tg-43: formalism for brachytherapy

Brachytherapy Dose Calculation Formalism, Dataset Evaluation, and Treatment PlanningDataset Evaluation, and Treatment Planning

System Implementation

Mark J. Rivard, Ph.D. and Christopher Melhus, Ph.D.Tufts University School of MedicineTufts University School of Medicine

Jeffrey F. Williamson, Ph.D.Vi i i C lth U i itVirginia Commonwealth University

AAPM Summer School: Clinical Dosimetry Measurements in Radiotherapy23 June 2009 at Colorado College, CO

Full Disclosure

As related to the contents of this presentation,

Rivard has received research support from:

Varian Medical Systems and IsoRay Medical

Williamson has received research support from:

Varian Medical Systems and Philips Medical Systems

Learning Objectives / Purpose

l i t b h th d l l ti f li• explain current brachytherapy dose calculation formalism

• dosimetry parameter evaluation methodsdosimetry parameter evaluation methods

• consensus data formulation

• clinical implementation

• background on AAPM advances

Low-Energy Brachytherapy Sources

0.8 mm

4.1 mm

4.5 mm

Laser Welded Ends(0.1 mm wall)

Inorganic Substrate w/Cs-131 AttachedGold X-Ray Marker (0.25 mm diameter)

Titanium Case (0.05 mm wall)

4.0 mm

4.5 mm

IsoRay model CS-1 Rev2

High-Energy Brachytherapy Sources

Brachytherapy Dose Calculation Geometry

2004 AAPM TG-43U1Brachytherapy Dosimetry Formalism (2-D)

LG r,

LK L

L 0 0

,D r, S g r F r,

G r ,

dose rate to water in water at point P(r,S air kerma strength D r,

SK air kerma strength dose rate constant( ) di l d f tigL(r) radial dose function

GL(r,) geometry function (line source approximation)F(r,) 2-D anisotropy function

2004 AAPM TG-43U1Brachytherapy Dosimetry Formalism (1-D)

0( , )LG r 0

0 0

( , ) (r) ( )

( , )L

K L anL

G rD(r) = S g r

G r

dose rate to water at point P(rS air kerma strength

D(r)

SK air kerma strength dose rate constant( ) di l d f tigL(r) radial dose function

GL(r,) geometry function (line source approximation)( ) 1-D anisotropy function( )an r

Comparison of 1-D Formalisms( )G r BAD 0

P0 0

( , )g (r) ( )

( , )L

K anL

G rD(r) = S r

G r

BAD 2

0K L

rD r S g r ( r ) BAD

2

K L anD r S g r ( r )r

GOOD 2

0K P an

rD r S g r ( r )

r

BEST 0( , ) (r) ( )LG rD(r)= S g r

r

BEST 0

0 0

(r) ( )( , )K L an

L

D(r) = S g rG r

Intent of 2004 AAPM TG-43U1

• provide a revised definition of air-kerma strength;

• eliminate apparent activity for specification of source strength;• eliminate apparent activity for specification of source strength;

• eliminate the anisotropy constant in favor of the distance dependent 1-D anisotropy function;

• provide guidance on extrapolating tabulated TG 43 parameters• provide guidance on extrapolating tabulated TG-43 parameters to longer and shorter distances; and

• eliminate minor inconsistencies and omissions in the original protocol and its implementation.

