IAEA-TECDOC-896
XA9642841
Radiation dose in radiotherapy from prescription deliveryto
INTERNATIONAL ATOMIC ENERGY AGENCY The originating Sectio f thino s publicatio IAEe th Ann i was : Dosimetry Section International Atomic Energy Agency Wagramerstrasse5 0 10 x P.OBo . A-1400 Vienna, Austria
RADIATION DOSE IN RADIOTHERAPY FROM PRESCRIPTION TO DELIVERY IAEA, VIENNA, 1996 IAEA-TECDOC-896 ISSN 1011-4289 IAEA© , 1996 Printed by the IAEA in Austria August 1996 The IAEA does not normally maintain stocks of reports in this series. However, microfiche copie f thesso e reportobtainee b n sca d from
IN IS Clearinghouse International Atomic Energy Agency Wagramerstrasse 5 P.O. Box 100 A-1400 Vienna, Austria
Orders shoul accompaniee db prepaymeny db f Austriao t n Schillings 100,- for e for e chequ a th f m th IAE mf o n o i n i r Aeo microfiche service coupons which may be ordered separately from the INIS Clearinghouse. FOREWORD
Cancer incidenc increasins ei developen gi developin weln s i d a s la g countries. However, sinc somn ei e advanced countrie cure sth e rat increasins ei g faster tha cancee nth r incidence rate, the cancer mortality rate is no longer increasing in such countries. The increased cure rate ca attributee nb earlo dt y diagnosi improved san d therapy othee th rn handO . , until recently, in some parts of the world - particularly in developing countries - cancer control and therapy programmes have had relatively low priority. The reason is the great need to control communicable diseases. Toda yrapidla y increasing numbe thesf ro e disease undee sar r control. Thus ,expectee canceb y becomma ro d t eprominena t proble thid msan will resul publin i t c pressure for higher priorities on cancer care. The creation of adequate treatment facilities and the training of the necessary personnel will take time, in some cases 10 to 15 years.
In some relatively advanced developing countries radiation therapy is applied in about detectel 50al f %o d cancer cases. Approximately hal thesf o f e treatments have curative intent. Surgery and radiotherapy applied individually or combined result in the cure of about 40% of all patients. The application of chemotherapy alone has curative effects only on a small percentag e canceth f eo r patients. Moreover, palliative radiotherap s oftei y n excellenn i t providing prolonged increased lifan e d life qualit patientr yfo s with incurable cancer.
encouragins i t I noto gt e tha resulte tth s achieve radiatioy db n therapy show continuous improvement. This can be traced back to a number of developments: increased knowledge regarding rumour and normal tissue response to radiation, early diagnosis with improved tumour localisation, improved dosimetr dosd yan e planning e introductioTh . f moderno n equipment (CT-scanners, ^Co-units, linear accelerators, computerised treatment planning systems, etc.) has been crucial in these developments and makes possible a more accurate target delineation, better treatment planning resulting in irradiation of the Planning Target Volume (PTV) with a highly uniform dose and, simultaneously, a reduction in dose to healthy tissues outside the PTV.
Experience shows that high quality radiotherap onln achieveye ca y b conductes i t i f di d by a skilled team working closely together with good communication between various categories of staff. The team must consist of radiation oncologists, radiation physicists and radiographers alss i t oI . shown that dose prescribe dosd dan e delivered hav agreo et e within +/- 5 % in the PTV to achieve a controlled cure rate without excessive complications to normal tissue. Due to the increasing demands for high accuracy in dose delivery, one of the goals of the seminadeao t s differene l witth wa r l hal t step treatmenn si t procedure fro decisioe mth n of treatment strategy to the quality assurance of the treatments.
In some advanced developing countrie sespeciall- Latin yi n Americs i tren e w th ad- no to move from ^Co-unit lineao st r accelerators. Absolute dose wels sa doss a l e distributions fro lattee mth r typ therapf eo y machine easiln s ca alterevar e yd b yan d through service actions or faulty parts. Therefore, seminar trainind san g courses coverin aspectl gal qualitf so y control in radiotherap d dosimetran y f greao e ty ar importanc should ean e helb d d regionallr yo nationall regulaa n yo r basis. EDITORIAL NOTE
preparingIn this publication press,for IAEAthe staffof have pages madethe up from the original manuscripts submittedas authors.the viewsby The expressed necessarilynot do reflect those governmentsofthe nominating ofthe Member nominating the States of or organizations. Throughout textthe names Memberof States retainedare theyas were when textthe was compiled. The use of particular designations of countries or territories does not imply any judgement by publisher,the legalthe IAEA,to statusthe as of such countries territories,or of their authoritiesand institutions delimitation ofthe or theirof boundaries. mentionThe of names of specific companies productsor {whether indicatednot or registered)as does implyintentionnot any infringeto proprietary rights, should construednor be it an as endorsement or recommendation on the part of the IAEA. The authors are responsible for having obtained the necessary permission for the IAEA to reproduce, translate materialuse or from sources already protected copyrights.by CONTENTS
L ACCURACY REQUIREMENTS IN RADIOTHERAPY
Tumo normad an r l tissue response fractionateo st d non-uniform dose delivery ...... 9 . P. Kallman, A. Âgren, A. Brahme Converting dose distributions into tumour control probability ...... 27 A.E. N ahum Definition of treatment geometry in radiation therapy ...... 41 P. Acdtonen Dosimetric precision requirements and quantities for characterizing the response of tumors and normal tissues ...... 49 BrahmeA.
DL EQUIPMENT REQUIREMENTS
Lessons learned from accident radiotherapn si y ...... 9 6 . Ortiz-Lopez,P. Novotny,J. HaywoodJ. Review of WHO/PAHO/IAEA recommendations concerning radiotherapy facilities . . 83 O.P. Hanson Alternative design megavoltagr sfo e machine r cancesfo r treatmen developinn i t g countries ...... 3 9 . C. Bonus, Svensson,H. G.P. Hanson Simulation and radiation treatment in external radiotherapy ...... 101 E. Singer Analysis of variations in the dose delivered in radiation therapy ...... 107 D.B. Feld
m(a). INTERCOMPARISON role f SSDL-Helsinkeo Th r dosimetrfo i qualitd yan y audi radiotherapn i t y ...... 3 11 . P. Aaltonen SSDL Argentina: Dosimetric intercomparison programm r cobalefo 0 6 t therapy units ...... 123 Saravi,M. Papadopulos,S. MugliaroliH. A program on quality assurance and dose calibration for radiation therapy units in Venezuela ...... 135 M.C. de Padilla, L. Carrizales, J. Diaz, F. Gutt, A. Cozman
IIKb). DOSIMETRY PROCEDURES
Radiation dosimetry with plane-parallel ionization chambers: An International (IAEA) Code of Practice [Invited Paper] ...... 143 P. Andreo, P.R. Almond, O. Mattsson, A.E. N ahum, M. Roos An algorith mincludo t bremsstrahlune eth g componen determinatioe th n i t e th f no absorbed dose in electron beams ...... 159 S.C. Klevenhagen Clinical dosimetry with plastic scintillator salmos- t energy independent, direct absorbed dose reading with high resolution ...... 165 U. Quast, Flühs,D. KolanoskiH. Absorbed dose beam quality factors for cylindrical ion chambers: Experimental photoV determinatioM 5 n1 beamd an s6 t na ...... 1 17 . C. Caporali, A .S. Guerra, R.F. Laitano, M. Pimpinella Displacement correction factor versus effective poin f measuremeno t deptn i t h dose curve measurements at 60Co gamma rays ...... 179 Bruno,A. G.R. Vêlez, BrunetteM. An analysi f somso e aspect attenuation-scattee th f so r function brachytherapn i s y dosimetry ...... 185 S.C. Klevenhagen Standardization of indium-192 coiled source in terms of air kerma output...... 199 Shanta,A. Unnikrishnan,K. U.B. Tripathi, Kannan,A. P.S. lyer Calibratio 192f no Ir high dose rate brachytherapy sources ...... 3 20 . M.H. Maréchal, C.E. de Almeida, C.H. Sibata Quality control of Ir-192, Cs-137 and Ra-226 sources for use in brachytherapy ..... 207 C.H. Oyarzûn Cortes, PalmaD.,AM. PenalozaH. TomicicM. C.,
. QUALITIV Y ASSURANCE NETWOR RADIOTHERAPKN I Y
Quality Assurance Network: the European pilot study ...... 213 J. Chavaudra, Dutreix,A. Deireumaux,S Schuerender Brider,A. van E.
V. QUALITY ASSURANCE PROGRAMM RADIOTHERAPEN I Y
Minimum requirement prograA Q a radiatiomn n i so n oncology ...... 7 23 . P.R. Almond Accurac radiosurgeryn yi influence Th : f collimatoeo r diameterc ar d san weight dose th en distributioso r singlnfo e target ...... 1 25 . M.C. Plazas, D. Leflcopoulos, M. Schlienger, L. Merienne Dosimetry of breast cancer ...... 259 G. Ramirez C., J. Restrepo, CA. Aguirre Platon V2.0 & BRA VI.0 system for teletherapy and brachytherapy ...... 263 C. Aries, Coscia,G. LuongoA. Thermoluminescence dosimetry applie qualito dt y assuranc radiotherapyn ei , brachytherapy and radiodiagnostic ...... 267 G. Marinello
SUMMAR CONCLUSIOND YAN S ...... 1 28 .
LIS PARTICIPANTF TO S ...... 7 28 .
RECENT IAEA PUBLICATIONS ON RADIATION DOSIMETRY ...... 293 L ACCURACY REQUIREMENTS IN RADIOTHERAPY
>"&•••My ' yw ::EXI left BLAKK I
r- *-!•••;. "•i-ayS"^'"'^'* TUMO NORMAD RAN L TISSUE RESPONSE O ST XA9642842 FRACTIONATED NON-UNIFORM DOSE DELIVERY
P. KÄLLMAN, A. ÄGREN, A. BRAHME Radiation Physics Department, Karolinska Institute, Stockholm, Sweden
Abstract
The volume dependence of the radiation response of a tumor is straight forward to quantify because it depends primarily on the eradication of all its clonogenic cells. A tumor therefor parallea s eha l organizatio surviviny an s na g clonoge principln ni n eca repopulate the tumor. The difficulty with the response of the tumor is instead to know the density and sensitivity distribution of the most resistant clonogenic cells. Th tumo% e 50 increas re controth n i ee decreas lth e maximu dosth d an en i e m normalized dose slopth ef e o respons e relation presencn i , y , f smaleo l compartments of resistant tumor cells have therefore been quantifie describo t d e their influencn eo the dose response relation. Injury to normal tissue is a much more complex and gradual process. It depends on earlier effects induced long before depletion of the differentiated and clonogenic cells that in addition may have a complex structural and functional organization e volumTh . e dependenc e dosth ef eo respons e relatiof o n normal tissues is therefore described here by the relative seriality, s, of the infrastructur organe th modee f ealso Th n . generalizee oca lb describo dt response eth e of heterogeneous tissues to non uniform dose distributions. The new model is compared with clinical and experimental data on normal tissue response, and shows good agreement both with regard to the shape of dose response relation and the volume dependenc isoeffece th f eo t dose response Th . tumorf eo normad san l tissues quantifiee ar arbitrarr dfo y dose fractionations usin lineae gth r quadratic cell survival parameters a and ß. The parameters of the dose response relation are derived both for a constant dose per fraction and a constant number of dose fractions, thus in the latter case accountin uniforn g no als r omfo dose delivery.
1. INTRODUCTION
The major lines of progress of radiation therapy has during the last decade been: the development f threo e dimensional (3D) diagnostic methods; accurat treatmenD e3 t planning techniques; and new accelerator and isotope devices for precise 3D dose delivery r understandinOu . e developmenth f go d spreaan t f tumoro d s alsha so increased throug e identificatiohth f oncogeneo n e developmenth d an s f traceo t r techniques. The importance of fractionated radiotherapy has also become understood, n improvea o t e ddu knowledg e timth ef o edependenc f repaio e r processen i s malignan d normaan t l tissues. Equally importan e developmenth r fo t f modero t n radiotherap modeline th s yi radiatiof go n responses. This knowledg fundamentas ei l for an accurate evaluation of treatment response, and for the determination of that optimal dose distribution which will eradicat giveea n tumor with minimal adverse reactions in normal tissue [1, 2, 3]. Basically there are two different ways to approach the prescription of the absorbed dose for a given tumor location. If the tumor type is well known and not too radiation resistant the dose level is given mainly by its sensitivity and size. However, for most tumors the situation is more complex. There may be a resistant population of tumor cells prescribee Th . d dose leve thes i l n largely determine acceptable th y db e levef o l complication surroundine th n i s g normal tissues. Morphologicallye tissub a , n eca modelle complea s da seriaf x o paralle d an l l structures mose Th . t trivial case firsf so t order are purely serial or parallel structures as shown in Fig 1 a and b respectively. In this wor e simpleskth trivian no t l second order textur f paralleeo l serial structures have been use s describeda relative th y db e serialitcf, . Fign m/(ml/ & = s = e .y l ( ) n • Eq.(16) below) to describe the volume dependence of normal tissues. However, most organs hav organization ea f sucno h structure eveo st n higher degree complexitf so y as illustrated by the kidney in Fig. Id. For tumor e volumth s e dependenc e radiatioth f o e n respons s fairlei y simplo t e handle as it basically depends on the eradication of all clonogenic tumor cells. From a structural point of view the tumor is a parallel tissue because all clonogens have to be eliminate controo dt tumore th l . Instea difficulte dth y wit tumoe knoo hth t e s i wrth
m a) d)
b)
n
m c)
n
Figure l Schematic examples of tissue organization structures in the parallel- serial model. A serial string of subunits as described by equation (9) is parallee th show , a) ln ni structure , equatio finalld n an e (10)) yth b n ,i serial-parallel structure in c) is a more realistic organ model. The properties of the organ are now controlled by the parameters n and m, indicates a b figuree d th aan n d i , where 0 10 densit radiatiod yan n sensitivit mose th f ty o resistan t clonogenic cells. Toda wels i t yi l known that they may determine the result of therapy. It has recently been shown that even if they only make up as littip as 10~4 - 10~5 of the clonogens their distribution will determine the shape of the optimal dose distribution and the required dose level for tumor eradication[l. 2] , Normal tissuese otheth n r o ,hand , canno wele b t l describe purela y db y parallel structure d injuran , s inducei y d befor e depletioth e l steal mf no cells. Modelr fo s normal tissue reactions should therefore consider various degrees of injury to substructure f higheo s r complexity e developeTh . d mode s generalizei l o t d fractionated non uniform dose delivery and finally compared to clinical and experimental dose response data. . 2 THEORY 2.1. Classical dose response relationships The dose response relation for tumors and normal tissues has been described by a variety of mathematical functions. To allow a precise comparison (Fig. 2) we will writ e threth e e most commo f thea nstandaro mn i d format usin% g 50 onl e yth response dose, D50, and the maximum value of the normalized dose response Probit/• x- Logtf/Poisson .8 .6 .4 0 8 D/Gy 0 6 0 4 0 2 0 three Th e Figurdifferen e2 t dose-response curves base equationn do s (l)-(3). The solid curve is the Logit, the dashed curve is the Probit and the dotted curve is the Poisson expression. In this case DSQ = 50 Gy and y = 2-5. 11 gradient descriptorss a , y , Probite .Th , Logi Poissod an t n models respectively then take formse th : (1) 2 U5Q = l [ey(l-D/D50)] e 3U (3) Only Eq.(3) has a strict radiobiological background since it is based on the Poisson statistical model of cell kill. The Probit and Logit models may be used to approximate radiobiologicallshape e th th f eo y more relevant Poisson model Probie Th . t modes i l simple to use to estimate the influence of dosimetric and biological uncertainties [4, 5, Probie Logid Th . 6an t7] , t model mathematicalle sar y wheeasiee us no rt analyzinga large clinical material, but less desirable as these models have no biological base and merely approximates the sigmoidal curve shape. For example is the maximum value normalizee oth f d dose response gradient dP/dDD = y , , just abov Probie e th DS r Qtfo model, precisely at D$Q for the Logit model, but just above the 37% probability level Poissoe foth r n model Fige Furthermorese . , 2 . curve higth d , ean h shap w effeclo t ea t probabilities, whic importante har normae th r fo l tissue d tumorssan , respectively, may deviate considerably. 2.2. Volume dependenc dosf eo e response models The effect of a homogeneous dose delivery to an entire tumor or organ at risk may be approximated by Eqs.(l-3). In radiation therapy, only a fraction of each organ will normally be irradiated at a high dose, a situation not directly covered by these equations. To calculate the injury caused by partial irradiation, the organ could be organized in a structure of sensitive subunits. For tumors the classical model is to assum eunifora paralled man l structur probabilite th o es controo yt fractioa l e th f no whole tumor volume, v = V/Vref, of known response P(l) for the reference volume gives i ) : 1 nVreby = v ( f P(v) = [P(l)f (4) For the Poisson model, Eq.(3), the dose and volume dependent expression becomes especially simple: > + (5a) This expressio rewrittee b y n ma expres o nt volume sth e dependencd an y e th f eo values: (6) 12 where and °50,v =° The normalized dose response gradient and the dose giving 50 percent control probability increases with the logarithm of the volume or the number of tumor cells above Th . e 6] relations , [5 , derive uniforr dfo m tumorsgeneralizee b y ma , modeo dt l e volumth e dependence alsnormar ofo l tissue r structureo s d tumor insertiny b s ga constant A: in front of the logarithmic terms in Eqs.(5a), (7) and (8). For normal tissue Eq.(5a) then becomes D,n) + fclnvl J ' (5b) The constant k will be equal to unity for uniform tumors but will generally have a negative value for normal tissues. This is a mathematical way handling the decreased risk of causing injury when a smaller volume of normal tissue is irradiated as it is radiobiologically comparable to increasing the effective clonogen number NQ and y (y = InNo/e, [5]). Data on DSQ, 7 and k for some representative tissues are given in Table I by fitting to recent clinical data from Emami et al. [S]. By fitting Eq.(5b) to clinical tumo rk value dat] 1 a[ s thalese ar ts than unity (0.5 - 0.5their 0 fo 5 r data oftee ar ) n obtained indicating thadensite th t clonogenif yo cuniforn cellno s si m ove tumoe th r r volume simplicitr Fo . y this mode calles -modei lk e dth l below. A more relevant descriptio organ a f nobtaines o ni dividiny db t ginti o morphological or functional subunits. Modelin organ ga comprises na subunity db bees sha n dony eb several workers. Morphologically, the elementary compartment can be structurally well define r undefinedo d e termTh . "functional subunit "s suitei (FSU ) r [9 dfo , structurally well defined tissue compartments liknephrone er th structurall fo t bu , y quasi homogeneous tissues likskine "tissue th th , e rescuing unit" (TRUe , th [10] r o ) "regenerative unit" (RU, mora [11] s i )e valid descriptor. When definin functionaga l subuni e shoulon t d kee minn pi d thatargee th t tt onl organ celle a no yth f s o nsi functional cells but the tissue regenerating cells may be even more important. The divisio organ a f no n into functional subunit therefors si e quite comple centere th f xi s of function and regeneration do not coincide. Tissue regeneration can follow one of two different pathways: either 1) through proliferation and diffusion of progenitor cells that differentiate to functionally mature cells. This is the most frequent mode of regeneration and takes place in for example epithelial and hematopoietic stem cells; or 2) through proliferation of differentiated, functional cells until tissue damage is repaired. Typical cell types belonging to the latter group are hepatocytes and endothelial cells. Wheldo [12. al ]t ne calle d these different mode celf so l renewa- Pi s a l type (hierarchical) F-typd an , e (flexible), respectively discussed an , d their responso et radiation. In most tissues the regenerative or functional subunits all contaiu both functiona progenitod an l r cell botr sfo h mode f regenerationso . Hence observatioe th , n that permanent tissue damage arise from injur regeneratino yt g cell alss si o consistent with the presence of subunits. The subunit arrangee sar d structurall giv o yt e functiona eth l propertie organn a f so . Figure la shows an organ containing m subunits in series. The response probability of 13 the entire orga madnpurelf P o p eu y serial subunits depend locae th ln sresponso f eo subunits it describe, sPf Eq.(3)y db , accordingo t ) (9 !/>= - ft (1-/V /' = 1 orgae Ith f structures ni subunitn s da paralleln si , (Fig. Ib) response th , insteas i eP d given by P= ft P; (10) y = l J Many organs f havserialo a e , parallel and/or cross-linked organizatio f theio n r subunit varyina o st g degre complexitf e o simples e Th trivia. n d) no yt d (Figl an e .l second order structur n-ma s ei matri parallef xo seriad an l l subunits (Fig. le), giving followine th g response n m n i-1=n 1a-/>,,v) When the sensitivity of all subunits is assumed to be identical and the absorbed dose distribution is homogeneous, P.. =PEq. (11) becomes: A/ giving probabilite Th inducinf yo g injur fractioa o yt whole n th a-bf o e b orga w no n y Pfl^ma obtained from r i b 'n \ , / a i b •" P _i n p ïa 'm\ -n n plln) Pab-[ J l'(l'P-L A> 1 ^1-^ J > (14) e composefoth r d serial-parallel tissue ,e relativ wherth a However, when numerically calculatin responsee gth elementare th , y compartment will not be a functional subunit, but a volume element or voxel. Such a subunit is not 14 equivalent to an FSU and the response of a voxel is a purely computational entity, use determino dt integrae eth l respons orgae givea th r f nfo eo n dose distributios na described by the model, and it's response cannot be interpreted simply as a local probability of injury. For structurally undefined tissues like the skin and neural tissues this is particularly so, because the survival of a single subunit is affected also by other factors than DNA repair mechanisms and the regenerative capacity, for instance cell diffusion from neighboring volumes. A refined analysis should include all causes of variation e numbeth n i sf functiona o r le irradiate cellth n si d volume, also second order effects as cell diffusion and the transformation of differentiated cells to an altered stat f differentiationeo r examplfo , e from astrocyte oligodentrocytesd san A . critical volume model that takes diffusion into account have been developed by Yaes and Kalend [14]. However, in the first approximation cell diffusion is taken into account in the present model as the size of the subunits are for some tissues influence effective th y db e diffusion distanc examplr fo f eo e progenitor cells. No tissue is purely serial or parallel, but a large group of normal tissues are preferable parallel [9, 11] as the liver, the lungs and all tumors. In the serial extreme there are the spinal cord esophagusdan nexe th t n I sectio. generalizena d descriptio f tissuno e organization is introduced that represent a further simplification in that it does not requir edetailea d knowledg geometricae th f eo l arrangemen subunitsf o t . 2.3. Respons heterogeneour efo s tissue dosd sean distributions 2.3.1. Normal tissues The above described models for tissue response allow the calculation of the response of one sub-compartment PD when the structure and response of the entire organ is know r unifornfo m irradiation purpose th r f findinFo .eo gdescriptoa f treatmeno r t success which will alloseleco t optimae s wu th t l beam configuratio irradiatiod nan n geometry (cf. [2])d calculatioe secan ,th 4 . 2. probabilite th f no benefif yo injurd an ts yha to be performed in the patient's anatomy with heterogeneous tissues and dose distributions. The calculation must also be reversible, i.e. when the response probability subunite distributio integrae th l th al n knowns f si o ca l y responsnw P ho , e of the entire organ P be calculated? One of the computational problems experienced then is that a completely serial tissue as described by Eq.(9) gives P = 1 if Pj = 1 for any subunit i. Similarly the parallel organ subuniy strone an r Th . fo gt ;' 0 influenc= j P r singlcola f o i f t dgivee0 o espoho = sP t respectively t clinicallno s i , y realisti mosr cfo t normal tissues. Possible exceptione sar extremely homogeneously serial tissues. This problem is decreased and almost eliminated if a structure combining serial and parallel subunits is used as described by Eqs.(15-16) below. However, an inconvenience with this, or any, structured tissue model is that the respons s criticallei y dependen geometricae th n o t l alignmen orgae th f no t structures relative incidenth o t e t beam. Irradiatin a gcertai n percentag e seriaon f lo e compartment wile samt resulth no l en i tprobabilit f damago y s irradiatina e ga corresponding fractio parallee th f no l compartments. Physiologically, thi reasonabls si e t demandi t bu s considerably improved knowledge abou orgae th t n fine structurd ean behavior to make it useful in therapy planning. Furthermore, the exact location of the various structures has to be known in relation to the therapy beam. This will be a major obstacl ordinarn ei y therapy plannin performes i t i s ga d today, even thougt hi may be needed for a precise description of certain heterogeneous organs. It should 15 therefor muce eb h more preferabl response th f ei e coul orgabasen e da b n ndo model without a too complex internal geometrical structure, but with a correct integral and partial volume response. A special degenerated case of the parallel-serial model described above is obtained by defining a parameter, s, which essentially describe the relative seriality of the tissue with the additional simplifying assumption that b = I. The response Pv of a subunit v of a tissue can then be determined using Eq.(14) if the relative seriality of the tissue is known: n -m l s Ss v ' Pv=V-d-P ) ) (15) where P is the response calculated from Eq.(3) and v as before is the relative volume fraction of the entire organ. This definition of a subunit makes non-zero response values possible, even for partly non-irradiated parallel organs. Inverting this equation and assuming a homogeneous dose distribution to all functional sub-units with survival Pv at dose D leads to s l/ s 1/v (l-(l-PP= ) v) (16) here Iw fe a generalizeregar s a ds d parameter describin e relativgth e seriality, independen f irradiatioo t n geometry, then Eq.(16) fulfill e requirementth s s stated above sinc t representi e a structureless s model with decreased influenc f locao e l extreme low or high dose values. The next step is now to express the response of the entire organ when the dose distributio n e celuniforth no l d s densiti nm an s homogeneousyi fractionae Th . l volume of a volume element or voxel, Dv> will be equal to one over the number of voxels, M, of the organ in the diagnostic image, and we obtain (P = (1 - (1 - generalizee whicb y hma non-unifora o dt m dose distributio: nby (17) By expressing PAV(DI) whole th inr efo term organP f so , usin . (15)e gEq b , y thisma rewritte: nas p = i- nM 1=1 j (18) wher response e th entire uniforth n f eno o orgaa mo t dosnP e distribution ca w nno be described as a function of the response of the whole organ for the dose Df in each compartment i. The sensitivity of normal tissues is generally not heterogeneous within an organ, so even though the model can include heterogeneity, this is generall requiret yno normar dfo l tissues simplicitr Fo . presene yth t model, Eqs.(15-18), will be referred to as the s-model below. 16 2.3.2. Tumors Modellin e responsgth f heterogeneoueo s tumor non-uniforo t s m dose delivers yi rather simple as all tumors a priori have a purely parallel structure [6]. The clinical problem is to obtain sufficient knowledge of the heterogeneities with regard to the effectiv esensitivitd densitan ) (r y« (r), yD (effective D0)o clonogenie th f c tumor cells. If these distribution knowne studie sSPECar T y b . predictivd PE ,r eg sT an o e assayss i t i , relatively straight forwar expreso t d probabilite sth controo yt tumore th lmeae Th . n numbe survivinf o r g clonogen volum e gives si th y nb e integral: N =(nCr)e^/D^3r (19) o Under the assumption that Poisson statistics holds ([5, 6] the probability to control the tumor thus becomes: (20) The relative steepness y of the dose response relation may now be derived from its definition 3D where P(D) is given by Eq. (20). To get a feeling for the implications of these equations we will apply them on the simplest possible heterogeneous tumor consisting only of two different cell populations, the first of normal radiation sensitivity (De = DI = 2.5 secone th Gyd d)an radiation resistant (eg) . 3 hypoxi= R cOE celln a t sa wit, - 7. I 5Gy hD and a uniform delivered dose D. Eqs. (20) and (21) then reduce to and 7'1+2i .T(OV y) = P !1+| .ol-r—2m / - 6 D« 1 +-=r~ C 2 \ D ^^ /oo\ where the corresponding quantities for each cell population alone becomes: (24) (25) (26) (27) 17 0 20 80 D/Gy Figure 3 The dose-response relationshi a heterogeneou r fo p s tumor PI+2 consisting of 108 'normal' tumor cells with DO = 2-5 Gy and 103 hypoxic cells with DO = 7-5 Gy (OER = 3) with a dose response according to ?2curved an respectively I sP s see i resistan t w I n. fe thae th tt hypoxic cells lower the slope of the dose-response relationship to a value mose thas th a f talmos o tw resistan lo s a t t cell fractions. Using these later notations Eqs.(22) and (23) may be reduced to: P —P jp i .o "™" * i « * o (28) (29) As mentioned above the maximum relative steepness of the dose response relations ma approximatee yb : dby (30) Similarly correspondine th , g D5Q values accordin become ) Eqs.(2o gt 25 d 4an : (32) 18 From the last two relations it is clear that if there are even a few resistant cells in comparison to normal cells, their larger De value may cause a similar or even higher DSO value. If NI is equal to 108 and N2 is only 103 then DSQ,! = 47 Gy and Dso,2 = 55 Gy 2d Yz-5~ resistan an -w Thu8 fe 6. t e YI~ sth bu t cells will dominat tumoe eth r control as their DSO valu highes ei accordind ran Eq.(28o gt ) thu illustrates sa Pi+ 2 P 2- y db the curves in Fig. 3. And according to Eq.(29) their low y value will strongly influence heterogeneoue th s tumor e 100.00eveth f ni 0 times larger numbe f normao r l cells have a much steeper response as PI is close to unity and P2 is small over the steep portio dose th ef n o respons e curve. Obviously there are a large number possible explanations for the shallow dose response curve observed clinically for many tumors. A number of workers have suggeste dmora lesr eo s random sprea sensitivitn di y aroun meae dth , n15 valu, 5 , e[4 16 causo ]t shalloe eth w clinical dose responser fo 7 s - observeinstea6 5 - = y 2 f d= o d(y a homogeneous tumor, [17, 18])above Th . e analysis indicates tha vera t y small compartmen f resistanto t cellcausy ma seven a e n shallower dose response thana wide spread around a mean value. The resistant cells could be repair efficient cells [1, r hypoxio ] 20 , c cell19 s remaining after reoxygenation, eve f beforni e treatmene th t fraction of hypoxic cells could have been quite large [21, 22]. From our present knowledg clinicae th f eo l rol hypoxif eo c cells [23s i likel t ]i y that plathey ya ma major role in causing a shallow dose response for some tumors even if their values are not influenced so much as seen in Fig. 3. 2.4. Complication free tumor control In order to judge the clinical merits of a given dose distribution it is important to be able to compare its advantages in terms of tumor control with its disadvantages in the form of normal tissue complications. In the general case this is difficult since the radiation effect differenn i s t tissues generall incommensurable yar e entities. However, for fatal normal tissue injuries that cannot be salvaged by surgery, a strict comparison is possible as this end-point is as undesirable as an irresectable tumor recurrence. If we define PB as the probability of getting benefit from the treatment (i.e. tumor control) e probabilitth s a I f P causinyo d an g injur normao t y l tissues thee generath n l expressio probabilite th r nfo achievinf yo g complication free tumor contro gives i + P nl by: +*-*l (34) generae Inth l cas patiente efractioa th f o statisticall e n8 sar y independen+ t P [1] d ,an may be approximated by: B (35) The parameter 6 (= 0,2) specifies the fraction of patients where benefit and injury are statistically independent end-points. Here PB could be taken from Eqs. (20 or 22) and PI from Eq.(18). If several organs at risk are present then their total probability of injury ca calculatee nb d using Eq.(9) thao s , t total injury result f onl si orgae yon fatalls ni y affected [24] lesr sFo . fatal normal tissue endpoint reductiosa n operato appliee b y drma on their individua consideo t I P l r their influenc treatmene th n eo t outcome. 19 (a) 0 60 80 D/Gy (b) 0.2- 0. °0 D/Gy 20 (c) 0.2- 0.0 D/Gy P 1.0 0.8- Esophagus 0.6 (d) 0.4 0.2 O.OL- 20 40 60 80 D/Gy Figure 4 The volume-dependent dose-response functions compared with clinical data. Here the s-model has been applied to the NCI data set for: (a) lung ) liver (b ) heart, (c ,) oesophagus (d , e soliTh .d ) line(c n i s represent the s-model and the dashed lines are the fit with the k- model. 21 3. COMPARISON WITH CLINICAL DATA The dose response relation using the "s-model" in Eq.(16) has been applied to the recen clinicaI NC t normar l fo dat t ase l tissue complication t includesse [8]e Th . s many organhumae th n i sn body d containan , probabilit% s dos50 d e an f value yo 5 r fo s injury when irradiating 1/3, 2/3 and 3/3 of a given organ. A comparison of DSQ, y and s value mads si severar efo l organ abdominae s th dati e t n si Th ase l region. based more on clinical experience than solid empirical investigations, therefore both the probabilit complicatiof yo specifiea t na d dos irradiatee eth leve d an l d volume ear uncertain. With a suitable selection of parameters and number of unknowns it will always be possible to fit a model to the available data. The aim must be to find a model tha bots i t h simpl abld epredican o et t dose volume responses fro mlimitea f o t dse clinical data. Fig. 4 shows that Eq. (16) conforms quite well to clinical data, without loss of simplicity. The curves represent a least-square fit for variables DSQ, /and s. In all cases a higher weight has been applied to the values for the irradiation of the entire organ, because of their expected higher accuracy. Basically, normal tissues are parallel, which is seewherd Fign na- i lungr fit e 4 . efo sth , heart, liveesophagud an r showne sar e Th . uncertainty of the clinical data [8] is sometimes quite large and the observed deviations ma acceptablee yb . Esophagus would intuitivel expectee yb behavo t da seria n i el manner. The s-values for this organ in humans are even exceeding one, indicating that each subunit in the serial chain contains less than one regenerating unit. Extreme seriality leads to an negligible volume dependence, as seen from Fig. 4 d, where the effec f irradiatino t g e esophagu100th f %o irradiation% s similasi 33 e th thao t rf o t. Parallel tissustrona s eha g volume dependenc uncertaintiee th d ean volumee th n si s irradiate probable ar d y quite large fc-modee Th . l from alss Eq.(ha o ) bee5b n applieo dt the clinical data and the resulting ^-values are listed in Table I. The fc-model conforms quite well to the data for those organs where the y value increases slowly with decreasing irradiated volume. Similar to the method of Lyman [25] and Burman [26] the e dosshapth ef o erespons e curv s essentialli e y retained whe e irradiateth n d volum increasings ei exampln A . hearte fc-modele thif th eo r see s Th s.i fo Fig n ni c 4 . , however functionalite t baseth no n s di ,o structurd yan organe onl n th ca f yet o I . wit h difficulty be generalized to heterogeneous dose distributions, and is strictly applicable only to rumors. The clinical data is specified as a given volume fraction irradiated to a homogeneous dose level, wheorgae resth e f nnth o t receives zero dose. Clearly, this i s an unphysical assumption, and in the future there is a need for a more refined analysis of treatment plans and complication probabilities with heterogeneous dose distributions. This coul achievee db usiny d b treatmen D g3 t plans, accurate biological models and accurate diagnostic information. 4. DISCUSSIO CONCLUSIOND NAN S The radiobiological propertie f tumorso normad san l tissue gradualle sar y becoming better understood. Considerin uncertainte gth availablf yo e clinical dat e volumath e dependenc e dosth ef o erespons e models presented her e quitar e e realistic, As . expected a stron, g serial behavio s showi r organy nb s like spinath e l cord dan esophagus markea d an , d parallelit lungse th showlives ye i d .th ran Despity nb e this clear tendency, more reliable clinical information on the response of normal tissues is still needed. There is need for a more refined analysis concerning the quality of life of the remainine patienth d an t g functionalit injuren a f yo d organ whe s partialli t ni y injured. Total eradicatio kidnee lune on stiln on r gyca o lf no lea patieno dt t survival, 22 and thus a 100% probability of injury to one of two paired organ can be allowed if it decreases the dose burden to other critical structures, and increases the tumor control probability. Ultimately e difficultth , y wil coursf o lquantifo t e eb e impace th yth n o t quality of life and this evaluation should be an integral part of the normal follow-up of the patient. When the integral functionality of the organs after irradiation and the associated influenc tumoe th n reo control probability have been calculated, they coul usee db o dt fin scalada r measure that quantifie usefuw ho s give a l n treatmen suce t plaOn h . nis measure is the complication free tumor control P+, as given by Eqs.(34, 25). The aim of treatment planning will maximizo thet e nb e that measur treatmenf eo t succes. 3] , s[2 The evaluatio f differenno t treatmen f ofteo e n us quit te planth e y complesb x dose volume histograms is a complex and often subjective method. With accurate radiobiological model selectioe s th bes e th t f treatmenno t pla simpls ni y reducee th o dt compariso f scalano r probabilitie svalues + sucP outcome s ha .Th sucf eo htreatmena t optimization will depend on the accuracy of the response parameters of tumors and norma delineatioe th l n tissuetargee o th d f no tsan volume. Dat dosn ao e response relationships have been gathered by several workers and the implementation of more refined radiobiological models in treatment planning will also facilitate and promote the collection of such data. The information available today is accurate enough for many disease clinicalle b o st y usefu treatmenn i l t optimization developmene Th . f o t improved radiobiological dose response models and tumor imaging techniques will be of paramount importanc future th r eefo improvemen f radiatioo t n therapy. REFERENCES [1] ÂGREN, A., BRAHME, A., TURESSON, I., Optimization of uncompli-cated control for head and neck tumors, Int. J. Radiât. Oncol. Biol. Phys. 19 (1990) 1077-1085. [2] KÄLLMAN, P., LIND, B.K., BRAHME, A., An algorithm for maximizing the probabilit f complicatioyo n free tumor contro radiation i l n therapy, Phys. Med. Biol. 37 (1992) 871-890. [3] LIND, B.K., Radiation therapy planning and optimization studied as inverse problems, Thesis, Stockholm University, Sweden (1991). [4] SVENSSON , WESTLINGH. , , LARSSONP. , , L-G., Radiation induced lesions of the brachial plexus correlated to the dose time fraction schedule, Acta Radiol. Ther. Phys. Biol (19754 .1 ) 228. [5] BRAHME, A., Dosimetric precision requirements in radiation therapy, Acta Radiol. Oncol. 23 (1984) 379-391. [6] BRAHME, A., ACREN, A-K., Optimal dose distribution for eradication of heterogeneous tumors, Acta Oncol. 26 (1987) 377. [7] LINDBORG, L., BRAHME, A., Influence of microdosimetric quantities on observed dose-response relationships in radiation therapy, Radiât. Res. 124 (1990) S23-S28. 23 [8] EMAMI , LYMANB. , , J. BROWN, , COIAA. ,, GOITEINL. , , M. , MUNZENRIDER, J.E., SHANK, B., SOLIN, L.J., WESSON, AM., Tolerance of normal tissue to therapeutic irradiation, Int. J Radiât. Oncol. Biol. Phys. 21 (1991) 109-122. [9] WITHERS, H.R., TAYLOR, J.M.G., MACIEJEWSKI, B., Treatment volume and tissue tolerance, Int. J Radiât. Oncol. Biol. Phys. 14 (1988) 751-759. [10] HENDRY , THAMESH. . J , ,tissue-rescuine H.D.Th , g unit, Brit RadiolJ . 9 5 . (1986) 628-630. [11] ARCHAMBEAU, J.O., SHYMKO, R.M., Tissue population configuratioa s na modifier of organ response, Int. J Radiât. Oncol. Biol. Phys. 15 (1988) 727-734. [12] WHELDON, T.E., MICHALOWSKI, A.S., KIRK, J., The effect of irradiation on function in self-renewing normal tissues with differing proliferative organisation, Brit. J Radiol. 55 (1982) 759-766. [13] FAJARDO, L.F., Patholog Radiatiof yo n Injury, Masso Yornw InckNe . (1982). [14] YAES, KALENDJ. . R , , LocaA. , l stem cell depletion mode r radiatiofo l n myelitis, Int RadiâtJ . . Oncol. Biol. Phys (19884 .1 ) 1247-1259. [15] DUTREIX, J., TUBIANA, M., DUTREIX, A., An approach to the interpretation of clinical data on the tumour control probability-dose relationship, Review Article, Radiother. Oncol. 11 (1988) 239-248. [16] BENTZEN, S.M., THAMES, H.D., OVERGAARD , DoeJ. , s variation i e th n ni vitro cellular radiosensitivity explain the shallow clinical dose-control curve for malignant melanoma? Int RadiâtJ . .(1990) Biol(1 7 5 .) 117-126. [17] SUIT, H.D., WALKER, A.M., Predictor f radiatioo s n respons e todayus n ei . Criteria for new assays and methods of verification. In Prediction of tumor treatment response. Pergamon Press (Banff proceedings) (1988). [18] BRAHME, A.(Ed.) et al., Accuracy requirements and quality assurance of external beam therapy with photon electronsd san , Acta Oncol. Suppl (1988)1 . . [19] YAES, R.J., Tumor heterogeneity, tumor sized radioresistancean , , IntJ . Radiât. Oncol. Biol. Phys. 17 (1989) 993-1005. [20] WEICHSELBAUM, R.R., BECKETT, M.A., SCHWARTZ, J.L., DRITSCHILO, A., Radioresistant tumor cell presene ar s hea n i nectd dan k carcinomas that recur after radiotherapy, Int RadiâtJ . . Oncol. Biol. Phys (19885 1 . ) 575-579. [21] DISCHE , HypoxiS. , locad aan l tumor control . RadiotherII : . Oncol. Suppl0 2 . (1991) 9-11. [22] GONZALEZ, G. D., Hypoxia and local tumor control: I. Radiother. Oncol. Suppl (19910 .2 ) 5-7. [23] BROWN, J.M., Management and significance of hypoxic cells. Annual meetin f ASTROgo , Miami, Florida (1990) 24 [24] WOLBARST, A.B., Optimizatio f radiationo n therape critical-voxeTh : II y l model, Int. J Radiât. Oncol. Biol. Phys. 10 (1984) 741-745. [25] LYMAN, J.T., Complication probabilities as assessed from dose-volume histograms, Radiât. Res (19854 .10 ) S13-S19. [26] BURMAN , KUTCHERC , , G.J., EMAMI , GOITEINB. , , FittinM , f normago l tissue tolerance data to an analytic function, Int. J Radiât. Oncol. Biol. Phys. 21 (1991) 123-135. EEXT PAGE(S) I left BLANK CONVERTING DOSE DISTRIBUTIONS INTO TUMOUR CX)NTROL PROBABILITY A.E. NAHUM Joint Department of Physics, The Royal Marsden Hospital, London, United Kingdom Abstract The endpoints in radiotherapy that are truly of relevance are not dose distributions but the probability of local control, sometimes known as the Tumour Control Probability (TCP) and the Probability of Normal Tissue Complications (NTCP). A model for the estimation of TCP based on simple radiobiological considerations is described. It is shown that incorporatio f inter-patienno t heterogeneity intradiosensitivite oth y paramete througra resul n clinicalla ca n . i th5 y realistic slope for the dose-response curve. The model is applied to inhomogeneous target dose distributions in order to demonstrate the relationship between dose uniformity and s,. The consequences of varying clonogenic density are also explored. Finall modee yth s applie i ltarget-volume th o dt e DVH r patientfo s clinicaa n si l tria f conformao l i pelvic radiotherapy, the effect of dose inhomogeneities on distributions of TCP are shown as well as the potential benefits of customizin targee gth t dose accordin normal-tissugo t e DVHs. 1. INTRODUCTION Physicists working in Radiotherapy spend a lot of their time measuring doses in phantoms and then calculating the dose distributions in patients due to a particular arrangement of beams. This is becaus radiotherapise eth t prescribe treatmene sth termn i t(uniforma f so ) targete dosth o et volume accompanie somy db e r morsoro f constraino e te dose on organs th o et n o t trisk. a Howevere th , endpoints in radiotherapy that are truly of relevance are not dose distributions but the probability of local control, sometimes known as the Tumour Control Probability (TCP) and the Probability of Normal Tissue Complications radiotherapise th f o (NTCP)m ai maximiso e t whil P s ti .Th e TC e th e eth NTCP remains below some "acceptable" (usually very low) level. This lecture will deal with the biological modelling of Tumour Control Probability in terms of the spatial distribution of the absorbed dose within the patient but not the temporal distribution i.e. the difference between different fractionation scheme. It should be noted that the reference list is provided furtheo t d primarilai r n readina s ya g rather tha attempn na acknowledgo t t originatore eth e th f so various concepts and ideas. desirabls i reasonSomP e modea th listeTC y e f er o ear slfo wh d below (the references citee dar not intended to be exhaustive): • Dose distribution inherentle ar D 3- y n si ver y comple assimilatinf o somd y xan e wa g this vast amount of information is needed [1,2] meana s A quantifyinf so M g treatment plan comparisons [1,2] estimatinf o y wa effece a g non-uniformitieth s f o tA K tumoue th n si r dose distributio] n[3 • Enable mako t e e on sestimate effece f dosth patiend o tf eo an s t position uncertaintien o s therapy outcome [4-6] m Optimization is beginning to be done in terms of TCP and NTCP [7,8] 27 0.8 CO < 00 O aDC O 0.6 - Œ O Ü £E 0.4 0.2 0 8 0 6 0 4 20 100 120 140 160 DOSy G E / 9 curvesP FigurTC . e1 from Equ. r clinicallyfo 4 realistic ranginga valuesd an fromofN10 1 = 0. 0 to 1.0 Gy1. Note the unrealistically steep slopes. As a way of using the results of biological assays for o(tumour) etc. [9-11] As the only (?) way (short of doing a full clinical trial) of quantifying the benefits of improvements in dose distributions through new technology e.g. Multileaf Collimators (MLCs), 3-D planning etc. [12] clarito t d thoughf ai y Ao n sa external-bean i t m radiotherapy [13] 2. A SIMPLE BIOLOGICALLY BASED TCP MODEL 2.1. General It is well-known that the so-called Dose-Response curve has a sigmoid shape e.g. [3]. Several authors have fitted mathematical functions to this curve. However, it is not easy to see how changes in basic parameters suc s tumoua h r cell radiosensitivity, inhomogeneitie e dosth en i sdistribution , variatio tumoun ni r volumclonogenin i d ean c cell densit yaccommodatee b etc n .ca empiricay db l curve-fitting approaches e cas th f Tumou eo n I . r Control contrasn i , thao r t Complicationt fo t n i s Normal Tissues e.g. [14], it is possible to develop a model starting from the response of cells to 28 radiation. Nimierk Goiteid oan n [15] have recently described suc hmodela , whic basicalls hi y identical developee on e th o dt here [12,16] this i st I . latter model whic describes hi d belo res e used f th w to an n d i analysise th . 2.1.1. Basic Cell-Survival Curves Numerous radiobiological experiments have demonstrated beyond doubt that the killing of cells by radiation can be described by an expression of the form exp(-aZS= - ) (1) where S is the surviving fraction after a (uniform) dose D of radiation to a population of cells. The parameter characterisb d an sa initiae eth l slop degred ean curvaturef eo , respectively survivale th f o , curve. Thi knows sLinear-Quadratii e th s na modeQ L celf r clo o l killing [10,17-19]. Whe irradiatioe nth fractionates ni external-bean i s da m radiotherapy (Fig : ful2 . l curve)r fo , the almost universally adopted 2-Gy (per day) fraction scheme the effective slope of the survival curve vers i y nearly valu e givealonea th f ey o nwrite b n . Thuca e : son N, * No exp f-aDJ (2) where JV is initial number and N the surviving number of clonogenic cells, assumed here to be 0 t irradiated uniformly and to have uniform radiosensitivity a (Gy').1 Note it would be not be difficult to reinstate the b term; Nimierko and Goitein [15] include it in their model. 1.0 0.8 cr0 = 0.08 - cra = 0.00 0-6 Ü 0.4 Tumour Vol = 320 cm3 Clngc. cell density = 107 cm"3 0.2 0.3= a 5 Gy"1 0.0 40 60 80 100 120 DOSE / Gy Figur . Tumoure2 Control Probability (TCP) Junctiona s a of target dose,, derived6 d an 5 from s Eq 7 -with a - 0.35, rcl = JO for a 320 cc volume, for sa = 0.0 and for the clinically more realistic sa = 0.05 (adapted from [12]). 29 PoissonThe 2. 