Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology, Warsaw, Poland 8th ECMP, Athens, 11-13.09.2014 Dosimetry audits in radiotherapy

Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology, Warsaw, Poland

8th ECMP, Athens, 11-13.09.2014

Dosimetry audits in radiotherapy in Poland

W. Bulski, K. Chełmiński, J. RostkowskaMedical Physics Department

Bielsko--Biała

Rzeszów

Kielce

Katowice

Olsztyn

Bydgoszcz

Szczecin

Zielona Góra

Poznań

Łódź

Warszaw a

Gdynia

Gdańsk

Białystok

Lublin

Brzozów

Kraków

W rocław

Opole

Gliwice

Słupsk

Tarnów

Wałbrzych

Radiotherapy centres in Poland as of 2013:30 radiotherapy centres

0102030405060708090

100110120130

1985 1990 1995 2000 2005 2010 2015

lata

liczb

a

Co-60

linacs

total

Radiotherapy equipment in Poland as of 2013:About 115 accelerators in 30 centres

For over 20 years a postal audit (IAEA methodology) has been carried out to check the TPS dose calculations in a homogenous

phantom (water) in reference and variousnon-reference conditions.

SAD

d=10 cm

TLD

SSD

d=10 cm

TLD

Non-reference conditions,Non-reference conditions,on axison axis, , MLCMLC shaped fields shaped fields

reference (1)

„small” (2)

„irregular” (5)

„inverted Y” (4)

„circular” (3) „irregular”+wedge (6)

MLC shaped photon beams from linear accelerators; dose calculated with TPS vs. measured with TLD. D(stated) - dose stated by the participant,

D(TLD) - dose determined by the SSDL

Results of the postal audit in 2012:

0,90

0,95

1,00

1,05

1,10

0 10 20 30

RT centre code

D(T

LD

) / D

(sta

ted

)

10x10

small

circular

inverted Y

irregular

irregular with wedge

In the above example all results were within the ± 5% level. Only 10 out of 168 results (6%) were over ±3.5% level.

This allowed us for adoption of the levels of the evaluation of the results:acceptance level ± 3,5%, and intervention level ± 5%.

Such levels are adopted by the IAEA only in the case of Secondary Standard Dosimetry Laboratories (SSDL).

However, all the measurements were done in a homogenous (water) phantom.

In the radiotherapy Treatment Planning Systems (TPS) various calculation algorithms are used and they accuracy of dose calculations in heterogenous medium has to be verified.

For this reason a heterogeneous cubic-shape phantom has been designed within a Coordinated Research Project of the IAEA.

Materials and methods• The heterogeneous phantom was developed in the

frame of an IAEA Coordinated Research Project. • The phantom consists of frame made with

polystyrene and of bone and lung inhomogeneity slabs.

• Special inserts allow to position TLD capsules within the polystyrene below the bone or lung material and also within the lung equivalent material.

• There are also inserts for positioning an ionization chamber.

• The comparisons were performed for a number of various TPS and for a number of various linear accelerators in radiotherapy departments in Poland.

15 cm

15

cm

10

cm

2 cm

2 cm

5 cm

2 cm

2 cm

2 cm

15 cm

15

cm

5 cm

2 cm

2 cm

2 cm

2 cm

2 cmBone

15 cm

2 cm

2 cm

2 cm

2 cm

2 cm

10

cm

15

cm 5 cmLung

15 cm15 cm

15 cm

10

cm

15 cm

15

cm

10

cm

2 cm

2 cm

5 cm

2 cm

2 cm

2 cm

15 cm

15

cm

10

cm

2 cm

2 cm

2 cm

2 cm

5 cm

2 cm

2 cm

2 cm

15 cm

15

cm

5 cm

2 cm

2 cm

2 cm

2 cm

2 cmBone

15 cm

2 cm

2 cm

2 cm

2 cm

2 cm

10

cm

15

cm 5 cmLung

15 cm15 cm

15 cm

10

cm

The phantom consists of a frame made of polystyrene and inhomogeneitis - bone or lung equivalent exchangeable slabs.

polystyrene

bone

lung

IAEA heterogeneity phantom with cassettes for TLD or films and ionization chamber cavity inserts.

