen
tage
do
se
Measured CCE Model 0 PBE Model 0
Figure 25: The y-axis profile normalised at the central axis position, calculated by the Collapsed Cone and
Pencil Beam algorithms as compared with that measured.
Gamma analysis was performed to quantify the differences between the profiles calculated
by the TPS and the measured profile:
Penumbra Penumbra Inner
Left Right
Dist tol (mm) 2 3 3
Dose tol (%) 2 3 3
G_max 1.47 1.28 1.25
G_mean 0.27 0.35 0.33
Pass rate (%) 96.53 94.19 98.81
Gamma norm type Locally Central axis Central axis
Table 5: The results of gamma analysis performed to compare the Pencil Beam calculation of the profile and
the measured data.
42
Penumbra Penumbra Inner
Left Right
Dist tol (mm) 2 3 3
Dose tol (%) 2 3 3
G_max 0.65 1.61 8.41
G_mean 0.16 0.29 0.54
Pass rate (%) 100.00 98.84 92.68
Gamma norm type Locally Central axis Central axis
Table 6: The results of gamma analysis performed to compare the Collapsed Cone calculation of the profile
and the measured data
The above comparison favours the collapsed cone calculation, however the beam profile
is not considered a major factor in selecting a suitable calculation beam model. The
differences are only significant near the edges of the field where the beam will in most
clinical cases overshoot the patient.
2.7 Cable signal
The inclusion of the ionisation chamber cable in the path of the beam can result in a
spurious component of the measurement which is known as cable signal. To quantify this
signal, the chamber was irradiated under reference conditions (100 cm SSD for a 10 cm x
10 cm field) with a standard length of cable within the path of the beam and the charge
recorded; this was repeated this time including an equivalent length of cable as would be
in the path of the beam for the measurements made at 400 cm SCD. The percentage
variation in signal was recorded as 0.8% for the approximate length of cable that would be
present in the treatment beam for the Farmer chamber in the simple solid water phantom
described in section 2.8. The cable effect constitutes a reduction of signal 0.8% for the
Farmer chamber and an increase in signal of 7% for the Pinpoint chamber (used in
following chapters).
2.8 Simple solid water phantom
A small solid water phantom similar to the phantom shown in figure 12 was also used in
the investigation of the variation of the midline dose as a function of separation (thickness
in the beam direction). This was performed to assess the system’s ability to correctly
predict the dose at the prescription point for a variety of separations. The phantom was
43
placed on the treatment bed at constant 400 cm source to chamber distance, with the
chamber reference point on the beam axis, and doses were measured at this point, while
the phantom thickness surrounding the chamber was increased.
The ion chamber readings were converted to dose for 400 MU delivered. This process
assumes that the chamber perturbation factors in solid water do not differ significantly
from those in water; which are accounted for by the kQ factors in the calibration chain†.
Depth (cm) Dose (cGy) Stdev uA
5 24.74 0.01 0.01
14 19.90 0.01 0.01
16 18.90 0.01 0.01
21 16.42 0.01 0.01
26 14.23 0.01 0.01
Table 7: The doses as measured by converting the charge collected in the simple solid water phantom to
dose, where the standard uncertainty (uA) is derived according to TRS398[72] as the standard deviation of
the mean.
The dose on the phantom side opposite to the beam incident surface (also known as the
exit dose) was looked at as part of a comprehensive dosimetry investigation (section 2.12).
2.9 Simple solid water phantom comparison
The simple solid water phantom geometry was replicated in the TPS, by creating an empty
image CT series and recreating the phantom as a region of interest with a defined density
of 1 g/cm3. The dose was calculated using both pencil beam (PBE) and collapsed cone
(CCE) calculation algorithms at the midpoint for 400 MU. This plan does not include the
BUS; as discussed previously the inclusion of this screen in the plan is not clinically
achievable.
† For more details see TRS-398[73]
44
Figure 26: The simple solid water phantom replicated in the TPS for the 14 cm separation arrangement.
The doses calculated by the TPS are compared with the doses measured in the solid water
phantom and the differences are found in table 8.
PBE CCE PBE – CCE
Depth Model 0 Model 0 Model 0
5 7.3% 4.8% 2.5%
14 8.6% 5.2% 2.4%
16 8.7% 5.8% 2.9%
21 9.1% 5.3% 3.8%
26 8.5% 6.6% 1.9%
Table 8: The comparison of doses calculated by OMP using the Model 0 data with the doses measured by
the Farmer chamber in the simple phantom. The tabulated values are the percentages by which the
calculated dose exceeds the measured dose.