2.Statistics1. Result nexe Th t incorporatsteo t s pi endpoine eth eradicatioe t th i.e. tumoue th f no r intmodele oth . Ther considerabls ei e radiobiological evidenc statemene th r efo t tha tumoua t onls i r y "dead" when every single clonogenic cell (i.e. cells wit potentiae hth uncontroller fo l d division bees )ha n eliminated. Thu quantite sth y tha require Probabilitye w t th s ei that Singleo N Clonogenic Cell Survives; equating this with Tumour Control Probability (TCP), we then exploit the Poisson Statistics Relation e.g. [20]: TCP = exp f-Nj (3) substitutw no e Nw r fef I fo intEqu 2 arrive o. w Equ3 t ea . TCP= A plo thif o tfunctioa s s expressioa P Dosf no TC produc eD r nfo well-knowe eth n sigmoidal 9 curve. Usin grealistia c numbe e valuth r einitiaf fo ro l clonogenid [19an ordee ]10 th c f f ro cello 0 sN realisti Gy"c1 value 1 o t froa (e.g 1 f . ms. o 0 obtain e [10]on famile ) sth curvef yo s show Figurn i . e1 2.1.3. Inter-patient variability in radiosensitivity Theoretical models mus comparee b t clinicao dt l data wherever possible. There e existth n i , literature, a number of Local Control vs Tumour Dose studies e.g. [21-25]. Despite the limitations associated with such data i.e. uncertainties in the dosimetry, inadequate patient numbers, imprecise clinical definitio Locaf no l Control etc. ther almose ear Tumouo tn r Contro Doss v l e curves with slopes anything like as steep as the ones in Figure 1. This has led several investigators to favour an empirical model to fit these clinical Dose-Response curves e.g. [3,8]. Various hypotheses have been advanced over the years to explain the shallowness of the clinically observed dose-local control curves e.g explanatio.e [26]Th . n tha currentls i t y thoughe b o t mose th t inter-patients likeli e yon heterogeneity intrinsie th n i c radiosensitivit tumoue th f yo r cells i.e. valuea e th s [4,10,13]. contrasn Thi i possible s th i s o t t e heterogeneit f radiosensitivityo e th f yo clonogenic cell patient'se withion y snan tumour. Hypoxia, long considere majoa e b r o dt caus f eo failure to achieve local control in radiotherapy, is not now thought to play such an important role [17,19]. Thus explanations effec basee smala th f o n td lo hypoxic thereford an , e radioresistant, fraction cellf o s have fallen into disfavour e.g. ver[27]A y. interesting analysi clinicae th f so l Dose-Response data has recently been published by Brenner [28]. He demonstrated that it was possible to explain the wide variations in the dose required to achieve local control for a number of different lesions solely in terms of variations in a from one lesion type to another and the variation in the number of clonogenic cells assumee ,th d proportiona volume lesione th bottomth e o lt f eo . Th line thif o s stud thas e yton wa did not need to invoke any assumptions on e.g. the variation in the hypoxic cell fraction with tumour size. Thu Brennee th s r analysis lends suppor base P modelo t tonln TC parameterso d o r ytw fo s , intrinsic tumour-cell radiosensitivit clonogenid an a y c cell density, that vary with tumour type. However, Brenne built no dd intr di analysi s ohi inter-patieny san t variatio radiosensitivitn i a s a d yan consequenc e numbeth e f clonogenio r c e clinicacell 0 th require N st fi l o t ddat as a cam t ou e unrealistically small. P modeThTC e l described here, whic I hbelieve soundl b o t e y base n meaningfuo d l radiobiological parameters, explicitly incorporates inter-patient variatio y assuminb n g s i tha a t distributed normally amongst the patient population, with standard deviation s» [12,16]. As one increase dose-respons e slope th th valu e f , eo s th f eo e curv thinkinef o decreases y wa ge abouOn . t this is to regard the resulting curve as the sum of the curves for different a values in Fig. 3. There is thus a group of patients with low a values who will never be cured (TCP = zero), another group with high a value wilo ls wh alway cured, e groua sb d pan with intermediat evaluea whor sfo tere mth m stochastic fraction bees ha n coineoutcome th s d thesa r efo e patient literalls si ymattea chancf ro e [29]. 30 TCP AS A FUNCTION OF DOSE UNIFORMITY 0.5 n Meay G n 0 Dos6 = e Mean Alpha - 0.35 No. Cells = 10**9 5 10 15 20 [STD. DEW DOSE] x 100 Figure 3. The effect on TCP of non-uniformity in the target dose distribution, expressed as Sc/D, for different valuesinter-patientthe of radiosensitivity variation parameter meanthe ; a s target doseis mean 0.35the ~ 60a Gy, (corresponding bladderto tumours [10])- andN JO= . 9 0 2.1.4. Application clinicalto dose-response data The present model has been applied to the case of bladder tumours; through the irradiation of human bladder tumour cells grown in vitro Deacon et al [10] determined a mean a value of 0.35 Gy"1 which was adopted. Local control vs dose curves were then computed from TCP(D)= (5) where TCP(a£>Jf give s fractioi a 0 )Equd y npatientb e an n.th 4 #f o ssuc, hava h= ethaa t exp[-fa,-a//.2ac o . g 7] a (6) and Sgi ~ 1. The initial number of clonogenic cells, N* has been estimated from the product rd x V^ with the clonogenic cell density taken to be 10 [197 ] and the mean value of the target volume V = 320 cm3; this latter valu derives ewa d fro analysin m a actuae th f so l target volumesf o outlines , a , CT n do v patients entered intongoine oth g Royal Marsden clinical tria f conformao l i pelvic radiotherapy [30]. curveo tw Figure n si Th have2 e been calculated usin above gth ee datath t botn a I . P h caseTC e sth actual clinical dose use thi n d2-Gi x s2 hospitaly(3 fractions)y G 4 6 , , just coma t t ebeloou w 0.5. This is consistent with clinical findings and lends some confidence to the model. The value of 5, = 0.08 was arrive adjustiny b t da , untigJ curv e lth e "fitted dose-response "th e dat [22]n ai ; thu sentirel n thia s si y empirical value. It has been used in subsequent analyses e.g. [16]. 31 2.J.5. Inhomogeneous Dose Distributions modee Th l develope s assumed ha thu r fa s d tha l cellal t s receive exactl same yth e dosen I . radiotherapy practice this will neve casee rb . Thu sneedes i som y incorporateo dwa t e dose distributions modelP date TC int e aTh . o th tha requires i t numbee th s d i clonogenif ro c cell thai sN t receiv edosa e Thi- A moss si t conveniently obtained0 fro mDose-Volumea Histogram (e.gH . [1,31]DV r o ) generated by the planning computer. Strictly what is required is the differential dose-volume distribution, dF/d£> from which the more familiar cumulative DVH is calculated. Thus one generalises Equ. 2 to: #* = î>* exP f'a £>J lml (7) wher summatioe eth carries n i DVH e bin n ovet th e dn sou i .r th This expression shoul alse db o usen di Equ. 5 in order to take account of the effect of both dose inhomogeneities and inter-patient a variability [16]. Strictl ytargete speakin th t insteat thar bu no fo t s e tumourrelevane i d g th th tha H r fo t DV t volume (GTV in ICRU 50 terminology [32]) i.e. one should not include the margin added to account for patient movement. However, it is currently not possible to be more precise about such issues. A more serious limitation is probably the implicit assumption that the clonogenic cell density rci is constant right out to the edges of the tumour or target volume. This is discussed in more detail below. Brahme [3] applied a TCP model, which differed only slightly from the one described here, to the question of the effect on TCP of both inhomogeneities in the target dose distribution and also uncertainties in the absolute absorbed dose determination. A similar exercise has been carried out here particulae foth r re bladde casth f eo r tumour datae dosTh .e inhomogeneit targee th n i yt volume, consisting of 109 clonogenic cells of uniform radiosensitivity, was assumed to follow a normal distribution i.e. N0j was varied normally as a function of A, with a variable s^/D (in Equ. 7). The results of this exercise are shown in Figure 3. The calculation was carried out separately for different values of 5. The mean dose was set to 60 Gy, which is close to what is employed clinically (see above). Fo groua a r patientf po s with tumour f exactlso same yth e radiosensitivit eve0 y0. ni.e= , .s small inhomogeneities in dose have a disastrous effect on the TCP; this corresponds to the very steep dose-response curves in Figs. 1 and 2. More realistic values of s» e.g. 0.10 result in a much less dramatic reduction in TCP as the dose inhomogeneity is increased. The message of this study is that the appreciable inter-patient variabilit radiosensitivitn yi y indicated clinicall manr yfo y type f tumourso s considerably reduces the consequences of even moderate deviations from target dose uniformity. The corollary of this is the conclusion reached by Brahme [3] that for certain classes of tumours with steep dose-response slopes, notabl laryne th n i yx (the normalised dose gradien) onl4 y> verg t y small uncertainties in the absolute dose determination can be tolerated. 2.1.6. Variation clonogenicin cell density If the model as described thus far is applied to the DVH of the target volume (the PTV in ICRU50 terminology [32]) then implicitl assumptioe yth mads ni e thaclonogenie th t c cell densits yi constant over the whole of the PTV i.e. one calculates the number of clonogenic cells at dose A, NO.» in Equ fro 7 produc.e mth r d c(f V^.o t an However actuallV PT e y ,th involve margisa microscopir nfo c spread plu seconsa d margi geometricallr nfo y inaccuracies. Thus clearl assumptioe yth constanf no l c tr is quite unrealistic. Figure 4 illustrates this point. Whilst there is presently no clinical data on exactly how the cell density does vary throughout a radiotherap ymodee targeth assese o t l us teffec e volumn sth ca t e thaeon t such variations might have predictee o nlookinf th o y calculat o t thidgt wa a TCPe s si changOn e .e th dosn ei thaeD t corresponds changa o t celn ei l densit d wheyr requiree non remainP s TC thae s th tunchange givea r dfo n volume element of cells. Taking the clonogenic cell density at the tumour centre, say, to be r(0) and the 32 TARGET VOLUME (PTV) DOSE | PLOTTEc p , D ALONG THIS LINE MARGIN (Mvmt. + microscopic spread) > 1 • ^, LJUO 1 X <7) X J/. ._ ^_:l____: _,A<_ ^__ ,^-rrr. / — « • _» •« UJ 1 ~~~~ A^Q Q , /nOO _J \ _J 1 j UJ 1 i POS Ü i \^^~ i / \ o i i LU 1 C/5 ( i i O i , i ' TARGETAI T CONTOUR Q , i • i y \ ' > X y POSITION MARGIN Figure 4. Schematic drawing illustrating the problem of the variation ofclonogenic cell density at the edge of the PTV; a hypothetical dose and cell density profile through the centre of the PTV is shown. corresponding quantity at some position r to be r(r), then it is straightforward to show that the change same th r ei nyiel o fo siz t dos sam e volumf r P e do th t e a e TC e elemen gives ti y nb log— , = D -A P(0) a P(r) (8) where we see that the product aDD is proportional to the logarithm of the ratio of cell densities. Figure 5 gives the dose change for three different values of radiosensitivity a. Webb and Nahum [16] have attempted to address this problem by extending the present TCP model to account for variations in rc! as a function of position in the tumour. Figure 6 is taken from their paper. This is an attempt to illustrate the practical consequences of Equ. 8 on a tumour where the decrease in clonogenic cell density follows the (entirely hypothetical) full curve in Figure 6. The main messag considerabl a tha s ei r fo t e decreasallowablee th i c r n ei dose decreas vers ei y modest. Similar conclusions were drawn in [33]. 33 ISO__TCP DOS EF CLONOGENI O CHANG N F' S EA C CELL DENSITY 0 -—.... a = 1.0 LU ISO-TCP DOSE AS FUNCTION OF CLONOGENIC CELL DENSITY 4.5 5.0 5.5 6.0 1.0 1.0 C> 0.8 0.8 CELL DENSITY x DOSE a 0.6 0.6 O 0.4 0.4 oc D(OJ = 64 Gy O a = 0.35 0.2 x cc 0.0 0.0 0 6. 5 5. 0 5. 5 4. DISTANCE FROM CENTRM C E/ variatione Figur Th dosee . th e6 n i corresponding e iso-TCPth o t condition tumoura r fo witha variation clonogenicin cell density full giventhe curveby (reproduced with permission from [16]). 34 DVH/SIGMA 0.08/NVV 40 r mean TCP = 0.40 Std. Devn. = 0.06 30 V) *-» 0) •t-f «O CL 20 "o O 10 .1-.2 .2-.3 .3-.4 .4-.5 .5-.6 .6-.7 Tumour Control Probability UNIFORM (MEAN) DOSE/SIGMA = 0.08/NVV 40 mean TCP = 0.47 Std. Devn. = 0.04 30 c: — '*-» «3 20 • O 10 .1-.2 .2-.3 .3-.4 .4-.5 .5-.6 .6-.7 Tumour Control Probability Figur a,be7 . Distributions subsetf a TCo r Pfo of treatment plans taken from RMe th H pelvic trial. The ones shown here correspond to the conventional arm and plans -which have no non-zero volumes in dose bins below 70% of the isocentre dose. a - DVHfrom the PTV, natural volume variation (NW) b - Uniform mean dose, NW. 35 DVH/SIGMA 0.08/FX= V 40 r mean TCP = 0.38 Std. Devn 0.0= . 6 Fixed Volume 30 W *—er' CO EL 20 O • O 10 .1-.2 .2-.3 .3-.4 .4-.5 .5-.6 .6-.7 Tumour Control Probability UNIFORM (MEAN) DOSE/SIGMA - 0.08/FXV 40 = 0.4 meaP 6 TC n Std. Devn 0.0= . 4 30 Fixed (mean) volume 0> *•"• «3 a, 20 **«, o d 10 .1-.2 .2-.3 .3-.4 .4-.5 .5-.6 .6-.7 Tumour Control Probability Figure 8 a,b. Distributions of TCP for a subset of treatment plans taken fron the RMH pelvic trial. The ones shown here correspond to the conventional arm and plans which have no non-zero volumes dosein bins below of the70% isocentre dose, DVHfrom- a PTV,the fixed mean volume (FXV) - Uniform b mean dose, FXV. 36 3. THE TCP MODEL APPLIED TO A LARGE BODY OF PATIENT DATA The Royal Marsden pelvic trial [30,34] has furnished DVH data for over 200 treatment plans. modeP TC le describeTh d her bees eha n applie targeDVHe e th th r do tt sfo volumes (PTVs). Originally calculationP TC 0.0d an 8 s, 0 equa (se 5 wer 0. r eo fo t l eearlie t 0.3valuecarrier = bu 5a fo t f rso d ou section); the dose at isocentre was assumed to be 64 Gy in all cases. As expected the clinically unrealisti calculation0 0. c = s s yielde averagw lo da evera valud y an broa eP fodrTC sprea valuesn di . t 0.0= t 8s e calculationsTh , which correspon mora o dt e realistic dose-response i.e. less steep dose-response relationship, also yielded a surprisingly broad spread in TCP values, with some plans having TCP values below 0.2, whereas the mean value of around 0.4. This was subsequently traced to errors in the DVH calculation by the treatment planning system. There were some plans which had non- zero values even in dose bins below 70% of the isocentre dose, effectively corresponding to severe "cold spots" i.e. underdosag PTVe th .n e i Another aspec thif to s proble positioe th ms i thesf no e "cold spots". If one assumes that these were located close to the edge of the PTV then in reality there would be clonogeniy unlikelan e b o yt c cells involved. However discusses a , d earlier modee th , s appliea l d here assume uniforsa m densit clonogenif yo c cells everywher PTVe th .n ei Clearly then resultine th , P gTC values will represent some kind of most pessimistic estimate. In an attempt to get around the above problem, an alternative set of calculations were carried out in which it was assumed that the dose was uniform in the PTV, with the value of the single dose given by the mean dose in the PTV. These calculations yielded a much smaller spread in the TCP. One expecn ca t tha "truthe tth " lies somewhere in-between thes extremo etw e calculations. 14 DOSE OPTIMIZED FOR NtCP = 0.05 12 £ 10 «3 8 Q_ "o 6 o• 4 0 .5-.3 .3-.2 .2-.1 .1-0 0-.1 .1-.2 ,2-.3 -ve Change in TCP •+ +ve Figure 9. A frequency distribution of the changes in TCP that would result from customizing the target dose such complicationthat5% a there was probability eachfor patients of51 (adapted from [12]). 37 A final "tweaking" of the parameters has been done in order to investigate the effect of the range tumour volumes (actuall range yth spreaPTVse n ei th patiente TCPsn n th de i o ) r th n Fo .i s RMH pelvic trial these range from 100 to 1680 cc with a mean value of 360 cc and a standard deviation of 277cc. Does the TCP distribution become narrower if all the PTVS are forced to be the same volume Figurs ?A showse8 factumoun e i , th t r volum standare th almoss n eo ha l effecdo al tn t a t deviation of the TCP. FUTURA . E4 PROSPECT Nahum and Tait [12] carried out a modelling study on 51 patients treated conformally in the Royal Marsden clinical tria r pelvifo l c radiotherapy (about equal number f bladdeo s prostatd an r e tumours). Firstlassumes wa t yi d tha l patienttal s were treate 2-Gn (i dy witfractions)G 4 h6 . Rectal and small-bowel (late) complications were estimated using the Kutcher-Burtnan model for NTCP [14] together with the Burman et col [35] values of TDy> etc. A very broad spectrum of NTCP values resulted, spectrue whic th s entirelo t h wa f normal-tissu me o ydu e DVHs foun r thidfo s grouf po patients. The prescribee nth d dos eacr efo h patien s adjustewa t d unti NTCe th l P equalle , thud5% s yieldin a spectru w gno f dosesmo . Figur showe9 changee thaP th s tTC wouln i s d have resulted compared to the standard 64-Gy prescription; the increases in TCP easily outweigh the decreases. Could customized dose prescriptio optimizee pare th n b f to d radiotherap futuree th f yo ? 5. CONCLUDING REMARKS The TCP model described here is mathematically very simple and yet it is a reasonably complete descriptio procese th f no f tumou so r eradicatio irradiationy nb , apart from proliferation effectse Th . lace mai clinicaf th k o n e ar limitation le datus radiosensitivitn s ao it n si clonogenid yan c cell density. Thus the absolute values of TCP predicted must be treated with caution. Nevertheless this and similar modelusee b gaio dn t nsca insigh certaif o P t intn effece TC feature oth n o t dosf so e distributions such as inhomogeneities [36], patient movement [37] and of differences in patient radiosensitivities. It is to be hoped that the biological assays for the latter currently under development [9,11] will provide data to enable treatments to be individualized both biologically as well as physically. ACKNOWLEDGEMENTS The financial support of the Cancer Research Campaign, UK is acknowledged. REFERENCES [1] GOITEIN, M., The comparison of treatment plans, Seminars in Radiation Oncology 2 (1992) 246-256. [2] WEBB, S., The Physics of Three Dimensional Radiation Therapy: Conformai Therapy, Radiosurger Treatmend yan t Planning, Institut Physicf eo s Publishing, Bristol (1993). ] [3 BRAHME , DosimetriA. , c precision requirement radiation si n therapy, Acta Radio!. Oncol3 2 . (1984) 379-391. [4] BOYER, A.L., SCHULTHEISS, T., Effects of dosimetric and clinical uncertainty on complication-free local tumor control, Radioth. Oncol (19881 1 . ) 65-71. [5] MIJNHEER, B.J., BATTERMAN, J.J., WAMBERSIE, A., Reply to: Precision and accuracy in radiotherapy, Radiother. Oncol (19894 1 . ) 163-167. ] [6 AWWAD, H.K., "The impac f qualito t y assuranc optimizatioe th n eo f Tumouno r Control Probability", Radiation dos radiotherapn ei y from prescriptio deliveryo nt , IAEA-TECDOC- 734, Vienna (1994). 38 [7] KÄLLMAN, P., Optimization of radiation therapy planning using physical and biological objective functions, PhD Thesis, Stockholm University (1992). [8] MOHAN, R., MAGERAS, O.S., BALDWIN, B., BREWSTER, L.J., KUTCHER, G.J., Clinically relevant optimization of 3-D conformai treatments, Med. Phys. 19 (1992) 933-944. [9] PETERS, L.J., BROCK, W.A., CHAPMAN, J.D., WILSON, G., FOWLER, J.F., Response predictors in radiotherapy: a review of research into radiobiologically based assays, Br. J. Radiol. Suppl 22 (1989) 69-108. [10] DEACON, J., PECKHAM, M.J., STEEL, G.G., The radioresponsiveness of human tumours and the initial slope of the cell survival curve, Radiother. Oncol. 2 (1984) 317-323. [11] WEST, C.M.L., HENDRY, J.H., SCOTT, D., DAVIDSON, S.E., HUNTER, R.D., 25th Paterson Symposiu thers i m - futurea radiosensitivitr efo y testing? . CanceJ . (1991Br ,4 6 r ) 197-199. [12] NAHUM, A.E., TAIT, D.M., "Maximising Local Control by Customised Dose Prescription for Pelvic Tumours", Advanced Radiation Therapy: Tumour Response Monitorind an g Treatment Planning (BREIT, A., Ed.), Springer, Heidelberg (1992) 425-431. [13] DUTREIX, J., TUBIANA, M., DUTREIX. A., An approach to the interpretation of clinical tumoue datth n ao r control probability-dose relationship, Radiother. Oncol (19881 1 . ) 239-248. [14] KUTCHER, G.J., BURMAN, C., Calculation of complication probability factors for non- uniform normal tissue irradiation: the effective volume method, Int. J. Radiât. Oncol. Biol. Phys. 16 (1989) 1623-1630. [15] NIMIERKO, A., GOITEIN, M., Implementation of a model for estimating tumor control probability for an inhomogeneously irradiated tumor, Radioth. Oncol. 29 (1993) 140-147. [16] WEBB, S., NAHUM, A.E., A model for calculating tumour control probability in radiotherapy includin effecte gth inhomogeneouf o s s distribution dosf clonogeniso d ean c cell density, Phys. Med. Biol (19938 3 . ) 653-666. [17] STEEL, G.G., PEACOCK e som, ar J.H. ey huma,Wh n tumours more radiosensitive than others?, Radiother. Oncol (19895 1 . ) 63-72. [18] STEEL, G.G., Cellular sensitivity to low dose-rate irradiation focuses the problem of tumour radioresistance, Radioth. Oncol. 20(1991) 71-83. [19] STEEL, G.G. (Ed.), Basic Clinical Radiobiology, Edward Arnold, Sevenoaks, UK (1993). [20] PORTER, statistice E.H.Th , dose/curf so e relationship irradiater sfo d tumours . RadiolJ . Br , . (1980)336-3453 5 . [21] MORRISON, R., The results of treatment of cancer of the bladder - a clinical contribution to radiobiology, Clin. Radiol (19756 .2 ) 67-75. [22] BATTERMAN, J.J., HART, A.A.M., BREUR , Dose-effecK. , t relation r tumousfo r control and complication rate after fast neutron therapy for pelvic tumours, Br. J. Radiol. 54 (1981) 899-904. [23] MOORE, J.V., HENDRY, J.H., HUNTER, R.D., Dose incidence curve tumor sfo r controd an l normal tissue injury in relation to the response of clonogenic cells, Radiother. Oncol. 1 (1983) 143-157. [24] HANKS, G., MARTZ, K.L., DIAMOND, J.J., The Effect of Dose on Local Control of Prostate Cancer, Int . RadiâtJ . . Oncol. Biol. Phys (19885 1 . ) 1299-1305. [25] WILLIAMS, M.V., DENEKAMP , FOWLERJ. , , J.F., Dose-Response Relationshipr fo s Human Tumours: Implication Clinicar sfo l Trial f Dosso e Modifying Agents, Int . RadiâtJ . . Oncol. Biol. Phys. 10 (1984) 1703-1707. [26] BENTZEN, S.M., THAMES, H.D., OVERGAARD, J., Does variation in the in vitro cellular radiosensitivity explai shalloe nth w clinical dose-control curv malignanr efo t melanoma?, Int. J . Radiât. Biol (19907 5 . ) 117-126. [27] YAES, R.J., Tumor Heterogeneity, Tumor Size and Radioresistance, Int. J. Radiât. Oncol. Biol. Phys. 17 (1989) 993-1005. [28] BRENNER, D.J., Dose, volum tumour-controd ean l prediction radiotherapyn si , Int . Radiât.J . Oncol. Biol. Phys. 26 (1993) 171-179. 39 [29] ZAGARS, O.K., SCHULTHEISS, T.E., PETERS, L.J., Inter-tumour heterogeneitd an y radiation dose-control curves, Radiother. Oncol (1987.8 ) 353-362. [30] MAYLES, W.P.M., CHOW , DYERM. , , FERNANDEZJ. , , E.M., HEISIG , KNIGHTS. , , R.T., MOORE, I., NAHUM, A.E., SHENTALL, O.S., TAIT, D.M., The Royal Marsden Hospital pelvic radiotherapy trial: technical aspects and quality assurance, Radiother. Oncol. 29(1993)184-191. [31] CHEN, G.T.Y., Dose volume histograms in treatment planning, Int. J. Radiât. Oncol. Biol. Phys. 14(1988)1319-1320. [32] INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS, Prescribing, Recording Reportind an , g Photon Beam Therapy, Repor , ICRU50 t , Bethesda, (1992)D M . [33] BRAHME, A., ÂGREN, A-K., Optimal Dose Distribution for Eradication of Heterogeneous Tumors, Acta Oncol. 26 (1987) 377-385. [34] TATT, D., NAHUM, A.E., RIGBY, L., CHOW, M., MAYLES, W.P.M., DEARNALEY, D.P., HORWICH , ConformaA. , i radiotherap pelvise th f yo : assessmen f acuto t e toxicity, Radiother. Oncol. 29 (1993) 117-126. [35] 'BURMAN , KUTCHERC. , , G.J., EMAMI , GOITEINB. , , FittinM. , f normao g l tissue tolerance data to an analytic function, Int. J. Radiât. Oncol. Biol. Phys. 21 (1991) 123-135. [36] radiatioGOITEINn i utilitT e C Th f n y, o therapyM. , estimatn a : outcomef eo , Int . RadiâtJ . . Oncol. Biol. Phys (19795 . ) 1799-1807. [37] WEBB, S., The effect on tumour control probability of varying the setting of a multileaf collimator with respec plannine th o t g target volume, Phys. Med. Biol (19938 .3 ) 1923-1936. 40 XA9642844 DEFINinO TREATMENF NO T GEOMETR RADIATION YI N THERAPY P. AALTONEN Finnish Centre for Radiation and Nuclear Safety, Helsinki, Finland Abstract When accurate system qualitr sfo y assuranc treatmend ean t optimizatio employede nar precis,a e syste fixatior mfo d nan dosimetri portad can l verificatio s importana e n ar continuea s a t standardized dan d cod f practiceo dosimetrr efo d yan patient follow-up, including registration of tumor responses and acute and late normal tissue reactions. To improve the accuracy of existing dose response relations in order to improve future therapy the treatment geometry and dose delivery concepts accuratelhave b o et y define uniformld dan y employed Nordi.A c workin 199n i Nordiy p 1u (b gt grouse c s pwa Associatio f Clinicano l Physics standardizo )t concepte eth quantitied san s used durin whole gth e radiotherapy process in the Nordic countries. Now the group is finalizing its report "Specification of Dose Delivery in Radiation Therapy". repore Th t emphasizes tha treatmene tth t geometry shal consistene b l t wit geometre hth y used durin diagnostie gth c work up. The patient fixation is of importance early in the diagnostic phase to ensure that the same reference points and patient position wil e useb l d both durin diagnostie gth c wor , simulatiokup treatmend nan t execution. Reference Coordinate System of the patient is a concept based on defined anatomic reference points. This Patient Reference System is a local system whic s validit tissuese hha th r yfo , organ volumed san s defined during radiotherapy reference .Th e pointe th f so Patient Reference System should in turn be used for beam set-up. The treatment geometry is then defined by using different concepts describing tissues which are mobile in the Patient Reference System, and finally, volumes which are fixed in this coordinate system. A Set-up Margin has to be considered for movements of the volumes defined in the Reference Coordinate System of the Patient in relation to the radiation beam. The Set-up Margin is dependent on the treatment technique and it is needed in the treatment planning procedure to ensure that the prescribed dose to the Target Volum delivereds ei . . INTRODUCTIO1 N performance Th externaf eo l beam radiation therapy accelerator othed san r radiation therapy device bees sha n developed considerably durin decadelaso e gth tw t s with regarqualite th d o dt yan precisio beame th f no s throug targew filted hne an t r designs, improved stabilit acceleratorse th f yo , increased flexibility in beam collimation systems and compensation techniques and improved dosimetri geometrid can c treatment verification methods powerfuw Ne . dimensina3 l l diagnostic equipment has been developed starting from computed tomography to magnetic resonance imaging and spectroscopy. Simultaneously computerized treatment plannin dosd gan e delivery optimization methods have been developed considerably. Partly due to the inaccuracies in basic definitions one of the weak links in this development has been the way we define our target volumes and specify and prescribe the dose delivery. An investigation amon Nordie gth c radiotherapy centre 199n si 1 confirmed that inconsistent use of dose and volume concepts is jeopardizing the high standard of radiation therapy [1]. A Nordic Working group was set up by NACP to standardize the concepts and quantities used throughout the whole radiation therapy process. Now the group is finalizing its report "Specification of Dose Deliver Radiation yi n Therapy maie th nf " o subject [2]e definitioe .On th s si treatmenf no t geometry in radiatio bees nha recommeno n t therapyconceptf m o ai e e us e Th .s dth base recenn do t scientific developement in the field of radiation therapy which are needed for the developement of daily clinical practice. 41 The principal aim the draft report is to treat the situation at clinics with state of the art equipmen proceduresd an t r obviouFo . s reason NACe sth P repor alss i t o written primarily wite hth fairly uniform equipment situatio Nordie th n ni c countrie mindn si . However r firmou s i belie t i , f that once general high quality procedures have been develope advancer dfo d equipment the alsn yoca be transferred and adopted to more traditional equipment once the basic underlying principles have been developed. definitionMane th f yo s introduced have obvious counterpart classican si l radiation therapy procedures, even though they are not always coinciding with all established methods since some new proposals have had to be made for new irradiation techniques. These proposals have evolved in discussions among radiotherapists and physicists in the Nordic countries during the last five years f coursO . e they have also been considerably influence discussiony db s with word an ,f ko (ICR [3]0 Uinternationae 5 th ) l radiation therapy community. . DEFINITIO2 REFERENCF NO E POINT TREATMEND SAN T GEOMETRY followine Inth methode gth f defininso treatmene gth t geometry suggeste NACPe th y db - draft repor presentede ar t . 2.1. Reference point alignmend san t markings 2.1.1. Reference coordinate system patient ofthe The concept of a reference coordinate system of the patient is based on defined anatomic reference points e patienTh . t reference syste mlocas i l system whic s validittissuese hha th r yfo , organs and volumes defined during radiotherapy. There does not exit a general patient reference system, sinc humae eth n rigidt bodpatiene no s Th y.i t reference syste mdefines i e th df o wite hon anatomic reference points and the other reference points (or markings on the skin) are for orientation systee oth falignmen d man patient e th f o tcervi a . Thir fo illustrate xs si 1 cance g Fi rn d i patient . Note that alignment markings for patient set-up can not always be firmly connected to the reference point se symphysi sucth s sternae ha th r so l notch e coordinatTh . e system with external reference points should therefore preferabl usee bear yb dfo m set-u orden i p havo t r e"mora e rigid" relation to the target volume than the more uncertain skin mark often used in radiotherapy. The reference points are preferably defined already during the diagnostic stage so that they can be used for a l imaginal n cohereno gp treatmenu e modalitie t th se tn o d t san machines reference .Th e points will then be able to work as markers for image matching and fusion and to form an accurate integrated diagnosti plannine basth a r cs e a fo dat gt procedurease . During dose plannin reference gth e points locatioe useth e r ar definitiod fo n an isocente e th f no r relativ reference th o et e poin defines a t d dan indicated on the dose plan. The tissues, organs and volumes delineated for radiation therapy planning should be defined in relation to the reference point of the patient coordinate system. These reference points should in tur usee nr beab f usindfo o mm gset-up ai referenc e Th . e point s thasi definitioe th t targee th f tno volum bead ean m sam e set-uth ed locapan e referl on coordinat o st e system clinican I . l practice eth reference point should be located on the surface of a bony structure, or on the skin close to bony structures reference Th . e point should possible b ) 1 : visualizo et simulaton eo verificatior o r n films and beams-eye-view srigis a possiblplotss e da b ) 2 , relation e i targe e th o ntt tissues locatee b ) 3 , d as clos s possiblea targee th o tet tissue orden i s minimizo t r e beam set-up patieno errort e du st misalignment. Unfortunately rigia , d connection betwee targee nth t reference tissueth d san e point s rari e situation. This implies tha anatomican a t targee e addelb th margio o t tdt tissues nha o st account for the movements of the tumor in relation to the reference points when specifying the target volume. 42 2.1.2. Internal reference points e internaTh l reference point locatee sar d insid bodye e th interna n .A l reference poin uses i t d for beam simulatore set-uth n o p , befor firse eth t treatment internae .Th l reference point should thus selectee seee b b diagnosti n no n thao ca ds t i t c radiograph simulatore th t sa . This make t possiblsi e to hav ea ver y accurate beam simulatore set-uth t a p t als bu tako ,ot e simulator film whico t s h treatment unit verification films can be compared. 2.1.3. External reference points External reference point palpable sar visiblr eo located surfac e bode ean th th r n yo f do eo e surfaconth f fixatioo e n devices tha t closelexterioe fi tbode th th o yt yf o r(e.g . facemaskd an s shells) e externaTh . l referenc palpablee b poin y tma e bony structure skia , n markin alignmenn a r go t tattoo preferably where the skin is tight over a bony structure as for example on sternum. The external reference point is used for beam set-up both at the simulator and the treatment unit. The external reference point e normallar s y palpable bony structures whic e e easth har findo r t y Fo . extremitie largse th thi ef trune s o bones th wild r cen l pelvie pointFo e .typicall th th n t ca o s e yb bones can be used for the lower part, and the sternum for the upper part. For the head the lower point of the mandibula and upper point of the nose can be used for the sagittal plane, and the ears laterae foth r l points. 2.1.4. External reference systems e externaTh l reference surface pointth n f fixatioso o e n devicedevelopee b y e th sma n di forlocaa f mo l stereotactic syste defind m an externan a e l reference coordinate syste whicmn i e hth target volum s describeei definedd dan . Several stereotactic system s bees ha head e n th use r . dfo Reproducible systems for the abdomen has also been used. It is important that the same external referenc treatmen e simulatoe th th t t a ea d , systetran unit CT coordinate e e muses i Th .th t da e th f so isocenter can then be defined in the external reference system during dose planning. Similarly, the internal and external anatomic reference points define the local coordinate system of the patient in which the target volumes and organs at risk should be delineated. 2.2. Treatment geometry e followinTh g definition e mad ar so genera s e l that they pertain bot o curativt h d an e palliative treatments. For curative therapy the terms target tissues or target cells can be replaced by the clonogenic tumor cells as normal tissues are not generally the target for the treatment. However, the target volum econtai o oftet s nha n normal tissue ensurr so curativea e tumol dosal o et r cellsr .Fo postoperativa e treatmen grosdefinitioo e tn th targese leftd e tumob th an ,f y nto tissuerma s consists of the remaining microscopic disease. In some regions, e.g. in the head, the target tissues may be delineated partl osseouy yb s barries, partl surfacesy y b clinica n o r o , l grounds include areas with known probability of métastases. Thus, by the target tissues is understood the tumorous tissues with a sufficiently high probability of tumor cell spread to be considered for radiation therapy. DEFINITIONS: Gross Tumor The Gross Tumor consists of solid demonstrable malignant tissues in the patient (see Fig. 1) and it is often mobile in the local coordinate system of the patient. 43 Verified Disease The Verified Disease include demonstrabll sal e macroscopic malignant tissue patiene th n si t (see Fig. 1). The verified disease includes all the Gross Tumor and verified nodes and it is often mobile in relation to the local coordinate system of the patient. Presumed Microscopic Disease The presumed Microscopic Disease (see Fig hig) containa 1 .s h ha ris r f containinso ko g clonogenic malignant cells to be eradicated. It is often mobile in the local coordinate system of the patient. Target Tissues The Target Tissues contai l verifienal d and/or presumed diseas treatee b prescribea o e t o dt d time-dose pattern (see. Fig. 1). To be more precise in radical radiotherapy the target cells are the clonogenic tumo rgrose cellth f ss o tumo associated an r d microscopic disease Targee Th . t Tissues oftee ar n mobil locae th n lei coordinate syste patiente th Targef me o Th . t Tissues, when treatinn gno malignant disease, will include benign tissue treatee b o examplr st dfo e wit palliativha e intent. Anatomic Margin The Anatomic Margi a margi s ni n aroun Targee th d t Tissue accouno t s r expectefo t d movements and/or changes of shape and size of those tissues or cell structures in relation to the reference points in the patient. The anatomic margin also contains possible uncertainties in micros- copic spread. The outer boundary of the Anatomic Margin specifies a fixed volume in the local coordinate system of the patient. Target Volume The Target Volum anatomicalln a s ei y defined volume coordinat e fixeth n di e systee th f mo patient, which contains or has a high risk of containing tissues or cells to be treated to a prescribed time-dose pattern. This volum s defineei enclosed outee an dth y r db boundar Anatomie th f yo c Margin. The Target Volume is therefore specified in relation to internal and external anatomic reference points (see Fig ) whic.1 h preferably shoul rigidle db y relate eaco dt h other through bony structures. Organs at Risk The Organ t Ris sa normae kar l tissues whose presence influence treatment planning and/or dose prescription targee . Likth r te fo volume locatioe organe th , th f nt riso sa k shoul definee db n di relatio anatomicae th o nt l reference points. Treated Tissues The Treated Tissue e tissuear s s enclose isodosn a y db e surface cumulateth n i e d dose distributio e patienth n i nt being representativ r tumofo e r eradicatio r palliatioo n n (e.g. 0,95x prescribed dose). 44 Beam Edge £50) %D External Anatomic Set Up Penumbra Region Reference Point Margin 1 m-80% Beam Anatomic Reference Internal Extema Fig 1. Illustration of the different anatomic reference points, margins, and volume concepts for an advanced cervix tumor internae .Th l anatomic reference point essentiae sar simulationr fo l s and portal verification wherea e externath s l reference point e primarilar s y intendeo t d improve the precision in patient and radiation beam set-up. 45 Irradiated Tissues e IrradiateTh d Tissue tissuee ar s s which receiv ea dos e tha s considerei t d significann i t relation to the biological effects of normal tissues. Cold Region ColA d Regio voluma s ni e insid Targee eth t Volume, which receive dossa e lower thae nth prescribed Targe e dosth o et t Volume r quantificatioFo . n purpose recommendes i t si d thadose th t e quoted should be the mean value in a volume of 0.1 cm3 or less. Hot Region A Hot Region is a volume, which receives a dose larger than the prescribed dose in the Target Volume. For quantification purposes it is recommended that the dose quoted should be the mea t largerr Regiono valuHo 3 voluma A .cm n n ei 2 outsid f eo targee eth t volum oftes ei n called t SpotaHo . Set-up Margin The Set-up Margin is a margin for movements of the Target Volume or Organs at Risk in relatio radiatioe th o nt n beam Set-ue Th . p Margi dependens ni treatmene th n o t t techniqus i t i d ean needee treatmenth n di t planning procedur r examplefo ensuro et e thaprescribee th t de dosth o et Target Volume is delivered and the dose to healthy normal tissues is as low as possible. This margin ha accouno st uncertaintier fo t patien) 1 n si t positioning (interfractional movements) ) movement2 , s of the patient during each treatment fraction (intrafractional movements), 3) dose planning and treatment techniqu generan e) treatmeni 4 d an l t unit performance characteristics. RECOMMENDEE TH CONCEPT. E 3 TH F O E SDUS To deliver the right dose distribution to the target tissues would be no great problem provided there were 1) no uncertainly in microscopic tumor spread, 2) no positional uncertainties due to motion f internaso l tissues uncertainto n ) 3 e alignmen, th patiene n yi th f o tt wit therape hth y beams and finally, 4) no uncertainty in the delivered dose distributions. All these four categories of uncertainties decrease the probability of achieving complication free control of the tumor growth, especiall t accounteplannine no f they th i e n i yar r gdfo procedure. Obviously, the best thing would be if one could eliminate as far as possible the positional and set-up uncertaintie gooy sb d fixation techniques, possibly combined with synchronizatioe th f no irradiation with breathin other go r internal motions. However uncertainte th , microscopin yi c spread s veri y har eliminato dt patieneo t bote hdu t individua finite l th pattern eo t e f spreadu so d dan resolutio diagnostie th f no c methods firse th t n approximatioI . woule non d thin does k i mattet sno r much whethe uncertainte rth locatioe targee th th n y i f nuncertaintieto o t tissue e du s si microscon si - tumoc pi r spread, organ motions radiatior ,o n beam set-up. However, accurate patienrequirep u t tse s f externao e theus l reference point thud san s separatio f internano l (organ motion r microscopiso c spread) and external (set up) uncertainties. 46 When the target volumes and organs at risk have been accurately delineated relative to the reference points the dose delivery technique has to be considered. If there is no reason to expect different sensitivitie verifiee th r sfo d gross diseass presumeit d ean d microscopic extensioe th d nan tumor cell densities are not too different, a single target volume and a uniform dose delivery may sufficiente b . However verifiee th f i , d gross diseas contaiy ema n more resistant cell compartments such as hypoxic tumor cells, and/or if the density of tumor clonogens is considerably lower in the sub clinical region, different dose level generally ma s e desirableyb r more o definitio o Th .e tw f o n distinct target volumes is then called for and the most suitable dose level for each must be specified. REFERENCES [1] P. AALTONEN: "An inventory of dose specification in the Nordic Centres and a suggestion fo standardizea r d procedure Proceedingn "i Interregionae th f so l Semina Radiation ro n dose in radiotherapy from prescriptio deliveryo nt , organize IAEAe th y db , IAEA-TECDO4 C73 (1994) 83-89. [2] NORDIC ASSOCIATION OF CLINICAL PHYSICS: Specification of Dose Delivery in Radiation Therapy, Draf-repor Nordie th y b ct Working group (1994). [3] INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS, Prescribing, Recording and Reporting Photon Beam Therapy, ICRU Report 50, Bethesda (1993). 47 DOSIMETRIC PRECISION REQUIREMENTS AND QUANTITIES FOR CHARACTERIZIN RESPONSE GTH TUMORF EO NORMAD SAN L TISSUES A. BRAHME Radiation Physics Department, ""XA9642845 Karolinska Institute, Stockholm, Sweden Abstract Base simpln do e radiobiological model distributioe effece th s th f o t absorbef no d dosn ei therapy beams on the radiation response of tumor and normal tissue volumes are investigated. Unde assumptioe rth n tha dose tth e variatio treatee th n i d volum smals ei l it is shown that the response of the tissue to radiation is determined mainly by the mean dose to the tumor or normal tissue volume in question. Quantitative expressions are also increasee giveth r nfo d probabilit normaf yo l tissue complication decreasee th d san d probabilit tumof yo r contro functioa s la increasinf no g dose variations -aroun meae dth n dose leve theso t l e tissues. Whe dose nth e variation large minimusar e eth m tumor dose siz3 (toecm volumes) will generall bettee yb r relate tumoo dt r highese controth d tlan dose to significant portions of normal tissue correlates best to complications. In order not to lose more than one out of 20 curable patients (95% of highest possible treatment outcome requiree )th d accurac dose th en y i distributio n delivere targee th o dt t volume should be 2.5% (la) for a mean dose response gradient /in the range 2-3. For more steeply responding tumors and normal tissues even stricter requirements may be desirable. . 1 RADIOBIOLOGICAL MODEL 1.1. Statistics of tissue damage The classical theories for the killing of a cell assume that single or multiple hits are necessary in one or several targets in this cell. The targets are often considered to be locate cele th l nucleusn di achievo .T efaia r agreement between experimental resultd san theory rather complex mult multt ihi i target combinations havusede b o et . Furthermoret ,i is also known today that the curvature of the dose response relation is mainly caused by repair processes and to a lesser extent to target structure. For this reason the mathematically more simple semi empirical linear quadratic expression has been used more extensively during recent years for comparison with experimental data. However, when the survival of a certain organ or a tumor is considered the single hit multi target model may still be of interest as illustrated schematically in Fig. 1. The N targets are now interpreted as the functional sub units or clonogenic cells making up the e tumororgath r no , insteavarioue th f do s subtarget cele th l nucleun si s previously assumed to cause cell death when being hit. For simplicity it is also assumed that each clonogenic cell is inactivated when an ionizing particle hits the sensitive area of the cell, as quantified by the inactivation cross-section 49 ABSORBED DOSE: Total Inactivation number j- cross-section of cells 0 per cell Total Mea t numbenhi r cross-section per cell /V(TJ0(/) Mean number of hits in organ Fig. 1 The model used to calculate the response of an organ consisting of a large number of individual cell eac) s(N h wit inactivation ha n cross section (GQ/cm cell)r 2pe , when exposed to a radiation beam which is specified by its fluence ( fluence wit meae hth n restricted collision mass stopping powe electroe th rr L*fo pnl spectrum, in question. At a dose level of 1 Gy the fluence is typically 3-10^ electrons per cm^, assuming a mean stopping power typical for high energy electrons and photons of about 2 MeV cm^g-l. Similarly, the density of ionizing events may be calculated based on a mean inactivation event uni sizn i abouf etV o e densit 0 t6 densitye materialth y yG 1 t .A becomes 10^4 events per cm^. This corresponds to a mean distance between events of about 0.2 fim in agreement with the fluence value just calculated. When the inactivation cross section and the initial number of cells are known the mean cell survival as a function of absorbed dose or fluence can be calculated somewhat in analogy wit nucleae hth r reaction probabilit irradiaten a n yi d medium. Sinc probabilite eth a r yfo given cell to be hit by a given particle is extremely small but the beam consists of a huge number of particles, the mean hit number is finite and the problem is ideally_suited for Poisson statistics [1,2]. Thus the mean number of inactivated clonogenic cells, dN, due to a fluence increase, d J (1) 50 differentiae Thith s i s l equation resultin traditionae th n gi l singl t exponentiaehi l cell survival: ) (2 oe-OQN - = QC-D/DQN N ^= where ^is the mean number of surviving cells and NQ_is the initial value. The relation betwee inactivatioe nth Dd Oan n cross-section, OQ, usin gives Oi = p Ll gD y nb (3) A more detailed discussion on the dependence of OQ on radiation quality is given by Zaide Rossd an r i inactivatioe [3]Th . n cross-sectio eacr nfo h cel vers i l ye smalth d an l fluence of particles very large so that the product of them, the mean hit number per cell, CTQ Ph(v)= Thus the probability for no hits, Ph (0), or the probability that a given cell survives, Ps, becomes: D/Z />s = Ph(0) = e-^=e- e (5) The probability tha a singlt e cel s killei l r dmor o (i.ee e on .hits s therefori ) e Ph(0)- 1 P e= . Provide killine d th cel a statisticallf s gli o y independen whaf to t happeno st every other cell, the probability that a tissue or tumor consisting of NO cells is completely eradicated by killing all its NO clonogenic cells is given by the conditional as expressed by O (6, The survival probability for a tissue consisting of NO cells is thus I - Pe which is recognized as being mathematically similar to the traditional single hit multi target survival curve equation. If we again apply Poisson statistics a very useful alternative expression for the probability of eradication of a given organ can be derived from the probability of having precisely v surviving cells: eNNv ) (7 P(v)= where N is the mean number of surviving cells as given for example by Eq.(2). From this expressio probabilite nth survivalso n f yo , eradicatioe Ps(0)th r o / n probability, Pe, becomes simply: 51 This expressio verns i y closely relate Eq.(6o dt ) particularl higt ya h dose levels sucn o s ha over the sigmoidal part of the dose response curve. Eq.(8) always gives a slightly larger value than Eq.(6) as can be demonstrated by power expansion of Eqs.(6) and (8) for dose values both large smalled an r r However. thanDQ more th , e basic Eq.(6) shouln di principle be more accurate, but the difference is never clinically significant for large values of NQ. Due to the greater mathematical simplicity, Eq. 8 has been extensively used over the year accuratelo st y describ shape dose eth th ef e o respons e relation[l. ,4] 1.2. Cell survival curve Fo givera n cell populatio vitrn i relationshi e oth p betwee survivine nth g cell fraction, S , anabsorbee dth d dose deposite singla n di e irradiatio describes ni cele th l survivay db l curv above e (cfTh . . e2) Figequation. derivel al approximatioe e th sar n di single th f neo event cell kill without consideration of repair mechanisms. The simplest way to generalize the above equations to take such processes into account in an approximate manner is to replace DO by De the effective DO {4,5]. An even better fitting of experimental cell survival data over widea r rang f doseeo obtaines si simple th f di e exponential cell survival Eq.(5n i 0 replaces )i more a P= y d b e precise expressio equationn i (8)d .san lik) (6 e shapcele e th lTh f survivaeo l curve takin effece g th repaif o t r processes into accouns i t generally very well described by an equation of the type : The coefficient a of the linear term in absorbed dose determines the slope of the survival curve at small doses. The coefficient ß of the quadratic term is a measure of the shape of shouldee th survivae th f o r l curve collectioA . survivaf no l parameter dat humar afo n tissues have been published by Thames and Hendry [6]. An approximate but more simple cell survival curve mode giveexpressios ln i a y n b typ e th ef no (10) where / is the extrapolated cell fraction at zero dose assuming the slope of the almost linea celre parth l f survivao t l curv logarithmia n eo c e constantscale b th s o ei t O D . absorbed dose that reduces the proportion of surviving clonogenic cells to e'l around the mean dose of interest D (see Fig. 2). This relation has the advantage of being more accurate tha ns stili Eq.(5t i l t lineabu ) dos n i rthereford ean e valid als fractionater ofo n no d uniform dose delivery with the total dose as sole variable. The parameters DO and/can be related to a and ß through simple relations if, for the dose range of interest, the logarithm of the cell survival around the mean dose, D can be approximated by a straight line. The shape of the survival curve as determined by the above factors will in the first approxi- mation describ response eth cela f elo syste orgamr o fractionateno t d irradiation. 1.3. Dose response relation shapcelBasee e th th l f survivan edo o l curve afte singlra e irradiatio vitrnn i response oth e o tumoa orgafn a r multiplo no rt e irradiation estimatede b viv n si n oca detaileA . d analysi quits si consideo et comples capacite ha t rrestinth i e s xth a f yo g cell enteo st e rth cell cycle, the growth delay of the different phases of the cell cycle, the efficiency of repair processes and influence of oxygénation and nutritional factors, the condition of the vascular system and the radiation modality used (e.g. particle type, energy and dose rate). For the present purposes only the gross radiation effects are included' in order not to complicate the mathematics. 