• literature review of experimental and Monte Carlo Consensus Dataset Formulation Methodology

dosimetry results for brachytherapy sources• comparisons of all candidate datasets• average MC and average EXP from literature

CON = (MC + EXP)CON (MC EXP )2

• g(r) and F(r ) candidate datasets transformed using• g(r) and F(r,) candidate datasets transformed using common L, possibly with Leff = S × N

• g(r) and F(r ) typically taken from Monte Carlo• g(r) and F(r,) typically taken from Monte Carlo• (r) calculated from consensus F(r,dataset

fi l lt t b l t d ith h• final results tabulated with common mesh

Example Dosimetry Parameter Dataset

prefer AAPM-recommended datasetsotherwise use AAPM/RPC Registry data and original pubswebsites (ESTRO, Univ. Carleton, etc) also post datasets

Dosimetry Parameter Data Trail Part 1

2( )S K d d

Revised Air Kerma Strength Definition

( )KS K d dKS air kerma strength

air kerma rate in vacuo at specification

l h t t ff i l d d

( )K d air kerma rate in vacuo at specificationpoint d with energy cutoff , typically 5 keV

low-energy photon cutoff includedmeasurement conditions specified

NIST WAFACNIST WAFAC

Seed Calibrations at NIST + ADCLs

NIST WAFAC Bx source HDR-1000+

Correction of Errors and Inconsistencies

• g(r) presented for dimensionless units, consistency with investigator g(r), and 5th order polynomial

• explicit contraindication for erroneous 1-D equation

( )G r 0L

0 0

( , ) g (r) ( )

( , )P

K anP

G rD(r) = S r

G r

• elimination of Aapp and anisotropy constant

• methodology to extrapolate dose calculations for large and small distances

Need for Uncertainty Analyses

Recent Monte Carlo uncertainty analysis for Cs-131Rivard, Med. Phys. 34, 754-762 (2007)

Removal of Antiquated Terms

apparent activity: Aapp anisotropy constant: anchoice of ()x may lead to dosimetric errors cannot accurately reproduce dose r < 1 cmAAPM specifies SK for source strength errors result under specific circumstances

Recommendations to Dosimetry Investigators

• experimental measurement descriptors– description of internal and external source geometry– source irradiation geometry, orientation, irradiation timeline– detector calib. technique & energy response function, E(r)

di ti d t t d d t t– radiation detector and readout system– measurement phantom composition– phantom dimensions and use of backscatterphantom dimensions and use of backscatter– estimation volume averaging effect at all detector positions– # of repeated readings with standard deviation, # of sourcesp g ,– NIST SK value and uncertainty for measured source– uncertainty analysis section (statistical and systematic)


• Monte Carlo calculation descriptors– radiation transport code, version, and major options– cross-section library name, version, and customizations– radiation spectrum of source

i hi h d t t d i k t th– manner in which dose-to-water and air-kerma strength are calculated (i.e., tally used)

– source geometry, phantom geometry, and sampling spaceg y, p g y, p g p– composition and mass density of materials in the source– composition and mass density of materials in the phantom– physical distribution of radioisotope within the source– uncertainty analysis section (statistical and systematic)


• Monte Carlo recommended good practices– primary calculations in 30 cm diameter liquid water phantom,

ith t l t 5 f b k tt t i lwith at least 5 cm of backscatter material– use sufficient histories to limit statistical uncertainty

1 2 % for r < 5 cm in water; 1 1 % for s calculations1 2 % for r < 5 cm in water; 1 1 % for sK calculations– modern cross-section libraries should be used – verify manufacturer’s source dimensionsy– Volume averaging effects should be limited to < 1 %– model k(d) as a function of polar angle for sK simulation– point source modeling is unacceptable– mechanical mobility of internal source structures

2004 AAPM TG-43U1 Includes Reference Data

NIST-specified source spectra, half-lives, and atomic composition for air and water

MBq–1MBq–1

2008 AAPM Calibration Recommendations

Medical PhysicsThird-party brachytherapy source calibrations and physicistresponsibilities: Report of the AAPM Low Energy BrachytherapySource Calibration Working Group

Compiling and clarifying recommendations established by previous AAPMTask Groups 40, 56, and 64 were among the working group’s charges,which also included the role of third-party handlers to perform loading andassay of sources. This document presents working group findings on theresponsibilities of the institutional medical physicist and a clarification ofthe existing AAPM recommendations in the assay of brachytherapy

Th AAPM l it t th di ti f th i tit ti l di lsources. The AAPM leaves it to the discretion of the institutional medicalphysicist whether the manufacturer’s or institutional physicist’s measuredvalue should be used in performing dosimetry calculations.