From the above data although the percentage depth dose for this beam model is not
considered clinically acceptable for dose calculations, when calculating the dose for this
phantom the collapsed cone algorithm is sufficiently accurate; 2.8% of these differences
can be attributed to the absence of the screen in the calculations.
45
2.10 Thorax phantom data acquisition
A thorax phantom (Computerized Imaging Reference Systems Inc., Norfolk, VA, USA)
was used to verify the dose (to water) in tissue equivalent, bone equivalent and lung
equivalent materials, using a small volume ion chamber‡. This chamber was used because
the phantom chamber cavity had been milled for this particular chamber; a chamber with
a larger volume would have been preferred to improve the signal to noise ratio and reduce
the effect of the cable signal.
The thorax phantom was arranged on the treatment bed as pictured in Figures 27 and 28,
with the mid-sagittal plane of the phantom positioned 400 cm from the source (30 cm
from the surface of the BUS to the mid-sagittal plane) and the highest point on the
anterior surface of the phantom several centimetres below the beam central axis. The
distance from the wall to the closest side of the phantom was ~ 82 cm. The 40 cm x 25
cm beam was used to deliver 500 MU. The ion chamber was placed in cavities milled
within plugs which fit in the phantom at key points (as shown in figure 27). These plugs
are made from substances that replicate the physical properties of lung, bone and soft
tissue. Plugs for this phantom that do not contain chamber cavities are used where the
chamber is not required and are also employed in the simulation process, that is, when a
CT is obtained of the phantom the chamber is not present and so does not perturb the
average 7.0% average 14.5% average 16.0% average 16.0% average 14.2%min -13.3% min -7.3% min 8.5% min 8.5% min 6.3%max 15.4% max 23.5% max 23.5% max 23.5% max 25.0%
TLDsPBEModel 0 Model 1 Model 2 Model 3 Model 4
Table 41: The doses calculated using the PBE algorithm for each beam model, at each TLD position, and a
comparison with the doses measured by the TLDs. For each model, the first column shows the calculated
dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the
TLD-measured dose at that point for the same number of MU.
average 1.9% average 9.0% average 10.5% average 10.5% average 8.7%min -17.4% min -11.7% min 3.4% min 3.4% min 1.3%max 9.9% max 17.7% max 17.7% max 17.7% max 19.0%
TLDsPBEModel 0 Model 1 Model 2 Model 3 Model 4
Table 43: The doses calculated using the pencil beam algorithm for all models and the percentage differences
between the measured and calculated values. For each model, the first column shows the calculated dose in
cGy for 1000 MU, and the second column shows the percentage deviation of that dose from the revised
TLD-measured dose at that point for the same number of MU.
average -1.6% average 7.3% average 8.8% average 8.7%min -15.9% min -8.5% min 0.5% min 0.3%max 4.2% max 13.7% max 14.3% max 13.8%
TLDsCCEModel 0 Model 1 Model 2 Model 3
Table 44: The doses calculated using the Collapsed Cone algorithm for all models and the percentage
differences between the measured and calculated values. For each model, the first column shows the
calculated dose in cGy for 1000 MU, and the second column shows the percentage deviation of that dose
from the revised TLD-measured dose at that point for the same number of MU.
99
REFERENCES
1. Brown, J., et al., Increasing Incidence of Late Second Malignancies After Conditioning With Cyclophosphamide and Total-Body Irradiation and Autologous Bone Marrow Transplantation for Non-Hodgkin's Lymphoma. Journal of Clinical Oncology, 2005. 23(10): p. 2208-2214.
2. De Neve, W., M. Lybeert, and J. Meerwaldt, Low-Dose Total Body Irradiation in Non-Hodgkin Lymphoma: Short-and Long-term Toxicity and Prognostic Factors. American Journal of Clinical Oncology, 1990. 13(4): p. 280-284.
3. Mukai, H.Y., et al., Successful treatment of a patient with subcutaneous panniculitis-like T-cell lymphoma with high-dose chemotherapy and total body irradiation. European Journal of Haematology, 2003. 70(6): p. 413-416.
4. Clift, R.A., et al., Allogeneic Marrow Transplantation in Patients With Acute Myeloid Leukemia in First Remission: A Randomized Trial of Two Irradiation Regimens. Blood, 1990. 76: p. 1867-1871.
5. Clift, R.A., et al., Allogeneic Marrow Transplantation in Patients With Chronic Myeloid Leukemia in the Chronic Phase: A Randomized Trial of Two Irradiation Regimens. Blood, 1991. 77(8): p. 1660-1665.