52 2.00 0.01 D/Gy Fig. 2 Comparison of different cell survival curve models. The solid line curve is the linear quadratic relation wherea dashee sth d line correspond tangene a th o st d tan dash dotted a secant through the mean dose per fraction. Usin e lineagth r quadratic mode describo t l e cell survival afte totaa r le dosth D e probability to control a tumor is given by: P(D)= (ID where NO is the initial number of clonogens. This model makes it possible to accurately predic effece th tdosf o t e fractionatio replaciny dose n b y fractionr th b e, pe y b n gD D/ , writing: (12) 53 To illustrate the effect of the fractionation, the tumor control given by Eq.(12) is shown in functioa s totaFige a dosth e 3 .f r l fraction th no dos e pe d ean thin I . s diagra sees i mt ni totae th l w dosho e require controo dt tumoa l reduces i r dose r fractioth epe s da s i n increased. When the dose per fraction is decreased and the number of fractions is increased the response surface becomes very smooth with minimal effects of individual dose fractions. The general curve shape in Fig. 3 is therefore determined by the effective mean cell kilr dospe l e fraction cele lTh . surviva oven ca l clinicallra y relevant dosr epe fraction interval aroun accordinn dDi Eq.(lOo gt accuratele b ) y approximate: dby D 5=e n (13) n nDf Din /Gy 0 0 80 0 Fig. 3 The variation of the probability to control a rumor consisting of 108 clonogenic cells with the total dose and the dose per fraction at an a/ß = dosd Od e . fractionGy 5 shadee sar illustrato dt effece e th fractionation f o t s i t I . see dosne fractior thath epe s ta n decrease isoeffece sth t dose increases solie d.Th parallel and oblique lines correspond to a fixed dose per fraction of 2 Gy and a fixed numbe fractionf ro s (12), respectively. 54 In order to make both the cell survival and the slope of the dose response relation equal at D=D, the effective extrapolation number,/, and the effective DO, De, are given by: (14) (15) n Here D is the mean dose in the organ in question and a and ß are the cell survival parameter lineae th f sro quadratic model advantage .Th above th f eo e transfor thams i t i t will hold quite well also for non uniform dose delivery as discussed by Brahme [4] since it is linear in the total dose, D. This is illustrated by the two intersections in the dose response surface in Fig. 3. The dose response curve cut out by a vertical plane parallel to totae th l dose axis correspon constana o dt t dos jusf e o fractionr t abovpe y G e.2 However, oblique th correspondt ecu fixea o st d numbe dosf ro e fractions delivere targee th d do t an has the important property of being practically independent of dose distribution fluctuations in the dose range of ±25 per cent around D (cf. Fig. 2 and [4]). Using the above notation, Eq.(12) may be simplified according to where for simplicity the effective clonogen number, / nNQ, is denoted Ne- Eq.(16) is a simpl d usefuean le dos forth ef m o respons e relatio r tumonfo re b control n ca t I . generalized to describe the radiation effect in any tissue by expressing the variables Ne and D normalizeterme n ei th f so d dose respons edose gradientth ed causinan , y % , g50 probabilit effectf yo , assumes i DSQ t i f .I d tha numbee tth fractionf ro constans si dose tth e response gradient at the inflection point becomes simply: (17a) In Figconstana 4 . t dos r fractioepe s assumenwa d instead correspondine Th . g dose response gradient is slightly lower as given by (d£| =_L_.2Unf (17b) > eZ Q eD \dD!D Thus, at the steepest part of the dose response curve a dose increase of DO increases the tumor control by about 37 per cent. However, the dose increase should preferably be measured relative to the total dose when effects of dosimetric uncertainties are investigated morA . e important quantity with regar precisioo t d n requirementn i s radiation therapy is therefore the normalized dose response gradient, y, defined by: (18) This paramete dimensionlesa s ri s number which describe largw sho change a tumon ei r control probabilit expectee b o givet a ys r i dfo n relative increas absorben ei d dose facn .I a t 55 P(D) 1.0 N(0)- 108 Q 50 D/G Fig. 4 The shape of the dose response relation given by Eq. (9) for 108 and 109 tumor cells dottee .Th d lincumulativa s ei e normal probabilitye curvth o t ee fitteey y db larger cell population. dose increase of one per cent on the linear part of the dose response curve will result in an increas tumoen i r control probabilit preciself yo y /per cent. Unde assumptioe rth n thae tth number of fractions is constant and that only the dose per fraction is changed the normalized response gradient at the inflection point, )n, becomes: luN +n In/ , __c___ (19a) If instea e dos r dfractioth epe keps i n t constan normalizee th t d response gradiens i t reduced simply to: lnA (19b) These expressions show, under the present simplifying assumptions, that for uniform tumors the normalized gradient increases logarithmically with tumor size. It is also seen that a fixed number of dose fractions results in a steeper dose response curve as also illustrated in Fig. 3. 56 If we replace DC by DSQ the uppermost exponent of Eq.(16) has to be equal to ln(ln(2)) when D is equal to DSQ, thus Eq.(16) may be rewritten as = 2_e[re-ln(ln(2)J Eve thif ni s equatio derives nwa tumor dfo r controthin i ss i genera t li l form with /and DSO as only clinical variables, also applicable to normal tissue injury, as discussed in detail by Källman et al. [7]. From Eqs.(14) and (15) combined with Eqs.(19b) and (20), respectively, it is possible to derive how /and t>50 will vary with the number of fractions, n: 2 _lnN ßD+ e /n ' e lnNe D50 —— a+2ßD,n From Eq.(21 cleas i t unit )i increas w ry tha fe valuse ma a whe7 tth y f eb e o n going froma constant dose per fraction to a constant number of fractions (cf. Fig. 3). It is also clear how the DSO value increases with increasing number of fractions or decreasing dose per fraction. Unfortunately increasee th , valud7 e wit constanha t fraction number wilt no l generally make the therapeutic window wider even though the same effect will influence both tumor normad san l tissues. Thi becauss si difference eth e betwee DSe nth Q values will decrease simultaneously. 2. INFLUENC ABSORBEF EO D DOSE DISTRIBUTIO LOCAN NO L TUMOR CONTROL AND NORMAL TISSUE COMPLICATIONS 2.1. Fundamental considerations In the preceding section it has been assumed that the absorbed dose distribution was perfectly uniform over the entire tissue volume. In practice the dose distribution is seldom unifor d differenman t orgapartn targea a f r sno o t volum alsy oema have different sensitivities. Variations in local sensitivity and delivered absorbed dose distribution will influence the tissue effect in a similar way during fractionated therapy since local variations in the surviving fraction after each treatment are multiplied to give the net effect of a complete series of treatments. dependence Th probabilite th f eo radiatioa f yo n effectchanga absorbee n o th n ei d dose unifore th n i leve mn caslca e easil expressee yb d under assumption tha dose tth e variation sufficientla fall n o s y linear portio dose th ef no respons e curve increase Th . effecn ei t probability due to a dose increase AD may then be approximated by the first few terms of the Taylor expansion: 7 P(Z= P(+ »D +AD)D P(Z>= +A ) (23) doswhere th es responseyi e gradient denne Eq.(18)n di . 57 Let's first look at a tumor. Assume that the initial clonogenic tumor cells, N, can be , dividef equapopulationo o , tw l n Nb di sensitivity d an f cellsso a N , , irradiateo dt somewhat different respectively, dosDb e d levelsan controe a ,D . Th l probabilit eacf yo h cell population is then given by an expression of the type: (24> where S(Da,i) is the survival after each dose fraction, Da/i, to the population a, and with £ £>a,i = #a (compare Eq.16). The probability to control the total cell populatio conditionae gives ni th y nb l probability tha, controlles i ,P tb d whe knows i na n controllede b o first a s ta , approximation.If assumes i t e i , ar b d d tha actione an th ta n so statistically independent, P is expressed by: P = Pa Pb (25) which may be rewritten: n (26) This expression is in agreement with Eqs.(8) and (16) as the terms inside the square brackets give the expected mean number of surviving cells. If, for a moment, it is assumed that the dose delivery is the same for both populations Eq.(24) may be rewritten: PPN=Pma= (27) where the last step is based on the assumption that the cell mass is constant throughout both cell population respectivel, mb massef d so an a ysm (ra=ma+mb) sees i * I n. that this expression is consistent with Eqs.(25) and (26) because ma/ra+rab/wi=l. This is an explicit expression showing generan i , l agreement with Fig , thados.e 4 th t e response curvs ei shifte higheo dt r control rate lowed san r numbedosee th s s a cellsf ro tumo e ,th r massr ,o the volume decreases. Under assumption of a constant tissue density a related expression was used by Goitein [8] and Schultheiss et coll. [9]. effece overTh f o tunded 2.2-.an r dosage increase controe Th th f eo l probabilit whole th f yo e tumor whee th nfractiof a o m D n tumor masreceivins i sm expresse e excessivn b g a w no combinin y dn b e ca dos D eA g Eqs.(23) and (27) ydm/m (28) 58 which may be rewritten: m r P(D) , (29) If it is now assumed that AD/D and Am/m are so small that a power expansion is valid the change in control probability may be approximated by the first order term given by: (30) D ™ ë which clearly illustrates how dose and mass errors combine as an integral dose error in the first approximation. The second equality is based on the definition of the related quantity: the mean energy imparted, e, [10]. In the first approximation it is thus the relative changes fro desiree mth d mean energy imparte tumoe th n di r volume that alte e controth r l probability. This resul importans ha t t consequence dosr sfo e specificatio radiation ni n therapy. If the dose variations are not too large the mean absorbed dose to the tumor volume should be closely related to the therapeutic effect. The mean absorbed dose is defined by the expression: f D(r)dmD(r)dm( \ D(r)p(r)aV 5.I.L——.1 m i f (OL)(31) / p(r)dV wherabsorbee th e er(r)d ar D(r) an d dos densitd tumoe an s i th f positioyt V o ra d d an nr the differential volume element The use of D will make the increased and decreased cell d colan d t killinareasho n i g, respectively, compensate eac he firsotheth t n i r approximation tumoe part th . becausf Thio ho to s r e s volumsi th n ei e Jhe erro mean ri n energy imparted, Ae^ larges i r than zero wherea cole th d n sportioi n Aeless i s than zero and according to the definition of mean dose Ae = Ae +As=0 .In the first approximation according to Eq.(30) AP is therefore equal to zero and the tumor control should be quite close to that expected for a uniform dose distribution. When the density of the tumor, p(r), is constant Eq.(31) may also be expressed as a volume average: _ f D(r)dV J f (32) I dV The simple "two volume case" treated in Eqs.(28) to (30) may serve as a clear illustration as thiDn i s cas gives ei : nby ^ Am AD\ (33) By comparing this expression with Eq.(30 sees i t )ni tha change tth contron ei l probability is in the first approximation proportional to the change in mean dose to the tumor volume. It can thus be concluded. that when the dose variations in the beam are small (AD«D or 59 more exactly when Ue Lil= e< uniforn No 2.3m. dose distribution generan si l Under assumption tha meae tth n absorbed dose accordin Eq.(31go t uses )i referencs da e fino t mora dy tr ewile w generaw no l l expressio controe th r nfo l probability. This si possible by generalizing Eq.(29) in such a way that it holds for arbitrary dose distributions b numbeye th allowin d masf an ro m sg A fraction decreaso t s increased ean , respectively, without boun takind dindividua e an produc e th g th l al f to l probabilities. After takine gth logarithm and conversion of the sum to an integral the control probability for this general case takes the form: P(D(r) )= (34) If it is assumed thatjp is on the linear part of the dose response curve and D(r) does not deviate much fro Eq.(31usee mD b expano dy t )ma d Eq.(34) integrae Th . Eq.(34n li y )ma thus be approximated by: D(r)-D 1 + y „_ y-j I2 dm (35) where the first terrrj^inside the square bracket gives no contribution to the integral owing definitioe complete th Th o . t D f eno Eq.(34 thuy secono s)ma t d orde approximatee rb : dby P(D(r)) = P(D) - -r (36) ^ 2P(D)\D o where cC is the variance of the dose distribution in the tumor volume as defined by: f(D(r)-Dfcdm ^——7—— (37) I dm The negative sign of the quadratic term in Eq.(36) shows that all variations in the dose distribution that introduce deviations from the mean dose level reduce the control probabilit givea r yfo n mean tumor dos r meaeo n energy imparted. This shoule db expecte increasee th s da d surviva dosw lo e n areali neven s ca completel e rb y compensated by the decreased survival in high dose areas. By setting Eq.(34) equal to P(Dff) the e 60 effective total dose assuming uniform dose deliver calculatee b n t becomei y ca d dan s simply: 1 ——-=- \^èé (38) 2P(D)\D This result again clearly shows thaeffective th t e dos slightls ei y lowen a r y thab nD amount determine relative th y db e varianc dose th ef e o distribution . The above relations were all derived in terms of the probability of achieving tumor control insteae w f interesteI e . dar effece th norman o n tdi l tissue injur resule yth s i t different particularly with regard to non uniform dose delivery. This is so since doses above the mean dose increase the complications more than doses below the mean dose decrease them. Thus, quite generall probabilite yth complicationr yfo norman si l tissue D/Gy Hea Necd& k Target Volume P/% 70 100 45 270 3606° Fig. 5 The variation of P+ and PI with the angle of incidence, ß, for a P+ optimized treatmene lympth e s th i poin e f p o thth necks nodee ti D9 dos, th p p n D .t eso a mean value of 9 neighboring points over an area of 3 x 3 pixels, Dmax/ Dmm and D are the maximum, minimum and the arithmetic mean value of the dose distribution in the target volume. The strong correlation between P+ and D is evident! 61 t pa 0 Advance5 . ca N dH& 2 fr /day 2.1-2.3-2.4 Gy/fr 5.5-6.n 0.i r 8f 8 52 weeks /PI ,_Q •* ' y __o ; —O if—O r D o=58- / 0.6 5 "Clinically Optimal- Total Dose :59Gy 0.4 Acuracy ±0.4Gy 0.2 0 20 40 60 80 D/Gy Fig. 6 Clinically established dose response relations for advanced head and neck tumors [12,13] dashee .Th d curves labelle (Benefit (InjuryI B dP P d an )) correspone th o dt tumor control and severe normal tissue damage probabilities, respectively. The solid curve represents the probability that patients are cured without inducing severe complication healthn si y normal tissues sees i t n.I tha absorbee th t d dose should be within about 0.5 Gy from the clinically optimal dose in order not to lose more tha percenn5 patiente th f to s that potentiall curede b n y.ca increases with increasing fluctuations around the mean dose level! If the same analysis as abov normamade s ei th r efo l tissue obtaine sw : (39) Thu firse sth t order effec similas i t normar rfo malignand an l t relates tissui e d th eo an dt mean tissu e dosth questiono n et i . However, dose fluctuation norman si malignand lan t tissue will increase complications and decrease tumor control, respectively. Thus, in genera probabilite lth achievyo t e complication free tumor control wil, ,P+ l decreaso t e edu both these effects. This is so since in the first approximation [12]: ' PB - PI (40) 62 To illustrate the importance of the various dose distributional parameters the dose at a central tumorpoine th arithmetie n ti th , c mean dose minimue ,th maximud man m dose showe functioa ar angls e a Fig n th i 5 inciden f .f e no o t beam hea a nec d n dso an k tumor. It is seen that all dose concepts varies substantially with the angle of incidence. However, the mean dose always seems very well correlated to the probability of achieving complica- tion free tumor control (P+). This figure clearly show importance sth usinf eo righe gth t dosimetric quantities when prescribing the dose delivery in radiation therapy. 2.4. Precision requirement radiation si n therapy e abovth n eI discussion several expressions were give ne probabilit th bot r hfo f yo achieving tumo complicationd an r ) contro36 , 20 l , norman 8 (Eqssi , 6 . l tissues , (Eqs20 . 39). Wha moss ti t importan radiation ti n therap tha s yprobabilityi e tth , P+, that patients achieve tumor control without severe complication norman si l tissues largs a s s i e,a possible . 40)sees (cf. A FigEq .n n i thes 6 . e patients follo wbela l shaped curva s ea function of dose with a maximum approximately half way between the doses causing 50% probabilit tumof yo r contro thad an lt causin norma% g50 l tissue injurys recentlha t I . y been shown [13] that this curv wels ei l describe Gaussiaa y db n distribution accordingo t P+=P+(D)e ^ " ) (41) where D is the optimal mean dose to the patient. Basically, this is due to the fact that the difference betweecloselo tw e yn th space d sigmoidal tumo normad an r l tissue injury curveapproximatee b n sca derivative th y sigmoide b th f eo d provided thei valuery e sar fairly similar. From this show e relatiob n nca wante t nthai on f havo ti f st o t leasea % t95 the maximum possible complication free tumor control P+(D) the dose delivered should be within AD from D as given by t (42) lOy This means that for steep dose response gradients y the clinically acceptable dose interval quits i e narrow r typicaFo . e requirelth clinica4 = dy l d datan ay assuminG 4 6 = gD accurac dosn yi e deliver lesr o s y abou ys tha i G dos e 6 non e1. t fraction. This corresponds accuracn a o t dosyn i e deliver abouf yo t 2.5% whic quita hs i e demanding value compared with what is generally achieved in clinical practice [4, 11]. 3. CONCLUSIONS The accuracy of the dose distribution and in particular of the mean absorbed dose delivered to the target volume need for many steeply responding rumors be as high as 1 G ymaie see(las a Th nFig )n . i reaso6 . n being thawane tdeliveon o t dosra e whics a hs i hig possibls ha t preferablebu highet yno r thabele peae th n lth f kshapeo d complication free tumor control curve. At higher doses the probability of inducing fatal injury increases very steeply so the optimum clinical dose is just below the peak of the curve. In order not to lose more than one out of 20 curable patients (95% of highest possible treatment outcome requiree th ) d accurac dose th en yi distributio n delivere targee th o dt t volume 63 shoul 2.5e db % (lameaa r )fo n dose response gradienrange th n ei 2-37 t r morFo . e steeply responding tumor d normaan s l tissues even stricter requiremente b y ma s desirable. When onlsingle yon e dose quantit requires yi analyzo dt e dose responseo t datd aan prescrib repord ean t dose deliver meae yth n targee dosth moso e et th volum s ti ) e(D relevant concept at least in external beam radiation therapy where the dose variations over tumoe th generan i r quite ar l e small (cf. 38)d mea e Eqsan .Th 6 n.3 dos s alsei o most relevant for normal tissue damage under the same assumptions (Eq. 39). To get a physical radiobiologicad an l feelininfluence th dose r th gfo ef eo heterogeneit tumoe th d n yi ran organs at risk the relative standard deviation of the dose around the mean dose is the most useful concept (Eqs. 36,38,39). In addition to the mean dose and its standard deviation the ICRU point dose shoul statee db alloo dt w comparison with other recommendations [14, 15]particulan I . r whe dose nth e variation large sar e maximue (>5th %) la m andn i , particula tumorsr rfo minimue ,th m dos highle ear y relevant too. REFERENCES [1] MUNRO ,T.R., GILBERT, C.W. relatioe ,Th n between tumor lethae l th dos d ean radiosensitivit f tumoyo r cells, Brit Radiol. .J (19614 .3 ) 246. [2] BRAHME, A. and ÂGREN, A., On the optimal dose distribution for eradication of heterogeneous tumors, Acta Oncol. 26 (1987) 377. [3] ZAIDER, M., ROSSI, H.H., Microdosimetry and its applications to biological processes, In: Radiation dosimetry, p. 171. Ed.: C. G. Orton. Plenum, New York (1986). [4] BRAHME , DosimetriA. , c precision requirement radiation i s n therapy, Acta Radiol. Oncol. 23 (1984) 379. [5] WITHERS, H.R., PETERS, L.J., Biological aspects of radiation therapy, In: Textbook of radiotherapy, Third edition, p. 103. Ed.: G. H. Fletcher. Lea & Febiger, Philadelphia (1980). [6] THAMES, H.D., HENDRY, J.H., Fractionatio radiotherapyn i , Taylo Francis& r , London (1987). [7] KÄLLMAN, P., ACREN, A-K., BRAHME, A., Tumor and normal tissue responses fractionateo t uniforn dno m dose delivery, Int Rad.L . Biol (19922 .6 ) 249-262. [8] GOITEIN, M.,The utility of computed tomography in radiation therapy. An esti- mat outcomef eo , Int Radiât. .J . Oncol. Biol. Phys (1979.5 ) 1799. [9] SCHULTHEISS, T.E., ORTON, G.G., PECK, R.A., Models in radiotherapy. Volume effects, Med. Phys. 10 (1983) 410. [10] ICRU: Radiation quantit unitsd yan . Repor , Internationa33 t l Commission no Radiation Unit Measurementsd san , Washington, D.C. (1980). 64 [11] BRAHME, A. Ed. Accuracy requirements and quality assurance of external beam therapy with photons and electrons, Acta Oncol. 27 Suppl. 1 (1988). [12] ÂGREN, A-K., BRAHME, A., TURESSON, I., Optimization of uncomplicated con- trol for head and neck tumors, Int. J. Radiât. Oncol. Biol. Phys. 19 (1990) 1077. [13] TURESSON, I., BRAHME, A., "Clinical rationale for high precision radio-therapy". ESTRO, Malrno, Sweden (1992). [14] ICRU: Dose Specification for Reporting External Beam Therapy with Photons and Electrons, Report 29, International Commission on Radiation Units and Mesurements, Washington (1978). [15] ICRU: Prescribing, Recording and Reporting Photon Beam Therapy, Report 50, International Commissio Radiation no n Unit Mesurementsd san , Washington (1993). Î,"EXT lef5 t6 BLAN I K t EQUIPMENt T REQUIREMENTS :EX r P4G£(sf LESSONS LEARNED FROM ACCIDENTS IN RADIOTHERAPY ORTIZ-LOPEP Z Divisio f Nucleano r Safety, XA9642846 International Atomic Energy Agency, Vienna J. NOVOTNY " University Hospital St. Rafaël, Leuven, Belgium J HAYWOOD Cleveland Medical Physics Unit, South Cleveland Hospital United Kingdom Abstract Radiotherapy is the only application of radiation which intentionally delivers very high doses to humans. A gross deviation from the prescribed dose or dose distribution can have severe, or even fatal consequences. Since the patient is placed directly in the beam or sources are inserted in the body, any mistake made with the beam or the sources leads almost certainly accidentan a o t l exposure. Lessons learned from previous incident usee b teso de n t tsth ca vulnerabilit givea f yo n facility, provided that thes adequatele ear y disseminated purpose Th . e of this paper is to present a summary of the lessons learned from a relatively large sample of events analysie Th . bees sha n presente shora s da t description followe identification a y db f no triggerine th contributinge eventh d tan g factors. These have been groupe followss da : errors in commissioning or calibration machines and sources affecting many patients; mistakes affecting individual patients suc irradiatins ha wrone gth g patient wronge th , , fiel siter d do an , mistakes when entering data int readinr oo g fro patient'e mth s chart; errorunusuao t e sdu l treatment situationsr so ; equipment failur humad ean n machine problems, including maintenance. INTRODUCTION There are situations that are unique to the medical use of radiation sources: patients are exposed to direct radiation beams and radiation sources are incorporated to their bodies as diagnosie parth f o ttreatment d san . In therapeutical applications, doses are very high and a departure from the prescribed doses may have severe or even fatal consequences. Not only overexposure but also doses belo intendee wth d one accidentae sar l exposure patientse case th f th eo n si . Accidental exposures also includ treatmeny ean t delivere wrone th o dgt patien wrone th r to g tissuer o , usin wrone gth g sourcwrone th r ego radiation beam witr o , dosha dosr eo e fractionation differing significantly from the values prescribed or which may lead to undue secondary effects. Lessons learned from previous accidents can avoid reoccurrence. For this purpose, the IAEA has collected information on accidents and made a review of 55 events. The result will publishee b IAEn a s Ada special publication. Present address: Radiotherapy Institute, Ministry of Public Health, Natruhlarce 100, CS-18000, Prague 8. 69 The sources of information have been: papers published in scientific journals information provided by professional associations periodical reports available from national institutions, such as the USNRC lessone Th s learned from this sampl directleventf e eo b n sca y usechecklisa s da r fo t testing the vulnerability of a given facility against the initiating events which triggered the accidents reviewed. An effective way of learning concrete lessons is the collection of contributing factors which made possible the initiating event culminating in an accident. To illustrate the method of review, a few examples are given in the following section. It should be noted that, due to the length of this paper, the examples are extremely simplified. The review also includes two accidents involving the public and the environment. EXAMPLES OF ACCIDENTS, INITIATING EVENTS AND CONTRIBUTING FACTORS Event 5A Co-6w Ane 0 sourc installeds ewa mistakA determinatio. e th n ei outpuf no t (dose rate mades dosee )wa Th .patiento st s werhighe% e25 r than intended7 tota20 A .f lo patients were affected. Initiating event: Miscalculation of dose rate from measurements Contributing factors: 1) There was no independent calibration of output 2) Subsequent reviews did not detect the mistake. Only an intercomparison exercise discovered it ) 3 sufficieno Thern s ewa t investigatio unusuallf no y slow healin skif go n effects Event SB A Co-60 decay curv wrongls ewa y drawn. calibratioo Thern s ewa verificatior no f no outpue th t durin months7 g2 . Overdoses increased wit resule a h th s 40%time o a f t to ,p u , departure of the curve from the real decay. A total of 450 patients were affected. Initiating event: Mistak determinatioe th n ei decaf no y curve Contributing factors: lineaf o e r Us scale1)s instea simplef do r semi-log scales 2) No independent determination 3) No beam verification during 27 months 4) Priorit acceleratorw givene s a y wa no t thao s , t Co-60 assigneunis twa d lower priority 70 Event 7 A computerized Treatment Planning System (IPS) was commissioned. A manual correctio distancr nfo isocentries i c treatment introduceds swa correctioe Th . n factos rwa already include software th TPS e dose n isocentrie dl i th al f Th .eo n s i c treatments were lower than intende mucs a y 30%s dh b a mistak e Th . e remaine eighr dfo t years. More than 1,000 patients were affected. Initiating event: A TPS was incorrectly commissioned. A distance correction was introduced twice. Contributing factors: ) 1 dosimetrio Thern s ewa c S validatioTP e th f no 2) Lack of written procedures made it likely that mistake remained undetected for years Even6 t1 The wrong patient received a teletherapy fraction of 2,5 Gy to the lumbosacral spine. Initiating event: A patient responded when another patient was called Contributing factors. 1) Procedures for patient identification (photograph) were not followed 2) Procedures for confirming the treatment site were not followed (tatoos) 3) Patient's objections were not given sufficient attention (by neither the technologist nor the oncologist) Event 35 patienA prescribes twa teletherapda y treatment receivet bu , treatmenda t with Sr-90, surface lOGrighe th th o ytf t eo eye . Initiating event: wrone Th g treatmen choses twa n Contributing factors: The information on the patient chart was not considered or was mistaken Event 19 interloce Th k syste lineaa mf o r accelerator stoppe beame dbeath o availables N .m wa . In order to obtain the beam, a decision to treat in "physical mode" was made. Twelve patients were treated normally. An equipment failure occurred when the interlock system ("non clinical mode") was disabled and one patient died. Initiating event: Equipment fault beao N .m available from accelerator Contributing factors. 1) Decision to treat patient in "no clinical" mode 2) A second failure (power supply) disabled both, the x ray positioning system (target, cone, ionization chamber warnine th d )an g signa theif o l r incorrect position 71 Event 20 The interlock system of an accelerator stopped the beam. No beam was available. An incorrect repai mads rwa e without notifyin radiatioe gth n physicists controe Th . l panel showe energx fi MeVd6 a y(3 ) permanently regardles selectee th f so d energy. Twenty-seven patients were treated wit wrone hth g energy, dos dosd ean e distribution. Several deaths have been admitted to be caused by the wrong treatment. Initiating event: Failure of the equipment and beam was stopped by the interlock Contributing factor: 1) The repair did not correct the real fault. Energy fixed at 36 MeV. Energy selector at console disabled radiatioe Th n ) physicist2 s wer t notifiedeno . Treatments resumed without verificatio beae th mf no 3) Staff assumed that the fix energy displayed at control panel was defective (the display was correct) 4) It was possible to operate the accelerator with the energy selection disabled (equivalen "clinican no o t l mode") fro normae mth ly "beake " mon Event 21 An accelerator equipped with verification of the treatment parameters by software. The automatic verification process took about 20s. Selected parameters were modified during automatie th c verification process, resultin operation a gn i n with hybrid parametersx Si . accidents involving overdoses occurred. Three patients died. Initiating event: An accelerator operated with hybrid parameters after the technologist changed previously selected parameters Contributing factors: 1) Equipmen t teste tno operatin r dfo g conditions that occurre practicn di severaen i l facilities. Quality control and quality assurance of the software was not sufficient to avoi occurrence dth quicf eo k chang selectef eo d parameters ) 2 Commissionin includt no d gedi testin operatingn i g conditions, that occurren di practice 3) The problem was identified by manufacturer only after occurrence of six accidents involving overdose foun si r different hospitals firse Th t . accident occurre 198n di d 5an 198n i lase e 7th on t Event 27 The wrong sourc use s brachytherapa en dwa i y treatment. Initiating event: wrone Th g sourc selectes ewa d Contributing factors: 1) One drawer contained sources of two different activities 2) Labelling of source not adequate 3) No effective verification of sources prior to implant 72 Event 28 A treatment was performed with the wrong source. Initiating event: wrone Th g sources were selecte treatmenr dfo t Contributing factors: 1) the colour coding was fading sufficieno n t) 2 verificatio sourcee th f o n s prio implano rt t Event 25 same Th e event occurre differeno tw n di t hospitals. Source ribbon- Ir 0.7i f so 9mC 192 each were ordered. Ribbons with activitie 0.7f so 9 mg-Ra-equivalent were delivered. The wrong source was implanted into one patient in each hospital. Initiating event: The wrong activity was delivered Contributing factors: 1) Different units of activity were used by the hospitals and the supplier 2) Insufficient verification was performed (only the number 0.79 but not the unit) sourco N e verification/calibratio ) 3 usere th sy nb Event 36 mCi3 A5 , Cs-137, radiation sourc discharges ewa d afte brachytherapa r y treatment Initiating event: A radioactive sourc discharges ewa d with inactive waste Contributing factors: 1) Control of sources returned to the container after treatment was insufficient 2) Monitoring of the area after removal was not performed or was unsuccessful Even9 t3 Brachytherapy treatment with high Dose Rate (HDR) Equipment sourcA . e remained in the patient. The control panel indicated "safety" but the are monitor was sounding an alarm, indicating that the source was not returned into the container. The patient died. Member publie th f sco wer sourceeR exposeHD .a o dt Initiating event: The source became dislodged from the equipment Contributing factors: 1) Misinterpretation of two conflicting signals. Monitor alarm was ignored ) 2 Failur ensuro et e tha sourcee tth s wer t outsideno equipmene eth y tb monitoring of patient, clots and working area 73 Even0 t4 Remote control afterloading. Equipment faile completo dt e transferrin sourcee gth o st the equipment. The sources remained for an unknown time somewhere near the patient's leg. Initiating event: source Th e became disconnected from drive mechanism Even0 t5 A nursing mother was given 4.89 mCi of 1-131 which resulted in an untended radiation infant'r he o t doss y thyroi somG f e o 0 ed30 gland infane Th . t will require thyroid hormone medication for life. The mistake was detected when a whole body scan of the mother was done, which indicated an unusually high breast uptake. Initiating event: gives nursina wa o ni t 4.8gmC 9mothe r Contributing factors: (neithee on referrino e rN th g ) physician1 nuclea e stafe th th n fi r rno , medicine station) nursing s aske mothewa e usua e de th Th sh . lf ri measure thir sfo s kin situatiof do n wer taket eno n Even (involvin2 t2 g publi environmentd can ) A radiotherapy department moved to new premises and left a teletherapy unit in the old hospital. Members of the public had access to the unit, dismantled it and broke the Cs-137 source capsule sprearesule a th s radioactivf f A dt.o o e material, 112,000e personb o t d sha monitored for possible contamination, 249 individuals were found to have some contamination person4 , monitorese b died house 9 o t 15 ,d contaminationr sha d fo house2 4 , s required decontaminatio 35,00d wastef nan o 3 0m d were generated. Initiating event: A teletherapy unit was dismantled and the source capsule was broken. Contributing factors: 1) There was no proper decommissioning of the facility 2) The teletherapy unit was left in unsafe storage conditions nationae Th l regulator ) 3 y authorit t notifie discontinuee no th s f do ywa d operation of the radiotherapy facility 4) The unit was not recognized by the members of the public as something dangerous 5) The chemical form of the source facilitated the spread of the contamination Even (involvin3 t2 23 (involving publi environmentd gc an publi c an)d environment) A f~*f\ £*r\ +ala4-\\^i **v\\i •!«•«*+ «troc fllASTQllir +*•« A Co-60 teletherapf y unit was illegally transported, imported into a country, stored durin yearx gsi unsafn si e conditions maintenancA . e technicia t acces e abls ge th no ewa t o s t unit, to dismantle the source driver, and to break the source capsule. Metal parts were sold to scraa p metal compan thid meltintura n ys an i o nt g facility. The result was that 30,000 metallic table bases made form contaminated material were distributed wels a ,6,00 s a l 0 ton reinforcef so d rod buildingsr sfo . 17,600 houses were checked 74 for contaminated rodshouse4 81 . s were partiall totallr yo y demolished. 16,00f o 3 0m contaminated earth were generated. Initiating event: A perso acces teletherapd e nha th o st y head, dismantle unibroke d d th sourcan t e eth e capsule. Contributing factors: 1) Illegal import, transport and unsafe storage of a radiotherapy unit 2) The unit remained stored unsafely for six years recognizet no unis e techniciae th wa tTh y db somethins na ) 3 g dangerous scrae Th p meta ) l4 company received contaminatet d no materia d di d lan detect it LESSONS LEARNED FROM INITIATING EVENT AND CONTRIBUTING FACTORS 1. MISTAKES MADE DURING COMMISSIONING, BEAM AND SOURCE CALIBRATIONS, SUC: HAS Determination of the dose rate for Co-60 teletherapy units and of the dose per monitor unit for accelerators Decay table calculatiocurver e so th r sfo irradiatiof no n tim fractior epe n Determination of wedge factors Validatio treatmenf no t planning systems Verification of the activity for sealed sources (brachytherapy) and unsealed sources (metabolic therapy) The lesson learned from these accidents is that the consequences may be severe or even fatal and affect a large number of patients. Examples are 207, 450, 1045 patients. Recommendation accidenn so t prevention and/or mitigatio consequencesf no : Human redundanc independend yan t determinatio calculatiod nan commissioninr nfo g of equipment and facilities Periodic interna externad lan l independent audits Dosimetric verificatio phanton ni m under real working condition witd normae san h th l operator using the equipment. This includes the validation of the computerized treatment planning system When and where practicable, "in vivo" dosimetry (at least at the first fraction) Implementatio qualitf no y assurance with periodic constancy checks 75 ) 2 MISTAKES CONCERNIN WRONE GTH G PATIENT WRONE TH , G FIELDE TH , WRONG BEAM OR SOURCE These mistake relativele sar y frequenusualle ar d yan trelate procedurao dt l mistakes (either non written procedures, not well understood, not well verified or violated procedures) Recommendation for prevention and/or mitigation. (In many teletherapy cases the mistake affects only one or two fractions. Therefore the severity can be kept low if the mistake is discovered by means of frequent verification) Patient identification with photograph on the patient's chart Communication wit patiene hth t Clear and unambiguous procedure for tatoos Human redundance, clea descriptionsb rjo , clearance with signature operatine th y sb g staff for each treatment/fraction 3) MISTAKES RELATE ENTERINO DT READINR GO G DATA FROE MTH PATIENT'S CHART These mistakes often deal with enterin wrone gth g fraction dose, accumulatioe th f no total dose, registe wrone th f rgo beam, wrong identificatio wedgesf no wrone ,th g wedge factors, or wrong activity or dose units Recommendation preventior sfo n and/or mitigation: Redundan independentd tan , frequent revie patient'f wo s charts (two persons, twicea week) Clear, concis unambiguoud ean s written procedures Clear job descriptions. Cross check of job descriptions to avoid gaps in responsibility Clearance with signatures of the operating staff 4) MISTAKES SPECIFIC TO BRACHYTHERAPY The wrong radionuclide wrone th , g activity wrone th , g unit activityf so , mistakes made identificatioe inth sourcese th determinatio e f no th n ,i treatmene th f no t timee th , position of the sources in the applicator, the incomplete retrieval of the sources (sources left in the patient), mistakes in the accountancy of the sources at the storage, source damag losr eo t sources. Recommendations for accident prevention or mitigation: Clear identification of sources, double verification Registration of all source movements without gaps 76 Accountancy and verification of sources before and after treatment Monitorin patiente gth , cloth ared saan before discharginm ghi Clear definition of function of staff with regard to all steps: ordering, reception, labelling storage, retrieval for use, transfer between staff and return to storage Clea descriptiob rjo clearancd nan meany eb signaturef so s Quality control of remote control afterloading 5) COMMUNICATION MISTAKES The mistakes are: Lack of communication Incorrect communication wrone th o wrongt e r persoth y o , y gb wa r no Oral communicatio criticaf no l information wrongly understood Mistakes when reading or transferring information Unreadable or confusing handwritten communication, informal expressions or use of jargon whic necessarilt no hs i y understoo everyony db exactln ei same y th ewa Equipment instruction in a foreign language Telephone-only communications Interpersonal problems Noisy environment, prone to distraction or to loose concentration Insufficient dedication (non availability) of key positions personnel in the process of communication Recommendation accidenr sfo t prevention and/or mitigation: Identificatio listin d safetl nan al f go y critical communication Clear and concise written procedures for the safety-critical communications identified in the list Clear assignment of responsibilities to the staff involved in the process of communication Written and signed safety-critical communications Communications check-lists 6) SPECIAL TREATMENTS, SPECIAL SITUATIONS, PERSONNEL CHANGES Change of supplier of radioactive material Non-typical dose for a specific treatment, unusual area or treatment with the patient in difficult and non-usual position Change in units of activity 77 Recommendations for accident prevention and/or mitigation: Anticipation of these situations by means of a safety assessment of all procedures Clea concisd ran e written formulation Additional, specific training for special situations Clea descriptiob rjo clead nan r assignmen responsibilitief to s 7) EQUIPMENT FAILURE percentage Th contributinf eo g factor overale th o st l causeequipmeno t e sdu smals ti l reviee inth thesf wo e histories. However consequencee ,th vere b yn seversca d ean affect many patients. These contributing factors are: No sufficient redundancy (single fault criterion) (interlock failure) Software problems Hardware incompatibilities Recommendation accidenr sfo t prevention and/or mitigation: Implementation of basic design rules, such as: Safet deptn yi termn h(i engineerinf so g thi cals si l defenc depthn ei ) Note: defenc deptn ei h consist overlappinf so g safety provisions, suc physicas ka l components, procedure combinatior so bothf no thao s , ver tprobabilita w ylo f yo failure can be achieved by combining protective layers such that the probabilities are multiplicative. Single fault criterion, (e.g. single ,on e faulty component should never caus accidentn ea n A . accident should be only possible if two systems or components fail simultaneously. By repairin faulte gth y component befor secone eth d fault appears, accident preventede b n sca . The "single fault criterion" is a simple design approach to obtaining a minimal redundancy). Redundancy, independenc diversitd ean safetr yfo y critical components faulf o e t treUs e analysi desigr sfo n Systematic use of safety assessment methodologies Equipment tests including possible operating mistakes (quality contro manufacturine th f lo g process including clinical conditions, and real machine operators) including special or extreme conditions and handling. Whe equipmene nth "non i s ni t clinical mode" (reductio interlockf no eliminatiod san layerf no s of the safety in depth), it should be made impossible to meter a "beam on" order from the normal key in the keyboard. Rather, validation exercises of the whole, including clinical conditions shoul made db e before starting treatments with real patients. 78 Manufacturers should investigate promptly and thoroughly any reported incident or unusual event and notify authorities and users of the findings and/or preventive measures. To achieve this, a system of formal reporting should be established in each country and disseminated internationally. 8) PROBLEM HUMAN-MACHINF SO E INTERFACE Mistaken interpretatio signalf no displayd san s Mistaken decisions of contradictory signals. Tendency to assume as a good signal, the one which fits expectations (tendency to accept unsafe conditions as those which allow resumption of operation). Recommendations for accident prevention and/or mitigation: Training of personnel to recognize abnormal situations (understanding of safety assessment). Training which includes both norma abnormad an l l contradictory signal generan s(i d lan specifi particulae th o ct r equipment) Learning from previous accident unusuad san l events (including maintenance personnel) Training promotin gquestionina learnind gan g attitud safete pars th e a f yto culture. (See point 12) ) 9 BYPASSIN INTERLOCKF GO OPERATIOD SAN "NONN I N CLINICAL MODE" The bypas interlockf so s reduces safety drastically The decision of operating in a "non clinical mode" is often due to frequent or non-resolved maintenance problems intermittenr ,o t faults non-definitely resolved, which causee sth equipment to stop, thus disturbing the patient's treatment consequencee Th operatiof so "non i n clinical mode oftee ar "n fata quitr o l e severe Recommendation accidenr sfo t prevention and/or mitigation Priority shoul givee db pla no implement d nan t efficient maintenanc avoid ean d improvisation as far as practicable Never treat patients in "non clinical mode" Equipment design should not allow "beam on" from the normal key in "non clinical mode" (it should only be possible by using special tools or with special computer codes, for maintenance or physical work). 79 10) MAINTENANCE PROBLEMS Mistakes made in maintenance can also be severe, even fatal, or affect many patients Recommendations on accidents prevention and/or mitigation trainine Th maintenancf go e personnel should includ knowledge eth consequencee th f eo f so any manipulation, adjustment in all components. The training should be the result of an exhaustive safety assessment There should be unambiguous communication procedures for transferring the machine, for initiating maintenanc returnind ean machine g th staf e th fo e t responsibl radiotherapr efo y physics Clea unambiguoud ran s responsibility definition ) 11 PROBLEMS WITH DECOMMISSIONIN SOURCEF GO EQUIPMEND SAN D TAN FACILITIES, AND UNSAFE STORAGE Sources out of control can lead to catastrophic results Recommendation accidenr sfo t prevention and/or mitigatio resultf no s Formal procedures for regulatory control of sources not yet in use or no longer in use, from import into the country to the proposer disposal. Clear, unambiguous and authorized procedures making it mandatory to retrieve the source and storo t temporarilt ei definitelr yo y until disposal. Formal control during this perio timf do e Seal-off of equipment out of use to avoid connection to the electrical power supply Storage of sealed sources after leakage or damage Safe disposal of brachytherapy and nuclear medicine sources Verification and clear labelling of empty containers SAFETY CULTURE In the preceding section, recommendations on accident prevention emerging from lessons learned from previous accidents hav neeee clear lea th d descriptionsfo b do t r jo , clear assignmenf to responsibilities, clear written procedures, compliance with procedures documented by signatures, written communication, safety assessment and training of personnel on anticipated problem situations. All this constitutes good practice. However, even with strict procedural compliance there may be situations, events or combinatio eventsf no exactlt ,no y define proceduree th n di s where strict observance b t no y ema sufficient thesn I . e case effectivenese sth dealinn si g with initiating culminatio e eventth d san f no accidents strongly depend human so n attitud judgementd ean . Therefore, good practic essentias e sufficienti t no t lbu . Ther beyon neeo a g es i do t e dth implementatio gooa f no d practic orden ei ensuro rt e that both personal attitude habitd san thoughf so t and organizational approaches and priorities are oriented toward the goal. This goal is that all duties 80 relate safetdo t carriee yb correctlyt dou , with alertness thoughe du , fuld ltan knowledge, sound judgemen propea d tan r sens accountabilityf eo . Furthermore, it implies a learning attitude in all organizations concerned, taking into account all relevant operating experience as well as new research results as a basis for safety improvements and reassessments. Internationae Th l Basic Safety Standard Radiatior sfo n Protectio Safete th Radiatiod f yo n an n Sources [1] establish that the Regulatory Authority is to require all parties involved to develop a safety culture. safete Th y culture involves therefor l levelseal : regulators, commitmen managemenf to d tan response of individuals. Measures to encourage a questioning and learning attitude and to discourage complacency should be put into place. INTERNATIONAL DISSEMINATION OF INFORMATION ON ACCIDENTS AND UNUSUAL EVENTS A questionin learnind gan g attitude, combined with informatio accidentn no unusuad san l events, their initiating event contributind san g factor efficien n drasticallo a t s si y wa t y reduce eth probabilit furthef yo r accidents. The lis contributinf to directle gus factore b tes vulnerabilityo e t n tth sca radiotherapy an f yo y department. The number of reported events in a single country is in most cases insufficient to provide a significant number of lessons in a reasonable time. Therefore, a compilation of accidents at the international level would allow all countries to benefit from the lessons learned by each of them. Moreover, unusual events which did not culminate in an accident can build up a body of operational knowledg avoio et d real accidents. Therefore systee th , m should allo anonymour wfo s report thesf so e cases that otherwise are not published in any scientific journal and would never reach the interested community. For this reason Internationan ,a l Reporting Syste Unusuamf o l Event bees sha n proposen a y db advisory group reportine Th . g system will consis questionnaira f to narrativa d ean e sectione Th . database will include protected field confidentiar sfo l identity data which shoulincludee b t e dth no n di output of the system (reports). REFERENCE [1] INTERNATIONAL ATOMIC ENERGY AGENCY, International Basic Safety Standard Protectior sfo n against Ionizing Safet e Radiatioth r yfo d nan of Radiation Sources, Interim Edition, Safety Series No. 115-1, IAEA, Vienna (1994). REVIE WHO/PAHO/IAEF WO A RECOMMENDATIONS CONCERNING RADIOTHERAPY FACILITIES """" XA9642847 O.P. HANSON Radiation Medicine, World Health Organization, Geneva, Switzerland Abstract 1960d Sinc Worlmi e se th eth d Health Organization American Pa e ,th n Health Organizatione th d ,an International Atomic Energy Agency have provided recommendations concerning radiotherapy services, including organization, staff requirement facilitiesd san . Thes containee ear varioun di s report WHOf so , PAH IAEAd Oan , whic reviewee har summarizedd dan . INTRODUCTION Beginnin 1960se th gn i , international organizations with interes planninge th n ti , organizationd an , provision of radiotherapy services have provided guidance for their member states. These are contained in various report Worle th f sdo Health Organization American Pa e ,th n Health Organizatio Internationae th d nan l Atomic Energy Agency. For example WHO Technical Report Series Nos. 322 Cancer Treatment, 328 Planning of Radiotherapy Facilities (a joint WHO-IAEA meeting) and 644 Optimization of Radiotherapy. The earlier recommendations have been reconfirmed in more recent meetings convened by IAEA and WHO and today are still valid in many parts of the world. The need for a scientifically sound, robust, reliable treatment machine capable of high-quality performance has been universally recognized. Up to the present time the consensus of opinion is that a cobalt-60 machin preferabls ei linead ean r accelerators generall e coulb t dno y recommende developinn i e us r dfo g countries. Because of difficulties related to the timely replacement of cobalt-60 sources, including the proper disposal of used sources as well as the difficulties in providing the proper infrastructure to effectively utilize linear accelerators, IAEA, PAHO, UNIDO, WHO recently have organized (Washington D.C., December 1993 Advisorn )a y Group Consultatio consideo nt "desige rth n requirementa r sfo megavoltage x-ray machine which could overcome the disadvantages of both the cobalt-60 unit and currently available linear accelerators. REVIEW OF RECOMMENDATIONS Cancer Treatment (1966)2 Exper32 O . Repor(1)WH tNo :Committeea S f o t TR O WH . EQUIPMENT, Treatment Unites Recommended for a Radiotherapy Department or Centre A. For conditions lying close to accessible body surfaces: Low voltage x-rays, 60 - 100 kV . B Megavoltage radiatio deep-seater nfo d lesions: For x-ray therapy, 3-8 MV C. Kilovoltage radiation: S.S.D m cesium-13r c o 0 V 4 . k - 0 0 3 25 t 7a D. Electron therapy, not essential for every radiotherapeutic department. 