© 2008 American Association of Physicists in Medicine

Butler et al., Med. Phys. 35, 3860-5 (2008)

2008 AAPM Calibration Recommendations

Systematic Approach to RTP QA

Example datasets (from TG-43UXX or published literature)

Methods for data entry (HDR & LDR)

Check RTP for list of key functions

Tests of all RTP field entries (SK linearity, test of decay, dose superposition, shield attn factor for HDR, etc)

Impact of coarse data grid (interpolation technique & errors)

QA comparison of hand calc with RTP point dosesQA comparison of hand calc with RTP point doses

Compare printed/plotted isodose images

Documentation (use e-records)

AAPM Therapy QA Guidance Documents

TG-40: General Radiotherapy QA

TG-43: Bx Dosimetry Parameters & Dose Calc. Algorithm

TG-53: RTP QA

TG-56: Bx Code of PracticeTG-59: HDR Tx DeliveryTG-64: LDR Prostate Tx DeliveryThese reports focus on treatment-related QAThese reports focus on treatment-related QA

AAPM TG-40 Guidance

Source strength decay of inventory

Global scaling of dose rate with SK and

Test source localization tools over full range of clinical use

Accuracy and implementation of shield attenuation factor

I t f id iImpact of grid size

AAPM TG-43 Guidance

Dosimetry parameters: SK and gL(r), F(r,), and (r)taken from consensus datasetstaken from consensus datasets

Parameter units: r [cm] and [degrees] except for [radians][ ] [ g ] p [ ]ensure consistency with RTP

Dosimetry algorithm:

LK L

G r,D r S g r F r

K L

L 0 0D r, S g r F r,

G r ,

AAPM TG-53 Guidance

RTP specs and ATP

Hardcopy printout accuracy

Temporal dose rate correlation with radionuclide decay

Accuracy and consistency of parameter units

C t t t d itComputer systems storage and security

AAPM Committee Structure

Chairs and LiaisonsBrachytherapy Subcommittee (BTSC) Mark RivardBrachytherapy Subcommittee (BTSC), Mark Rivard

Special Brachytherapy Modalities Working Group (SBM), Bruce ThomadsenHigh-Energy Brachytherapy Source Dosimetry Working Group (HEBD), Jose Perez-CalatayudRobotic Brachytherapy Working Group (RoboBx), Bruce Thomadsen and Yan YuLow-Energy Brachytherapy Source Dosimetry Working Group (LEBD), Michael Mitchgy y y y g ( )Brachytherapy Source Registry Working Group (BSR), Geoff Ibbott

TG-129 Eye Plaque Dosimetry, Sou-Tung Chiu-TsaoTG-143 Dosimetry for Elongated Brachytherapy Sources, Ali MeigooniTG-144 Dosimetry & QA Procedures for 90Y Microsphere Brachytherapy for Liver Cancer, Andy DezarnTG 137 P t t C D P i ti R d ti R i d N thTG-137 Prostate Cancer Dose Prescription Recommendations, Ravinder NathTG-138 Brachytherapy Dose Evaluation Uncertainties, Larry DeWerdTG-167 Investigational Brachytherapy Source Recommendations, Ravinder NathTG-182 Electronic Brachytherapy Quality Management, Bruce ThomadsenTG-186 Model-Based Dose Calculation Techniques for Advanced Dosimetry, Luc Beaulieuq y,

American Brachytherapy Society (ABS), Zoubir OuhibEuropean Society for Therapeutic Radiology & Oncology (ESTRO), Jack Venselaar

AAPM Plan to Address Dosimetry Advances

TG-186 charged to review– next-generation dose calculation algorithms– studies evaluating advanced algorithms for