6. Alexander, B.M., et al., Utility Of Cranial Boost In Addition To Total Body Irradiation In The Treatment Of High Risk Acute Lymphoblastic Leukemia. International Journal of Radiation Oncology Biology Physics, 2005. 63(4): p. 1191-1196.
7. Inagaki, J., et al., Bone marrow transplantation in children with severe aplastic anemia using a conditioning regimen containing 3 Gy of total body irradiation, cyclophosphamide with or without antithymocyte globulin. Pediatric Transplantation, 2007. 11(2): p. 180-186.
8. Van Dyk, J., et al., The physical aspects of total and half body photon irradiation: a report of Task Group 29, Radiation Therapy Committee Association of Physicists in Medicine. 1986: American Association of of Physicists in Medicine.
9. Harden, S.V., et al., Total body irradiation using a modified standing technique: a single institution 7 year experience. British Journal of Radiology, 2001. 74(887): p. 1041-1047.
10. Maynes, P., A. Nahum, and J.-C. Rosenwald, Handbook of Radiotherapy Physics: Theory and Practice. 2007, Boca Raton, FL, US: Taylor and Francis Publishing Group
11. Shank, B., Total Body Irradiation for Marrow or Stem-Cell Transplantation. Cancer Investigation, 1998. 16(6): p. 397 - 404.
12. Metcalfe, P., T. Kron, and P. Hoban, The physics of radiotherapy x-rays and electrons. 2007, Madison: Medical Physics Publishing. xix, 905 p. :.
13. Khan, F.M. and e. al, Treatment Planning in Radiation Oncology. Second ed, ed. F.M. Khan. 2007, Philadelphia, PA, USA: Lippincott Williams and Wilkins.
14. Van Dyk, J., The Modern Technology of Radiation Oncology. Vol. 1. 1999, Madison, Wisconsin, USA: Medical Physics Publishing.
100
15. Millar, M., et al., ACPSEM position paper recommendations for the safe use of external beams and sealed brachytherapy sources in radiation oncology. Australasian Physical and Engineering Sciences in Medicine, 1997. 20.
16. Mayles, W., et al., Physics aspects of quality control in radiotherapy. 1999: Institute of Physics and Engineering in Medicine.
17. Hui, S.K., et al., CT-based analysis of dose homogeneity in total body irradiation using lateral beam. Journal of Applied Clinical Medical Physics, 2004. 5(4): p. 71-79.
18. Lagrange, J.L., et al., CT Measurement of Lung Density: The Role of Patient Position and Value for Total Body Irradiation. International Journal of Radiation Oncology Biology Physics, 1987. 13(5): p. 941-944.
19. Hurkmans, C., et al., Limitations of a pencil beam approach to photon dose calculations in the head and neck region. Radiotherapy and Oncology, 1995. 37(1): p. 74-80.
20. Knöös, T., et al., Limitations of a pencil beam approach to photon dose calculations in lung tissue. Physics in Medicine and Biology, 1995. 40(9): p. 1411-1420.
21. Wieslander, E. and T. Knöös, Dose perturbation in the presence of metallic implants: treatment planning system versus Monte Carlo simulations. Physics in Medicine and Biology, 2003. 48(20): p. 3295-3305.
22. Knöös, T., et al., Comparison of dose calculation algorithms for treatment planning in external photon beam therapy for clinical situations. Physics in Medicine and Biology, 2006. 51(22): p. 5785-5807.
23. Mackie, T.R., et al., Generation of photon energy deposition kernels using the EGS Monte Carlo code. Physics in Medicine and Biology, 1988. 33(1): p. 1-20.
24. Wieslander, E. and T. Knöös, Monte Carlo based verification of a beam model used in a treatment planning system. Journal of Physics: Conference Series, 2008. 102: p. 012027.
25. Knöös, T., et al., The dosimetric verification of a pencil beam based treatment planning system. Physics in Medicine and Biology, 1994. 39(10): p. 1609-1628.
26. Aspradakis, M.M., et al., Experimental verification of convolution/superposition photon dose calculations for radiotherapy treatment planning. Physics in Medicine and Biology, 2003. 48(17): p. 2873-2893.
27. Aspradakis, M.M., H.M. McCallum, and N. Wilson, Dosimetric and treatment planning considerations for radiotherapy of the chest wall. The British Journal of Radiology, 2006. 79(946): p. 828-836.
28. Ahnesjö, A. and M.M. Aspradakis, Dose calculations for external photon beams in radiotherapy. Physics in Medicine and Biology, 1999. 44(11): p. R99-R155.