83 "The linear accelerator or the cobalt-60 teletherapy unit (whose gamma-rays are equivalent to 3-million-volt X-rays) must therefore be regarded as a standard piece of equipment for a modem department. As to which of thes typeo etw machinf so preferree b o t s ei d ther mucs ei h argument. Linear accelerator producn sca e rather more penetrating radiation and therefore make the treatment of the large patient somewhat simpler. They also provide sharper beams, which may be clinically advantageous. On the other hand, they require much more skilled technical attentio ensuro nt e their continued steady running cobalt-6e th d ,an 0 machines do provid enon-fluctuatina g sourc radiationf eo . Offsetting this coursef ,o fac e th t s thai , cobale tth t radiation can never be switched off, which introduces extra protection problems, and furthermore the natural decay of the radio-active source means that replacement has to be undertaken about once every three years, an expensive and somewhat troublesome business. i On balance, it is probably fair to say that if regular expert technical assistance is not readily available the telecobalt unit is to be preferred. Elsewhere the choice may depend on local opinion and can be allowed to do so since the differences are marginal. One point worth mentioning is that financially there is also little to choose betwee machinese nth , what advantage exists seemin wite lineae li hgth o t r accelerator, provided uses ii t capacity."...... do t . pag2 e5 STAFF; full time dedication to radiotherapy of medical and auxiliary staff (physics, radiotherapy technicians, statistician). Plannin Radiotherapf go y Facilities. Repor Joina f to t IAEA/WH8 32 . No OS MeetingTR O .WH EQUIPMENT; Choic Radiotherapf eo y Equipment I Deep-seated Lesions (followin availablee gar ) cobalt-60 2 MV resonant transformer Grafe d n f Va V 2M 4-8 MV linear accelerator x-ra V electrond M y an 2 4 o sV t Betatrons abovp u Me 0 e2 , II Lesion Surfacf o s withim c ew nfe 200 - 400 kV x-rays medium distance gamma-ray beam units electron beams,V 6-2Me 0 I II Superficial Lesions (Ski Bodn i r nyo Cavity) short distance gamma-ray beam units electroV nMe beam6 o t p su RECOMMENDED: ONE UNIT FOR EACH CATEGORY AS BASIC EQUIPMENT FOR A RADIOTHERAPY DEPARTMENT 84 Considerations Affecting Choice or Specific Equipment within a Given Category "Several physical aspect competitivf so e devices shoul comparede db , including: eas e uniformit(ad th )e an y with whic hselectea d tumou deliverede r b dos n eca ; (b) the reliability and ease of maintenance: (c) the versatility, i.e., ability to provide treatment in more than one of the above categories; (d) the radiation safety; capitae (eth ) l cost; econome (fth ) operatiof yo maintenance.d nan " ... page 14 Treatmen Deep-Seatef to d Lesions "The most important single piece of apparatus to be selected for a new radiotherapy department is the supervoltage unit. Multiple-fiel rotationar do l treatment plans, using X-rayr o V sM rangin0 3 o t g V froM m2 gamma rays from cobalt-60, do not differ remarkably. In all cases, adequate doses of sufficient uniformity are deliverabl volume th o et e containin lesione gth thesn I . e circumstances, reliability, eas maintenancf eo cosd ean t becom decisive eth e factors." ...... pag4 e1 Reliability of Supervoltage Equipment "Probably the two most reliable types of supervoltage radiation sources are the 2-MV resonant transformer and the cobalt-60 teletherapy units. The former have been known to run for 10 years without being opened and without replacement of the glass accelerator tube. Tube lives of over 10 000 hours have been recorded." "In the same way, modern cobalt-60 units frequently give continuous service for periods of several years between source replacement." "It is generally recognized that Van de Graaff machines, linear accelerators and betatrons require the rapid availabilit skillea f yo d maintenance enginee technicianr ro . Without such skilled personne withoud lan t preventive maintenance routines, it would hardly be possible to operate these machines and not incur excessive loss of treatment time. Neithe resonane rth t transformer teletherape th r sno y units requir immediate eth e availabilit speciallf yo y trained maintenance personnel. It is therefore concluded that reliability dictates the installation of one of these types of equipment in the developing countries." ...... pages 14/15 Economics of Supervoltage Equipment Comparative estimates of operational cost per 200-rad (2 Gy) tumour dose at 10 cm depth. ASSUMPTIONS All units operated under condition durinfule f so us l gtreatmena t day. Same or comparable source-surface distances. Constant set-up tim patientr epe . Note: Comparisons do not include cost of personnel (personnel costs considered constant per treatment given). 85 "It is therefore recommended on pounds of reliability and economy that the supervoltage unit of choice for a developing country is a large cobalt-60 teletherapy unit." ...... page 15 STAFF; Medical radiotherapist, Radiological physicist, Radiotherapy technician, Physics technicians. "Unless a fully-qualified radiotherapist and fi radiological physicist will be available for staffing a new department wisdoe th , establishinmf o shoult gi reviewed.e db " Optimizatio (1980)Radiotherapyf 4 no 64 . No S MeetingO TR . ReporO WH a WH .f to (3) EQUIPMENT; Choice of Radiation Energy for Radiotherapy with Photon Beams "High-energy electron accelerators such as linacs and betatrons, when the electrons are directed at a metal target, give the radiotherapist a beam of X-rays of higher energy and sharper delineation than may be obtained with cobalt-60 gamma rays. Accelerators have other advantages also, suc highes ha r radiation abilite outputh d o yt tan produce electron beams of specified energy (see section 3.1.2), but they are secondary to the central feature of higher radiation energy coupled with a sharper beam." ...... page 16 "From the available evidence the conclusion is inescapable that an electron accelerator (preferably with electron-beam capability) offers considerable advantages ove cobalra treatmene t unith r tfo largf to e patientd san patients with lesions tha "difficulte tar " owin sizego t , shape locationr o , . Altogether, these patient constituty sma e 10-15% of the total. For the remaining 85-90%, the accelerator may offer a slightly better treatment than the cobalt unit but this is in any case offset by problems of maintenance and dosimetry, which are particularly acute in developing countries." ...... page 17 Role Electrof Th eo n Accelerator Developinn si g Countries "While the Meeting did not consider that high-energy electron accelerators could be generally recommende developinn i e us r dfo g countrieswist no discourag ho t d di t i , purchase eth sucf eo h machines where certain conditions are met. They can be justified for a centre of excellence in a country that has already achieved an acceptable standard of radiotherapy covering the majority of the population". "An accelerator should be regarded as a machine of different capability from that of a cobalt unit - a machine able to treat patients who cannot be treated properly with a cobalt unit." "It follows thatacceleraton a f purchasedi ,e b o t s ri shoult i , machina e db e capabl deliverinf eo g 8-1V 0M X-rays and an electron beam of variable energy up to at least 20 MeV and perhaps even 45 MeV." "However, quite apart fro vere mth y large capital cos sucf to roo e hmachina th mf o require d ean houso dt e it, serious consideration should be given to the following prerequisites for the successful use of an accelerator." ...... pages 22/23 Prerequisites for the Successful Use of an Accelerator (1) Expert personnel mus availablee tb , including radiotherapists specially traine high-energn di y X-ray therapy, radiation physicists, a sufficient number of specially trained medical radiological technicians (more staff are needed cobalthaa r nfo t unit) engineerd ,an electronir so c technicians. infrastructure (2Th ) radiotherape th f eo y department mus excellente tb . This include reliablsa e electricity supply, not subject to interruptions or severe reductions in voltage, a good water supply at reasonable pressure, and acces gooa o st d machine sho othed pan r back-up facilities. (3) Good communication with readd an , y acces servica , sto e agenc manufacturee th f yo essentials i r . Eithe agenc e radiotherape rth th r yo y centre itself must kee pgooa d suppl importanf yo t spare part circuid san t 86 boards. Telephone communication with the service depot should be good and not liable to frequent interruption, arrangement shoul drapie existh r d tfo passag e through custom sparf so e parts that obtainehave b o et d from abroad. nighe th f acceleratoe tI cos th ) justifiedf e t(4 o b o t s ri must i , usee tb d intensivel efficientld yan larga r yefo patien. t load. This implies tharadiotherape tth y centre mus efficientle tb y organize permio t s a higta o d s h throughput of patients. If the treatment room remains unoccupied for even 5 minutes between each patient, the economic advantage of the high radiation output of the accelerator is lost. Furthermore, the "back-up" structure of the department (diagnosis, localization, treatment planning, dosimetry, follow-up, records) must be commensurate wit hhiga h throughpu treatmene th n i t t rooms. These lasmann i tt conditiony me radiotherap t no e sar y centres, even in industrialized countries sucn i d h centre,an s accelerator employee sar belor dfo w their economically optimum potential. (5) Excellent dosimetry and a daily check of the machine are further prerequisites for operating accelerators. If accelerators are not calibrated and operated with great care they can be extremely dangerous, since patient dosage can be in error by up to an order of magnitude. Recommended Teletherapy Machines for Developing Countries "The undoubted advantage high-energf so y X-ray perhapd san s electrons relativ cobalt-6o et 0 gamma rays are offset by the fact that an electron accelerator (linear accelerator, betatron, or microtron) entails considerably higher capital and annual expenditures. In addition, the accelerator is more liable to break down and more difficult to repair and to maintain in good running order. The Meeting therefore felt unable to recommend accelerators for genera developinn i e lus g countries consideret I . d that nexe least th ,a tr 5-1tfo 0 years, radiotherap thesn yi e countries should rely principally on cobalt units and only in certain circumstances can accelerators be recommended for countries." ...... pag9 e1 OTHER REQUIREMENTS Report of Research Coordination Meeting on Testing of Dosimetrv Equipment IAEA. Vienna. 27-29 November 1989 (Internal IAEA Report)'4'. EQUIPMENT Minimum Requirement Perforo st m Radiotherapy Megavoltage units are strongly recommended for curative radiotherapy. When setting up a new facility it is strongly recommended tha firse tth t therapy cobalunia e unie b t tTh t unit shoul. ,cm isocentrid0 8 f o D c witSA n ha includ emovabla e collimator, mechanica motionse th l l scaleal r mechanicaa , sfo l distance indicator automatid an , c necessare timeth d ran y safety devices. additionas A l equipment minimue th , m requiremen: is t wedgf o t se e filtera s traa y with standard shielding blocks a convenient couch" ...... page 1 DISCUSSION After completion of the present report, the Committee had the opportunity to consult the WHO technical comparo t d repor an suggestions eit 4 64 t s wit recommendatione hth s mad 1980n ei . presene Th t recommendation excellenn i e sar t agreemen treinforcn witca hd conclusioe reporan eth 4 64 t n of 1980 onle yTh . differenc tha s eCommittei e tth e has purposen ,o , considere vere dth y minimum requirementd san optimunoe th t m ones." ...... page3 87 Organization and Training in Radiotherapy for Developing Countries in Africa. IAEA TECDQC-614 (1991)(S): EQUIPMENT; Recommendations "The discussants recognized that African countries are in dire need of radiotherapy equipment. While most of the available equipment is sophisticated and expensive, they recommended that as much as possible the design of simple, sturdy, safreliabld ean e cobalt machine radiotherapr sfo y shoul encouragede db . Such equipment shoule db chea affordabled pan , whilsame th t eea tim wilt ecompromist i no l e safety, reliabilit efficiencyd yan believes i t I . d that this is feasible if such a machine is devoid of costly and sophisticated electronic and mechanical parts." ...... pages 219/220 "Adequate care must be taken in Africa to ensure optimal suitable power and air conditions wherever any major radiotherapy equipment is to be installed in order to reduce the risk of damage from the harsh atmospheric conditions, like temperature, humidity, dust, etc." ...... page 220 Radiation Dose inRadiotherapv from Prescription to Delivery. IAEA TECDOC-734 (1994) : (6) Chapter VIII - Conclusions, Recommendations, and Future Work Summar Rounf yo d Table Discussion "The need for a scientifically sound, robust, reliable, treatment unit capable of high-quality performance wit initiaw operatinhd lo lan g universallcoss wa t y recognized characteristice Th . sucf so hunia t were considereo dt be: ) (a Mechanically robust, incorporating onl yminimua electricaf mo r electromechanicao l l features witd an ,h careful selection of components for resistance to deterioration due to high levels of heat and humidity. Mechanical scales must be provided for all motions. ) (b Isocentric mountin source-axig m witc 0 h8 s distance. (c) Automatic timer (electronic). ) (d Adjustable collimator capabl rotatiof eo n wit l movementhal s manually operated witd an ,hfiel a d 0 siz3 f eo x 30 cm at 80 cm source-axis distance. (e) Accurate light field and distance indicators (electrical), with mechanical back-up must be incorporated. (f) The radiation sources must be cobalt-60 and must provide at least 1 Gy/minut isocentert ea . (g) Accessories must include a wedge filter holder and a beam-block holder. treatmene Th t ) couc(h h shoul mountee db isocentrict no railn d o san t shoul I . electricalle db y driven with mechanical back-up features, and should be constructed so that it is essentially radiotransparent for treatments wit beae hth m directed upward throug. hit (i) Necessary safety feature complo st y with national and/or international requirements mus incorporatee tb n di the design." . pag.. 3 e38 88 National Cancer Control Programme. Report of a Working Group on National Cancer Programmes. Geneva. 25-29 November 1991. WHO/CAN 92.1. Limited Distribution. 1992(7) TREATMENT OF CANCER - "Fo lons ra preventios ga cancef no r canno fulle tb y achieved, treatment will remain important." "Whil basie th e c principle f treatmen o l sworle samal th n i ede ar tregions e emphasith , s accordeo t d treatment will depend upo locae nth l patter f cancerno , i.ecommonese th . t type f canceso e r th seen d an , relative proportion of early and late stage cancers." "The primary objectives of cancer treatment are: Y cure; Y prolong useful life; Y improve quality of survival." " "The primary treatment modalities include surgery, radiation therapy, and chemotherapy. It is unlikely that newer approache treatmeno st t wil applicable b l lese th s o developeet d countrie e immediatth n si e future, since they are not only costly, but also, at the present time, of unproven or minor efficacy. " ... page 55 "Radiotherap e curativb n r somca yfo e e cancer heag d neck(e s an d , cervix) and~ provide substantial palliation for others. Treatment policies should be established." "Relatively inexpensive cobalt therapy machines wil easiee b l o maintait r d wilnan l provide adequate therapy or palliation for the majority of patients without resorting to expensive and service-demanding linear accelerators or other high energy machines. For the great majority of treatable cancers in developing countries, linear accelerators offer no advantage over cobalt therapy." ... page 56 "Manpower needs should be reviewed. Where possible training should be sought in programmes with patients, training and equipment relevant to the needs of the country ..." Advisory Group Consultation on the Design Requirements for Megavoltage X-rav Machines for Cancer Treatment in Developing Countries (Washington P.C.. 6-10 December 1993). to be published by PAHO(8). EQUIPMENT The required machine dimensiona bead an lm performance parameter meeo st e need th t f developino s g countries were define founverd e b d an y o dt simila thoso rt developer efo d countries agrees factn wa I .t i ,d that such a machine must be suitable for use in developed countries. The major dimensional preferences were: (a) Low isocenter height, not more than 130 cm, 115 cm source-axim c preferred 0 10 ) s(b ;distanc e preferred acceptablem c 0 8 ; ) Couc(d ; h vertica l belom travec w0 7 l isocenter. Rotation + or - 90 degrees; (e) Field size at isocenter at least 30 cm x 30 cm. An x-ray energy of about 6 MV is preferred. There was a strong preference expressed that if the machine is acceleratorn a te ob must i , t provide significantly greater beam penetration than cobalt-60. WHO Manual of Radiotherapy and Cancer Management (Manual being prepared, expected publication. 1995)C9). i MODALITIES Medium energy (orthovoltage) x-ray machines with generating potentials in the range of 100-300 kV which originally were used to treat deep-seated tumours are no longer recommended for that purpose, having been replaced 89 by "megavoltage" machines operating at effective energies equal to or above the energy of cobalt-60 (1.25 MeV). advantagee Th higher-energe th f so y radiatio universallnw whicno e har y accepted are: skin-sparine th 1 electronbuilf .e o th g p d o u effect e s du tbelo surfacee wth . greatee th r powe2. penetratiof ro hencd nan increasee eth d percentage depth-dose. decreasee th 3d. scatter sideways fro directioe mth bea e consequentld th mf nan o y sharper delineation of the beam. 4. the smaller increase in specific absorption in bone compared to soft tissue. ACCELERATORS Advantages: 1. Sharper beam delineation (small physical penumbra) than cobalt-60). 2. Higher radiation output than cobalt-60. • . 3 Higher penetratio e photonn tissui th n f o e s than cobalt-60 s onl.ha y CommentV M 6 4- : moderately higher penetration than cobalt-60. Higher photon energies (over 6 MV) often give some advantage deep-seaten si d tumours. 4. Greater ease in treating large patients, and lesions that are difficult to treat because of size, shape, or location patientl al f o .s % Commenttreated15 o t .% 10 : e Thesb y ema . 5. Unit available sar e whic producn hca e electron beams. Comment purchasf I : beins ei g considered, an electron bea variablmf o ecover energV abouo t Me sp 0 ymosu t2 t applications. ACCELERATORS Requirements: 1. Large amount of resources for the initial purchase and for construction of the treatment room. 2. Reliable electricity supply, not subject to voltage reductions or interruptions. 3. Availabilit f gooo y d service froe manufacturemth r suppliero r , including spare parts locally available, and easy access to telephone communication. 4. Excellent radiation dosimetry and quality assurance capabilities in the department. 5. Well trained staff especially qualified to work with high-energy photons and electrons. Comment verA : y sharp beam delineatio onln utilizee nca yb d treatmene whe th e res f nth o t t procedur vers ei y accurate examplr fo , e accurate diagnostic outlinin tumoue th f go r volumed an , careful setting-u positionind pan patiente th f go . COBALT-60 TELETHERAPY MACHINES Advantages: 1. Provides high-energy photons (1.25 MeV average). 2. Many years of experience in the use of these units has proven their dependability. mosr Fo t practica3. l treatment situations adequate dose distribution obtainede b n sca . 4. The initial capital investment is moderate, and routine operating costs are moderate. 90 5. Installatio achievee b n relativelna ca n di y short time with moderately skilled workers. totaw Fe l staff stafd an ,f wit 6 h moderate level traininf so sufficiene gar routinr fo t e operation. 7. Maintenanc requiret repaid no ean e rar d frequentl whed yan n neede moderate dar cosn ei t COBALT-60 TELETHERAPY MACHINES Requirements: 1 Adequate premises and shielding are necessary and must be provided. 2 Provision mus made b t replacemenr efo cobalt-6e th f to 0 sourc t periodiea c intervals 3. Trained staff in sufficient numbers are necessary. 4. For curative treatment of certain cancers (for example, very deep-seated lesions or lesions near critical tissues), more care in planning is required (as compared to accelerators) because of the less penetrating radiation, and the larger penumbra. SUMMAR CONCLUSIOD YAN N In view of the global range of situations now existing, most of the recommendations which WHO, PAHO, and IAEA have made concerning radiotherapy equipment and facilities since the 1960s, are still applicable in some place in the world. In 1994 in many places, reliable electrical power (voltage, current, frequency), water supplies (quantit qualityd yan environmentad )an l modification systems (heat, humidit dusd yan t control unavailablee )ar r o , only available in a few large cities. Additionally the availability of adequate maintenance services, and spare parts and/o funde rth provido t s themr efo severels i , y limite mann di y locations. Furthermore prerequisitee th , s remain unchange comple y successfue an th f r o d fo xe equipmentus l , cobalt-6 acceleratorr 0o ; expert personnel, excellent infrastructure, good communications, large patient load, excellent dosimetry, efficient organization and supporting structure. Consequently, taking into accoun e existinth t g situatio d modificationan n s whic e feasiblar h d an e sustainabl economi e vien eth i f wo sociad can l situation, priority shoul givee db acquirino nt kine g th equipmenf do t which is most likely to be able to function in the local environment (climate, staff, supporting services, operating resources usede )b whero t . s i t ei REFERENCES 1. Cancer Treatment, WHO Technical Report Series No. 322, WHO, Geneva, 1966. 2. Planning of Radiotherapy Facilities, WHO Technical Report Series No. 328, WHO, Geneva, 1966. 3. Optimizatio Radiotherapyf no TechnicaO WH , l Report Serie 644. sNo , WHO, Geneva, 1980. 4. Research Coordination Meetin Testinf go Dosimetrf go y Equipment, Internal Report, IAEA, Vienna, 1989 . 5 Organizatio d Traininnan Radiotherapn gi r Developinyfo g Countrie n Africai s , IAEA-TECDOC-614, IAEA, Vienna, 1991. . 6 Radiation Dos Radiotherapn ei y from Prescriptio Deliveryno t , IAEA-TECDOC-734, IAEA, Vienna, 1994. 7. National Cancer Control Programme - Policies and Managerial Guidelines. A Handbook produced by the World Health Organization Global Programm Cancer efo r Control, WHO/CAN/92, WHO, Geneva, 1992 (limited distribution). . 8 Advisory Group Consultatio Desige th n no n Requirement Megavoltagr sfo e X-ray Machine r Cancesfo r Treatmen Developinn ti g Countries, expected publicatio American Pa e th y nnb Health Organization, 1995. 9. WHO Manual of Radiotherapy and Cancer Management, expected publication, 1995. 1 9 T EX left ALTERNATIVE DESIGN MEGR SFO A VOLTAGE MACHINES FOR CANCER TREATMEN DEVELOPINTN I G COUNTRIES C BORRAS Pan American Health Organization, XA964284i3 Washington, D.C., USA H. SVENSSON Imtemational Atomic Energy Agency, Vienna, Austria O.P. HANSON World Health Organization, Geneva, Switzerland Abstract In developing countries radiation therap oftes yi n performed with antiquated cobalt-60 units, the radioactive source f whico s e lonhar g decayed and, thus, treatment e ineffectivear s . Furthermore the cost involved in the disposal of spent radionuclide sources discourages owners from proper removal and storage, and accidents occur. Although present design of microwave electron linear accelerators provide excellent beam characteristics, developing countrie mann si y location t hav infrastructure no eth o sd maintaio t e n such machineso T . explore the possibilities of designing, taking advantage of the latest advances in technology, a more elementary electrical teletherapy machine, inexpensive in first cost and maintenance, American thePa n Health Organizatio n/ Regiona l Worle Officth f do e Health Organization Americae foth r s organize Washington di Advisorn na y Group Consultatio Desige th n no n Requirements for Megavoltage X-Ray Machines for Cancer Treatment in Developing Countries, with the collaboration of the World Health Organization, the International Atomic Energy Agency and the United Nations Industrial Development Organization. It was attended by 40 radiation oncologists, physicists, technologists and engineers representatives from radiotherapy equipment manufacturers. Afte analysin a radiotherapre th f so y situation world-wid eespeciall- y fro viewpoine mth maintenancf o t consensua e- reaches e swa th n do radiotherapy equipment performance requirements. To meet these requirements, several accelerator designs were considered. Among the most promising new designs were the klystron/linac and the high frequency linear accelerator, the microtron in a radiation head, the high frequency betatron -alsa radiatio n oi nacceleratorsC D head- d an , . Possible treatment designs, including thos f modulaeo r nature, were presented. Sinc estimates i t ei d that by the year 2015 - barring a dramatic and unforseen cure for cancer - a total of 10,000 machines wil needee b l provido dt e treatmen estimaten a r fo t millio0 d1 case w r yeanne s pe r in developing countries, the impact of such high technology simple machine could be substantial in providing equity and quality for the management of cancer patients. 1. INTRODUCTION Accordin estimatesO WH o gt , currently ther approximatele ear y nine milliow ne n cancer cases per year, worldwide. This number is expected to increase to about 15 million new yeae caseth r y 2015sb , with about two-third f thesso e cases occurrin developinn gi g countries. 93 Radiotherapy will r year fo ,mose comeo st th te b ,importan t therapy approacr hfo mos f theso t e tumors, bot curr palliationd hfo ean . Surgere th o limites t y i rols e it edu n d i advanced stagediseasee th f so s encountere developinn di g countrie chemotherapd an s s yi expensive. In developing countries cance w typicae ne th , r0 patientl 15 incidenc o t r 5 spe 7 s ei 100,000 population. To serve a current population of 4.4 billion, assuming 4.4 million new cancer cases per year - 50% requiring radiotherapy - and one machine per 500 new cancer cases treated currene th , t neetotaa 4,40f s di o l 0 machines yeae th ry 2015B . , barringa dramati unforseed can n cur cancerr efo totaa , f 10,00o l 0 machines wil needee b l provido dt e treatment for an estimated 10 million new cancer cases per year in developing countries. Presently it is estimated that in developing countries approximately 2,300 megavoltage teletherapy units are installed, primarily cobalt-60. Unfortunately many of these units are antiquated, have received very little maintenance over the years and their radioactive source lone sar g decayedeconomicao t e Du . l constraints, source replacement intervaly sma years0 1 o t , p especiallu e b privatn yi e institutions [1]. Data obtained throug e 199hth 2 IAEA/WHO postal dosimetry intercomparison program for high energy radiotherapy units, show tha tunite morth sf o testee tha % Latin di n50 n America Caribbead nan n countries would require treatment times of over 4 min to deliver 2 Gy to the tumor. To compensate for the low absorbed dose rates at the treatment distance and still treat a very large number of patients becoms ha t i , ecommoa n practic shorteo et treatmene nth t distance -often without correcting the percentage depth dose tables in clinical use - and to give lower doses than necessary. In no case are doses increased to compensate for the low dose rates. The consequence f theso s e practice t onlno y e inaccuratar s e doses being delivered (the 1992 intercomparison showed error f moro s e than e fac38%!)th t t thabu , t treatmente ar s ineffective, fosterin concepe gth t that cance incurables i r . Thu healte sth h authoritiet no o sd assign proper budgets to radiotherapy services. Furthermore, the cost involved in the disposal of spent radionuclide sources discourages owners from proper remova storagd an l accidentd ean s likonee eth Ciudan si d Juarez, Mexico [2], and Goiania, Brazil [3] occur. The industrialized countries have started the process of replacement of cobalt-60 units and most radiation oncology departments in the United States and in Europe have switched to electron accelerators. However, some of these units are very expensive and difficult to maintain and the infrastructure to properly use them is often lacking in developing countries. Thus, for the purpose of improving the availability of radiation therapy for cancer treatment, manufacturers and research laboratories are being encouraged, taking advantage of the latest advances in technology, to consider the design and development of a megavoltage x-ray machine much simpler than present microwave electron accelerators, a machine that could be used both in developed and developing countries. addreso T s thi American s issuPa e eth n Health Organization/Regional e Officth f o e World Health Organization for the Americas organized in Washington an Advisory Group Consultation on the Design Requirements for Megavoltage X-Ray Machines for Cancer Treatment in Developing Countries, with the collaboration of the World Health Organization, e Internationath l Atomic Energy e UniteAgencth d dyan Nations Industrial Development Organization. It was attended by 40 participants: radiation oncologists, physicists, technologists and engineers representatives from radiotherapy equipment manufacturers. 94 Afte analysin a radiotherapre th f so y situation world-wid eespeciall- y fro viewpoine mth f o t maintenance - the Group reached a consensus on performance requirements and proposed various novel accelerator designs. 2. PERFORMANCE REQUIREMENTS To reduce complexity and improve safety a single energy photon unit without electrons is recommended. (It is assumed that there is access to at least superficial x ray machine treatmense witth r hfo energie tumorf depta V to k o t f h0 o p su s30 betweed an 0 n10 3cm.) - Treatment time (average 2 fields/patient): 10-15 min/patient - Patients treated/8 hours day: 32-48 - Dose rat isocentet ea r (dept dosf ho e maximum): minimu Gy/mi8 m0. n recommended 2-3 Gy/min 2.1 Mechanical data - Isocentric design recommended - Isocentric height above floor level <130 cm, preferably 115 cm - Isocentric clearance (with all devices) >35 cm - Source-isocenter-distanc , preferablcm m c 0 0 e>8 y10 The floor surface should preferably be flat (no pit). (A small depression is acceptable). Collimator jaw and distance indication: mechanical or electrical with mechanical backup. 2 2. Couch motion radiatiod san n field size - Isocentric, rotation is preferred. - Angle of rotation ±90° - Lateral range ±20 cm - Vertical belo rangm c w0 e 7 isocenter preferred. patiene th t a thic tm surfacc kcm 5 patient2 2 4 f - eo Fiel X 2 ,d4 shoulsizeo t p su d 2 be available from above with lowered table (Maximum Field Size at isocenter >30x30cm2) 2.3 Radiation Beam Quality beae Th m qualit defines yi parallea n di l opposed 1 beam0 configuratio1 x 0 1 a r nfo cm 2 field size and a patient thickness of 25 cm using equal beam weights. In this configuratio followine nth g should hold: - Depth of superficial 90% isodose2 <5 mm - Hot spot relative a central target3 =115% - Penumbra width < 1 cm, and preferably < 8 mm % - ±3 Uniformit fielf yo dove % (IECr80 ) highls Ii t y desirabl t greate t spono e ho s i ttha e r th ttha n 110% relativ centraa o et l target. (See Table I for Dose nuxmum/Dose »* for various beam energies). 95 About 65% of all radiation therapy is with two opposing fields and there is a strong preference for the higher energy (6 MV) instead of using more than two fields with lower energy (e.g., 2.5 MV equivalent to cobalt-60). (See Table II). To treat superficial lymph nodes. To avoid fibrosis. Table I DM/DA RATIO VARIOUR SFO S BEAM ENERGIES Patient Thickness Machine Depth(mm) SSD 20cm 25cm DM DM (cm) DM/DA DM/DA Co-60 5 1.8 80 114% 127% Co-60 5 2 100 111 123 4MV 10 4 80 110 120 4MV 10 4 100 108 117 6MV 15 7 100 106 112 8MV 20 9.3 100 104 110 V M 0 1 25 11 100 102 107 Table II 115% DM/DA REQUIREMENT 25 cm thick patient 1:1 Parallel Opposed Fields Photon Energy Co-60, 4 MV No 5MV treatmentD SS K sO 6MV SAD treatments OK 8MV Better, DM/D A110= % Electron Beam Curren GyMn,SAD)n 3 ( ta 0 10 6MV 100 4MV 200 96 4 Device2. available b o t sradiotherapr - efo y treatment - light indication for field size with central axis indication - distance indication with mechanical backup - isocentric indication with mechanical backup - wedges 15°, 30°, 45°, 60" light field preferably visible after insertion, orientatio wedgd nan e angle interlock. - shadow tray(s) for standard and customized beam blocks. - possibilit tako yt e megavoltage port films 5 Safet2. y - Compliance with FAO/IAEA/ILO/OECD-NEA/PAHO/WHO International Basic Safety Standard r Protectiofo s n against Ionizinge th Radiatio r fo d nan Safet Radiatiof yo n Sources wels a ,nationas a l locad lan l safety regulations. 2.6 Quality and Maintenance - Uptime >95% - Service interventions* < I/month - Preventative maintenanc 4/yea< r e - Self diagnosis is recommended - Component potential failure status read-out recommended - Because clinical treatment cannot continue 2.7 Serviceability/Reliability Specifications 2.7.1 Preventative Maintenance There shall be specified intervals for Preventative Maintenance. The integrity of e machinth guaranteee b shal t manufacturer'e no l th f di s schedul t followedno s i e . These specifications consume other maintenance thae regulath n r (daily/weeklyA Q ) checks on the equipment. 2.7.2 Failures requiring intervention These shall be classified according to the level of skill required to rectify the fault. With good education programs, som thesf eo e intervention handlee b y house"n sma d"i . a) First Line - This is a local engineer trained by the manufacturer or the manufacturers local organization as appropriate. Diagnose problem to unit level using standard fault finding practices mosn I . t cases isolate faul o Printet t d Circuit Board level .problems f o Solv % e90 . b) Second Line - Manufacturer's engineers usually at a regional rather than hospital levels. System oriente remainine f problemsth o df abl- o solv% o et % g10 e80 , i.e. 98%. c) Third Line - Manufacturer's engineer of a senior level usually at the Head office. 97 2.7.3 Meantime between failures Failures capabl monthfirsf 3 eo t> linse repair Failures capable of second line repair > 1 year Failures capable of third line repair > 10 years 2.7.4 Target Planned Maintenance Schedules requirey da 1 d ever monthy3 s day3 s required every year 2.7.5 Spare parts A stocking strategy should be defined, e.g.: First line repairs on site Second line repairs regional Third line repairs manufacturer 3. ALTERNATIVE NEW APPROACHES varioue Th s accelerator technologies considered were subjectivelyr o ranke , B , dA C depending upon their practicality and likely ability to meet the above requirements. Ranking Definition A Technology is most likely to meet the requirements and merits further exploration B Technolog probabls yi y relevant C Technology is not likely to meet the requirements, and is not recommended for exploration at this time but should be retained for future reference. 1 3. CategoryA Microwave linac Microwave power source - linac Microtron in radiation head 2 Beam klystron/linac accelerator Modular design linac 3.2 Category B Betatro radiation i n head Rhodotron Continuous wave rf linac or microtron EC Accelerators 98 Cascade voltage multiplier Laddertro Pelletror no n charged electrostatic accelerator Nested high voltage generator Transformer coupled high voltage generator 3 3. CategoryC Plasma wave accelerator Induction linac Interlaced accelerator structure Superconductig linac using superconducto berylliur ro m coated cavities Small synchrotron Multipl powew elo r magnetron Accelerator activated short lived isotope . 4 BRIEF DESCRIPTIO PREFERREF NO D UNITS 4.1 Hie Klystron/Linac and High Frequency Linear Accelerator Each of these offered the possibility of a compact 6 MV gantry mounted accelerator klystron/linae Th . integratee th d can d klystron/accelerator waveguide system t requirno o d eparticulaa r resonance frequency. Higher frequency accelerator f eitheso r the linac or microtron type would be less massive. 4.2 "Mini-Microtron" This design allows the possibility of a compact (35-40 cm diameter) 6 MV radiatioe th microtof o mountep ne b to head gantrsmal e e o ne t th Th th n .ldt i y a siz f eo the accelerator can be achieved either by use of a high frequency or by a high energy gain per turn. It was considered that an integrated magnetron/cavity design might be advantageous. The microtron layout allows the field flattener to be in the fringing magnetic field thus reducing secondary electron emission photoA . n beam e spoileb y rma employed for control of dose build-up. The accelerator should fit in a 80 cm SAD gantry gantry,D possiblSA .m c y0 eve10 na 3 4. "Mini-Betatron" In this design a small, 25 cm diameter donut betatron mounted in the gantry at the top of the head provides a compact 6 MV machine. A high frequency (possibly 10 Khz) is required to provide adequate output. A DC bias can be used to double the energy gain. 4.4 DC Accelerators A relatively compact gantry mounted accelerator can be achieved using DC accelerator principles. This machine was considered to be potentially highly reliable. However, the 2-3 MV energy achievable was generally considered lower than required by most users. Heavy filtratio f thino s beam would allow beam characteristics similao t r conventionaa acceleratorV M 4 adequacr e o beaTh 3 le . th m f y o curren t obtainabls ewa questioned. 99 4.5 Classical linacs (rf) Commercia linacV f M compaco s 6 4- l t desig e widelar n y available, though general concern cosf reliabilitsd o tan y were expressed. 5. CONCLUSIONS e currenTh t manufacturer f electroo s n linear accelerator microtrond san s should be encouraged to design and prototype a super-reliable 6 MV x-ray system subject to the established performance specifications. This encouragement should come from accelerator designers who might cooperate with the manufacturers as well as representatives of the developing nations who can best make the case for their needs. REFERENCES [1] BORRAS, C. "Cobalt Therapy Source Replacement and Disposal Problems in Latin America", Proceedings World Congress on Medical Physics and Biomédical Engineering Janeiroe d o Ri ., Brazil (1994). [2] SECRETARÎA DE ENERGÏA, MINAS E INDUSTRIA PARAESTATAL, "Acci-dente por Contaminacion con Cobalto-60". Mexico (1984). [3] INTERNATIONAL ATOMIC ENERGY AGENCY, "The Radiological Accident in Goiania, Brazil", Vienna, Austria (1988). [4] PAN AMERICAN HEALTH ORGANIZATION, WORLD HEALTH ORGANI- ZATION, INTERNATIONAL ATOMIC ENERGY AGENCY, UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION, "Design Requirement r Megavoltagfo s e X-Ray Machine r Cancefo s r Treatmenn i t Developing Countries". In preparation. 100 SIMULATIO RADIATIOD NAN N TREATMEN TN I XA9642849 EXTERNAL RADIOTHERAPY E. SINGER Mevaterapia Medical Center, Buenos Aires, Argentina Abstract It is well known that in order to obtain a uniform dose in the treated volume as defined in ICRU 50, there should be a 10% maximum difference between maximum and minimum dose values in treatment planning. Clinical target volume (CTV) should be related to external areas of body sections where tumour is located. These areas are important because different radiation beams enter through them. Therefore, verification plannine th f o g target volume (PTV) throug externae hth l area highls si y significant. thin I s worimportance pointh e t kw ou t controllinf eo irradiates i g thaV PT t s da planned considering some error sources usually found in radiotherapy practice with equipment thas beeha tn intensively a lonuse r gfo d time. Moreover I thin, k this experience will be helpful for those centers around the world where radiation treatment is carried out with reconditioned units. 1 PATIENT POSITION WITH RESPEC TREATMENO TT T TABLE D SIMULATOAN R EQUIPMENT TABLE severas A l step involvee sar simulatioa n do n table before treatmen starteds ti e th , differences between simulator table and different treatment tables must be considered in orde avoio rt d incorrect transferenc daile th yo et set-up . The treatment table and the simulator tables are built differently and what is more important with different pre-alignments. They are located in different bunkers, and many times the simulator room is smaller than the treatment one or vice-versa; so the simulato treatmene th r o r t tabl limites e i movementss it n di , generating differene us t modes. Frequency of use is approximately 10 times higher in the treatment equipment simulatoe thath n i r causing deformation firse th t n si one . e cas th f intensivelo en I y used machine e treatmenth s t flatno tabl ,s i e longitudinally and transversally, due to the fact that patients always use the same side greated an , o climt it r weighn o b s exertee i aret th an do whero n table s eth ha e support. An example of treatment table's deformation is observed in large fields such as the mantle field which encompasses almost the entire patient width. In this situation we foun t difficuldi encompaso t t s equally both mastoid epiphysi wels s a axilla s a l r areas, without modifying patient's position by rotating his/her head one or two degrees, and slightly widening the treatment area, since these areas are included in the PTV. furtheA r proble face th t ms thai t while many tables rotat degree0 e18 havd san e sections with different materials this is not the case on a simulator or on other treatment tables. For instance, when irradiating a mammary volume, the patient can be set-up on the center of the couch if it has a mylar foil without metallic parts on the borders; otherwis e patieneth t doe t ressno t couce firml th tendd n h yan o mov o st r eo rotate. 101 The weight of the patient's body on the mylar, racquet or on a rigid table as is the case with a simulator, changes the anterior-posterior diameter and the incidence of the lateral beams on the patient. An exampl differen e gives ei th y nb t contour same th f eso patient e taketh t na same sectioe treatmenth n no t equipment simulatore th , e scanneth , a r n tablo r eo NMR used to determine theCTV. e contouth factore f I th s modifiei r o t s e jusdu dt mentioned isodoses values around the PTV will become altered. If the contour does not coincide with the one that correspond scannerT erroC % e location 3 i r th a , o st n and/or siz lesiof eo n coule db produced. Furthe face rth t differencethao t e t treatmendu e ar s t table usualle sar y narrower than simulation ones. Let us analyse the case of a mantle treatment in hyper extension position. Her patiene eth abls i t movo et armss ehi . This cause transversasa l expansion in the lungs seen in successive films. It sometimes originates an effect by which lung protections coincide with the film in which they were outlined, but not always with the verifyin gequipmente filth mn i . In order to treat the mantle (inverted "Y") in the same position as it was treated before wit normae hth l mantle rotatee b couco e t dth ,s throughs ha a ; ° angln ha 90 f eo it is usually not perfectly isocentric, the patient has to be repositioned. Besides, the chang armn ei s positio change th d patienn eni an t position when film cassett places ei d under him/her , differenp makega e sth t from that mathematically calculated. A further problem easily solved is due to the quite common practice of using non-tnountable mats to cover couches, and when removing one part of them where the beam impinge flae th t , suppor sbodon e th deformedf ys i o t . In summary, the three main error sources when irradiating mantle fields are: 1) the differenc position i e patient' e th f no s arms, 2)couch rotation aroun isocenters dit , and 3)placemen file mth f cassetto t e unde patiene rth obtaio t nbettea r image. 2 RELATIVE POSITION BETWEEN THE PATIENT AND ACCESSORIES IN CONTACT WITH HIM same Th e accessorie streatmene mussimulatoe th useth e n b to n do d t ran unit . Pillow se patient' musth t fi t s head, neck bacd r shoulderan ,ko s fixin e patiength o s t that uncontrolled movemen avoidede b n ca t , allowing repeatability mose Th . t common shapes should be available in small, medium or large sizes. For irradiation with electron beams, where surface flatness is very important, these supports are essential. Immobilises They are particularly useful for treating head and neck and essential for children r patientfo r o s with involuntary movements. Immobilisatio prevenn ca n t lateral head rotation more easily tha anterior-posterios nit r movement. Lateral head rotatioy nma be controlled by lateral lights and by entrance and exit of radiation beams while these t registebondno o santerior-posteriod n ra r movement requiring tattoo markr se o th n si radiation field vertexes. Given the fact that in certain cases each unit must handle as many patients as possible, time is saved and doses are accurately administered, if the patient is immobilise correce th n di t way. These immobilises shoul ease db placeo yt . Theo yd not only prevent movemen t alsbu to make each treatment eas reproduceo yt , avoiding uncertainte th thud irradiatioe yan s th volumef no s larger than necessary. 102 Lowering shoulders device This is useful for lateral fields treatment where fields should not be discontinued taking into account lesion characteristics. Careful attention must be paid to possible deformations of head position and these devices should differ depending on whether they are fastened to the patient or on to the couch. When the lower part of a cervical field runs through a shoulder, the source surface distance (SSD) may vary within 10 shouldere th n i dose cm% smorer th dependino 70 .e d o riset an typ,e p sth f u eo n go machine. Belly flattening device This should be used for those patients with overlapping adipose tissue due to which the maximum dose on the skin takes place on skin folds and produces skin lesions adipose Th . raisee eb tissuordeo n t d i s avoio erha t irradiatione s db it o t s ha t I . repeatee b donn thay ca suc n det i i twa dailyh a d keepinan , g fixed references when these tissues move. Wide, stiff bandages fastened to the side of the couch with a Styrofoam wedge under patient's buttock usede ar s . References mus e markeb t n di non-movable patiente areath f so , possibl t withiyno n treatment field t whicbu , h clearly define withi admissibln na e margi daile nth y set-up. Such a modification must be carried out locating the PTV on a contour taken with this new external configuration. Support for the arm in breast treatment The main objective is to achieve a position easy to reproduce daily in order to irradiat e movableth e mammary volume which dependpositionm ar n so , particularly when ganglion areas are to be irradiated. The breast support must be different for the different treatment units, depending on the type of table, the telemeter position and the SSD. At an SSD of 80 cm. there is collision risk, therefore it is convenient to place the lateral support behin patient'e dth s arm. Support mus linkee angulan b ta o dt r scale with respect to the longitudinal axis of the table. The main problem arising here is that, when using the same device for the different treatment units, in some cases it hides the distance scale. Notwithstanding the cumbersome procedure it is better to use it, than to place the patient's arm under the neck to irradiate the mammary volume and modify its position to irradiate the ganglion areas, as the PTV is defined for a specific arm position. differene commoe th f Th o e t naccessorieus s make t convenieni s havo t t T eC scans taken at treatment position and with the accessories in place. Bolus Different bolus types are: t gauzwe ) ea bolus: skin attachmen goods i t . However, tissue equivalence differs accordin wateo gt r absorbed the d difficule an y,ar adapo t t widthr fo t excesn si 1 f so cm. Width errors plus water absorbe dosn i d errore reac% y delivered2 o hma t , . b) wax bolus, tend to harden and must be heated to shape them adequately so they stick to skin; they are helpful in shaping sections to eliminate oblique incidence. Tissue equivalence mus calculatede tb . 103 tissu) c e equivalent bolus somn i bese t :e th thebu t case e yar s ther areae ear n si which intermediate air cavities are difficult to avoid, due to lack of adherence. Width, shap typd boluef an e o s shoul takee db n into consideration when carrying out planning. 