• phantom size effect• inter-seed attenuation• material heterogeneities within the body• interface and shielded applicators• interface and shielded applicators• directional brachytherapy (azimuthal asymmetry)

– commissioning issuescommissioning issues– patient-related input data– potential clinical issue risks and limitationspotential clinical issue, risks, and limitations

Recently Published Vision 20/20 Paper

M di l Ph iMedical PhysicsThe evolution of brachytherapy treatment planning

Mark Rivard,1 Jack L. M. Venselaar,2 and Luc Beaulieu3

1Department of Radiation Oncology, Tufts University School of Medicine, Boston, Massachusetts, USA2Department of Medical Physics, Instituut Verbeeten, P.O. Box 90120, 5000 LA Tilburg, The Netherlands3Département de Radio-Oncologie et Centre de Recherche en Cancérologie de l’Université Laval, Quebecp g g , Q

Brachytherapy is a mature treatment modality that has benefited from technological advances.Treatment planning has advanced from simple lookup tables to complex, computer-based dosecalculation algorithms. The current approach is based on the AAPM TG-43 formalism with recentg ppadvances in acquiring single-source dose distributions. However, this formalism has clinicallyrelevant limitations for calculating patient dose. Dose-calculation algorithms are being developedbased on Monte Carlo methods, collapsed cone, and the linear Boltzmann transport equation. Inaddition to improved dose-calculation tools, planning systems and brachytherapy treatmentp , p g y y pyplanning will account for material heterogeneities, scatter conditions, radiobiology, and imageguidance. The AAPM, ESTRO, and other professional societies are coordinating clinicalintegration of these advancements. This Vision 20/20 article provides insight on these endeavors.Med. Phys. 36, 2136-2153 (2009)Med. Phys. 36, 2136 2153 (2009)

Sensitivity of Anatomic Sites to Dosimetric Limitations of Current Planning Systems

anatomic site photon energy

absorbed dose attenuation shielding scattering beta/kerma

dose

prostatehigh

prostatelow XXX XXX XXX

breasthigh XXXlow XXX XXX XXX

GYNhigh XXXlow XXX XXX

skinhigh XXX XXX

skinlow XXX XXX XXX

lunghigh XXX XXXlow XXX XXX XXX

penishigh XXXlow XXX XXXhigh XXX XXX XXX

eyelow XXX XXX XXX XXXRivard, Venselaar, Beaulieu, Med Phys 36, 2136-2153 (2009)

Photon Attenuation and Absorption

Rivard, Venselaar, Beaulieu, Med Phys 36, 2136-2153 (2009)

High-Z Photon Attenuation

Rivard, Venselaar, Beaulieu, Med Phys 36, 2136-2153 (2009)

Phantom Size and 192Ir Photon Scattering

Melhus and Rivard, Med Phys 33, 1729-1737 (2006)

Conventional TPS Fails to Accurately Calculate Brachytherapy Dose

air water?

tissue water?

contrast impact?contrast impact?

source superposition?

source shielding?source shielding?

radiation scatter?

Is the Future Monte Carlo-based TPS?

Medical PhysicsAn approach to using conventional brachytherapy software for clinical treatment l i f l M t C l b d b h th d di t ib tiplanning of complex, Monte Carlo-based brachytherapy dose distributions

Mark Rivard,1 Chris Melhus,1 Domingo Granero,2 Jose Perez-Calatayud,2 Facundo Ballester3

1Department of Radiation Oncology, Tufts Medical Center, Boston, Massachusetts 021112Radiation Oncology Department, “La Fe” University Hospital, Valencia, Spain3Department of Atomic, Molecular, and Nuclear Physics, University of Valencia, Spain

Med. Phys. 36, 1968-1975 (2009)

Summary

• current dose calculation formalism presented

d i t t l ti d• dosimetry parameter evaluation and consensus methods for uniform datasets derivation

li i l i l t ti b d AAPM t• clinical implementation based on AAPM reports

• AAPM current status and areas for future advancement