Beam Radiationunit

TLD set #

User stated (TPS)

dose [Gy]

IAEA (measured)dose [Gy]*

deviation relative**to IAEA

mean dose [%]

IAEA mean dose

/User stated

dose

6 MV Clinac 2300CD

P(Polystyrene)

2,00 1,99 0,7 0,99

BP(Bone)

2,00 1,99 0,3 1,00

LP(Lung on-axis)

2,00 2,04 -1,6 1,02

LL(Lung off-axis)

2,28 2,22 2,9 0,97

Pencil beam algorithm, X6 MV, Warsaw, Poland

EXAMPLE RESULTS OF IAEA/WHO TLD POSTAL DOSIMETRIC QUALITY AUDIT

IAEA – HETEROGENEITY PHANTOM AUDITRESULTS FROM PILOT STUDY IN POLAND

PERFORMED BY SSDL-WARSAW

• Irradiaton of TLD capsules• Ionizing chamber measurements using dedicated inlet

Irradiation conditions (same for all detectors):• beam quality: 6MV• field size: 6 cm x 6 cm• SSD = 90 cm

Ten radiotherapy centers (of 30 total in Poland), six TPS types with alternative algorithms were examined giving 15 TPS/Algorithm/Linac combinations

No. TPS Algorithm Linac

1 Panther 5.01 EP Siemens Artiste

2 PrecisePlan 2.16 PBC Elekta Synergy

3 Eclipse 7.3 PBC Varian Clinac 2300

4 MasterPlan 4 PBC Varian Clinac 2300

5 Eclipse 8.2 PBC Varian Clinac 2300

6 Panther 5.01 CCC Siemens Artiste

7 CMS XiO 4.62 CCC Elekta Synergy

8 MasterPlan 4 CCC Siemens Primus

9 MasterPlan 4 CCC Varian Clinac 2300

10 MasterPlan 4 CCC Elekta Synergy

11 Eclipse 8.2 AAA Varian Clinac 2300

12 Eclipse 10.0 AAA Varian Clinac EX-S



15 Monaco 3.2 MC Elekta Synergy

PBC – Pencil Beam Convolution

AAA – Analytical Anisotropic Algorithm

MC – Monte-Carlo

EP – fast photon effective path

CCC – Collapsed Cone Convolution)

TLD – POLYSTYRENE

The TLD capsules were located in the polystyrene material at 10 cm depth. The plan was normalized to 100% for 2 Gy at 10 cm depth..

TLD – BONE

The TLD capsules were located under the bone tissue equivalent material at 10 cm depth. The plan was normalized to 100% for 2 Gy at 10 cm depth.

TLD – LUNG

In case of lung tissue equivalent material the TLD capsules were located in two positions:

in lung tissue, under lung tissue slab at 10

cm depth.

The plan was normalized to 100% for 2 Gy in the point at 10 cm depth.

Results

• Ten Polish radiotherapy centers (of 30 in total) were audited. • Six different TPSs and eleven calculation algorithms were

examined. • Generally, most of the results from TLD measurements were

within 5% tolerance. • Differences between doses calculated by TPSs and measured

with TLD did not exceed 4% for bone and polystyrene equivalent materials.

• Under the lung equivalent material, on the beam axis the differences were lower than 5%, whereas within the lung and off the beam axis – in some cases were around 7%.

• For algorithms which use point kernel convolution/superposition and density variation in 3D with modeling of lateral electron and photon transport (CCC, AAA) the calculated doses were usually underestimated compared to the TLD measurements.

National audit system of TPS in Poland within the IAEA pogramme with the CIRS

Phantom

Ten radiotherapy centres , seven TPS systems with various algorythms. Test cases were planned and measured for 31 combinations of beams of various energies and various calculation algorythms.