3 RELATIVE POSITION BETWEEN PATIEN EXTERNAD TAN L ACCESSORIES Some other widely utilised accessories will be analysed in terms of how their variations and/or deficiencies affec alter o t r treatment plannin wels g a dail s a l y dose. Wedges In some machines they may become loose with respect to their support and displaced even though they are well hooked up. Should wedge with its nominal maximum field siz usede eb somn i , e cases therappeay ema slia rwhic n i t dose hth e value is much higher, thus altering dose uniformity. In addition, technicians must be require placo dt e wedges last, onc gantre e th bees yha n rotate incidence th o dt e angle, and the collimator angle and depth have been controlled. Protections Routine protections must be fixed by means of threads and screws. The wearing out of the threads may generate displacement of the protections due to their weight when the equipment is rotated. They should have no rim which might deform shadows. The Pantograph should be checked mechanically at regular intervals. Patients' protections must be controlled on the basis of the X-Ray film on which they were built, and verified by x-raying. The simulator must have trays at the same distance of each unit and similar shields made of wood or Styrofoam. X-ray film Whe e cassettth n s placei e d unde e patientth r , his/her positio becomy ma n e deformed r example Fo mantle . th n i ,e conveniens fieli t di placo t t e acrylic platef so the same depth as the cassette under the regions where the cassette does not reach; so the patient is at the same hyper extension position as before. In oblique field it is convenient to utilise a lectern with angles allowing the placement of the X-Ray film perpendicularl beame th o yt . 4 RELATIVE POSITIO HEADF NO , ARM LEGD SAN S To fix a patient's body in space the number of degrees of freedom should be considered, and the bond condition should be fixed. Usually the different immobilisation modes for head and neck take into consideration a transversal axis to the couch. To control rotation around a longitudinal axis is more complex and can be done taking the SSD at two fixed points as references. It is important to position the patien e simplesth n i t t possible ways, tryin movo gt e couch, gantr collimatod yan r instea movinf do patiene gth t especially when using electron applicators. 104 Treatment position with arms above head is frequently used in breast and oesophagus irradiation but must be avoided unless absolutely necessary. This position promotes patient movement, especially rotation and is less comfortable. When irradiating the ganglion areas and the mammary volume in the case of the breast, a new erro addes ri d when field overlappee ar s t waisa s i t r disjoinedm levedo r ar fo e l th f i , the ganglion field and under the head for tangential fields. In such cases it is more difficul transfeo t t r data fro simulatoe mth treatmene th o rt reproduco tt unid an t e daily treatment accurately. It is very important to register patient positioning with a photograph, controlling tattoos in different planes and the SSD in different points of fielde th , where skin surfac t perpendiculano s ei beae th mo rt axis 5 POOR NUMBER OF TATTOOS AND THEIR MOBILITY Parallel, opposite d lateraan ,r slantino l g fields rotating aroun n axie a dar s erroneously defined by only one point at an angle of 0° If this is the initial position and patientattooe o n th t clearl s no fiel ha tn s i i yd D registereentranc SS cas n i e ed th ean d patient'e inth s treatment report sheet vers i t i y, difficultechniciae th r fo t knoo nt w whethe treatmene th r t fiel wels i d l located. This method saves tim r eacefo h patient, and is usual when there is only one technician per team Coincidence of tattoos, depths, protection and wedges cannot be verified, This way of working is the simplest one but it leads to mistakes; the appropriate way is to train the staff in order to improve their skills for such a job Otherwise conflicts tend to arise between physicists and technical staff, usually closely involve n thesi d e issues. Tattoos shoule b d registere photo n dfieli n i dpermanentld e b films an d san y availabl eacr efo h patient 6 THE MOBILITY OF THE ORGANS INVOLVED IN THE TREATMENT e followin e mosTh th te importanar g t causes which makconsiderablV PT e y higher than CTV. movemene th ) diaphrage a th f o t m muscles when diaphragmatic couple muse b t irradiate ovarn di y treatment b) lung movement due to pleuro-pulmonary synchrony. e followinTh g organ y fall ma s inte irradiateth o d volumo different e du e t movements: kidne) a y position accordin patient'o gt s positioning examplr fo , e when they must protectee b ovarn di y treatmen markee ar d dan t ecographically with patient standing. bon) b e marro heaw n neci d dan k treatment when ther nece ear k movements. These o caselasmodify tw t ma s y target volume position with respeco t t treatment portal volume and therefore, should be taken into consideration in the original planning. 7 CONCLUSION Give e typd amounth n an e e possiblf somth o t f o e e cause a differenc f o s e between dose delivered and dose prescribed, an adequate and close relationship among operators, physicists, and radiotherapy physicians is crucial Regular staff meetings for patient control in treatment rooms, as well as regular workshops, are highly recommendable KEXT PAGE(S) I 105 loft BLANK ANALYSIS OF VARIATIONS IN THE DOSE DELIVERED IN RADIATION THERAPY XA9642850 D.B FELD Comisiôn Nacional de Energia Atômica, Radiaciôn-Gci Aplicacionese ad , Division Dosimetria Buenos Aires, Argentina Abstract The outcome of radiotherapy in cancer care is heavily dependent on the quality of the treatment. This work present dailreviesa w yho f wpractic o currene th d etan availabilit equipmenf yo r fo t treatment planning and simulation as well as a number of other factors affect the radiation therapy qualit Argentinan yi e establishmenTh . f refreshino t g coursel typ al f staf o er fo sf involved in the treatments, modernisation of equipment and strict routines in patient set up and quality control would give a significant contribution to a higher quality in radiation therapy. 1 INTRODUCTION In Argentina radiation therapy is mostly accomplished with Co-60 machines and linear accelerators. The number of linear accelerators is increasing very fast; the first one was a 6 Mv installe 1979linacsn 3 d i ther2 thef w o e m8 eNo ar ,. have been installed between 199d 3an 1994. Hal thef fo m X-Ra v havM e6 y beams othere ;th s hav5 1 e X-Rad an v y M beam0 1 f so electrod an v nM beams with energiethif o sd yeaen se r3 Mev ranginth 8 1 r o .Fo t gv froMe m4 new linacs wil operatinge lb othee th n r O .hand numbee th , f simulatorro t increaseno d sha n di the same way. the first one was installed in 1978 in a center with a Co-60 machine as the high energy unit; now there are 11 simulators all over the country. Medical centers operating with linacs, hav t leas emedicaa e on t l radiotherapy physicist working . Exceptionalliit n y centers tak morn eo e thae physicistnon , even when they have several treatment machines. Sometimes physics technicians (or dosimetrists) are taken on in Radiotherapy centers which only have Co-60 units mant Bu . y centers havin gCo-6a 0e unith s ta main treatment machin operatine ear g wit physico hn s till staflal nowt fa . Manvariatione th f yo s dose inth e delivered, analyze thin di s presentation, have nothin wito d hgo t typ machinf eo t ebu with the number of patients per hospital and with the staff. So, the first and essential task is to get a staff including radiotherapists, physicists, dosimetrists and technicians (machine operators) working close together for the accurate performance of any Radiotherapy Center. 2 ANALYSIS OF DIFFERENT FACTORS CAUSING DOSE DELIVERED DIFFERENT FROM DOSE PRESCRIBED In this presentatio startine nth g assumptiopoine th s ti n that routinary mechanica dosimetrid lan c checks have already been accomplishe resultd an de withi ar s n tolerance values. Also that treatment planning for each patient have been already checked. So, the analysis refers to daily errors which can be named "daily accidents". These daily accidents are more difficcult to record an solvo dt e than r instancefo , , calibration uncertainties where b en figureca % 3 s r liko e2 established 107 3 SIMULATION e pooth ro t numbe e Du f simulatoro r s operatin t presenta g , verpatientw fe y e reallar s y simulated. If the center is operating without simulators, the treatment machine is used as a simulato r somfo r e selecte dadvantage patientsth s ha f t havinI eo . g identical conditionr sfo simulation and treatment (e.g. the treatment table), but the disadvantage of poor definition, particularl case numbe th w f Co-60 eo fe n y i f patientd o r an , s selecte r simulationdfo e th n O . other hand simulator t alwayno e ssar properly used field) referencea :d an s s localized wite hth simulator are not always verified on the treatment unit, giving rise to different errors in field size, patient position, etc causind .an g change volume th n si e encompasse e isodosth y db e curve already selecte radiotherapise th y db t (treated volume) [1] radiotherapisusualle ) ;b th s i t yi o twh defines fiel donle h sizes yd takeAn . s into account wha "seese th " from localization X-Ray films and fro mradioscopia Clinicae th esimulatory e viesa th l Targeo n wt o s i t i , t Volume (CTV] [1 ) without considering that the treated volume may be different for different machines and different treatment techniques. In the very frequent and simple case of parallel opposed fields, when varying irradiatio becomy n conditionsma eV underdoseCT e th , d nea bordee rth r [2]. This dose gradient within the CTV varies between 5% and 10%, depending on type of machine, separation distance betwee field ndan fields sizeD .SS , Whe linana c stop repairede b so t workin s ,a ha d gan great numbe f patient o rwitn o ho sg thei r treatmen anothen i t r machine ,timee mosth f so t different from the first one. Field sizes are not changed and dose distribution will be different. It is importan emphasizo t e that ther center e Co-6d ear ol e s0 havin on lina v d M c gan 6 onle yon unit which operates for instance at an SSD of 60 cm. Usually with an SSD of 60 cm,the isocenter technique is not used, and so field sizeis defined on surface. With the same field size for linac and Co-60, dose gradient may reach to 12% within the CTV 4 TATTOOS AM) REFERENCE POINTS Very ofte radiatioe nth n fiel patiene set-us i dth n po t having onltattooe yon , usualle th n yi techniciacentee fielde th t man th f f I .no ro ys nindicationha patienr sfo t positioning t oftei s n(a , occurs) the patient may happen to be rotated with respect to the right position; the effect is the same as having the collimator rotated a few degrees; if it happens many times along treatment, the result is a widening of the penumbra and a different treated volume, giving additional dose gradient to the PTV. 5 REPEATABILITY In orde assuro t r repeatabilite eth treatmena f yo t schedule, technician need havo t s e clear reference points and enough number of tattoos as well as SSD or depths clearly indicated. When using isocenter treatment techniques, with lateral or oblique incidence, it is very dangerous (but it happens havo t ) e onltattoe yon o whicreference th s hi e whe gantre t n0°.Usuallth a s yi y technician deptr doe o positionint verifd D shno an oblique ySS th r gfo e fieldsy . ma Agai V nCT receivine b t no righe g th irradiate e t th dos d ean d greatevolume b y rema thanecessarys i t ni . There are a number of different "daily accidents" which have to do with the great number of patients per treatment unit. Some of them are; 1) Error in 0° gantry positioning due to parallax; an error of 1.5° in the gantry angle gives a lateral field displacement of 2mm. This displacement happens always in the same direction as the technician looks the 0° gantry angle always from the same side. 2) to leave the block tray for patients who dont need it; if calculations for these patients have been done without tray attenuatioe th f i mucs % na 6 factor e s hb a erro e y th , rma tray is an entire one. 3) error in SSD when it is read after the entire block tray is put in place; light scal alteres ei d when tra places yi d between focu patientd differenca 6 san r d o an ,5 f o e mm can happen. 108 6 VERIFICATIONS Agreement between treated volume and PTV must be checked with X-Ray films during treatment wels a , vers a l y simple changes like variation patienn si t diameter differenca , m 1c f eo in tumor depth results in an error of 6% in the dose delivered for a 10cm x 10cm field and a Co- 60 unit [3]. When there is only one medical physicist doing all the treatment planning and calibration, verification scarcele sar y eve d carriean n t importandou t accident happenn sca . followine Th g examplcasn a es i thisf eo : treatmenA t plan wit hfixea dwedg ° a use compemsatofiela d 30 s dda f an e o decides rwa r dfo a patient; and a series of mistakes were committed: 1) The physicist chose a wedge shorter than needed. He/sh) 2 e forgo indicato t t collimatoe eth techniciae th wedg e ro th s angl t placd n ei npu e an e with the angle in the opposite direction. 3 ) As the physicist had a lot of work to do he/she had no time to verify field set-up on patient. 4)Tthe technician realized that the wedge was very short and decided to change it; he looked for another wedge longer than the first one, but he didn't take into account the wedge angle and he put a longer wedge but of 45° instead of 30° without notifying the physicist about the change. Half of the dose was delivered to the patient in the wrong way 7 FRACTIONATION Usually treatment plannee s ar fraction 5 r dfo weeksoma sn i t e;bu circumstances patients miss some sessions. For instance, a) centers with only one treatment machine are not able to send all the patients for treatment to another hospital when machine comes out of service; b) people with economic diffîccultie sometimee ar s troubln si atteno et d hospital everydaye th f o particulad n i , en e th t ra month. Therefore thesn i , e cases effective dose delivere patiene th do different s ti t from dose prescribed. 8 EDUCATION AND TRABSING In Argentina ther regulatione ear operatinr sfo gRadiotherapa y Center more Th . e important are: Physicist) 1 s (wit hdegrea Physicn ei Engineeringr so physiciand )an licensa st neege n ei o dt order to be registered as radiotherapy medical physicists and radiotherapists respectively. Only those who have the license can be in charge of a Center. Both have to pass the Course "Dosimetry in Radiotherapy" carried out by the Atomic Energy Office (CNEA); physicists have to pass in addition the Course "Physics in Radiotherapy" also carried out by CNEA after passin coursese gth , radiotherapists must fulfil trainina l g perio year3 physicistf d do s an 1 f so year, carried out in Radiotherapy Hospitals. requiremente Th beinr sfo gdosimetrisa thav o aret ) esecondar:a a y ordinary degree paso t ) s;b the Course "Dosimetry in Radiotherapy" and c) to get practice in dosimetry, in radiotherapy hospitals, unde supervisioe rth medicaa f no l physicist durin yearg1 . radiotherapA ) 2 y hospital wher linaea operatins ci includo t e staf s th medicaa fgha n ei l physicist having the corresponding license. 3) This license is renewed each 5 years in order to be sure that they have continuity in radiotherapy performance. ) Eac4 h hospitao takt s e ha par l postan intercomparisonsi D t TL l D TL . 4 Ther e ar e intercomparison yeasa thao rs t each center take sleastt para intercompariso e , in ton , yearr npe . 109 Intercomparison SSDe th carriee y Lb sar (CNEAt dou anonymou n a n )i s way resulte ;th e sar reporte eacdo t h center wit codha e number. 5) To install a new radiotherapy center, requirements for equipment and staff have to be fulfilled: A) minimum equipment include treatmen) a : t machines with high energy photon beams (Co-60 units with an isocenter distance of, at least, 80 cm and/or linacs of, at least, 4 Mv X-Ray beam); b) an X-ray treatment machine operating at 50-90 Kv or a linac with electron beams having energies between 6 and 8 Mev; c) afterloading intracavitary applicators and sources for brachytherapy simulatoa ) ;d r machin diagnostia r eo c X-ray machine adapte localizationr dfo . B) a licensed medical physicist or, at least, a dosimetrist supervised by a physicist, must be part of the staff for new radiotherapy centers where a Co-60 unit is the main machine. e minimuth , So radiotherapw m ne staf r fo f y centers having Co-60 unit r linaco s s maia s n machines include radiotherapists, medical physicists and/or dosimetrist techniciand an s s with their corresponding licenses. Unfortunately there are not, till now, regular radiotherapy courses for the education of the technician operato swh treatmene eth t machines; ther coursee ear s intende thosr wano dfo ewh t to be diagnostic radiologist technicians; some radiotherapy elementary notions and practice are imparted along these courses, but they are not enough. Refreshment courses partially solve this techniciane th l situatioal t regulat theno t no s d ne attenbu y ar ran d these courses. 9 CONCLUSION cleas Ii t r that each radiotherap ypracticn i hospita t pu qualitea r centeo t o l s yha r assurance program. It is clear also that regular courses are necessary but not enough to assure quality in daily practice. Regular meetings among all the members of radiotherapy staff are very helpful to coordinate treatment schedules analyzo t , e different error mistakesd san importans i t ;i insiso t t on the need of precision in radiotherapy and how it can change treatment results; verification films must be analyzed also with the technicians in order they can realize the different errors and changes which happen in daily practice. And it is important to emphasize that every indications writtee hab recordeo sd t nan eacn do h patient treatment sheet Partial replacemen treatmend ol f o t t Argentinaunitn i equipmen w comins sw i ne no t n gBu .o t are in many cases not new but repaired machines and they look as new machines and are very often out of service.The other important point is that the great number of patients per machine and the few number of physicists and technicians, conspires against quality. When technicians are accostumed to work very fast, to be only one for each unit and to reduce time for the set-up, vers ii t y difficcul reverso gooa t t s i dhabitse t i ethin th , So remin.o g t workine dar tha e w t g with patients to whom high doses are delivered; when a wrong dose is administered to a wrong volume, little can be done after. REFERENCES [1] ICRU REPORT 50,1993. [2] Radiographie tumor localization~J.C.Rathe,M.D. and P. Elliott M.S.~Warren H. Green, Inc., St. Louis, Missouri, USA. Qualit] [3 y assuranc anatomican eo l data (Table I)—A. Dutreix Worksho Qualitn po y Assurance, BarilocheS.Ce d . , Argentina, 1988 110 (a). INTERCOMPARISON .'.JAS left THE ROL SSDL-HELSINKF EO DOS1METRR IFO D YAN QUALITY AUDI RADIOTHERAPTN I Y XA9642851 P. AALTONEN Finnish Centr r Radiatioefo Nuclead nan r Safety, Helsinki, Finland Abstract Quality and dosimetry audit in radiotherapy has in Finland been implemented through inspections carried out by the Finnish Centre for Radiation and Nuclear Safety (STUK). In connection with the Radiation Metrology Laboratory Centree oth f SSDL-Helsinkie th , , ther sectioa s ei radiotherapr nfo y supervision inspectioe Th . STUy nb n a Ks i independent revie qualite th dosimetrd f wo y an y control system callee whicb n dh ca qualit dosimetrd yan y audiy b t site visits. STUK is the responsible authority for the supervision of all use of radiation in Finland and that is why it also can set up requirements on the basis of results of the review. The disagreement of the measuring results between STUK and the radiotherapy department, of more than a given action level, will always lead to a through investigation of the reason and to a discussion on the most reliable results to be used for the treatments. The inspections include dose calibratio conventionar nfo l X-ray therapy equipmen dosd tan e comparison, including field size dependence, for high energy equipment. For afterloading equipment the reference air kerma rate is checked. Additionally inspectione ,th STUy sb K include check performance th f so e characteristic equipmene th e f so th d an t accomplishment and the results of quality control procedures. Further, methods are currently being developed to supplemen direce th t t measurement TL-measurementy sb specian i s l phantom ordern i s includo t e wholeth e treatment chain (e.g treatmene .th t planning system audite th n )i . . INTRODUCTIO1 N Quality audit is a review of the quality control system and dosimetry audit is a review of absorbed dose determination, performed by an independent person or body which is not responsibl performance th r e fo produc e th f eo procesr to s under review audie .Th t shoulde th n i , final stage of its implementation, cover all steps of the radiotherapy procedure, i.e. the final aim should be to ensure that the prescribed dose is given to the patient. The quality and dosimetry audit can be established by three different ways: (1) postal measurements (i.e. by mailed dosimeters) (2) site visits (3) combination of postal measurements and site visits. Audi sity b te mosvisite th ts s i efficien t method, sinc l interestineal g parameters can be easily checked and the necessary actions can immediately be started. During a site visit all aspects of the work can be discussed with local personnel and a comprehensive review of the overall accurac possibld yan e problem attaindede b n sca . However, site visit eaco st h clinif co the area service "audie th y tdb centre becomy "ma e relatively expensive, especiall biggen yi r countries a suitabl d an , e combinatio f postao n l measurement sitd e moran s b visit n e ca s appropriate. Qualit dosimetrd yan y audi radiotherapn i t Finlann i s yha d been implemented through the inspections carried out by the Finnish Centre for Radiation and Nuclear Safety (STUK). For practical reasons, the quality audit operations have been connected with the standard dosimetry activities SSDL-Helsinkie ,th fact n botI e . th ,h activities form togethe speciara l sectio r "thno e laboratory for radiation metrology" at the Department of Radiation Safety of STUK. The combination ensures, among many other things e traceabilitth , f measuremento y e th d an s maintenanc f higeo h dosimetric competence. 113 While the inspection by STUK can be regarded as an independent review of the quality control systeradiotherape th f mo y department (quality audi sity b t e visit) t differi , s froe mth general principle f qualitso yaspectsrequiremento p u auditw t ) STUn se i t(1 :e n th K ca n o s disagreemenbasie th f result reviews) e o (2 th measurin e f d so th an ,f to g results between STUK andepartmente dth f moro , e tha givena n action level, will always leathrouga o dt h investiga- tion of the reason and to the discussion on the most reliable results to be used for the treat- ments. DEVELOPMENE TH . 2 AUDITINF TO G ACTIVITIE FINLANN SI D The Finnish Centre for Radiation and Nuclear Safety is responsible for the supervision f radiatioo e us l Finlandn oi al f . Becaus Finnise th f eo h legislatio nlicenca e issue STUy db K is required for use of equipment used in radiation therapy. A through inspection by STUK is carried out before any radiotherapy machine is taken into use, and then regularly (every two or five years) or whenever a significant repair has been made or a significant change in the output has been observed. The dosimetric competence and activities were concentrated at STUK in the 70's mainly because of the legislation, but also due to the pressure on the hospitals side. In the 70's STUK offered dose distribution measurement service n ordeo s f hospitalo r a computerize d an s d radiation beam scanne developes wa r t STUKda . In 1977 STU nominates Kwa membea s IAEA/WHde a th n ri O Networ Secondarr kfo y Standard Dosimetry laboratories. At the end of 70's the old ^Co-units and betatrons were replaced by linear accelerators and Quality Assuranc radiotherapn ei y became actual overale Th . l accurac targee th n ty o dos e of + 5 % was considered important in the international littérature [1,2]. Some international protocols contained detailed requirement 198n I recommendatione . 0th QA n so Nordie th y sb c Association of Clinical Physics (NACP) were published [3]. In 1981 a "radiation dose commit- tee" gav s reportit e , wher t proposei e ddetailea d progra r acquirinmfo radiotherapw gne y equipment within the next 10 years, and stated the need for QA according to the principles by the NACP. STUK cam conclusioo et n thacentrae th t l dose measurement system mus supplee b t - mented by the increased responsibility of the therapy clinics for dose measurements and QA of treatment units. In the beginning of the 80's the hospitals started their own dosimetric measure- ments. The hospitals suffered from a lack of adequate personnel resources to undertake the increased duties on QA. In 1985 a meeting of STUK and the representatives of each therapy clinic was organized to discuss both organizational and technical questions of QA. In the meeting it was agreed that the QA programs should cover all procedures which affect the accurac patient e f dosyth o o et . In 1987 the Radiotherapy section was founded in the Department of Inspection and Metrology (today: the Department of Radiation Safety) to be responsible for the supervision of radiotherap Finlandn yi . 114 . SUPERVISIO3 STUY NB K In 199 1radiatiow Finlanne a t n dgo legislation Radiatiow ne e t stipulate.Th n Ac s that a regular control of the performance of each piece of radiotherapy equipment has to be arranged. Under the Radiation Act STUK has issued instructions concerning quality assurance for radiotherapy equipment (ST-Guide 2.1) [4]. In the following the main principles of the established practic presentede ear . 3.1. Safety licens advancd ean e control A special safety license is needed for radiotherapy, for the use of each therapy equip- ment, according to Radiation Act. The list of equipment which are covered by the license is giveappendin a n licensee ni th o xt piec .A takee f equipmenb eo t n no intbefore y ous ma tn ea inspection by STUK has been carried out, unless otherwise specified in the license. To take into equipmenw ne a remov o t oned e r ol modificatioa us o t, n e a applie e b license o t th df s n.o eha for. conditioA license th f no e requires specifthao clinie t th s t organizatioe cha th y r nfo safety inside the clinic. In particular, the persons responsible for dose measurements, for the quality contro f radiotherapo l y equipment r radiatiofo d an , n safety arrangements e havb o t e specified. safete Th y licensmodifications it d ean grantee sar STUy db written Ko n applicatioy nb the clinic requestn O . , STUK will giv advancn ea e statemen radiatioe th n o t n shielding plar nfo radiotherapy rooms f neededi d an , , carrieadvancn a t sou e inspectio r radiationfo n shielding. Information on radiotherapy clinics and equipment is entered into a computer register maintained by STUK. The important inspection data, machine'faults, radiation accidents and abnormal incidents are also recorded in the register. 3.2. Inspections A new piece of radiotherapy equipment, the rooms and the compliance of the operation wit license hth inspectee ear STUy db K befor equipmene eth takes i t n into inspectioe useTh . n is carrie t alsdou o after significant repairequipmene locatios th it clini e f f so i th s d n cnha i an t been changed. Thereafte inspectione th r repeatee sar d regularl followss ya : At the minimum every two years: high-energy and conventional X-ray therapy equipment radiotherapy simulators At the minimum every five years: afterloading equipment The information received from the clinic may give rise to an extra inspection. 115 Table 1. Contents of the Inspection for Different Radiotherapy Equipment Object of inspection High energy Conventional Afterloading Radiotherapy treatment X-ray therapy equipment* Simulator equipment equipment* Compliance with the license A, R A, R A, R A,R Structural radiation shielding A A A A (Guides SS 2.8, 2.9 and 2.10, . ST Guide 3.6 when applicable for simulators) R , A Radiatio R , A n safety arrangement R , A s A, R (Guides SS 2.8, 2.9 and 2.10, and applicaple parts of machine standards) R , A Dose calibratio R , A n (see claus eR 3.5), A . Radiation beam characteristics A, R A, R A, R - uniformity, symmetrd yan penumbra - depth dose characteristics Dose monitoring characteristics A, R A, R A, R - repeatability and proportionality - dependenc gantrn eo d yan collimator angles R , A Mechnica l R characteristic , A s A,R - radiation field indicators - angle, field sizdistancd ean e indicators - isocentricity - treatment or imaging table movements - laser beams Imaging and fluoroscopic A,R** characteristic GuideT s(S 3 s3 anwhe4 d3. n applicable) Results' of quality control R R R R whe(S4 3. T nd Guidean 3 s3. applicable concerning simulators) R , A A Dos eR plannin, A g system firs= * tA wheinspection applicabln beforee taking inte ous ** when needed R = regular inspection 116 APPENDIX A Suggested Details of the Inspection for High-Energy Treatment Equipment. Gantry angle (GA) and collimator angle (CA) are 0° unless otherwise stated. Only values of parameters or machine settings which will be used are considered. Object of inspection A. First inspectio . RegulaB nr befor inspectioe takinn g inte ous 1. Total procedure 2. Measurements by STUK 1. Total procedure . Measurement2 STUy sb K Dose calibration All radiation qualities, l radiatioAl n qualities, dose As A2. As A2. dose rates, wedged an s rate wedgesd an s . Selected field sizes. field sizes. Radiation beam radiatiol Al n qualities. Field Minimum e settinon :r fo g All radiation qualities. Minimum e photoon : d nan characteristics , réfcm .0 1 x m sizec 0 1 s each radiation quality (i.e. Minimum: field sizes réf. one electron quality. For Uniformity fiele on d d sizan esizex ma , field size, GA, CA, measu- and max. Minimum: GA 0° each quality, minimume :on Symmetry betweelaso ttw onese nth . ring depth). and 90°. Minimum° 0 A C : setting (fiel, dCA size, GA , Penumbra Minumum: GA 0° and 90°. and 90°. Minimum: one measuring depth). Minimum: CA 0° and 90°, measuring depth. whe 90°A G n . Measuring depths: réf. depth (all GA and CA) and another depth (only GA 0°, CA 0°). Depth dose All radiation qualities, l radiatioAl n qualities, réf. l radiatioAl n qualities, As A2. characteristics wedges, fielde sizeson r Fo . field size. selected field sizes. electron energy, minimum: two values of GA. 00 APPENDI (contXA ) Object of inspection A. First inspection before taking into use B. Regular inspection . Tota. Measurement1 2 l procedur STUy sb e K . Tota1 l procedur• e . Measurement2 STUy sb K As Al. Dose monitoring charasteristics Repeatability l radiatioAl n qualities. Minimum e photoon : d an n Minimum: one photon and electroe on n energy. Mini- one electron energy. Mini- e mumsettinOn : g (GA, mum: four field sizes for CA). photon l fielal dd applisan - Proportionality Minimum: one photon and cator r electronssfo . electroe on n energy. Dependenc n gantro e y l Al radiation qualities. Minimum e photoon : d an n and collimator angles Minimum, 90°0° A ,G : one electron energy. Mini- 270°. Minimum: CA 0° and mumo fieltw d: sizer fo s As A2. 90°, when GA 90" or 270°. photons and one field size for electrons. Mcchanicaä characteristics As Al. Accuracy of radiation All radiation qualities. field indicators Minimum: four field sizes for photons and al! field As A2. . AsA2 applicators for electrons. As Al. All photon energies. Mini- AsAl. Minimum: one photon mum: two electron ener- energy. Minimum: two APPENDI (contXA ) Objec f inspectioto n A. First inspection before taking into use B. Regular inspection 1. Total procedure 2. Measurements by STUK 1. Total procedure 2. Measurements by STUK indicator furthe t a fielr d AsA2. Angle, field sizd ean anm dc 0 1 x m c size 0 1 s Angle indicators, minimum: distance indicators max, when GA 90° and CA one angle. Field size indi- 0°, 90°, 180° and 270°. cator, minimum o fieltw :d Distance indicator further, sizes. Distance indicator, As A2. whe 90"A .nG minimum: one distance. As Al. Location, sizconsistd ean - As Al. enc f o mechanicay d an l radiation isocentres. . AsAl Isocentricity l moveAl . - ment theid an s r indicators. Isocentricity Depend on the case. As Al. All lasers. Treatmen r o imagint g Selected movements. . A2 s A table movements Laser beams . AsAl Minimum: angle indicators As Al. ° intervalsa90 t , field size indicator at 5 cm intervals, d distancan e indicatot a r three distances. Field size AppendiTabld an e1 specifxA contente inspectionyth e th f so inse detailsn si th - r Fo . pections STUK maintains high quality equipment for absolute and relative dose measurements. protocoA eacn o l h inspection wit resulte hth f measurementso clinice gives si th o nt . 33. Quality control qualitA y contro lsubmitte e prograb o t s mSTUo dha t approvar Kfo l withi yeae non r fro date mth e piecwhew ne f equipmen eo na s takewa t n into use. International recommenda- tion appliecriterie e sar th r f approvadfo ao l [3,5,6,7]. However qualite th , y control measure- ment s e startee equipmenthavb th o t f ed o fro e e beginnins specifieus a m,th e th n f i dgo connection with the first inspection by STUK. The quality control program of a high-energy equipment shall always include dose calibratio e equipmentth f no e accomplishmenTh . f o t quality control procedure resulte th inspectee d sar s an STUy db K durin regulae gth r inspections of the equipment. 3.4 Dose calibration of treatment equipment 3.4.1 Conventional X-ray units The dose calibration of a conventional X-ray therapy unit is defined as the deter- mination of absorbed dose rate produced by the equipment. "Absorbed dose" for this case is de- e absorbefineth s a d d dos o surfacwatet e e th t a rf wate eo . r phantomm 5 depte 0. th f mt ho a The dose calibrations of conventional X-ray units are carried out by STUK. The dose rat measures ei tubue on r s d(deefo p therapy separatelr )o eaco yt h tubus (superficial therapy). resulte dose th Th ef scalibratioo givee Dosna ar n i e Table, wher measuree eth r calculatedo d dose rate is given for each tubus or as a function of field size. 3.4.2 High-energy treatment equipment For radiotherapy electron accelerators, the dose calibration is defined as the deter- mination of the relation between the dose produced and the setting of monitor units. For gamma beam therapy equipment, the dose calibration is defined as the determination of the dose rate. "Dose hers "i e defineabsorbee th s a d d dos radiatio e wateo et th n o r n beam dept e axisth t ha , of dose maximum in a water phantom, when the surface of the phantom is at the normal treatment distance dose th er calibrationFo . above th , e relatio determines ni oper dfo n fiels a d well as for each wedge field, using the reference field size. In addition, the dependence of this relatio fiele th d n nsizo s determineei r checkeddo botr fo , h ope wedgd nan e fields e checTh . k of the uniformity of the field shall always preceed the dose calibration. t measuremente Th dose th er calibratiosfo carriee STUy nar b t connectioKn di ou n with each inspection of the equipment (see Table 1 and Appendix A). The dependence on the field size, which the clinic has to determine before taking into use the equipment, is then checked at selected values of field size. The results of measurements at the reference field size are compared wit correspondine hth g results obtaineinspectione timclinie e th th f th eo t y ca db n I . the comparison, the following action levels are applied: 120 consistency of dose in open field at the reference field size % 1 photons % 2 electrons consistenc% 1 wedgf yo e factor consistency of field size dependence 1 % difference Ith f f resulteo s doe exceet sno actioe dth n level resulte clinie th ,th e y car sb used to prepare a Dose Table. In a Dose Table, the monitor unit setting which produces the dose of 1 Gy at the given depth in water, is given as a function of field size. The Dose Table is the ultimate result of the dose calibration. If the action level is exceeded, the reason will be examined mose th td reliablan , e resul measuremenf o t Dose basi e taketh s i t th ef s o Tablen a . 3,4.3. Afterloading equipment dose Th e calibratio f afterloadinno g equipmen hers i t e define checkins da refe -th f go erenc r kermeai a rate. This chec mads ki STUy eb K regularly. 3.5. Requirement givinr sfo g information e RadiatioTh n Degree state a numbes f mattero r n whico s e clinith h s responsibli c r fo e informing STUK. All changes of information included in the license, considerable changes in the intended use of the treatment equipment or in the approved quality control programs, important machine fault repairsd san radiatiod ,an n accident incidentr so s affectin radiatioe gth n safety shal informee lb STUKo dt . STUK must als informede ob resule th f dosf i to , e calibration for a radiotherapy electron accelerator (dose per monitor unit), taking into account all possible adjustments of the calibration, differs by more than 5 % from the value agreed on during the latest inspectio STUKy nb . 3.6. Calibratio radiotherapf no y dosemeters STUK maintains a calibration service for radiotherapy dosemeters. The dosemeters for dose calibratio radiotherapf no y equipment shal recalibratee b l d every three years. Calibrations are performed agains Finnise th t h national standard, whic secondara s hi y standar calid dan - Internationabratee th y db l Burea Weightf uo Measured san s (BIPM) resulte Th . calibratiof so n are give Certificata n i calibrationf eo , wher calibratioe eth n facto absorber fo r d dos wateo et r is tabulated as a function of radiation energy. 3.7. Supporting activities Supervisory activitie STUf so backee Kar traininy db g activitie continuoud san s devel- opmen f methodo t proceduresd san . STUK gives trainin advicd gan Qualitn eo y Assurancr efo radiotherapy equipment, both through direct contacts with individuals and through special oc- casions maintaio T . hige nth h expertism require works it r d,fo STUK undertakes also research on dosimetry and on other aspects which affect the overall accuracy of the dose to the patient. STUK maintains contacts with nationa internationad an l l organization dealing wit same hth e subject area participated an , nationan i s internationad an l l cooperation e knowledgTh . d an e information gained STUK transmits furtheclinicse th o t r . 121 . RESOURCE4 FUTURD SAN E DEVELOPMENTS The above quality and dosimetry audit system concerns about 25 accelerators, 10 afterloading units, 11 simulators and 11 conventional X-ray units in 9 radiotherapy centres in Finland. Abou physicis2 t t manyear requires si operato dt systeme systee th th cose f mTh o .t is approximately US $ 250 000 per year, most of which is due to the salaries. The operation is mainly financed by annual charge for the hospitals. The dose planning systems have not been properly included in the current quality audit procedures by STUK. A project has been undertaken to develope a special phantom, which could be used to check the whole chain of radiotherapy: from CT-scanning to calculation of the dose distributio targen i t volume teste Th .s wit phantoe hth m would then supplemen othee th t r measurement r qualitsfo dosimetrd yan y audit. The benefit of the quality and dosimetry audit is quite evident based on the experience throug site hth e visits from several years discrepanciee .Th s observed includ casesf o t elo , where the observation would not have been possible or would have been less probably if only postal procedures had been applied. REFERENCES ] INTERNATIONA[1 L COMMISSIO RADIATION NO N UNIT MEASUREMENTD SAN S (ICRU), Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Ray Radiotherapn si y Procedures, ICRU Repor (1976)4 2 t . ] WORL[2 D HEALTH ORGANIZATION (WHO), Optimizatio f Radiotherapyno , Repora f o t WHO Meeting of Investigators, WHO Technical Report Series 644 (1980). ] NORDI[3 C ASSOCIATIO CLINICAF NO L PHYSICS (NACP), Procedure n Externai s l Radiation Therapy Dosimetry with Electro Photod nan n Beams with Maximum Energies Between 1 and 50 MeV, Acta Radiologica Oncology 19 (1980) 55-79. ] FINNIS[4 H CENTR RADIATIOR EFO NUCLEAD NAN R SAFETY, Quality Assurancr efo Radiotherapy Equipment, ST-Guide 2.1 Helsinki (1993). [5] BRAHME, A., Accuracy Requirements and Quality Assurance of External Beam Therapy with Photons and Electrons, Acta Oncologica Supplementum 1 (1988). ] INTERNATIONA[6 L ELECTROTECHNICAL COMMISSION (IEQ, Medical Electrical Equipment, Medical Electron Accelerators - Functional Performance Characteristics, International Standard IEC 976 (1989). ] INTERNATIONA[7 L ELECTROTECHNICAL COMMISSION (IEC), Medical Electrical Equipment, Medical Electro - Guideline V nMe Accelerator0 r 5 sfo o t RangV e th Me n si e1 Functional Performance Characteristics, Technical Report IEC 977 (1989). 122 SSDL ARGENTINA: DOSIMETRIC INTERCOMPARISON PROGRAMM COBALR EFO THERAP0 T6 Y UNITS M SARAVI, S. PAPADOPULOS, H. MUGLIAROLI Comisiön Naciona Energie d l a Atômica, Buenos Aires, Argentina Abstract Thermoluminescence dosimeters (TLD widele )ar y useo verift d y absorbed dose delivered from radiation therapy beams. The Secondary Standard Dosimetry Laboratory (SSDL) of Argentin s maileit r fo d a usedosD TL se intercomparison programm r cobalfo e 0 6 t radiation therapy units. Results obtained since 197 wels 8 a causes a l f dosso e discrepancies greate analyzede rar tha % n5 . Result externae th f so l quality control performe IAEe th Ay db for this programme indicate that the dose evaluated by the SSDL TLD service for the participating centers is about 1% lower than that evaluated by the IAEA TLD service. This deviatio acceptes ni d takin accounn go dos% t2 thae± uncertainta t dosimetrD TL r yfo s i y reasonable. 1. INTRODUCTION Regionae Th l Reference Center (RRC Dosimetrr )fo f Argentinyo Secondara s ai y Standard Dosimetry Laboratory (SSDL) belongin IAEA-WHe th o gt O SSDLs Network s similaA . r laboratories existin othen gi establishes r wa countries necessite C th o 196RR t n d i e e yth ,8 du f improvino dosimetre gth treatmend yan t planning radiatioe th n si n therapy centero t d san increase the participation of physicists, specially trained in this field, within the staff of these centers. The most relevant activities performed by the RRC in the field of radiation therapy are: dosimeter calibration service, calibration programm cobalt-therapr efo y e countryunitth n si , advisory about clinical dosimetr r radiatioyfo n therapy centers, organizatio f post-gradno e courses for physicians and for physicists specialized in radiation therapy, and a dosimetric intercomparison programme for cobalt 60 therapy units. In 1977, throug IAEe hth A Research Contrac 1791/RBstarteC C R t RR develop o dt e th , ea national postal dose intercomparison programme or cobalt-therapy units using TL-dosimeters. e firsTh t dosimetric intercomparison took plac 1978n i e . Since then, abou 4 dosimetrit c intercomparison year spe r have been made, includin caesiu8 d cobal 8 therapg7 8 an m0 13 6 t y units belonging to public and private centers. s participateha C e IAEth RR n Adi e dosimetriTh c postal intercomparison programmr efo SSDL cobalr sfo beams0 6 t mord an , e recentl similaa n yi r programm higr efo h energy X-ray beams with deviations lower tha. n1% In 1992 the RRC participated in the Coherent and Accurate Reference Instrument (CARE) Programme for the IAEA/WHO Network of SSDLs. During 1992 and 1993 the RRC participate e IAEth An i d Quality D postaControTL l C l RR Programm e th r fo e intercomparison servic orden ei teso rt t this intercomparison system. From these results i t si possible to evaluate the present situation and trends of the SSDL of Argentina with regard to its postal dosimetric intercomparison programme. 123 . METHO2 D The method and technique employed by the RRC for the postal dose intercomparison programme were described in previous papers [1], [2], [3]. Briefly, the dosimeters consist of LiF powder contained in plastic capsules. A batch of capsules containing annealed powder TLD-70sen d posy an b tradiatio o t C prepare0s i RR e n th therap y db y centers. Each center receive dosimetris3 c capsule irradiatior controsfo e on d ln an capsul e th e t irradiatea y G 2 o dt RRC. The participating center is requested to irradiate each of the three capsules separately to a dose of 2 Gy to water, in a water phantom, at the central axis of a vertical irradiation beam, at t eithea source m c rth 0 1 surfaco e t depthx m fielc e m c e5 d Th .0 use distance 1 sizb s di o e t e (SSD) or the source to capsule distance, depending upon the usual technique employed at the center. The irradiation is coordinated so that all participants and the RRC irradiate during the same week in order to avoid any fading correction. The participants have to filll in a data sheet givin methoe gth absorbede useth r dfo d dose determination. This help finreasono e st dth f so dose discrepancies betwee dose nth e quote participane th dose y dth b ed evaluatean t RRCt da . calibratioe th r Fo TL-dosimeterf no use IAEC e sth RR A e Internationasth l Cod Practicf eo e [4] for absorbed dose determination: (Sx w =MD ,,ir)D u N p x ux where: absorbe Nth Ds i e dose chamber factor; M« is the meter reading, (corrected for ambient parameters). (Sw,,ir)u is the mass stopping power water to air ratio and perturbatioe th Ps ui n correction factor. measuremente madTh e watea ar n ei C r phantoRR t s a centra e th mt a l verticaaxia f so l cobalt 60 beam, with the ionization chamber centered at 5 cm depth and correcting for effective point of measurement. A 10 cm x 10 cm field size at surface is used. Once the TL-dosimeters return to RRC 11 measurements corresponding to a batch are made on the same day. From each capsule 3 TL-readings are obtained. The mean value is determined for each capsule being the standard deviation of these readings better than 1.5 % . The TL- readings are normalized to reference powder readings. The calibration line is obtained and the dose delivered by the participant is determined by interpolation in this straight line. The total TLD-systeuncertaintC . RR e 2% th f ± ymo s i After the dose delivered by the participant is evaluated at RRC, the per cent deviation, Dev(%), betwee dose nth dose quoteth particiant e ed th evaluatean y d, b RRCe th QD , y d,b calculates i , ED eacr dfo h participating center: Dev(%) = (QD - ED) x 100/ED 124 3. RESULTS Results for the 88 cobalt 60 therapy units in operation in Argentina are summarized in Figure 1 to 3. The results obtained during the first participation of each unit (year 1978 to 1981) are show Figurn i . Onle1 unitf yo abous% delivere5 4 t dose dth e withi intervae nth l ±5%e .Th mea11.7%) = 1.90 = o n( D dosy 8s ,G wherewa e centers with dose discrepancies greater than ±30% wer t considereeno calculatioe th r . Figurdfo D f no summarizee2 resulte sth r sfo secone th d participatio centerf no thin si s programme (year 198 2198o t f o 3. Abou) % 0 7 t participants obtained dose deviations lower than ± 5%. The mean dose was D= 1.970 Gy (a = 6.4%) abreviateo T . , Figur summarizee3 resulte sth s obtained during year 199 o 19921t . About 80% of participants delivered the dose within the interval ± 5%. The mean dose was D= 1.99 wit 3.4%y = 0 G ho . With data sheet informatio calculato t C reapossible RR s i e th e t l ni th dose r efo , (QD)*, given capsuleo t participantse th ceny r sb pe t e deviatioTh . n between (QD)evaluatee th d *an d dose ED can be calculated: Dev *(% 100/E(QD)x ( ) )= ED D* - Figures 4 and 5 show the Dev* (%) distribution for first and second participation of centers in the intercomparison programme. Results improve significativel f Dev*(%i y s consideredi ) . This means tha mann i t y cases dose discrepancie erroro t e dosn si du e sar calculations n I . order to correct this problem the RRC sent to each center the corresponding information, the method for dose calculation and a list with the last recommended factors for dose evaluation. Nowaday significano sn t differences between Dev(% Dev*(%d )an founde )ar . According with the information given in the data sheet about 45% of centers in Argentina applies the IAEA Internatinal Code of Practice [4] for dose measurements. About 30% of centers use watesa r usephanto% s25 measurementd m an r dos fo er ai determination n i s s according to recommendations given in ICRU 23 [5]. 4. DISCUSSION 198n I authoritiee 0th Nationaf so l Health Ministr Nationae th d yan l Commissio Atomir nfo c Energy approved regulations for radiation therapy centers. These regulations include equipment and staff requirements for the authorization of operation of those centers. The participation of physicists within the staff of centers is required . The participation in the RRC dose intercomparison programm s considerei e d within this normative too e center:th e ar s oblige participato dt t leasea t onc nationae yeaea th n ri l dose intercomparison. Nowadays about 50 physicists specially trained in radiation therapy are included within the staff of radiation therapy centers. Through the intercomparison programme it has been noted centerf o tha% s90 t with physicists delivere dose dth e withi acceptee nth d interva . Dos5% el± discrepancies greate easile ar ry % thacorrecte5 n± thosn di e centers wit hphysicisa e th n i t staff. The use of the IAEA International Code of Practice for dose measurements [4] is linked with the presenc physicistf eo t radiatiosa n therapy centers above Th . e mentione f centero % sd45 that uses this Cod dosr efo e measurements hav physicisea t withi staffe nth . 125 Frequenc] y[% o r 25 ° r- O C2O, ent o , H _i O O (a 2n. « ! ~' g, cn O 0 1 (D ""*• < o O «-*• v^O Ï 0s*11 ——— - — — »A1 o 1 I II 1 << c 5. cn «-*• c« Ï •1o. o o •5' o* 3 « •_. ro o Frequency [%] -s. ro O > co o ro 4ä. o oo l __ l __l_____l l ~ hfl Sf M O 5» n H r1 O o n o" o> Öl 8 ü §- (D EL < » • o O r—i n 3 01 y C/3 O O •O •5' § o 01 Î o 8ZI Frequency [%] 3) 00 O j___i___i ' rio I o 8» ÇA €•*• CO O O 'f'f.f-ffff.ffrfffff g ;f-'tF~4.fc&^-f,^&if,ffJi*f.~&è O* n 01 o O 8- (D O\ ï O • o 'fMfffMfàâ I«•^- C rV I» 0> »Hvot O VO vu O Frequency [%] o ro £» ) ix o> o Öi__i__D i _j__i__ö) 00 Oi l| LII o °£ oo f^ T >—' CO -4 l jjf O ™L o o O9 T 3 cn o 8- ö «» m _ o\ 7^- o ,, _. «. * O '' c 3 O O cn a5J O l' Ï ^ «3 o Frequency [%] o r8? If ° Wl o O n ü o !s n lu01 ° ?