29. Fotina, I., et al., Advanced kernel methods vs. Monte Carlo-based dose calculation for high energy photon beams. Radiotherapy and Oncology, 2009. 93(3): p. 645-653.
30. Paelinck, L., et al., Experimental verification of lung dose with radiochromic film: comparison with Monte Carlo simulations and commercially available treatment planning systems. Physics in Medicine and Biology, 2005. 50(9): p. 2055-2069.
31. Irvine, C., et al., The Clinical Implications of the Collapsed Cone Planning Algorithm. Clinical Oncology, 2004. 16(2): p. 148-154.
32. Ahnesjö, A. and A. Trepp, Acquisition of the effective lateral energy fluence distribution for photon beam dose calculations by convolution models. Physics in Medicine and Biology, 1991. 36(7): p. 973.
33. Ahnesjö, A., M. Saxner, and A. Trepp, A pencil beam model for photon dose calculation. Medical Physics, 1992. 19(2): p. 263-273.
101
34. Ahnesjö, A., et al., Beam modeling and verification of a photon beam multisource model. Medical Physics, 2005. 32(6): p. 1722-1737.
35. Ahnesjö, A., Collapsed cone convolution of radiant energy for photon dose calculation in heterogeneous media. Medical Physics, 1989. 16(4): p. 577-592.
36. Khan, F.M., The Physics of Radiation Therapy. 3rd ed. 2003, Philadelphia, PA: Lippincott Williams & Wilkins.
37. Van Dyk, J., The Modern Technology of Radiation Oncology. Vol. 2. 2005, Madison: Medical Physics Publishing.
38. Knoll, G.F., Radiation Detection and Measurement. Third ed. 2000, USA: John Wiley & Sons.
39. Podgorsak, E.B. and e. al, Radiation Oncology Physics: A Handbook for Teachers And Students. 2005: International Atomic Energy Agency.
40. Duch, M.A., et al., Thermoluminescence dosimetry applied to in vivo dose measurements for total body irradiation techniques. Radiotherapy and Oncology, 1998. 47(3): p. 319-324.
42. Deeg, H.J., et al., Single Dose or Fractionated Total Body Irradiation and Autologous Marrow Transplantation in Dogs: Effects of Exposure Rate, Fraction Size, and Fractionation Interval on Acute and Delayed Toxicity. International Journal of Radiation Oncology Biology Physics, 1988. 15(3): p. 647-653.
43. Cosset, J.M., et al., Single dose versus hyperfractionated total body irradiation before allogeneic bone marrow transplantation: a non-randomized comparative study of 54 patients at the Institut Gustave-Roussy. Radiotherapy and Oncology, 1989. 15(2): p. 151-160.
44. Cosset, J.-M., et al., Single Dose Versus Fractionated Total Body Irradiation Before Bone Marrow Transplantation: Radiobiological and Clinical Considerations. International Journal of Radiation Oncology Biology Physics, 1994. 30(2): p. 477-492.
45. Cox, J.D. and K.K. Ang, Radiation Oncology: Rationale, Techniques, Results. Eighth ed. 2003, St Louis, USA: Mosby Inc.
46. Keane, T., J. Van Dyk, and W. Rider, Idiopathic interstitial pneumonia following bone marrow transplantation: the relationship with total body irradiation. International Journal of Radiation Oncology* Biology* Physics, 1981. 7(10): p. 1365-1370.
47. Van Dyk, J., et al., Radiation Pneumonitis Following Large Single Dose Irradiation: A Re-evaluation Based On Absolute Dose To Lung. International Journal of Radiation Oncology Biology Physics, 1981. 7(4): p. 461-467.
48. O'Donoghue, J.A., Fractionated versus low dose-rate total body irradiation.Radiobiological considerations in the selection of regimes. Radiotherapy and Oncology, 1986. 7(3): p. 241-247.
49. ICRU, ICRU Report 24: Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Rays in Radiotherapy Procedures. 1976, International Commission on Radiation Units and Measurements (ICRU).
50. ICRU, ICRU Report 50: Prescribing, Recording and Reporting Photon Beam Therapy. 1993, International Commission on Radiation Units and Measurements (ICRU).
51. Safwat, A., et al., The clinical, physical and radiobiological aspects of total body irradiation. European Journal of Cancer, 1995. 31(Supplement 6): p. S33-S33.
52. Humphries, S.M., et al., Comparison of super stuff and paraffin wax bolus in radiation therapy of irregular surfaces. Medical Dosimetry, 1996. 21(3): p. 155-157.