IAEA SUPPORTED AUDIT OF TPS IN POLAND

CC – Collapsed ConeAAA –Anisotropic Analytic AlgorithmPBC – Pencil Beam ConvolutionFPH – Fast Photon AlgorithmFFT – Fast Fourier Transform Convolution

31 combination of TPS x algorithm x beam energy were tested in 10 audited centers

0% 20% 40% 60% 80% 100%

CC4

FFT4

PBC4

AAA 6MV

AAA 6MV

AAA 6MV

CC 6MV

CC 6MV

CC 6MV

CC 6MV

CC 6MV

FFT 6MV

PBC 6MV

PBC 6MV

PBC 6MV

PBC 6MV

FPH 6MV

AAA 15MV

AAA 15MV

AAA 20MV

CC 15MV

CC 15MV

CC 15MV

CC 15MV

CC 15MV

FFT 15MV

PBC 15MV

PBC 15MV

PBC 15MV

PBC 15MV

FPH 15MV

TP

S c

alc

ula

tio

n a

lgo

rith

m, b

ea

m e

ne

rgy

% OF CASES PASSING THE TESTS


CC – Collapsed ConeAAA –Anisotropic Analytic AlgorithmPBC – Pencil Beam Convolutionother:FPH – Fast Photon AlgorithmFFT – Fast Fourier Transform Convolution

31 combination of TPS x algorithm x beam energy were tested in 7 audited centers

0%

20%

40%

60%

80%

100%

CC AAA PBC other

TPS calculation algorithm

% O

F R

ES

UL

TS

PA

SS

ING

TH

E

TE

ST

S

4-10MV

15-20 MV

0%

5%

10%

15%

20%

25%

30%

35%

40%

CC AAA PBC other

TPS calculation algorithm

FR

AC

TIO

N O

F R

ES

UL

TS

OU

TS

IDE

A

CC

EP

TA

NC

E

4-10MV

15-20 MV


CC – Collapsed ConeAAA –Anisotropic Analytic AlgorithmPBC – Pencil Beam Convolutionother:FPH – Fast Photon AlgorithmFFT – Fast Fourier Transform Convolution

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Case 1 Water Point 3 (0°)

Case1 Lung Point 9 (0°)

Case1 Bone Point 10 (0°)







Case 4 Water Point 5 SUM

Case 4 Lung Point 6 (0°)




Case 4 Lung Point 6 SUM

Case 4 Bone Point 10 (0°)




Case 4 Bone Point 10 SUM










Case 8 Water Point 5 (30° Table 90°)

Case 8 Water Point 5 (G90)

Case 8, Water, Point 5, (Gantry 270°)


tes

t

% of Audits PASS


0

1

2

3

4

5

6

7

8

9

10

100% 80% 60%

%PASS

Nu

mb

er

of

au

dit

ed

co

mb

ina

tio

ns

o

f b

ea

m, T

PS

, alg

ori

thm

Conclusions

• Over the last 20 years the postal TLD audits fulfilled their role and remain the primary and well established dosimetric audit method;

• The measurements allow to the detect limitations of TPS calculation algorithms. The audits performed with the use of the IAEA heterogeneous phantom seem to be an effective tool for detecting errors in radiotherapy procedures.

• The variety of new sophisticated heterogenous phantoms allow for thorough testing of TPS performance, for detecting computing algorithm’s limitations and corrections in treatment planning procedures.

• New irradiation modalities (tomotherapy, CyberKnife, etc.) require specific methods of dosimetric audits which are now being elaborated in a number of radiotherapy centres.

Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology, Warsaw, Poland

8th ECMP, Athens, 11-13.09.2014

Thank you for your attention

W. Bulski, K. Chełmiński, J. RostkowskaMedical Physics Department

Documents

Maria Skłodowska-Curie Memorial Cancer Centre and Institute of Oncology, Warsaw, Poland 8th ECMP, Athens, 11-13.09.2014 Dosimetry audits in radiotherapy