* o ''f^^^^M^^^/^Jj^^^^^^iy^^^Ê ^j^^y^^^ ^mmm^mmi^"^"^ o co £L S 0 5. TEST OF THE METHOD Durin gparticipate C 199 199d RR 2IAEan e e 3th th n Adi Quality Control Programme th r efo RRC TLD postal dose intercomparison service. The QC programme included: a) reference irradiation at IAEA of TL-dosimeters from RRC; b) participation of IAEA as a radiation therap yIAEf servico center D n Aru TL paralel n e) i service; c D TL l witC . DistributiohRR f no capsule coordinates swa thay sucn di twa participatinha g centers irradiateC , IAERR d Adan durin same gth e week. y 1992IMa n , previou e irradiatioth o t s n window establishe r TL-capulesfo dC RR e th , receive IAEe dth A CARE system consistin electrometero tw n gi ionizatioo tw d san n therapy level chambers to be calibrated at RRC. These dosimeters were calibrated in a horizontal cobalt 60 beam, in air, using the Secondary Standard NE 2560 Therapy Level dosimeter with graphite chambe 256E N r 1 belongin RRCe th e calibratio o gTh t . n facto r termn i ai r f o s kerma, NK, was obtained for each CARE system. The CARE dosimeters were calibrated in a water phantom too, in a horizontal cobalt 60 beam at 5 cm depth on the central axis. The calibration facto termn i r f absorbeso d dos wateo et r ND.r eacfo Wh CARE dosimetes wa r determined. Results obtained in the IAEA CARE Programme participation are summarized in Tablel. The maximum difference between the NK factors obtained by the RRC and those reported by Ne Dth ,W IAEr factor-0.6s Fo A . maximu wa e 3sth % m deviatio valueC RR sf no wit h regard to those reported by IAEA was + 0.31% . TABLE I. RESULTS OF PARTICIPATION IN IAEA CARE PROGRAMME CA4 D10 CAD 105 Ion chamber TK02 chamben Io r TK02 ser. No 104 5 10 sero N . Mean air Kerma calibration det. by IAEA 1.599±0.8% 1.598±0.8% NK(Gy/V) Mean air Kerma calibration det. by the SSDL 1.589±0.9% 1.590±0.9% NK(Gy/V) Deviation -0.63 -0.50 Absorbed dose to water cali factor det. by IAEA ND>W (Gy 1.599±1.0% 1.599±1.0% Absorbed dose to water cali facto rSSDLNe detth y .b D,W( 1.599±1.1% 1.599±1.1% Deviation + 0.31 + 0.17 Deviatio (RRC-IAEA)xlOO/IAEn= A 131 The results of participation in the IAEA QC Programme for TLD postal dose intercomparison summarizee ar C RR servic Table n di th f wherI eeI o ratie eth o betwee dose nth e evaluatey db RRCe th , EÜRRCI participatine th r fo , g dose centerth ed evaluatesan IAEAy db , EÜIAEAr fo , the same center showe sar n .Accordin theso gt e result dose sth e evaluateTLDC RR -y db system is about 1% lower than that evaluated by IAEA TLD-system, beeing the standard deviation 1.1%. TABLE H. RATIO BETWEEN EÜRR EDiAEd C an A D E Participant RRC /EDiAEA TLD set No R002 0.9961 R003 0.9926 R004 1.0073 R005 0.9891 R007 0.9963 R008 0.9735 R009 0.9975 EL/1-93023 1.0055 EL/1-93024 0.9824 EL/1-93025 0.9931 EL/1-93026 1.0000 EL/1-93028 0.9664 EL/1-93029 0.9867 EL/1-93030 0.9888 EL/1-93031 0.9827 EL/1-93032 1.0054 Mea 0.9n= 95 1 CT =1. 1%, 6. CONCLUSIONS The results obtained by RRC in the IAEA CARE Programme and in the IAEA QC Programme for the RRC TLD system show the coherence of dose measurements. The average deviation -1.0% obtained by the RRC in the IAEA QC programme for RRC TLD-system is acceptable taking on account that the total unceratinty for dose determinations with the RRC TLD system is ± 2%. The RRC of Argentina has gained great experience in the TLD intercomparison programme for cobal 0 therap6 t y units. Increas f visito e o radiatiot s n therapy centers shoule b d implemente orden di improvo rt resulte eth thif so s programme. Increas physicistf eo s trained in radiation therapy will hel improvo pt dosimetre eth n yi those centers. 132 National regulation r operatiofo s f radiatioo n n therapy centers have give e necessarth n y sustain for succesfull application of the RRC dose intercomparison programme. Improvements in the dose delivered by radiation therapy centers have been obtained through this programmme. REFERENCES [1] R. GONZALEZ, M. SARAVI, " Dose Intercomparison Programme of the Regional Reference Center of Argentina", IAEA-TECDOC-249, Vienna (1981) 45-50. . SARAVIM . ] PAPADOPULOSS ,[2 . LINDNERC , , "Intercomparison Programmf eo Absorbed Dos Cobalr efo Therap0 6 t y Unit Argentina"n si , Dosimetr Radiotherapn yi y Vo , IAEA2 l , Vienna (1988) [3] M. SARAVI, S. PAPADOPULOS, H. MUGLIAROLI, "The SSDL Argentina Ready o Providt e Intercomparison Servic r HospitalFo e n Latin-AmericaI s , "SSD L Newsletter No 32, IAEA, Vienna (1993)15-23. [4] INTERNATIONAL ATOMIC ENERGY AGENCY, Absorbed Dose Determination in Photon and Electron Beams. An International Code of Practice, IAEA-TECDOC- 277, Vienna (1987). [5] INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS, "Measurement f absorbeso d dosphantoa n ei m irradiatea y db singl gammr eo bea X mf ao rays", ICRU(1973)3 2 o N , . A PROGRA QUALITN MO Y ASSURANC DOSD EAN E CALIBRATION FOR RADIATION THERAPY UNIT VENEZUELN SI A M.C. DE PADILLA, L. CARRIZALES, J. DIAZ, F. GUTT . COZMAA , N Laboratori Calibraciôe od n Dosimétrica, Institute Vénézolan Investigacionee ed s Cientificas, Caracas, Venezuela Abstract The results of a five year program (1988-90-91-92-93) on quality assurance and dose calibration for 12 Cobalt-60 units from public hospitals, which represents 30% of total radiation therapy units in Venezuela, are presented. The remarkable improvement in the general performance of these units can be seen in the IAEA/WHO Postal TLD Intercomparison results which gave 100% 199n i 1992 n 1992i d withi 0% an % 5 ,199n ,44 ni wit± whil d 0% han eerror63 o t p su 37% were obtained for the participants not included in the program. The difference between the two groups lead the government to décrète through the Gaceta Oficial de la Republica de Venezuela, Resolution G-1397 on March 3, 1993, that quality assuranc d dosan ee calibration programs shal e establisheb ll radiatioal r fo d n therapy installations in Venezuela. The project for the standards was developed by the SSDL physicists alreads anwa t di y approbate Healte th y dhb Ministry expectes i t I . d thaNorme th t s will enter into effect by the end of 1994. 1. INTRODUCTION In 1988 prograa qualitmn o y assuranc dosd ean e calibratio startes n wa IVIC-SSD y db L 2 Cobalt-6o1 n 0 units from public hospitals, which represent f totao unitT R ln % i s 30 s performes wa SSDe Venezuelab th jo y Le db tasth l kprograe Al .grou th d sponsores pman wa d by the Health Ministry. This paper present method e resulte sth th d s san relate basio dt c physic dosimetrid san c checks performed through postal dose intercomparison durind san g on-site visits between 1988 and 1993, the SSDL projects for the next years as well as the main aspects covered by the Norms that will regulate radiation therapy installations. 1.1. Situatio 198y nb 8 1.1.1. No unit had preventive maintenance. 1.1.2. Som unitT eR s worked only wit certificate hth e provide source th y deb manufacturer. No beam calibration. 1.1.3. Most units were calibrated just once over several years. 1.1.4 programA . Q Nona d .eha 1.1.5. Most institution empiricad sha l techniciens. 1.1.6. Many radiation therapy physiciens have never had a course on radiation physics. 1.1.7. Of 36 institutions (25 Co-60 and 16 LA), only 3 had a hospital physicist (part-time). 1.1.8 rolhospitaa e f eTh o . l physicis t recognizedno s wa t . Often, salaries were lower thar nfo techniciens. 1.1.9. As there were no legislation in this field, some private and public institutions never participate Posta n di Intercomparisons D lTL . 135 2. METHODS 2.1. On-site visits On-site visits were performe radiatio2 1 e t eacth da nf ho therap y departments oncea year. At each visit, a total QA inspection of unit parameters was carried out along with beam calibration for all field sizes. Two staff members of SSDL working 2-3 days were needed for each visit. 2.1.1. Mechanical verification bead sman alignments followine Th g main parameters with their indicated tolerances [1-7] were verifiet da each Co-60 unit (TABL. EI) TABLEI: Main unit parameters with their tolerances Rotational system axis: isocenter and scales ± 2mm/0.5° Field size congruence ± 2mm Collimator symmetry and stability ± 2mm m 2m ± Couch parameters Radiation field m alignmen2m ± homogeneitd tan y Source transit time* ± 0.5% Virtual source distanc isocenter**______o et + 0.5% * influence for a dose of 2 Gy ** within consecutive calibrations During 1993 SSDe th , L also conducted work session radiatio2 1 e eact sth a f nho therapy department ensuro st dosd e an etha procese calibratioA th t Q IAEA/WHe f so th d nan O Postal TLD Intercomparison were fully understood by the physiciens, techniciens and dosimetrists. 2.2. Absorbed dose determination Determination of absorbed dose in water phantom. The measuring procedure and calculations were performed according to the recommendations containe Internationae th n di l Cod Practicf eo e publishe IAEy db A [8]. A cylindrical ionization chamber cm6 , 0. typ 3E , connecteeN electrometen a o dt r type NE Farmer, calibrated at SSDL in a Co-60 beam, against a secondary standard dosemeter, in terms of absorbed dose to water and air kerma were used for all Co-60 beams. The constancy of the response of the chamber and electrometer was checked daily during the visits, with a Sr-90 reference source. Absorbed dose determinatio IAEn a carrie s n Ai n wa t watedou r phanto reference th t na e depth of 5 cm, and at dmax, at the usual treatment distance, for all field sizes [4-6,8-13]. When available, absorbed dos wateeo alst s orwa measured wit brease hth t tray. 136 Determination of absorbed dose to air. e Absorbe th fielm c t a d0 1 d r measuremenx dos ai 0 o 1 et a r s carriefo wa t dou treatment isocentedistance th t a r eo r wit chambee hth witr build-ue ai hth n r i p cap. Other parameters verified were: - unit maintenance - general mechanical performance - head leakage - radiological safety and area survey accesor- y factors - personnel dosimetry requirement- sourcn so unir eo t replacement 2.3. Postal TLD Intercomparisons WHO/IAEA Postal TLD Intercomparisons [14], coordinated by the SSDL, started in Venezuela in 1983. They were regularly performed in 1988, 1990 1992 and 1993 (TABLE IE includes results up to 1992) with all the units under control and a small number of units not included in the program, even though TLDs were sent to all institutions nationwide. 3. RESULTS 3.1. Situatio 199y nb 3 unit2 1 3.1.1 se undeTh . r control plus othe werr2 e already working unde contracra t between the government and the SSDL for annual QA and dose calibration, and with private companie regular sfo r unit maintenance. 3.1.2. For Postal Intercomparison, any TLD with a discrepancy exceeding 5% was pursued by reirradiatioD TL n anddosimetry b , y review discrepanc e visith f ti unresolvables ywa . 3.1.3. Five institutions calibrated their units and participated in Postal TLD Intercomparisons firsfoe th rt time. 3.1.4. For the rest of the RT units, the situation remained without changes. 3.1.5. For medical physicists, the situation remained without changes. 3.1.6. Consequently to on-site visits, a better uniformity among the visited centers was reached. 3.1.7. The remarkable improvement in the general performance of the Cobalt-60 units can be IAEA/WHe seeth n ni O Posta IntercomparisoD TL l n result 199n si d 1990an 2 where 100% of the units under control were within ± 5 %, while 63% in 1990 and 44% in 1992, with errors up to 37%, were obtained for the participants not included in the program (TABLES n and ffl). 3.1.8. The difference between the two groups (Fig. 1) lead the government to décrète on March 1993 that quality assuranc dosd ean e calibration programs shal establishee b l l al r dfo radiation therapy installations in Venezuela. The project for the standards [5] was prepared by the SSDL physicists and it was already approbated by the Health Ministry. It is expected that the Norms will enter into effect by the end of 1994. 137 PERCENT OF Rf UNITS WITHIN ± 5% Legend UNITS UNDER CONTROL OTHERS 1988 1990 1992 YEAR Fig. 1. Percent of beam calibration within criterion for the group under control and the rest of the units for 1988, 1990 and 1992 Postal TLD Intercomparisons. TABLEE: Percent of units under control with acceptable deviation CHECKS 1988 1993 Unit preventive maintenance 0 100 General mechanical performance 42 80 Head leakage 100 100 Radiological safet ared yaan survey 50 83 Rotational system axis: isocenter and scales 33 73 Field size congruence 43 73 Collimator symmetry and stability 50 73 Couch parameters 50 50 Radiation field alignmen homogeneitd tan y 75 82 Source transit time 90 90 Virtual source distance to isocenter 100 100 Personnel dosimetrv 8 50 138 TABL: EHI Posta IntercomparisoD lTL n results* Year______1983 1984 1985 1988 1990 1992 Numbe Participantf 4 ro s 17 13 21 13 26 within ± 5% 75 47 41 67 70 54 within ±5% with QA/CAL — — 57 100 100 withi withou% n+5 t OA/CAL— — — 79 63 44 * In 1986, 19 institutions participated on the Intercomparison. TLDs were not returned on time for processing. 4. DISCUSSION 4.1. SSDL Project neae th rr yearssfo : 4.1.1. To establish a post-graduate program with a Master Degree on Medical Physics, starting Venezuelan i b firse ye th on 1995t supplgoao e e t b s Th i l. o t y,medicaa l physiciso t every institution unit witT R .ha 4.1.2. To organize an national society of medical physics. wor4.1.o T 3k very closely wit Healte hth h Ministr superviso yt implementatioe eth e th f no Norms for RT installations in Venezuela. 4.1.4. To spread nationwide the use of SSDL protocols for QA and beam calibration for Co- , linea60 r accelerators (photo electrod nan n beams ortvoltagd an ) e units orden i , o t r achiev ebettea r uniformity among radiation therapy centers. 4.2. Mai dosd n an epoint coveree NormA calibratioe b Q th o st unitsr T y dR b fo s r : nfo Cobalt-60, linear accelerators (photon electrod san mediuo nt beams)w lo m d energan , y x-rays units [5]. 4.2.1. Requirement personnen so dosinstrumentd d an lan e calibrationA Q r sfo . 4.2.2. Requirements on personnel involved in RT installations. 4.2.3 Detailed descriptio test theiA d Q san r f ntoleranceso . Frecuenc f constancyo reproducibilitd yan yunitsT l checkR typ al f :r eo Co-60fo s , linear accelerators (photon electrod san n beams ortovoltaged )an . 4.2.4. Within the next 5 years, every institution with RT units must have at least one qualified physicist, minimum half journey per unit. 4.2.5. Mandatory participatio Posta Intercomparisonsn no D lTL . 5. CONCLUSIONS. The Postal TLD Intercomparisons and QA surveys conducted in these units clearly demonstrated the need for upgrading the standards applied to radiation therapy installations. The most important consequence of these surveys is the establishment of the Norms that will regulate radiaton therapy installation Venezuelan si . 139 REFERENCES ] SOCffiDA[I D ESPANOL FISICE AD A MEDICA (SEFM). Procedimientos Recomendados para la Dosimetrîa de Fotones y Electrones de Energias comprendidas entre 1 MeV y 50 MeV en Radioterapia de Haces. Espana, 1984. [2] SOCIÉTÉ FRANÇAISE DES PHYSICIENS D'HOPITAL. Contrôle de Qualité d' une Installation de Télécobalthérapie. 1° Partie. Vérification des Performances. Cahier N° 28. Paris. 1984. ] SVENSSON[3 . QualitH , y Assuranc Radiation ei n Therapy: Physical Aspects. IntJ.Radiat. Oncol., Biol. Phys. 10. pag 59-65. 1984. ] SSDL-IVIC[4 . Protocole par l controae calidae d l calibratiody unidaa un Cobaltoe e dd nd - pages7 1 . . 60 Caracas, 1989. ] SSDL-IVIC[5 . Proyecto: Disposiciones Légales par l controae calidae d l calibratiody e nd unidades de teleterapia: Fuentes de Cobalto-60, aceleradores lineales (modalidades fotones y electrones) y rayos X de energias baja y média. 52 pages. Caracas, 1993 [6] HOSPITAL PHYSICISTS' ASSOCIATION. Revised Code of Practice for the Dosimetry X-rayCaesium-13f V o d M 5 an , 3 o t Cobalt-6o d f2 7 an 0 Gamma-ray Beams, Phys. Med. Biol. 28. pages 1097-1104. 1983 [7] ORGANIZACION PANAMERICANA DE LA SALUD. Control de calidad en Radioterapia. Aspectos Clrhico Fisicossy . Publication Cientific 499° aN , 1986. [8] INTERNATIONAL ATOMIC ENERGY AGENCY. TRS N° 277. Absorbed Dose Determination in Photon and Electron Beams. Vienna, 1987 ] INTERNATIONA[9 L ATOMIC ENERGY AGENCY. Manua f o Dosimetrl n i y Radiotherapy. TRS N° 110. Vienna, 1970 [10] INTERNATIONAL COMMISSIO N O RADIATION N D UNITAN S MEASUREMENTS, Radiation Dosimetry: Electron Beams with Energies Between 1 and 50 Mev, ICRU Rep , ICR.35 U Publications, Bethesda , 1984MD , . [II] INTERNATIONAL ELECTROTECHNICAL COMMISSION, Medical Electrical Equipment, Dosimeters with lonization Chambe s Usea r n Radiotherapyi d StandarC BE , d Publication 731, IEC, Geneva, 1986 [12] NORDIC ASSOCIATIO CLINICAF NO L PHYSICS, Procedure Externan si l Radiation Therapy Dosimetry with Electro Photod nan n Beams with Maximun Energies 0 Betwee5 d an n1 Mev, ActaRadiol., Oncol. 19 Fasc. 1, 1980 [13] NORDIC ASSOCIATIO CLINICAF NO L PHYSICS, Protoco Determinatioe th r fo l f no Absorbed Dose from High-Energy Photon and Electron Beams, Med. Phys 10 (6), Nov/Dec, 1983 [14] INTERNATIONAL ATOMIC ENERGY AGENCY. Intercomparison Proceduree th n si Dosimetr Photof yo n Radiation 182° N S . ViennaTR . , 1978 140 ffl(b). DOSIMETRY PROCEDUÄES "•y'y:-" ••••••w PAGEfS) iof i BLAHK Invited Paper XA9642854 RADIATION DOSIMETRY WITH PLANE-PARALLEL IONKATION CHAMBERS INTERNATIONAN :A L (IAEA) COD PRACTICF EO E P ANDREO Radiophysics Department, Lunds Hospital, Lund, Sweden P.R. ALMOND Department of Radiation Oncology, J.G. Brown Cancer Center, Universit f Lousvilleyo , Lousville, Kentucky, USA O. MATTSSON Department of Radiation Physics, Sahlgrenska Hospital, Gothenburg, Sweden A.E. NAHUM Joint Departmen f Physicso t , Royal Marsden Hospital, Sutton, United Kingdom . ROOM S Physikalisch-Technishe Bundesanstalt, Braunschweig, Germany Abstract Research on plane-parallel ionization chambers since the IAEA Code of Practice (TRS-277) was published 1987 in expanded has knowledgeour perturbationon otherand correction factorsin ionization chamber dosimetry, and also constructional details of these chambers have been shown to be important. Different national organizations have published, or are in the process of publishing, recommendations on detailed procedures for the calibration and use of plane-parallel ionization chambers. An international working group was formed under the auspices of the IAEA, first to assess the status and validity of IAEA TRS-277, and second to develop an international Code of Practice for the calibration and use of plane-parallel ionization chambers in high-energy electron photonand beams. purposeThe thisof workdescribeto is forthcomingthe Codeof Practice. 1. INTRODUCTION. The advantages of using plane-parallel ionization chambers in the dosimetry of therapeutic electron beams have been recognised in all dosimetry protocols. The design characteristics, mainly regarding the shape and size of the collecting volume, make this instrument theoretically ideal for measurements in regions with large dose gradients in the beam direction. A numbe chamberf ro available sar thef e o today mw fe havin a , g completel designsw yne , with practically negligible perturbation effects in electron beams. Large correction factors have 143 been found, however other fo , r chambers, electromainlw lo t ya n energies. Ther stils ei l controversy on the use of plane-parallel chambers for photon beam dosimetry. Most chambers are far from homogeneous in their construction as, in general, materials with different scattering and absorption properties are used in the various walls. It is likely that these effects approximately balance other effects in electron beams, but measurements and calculations in photon beams have shown the need for correction factor accouno st differene th r tfo t material chambere th n si . This suggests that plane- parallel chambers should mainly be used for absorbed dose determinations in electron beams but onl r relativfo y e measurement photonsn i s remainine Th . g proble e calibratioth ms e i th f no chamber. The lack of details on dosimetry procedures using plane-parallel chambers, particularly regarding their calibration or a practical determination of the ND (Ngas) chamber factor, has been majoe th f ro criticisme on s IAEe madth f Aeo Cod f Practiceeo , TRS-27 ] wher7[1 e onla y referenc e procedureth o t e s describe s consideres made NACy wa wa db t ] I . P[2 d that these procedures were well establishe thereford dan erecommendede b stil o t l influence Th e . th f o e central electrode correction for cylindrical chambers in TRS-277, however, added an unexpected complicatio experimentao nt l determination Nf sDo base comparisoa n do electron i n beams [3]. Researc fiele th dn h i sinc e IAEA TRS-27 publishes 7wa expandes dha knowledgr dou n eo perturbation and other correction factors in ion-chamber dosimetry, and also constructional details chambere th f o s have been show importante b o nt . Different national organisations have published [4, 5] or are in the process of publishing [6, 7] recommendations including detailed procedures for plane-parallef o e us e th l chambers internationan A . l working grou formes pwa d unde auspicee rth s of IAEA, first to assess the status and actual validity of IAEA TRS-277 [1] and second to develop an internationa plane-parallef l o Cod e Practicf us e o e th r efo l ionization chamber high-energn si y electron and photon beams. The purpose of this work is to describe the new Code of Practice. Further detailpresene th n so t situation regarding correction factor quantitied san s briefly discussed foune b her [3]n n di e ca . COD OVERVIEW PRACTICEN F A ENE O . 2 E TH WF O . contente Code Th th Practicf f eo so observee shows eb i n Tablca n i t dI . thaeI t together wit ratheha r conventional distributio differene th f no t sections 1-9, Sectio contain0 n1 summarsa y TABLE I. CONTENTS OF THE IAEA CODE OF PRACTICE FOR PLANE-PARALLEL IONIZATION CHAMBERS 1. Introduction . Updat 2 informatioe th f eo TRS-27n ni 7 Equipmen. 3 t 4. Beam quality specification 5. Afc-based formalism and determination of Afo.a/r for plane-parallel ionization chambers 6. A^£>iM-,ß0-based formalism and determination of ND,W,QO factors for plane-parallel ionization chambers 7. Use of plane-parallel chambers in electron beams plane-parallef o e Us . 8 l chamber photon si n beams 9. The uncertainty in absorbed dose determination at the reference depth using plane-parallel chamber electron si n beams 10. A Code of Practice for the calibration and use of plane-parallel ionization chambers Appendix A. Worksheets Appendi . Stopping-powexB r ratio clinican si l electron beams. Appendi . ChambexC r perturbation factor electron si photod nan n beams 144 of all the procedures and data required; this Section is effectively the Code of Practice. The report also contains Appendices where different topic coveree sar detailn di ; they also include Worksheets. The new Code updates information in IAEA TRS-277 regarding recent developments in radiotherapy dosimetry. In most cases differences from existing values or the magnitude of new corrections, are within half a percent but developments (and clarifications) in the field are taken into account speciaf .O l interes calibratioplane-parallef e o th e r us tfo d nan l ionization chambere sar effece th f metalli • o t c central electrode cylindrican si l ionization chambers (include TRSn di - 277 as a global factor) has been separated into two components, one at the Co-60 calibration l (kcei= 1.00Farmer-typa r 6fo e chamber thereford an ) e entering into Nrj ) (1 ND, Nair= K(\-g)kattkmkcel • a procedure based on an absorbed-dose-to-water calibration factor, ND>W, is also introduced. This symbol was given in TRS-277 but in practice its use was restricted to low-energy X-rays. becominw no Is i t g availabl high-energr efo y photons t presenA . mose th t t common approach is to provide users with ND,W at a reference quality go. usually 60Co, and apply beam quality correction factors for other beam qualities. Users should be warned of the possibility of confusion arising from the notation N D used by AAPM TG-21 [9] for the ND,W factor. scalinw ne a g • procedur conversior efo depthf no ranged san s measure plastin di equivaleno ct t quantitie waten si givens ri ; thi concepe bases si th n d o f detour o t factors alternativn a s a o et ratio csdf so a ranges [10]. correctioa • non-mediue th r nfo m equivalenc chambee th f eo r wall material, pwaii- This factor has implicitly been assume unite b electron o yi dt n dosimetry protocol dateo st . Thers ei however considerable experimental evidence that this factor may not be unity for certain plane- parallel chamber designs; the probable mechanism here is backscattering differences between the material behind the cavity and that of the wall material. However only values for an overall perturbation factor pQ=pCa\> Pwall givene ar ; pcav - correctioe in replaceth e s th a r u nsp fo scattering cavitiess effecga n i t . calculationw ne • stopping-powef so r ratios water/air, sw>air, using several independent Monte Carlo codes where different density effect corrections were taken into account. Compared with the stopping-power ratios in TRS-277, differences are small for the electron energies most commonly use radiotherapyn di , being clos 0.5o et % t mosa t depths recommendatioe Th . r nfo smale th l chang justifies ei lace termn di f ambiguitth k o f so correctione th n yi e s th use d dan higher accuracy of the present set of data. • the determination of the recombination correction factor for plane-parallel ionization chambers usin "two-voltagee gth " metho bees dha n show havo nt e limitation mosr sfo t chambero t e sdu lace f th linearitko f saturatioyo n curveregioe th f n interestsi no orden I . decreaso t r e eth influence in the dosimetry procedure it is recommended to use the same voltage ratio for the determination of No,air and for the absolute dose determination. 1 e Note that the factor ND in TRS-277 is now denoted by Nrj>air in order to distinguish it from ND.W- th factor in terms of absorbed dos watero et . 145 Sectio providen3 s detailed descriptio phantomn no equipmend an s t available, with emphasi propertiee th n o s f plane-paralleso l ionization chambers bot r electrohfo photod nan n radiation. Chamber desigw ne f nso (Attix, Roos, etcincludee ar )compilatione th n di TRSn i s -A . 277, water is the recommended reference medium although plastics may be used for measurements electrow alo t n energies. Emphasi givens si , however hige th h o t accurac, y achievable today with modern equipmen positioninn ti g ionization chamber waten si r phantoms which thus minimizee sth plastie neeus o dt c phantoms. The uncertainty in absorbed dose determination at the reference point using the recommended procedure for determining No,air is treated in detail, separating the different steps of the dosimetnc procedur similaa TRS-27o t n ei y rwa t incorporatin 7bu updaten ga d evaluatiof no uncertaintie differene th n si t steps. Uncertaintie alse sar o evaluate alternative th r dfo e calibration methods base measurementn do photon si n beams. Further detail certain so n sections follow. 2.1. Beam quality specification The specification of the quality of the beams used for the calibration of plane-parallel ionization chambers follows the recommendations given in TRS-277. As mentioned in the introduction, absolute dosimetry is to be performed in electron beams only, as is the recommended calibration procedure (see below). r dosimetrFo y purpose becoms ha t si e customar specifo yt qualite yth f electroyo n beams in terms of the mean energy at the surface of the phantom, E0, determined from empirical relationships between electron energy and the 50% range in water, RSQ- £b is needed for the selection of different quantities and parameters in the formalism, and mainly affects the choice of stopping-power ratios wate airo rt , sreference With ai t ra i e depth, namely sWiai^E IAE0n ,i zs ref)-AA TRS-277 [1] and most dosimetry protocols, the recommendation is to determine E0 using the energy-range relationship (2) where C=2.33 MeV cm'1 and R$o is obtained from a depth-dose distribution measured with constant source-chamber distance wels i s l .A known , whe dose nth e distributio bees nha n obtained with a constant source-surface distance (SSD=100 cm) Eq. (2) is not strictly valid. As an alternative IAEA TRS-277 has provided tabulated data for determining E0 either from ionization curves measured at SSD=100 cm with an ionization chamber or from depth-dose distributions at SSD=10 , measure0cm r instancdfo e with solid state detectors. Thes fittee eb datn d aca wit e hth following second order polynomial: Eo = 0.818 + 1.935 R50* + 0.040 (Kyf? (3) for RSO* determined from a depth-ionization curve and ÊO = 0.656 + 2.059 R5(P + 0.022 (R5(P? (4) for the case of a depth-dose curve, RSOD- For energies above 3 MeV, Eqs. (3) and (4) yield stopping-power ratios, water-to-air, that on the average agree within 0.2% up to depths equal to 0.8 witp 0R h sw>air values obtained with Eo derived from TRS-277 Tabl , witeIV hmaximua m deviation of 0.4% close to 12 MeV. Improved energy-range relationships between Eo and RSQ, based on Monte-Carlo calculations for mono-energetic electron beams, have been developed[3, 11] but all yield EO values higher tha above nth e expression. This would resul lowen ti r stopping-power ratioreference th t sa e depth compare thoso dt e obtained with sWtair(E0, Ed Zref)0 an fro . (2)mEq . 146 2.2. Determination of No^ir for plane-parallel chambers Several different methods have been propose Mattssoy db l [12a r obtainint ne ]fo e gth absorbed-dose-to-air chamber factor #£»,<,/ plane-parallea r >fo l chamber. These methods fall into two broad categories. In the first one, a Standards Laboratory calibrates the chamber in terms of N K and then A/b,ai obtaines ri d theoretically. secone Inth d one usee ,th r determines No,air directl experimentay yb l intercomparison with a reference ion-chamber having a known No,air factor. Both chambers are alternatively positioned at a reference depth in a phantom and the unknown A/b,a«> is obtained from equating the absorbed doses with the two chambers. These procedures have been extensively discussed in Ref. [13] and in the recent TG-39 protocol of the AAPM [14]. Methods in the second category are generally performe user'e th n dsi beam, either C high-energr oo y electron notee sb [12]n d ca t .I 60 that this method can in principle be applied to determining #£>,<„> for any chamber that is to be used in electron or photon beams e.g. a second cylindrical chamber provided that Af£>,fl,> is already know referenca r nfo e chambe , 15]r[4 . Consequentl chambee yth calibratee b o rt d (not necessarily plane-parallel reference th d )an e chamber wil re d denotee fb l an respectively x y db . The primary recommendation is the use of a high-energy electron beam. Following the formalism in TRS-277, and equating the absorbed dose at the reference depth with the two chambers, the expression for ND^ for the chamber x to be calibrated, becomes M _ fjrefPwallPcayPcelMX , M n n n ™D,air ~ " D,air ,,* x „x „x M PwallPcavPcel (5) wher numeratoethe denominatoand r r corresponw determinatioD the dto n usingthe reference chamber (usually cylindrical chambed )an respectivelyrx stopping-powee th d an , r ratios cancel out. Md Affärre*an echambero readinge ratiotw th e f th so thosexternao f n st sa o f eo l monito tako rt e into account possible accelerator output fluctuations. They mus correctee tb e th r dfo polarity effect, for recombination, and for temperature and pressure. Note that p™£u for the reference chambe units ri recommendes ya d reference cylindrical chamber assumee sar havo dt e negligible wall effect electron i s n beams [16, 17] mosr .Fo t plane-parallel ionization chamberd san energiee ath t s recommende calibratione th r dfo factore th , s p*p^aiid avan are practically unity. factoe Th relevant r p£no s ei plane-paralleir fo t l ionization procedurchambere th s a t alsn sebu oca be extended to cylindrical ion chambers it has been retained in this Eq. For the case of x being a cylindrical chambe value rth f p*eo av shoul interpolatee db d fro date mth a from Johanssol a t ne [16] give TRS-27n i 7 Tabl. eXL phantoe Th m material should preferabl same th tha s ea e yb absolute t useth r dfo e dose determination. This automatically ensures thaoverale th t l perturbatioy effectan f o o st e ndu differences in backscattering between the material behind the cavity and that of the phantom (i.e. the component of PQ due topwau) will be minimized. Water is the preferred material. The energy of the electron beam should be as high as possible in order to minimise the perturbation due to the air cavity of the reference chamber. As a guide p*£*v should be within 2% of unity. For a cylindrical reference chamber with an internal radius of 3 mm (approximately Farmer type) this means that E0 should be no lower than 15 MeV but should preferably be as high as possible; the lower limit on E0 ma loweree yb chambee th f di r radiu smallers si depte Th . h reference shoulsame th th s e a d b e depth zref use absorber dfo d dose determinatio chosee th n i n high-energ shoulyD beamSS de Th . fiele th dd sizan em shoulc r large o 0 criticalt approximatele 10 m d thic rb - no e 2 s b si 1 .x m c 2 y1 chambere placee b Th o t de witsar h their respective effective point f measurementso e th , t Pa eff, same depth. A Farmer- type chamber, i.e. approximately 6 mm internal diameter and 1 mm electrode diameter reference th r fo , e cylindrical chambe recommendes ri d hergreaa s ea t deaf o l experience has been gained with such chambers and the correction factors can be said to be well known [18, 19]. The choice of a chamber with a radically different geometry, e.g. a very thick central electrode, can lead to larger uncertainties. 147 Alternative method r obtaininsfo absorbed-dose-to-aie gth r chamber factor Afo.uia r Vfo plane-parallel chamber based on measurements made inC ao beam have been introduced. They classifiee ar d categorieinto otw s generally dependin institutioe th n go n wher calibratioe eth s ni 60 performed in-phantome Th . metho s generalldi y performe user'e th n sdi Hospitale beath t ma , although it can also be performed at the Standards Laboratory. Measurements free in air are usually performed in a 60Co beam at the Standards Laboratory. calibratioe Th 60a Cn ni o bea t deptm a phantoa n hi s beemha n describe severay db l authors. First by Mattsson et al [12] and then in more detail by Attix [20]. The approach is based upon the determination of A^o.air from the knowledge of the absorbed dose in the phantom determined wit calibrateha d reference chamber, like that recommende electron-beae th n di m method, but in this case irradiated with a Co beam. The formalism yields: 60 60 1 where p^ p^perturbatiod e ar uan n factorreference th f o s e chambet ra Co ; p s ^,i unitgraphita r yfo e electrod 0.99 d Farmer-typa ean r 4fo e chamber °"Ca standarr e ofo Th . D dSS surface th t a e m shoulc 0 1 usede db x fielthia n I m .uni d c s d 0 methoan tsiz 1 f eo effective dth e poin f measuremeno t botr fo t h chambers shoul placee db referenca t da e cnr g depta 5 n i f ho phantom that matche plane-parallee sth l chamber materia minimis2o (t l e watee n p^j)i th r f o i r actor 60 Pwall f is known. For cylindrical chambers in Co beams Peffis positioned at a distance equal to 0.6 r from the centre of the chamber. importanIs ti noto t e that whe non-watena r phanto museds i , TRS-277 doe providt sno ea at 60 direct determination of the perturbation pwall Co as absorbed dose should only be determined in a water phantom. The perturbation factor of the reference chamber is determined according to the general equation that takes into account the thin waterproofing plastic or rubber sleeve normally used to protect the chamber in a water phantom (see also Refs. [21, 22]): ref a Swaiiair ( Ver/P)med,wall+* Ssleeve>air___ (Hen/P)med,sleeve+(l- where med is the phantom material and a and T the fractions of ionization due to electrons arising fro walwaterproofine d mth an l g sleeve respectively available th o t t efi a dat [23.A r afo ]s i given. By applying this fit to the combined thickness of the wall and the sleeve, and substracting a from this, an expression for T is obtained. This insures that 148 outer dimension chambee th s spreferabla d same ran th f e o materia e predominanyb e th s a l t material of which the chamber is constructed. The procedure yields the NK calibration factor of the plane- parallel chamber (8) K and if the product katt km is known Nß>air is determined according to the well-known expression where for plane-parallel ionization chambers keel is not involved. In principle this procedure is used together with a universal value of katt km for a given type of plane-parallel ionization chamber e limitationTh . f thio s s approach increase considerabl e estimateth y d uncertainty. 2.3. Determinatio ND,f n o plane-paralle r Wfo l chambers formalise Th determinatioe th r mfo f absorbeno d dos wateo et photon i r electrod nan n beams using a No>w-based calibration factor has been given in detail by Hohlfeld [26]. The absorbed dos referencwatee o et th t ra e chambee pointh f o t r (wher calibratioe eth n factor applies) in a phantom irradiated by a beam of reference quality Q0 is given by the simple relationship where NDIVV>QO is obtained at the Standard Laboratory from the knowledge of the standard quantity absorbed dos watepoino e t th measuremenf t ro ta calibratio e waten ti th r rfo n quality QQ. Effortt presena e ar s t being addresse providino dt g ND,W,Q calibration photor sfo n beams, mainly 60Co gamma-ray lessea o t rd extensan t high-energy photo electrod nan n beams [27-31]A . practical approach in common use is to provide users with Afo.w.cb. i- e- calibration at the reference quality 60 Coappld an , y beam quality correction factors othekc>r fo r beam qualities [26, 32]r Fo . beams other tha reference nth e qualit absorbee yth d dos wateo et thes ri n givey nb WDQ Av,,M w,ß o= ö kQ (11) where the factor kQ corrects for the difference between the reference beam quality Q0 and the actual quality being shoul usedQ k . Q d, ideall determinee yb d experimentall same th t eya qualite th s ya user's beam, although thi s seldoi s m achievable. Whe experimentao n l dat available aar n ea derivee b expression ca dg comparin/k r nfo . (11gEq ) wit formalise hth TRS-277mn i ; this ensures consistency with the A/D,a/> procedure when kQ is calculated with the data in TRS-277 [26, 33, 34]. In therapeutic electro photod nan n beam generae sth l assumptio f (Wno air)Q=(Wair)Q0 yielde sth equatioQ k r nfo \sw,air kQ = which depends onl ration yo f stopping-poweso r ratio perturbatiod san n factors t shoulI . notee db d thachamber-dependene th t t correction factorinvolvet no definitio e kd e cesth iar an k n d i att m f ,k no onlkQe .Th y chamber specific factors involve perturbatioe th e dar n correction factors pQPd Qffan 149 The connection between the N^air and the ND>W based formalisms is established by the relationship, (13) In principle Eq. (13) could be used to determine Afo.air independent of the factors katt, km and kcei. 60 The use of Co as reference quality for determining ND,\V,QO for plane-parallel ionization chamber attractivn a s i s e possibility, especiall r mosyfo t SSDLs. Usin formalise gth t othema r qualities (both high-energy electro photod nan n beams) requires, however knowledge ,th PQf eo t Oa 60Co in Eq. (12) which enters in fcg; this is the main drawback of this procedure. This is also the alternative casth r efo e option which enables user determino st e N^w ^ directl experimentay yb l intercomparison in a ^Co beam with a reference ion-chamber (cylindrical in this case, where PQ is r more precisely known) havin gknowa n N ffw g, factor. It is assumed that the water absorbed dose rate is known at 5 cm depth in a water phantom for 60Co gamma rays. The plane-parallel chamber is placed with its reference point at a depth of 5 watea c mn i r tank wher absorbee eth d dos knowwates o i et w N^d robtaineD nO an wQ d from r DW is obtained using a reference chamber having a calibration factor N ffw Co . The calibration plane-paralle e factoth r rfo l chamber becomes 05) where it is assumed that the centre of the reference chamber is positioned at the depth of measurement alternativee Th . f Ps alsi o eff ovalie a us d option l experimentaAl . l conditione sar identical to those for the determination ofNo.air in ^Co using in-phantom measurements. plane-parallef o e 2.4Us . l chambers f plane-paralleo e us e Th l ionization chambers bot electron hi photon i d nan n beams si considered in line with the introduction above. Reference.1. 2.4 conditions In electron beams, reference conditions conside referenca r e dept TRS-277n i h s Zref(a ) instead of the depth of maximum absorbed dose used in other dosimetry protocols. The absorbed dose to water is determined according to swD ,airN Q u P (Zref)w M D = (16) where P PcavPwallQ= (17) Note that the perturbation factor pu in TRS-277 is replaced here by PQ which is the product factoro oftw s (Eq. (17)) firste .Th , electro e pcavth s i , n fluence perturbation factor, identicae th o lt Pu facto TRS-27n i r 7change TablTh . symbon ei eXI l attempt emphasiso st e exclusivelthas i t i t y concerned with effects due to the air cavity, rather than the wall material, that is, a correction for the 150 effect know s in-scatteringna where electron track scatteree sar mediue th y db m towardr ai e sth cavity stresses i t I . d that pcav strictls i y know reference th t na e depth only secone Th . d factor pwaii takes into account the lack of backscatter of the back wall material compared to water, and has implicitly been assume unite b electroo dn yt i n dosimetry protocol dateo st . This facto discusses ri d in detail in Appendix C. It was not possible to make definitive recommendations regarding pwau due to the present lack of consensus in the literature [35, 36]. However, all experimental determination f perturbatioo s plane-parallen ni l ionization chambers have effectively e beeth f no overall factor PQ. Figure 1 shows the values recommended for various plane-parallel chambers in Codw Practicef ne e o e th . Regarding water/air stopping-power ratios calculationw ne , s have been performed [39] includin seto f density-effecso tw e gth waten i t r ICRU-3e giveth n ni 7 electron stopping power tables. The density-effece yar t corrections accordin e Sternheimer'th o gt smore modeth ed an l accurate calculations of Ashley based on semi-empirical dielectric-response functions (DRF). It was argued [39] that for electron energies used in radiotherapy, where the density effect in air is negligible, ODRp-based water/air stopping-power ratios provid mora e e accurat f datao t .se e Differences in stopping-power ratios due to the different evaluations of the density effect correction are within 1%. The information on the density-effect correction in the set of values actually in use in TRS-27 othen i d r 7an dosimetr y protocol , howeversis dat,f o ambiguous providet s ai se w ne d A . here base Ashlen do y density-effect correction waterr sfo , Tabl . CompareeII d wit stoppinge hth - power ratios in TRS-277 differences are negligible for the most commonly used range of electron energies in radiotherapy, being close to 0.5% at most depths and high-energies (see figure 2). The small change is justified in terms of the lack of ambiguity in the corrections used and higher accuracy of the present set of data. It is interesting to note that if the comparison is made with Sternheimer-based electron stopping-powers, differences would be larger at shallow depths (up to mosr -1.fo t0 % energies slightld )an y smalle deptht ra s beyon r2 0d.0. 1.01 0s 1-00 u a 0.99 •PTW/Markus (VdP) IB 0.98 .a • PTW/Roos (Roos 93) A PTW/Markus (Roos 93) a> a • PTW/SchuIz (Roos 93) 0.97 o Capintec (TG-39) 0.96 10 15 20 mean electron energyE , (MeV) variationFIG.The 1. perturbation ofthe factorseveralfor PQ different plane-parallel chambersin common use, relative NACPthe to chamber, indicated dashedthe by line drawn p1.00.Qat = All the measurements were depthmadethe at of dose maximum normalizedand quotientthe to test chamber/NACP in a high-energy electron beam. The full line is a fit to 3 separate measurement series differentn o accelerators using PTW/Markuse th chamber filled[37]e Th , data pointse ar measurements threen o different designsW PT taken from [38] re-normalizedd an , thato r s fo 1 p Q= the NACP chamber; the unfilled symbols are for the Capintec-PS-033 chamber as given in [14]. 151 As already mentioned specificatioe th , "qualitye th electro e f no th f "o n bea mtermn i f so meae th n electron energe phantoth t ya m surfac e "2.3 s basei eth 3n do approximation"d an , stopping-power ratios selected with sWiair(E0,z) using data from monoenergetic beams. The validity of these two approximations and their limitations is discussed in detail in Appendix B. In particular the influenc f electroo e d photoan n n contaminatio s demonstratedi n , showing maximum discrepancies up to 1% at zref between the sWiai^E0,z) method and full Monte Carlo simulations. Differences are usually larger at shallow depths due to the difference in slope of the sWiair (z) distributions obtained with the two methods and increase further if analytic expressions yielding E0 values larger than the "2.33 approximation" are used [3]. emphasizes Ii t d tha accurato n t e method exists toda predico yt se w 1.0 stopping-power ratios water/air for electron beams 0> % difference using 5w r(Ashley) £ 0.5 (0 10Me\7 -">*-_ _\-- -- 0.0 u 2 S -0.5 -1.0 0.0 0.2 0.4 0.6 0.8 1.0 dept waten hi o r r/ FIG. 2. Percent difference between the new stopping-power ratios for electron beams, sWiair, given in Table III and those tabulated by TRS-277 [1] and other dosimetry protocols. 2.4.2. Non-reference conditions f plane-paralleEmphasio e us e s givei sth o nt l ionization chamber non-referencn i s e conditions, especially to determine relative dose distributions. For electron beams the need to take into account the depth variation of different quantities and correction factorchamben io r sfo r measurement stresseds si . Thi significana s si t disadvantage compared with other detectors like TLD, diodes, plastic scintillators, synthetic diamonds, Fricke dosimeters or liquid ion chambers. A common mistak applicatioe th n ei TRS-27f no fielr 7fo d sizes smaller tha reference nth e field is to determine R$o for such fields and use equation (2) or alternative tables to determine E0, and the se w>anus ir(E0 ,z)seleco t t stopping-power ratiosTRS-27n i s A . t shoul7i emphasizee db d here thavalidite th t equationf yo s (2-4 alternativr o ) e table restrictes si largo dt e field sizes. Users 152 TABL SPENCER-ATTIk EI X STOPPING-POWER RATIOS (A=10 KeV). WATE(s R WillirELECTROR AI )FO O RT N BEAM FUNCTIOA S SA DEPTD Ê,F NAN WATER,O HN I . Density effect correction (8Ashley I-valued )an s from ICRU-3 electrod 7an n fluence Monte Carlo calculations from Andreo [39] usin EGSe gth 4 Monte Carlo system. deptn hi Electron beam energy E,> water(mm) V Me 7 V V Me Me 2 6 V V 1Me Me 5 V Me 4 V 3Me 8 MeV 9 MeV 10 MeV 12 MeV 14 MeV 16 MeV 18 MeV 20 MeV 22 MeV 25 MeV 30 MeV 40 MeV VMe 0 5 Rp(mm)' 3.6 8.8 14.0 19.1 24.3 29.4 34.39.56 44.7 49.8 59.969.9 79.9 89.8 99.6 109.3 123.8 147.7 194.1 238.8 0 1.117 1.088 .066 1.049 1.034 1.026 1.011.004 6 0.998 0.993 0.981 0.969 0.961 0.955 0.948 0.943 0.936 0.924 0.912 0.907 1 1.125 1.096 .072 1.055 1.040 1.030 1.018 1.010 1.002 0.996 0.985 0.973 0.965 0.959 0.951 0.946 0.938 0.927 0.914 0.908 2 1.131 .104.079 1.060 1.045 1.033 1.022 1.014 1.005 0.999 0.988 0.976 0.968 0.962 0.954 0.948 0.941 0.929 0.915 0.909 3 1.134 .111 .085 1.065 1.049 1.037 1.026 1.018 1.009 1.002 0.990 0.979 0.971 0.964 0.957 0.951 0.943 0.932 0.917 0.911 4 1.136 .117 .091 1.070 1.053 1.041 1.029 1.021 1.011 1.005 0.993 0.982 0.973 0.966 0.959 0.953 0.945 0.934 0.918 0.912 5 .12.0937 1.075 1.057 1.044 1.032 1.023 1.014 1.007 0.995 0.984 0.975 0.968 0.961 0.955 0.946 0.935 0.920 0.913 6 .12.1072 1.079 1.061 1.048 1.035 1.026 1.016 1.009 0.997 0.986 0.977 0.970 0.963 0.957 0.948 0.937 0.921 0.914 8 .13.1122 1.089 1.069 1.055 1.041 1.031 1.021 1.013 .0010.989 0.980 0.973 0.966 0.960 0.951 0.940 0.924 0.916 10 .13.1250 1.098 1.077 1.062 1.047 1.036 1.025 1.018 .0045.992 0.983 0.975 0.969 0.962 0.953 0.943 0.926 0.918 12 .127 1.107 1.086 1.070 1.054 1.042 1.030 1.022 .0083.995 0.985 0.978 0.971 0.964 0.956 0.945 0.928 0.920 14 .132 1.116 1.095 1.079 1.061.041 8 1.035 1.027 .0113.998 0.988 0.981 0.973 0.966 0.958 0.947 0.930 0.922 16 .135 1.123 1.104 1.087 1.069 1.054 1.041 1.031 .015 .001 0.991 0.983 0.975 0.969 0.960 0.948 0.932 0.923 18 .137 1.129 1.112 1.095 1.076 1.061 1.047 1.037 .018 .004 0.994 0.986 0.977 0.971 0.962 0.950 0.933 0.924 20 1.133 1.118 1.103 1.084 1.068 1.053 1.042 .023 .008 0.997 0.988 0.980 0.973 0.964 0.952 0.935 0.925 25 1.128 1.120 1.102 1.086 1.069 1.056 .034 .016 1.004 0.994 0.986 0.978 0.969 0.956 0.938 0.928 30 1.133 1.131 1.118 1.103 1.086 1.072 .047 .027 1.012 1.002 0.992 0.984 0.974 0.960 0.941 0.931 35 1.132 1.129 1.118 1.102 1.087 .060 .038 1.021 1.008 0.998 0.989 0.978 0.964 0.944 0.933 40 1.128 1.116 1.103 .074 .050 1.031 1.016 1.005 0.996 0.984 0.969 0.948 0.935 45 1.130 1.127 1.115 .088 .062 1.041 1.026 .012 1.002 0.990 0.973 0.951 0.938 50 1.125 .102.075 1.053 1.035 .021 .009 0.995 0.978 0.955 0.940 55 1.127 .114.088 1.065 1.045 .029 .016 1.001 0.983 0.959 0.943 60 1.124 .123 .100 1.077 1.056 .038 .024 1.007 0.987 0.962 0.946 70 .122 .120 1.099 1.078 .058 .041 1.021 0.998 0.969 0.952 80 .118 1.118 1.099 .078 .060 1.037 1.009 0.977 0.957 90 1.114 1.116 .099 .079 1.053 1.022 0.984 0.963 100 1.114 .098 1.071 1.036 0.993 0.970 120 .109 1.104 1.065 1.012 0.984 140 1.095 1.034 0.999 160 1.099 1.058 1.015 ISO 1.081 1.033 200 1.091 1.053 220 1.071 240 1.084 -1.6= - *5.2Rp 0.0085 + p 3E 4 Ep^, average from Monte Carlo calculation monoenergetir sfo c electrons usin EGSe gth ITSd 4an 3 systems shoul aware db e that stopping-power ratio almose sar t independen f fielo t d size Figure d se ,an , e3 using the incorrect approach just described to determine Eo will result in stopping-power ratios that correspond to a beam with a different energy. In photon beams, plane-parallel ionization chamber t recommendeno e sar r absolutdfo e determinations, but for relative measurements on the central axis only and for output factors. Perturbation factors in photon beams are very sensitive to the details of the construction of a chambe thed yan r canno predictee b t d wit acceptabln ha e uncertainty. Furthermore, small changes from chambe chambeo rt manufacturine th n ri g process rende "generalf o r invalie us e "dth factors for chamber same th ef o smake . Plane-parallel ionization chambers shoul avoidee db vern di y narrow beams suc thoss ha e use stereotactidn i c procedures. i——i——i——i field size dependence of s^.., for electron beams 0 15 50 0 10 200 250 depth in water (mm) Field-size3. FIG. dependence of water/air stopping-power ratios electronfor beams determined with Monte-Carlo calculations. broad Radiiand beam; mm shown figure 10 the MeV, 5 in are:for for 10 MeV, 10 mm, 20 mm and broad beam; for 25 MeV, 10 nan, 30 mm, 50 mm and broad beam; broadand beam.mm solid60 The curvesmm, 40 pertainmm, broad MeV,10 the 50 to beams.for 3. TESTING OF THE NEW CODE Tests at two different levels have been proposed to the IAEA by the working group: • Category A - for checking that the Code is clearly written so that the procedure can be unambiguously carrie t frod ou mpractica a l poin viewf to comparisoA . n with absorbed dose determinations using TRS-277 wil includee b l thin di s category groue Th . p includee sth (obvious) authore oc-testh y b t s followe ß-testy db s performe independeny db t persons. This category must be carried out before the new Code is published and it should not take more than two months t shoulI . undertakee db hospitay nb l physicist severan si l centres, som whicf eo h should not be in an English-speaking country. Categor• testinr fo - gyB thacorrece th t t absorbed dos wateo et obtaines i r followiny db e gth new Code of Practice. This category is a longer term project and represents a significant research project to be undertaken in a sophisticated centre or centres. 