102
53. Pla, M., S.G. Chenery, and E.B. Podgorsak, Total body irradiation with a sweeping beam. International Journal of Radiation Oncology*Biology*Physics, 1983. 9(1): p. 83-89.
54. Quast, U., Total body irradiation--review of treatment techniquesin Europe. Radiotherapy and Oncology, 1987. 9(2): p. 91-106.
55. Breneman, J.C., et al., A technique for delivery of total body irradiation for bone marrow transplantation in adults and adolescents. International Journal of Radiation Oncology*Biology*Physics, 1990. 18(5): p. 1233-1236.
56. Niroomand-Rad, A., Physical aspects of total body irradiation of bone marrow transplant patients using 18 MV X rays. International Journal of Radiation Oncology*Biology*Physics, 1991. 20(3): p. 605-611.
57. Miralbell, R., et al., Can a total body irradiation technique be fast and reproducible? International Journal of Radiation Oncology*Biology*Physics, 1994. 29(5): p. 1167-1173.
58. Jones, D., et al., An isocentrically mounted stand for total body irradiation. British Journal of Radiology, 2000. 73(871): p. 776-779.
59. Umek, B., M. Zwitter, and H. Habic, Total body irradiation with translation method. Radiotherapy and Oncology, 1996. 38(3): p. 253-255.
60. Gerrard, G.E., et al., Toxicity and dosimetry of fractionated total body irradiation prior to allogeneic bone marrow transplantation using a straightforward radiotherapy technique. Clinical Oncology, 1998. 10(6): p. 379-383.
61. Sampath, S., T.E. Schultheiss, and J. Wong, Dose response and factors related to interstitial pneumonitis after bone marrow transplant. International Journal of Radiation Oncology*Biology*Physics, 2005. 63(3): p. 876-884.
62. Gerbi, B.J. and K.E. Dusenbery, Design specifications for a treatment stand used for total body photon irradiation with patients in a standing position. Medical Dosimetry, 1995. 20(1): p. 25-30.
63. Papiez, E., et al., A Novel Technique for Total Body Irradiation using Gantry Rotation and Beam Modulation. International Journal of Radiation Oncology*Biology*Physics, 2008. 72(1, Supplement 1): p. S677-S677.
64. La Macchia, N.A., et al., The Development of a Segmented Intensity Modulated Beam Technique for Total Body Irradiation Prior to Hematologic Stem Cell Transplantation. International Journal of Radiation Oncology*Biology*Physics, 2007. 69(3, Supplement 1): p. S540-S540.
65. Quast, U., Physical treatment planning of total-body irradiation: Patient translation and beam-zone method. Medical Physics, 1985. 12(5): p. 567-574.
66. Harden, S.V., et al., Total body irradiation using a modified standing technique: a single institution 7 year experience. British Journal of Radiology, 2001. 74: p. 1041- 1047.
67. Sánchez-Nieto, B., F. Sánchez-Doblado, and J.A. Terrón, A CT-aided PC-based physical treatment planning of TBI: a method for dose calculation. Radiotherapy and Oncology, 1997. 42(1): p. 77-85.
68. Houdek, P.V. and V.J. Pisciotta, A comparison of calculated and measured data for total body irradiation by 10 MV x-rays. Physics in Medicine and Biology, 1987. 32(9): p. 1101-1108.
69. Curran, W.J., J.M. Galvin, and G.J. D'Angio, A simple dose calculation method for total body photon irradiation. International Journal of Radiation Oncology*Biology*Physics, 1989. 17(1): p. 219-224.
103
70. Podgorsak, E.B., et al., The influence of phantom size on output, peak scatter factor, and percentage depth dose in large-field photon irradiation. Medical Physics, 1985. 12(5): p. 639-645.
71. Low, D.A., et al., A technique for the quantitative evaluation of dose distributions. Medical Physics, 1998. 25(5): p. 656-661.
72. Andreo, P., et al., TRS 398: Absorbed Dose Determination in External Beam Radiotherapy: An International Code of Practice for Dosimetry based on Standards of Absorbed Dose to Water. 2004, International Atomic Energy Agency (IAEA): Vienna, Austria.
73. Andreo, P., et al., TRS 430: Commissioning and quality assurance of computerized planning systems for radiation treatment of cancer. 2004, International Atomic Energy Agency (IAEA): Vienna, Austria.
74. Poole, C., Using Monte Carlo simulation to predict dose at extended source to surface distance, in School of Physical and Chemical Sciences. 2008, Queensland University of Technology: Brisbane.