154 4. CONCLUSIONS forthcomine Th g IAEA Cod f Practiceo plane-paraller efo l ionization chambers should improv accurace eth electrof yo n beam dosimetry andlessea o t , r extent photof o , n beam dosimetry too. Whereas effort beine ar s g mad implemeno et latese th t t development ionization si n chamber dosimetry, the verification of the Code will show if they are to be preferred to previous methods or to procedures recommended in other recent protocols in the same field. It is hoped that changes in structure, compared with TRS-277, wilCodee th l f facilitato . e us e eth REFERENCES [I] IAEA INTERNATIONAL ATOMIC ENERGY AGENCY, "Absorbed Dose Determination in Photon and Electron Beams: An International Code of Practice", Technical Report Serie 277. sno , IAEA, Vienna (1987). [2] NACP NORDIC ASSOCIATION OF CLINICAL PHYSICS, Supplement to the recommendations of the Nordic Association of Clinical Physics: Electron beams with mean energies at the phantom surface below 15 MeV, Acta Radiol Oncol 20 (1981) 401. [3] ANDREO statue Th high-energf , so P. , y photo electrod nan n beam dosimetry five years aftee publicatioth r e IAEth f Ano Cod f Practico e e Nordith n ei c countries, Acta Oncologica 32(1993)483. [4] SEFM SOCffiDAD ESPANOL FfSICE AD A MÉDICA, Procedimientos recomendados para la dosimetria de fotones y electrones de energîas comprendidas entre 1 MeV y 50 MeV en radioterapia de haces externes, Rep. SEFM 84-1, SEFM, Madrid (1984). [5] NCS NEDERLANDSE COMMISSES VOOR STRALINGSDOSIMETRIE, Code of Practic dosimetre th r efo high-energf yo y electron beams, Rep. NCS-5, NCS, Amsterdam (1990). [6] THWAITES, D.I., Priv communication from the UK working party on electron beam dosimetry, (1995). [7] ALMOND, P., "Calibration of parallel plate ionization chambers. Status of the AAPM protocol (IAEA-SM-330/60).", Measurement Assuranc Dosimetrn i e y (Proc Symp. Vienna, 1993), Vienna: IAEA (1994) [8] MA, C.M., NAHUM, A.E., Effect of the size and composition of the central electrode on the respons cylindricaf eo l ionisation chamber high-energn si y photo electrod nan n beams, Phys. Med. Biol (19938 3 . ) 267. [9] AAPM AMERICAN ASSOCIATION OF PHYSICISTS IN MEDICINE, Task Group 21 : A protoco determinatioe th r fo l f absorbeno d dose from high-energy photo electrod nan n beams, Med. Phys (19830 1 . ) 741. [10] GROSSWENDT , ROOSB. , , ElectroM. , n beam absorptio n watei n solii nd ran d phantoms: depth scalin energy-rangd gan e relations, PhyBiod (19894 3 lsMe ) 509. [II] ROGERS, D.W.O., BIELAJEW, A.F., Differences in electron depth-dose curves calculated with EGS and ETRAN and improved energy-range relationships, Med. Phys. 13 (1986)687. [12] MATTSSON, L.O., JOHANSSON, K.A., SVENSSON, H., Calibration and use of plane- parallel ionization chambers for the determination of absorbed dose in electron beams, Acta Radiol Oncol 20 (1981) 385. ] 13 NAHUM[ , A.E., THWAITES plane-paralle, f 'Tho ,D. e eus l chamber dosimetre th r sfo f yo electron beam radiotherapy"n i s , Revie f datwmethodd o an a s recommendee th n di International Cod Practicef eo , IAEA Technical Report Serie 227. Absorbesn No o , d Dose determination in photon and electron beams: a Consultants meeting (Proc Symp. Vienna, 1992), IAEA, Vienna (1993. )47 [14] AAPM AMERICAN ASSOCIATION OF PHYSICISTS IN MEDICINE, Task Group 39: The calibration and use of plane-parallel ionization chambers for dosimetry of electron beams: An extension of the 1983 protocol, Med. Phys. 21 (1994) 1251. 155 [15] MEDIN, J., ANDREO, P., GRUSELL, E., MATTSSON, O., MONTELIUS, A., ROOS, M, Ionisation chamber dosimetry of proton beams using cylindrical and plane-parallel chambers. Nw versus Nk ion chamber calibrations, Phys Med Biol 40 (1995) in press. [16] JOHANSSON, K.A., MATTSSON, L.O., LINDBORG, L., SVENSSON, H., "Absorbed- dose determination with ionization chambers in electron and photon beams having energies (IAEA-SM-222/35)"V Me betwee0 5 d an n1 , Nationa Internationad an l l Standardizatiof no Radiation Dosimetry (Proc Symp. Atlanta, 1977), Vol. 2, IAEA, Vienna (1978) 243. [17] NAHUM, A.E., "Extensio Spencer-Attie th f no x cavity theor3-medie th o yt a situatior nfo electron beams (IAEA-SM-298/81)", Dosimetry in Radiotherapy (Proc Symp. Vienna, 1987), Vol. 1, Vienna: IAEA (1988) 87. [18] HOHLFELD , "Testin K. ,IAEe th f Ago Cod e: Absorbe d dose determinatio t Co-6na 0 gamma radiation", Review of data and methods recommended in the International Code of Practice, IAEA Technical Report Series No. 227, on Absorbed Dose determination in photon and electron beams: a Consultants meeting (Proc Symp. Vienna, 1992), Vienna: IAEA (1993) 67. [19] BOUTILLON, M., PERROCHE, A.M., "Comparisons and calibrations at the Bureau International des Poids et Mesures in the field of X and gamma rays (IAEA-SM-330/22)", Measurement Assuranc Dosimetrn ei y (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) 15. [20] ATTDC, F.H., A proposal for the calibration of plane-parallel ion chambers by accredited dosimetry calibration laboratories, Med. Phys (19907 1 . ) 931. [21] GILLIN, M.T., KLINE, R.W., NIROOMAND-RAD, A., GRIMM, D.F., The effect of thicknes waterproofine th f so g calibratiosheate th n ho f photono electrod nan n beams, Med. Phys. 12 (1985) 234. [22] HANSON, W.F., DOMINGUEZ-TINOCO, J.A., Effects of plastic protective caps on the calibration of therapy beams in water, Med. Phys. 12 (1985) 243. [23] LEMPERT, G.D., NATH, R., SCHULZ, R.J., Fraction of ionization from electrons arising in the wall of an ionization chamber, Med. Phys. 10 (1983) 1. [24] ROGERS, D.W.O., Calibration of parallel-plate chambers: resolution of several problems by using Monte Carlo calculations, Med. Phys (19929 1 . ) 889. [25] KOSUNEN , JÄRVINENA. , , SIPILÄH. , , "OptimuP. , m calibratio f NACno P type plane parallel ionization chambers for absorbed dose determination in low energy electron beams (IAEA-SM-330/41)", Measurement Assuranc Dosimetrn ei y (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) 505. [26] HOHLFELD, K., "The standard DIN 6800: Procedures for absorbed dose determination in radiology by the ionization method (IAEA-SM-298/31)", Dosimetry in Radiotherapy (Proc Symp. Vienna, 1987), Vol. 1, IAEA, Vienna (1988) 13. [27] BOUTILLON, M., COURSEY, B.M., HOHLFELD, K., OWEN, B., ROGERS, D.W.O., "Comparison of primary water absorbed dose standards (IAEA-SM-330/48)", Measurement Assurance in Dosimetry (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) 95. [28] ROOS , HOHLFELDM. , , "Statu primare K. , th f so y standar watef do r absorbed dosr efo high energy photo electrod nan n(IAEA-SM-330/45)"B radiatioPT e th t na , Measurement Assurance in Dosimetry (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) 25. [29] BURNS, D.T., MCEWEN, M.R., WILLIAMS absorbeL , NP A.J. n d,"A dose calibration servic electror efo n beam radiotherapy (IAEA-SM-330/34)", Measurement Assurancn ei Dosimetry (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) 61. [30] ROSSER, K.E., OWEN, B., DUSAUTOY, A.R., PRITCHARD, D.H., STOKER, I., BREND, C.J., "The NPL absorbed dose to water calibration service for high energy photons (IAEA-SM-330/35)", Measurement Assurance in Dosimetry (Proc Symp. Vienna, 1993), IAEA, Vienna (1994. )73 156 [31] ROGERS, D.W.O., ROSS, C.K., SHORTT, K.R., KLASSEN, N.V., BIELAJEW, A.F., "Toward dosimetra s y system base absorben do d dose standards (IAEA-SM-330/9)", Measurement Assurance in Dosimetry (Proc Symp. Vienna, 1993), IAEA, Vienna (1994) [32] DIN, "Dosismessverfahren nach der Sondenmethode für Photonen- und Elektronenstrahlung lonisationsdosimetrie", Deutsche Nor 680N mDI 0 Tei (Draft)2 l , Berlin (1991). [33] ANDREO, P., Absorbed dose beam quality factors for the dosimetry of high-energy photon beams, Phys Med Biol 37 (1992) 2189. [34] ROGERS, D.W.O., The advantages of absorbed-dose calibration factors, Med. Phys. 19 (1992) 1227. [35] HUNT, M.A., KUTCHER, G.J., BUFFA Electro, A. , n backscatter correction parallelr sfo - plate chambers, Med. Phys (19885 1 . . )96 [36] KLEVENHAGEN, S.C., Implication of electron backscattering for electron dosimetry, Phys. Med. Biol (19916 .3 ) 1013. [37] VAN DER PLAETSEN, A., SEUNTJENS, J., THIERENS, H., VYNCKIER, S., Verification of absorbed doses determined with thimble and parallel-plate ionization chamber clinican i s l electron beams using ferrous sulphate dosimetry, Med. Phys1 2 . (1994. )37 [38] ROOS, M., The state of the art in plane-parallel chamber hardware with emphasis on the new "Roos" and "Attix" chambers. Work commissioned by the IAEA plane-parallel working group, December 1993, Rep. Unpublished, Available fro authoe mth t PTBa r , 38116 Braunschweig, Germany (1993). [39] ANDREO , Depth-dosP. , stopping-powed ean r dat monoenergetir afo c electron beams, Nucl. Instr. Meth (19901 5 .B ) 107. 157 ALGORITHN A INCLUDO MT BREMSSTRAHLUNE ETH G COMPONENT IN THE DETERMINATION OF THE ABSORBED DOSE IN ELECTRON BEAMS S.C. KLEVENHAGEN "~""~XA96428Î55" Medical Physics Department, The Royal London Hospital, London, United Kingdom Abstract Currently used dosimetry protocol absolutr sfo e dose determinatio electrof no n beams from accelerators in radiation therapy do not account for the effect of the bremsstrahlung contamination of the beam. This results in slightly erroneous doses calculated from ionization chamber measurements. In this report the deviation is calculated and an improved algorithm, which accounts for the effect of the bremsstrahlung component of the beam, is suggested. 1. Introduction existinNone th f eo g protocol coder so practicf so high-energr efo y electron dosimetry (NACP 1980/81, AAPM 1983/94, HPA 1985, IAEA 1987, CFMRJ 1987, NCS 1989, IPSM 1992, etc.) takes any amount of the bremsstrahlung component always present in electron beams. This results in a systematic error in the derivation of the absorbed dose. The purpose of this study is to draw attention to this striking omission in dosimetric procedures algorithn A . mproposes i dealinr dfo g with this deficienc electron yi n dosimetry. . 2 Absorbed dose equation Routinely recommendee th , d calibration dept depte electronr peae th hfo th f h dos- ko t a es si curve (Klevenhagen 1994) thin I . s position ionizatioe ,th n chambe exposes ri boto dt h electrono t d san bremsstrahlung (fig 1) but only the dose due to electrons is accounted for as seen in equation 1. For the purpose of the beam calibration, an ion chamber is placed in a phantom so that its effective poin measuremenf o t reference th t a s i teabsorbe e deptth d han d phanto e dosth o et m medium obtaines i d fro familiae mth r expression Dw=M NDSw^r«Pt (1) where Dw is the absorbed dose to water, M is the mean value of the dosemeter readings corrected for recombination and polarity effects as well as for any differences between ambient conditions at the time f measuremeno e standarth d an dt condition r whic fo scalibratioe hth n factos rit appliesr (o N , D equivalent absorbee Nth gas s i ) d r (gas dosai o )et chamber factor ,stoppin e sth w>li s i r g power ratio water to air, obtained for the appropriate electron beam mean energy and rtp, is the product of the various correction factors applicable to measurement in phantom at user's accelerator. The calculatio e absorbeth f o n d dose involves applicatio f severao n l correction factors represented in equ. (1) by the product, Tip, but these will be neglected further on in this analysis as being outsid interese eth thin i t s case. 159 CHAMBEN IO R V Me 8 1 Electrons & X-Rays co N *E o 20 10 5- T T 2 5 10 15 20 Dept Waten hi r (cm) response Fi Th electron g1 a f eo n chamber place reference th t da e deptelectron a n hi n beams i due to both radiation components; the main electron beam as well as to the bremsstrahlung contamination (adapted from Rustg Rogerd ian s 1987) 2.2. Modified absorbed dose algorithm To allow for the bremsstrahlung component, equ. (1) may be rewritten as follows (2) where (VwrXi is the stopping power ratio for the electron fraction of the beam, (v«ir)br is the stopping powe rbremsstrahlune ratith r ofo bremsstrahlune th g s i fractio beame ß th d f no an , g fractioe th f no total radiation beam at the absorbed dose determination depth. The discrepancy, S, in the absorbed dose determination with accounting for and without accounting for the bremsstrahlung component can be defined as (3) 160 The first factor in the square brackets of equation 3 represents electrons alone, the second factor in the square brackets stands for the bremsstrahlung alone and the last component of the equation represents the total beam but for which only the stopping power ratio for electrons is employed. . 3 Estimatio absorbef no dbremsstrahluno t dos e edu g The magnitude of the error due to the bremsstrahlung omission can be evaluated using equ (3). The methods of obtaining the required parameters for the calculation of 5 are described below. Stopping power ratios sw, air for the bremsstrahlung. Determination of the absorbed dose fraction which is due to the bremsstrahlung component requires the knowledge of the stopping power ratios water-to-air for the photon energy concerned. An approach similar to that used with the conventional high-energy photon beams was adopted, namely using the beam quality index concept and by measuring the tissue-phantom-ratios TPRio20 which were performed in the tail of the appropriate electron beam. Thes valueR eTP s were then use findinr dfo appropriate gth e r watestoppinai o t r g power ratios (fig 2) from the data provided by Andreo and Brahme (1986). Stopping power ratios sw,»r for the electron beam. The stopping power ratios employed in the AAPM (1983) dosimetry protocol and featured in many other protocols are considered the best and were used for these calculations. The water/air stopping power ratios derived for the depth of electron dose maximum (reference depth) are given in fig 2 as a function of the mean electron beam energy at surface. 1.140 BREMSSTRAHLUNG va Ep,o , 9 1.120- r $a: 1.100 5ÜJ O ü. 1.080 O z o. 1.060 10 £ 1.040 ÏLECTRON a v SE . < I O V 1.020 Lu 1.000 0.980 10 15 20 25 ELECTRON ENERGYV Me , Fig 2 Stopping power variation with electron energy. For the electron beam the energy scale represent meae sth n electron energ t surfacey bremsstrahluna e th r fo , g componen t representti mose sth t probable energ t surfaceya . 161 ON K) TABL . UNDERESTIMATIOE1 ABSORBEE TH F NO OMISSODO T DOS E EF DU NO BREMSSTRAHLUNG BREMSSTRAHLUNG DATA FROM BREMSSTRAHLUNG DATA FROM GUR et al 1979 RUSTGI AND ROGERS 1987 NOMINAL UNDER- NOMINAL UNDER- ELECTRON ESTIMATION OF ELECTRON ESTIMATION ENERGYV ,ME F X-RAYO %m D T SA ABSORBED DOSE ENERGY, MEV % OF X-RAYS AT D. OF ABSORBED DOSE 4 1.7 0.05% 6 4.8 0.2% 6 2.2 0.1% 8 5.4 0.3% 8 3.5 0.2% 10 5.9 0.4% 10 3.8 0.3% 12 6.8 0.5% 12 4.2 0.3% 14 7.0 0.7% 15 4.8 0.5% 17 8.2 1.0% 18 5.7 0.7% 20 8.5 1.2% 22 7.5 1.0% Bremsstrahlung fraction at the peak of electron curve. For the purpose of this analysis one needs to know the bremsstrahlung fraction of the total radiation beam at the depth of the absorbed dose determination i.e. at the dose (ionization) maximum. thir Fo s analysis onle th ,y useful data bremsstrahlun e founth n do g conten maximue th t a t m electron depth dose curv e thosear e obtained experimentall Rustgy l 197b a d t d e 9an ian r Gu y yb Rogers (1987) botn I . h studies magnetic fields were use defleco dt beae e electronth mth tf o t ou s allowing the dose from X-rays alone to be measured. The data on X-ray contamination obtained from this work, expressed in terms of the percentage of the maximum electron dose, are given in table 1. 4. Results and discussion The discrepancy, 6, in the absorbed dose between the determination without considering the bremsstrahlun determinatiod gan n with inclusio bremsstrahlune th f no calculates gwa d using equation (3). The results for the two accelerators expressed as a percentage of the total electron beam (electrons plus bremsstrahlung givee )ar tabln i . e1 It is seen that the underestimation of the absorbed dose is small at low electron energies increasing however with energy clinicalle th n I . y most relevant energy range, MeV 5 5 betwee1 , d an n8 varies betwee 0.5o t 3 % nbotf 0. i h accelerator type takee sar n into consideration. Abov MeV5 1 e e ,th discrepancy reaches 1.2%. This is consistent with the data in fig 2 where the stopping powers for the bremsstrahlun electrond gan seee sar diverg no t e with increasing energy. 5. Conclusions This analysi shows sha n that omissio bremsstrahlune th f no g componen calibratioe th n i t f no electron beams lead systematia o st c underestimatio absorbee th f no d dose. This erro comparabls ri n ei magnitude with the existing correction factors and uncertainties pertaining to electron dosimetry and should therefore be taken account of in the dosimetric procedures. REFERENCES AAPM 1983. A protocol for the determination of absorbed dose from high-energy photon and electron beam. Med. Phs. 10, 741-771 AAPM 1994. The calibration and use of plane-parallel ionization chambers for dosimetry of electron beam: An extension of the 1983 AAPM protocol. Report of AAPM Radiation Therapy Committee Task Group No. 39. Phys. in Med. in print. Andreo P and Brahme A 1986 Stopping power data for high energy photon beams. Phys. Med. Biol. 8, 839-858 Berge 198J rM 3 Personal communicatio AAPMe th o nt . CFMR 2 1987 . No I, Recommendations Pou a Mesurl r a Dosl e ed e Absorbé Radiothérapien e . Monographie Bureau sd u Nationa Métrologie d l e Gur D, Bukowitz A G and Serago C 1979. Photon contamination in 8-20 MeV electron beams from a linear accelerator. Med. Phys. 6, 145-148. HPA (Hospital Physicists' Association). 1985. Code of practice for electron beam dosimetry in radiotherapy. Physics in Medicine and Biology, 30, 1169-1194. 163 IAEA (International Atomic Energy Agency), 1987. Absorbed dose determination in photon and electron beams Internationan :A l cod practicef eo . Repor (IAEA7 27 t , Vienna). IPSM 1992 Addendu code th f practic eo mo t r electroefo n beam dosimetr radiotherapn yi y (1985) Interim recommendation IPSe th Mf so electron dosimetry working party. Phys. Med. Biol. , 12137 , - 130. Klevenhagen S.C., 1993. Dosimetr d Physican y f Therapo s y Electron beams (Medical Physics Publications, Wisconsin USA). NACP 1980 Procedures in External Radiation Therapy Dosimetry with Electron and Photon Beams with Maximum Energies MeVBetwee0 5 d , an Act n1 a Radiol. Oncol , 55-619 . 5 NACP 1981 Electron Beams with Mean Energie Phantoe th t sa m Surface Belo MeV5 w1 , Acta Radiol. Oncol , 401-4320 . 0 (NetherlandS NC s Commissio Radiation no n Dosimetry) 1989. Cod f Practic eo do,simetre th r efo f yo high-energy electron beams. Report5 Rogerd Rustg an , 1987E S sbremsstrahlunJ e N i .Th g componen electro6-1n i V t 8Me n beam. Med. Phys , 884-8914 . 1 164 CLINICAL DOSIMETRY WITH PLASHC SONTILLATORS - ALMOST ENERGY INDEPENDENT, DIRECT ABSORBED DOSE READING WITH HIGH RESOLUTION U. QUAST . FLÜHD , S XÂ9642856' Division of Clinical Radiation Physics, Departmen f Radiotherapyo t , Essen D. FLÜHS . KOLANOSKH , I Institute of Physics, University of Dortmund, Dortmund Germany Abstract Clinical dosimetry is still far behind the goal to measure any spatial or temporal distributio f absorbeno d dose precisfasd an t e without disturbin physicae gth l situatioe th y nb dosimetry procedure 102E N .A plastic scintillators overcome this border. These tissue substituting dosemeter probes open a wide range of new clinical applications of dosimetry. e scintillatioTh n ligh transferres i t thina y db , multifibre plastic light guid vera o yet sensitiv stabld ean e miniature photomultiplier Cerenkoe Th . v light signal generate lighe th tn di guide is compensated by differential measurement [Flühs, 1989]. Multichannel photomultipliers allow simultaneous measurement with detector arrays. The plastic scintillator NE 102A is tissue equivalent for all energies of electrons and ß-rays as well as for photons above 100 keV. The detector can be prepared in any size and shape, so a very high spatial resolution can be achieved. This dosemeter syste larga s meha dynamic range detectoe Th . r doe t havy sno ean dependence of its response to dose, dose rate, temperature, pressure, or to the incidence of the radiation shor no ,w significana t chang responsf eo significana r eo t radiation damage. This versatil dosimetrw ene y system enables fast measuremen absorbee th f o t d doso et water in water also in regions with a steep dose gradient, dose to interfaces, or in partly shielded regions. It allows direct reading dosimetry in the energy range of all clinically used external photon and electron beams, or around all brachytherapy sources. Thin detector arrays permit higfasd an ht resolution measurement qualitn si y assurance, suc in-vivs ha o dosimetrr yo even afterloading dose monitoring maiA . n fiel applicatiof do dosimetrie th s ni c treatment plan- ning individuae th , l optimizatio brachytherapf no y applicators. Thus, plastic scintillator dosemeters cover optimall l difficulal y t field f clinicao s l dosimetry overvien .A w abou characteristics tit applicationd san gives si n here. 1. Introduction Radiooncology has done a big step forward by advanced methods of tumortherapy, such as intraoperative radiotherapy, afterloading brachytherapy, stereotactic treatment or con- formal therapy, total body irradiatio rotationar no l total skin electron radiotherapy. Improved 165 imaging technologie r localizationsfo NMR-based an - CT , d three-dimensional treatment plan- ning, Monte Carlo simulation based kernels for 3D-dose calculation, new real time verification methods, they all contribute to the success of radiotherapy. However, there are still open questions. Treatment planning algorithms are approxima- tions, only, coverin fieldl al radiotherap f t go s mosno t tbu y planning. Ther stils eneei e lth o dt determine dose regionn si s without secondary particle equilibrium steen i , p dose gradientsn i , the build-up region, at interfaces, in tissue heterogeneities, in partly shielded organs at risk, or in regions with contaminate mixer do d radiation beams. Clinical dosimetry toda stilbehins r yi fa lgoa e measurdth o t l spatiay e an r temporao l l distributio absorbef o n d dose directl precised yan . Clinical dosemeter t fulfi generae no th l o sd l requirement measurementr sfo physicae th , l situatio measuree b o nt ddisturbee musb t no t y db the measurement procedure itself. Dosemeter t tissuprobeno e e sar substituting e sizf eTh o . there sensitive volume is often too large. . 2 Plastic scintillator dosimetry The idea is, to use tissue substituting, direct reading detector probes for dosimetry with high response and high resolution [4]. Plastic scintillators are known since decades, but have been discovere clinicar dfo l dosimetry just recently [1-20]. Ther mane ear y materials availablet bu , the superior characteristics of the NE 102A plastic scintillator detectors indicate this material as an excellent dosimetry probe [4]. It has gained great importance for dosimetry. 102E N ATissu- 2.1 . e equivalent dosemeter probe 102plastiE a N s Ai c scintillator with high light output e materiaTh . s almosi l t fully tissue substituting (p=1.03 cm'2g , Ne-= 3.3 10x 9 cm' detectoe )Th . r doe t shosno wa significant change of response 3nor a significan23t radiatio3 n damage, as checked in a long time investigation oveyeae f continuouon ro r s irradiation withy b Cs-y-ray 160y G 0 s [4], .So NE 102A is suitable as dosemeter probe. 137 The NE 102A scintillator is well known in nuclear and particle physics for decades. higs Duit ho e t response, detector probepreparee b n s ca nearl n di r sizy fo ey an (e.gmm .1 brachytherapy [5-11,13-20]) and shap3 e (e.g. 0.1 mm thin for interface dosimetry [12,13]). Thus vera , y high spatial resolutio achievablens i . r researcInou h grou dosimetre pth y detecto uses i r d since seven year r differensfo t dosimetric applications [4] characteristics .It s opepossibilitiew nne f dosimetrso y with many physica clinicad an l l applications [4-20]. 2.2. The plastic scintillator dosemeter The scintillation light is transferred to a photomultiplier by a thin, multifibre plastic light guide, e.g cladde6 .1 wit0 d attenuatio n fibrem h a bundl a m n 2 i s f eo . nm) lengt6 1 f ho The Cerenko luminescencd van e light signal, generate e lighth tn di guid electrony b e f o s energies above 175 keV, is compensated by differential measurement in a bundle of parallel, but blind ending fibres [Flühs, 1989]. 1 As light detector, very sensitive (1 nA/1 mm scintillato3 r volume for 1 Gy h'), stable miniature photomultiplier tube usede sar . Their spectral sensitivity covers optimall emittee yth d rangm n scintillatoe 0 ligh th [4,8] 42 f o te .th Thin ri s dosemeter syste larga s meha dynamic range, more tha magnituden6 lineaf so r absorbed dose rate respons currene th n ei t mode [4]. Multichannel photomultipliers wit channel 6 since 1 hyears o us etw n r eves(i ) o n more separate channel channels6 s25 (e.go t preparation n i ,p . u ) allow simultaneous measurement 166 with many detectors. This enables linear or other detector arrays for 2D or 3D-dosimetry [5- 10,13-20]. 2.3. Physical application The plastic scintillator NE 102A is tissue equivalent in considering the interactions of electron ß-rayd energiel san al f photonsr o wels fo s a s a l(e.g sV abovke . rang0 e10 192f eo Ir), see Fig. 1. Compared to water, the mass absorption coefficient decreases slowly above 10 MeVpaio t e r production,du , while ther steea s ei p decreaskeV0 4 d , an e betweeV ke 0 n10 wit almosn ha t constant mass absorption coefficien(rangV ke 125f e5 o Id t an betweeV ke 5 n3 [8]. For protons (e.g. eye tumor treatment) the non linearity of response can be taken into ac- count during calibration. 1.5 1.3 TLD i i i Null IHI! 1 'i Ml/'i i mi 11 0.7 : I \A\\\\ I , 0.5 1 1 ! i / ! ! 1 . A I 1111 0.01 10 100 V energMe y/ Figure 1. Mass absorption coefficient for NE 102A plastic scintillators (S) relative to water, compared to TLD detectoe Th r doe t sho dependency sno wan responss it f eo doso et r doseo e rate (e.g. brachytherapy), to temperature (e.g. intraoperative radiotherapy) or pressure, nor to the inci- dencradiatioe th f eo n (e.g. stepping source afterloading). Thus detectoe th , suites ri mosr dfo t area physicaf so l application, dosimetr t steeya p dose gradients, at interfaces, at the surface, or in tissue heterogeneities like lung substitutes. They can be applied for electrons and ß-rays of all energies, for photons of external beam radiotherapy, as well as for those of brachytherapy. 3. Clinical application This versatile new dosimetry system allows fast measurement of the absorbed dose to water in water also in regions with a steep dose gradient (e.g. stereotactic treatment, brachytherapy), close to interfaces (e.g. in the build-up region), in tissue heterogeneities (e.g. in lung substitutes) or in partly shielded regions (e.g. the lung as organ at risk in TBI). 167 3.1. Basic dosimetry Faswitd tan h high resolutio dosl nal e measurements whic needee har basin di c dosime- try can be performed in a water phantom [4]. Different to ionization chambers, in electron beam dosimetry the absorbed dose is directly indicated. No energy dependent correction is needed. Plastic scintillator dosimetry allows not only direct reading basic dosimetry in the energy l rangclinicallal f eo y used external photo electrod nan n beams t alsbu ,o arounl dal common brachytherapy sources. 3.2. Dosimetric treatment planning Dosimetric scanning with a single detector or detector arrays permit fast 2D-dosimetry [5-20], e.g. for Ir afterloading applicator optimization [8,9,17-20], or for individual eye plaque preparatio 192n using I seeds, emitting photon ever o sn V belo SD-calibratioke 5 w3 f no 125 Ru/ Rh ophthalmic ß-ray applicators [5-8,10,13-17,20]. As the brachytherapy applicators ca 106optimizee nb 106 d durin minutew fe ga measurementf so , this techniqu usee b individs dn a eca - ual dosimetric treatment planning [5-10,13-20]. 3.3. Quality assurance The small size as well as the possibility to prepare detector arrays allow all measure- ments neede qualitn di y assuranc externaf eo l bea brachytherapd man y [8,9,14,15-20]. One main interes applicatiof o t in-vivs ni o dosimetry detectoe placee e Th .b th n t da rca body surface (in the build-up region). The detector array can be positioned in body cavities (no temperature dependence), in tiny catheters, e.g. in stereotactic therapy treatment (no directional dependenc response)f eo interstitian i r o , Ir-afterloadinl 192 g brachytherapy needles (no energy or depth dependence). dosimetrf o a er w openes yAi ne fasty db , high resolution plastic scintillator dosimetry. Afterloading dose monitoring is possible now [8,9,14-20]. A tiny detector array can be inte- grated into the afterloading applicator. Due to the high spatial resolution and the high time resolution the stepping source brachytherapy can be monitored independently and directly. To gain this technical information no in-vivo dosimetry is needed. All deviations in step position or step size in dwell time or stepping time can be indicated and measured directly. . 4 Conclusion Besides basic dosimetr maie yth n fiel clinicaf do l applicatio fase th t s ni 3D-dosimetri c treatment planning of brachytherapy applicators (e.g. eye plaques) and the high precision qual- ity assurance of afterloading applications. Thus, plastic scintillator dosemeters cover optimally all difficult field clinicaf so l dosimetr radiotherapyyn i . ACKNOWLEDGMENT This work is supported by the DFG, the German Research Foundation. REFERENCES [1] AS Beddar, T R Mackie, F H Attix, "Cerenkov light generated in optical fibres and other light pipes irradiated by electron beams". Phys. Med. Biol. 37 (1992), 925-936. [2] A S Beddar, T R Mackie, F H Attix, "Water-equivalent plastic scintillation detectors for high-energy dosimetry: I. Physical characteristics and theoretical considerations". Phys. Med. Biol. 37 (1992) 1883- 1900. 168 [3] A S Beddar, T R Mackie, F H Attix, "Water-equivalent plastic scintillation detectors for high-energy dosimetry: II. Properties and measurements". Phys. Med. Biol. 37 (1992) 1901-1913. FlûhsD [4,] "Aufba Einsatd uun z eines Szinü'llationsdetektorsystem Messunr szu Dosisverteilungen gvo n an einem Co-Teletherapiegerät", Physics Diploma work. Universit Dortmundf yo , 1989. FlûhsD [5HeintzM ], LübckeW , WieczorekC , 60 WieschoUekA , KolanoskiH , QuastU , , "Plastic scintillation detectors - Fast 3D-dosimetry based treatment planning of 125I and 106Ru ophthalmic applicators". Abstract. GEC-ESTRO Meeting, Venice 1993. [6J D Flühs, M Heintz, C Wieczorek, A WieschoUek, H Kolanoski, U Quast, "Plastikszintillationsdetektor Schnell- e 3D-Dosimetrie Bestrahlungsplanunfüe rdi 125106n d gRu/Jvo un 106Rh-Augenapplikatoren". Medizinische Physik '93, DGMP, 1993, pp. 358-359. [7] D* Fluhs, M Heintz, F Indenkämpen, C Wieczorek, H Kolanoski, U Quast, "Ein neues Dosimetrie- Konzept: Energie- und richtungsunabhängige, hochauflösende, schnelle 2D- und 3D-Dosimetrie mit Plastikszintilatoren". Medizinische Physik '93, DGMP, 1993 366-367. ,pp . [8] D Flühs, M Heintz, C Wieczorek, A WieschoUek, H Kolanoski, U Quast, "The measurement of absorbed dose by plastic scintillation detectors - A short theory and six years of practical experience". Tpublishee ob Medicadn i l Physics, 1994. FlühsD [9IndenkämpenF ,] HeintzM , KolanoskiH , QuastU , , "Tissue equivalent plastic scintillator probes: Fast, precise 192Ir-afterloading dosimetry. Dosimetric treatment planning and verification". Abstract, GEC-ESTRO-Meeting, Linz 1994, Radioth. Oncol. 31, Suppl. 1, 1994,26. [10] D Flühs, U Quast, H Kolanoski, M Heintz, "Dosimetric treatment planning - fast 2D/3D-dosimetry". Proc. Xlth ICCRT, Manchester, 1994, pp. 38-39. [11] M Heintz, "Einflüsse von Detektoren bei der Dosismessung in Gewebeinhomo-genitäten", Physics Diploma work, Universit Dortmundf yo , 1992. [12] M Heintz, D Flühs, U Quast, H Kolanoski, St Reinhardt: "Dosisbestimmung bei der Ganzkörperbestrahlung - Dosimetrische Konzepte und Monte Carlo Rechnungen". In: Roth (Ed.) Medizinische Physik '92. DGMP, Basel 1992, pp. 86-87. [13] U Quast, D Flühs, M Heintz, A WieschoUek, W Sauerwein, N Bomfeld, H Friedrich. "Concept of intraocular tumor treatment: Plaque preparation, 3D-dosimetry, treatment planning, and documentation". Abstract. GEC-ESTRO Meeting, Venic' e 1993. QuastU [14 ]Flühs D , HeintzM , LübckeW , KolanoskiH , concepw ne f dosimetryo A t" , : Energy independent, high resolution absorbe,D 3 rea d l an timd D dose2 e measurement". Abstract. ESTRO- Physics-Meeting, Praque 1993. [15] U Quast, D Flühs: "Real time 2D/3D absorbed dose measurement - A new concept of dosimetry in radiotherapy". Invited lecture, abstract. XTV AMPI-Conference, Nagpur, India, 1993, Med. Phys. Bull. (in press). [16] U Quast, D Flühs, W Sauerwein, W Friedrichs, "A general concept of treatment planning and confirmatio ophthalmin i n c tumor brachytherapy". Int. Symp. Intaocula Epibulbad an r r Tumors. Firence 1994,6. [17] U Quast, "A general concept of dosimetry, ideal for advanced radiotherapy techniques luce TBI, TSER, IORT, confonna bracbr o l y therapy. Proc. Int. Congr advancen o . d Diagnostic Modalitiew ne d san Irradiation Techniques in Radiotherapy, Perugia, 1994, pp. 179-186. [18] U Quast, D Flühs, M Heintz, F Indenkämpen, H Kolanoski, "Tissue equivalent plastic scintillator probes: Fast, precise 192Ir-afterIoading dosimetry. Dosimetric principles and clinical application." Abstract, GEC-ESTRO-Meeting, Linz, 1994 Radioth. Oncol. 31, Suppl. 1, 1994, 22. [19] U Quast, D Flühs, F Indenkämpen, " The build-in dose monitor - A new dosimetric concept for afterloading quality assurance". Abstract. ESTRO-Meeting Granada 1994. QuastU [20 Flühs] ,D Fast" , , high resolution, energy independent 3D-absorbed dose measuremenw ne A t era of dosimetry in radiotherapy". Abstract. World Conf. Medical Physics and Biomédical Engineering. Rio de Janeiro, Brasil, 1994. -x-'yvyH.y-'v-v ,—•/•••• v"-jf r.v-x ::EXT PÄÜEIS) I 169 r _laftJSI.A:.'X J ABSORBED DOSE BEAM QUALITY FACTOR CYLINDRICAR SFO L ION CHAMBERS: EXPERIMENTAL DETERMINATIOD AN 6 T NA 15 MV PHOTON BEAMS C. CAPORALI, A.S. GUERRA, R.F. LAITANO, M. PIMPINELLA ENEA-Casaccia, Departunento Amiente, Institut•w--.-i.-v-e Naaona* lItf Metrologi. 1*11a déli1e I Illll IIIIIHXA964285 1111 Hill 11111 !!••I I !1 II IBIIII I 7 Radiazioni lonizzanti, Rome, Italy Abstract Ion chambers calibrated in terms of absorbed dose to water need an additional factor conventionally ordedesignen i determing o t rk y db e eth absorbed dose quantite Th . y UQ depend bean so m qualit d chambeyan r characteristics. Andred Rogeran ] provide] s[1 o[2 d calculationQ k e th f so factors for most commercially available ionization chambers for clinical dosimetry. Experimental determinations of the kg factors for a number of cylindrical ion chambers have been made and are compared with the calculated values so far published. Measurements were made at 6 MV and 15 MV clinical photon beams at a point in water phantom where the ion chamber a Frick d ean sdosimete r were alternatively irradiatede Th . uncertaint e experimentath n yo factorQ k l s resulted about ±0.6%e Th . theoretica experimentad an l l kg valuefairln i e ysar good agreement. 1. Introduction The usual dosimetry procedure in which the 4onization chambers are calibrate termn di f air-kermso bases ai rathea n do r complex formalism and requires the determination of several parameters the uncertainty of which is not low. Thus systematic uncertainties and mistakes in data t negligibleno handlin e b y . gma Moreove e primarth r y standards upon which this procedur bases ei d are, everywhere, graphite cavity chambers. This makes it difficult to detect possible systematic uncertainties as these standard sam e basee th s ar e n do measuremen t procedures present A . t time however the majority of national standards laboratories have developed primary standards of absorbed dose to water, D, for the Co-60 gamma radiation. These standards are based on differenw t methods so that their intercomparison mak] s[3 t possiblei deteco et t possible systematic errors. The calibration in terms of Dw is based on a formalism that is conceptually more straightforward than that relevant to the air-kerma standards. Moreove e parameterth r s enterine gb intn e expressioca th o w D f o n determined principlen o , , wit hbettea r accuracy. When expresse termn di s of the absorbed-dose-to-water calibration factor, the dependence of Dw on the photon or electron beam quality is described by a factor originally uncertainte Th . denotekQ this n ya o s] factod [4 resul y t necessarilma r no t y kf i calculategs i w lo d fro singlvalue e mth th f eo e parameters upon which it depends. 171 AndreRogerd an compute] ] os[1 [2 value Q k r mane sfo dth y ionization chambers wit n uncertaintha o 1.5t %p u aboutf yo e possibilit Th . f yo achieving better accuracy levels in absorbed dose determination is importan decidinn i t g adop o whethet t tno calibrationr ro . termw n si D f so The present work was then undertaken with the aim of developing an experimental procedure to determine the kQ values with an accuracy appreciably higher than that resulting from the calculation of these factors. easile b expressioe n yca Th obtaine g k f no d just fro e definitiomth f o n absorbed-dose-to-water calibration factor, Nw, from which one has: DW=MNW (i) where M is the chamber reading corrected for ambient conditions and ion recombination. If the factor Nw is determined at the Co-60 calibration beam, equation (1) holds onl t thiya s quality othee th rn han.O expressioe whicdo th t w hD f no the dosimetry protocols refer for measurement at any beam quality, Q, is that air-kerme baseth n do a calibration factor (e.g, NK ., IAEA protocol [5]) an gives di : nby (2) where Nr>=NK(l-g)kmkattkcel (3) the other symbols having the usual meaning [5]. Equatio quality y hold) nan (2 t sa , includin calibratioe gth n quality The. c , n for the same ionization chamber having both the calibration factors Nw d irradiatean e calibratioD th N t a dd an n quality e obtainon , s from equations (1) and (2): (4) Finally fro findse mon ) equation: (4 d an ) s(2 Dw=MNwl———__-^=MJNwkQ (g) where , (vanPil (6} The factor kQ corrects the calibration factor Nw for beam quality dependence in radiation beams different from the Co-60 gamma-ray. 172 2. Experimental equipment experimentae Th l determinatio somr fo Q ek ionizatio f no n chambers wa s o photomadtw t a en , obtainebeamMV 5 qualities1 d frod an ma 6 , Siemens Mevatron MD Dual Photon linear accelerator. Type and characteristics of the ionization chambers are described in table I. Chamber calibrations were performe Co-6e th t 0a d gamma-ray quality usinga gamma irradiator mod. AECL Eldorad Co-6q TB 0 wito 0 6 source 6 ha e .Th chambers were irradiate n watei d r phantom with their individual waterproof sheath. Chamber sheaths were made of 0.5 mm thick PMMA accurately machine minimizo dt r thicknesai e eth s betwee chambee nth r sheathe th wald .an l Eac h sheat suitabls hwa y designe assuro dt e chamber cavity venting even during measurement in water phantom. Two different water phantoms were used. Measurement t acceleratosa r vertical beam were performe phantoa n di m realize sidm dc e wit 0 ope3 ha n Tabl Characteristic. I e cylindricae th f so l chambers use thin di s work, accordine th go t data taken from the manufacturer datasheet. chamber wall wall internal internal nominal central type material thickness length radius volume electrode (gcnr2) (mm) (mm) (cm3) material NE 2571 graphite 0.065 24.1 3.15 0.69 aluminium NE 2561 graphite 0.090 9.22 3.675 0325 aluminium Capintec PR-O6C C-552 0.050 22 3.2 0.65 C552 PTW M233642 PMMA 0.090 6.5 2.75 0.125 .... ENEA graphite 0.087 20 2 0.24 graphite ended cubic tank thicwitm c kh1 PMMA wall phantoe Th . accelerator mfo r vertical beam was also provided with two additional supports for two small cylindrical waterproofed chambers. This pai f chambero r s wa s positioned symmetrically with respect to beam axis. The aim of these two supplementary chamber accuratelo t s wa s y monitor beam outpud an t flatness watee Th . r phanto irradiatior mfo horizontae th t na l Co-60 gamma provides bea similas wa previoue mt wa th bu o drt e ) witson hthimm a 3 n( PMMA windo side th e n wwallo measuremenl Al . t data were processen di line by a computer interfaced with the electrometers and with the temperatur pressurd ean e probes. Measurement t Co-6sa 0 gamm wery ara e done with a 10 x 10 on2 field size at 100 cm SDD. The Co-60 gamma-ray dos emeasuremene ratth t ea t poin phanton i t abous mGy/minm0 wa 30 t . Measurement t acceleratosa r photon beams were made with higher dose rate f abouo s t 170 d 0160 mGy/mian ) 0 mGy/miMV 6 ( n MeV)5 (1 n , respectively SDDm photoc e 0 .Th 10 nt a ,collimato r settin t acceleratoga r 173 beam cm acceleratoalwaye 0 s 1 2sTh .wa x 0 s1 r beam qualit specifies ywa y db was the TPR2o/io parameter. The TPR2o/io determined by depth ionization curve waten si r phantom chamben io e keps fixeTh .n wa r i t d positiot na watee th d r an leve s varieD wa l SD d 10accordinm 0c procedure th o gt e recommende IAEy db A TPR e [5]Th . 2o/iod waan sV 0.67M 0.75d 6 3r an 7fo photonsV 15M , respectively l beaal t mA . qualitie ionizatioe sth n chambers (including the two monitor chambers) were positioned with their centre at the reference depth. At the Co-60 gamma beam the reference depth was 5 acceleratoe th r Fo . ron beam reference sth r fo m ec depth0 1 d san werm c e5 TPRe th 2o/i 0.67f o o 0.757d 3an , respectively. Absorbance measurementr fo s ferrous sulphate dosimetry were made by a precision double beam spectrophotometer, mod. Gar , interface1 y a compute o t d r datfo r a acquisitio processingd nan . 3. Methods The experimental determination e kgth e ionizatiof factoro th r fo s n chambers liste s basen tabli d wa measurementn o dI e s with ferrous sulphate (Fricke) solution. When absolutn usea s da e measuring method, Fricke dosimetry does not have sufficient accuracy if one needs to keep the uncertainty below ±1%solutioe Th . n adopte thin di sdesig o t stud s nywa an experimental procedur whicn ei knowledge hth f absorbeeo d doso et needee b o determinint r t dfo no wate d ha r k e ggth factors. Accordinglye th , only important characteristic required for the Fricke solution was a response reproducibility adequat objective th r efo e here considerede Th . Fricke solution prepared for this investigation was thoroughly tested for severa ls irradiate monthwa d san PMMn di A vials choicviae e th l Th . f eo materia Frickr fo l e dosimetr t immediateno s yi . Glas more s b vial n e sca easily cleaned thus allowin gmora e simple contro storage th f o l e effects. However the PMMA was definitely chosen because for this material the wall correction factor for possible non water equivalence is very close to e [6]on . Measurement f absorbancso e were mad bot n irradiatee o hth d t irradiateno f o t dsamplese controlsa d san . PMMA cylindrical vials were used. They were realize waldm m witl 5 thickness h0. internal m m 0 1 , diameter and 30 mm length. In order to remove residuals of substances used for the preparation of the vial (glue, PMMA particles, etc.), all the vials underwent a treatment consisting of an immersion for 30 minutes in an ultrasound bath, followe severay db l rinsings with distilled y wateb d ran a pré-irradiation at about 5 kGy. With the procedure above described, a reproducibility of better than 0.4% (1SD) on 5 samples was obtained with irradiations of not less than 30 Gy. A ferrous sulphate solution of about 5 liter prepares swa stored dara dan n dki plac t stablea e temperatura n ei glass container. Homogeneit stabilitd solutioye an th f yo n were periodically checke perioa r d fo abouf do mont1 t h during whic Fricke hth e irradiations e presenfoth r t study were performed time f permanencTh e.o e th f eo solution in the vials (about 2 hours) was kept constant and as short as reasonably possible f necessaryI . orden i , limio t rstorage th t e time/the solution was, at the end of the irradiation, transferred to accurately purified glass ampoules. This donwa s e mainl r irradiationfo y s carriet a t dou accelerator beams, located far from the authors' laboratory. The storage 174 tim eacn ei same hbotth r efo s via h wa irradiatel d sample controlsd san . The centre of the vial was positioned at the measurement point in water phantom same th t e A .dept phanton hi ionizatioe mth n chambers under investigation were positioned with their geometrical centre coincident wit measuremene hth t point procedure Th . determino et e experimentally thfactorQ e followink e base s th wa sn do g rationale absorbee th : d doso et water at any beam quality Q as measured with a calibrated Fricke solution is give: nby ) (7 {DW/F)Q=(AAF)QNW/F where (AAp)Q is the difference in absorbance between the irradiated and unirradiated solution, and NW/F is the calibration factor of the Fricke solution in terms of absorbed dose to water. The factor N/F/ determined Co-6e ath t 0 gamma radiation assumes ,i constane b o dt t sinc evidenco en e W of its dependence on electron energy (at least in the energy range of interest in this study) appears from the data so far available (e.g. ICRU 35 [7]). If an ion chamber calibrated in terms of D is used, the absorbed dose to water at any beam qualit s givei equatioy yQ nb ne sam th (5) en I . irradiatiow n conditions the calibrated Fricke solution and the calibrated ion chamber must measure the same Dw value when alternatively positioned at the same phantom depth. Accordingly hase on :) equatione (7 th d y an ,b ) s(5 Dw=MQNwkQ = tAAF)QNw/F (8} On the basis of their definition the calibration factors Nw and NW/F, determined at the calibration quality c, are given, respectively, by: (D„) Finally fro obtainse mon ) equation(9 : d an ) s(8 (AAF)Q MC I10) correctee th e ar dQ chambeM d wheran C absorbere ereadingM th o t e dsdu dose givin Frickge th ris n ei solutio difference th o nt absorbancn i e e (AAp) (AAp)Qd can , respectively. Accordin above th o get procedur equatioo t d ean n e factog (10)b k n e ca ,rth obtaine s functioda f quantitieo n determinee sb than ca t d experimentally with fairly good accuracy. Five series of measurements were made for each ionization chamber and beam quality. Afte monte on r h these measurements were then repeatey db repositioning the experimental setup. It was thus possible to check the long term reproducibility regarding the stability of measurement and the 175 mechanical equipment. Measurements with Fricke solution were mady eb five irradiations on five distinct vials for each beam quality and then takin meae gth n five valuth e f readingseo absorbee Th . d dos ferrour efo s sulphate irradiation was about 30 Gy. Also for Fricke dosimetry the series of measurements were repeated after one month about to assure the reproducibility of the overall dosimetric method. 4. Results and discussion The experimental value determineg k f so photoo tw t da n beam qualities are reported in table II for the ionization chambers listed in the first column. The results obtained in this work have been compared with the values computed (accordin equatioo gt Andrey Rogerd b ) an n6 ] o[2 s e [1]Th . differences amon e kgth g computeo t de valuedu e n tablar i s I I e difference sete parameterf th so n i s s considere thosy db e author their sfo r calculation. Andreo used stopping power ratios recalculated by himself and other parameters from the IAEA protocol [5]. Rogers used the basic parameters eithe e IAEr th ] protocols froAr [5 e AAPo mth ] M[8 , respectively above Th . e authors adopte thein di r calculation "prepe sth l approach". For data comparability the same approach was then used in this study. Measurement ionizatioy b s n chamber n watei s r phantom were made by positioning the chambers with their centre at the measurement point. Therefore the experimental kg factors of this work include the correction factor for water replacement, plp - To allow the comparison, the data by the other authors reported in tablre e n were interpolated to refer same th eo t TPR20/10 values use thin di sENEe workth r A Fo .ionizatio n chamber, the reference chamber of the italian AIFB dosimetry protocol [9], the theoretical kg factors were not computed by Andreo and Rogers but by the present authors, usin same gth e basic data considere thosy db e authors. The average deviation between the present results and the computed data of Andre Rogersd oan , respectively abous ,i t 0.6%. This shows tha thesl al t e determinations are rather consistent. Although the AAPM parameters are not the most updated, the data based on the AAPM parameters (column 3) resulslightle b o t t y experimentaclosee th o t r l ICQ values, tha othee nth r computed data. The experimental data either of this work or of other author alwaye ar s s slightly higher tha e computenth d ones. Actuallyt i , should be noted that all the data in table II are within the stated uncertainties. The uncertainty on the kQ factors is due to the experimental uncertainty of the individual quantities entering into equation (10). Accordingl overale yth l random uncertaintvalueQ k e s th determine n yo d in this study is about ±0.6%. The few experimental determinations of kg so far available are those determined at NPL [10] and PTB [3]. The comparison with the present result shows si tabln i Wit. eHI h respec result e presene th th o f t s o t work date th t TPR20/1dat ae a d froth B aan mfroPT L fairl°-670n m= i NP e y3ar good agreement, wherea deviatiosa abouf no t 0.8B % obtaines i PT e th r dfo result at TPR20/10 = 0.757. To this regard it should be taken into account dat baseB e a ar measuremenPT n do d an L parameterse NP tha th e f to tth s ,a N , r directlo D y entering into equatio . Thin5 s procedurmose th t t no s ei w w accurate because of the intrinsic uncertainty on Dw. Moreover primary 176 Table IL Experimental and calculated kQ factors for the various chambers used in this work. TPR20/1 0-= chamber experimental calculated calculated calculated type this work Andreol992 Rogers 1992-a Rogers 1992-b NE 2571 0.998 0.993 0.995 0.992 NE 2561 0.999 0.994 0.995 0.992 Capintec PR-O6C 0.999 0.995 0.998 0.993 PTW M233642 0.987 0.989 0.994 0.989 ENEA 0.998 0.993 0.996 0.994 = 0.757 chamber experimental calculated calculated calculated type this work Andreo 1992 Rogers 1992-a Rogers 1992-b NE 2571 0.983 0.980 0.983 0.978 NE 2561 0.982 0.982 0.983. 0.979 Capintec PR-O6C 0.985 0.980 0.986 0.977 PTW M233642 0.970 0.975 0.981 0.972 ENEA 0.983 0.979 0.985 0.979 standards of Dw in the various national laboratories have at present deviations of up to 1% [11-12]. Therefore the deviations among the data in wele tablar lI withieII experimentae nth l uncertainty pointes y A .b t dou Andre als] computee o[2 oth valueg dk uncertaintw lo s t havno ea f y o about ±1.5%. These computations, base equation do ) cannon(6 n o e b t principle very accurate becaus e intrinsith f o e c uncertaintiee th n o s individual parameter thin si s expression. The results of this work are instead based on measurements that can be performed with the highest accuracy and this procedure seems rather promising for determining the kg factors at any beam quality. 177 Table in. Comparison of the experimental kQ factors for the NE 2561 ionization chamber. NPL PTB, in TPR20/10 this work Owenl993[10] Boutillon 1993[3] 0.673 0.999 0.9973 0.998 0.757 0.982 0.9789 0.990 REFERENCES [I] Rogers, D.W.O., The advantage of absorbed dose calibration factors, Med. Phys. 19 (1992) 1227-1239 [2] Andreo , Absorbe,P. d dose beam quality dosimetrfactore th r sfo f yo high-energy photon beams, Phys. Med. Biol. 37 (1992) 2189-2211 [3] Boutillon , Coursey,M. , B.M., Hohlfeld Owen, Rogersd K. ,an . B , , D.W.O., Comparison of primary water absorbed dose standards, Document CCEMRI(D 793-33 (1993) page7 1 , s [4] Hohlfeld, K, The standard DIN 6800: Procedures for absorbed dose determination in radiology by the ionization method, in Dosimetry Radiotherapn i y Vol.1, Vienna, IAEA (1988) 13-22 [5] INTERNATIONAL ATOMIC ENERGY AGENCY, Absorbed-Dose Determinatio Photon ni Electrod nan n Beams Internationan A : l Cod Practicef eo , IAEA Tech. Rep. Ser 277. n . , IAEA, Vienna (1987) [6] Ma, C, Rogers, D.W.O., Shortt, K.R., Ross, C.K., Nahum, A.E. and Bielajew, A.F., Wall-correctio absorbed-dosd nan e conversion factors for Fricke dosimetry: Monte Carlo calculations and measurements, Med. Phys.20, (1993) 283-292 [7] INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS, Radiation Dosimetry: Electron Beams with Energies Between 1 and 50 MeV, ICRU Rep. 35, ICRU Publications, Bethesda, MD (1984) [8] AMERICAN ASSOCIATIO PHYSICISTF NO MEDICINESN I A , . protocol for the determination of absorbed dose from High-energy photon and electron beams, Med Phys. 10 (1983) 741-771 [9] ASSOCIAZIONE ITALIANA DI FISICA BIOMEDICA, Protocollo per la dosimetria di base nella radioterapia con fasci di fotoni ed elettroni con Emax fra 1 e 40 MeV, Roma, AIFB (1988) [10] Owen, B., private communication, (1993) [II] Shortt, K.R., Ross, C.K., Schneider Hohlfeld, M. , d Roos, an K ,. M , Perroche, A.-M. compariso,A absorbef no d dose standard highr sfo - energy x-rays, Phys. Med. Biol (19927 .3 ) 1937-1955 [12] Boutillon Perroched an , lonometri. M. M , . A , c déterminatiof no absorbed dose to water for Co-60 gamma rays, Phys. Med. Biol. 38 (1993) 439-454 178 DEPLACEMENT CORRECTION FACTOR VERSUS XA9642858 EFFECTIVE POINT OF MEASUREMENT IN DEPTH DOSE CURVE MEASUREMENTS AT "Co GAMMA RAYS A. BRUNA Facultad de Matemâtica, Astronomia y Fisica, Medina Allende-Haya de la Tone, Ciudad Universitaria G.R. VELEZ Hospital San Roque, Depaitamento de Radioterapia M. BRUNETTO Centre Médicx) Privado Dean Funes Cordoba, Argentina Abstract e discrepancieTh datn si a set valuef Displacemene so th f so t Facto p rrecommende differeny db t codef so practice r calibratiosfo n purpose still demand d further investigatio clarifo nt y this point thin I . s papere w , propose an experimental method to determine the displacement factor for cylindrical ionization chambers (thimble chambers) in photon beams. Measurement of Pd for several depths were performed for '"Co gamma rays. From these result calculatee sw effective shifth e df th o t e poin f measuremeno t t (z-Zeß r differenfo ) t depths. The results obtained in this work shown : (a) there is no significant change in pd from 2 cm to 17 cm ion-chambe a deptf r o fo valuwatere a hn i p th f ) eo (b , r Farmer type (inner pradius 0.988i d= ) 3.1= sr cm 5; (c) the shift of the effective point of measurement has a smooth variation with depth; (d) the value of (z-Zeg) recommendee th t a d calibration dept ^Cr h0.6fo s oi ) rbeam (witcm 5 s( : innehr r radiuchamber)e th f so . The result (b) confirms the value of pd suggested by the SEFM and NACP protocols and differs with that of e AAPMth value Th . e obtaine Zeff)(d- z ( vers r i )d fo y close thao dt t recommende IAEe th y Adb TRS-277. Finally, the results (a) and (c) suggest that it should be preferable to use the displacement factor instead of effective poin f measuremeno t perforo t t m measurement deptf so h dose curves f z,,«o e ,shoul sincus e deth take into account its dependence on depth. 1. INTRODUCTION The Bragg-Gray theory relates the ionization produced in a small gas-filled cavity place homogeneoun a n di s mediu absorbede th mo t e medium doseth n i . Howevere th , particle fluenc cavite ionizatioth a f n ei yo n chamber inserte mediua n di m wil longeo n l e rb representative of the fluence at the point of interest in the medium, because of the complex differences in attenuation and scattering of radiation due to the replacement of the medium material by the cavity material. The displacement factor pj, is one method of correcting for such perturbation, and so that it could be determined as the following quotient, for a given radiation quality: u , J«ir.u 179 wher cavitr e ai meaJiir,u An other metho f correctino d r thigfo s o effecdefint s i n effectivta e e poinf o t measurement, z«ir, radially shifted from the geometrical centre of cylindrical ion-chambers throug fronte hth . In order to have consistency in these definitions, both correction factors that are essentially the same perturbation correction, are related by the following equation : PDDfr) F PDD(z-(z-z«rr)) where z is the depth of the chamber axis (geometrical centre); Zeir the depth of the effective poin measuremenf o t chambere th f o tPDD(z d percen;an e th s )i t depth dose. There are significant discrepancies between the values of the displacement factor recommended by different authors and protocols. For example, the AAPM TG-21 [1] valuee th f o ssuggest calculatee us e sth Cunninghay db Sontad man g [2], which diffey b r about 0.5% from those measured by Johansson et al. [3] at Co. These last values of pj are recommende SEFy wels db a NACP[5] s ] a lM [4 the d consistene yan ,ar t wite th hshifa f o t effective point of measurement of (z - zeff)= 0.6r. IAEA TRS-277 [6] recommend (for Co) to use the value (z - Ze«r)= 0.5r, assigned to Johansson et al. [3]; however, in that paper, the average of (z - zdr) for different depths is 0.6r whiled there is no value of 0.5r for any depth. To clarify the discrepancies in the existing data sets for displacement factors, by means of experimenta r theoreticao l l techniques vers i , y important, mainl f calibrationyi termn i s f so absorbed adoptedose b nea e wateo eo t t th re n dfuturei r ar . 2. MEASUREMENT OF THE DISPLACEMENT FACTOR pd 2.1. Method The displacement factor migh e obtaineb t d fro equatioe mth measuriny b ) n(1 e gth mean ionization J»ir,u Hm Jair,iKr) Jiir,u(r) perforo T experience mth e exactl this ya s formula suggests t wouli , necessare db o yt have severa n chamberio l s with different interna e samlth radit e bu wali l materiad an l thickness. However, becaus r experimentaou f eo l limitations propose w , methoea e th r dfo determinatio measuremen e based th p f n ndo o ionizatioe th f o tchambe a n ni r locatee th n di centr severaf eo l cavities with different diameters maie Th n. assumptio made nw thas ei e th t ionization measured by the chamber will accuse the variations in the mean ionization of the air cavity, i.e. the variations in the ionization inside the cavities as varying their radii due to the lack of phantom material. 180 experimenr Ou soli a carries n i d wa tt phantodou m (acrylic orden )i achievo rt e high reproducibility in our measurements. The ion chamber was a NE 2571 (0.6cm and graphite wall). The phantom was constructed with 10 slabs of acrylic 200mm x 200mm x 20mm. One of these slabs has a cylindrical hole in order to allow the placement of different accessories. Figure 1 shows a scheme of these devices in the phantom. In this way, it is possible to ensure an accurate position of a thimble chamber in the centre of a cavity whose diameter could be varie y changinb d e acrylith g c accessory cylinder opposit e ion-chamberth o t e . Such cylindrical accessories consist in acrylic rods with wells of different diameters in one of its extremes in order to provide several cavities with different diameters surrounding the thimble chamber, as it is also shown in figure 1. To obtain electronic equilibrium (necessary for dose measurement airn si thic )graphitm a k2m (samp eca e wall material s mad)wa e which fite th s thimble diameter. -10cm- 10cm- carrie ionizatiof ro n chamber Fig. 1. Lucite phantom-with accessories and ionization chamber. Three cavities with radii r = 11.1, 12, and 16 mm were made to place in the phantom successively (mentioned wells in one end of the rods). The geometrical conditions of the irradiation were those recommended by the AAPM protocol [1] for calibrations in acrylic phantoms. 2.2. Results The valuedisplacemene th f so t facto r cavitier ai versu e radith e s sf th obtaineo i n di thi gammo sC wor t ak (a rays showe ar ) figurn ni togethe2 e r wit valuee hth s givey nb Cunningham [2] and Johansson [3]. Graphs of pd vs. r were obtained at several depths in the phantom: z = 1, 3, 5, 7, 9, 11, 13 and 15 cm. The curve shown in figure 2 corresponds to the average for all the mentioned depths, becaus excepcavite m th 1c thir e y= te fo son breakthaz f o t s througe hth build-up region. To illustrate the measurements of pd vs. depth (z), figure 3 show these values for r = . mm 5 3. 181 1.000 ----- Johanssonetal(1978) —- — Qmingia Sontad man g (1980) ^-^^ This v»fk " 0.975 234 Radu cavitr ai f syo [mm] Fig. Displacement2. factor radiusvs. Co. cavity ofair for 1.000 0.995 0.990 0.985 g 0.980 "g< 3 0.975 0.970 0.965 8 1 6 1 4 1 2 1 0 1 8 6 4 2 0 Depth in water z [cm] displacemente Th . 3 : factor depth. vs water n cavityi a r fo of radius 3.5mm= r 182 3. STUDY OF THE SHIFT OF THE EFFECTIVE POINT OF MEASUREMENT 3.1. Method The radial shift of the effective point of measurement was calculated from the values of the displacement factor pd obtained in this work, and using equation (2). The values of percentage depth doses were taken from the Br.J.Radiol. Supl.17 (1983) [7], for the corresponding field size and source-surface distance at MCo gamma rays. In our experience, referencd an m c e0 8 fiel = d S sizDP e 10cm 10cmx . Since there is no significant change of pd with z from 2cm to 17cm of depth in water, depthsl usee al w r dfo jus.value d p on tf eo 3.2. Results The values of the shift of the effective point of measurement versus depth is shown in figure 4. 0.70 • 0.65 \ ; t 0.60...... " N:; 0.55 ^"""•"B,^^^ 0.50 0.45 ! 1 6 1 4 1 2 1 0 1 8 6 4 C 2 Depth in water z [cm] Fig. 4. The shift of the effective point of measurement (z-zeff) calculated from displacementthe factor (equationpj depthand 2) dose dataBJR in suppl.17 [7]. 4. DISCUSSION. As it can be observed in fig. 2, the values of pd obtained in this work are in good agreement with those from Johansson et al. but there are differences higher than 0.5% with Cunningham & Sontag. Our results about no evidence of significant variation of pd with phantom depth (see fig. agre) 3 e with those from Johansson. From the conclusions mentioned above it is clear that the shift of the effective point of measurement has to vary, although smoothly, with phantom depth, as it is shown in Fig.4. 183 Consequently perforo t , m measurement f depto s h dose curve t depthsa s greater than (dept maximuf ho m dose t woul)i displacemenpreferable e db th e us o et t factor pa insteaf do using Zeff, sinc shoullase e eth on t d take into accoun dependencs tit depthn eo . ACKNOWLEDGEMENTS A special recognition to José McDonnell and Jorge Hisano for their helpful presence in the measurements and their instruments and equipment's. REFERENCES ] [1 AMERICAN ASSOCIATIO PHYSICISTF NO MEDICINE(AAPM)N SI A , protoco e determinatioth r fo l f absorbeo n d dose from high-energy photod an n electron beams. Med. Phys (19830 1 . ) pp.741-771. [2] CUNNINGHAM, J. R., SONTAG, M. R., Displacements corrections used in absorbed dose determination, Med. Phys (1980)7 . , pp.672-676. ] [3 JOHANSSON, K.A., MATTSSON, L.O., LINDBORG , SVENSSONL. , , H. , Absorbed-dose determination with ionization chamber electron si photod nan n beams having energies between 1 and 50 MeV, (IAEA-SE-222/35), pp.243-270 in National and International Standardizatio f o Radiation n Dosimetry l II(IAEA)Vo . , Vienna(1978). [4] SOCDEDAD ESPANOLA DE FÎSICA MÉDICA, Procedimientos recomendados para la dosimetria de fotones y electrones de energias comprendidas entere l MeV y 50 MeV en radioterapia de haces externes, Publication n° 1(1984) (Madrid SEFM). [5] NORDIC ASSOCIATION OF CLINICAL PHYSICS, Procedures in external radiation therapy dosimetry with electro photod nan n beams with maximum energies between 1 and 50 MeV, Acta Radiol., Oncol. 19 (1980) 55. [6] INTERNATIONAL ATOMIC ENERGY AGENCY, Determination de la dosis absorbida en haces de fotones y electrones, Côdigo de Prâctica International, Tech. Rep. Ser 277(1990. .n° ) (Vienna:IAEA). [7] BJR 17, Central axis depth dose data for use in radiotherapy, The British Institute of Radiology, London (1983). 184 AN ANALYSIS OF SOME ASPECTS OF THE ATTENUATION - SCATTER FUNCTIONS IN BRACHYTHERAPY DOSIMETRY S.C. KLEVENHAGEN Department of Medical Physics, XA9642859 The Royal London Hospital, London, United Kingdom Abstract An analysi presentes si attenuation-scattee th f do r functions radial dose functions) employe brachytherapn di y dosimetry which accounts for the interplay between attenuation and scattering along the radial distance from the source. Some of the characteristics of these functions are still not established with certainty and are subject of misinterpretation. Such issues like whether they shoul normalizee db r notdo , particularl relation yi e th o nt currently employed source strength specification in terms of air kerma, are not as yetagreed. In the literature, the function presentee ar s d eithe normalizes ra non-normalizer do differencee th t dbu s between the wrongle mar y interpreted as being due to either computational or experimental uncertainties. Furthermore, there is uncertainty about the attenuation-scatter ratio very close to the brachytherapy sources and , in the case of some functions, at larger radial distances. Although the function's value at close distance may seem of lesser dosimetric relevance, it is important if one wants the underlying physics to be correct. These problems were studied in this analysis on the basis of the available data. An experiment was also carried out in order to determine the scatter component e closth en i vicinit sourcee th e stud o yt dat e Th .s baser Iridium-19i yth afo n do e discussio th t 2bu d nan conclusions are relevant to all types of brachytherapy sources. It is concluded in this analysis that; i) it is incorrect to be comparing the normalised with non-normalised functions, ii) only non-normalised (the natural) functions suc s thaha t derive Meisbergey db l (1968a t e r Sakellio o ) t (1992a t ue e correc ar ) r dosfo t e calculation systems based on the recommended air kerma source specification iii) the function should not have a value of unity at r = 0 because of the scatter domination over attenuation in the space around the source and iv) the Van Kleffens-Star function is in error at larger radial distances. 1. INTRODUCTION A numbe f workero r s have investigate behavioue dth f radiatioo r vicinite waten ni th n i rf yo brachytherapy point sources, applying the results to the dosimetry of nuclides. In particular, large interest has been shown in the relationship between attenuation and scatter along the radial distance fro source mth e becaus significancs it f eo dose th en e i calculatio n algorithm generaln I . , scattering compensates partiall r attenuationyfo combinee th , d factor, A-S, attenuation-scatter function varying in magnitude slowly with distance and energy. This function is also known as radial dose function. In the days of radium therapy and until the early 1960s the attenuation-scatter contribution to dose was mostly ignored, only inverse square law was considered relevant. With the development of interest in the radium substitutes for brachytherapy this situation has changed. This analysis involves work done from this period onwards. The published functions up to now have been derived either experimentally or theoretically. Although they all show similarity in overall pattern of behaviour, they differ in details. Several papers written on the subject deal with the comparison of the functions obtained e varioubyth s workers wit e conclusiohth n thae differenceth t s between them, particularly evidenexperimentae th o t t sourcee e closdu th computationar e o o lt ar , l uncertainties. 185 e strikinOn g featur functionS noto t eA- s thai e th ts e presentepublishear r fa o dds eithe normalises a r unityo dt fro, m sourcee usuallc m th 1 r t non-normalisedya o , . Thus there normalizee ar non-normalized dan functionsS dA- . importann Thia s si t characteristia s i d can contributor to the differences in functions in the vicinity to a source both in values and shape of the curves (see fig.l). Because the numerical differences are not large, not exceeding 5%, they misinterpretee ar d (Glasgow 1981 Thomaso Higgind nan s 1989, Thomaso l 1991a t ne beins )a g dato t e a uncertaintiesdu . This vie wquestiones i thin di s analysis. 1.04 IRIDIUM 'DALE- - - - P 1.02 l- GEL O CO MEISB ^<, 1.00 o 0.98 Kornelson 0.96 0 234 6 RADIAL DISTANCE CM Fig. 1. Attenuation/scatter functions by various authors shown close to the brachytherapy source The tendency to normalize arises from the "hang-ups" in the approach to dosimetry in pase th t whe radiatioe nth n output constants (K-factor, specific gamm constany ara t etc.) used referree b . Dal distanca o t cm o e dt 1 (1983) Monts f eo hi n i e, Carlo work, introducee dth source dose constant concept (SDC for source strength specification which includes scatter component. This indeed involved radial function normalisation but this method of source specificatio dosimetryn i n e founus o dn . This subject has been discussed in several papers in which a comparison of the various available data have been analyzed (Glasgow 1981, Thomason and Higgins 1989, Thomason et l 1991a , Sakellio l 1992a t e u, Parke d Almonan r d 1992). However e normalizeth , d an d non-normalized functions were treate thef i s yda belonge dataf same o th t .eo dse t 186 purpose Th f thieo s analysi consideo t s i s e followinth r g aspect s normalisationi ) si , either at 1 cm or at zero distance, of the A-S function correct and consistent with the interpretatio f physicano l phenomena involved e functiont correci th s f i o ) o useii t y t,s dan normalize dosimetrn i m c I t yd a currentl e baseth n do y recommende kermr dai a calibrationd an , iii) is the Van Kleffens-Star fonction (1979), which differs in pattern from other functions, correc larget a t r distances. 2. PUBLISHED ATTENUATION - SCATTER FUNCTIONS The division of the A-S functions into two groups: non-normalized and normalized will be used as a criterion in the review of the published work. 2.1 Non-Normalized Functions firse th t f investigatioo e On n concerning radium substitute performes swa Meredity db h et al (1966) who studied the attenuationscatter relationship for six brachytherapy isotopes for distances between 2 and 10 cm. The results were presented in numerical form and not normalized. Meisberger at al (1968) used diffusion theory to calculate the ratio of exposure in water to exposure in air as a function of distance (1 to 10 cm) from isotropic point sources of several radionuclides. They then determine meae dth n valueratie th of so fro m experiment severay sb l author tood average san theoreticak e th th f eo mead an l n experimental value derivo st e curves which could be described by an empirical formula. This was the first mathematical function of this kind define found dan d usefu computationar fo l l purposes. Thomason and Higgins (1989) derived a radial dose function defined as the ratio of the measured dos waten techniqueei D r TL calculate e th o )t d (sourcr dosai n ei e activity, exposure constant, f-factor). A fit to a third order polynomial yielded a formula similar to the Melsberger"s r IridiuFo . functioe m th havin m c units i 5 n gvalua t ya f 0.96eo r 0.988o n 4i platinum and steel encapsulation respectively at 1 cm distance. In making a comparison with Meisberger and Dale work the authors concluded that the differences are indistinguishable when considering uncertainties involved. In 1992, Sakellio l carriea t t comprehensivue dou e Monte Carlo calculation f doso s e distribution around seven most popular brachytherapy source sphericar sfo l phantomd an 5 1 f so 20 cm radii. A polynomial expression has been derived for the A-S functions which show a very close agreement with the Meisberger polynomial. The important feature of their results is that the functions have not been normalised. That means that the function for a given isotope is unity at a distance at which the attenuation is compensated by scattering. That depth is a characteristic featur givea f eo n isotope. unfortunates i t I , however, tha derivinn i t e polynomiagth l expression Sakelliod uan colleagues set the first coefficient of the polynomial to unity without paying attention to the physical phenomena occurring in the close vicinity to source. This forces the polynomial to have a value of 1 at r = 0 for all nuclides considered. The Van Kleffens and Star function is discussed in detai section i l . n5 187 2.2 Normalized Functions One of the first Monte Carlo calculations of radial dose distribution was performed by (1979x WebFo severar d )fo b an l gamma-emitters date aTh . were normalizem c unito d1 t t ya Subsequently, they were used by Kornelson and Young (1981) as a basis for formulating an analytical attenuation-scatter functio typa f neo suggeste Evany db s (1955 absorben a r )fo d dose build-up. Thomason et al (1991) calculated radial dose factors (A-S functions) by Monte - Carlo for Ir-192 and Cs-137 normalizing at 1 cm and described them as undistinguishable from the Dale (normalized) and Meisberger (non-normalized) curves within the precision of the data. Melgoon Ravinded an i r Nath 1992 produced radial dose function r severafo s l isotopes by Monte Carlo code CYLTRAN. The functions were tabulated having been normalized at 1 functioe th cm2 solin 19 ni . Thur d I valurisin e r m wateth c s fo f 1.0 s e1 o 1.1o gt 2 t rha 0 a t 5a cm. . WHIC3 H APPROAC CORRECTHS I , NORMALIZE NON-NORMALIZEDR DO ? accepo t Ons eha t tha sensible describinf o th t y ewa strengte gth brachytherapa f ho y determinatioe sourcth y b s ei radiatios it f no n output approacn a , h long establishe externan di l beam therapy. The practice of specifying a source by a quantity defining its radioactive content (i.e. activity or the milligramme radium equivalent etc) is being phased out from use. The quantity recommended currently for the specification of the source is the air kerma rate at reference distance but its exact definition differs slightly by involving the square of the reference distance (CFMRI 1983 ; AAPM 1987) or not involving distance (BCRU 1984; ICRU 1985; NCORD 1991; BIR/IPSM 1993 This particular definitioe aspecth f o t , howeveris n , irrelevant to this discussion. What is relevant is that this type of source specification does not incorporate the scatter radiation. Helpful in this discussion is to consider the dosimetric situation shown in fig 2. This model is designed to focus attention on the issue of attenuation/scatter and disregarding other factors such as source anisotropy, effect of encapsulation, oblique filtration, inverse square law etc. Let's consider a calculation point close to the source the strength of which is expressed in terms of air kerma rate in air (or in vacuo as in BIR/IPSM 1993). e nexAth s te dosimetri th ste n pi c procedure e converte b r kerm ai y e ma ath ,o t d absorbed dose to water assuming that the point of calculation is surrounded by a small water phantom of minimal mass of air. Theoretically, this does not involve scatter production or attenuation betwee poine source th nf calculatioth o d t ean n sinc situatioe eth consideres ni d "free in air". Let's imagin thasourcw e th etno immerses ei wateplacen s di i r o rtissuen di dose Th .e a poin t f interesa o t e altereb t w wildno l being affecte attenuatioy b d n ove e distancth r e concerned. At the same time the scatter radiation will appears throughout the irradiated medium havin a gcompensatin g effec n attenuatioo t t shora n t distances. Over larger distances, attenuation will becom e dominaneth t effect suppressin compensatine gth g effec f scattero t . Where they are in balance, the function will have a value of unity. At any other distance the attenuation/scatter function which is used as a multiplying factor in the dose calculation will have a value different from unity. 188 W . 1 FRE. EIN _ _ „ __ y _ 9 - 1 6 a»" REFERENCE POINF TO SOURCE KERMR I E A AT AR CALCULATION • m c 0 10 T A • —— MINUTE PHANTOM air , v AROUND POINT 2. 'w=Kair(l-g) IN MEDIUM W (>I /P) 3. - g) " " (A-S) air (l^en/P)air Fig. 2. Application of the attenuation/scatter function in a dosimetric system based on air kerma 00 calibration conclusioe Th n from this discussio tha s nattenuationscattei e th t r function whethes ha t ri been derive theory db r measuremenyo e normaliset b mus t no t unito dt purposn yo e on t ea arbitrarily chosen distance, particularly not close to the source where absorption and scatter are balancen i t no . FUNCTIOE TH . 4 N VALU ZERT EA O DISTANCE An interesting issue is also the value of the function at zero radial distance Although this informatio s morha ne theoretical than practical valuimportans i t i e o considet te th r fo r physical correctness of the attenuation-scatter function. Unfortunately, there is no reliable information published on the A-S function value at a distanc . Thi0 understandabls = si e r e considerin experimentae gth l difficultie t distancea s s close to the source. The closest measurements reported so far were made at 1 cm from the source. It is surprising, however, that the Monte Carlo calculations published up to now do not cover this t providno o regiod e d solutionan thio nt s problem theoreticae th n I . l wor Sakellioy kb l a t ue (1992 polynomiae )th functionS A- l virtue s th hav f settiny eo b valuea 0 e g= f unitth e o r t ya first componen polynomiae e questioth Th f . o t1 whethes ni o t l r thi s correcti s o direcN . t answe foune b thin o dt rca s amon publishee gth d data either calculate measuredr do . Some deductions can be, however, made considering a hypothetical dosimetric situation shown in fig 3. This assumes a minute point like source and a point of absorbed dose determination positione closy ve ea vicinityt da . Under such circumstances, attenuation between the source and the calculation point is negligible. The scatter component, however does exist POINT OF CALCULATION AT 0.5 mm DISTANCE SOURCE SCATTER BUILD-UP BACKSCATTER ATTENUATION NEGLIGIBLE OVER SHORT DISTANCE Fig . Factor3 . s affectin imediat e e dosth th gn ei e vicinit brachytherapa f yo y source include forward scatter (large arrows) and backscatter (small arrows), the attenuation being negligible. 190 since the source is surrounded by a medium. There would be some backscattered radiation reaching this area. Thus at a point very close to the source, and indeed at r = 0, there would the unattenuated direct radiation plu backscatteree sth d radiation arriving fro mdirectionr - . Thun so physics grounds attenuation-scattee th , r functio expectee b n ca n hav o dt valuea e larger than unity. One can of course consider this issue also in terms of monte Carlo technique where the calculation model involves volume cylinders aroun sourcee dth . 4.1 Theoretical Experimentald an Evidence. An attempt was made to find support in published data for the above discussion and an experiment was designed to measure the scatter radiation in the immediate vicinity of brachytherapy sources. 4.1.1 Monte Carlo. The computatio Thomasoy nb l 1991a t ne ) provide de scatte th dat n o ar component calculate fractionae th s da l scatter alon radiae gth l distance fro source mth e starting from c m1 onwards. By fitting a polynomial expressions to the Thomason's Iridium and caesium curves (fig e dat r closeth afo ) 4 r distance e derivedb n e fractionaca sTh . l scatter (FS) variation with distance from sourc Iridiur followine efo th founs t mfi wa o dt g expression FSi, =0.0289 + 0.1034 r - 4.732 1 0-Sr2 (1) and for caesium FSc 0.0411+0.059= s 6 r-1.875 10-3r2 (2) Irseatt 0.0 01 234567 DISTANCM EC Fig. 4. Fractional scatter contribution to dose as a function of distance for Iridium and caesium sources (Thomaso l 199a t n1e extrapolate thiy db s cm0 autho = ) r o rt 191 The first coefficients in these equations yield the fractional scatter value at r = 0 which, as it is seenpercen1 Indiun 4. a ,r d caesiud amountfo an mt an 9 2. m o st source s respectively. This clearly indicates that the A-S function should have, at this point in phantom, a value larger than 1.0. This scatter component can be thought of as representing the build-up of scatter in forward direction (+r) due to the backscatter arriving from -r direction (fig 3 towards the origin of the calculation grid which coincides with the source geometric centre. 4.1.2 Experimental Evidence Measurements in the immediate vicinity of the source are very difficult to perform and there is no method of sufficient resolution and accuracy. Nevertheless, an attempt in this work was mad measuro et scatter-attenuatioe eth n function very sourcee closth r thio et Fo s. purposea , miniature ionisation "well" shape chamber was designed (fig 5). The chamber is a cylinder 7 mm in diameter and 15 mm in height with a sensitive volume formed by two concentric graphite SIGNAL ————— LEAD PTFE HT LEAD INSULATION GUARD RING POLARISING ELECTRODE COLLECTOR m 15m ELECTRODE SENSITIVE VOLUME 4 mm 7 mm dia Fig. 5. Schematics of a miniature well ionisation chamber designed to measure the A/S function t closa e distance 192 cylinders (polarising and colecting electrodes) separated by 1.5 mm air space creating the sensitive volume. The brachytherapy source was dropped, during the measurements, into the chamber central channel. The geometric source centre was about 2.75 mm from the center of the chamber sensitive volume enveloping the source. This design facilitated close distance measurement attenuation/scattee th f so r function. A numbe conflictinf ro g experimental parameter reconcilee b o t d sucsha n di hdifficula t design. Because of its small dimensions, the brachytherapy sources of conventional (low) activity measure e coulb t dno d with sufficient resolution i.e. high enough signal-to-noise ratio. The chamber could not be moved with distance, thus only one measurement point was possible obtain constructioe Th . chambere th f no , becaus miniaturs it f eo e siz vers ei y difficuls i d an t expensive. Using this chamber, the measurements of the attenuation-scatter function for an - HDR Indium-192 source were carried out by determination of the ratio of air kerma in water to air kerma in air using a 40 x 40 cm water phantom. The mean value of the A-S function obtained in determination4 2 foun1.01s e b swa o 2dt wit percentagha e standard deviatio meae th f f no no 0.3% . Thus, there is about 1.2 % scatter at the distance of 2.75 mm from the source centre. That is a valuw lo eIfhomaso e basi e compareth th f so n expectee o th n% o d t curve 9 ddifferenc2. e Th . e e explaineb e imperfecn th ca y b dairn "i "t measuring conditions e chambeTh . r body itself produces scatte e environmen th s wel a rs a l whicn i t e measurementh h t were performed. r kerm ai Therefore perfec th n i at e valuno t th e free-in-ais i e r value resultinn a n i g underestimation of the measured A-S factor. Nevertheless, the measured value is a conclusive evidenc scattee th f eo r dominance over attenuatio thit na s distance. 5. THE ATTENUATION-SCATTER FUNCTION AT DEPTH A function of interest is the Van Kleffens and Star (1979 function used in the Selectron HDR treatment planning software Nucletron, Netherlands). The function was published without giving details about the method of its derivation. Initiall functioe yth presentes nwa d withou parametere t produceth i t d an , o dcurv,a e not exceeding unity at distances- close to the source. Since this was subsequently found to be in variance with the Meissberger polynomial, the ô was added lifting the initial part of the curve to t-he value of 1.018. Kleffenn Va e Sta- s Th r function use Iridiur followine dfo th s mha g form; 2 2 ) (3 (A-S) ö(l+ar= r )/(l+ßr ) where for Iridium a = 0.0, ß = 0.0006 cm"2 and 0 = 1.018. The parameters for the original equation were limited to Cs, Co and Ra, for Indium they were added at a later stage. At depth, the slope of the Van Kleffens function differs significantly from all other functions. This is seen in fig 6 for Iridium in comparison with Meisberger and Sakelliou. The shape of the function is sensitive to the value of the arbitrary parameter P which is seen in fig 7. This suggests that this particular function may be an inappropriate algorithm to describe the attenuation/scatter relationship as required in a high doserate Iridium treatment planning system. Thialss oswa note Par y Almond b kan d (1992). 193 Iridium 1.05 IRIDIUM - 192 1.00 0.95 0.90 KLEFFENS 0.85 0.80 z 0.75 g KLEF-NDAk - < 0.70 MEISBERGE 0.65 SAKELLIOU 0.60 0.55 0.50 0.45 J_____L 0 8 10 12 14 16 18 20 22 24 26 RADIAL DISTANCM EC FigCompariso6. .kleffens-Sta Van the nof r function with Meisberge Sakelliorand Iridiuufor m Van Kleffen Stard An sr Function 1/(1+Beta*R2) 9 2 7 2 5 2 3 2 1 2 9 1 7 1 5 1 3 1 1 1 9 Distance From Source Fig. 7. Dependence of the shape of the Van Kleffens curve on the value of the parameter P. . CONCLUSION6 S It is concluded in this analysis that; i) It is incorrect to be comparing the normalised with non-normalised functions asigning the difference to computational or experimental uncertainties ii) Only non-normalised functions such as that derived by Meisberger et al (1968) or Sakelliou et at (1992) should be used in dosimetry systems based on air kerma source specification. However, functions from both studies are uncertain at zero distance, iii) The A-S function, for isotopes such as Iridium and caesium, should not have a value of unity at r = 0 because of the dominance of scatter over attei-iuation close to the source iv) The Van Kleffens-Star function is in error at larger radial distances, stude dat Th e bases yr Iridium-19i th afo n d o discussio e ) v th t 2 bu conclusion d nan e sar relevant to all types of sources. ACKNOWLEDGMENTS Thanks are due to my colleagues, Derek D' Souza and Colin Kerry for assistance in experiments . Fle D f Researc o danr M d h Worksho e patienth r pfo te constructio worth n o ke th f no ionization chamber. BIBLIOGRAPHY AAPM, 1987. Specifications of Brachytherapy Source Strength. Report No. 21. Task Group 32 Americae th f o n Associatio f Physicistno Medicinn i s e (American Institut f Physicseo w Ne , York). BCRU, 1984. Specificatio f brachytherapo n y sources. Memorandum froe Britisth m h Committe Radiation eo n Unit Measurementsd san . Radiol.J . Br ., 941-942 57 , . BIR/rPSM, 1993. Recommendation brachytherapr sfo y dosimetry. Repor Workina f o t g Party. Brit. Inst. Radiol. Londo. nUK CFMRI, 1983. Recommendations pou determinatioa l r s dosende s absorbée curietherapien se . Rappor Comitu d t é Français "Mesur Rayonnements ede s Ionisants (Burea1 o "N u Nationae d l Métrologie, Paris). Dale R G, 1982. Some theoretical derivations relating to the tissue dosimetry of brachytherapy nuclides, with particular referenc Iodine-125o et . Med. Phys (2)0 1 ., 176-183. Dale R G, 1986. Revisions to radial dose function data for 192Ir and 137Cs. Med. Phys. 13(6), p 963. Evan , 1955 D atomi e sR .Th c nucleus. (McGraw-Hill Yorkw 732-734,Ne p )p . Glasgow G P, 1981. The ratio of the dose in water to exposure in air for a point sorce of Ir 192: A review. In; "Recent Advances in Brachytherapy Physics". Editor D R Shearer AAPM Monograph no 7. p 104-109. 1% Kornelson R 0 and Young ME, 1981. Brachytherapy build-up factors. Birt. Journ. Rad. 54, 136. Masterson M E Simpson L D and Mohan R, 1979. Med. Phys. 49 Abstract. NCORD, 1991. Recommendation dosimetrr sfo qualitd yan y contro Almond Parkef an o l C drH , 1992R P . Evaluatio build-ue th f no p Ir-19n effeca f o 2t high dose-rate brachytherapy source. Med. Phys , 1293-129719 . . MeigoonNatd . an 1992h R S A i. Tissue inhomogeneity correctio bracgytherapr nfo y source Med. Phys , 401-40719 . . Meisberger L L, Keller J R and Shalek R J, 1968. The Effective Attenuation in Water of the Gamma Rays of Gold 198, Iridium 192, Cesium 137, Radium 226 and Cobalt 60. Radiology, 953-957p , 5 . No . Vol0 9 . Meredith W J, Green D and Kawashima K, 1966. The Attenuation and scattering in a phantom of gamma rays from some radionuclides used in mould and interstitial gamma-ray therapy. Br. J. Radiol. , 280-28639 , . Nath R. Meigooni Ali S. and Meli J, 1990. Dosimetry on transversal axes of 1-125 and 1-192 interstitial brachytherapy sources. Med. Phys. 17, 1032 - 1040. Ravinde , MeigoonJeromd rN , 1990an M S eA A .i Dosimetr transversn yo e axe 125f d so Ian Ir interstitial brachytherapy sources. Med. Phys. 17(6), 1032-1040. 192 Sakelliou L, Sakellariou K, Sarigiannis K, Angelopoulos A, Penis A and Zarris G, 1992. Dose rate distributions around Co., Cs, Au, Ir, Am, I (models 67002 and 6711) 60 137 198 I92 241 125 brachytherapy sourcenuclide th d sean "Te". Phys , 1859-187210 . . MedNo ., Biol.37 . l Vo , Thomason C and Higgins P, 1989. Radial dose distribution of Ir and Cs seed sources Med. 192 137 Phys. 16(2) 254-257. Thomaso , MackinC , Lindstro R Higgind T , ean 1991 D J dose P sm M .Th e distribution surrounding 192Ir and 137Cs seed sources. Phys. Med. Biol. Vol 36, No. 4. 475-493. Thomason C, Mackie T R, and Lindstrom M J, 1991. Effect of source encapsulation on energy spectr 192f l37ad o ICran s seed sources. Phys 495-505, 4 . . MedNo , .. 36 Biol l Vo . Van Kleffens and W.M Star, 1979. Application of stereo x-ray photogramometry in the determination of absorbed dose values during intracavitary radiation therapy. Int. J. Radiât. Oncol. Biol. Phys , 557-5635 . . Dose , 1979A wateTh n ei .R x r Fo surroundin Webd an bS g point isotropic ga=,a-ray emitters. Br.i. Radiol. , 482-48452 , . Iy107/ STANDARDIZATION OF IRIDIUM-192 COILED SOURCE IN TERMS OF AIR KERMA OUTPUT A SHANTA . UNNIKRISHNANK , , U.B. TRIPATffl, A. KANNAN, P.S. IYER Bhabha Atomic Reserach. Center Trombay, Bombay, India Presented rahmeB A. by Abstract ICRU (1985) recommended that the output of gamma ray brachytherapy sources should specifiee b termn di referencf so kermr eai a ratea t , a define r kerme ai th n s i d a r ratai o et reference distance of 1 meter, perpendicular to the long axis of the source, corrected for air attenuation and scattering. As these measurements are difficult to carry out in the routine clinica commoe lth uses i t i , n practic calibrato et re-entrane eth t ionization chamber with respect to open air measurements and use the re-entrant chamber for routine measurements. This paper e reportmeasurementth n o s s o correlatcarriet t ou de th e nominal r kermactivitai d a yan rat f o e Ir wire sources suppliee Boarth f y o db d 192 Radiation and Isotope Technology, Department of Atomic Energy. Introduction : One of the major factors which contributes towards dosimetric accuracy in brachytherapy is assessment of source strength used. According to ICRU Report 38 (1985), the specification of gamma ray brachytherapy sources should be in terms of reference air kerma rate , defined as kerma rate to air, in air at a reference distance of one meter from centee th f sourco r perpendiculad ean lone th gsourcee o t r axith lon e f so Th g. distance measurement geometry minimize dependence sth calibratioe th f eo n upo constructioe nth n e sourc th detector d f o an e e considereb s bota n , hca s pointa d d effec an sf obliqu o t e transmissio f gammo n a rays through source sheathing become negligible e lonth gt Bu . distance measurements under scatter free conditions are difficult to carry out in routine practice, especially with low activity sources. Hence it is a common practice to establish a calibration wel e factoth lr chambefo r r with respec opeo t measurementr nai e th e us d san well chamber along wit referencha e standar r routindfo e calibrations [1,2]. Undee th r EUROMET framework a progra, o confirt f wormo s L e initiate mth wa kNP t a d traceability to NPL secondary standard radionuclide calibrator of air kerma rate measurements made by Amersham International for wire sources of 192Ir [3]. Indian I , radiation source suppliee Boarsar e th Radiatiof y ddo b Isotopd nan e Technology (BRIT), Department of Atomic Energy. 192Ir wire sources used for interstitial therapy is supplied to hospitals in the form of cylindrical coils, the nominal activity of which is measure a re-entran n di t 4rc gamma chamber t int coie cu o s i Th requirel. d lengthe th y sb users. This paper deals with the measurements carried out to correlate the nominal activity of coiled source with reference air kerma rate measured for coiled as well as linear form of sources. 199 Material Methodd san s Ir wire sources supplie BRIy db T consist iridium-platinuf so mplatinu% cor 5 e(7 m& 192 25 % indium ) of 0.1 mm diameter with 0.1 mm thick platinum coating, thus making an coiles i activate d m an dc 0 overaly 10 dwirb e r lengtho n Th e i . 0 l5 diametef mm so 3 0. f ro thermal neutron irradiation in a reactor. The nominal activity of the coil is then measured in a calibrated well type ionization chamber and supplied to the users. Measurements r sourcI GBq(107 A3. n f leo Q ^ 0 mCi) produce exposurn a s e rat f aboueo t " 3.2210 5* Q A/Kg (12.5 uR/sec) at one meter. The current per unit volume works out to be about 4.17 * 10-.-1" 5 A/cm . To measure such low currents, a 400 c.c. ionization chamber coupled to a varactor diode amplifier with a calibrated capacitor in the feed back was used. The 400 c.c. bakélite chamber was calibrated against a spherical graphite chamber whose accuracy kermr oai f a rat measurementee th determinatio d an % traceabl e 2 sar 60 e s nCi th o o et therapy level primary standard whic intercompares hi d against international standards. Iridium wire cut into small pieces of 1.5 cm length, irradiated to an activity of 222 MBq/cm (6 mCi/cm) was procured from BRIT for the calibration of 400 cc chamber. The sources were arranged in a matrix of size 1.5 cm x 2.0 cm and aligned in level with the cente sphericaf ro l graphite chamber apar scattea m c n i t5 6 , r free geometry source Th . e- detector alignmen s verifiewa t d usin ga lase r beam e chambeTh . s connectewa r o t d varactor amplifie witp u hcalibratea t se r d capacito feedbacke th n i routpue Th . t voltage (V), ove timra et seconds measures wa , corrected dan backgrounr dfo d radiation, charge leakage, temperature and pressure. Corrections have also been applied for wall attenuation (K at1.066= t , )stoppin g power rati graphitf oo (S/r ai 1.015)= p o et . KCEd Pan wer r energfo e 0 assume1. y e correspondinb o dt r gammI 9 1 o gt a rays. Curren r unipe tt volum s evaluateewa correlated dan kermr ai o dat rategraphite Th . e chambe s thewa rn replaced with 400 ce bakélite chamber and measurement was repeated as before. After applying necessary corrections for background radiation, charge leakage, temperature and pressure, curren unir pe tt volum evaluatedsets o readingf estw wa o e ratie s th f oTh .swa o calibratiotakee th chambes c na c 0 n40 rfactoe th f o r The calibrate chambec c 0 thes d40 rnwa use standardizo dt e eIth r coils actually usen di 192 clinical practice. Two sets of measurements were carried out, one with the source in the coil form s supplieothee a , th usere rd th aftesan o dt r cutting into linear form uses a , n di clinical practice. 192Ir coiled source and the chamber were aligned in a scatter free geometry. The source chamber distances were kept as 75 cm and/or 100 cm. Measurements were carried out as before and after applying the necessary corrections discussed earlier, current was calculated and using the calibration factor of the chamber, air kerma rate was evaluated. Measurements were carried out for five different coils, 1 of 100 cm and 4 of 50 cm each. Correction for decay of source activity during the course of measuremen s appliewa t d assumin ghala f lif73.8f eo 3 days. Referenc r kermeai a ratr epe unit lengt evaluates hwa s uGy.h'Vm^cmda . eacm c h0 wer5 f o othee o theth t intrtw d ncu o an Threm c f thes0 eo 10 e f coilso e on , linear sources, . Channelvaryincm 0 lengtn 1 gi so t h wer m froc e m m6 c drille 1 t a d interval perspea n si nyloxe sheeth d n tubinan t g use implantatior dfo fixes nwa d into these. 200 The perspex sheet and the 400 cm ionization chamber were aligned in a geometry 3 identical to coil source measurement. Ir wires loaded in inner nylon tubings were then 192 inserted inte outeoth r nylon tubing e perspes th fixe n o dx sheet r kermAi . a s ratewa measure befores da accouno possible T . th r tfo e los f smalso l bit wirf so e while cuttinge th , actual lengt f wiro h e user measuremenfo d s determinewa t d from auto d X-raan - y radiographs Autoradiograph also helped to ensure the uniformity of activity. Reference air kerma rate per unit length was evaluated and correlated to that measured for coiled source. Discussion Dose computatio r lineanfo r sources require specificatio f sourco n e strengt termn hi f o s reference air kerma rate(RAKR) constant (u.Gy.h"'.m2.MBq'1). This was calculated from measuree th d referenc kermr eai a rate usin quotee gth d nominal activit source th d f yo ean assumin seriea coie s g th a l f ring so diametef so space m c 2 t equada 1. r l interval coir sfo l form and using Sievertfs line source dose function, for linear form. The values thus obtained for sources in coil and linear forms are given in Table - 1. The correction factor to be used for linear source output is given as ratio of RAKR constants in the last column of the Table. Table -1 Measured Output Correction Facto 192r rfo I r Coils Source Chamber RAKR Constant Rati RAKf oo R Distance (cm) (u.Gy.h~1.m2.MBq"1) Constant Coil Linear Linear/Coil 75.0 0.1025 ±0.0013 0.1 124 ±0.0019 1.101 100.0 0 1035 ± 0.0009 ±0.0020.6 115 1 1.118 It may be seen that the RAKR constant for linear form is significantly higher than that of coiled form. This coul attributee db higheo dt r inherent self shieldin r coil gfo clinican I . l practice , where long wires are used, the reference air kerma rate constant measured for linear form shoul usee dr b dos dfo e computations e publisheth s A . d value f RAKo s R constan r I92fo It r sources show large variations e measurementth , s t carriea t ou d respective centers shoul consideree db appropriates da . REFERENCES 1. T.P.Loftus, Standardization of Iridium-192 Gamma-Ray Sources in Terms of Exposure Researcf o . NBSJ .e th f ho , 85:19-25,1980. 2. A. Shanta, Dosimetry of Ir - 192 Wire Sources - Comparison of Theoretical and Experimental Values. Endocurietherapy/Hyperthermia Oncology, 7:2 , 1992733 . . 3 J.P. Sephton . WoodsJ M ,J Rossiter M. , . WilliamsT , , J.C.J. Dean. G.A. Basd san S.E. M.Lucas, Calibration of NPL Secondary Standard Radionuclide Calibrator for 192Ir Brachytherapy Sources. Phys.Med.Biol. :11578 3 , , 1993-64 . CALIBRATIO r fflGf m HF N O DOS E RATE BRACHYTHERAPY SOURCES1 M.H. MARÉCHAL Illllllllllllillllllillllllilllllllllllll mstitut Racüoprotecäe od Dozunetnoe a ™xSw2861 ALMEIDE D C.E. A Laboratôri Ciénciae od s Radiolôgicas, UERL Rio de Janeiro, Brazil SIBATH C. A Roswell Park Cancer Institute, Buffalo, New York, USA Abstract A method for calibration of high dose rate sources used in afterloading brachytherapy system describeds si calibratioe Th . n factor fo determineIrs i r interpolatiny db g Co gamma- rays and 250 kV x-rays calibratio 192 n factors. All measurement 60 s were done using the same build up caps as described by Goetsch et al and recommended by AAPM. The attenuation correction factors were determined to be 0.9903, 0.9928 and 0.9993 for Ir, Co and 250 kV x-ray, 192 60 respectively. A wall + cap thickness of 0.421 g.cnr2 is recommended for all measurements to 192 ensure electronic equilibrium foCr60 o anIrd gamma-ray beams. A mathematical formalism is described for determination of 1 INTRODUCTION High Dose Rate (HDR) afterloading systems using 192Ir source receivine sar g considerable attention throughou economicae worln a th t s a d safd an el optio r brachytherapnfo y specially where the patient load is high. At the present time, there are 8 units in operation in Brazil and 4 othe expectee rar neae th r n dfuturei . The source manufacturer specifies its activity with an accuracy of ±10% which is unacceptabl clinicar efo l purpose withoud san t traceabilit nationaa o yt l standard absence th n I . e of a standard for 192Ir calibration, the American Association of Physicists in Medicine (AAPM) proposes the use of an interpolation procedure [1,2] using the average calibration of a 0.6 cc ionization chamber obtained with 137Cs gamma-rays and 250 kV x-rays (HVL=3.2 mm of Cu, effective energy of 146 keV x-rays). A graphite cap thickness of 0.31 g.cnr has been recommended for all measurements. 2 An alternative method using a well-type ionization chamber [3] has been carefully analyzed and compared with the AAPM method and an excellent agreement was found. However calibratios it , n relie thimbla n so e ionization chamber calibrater dfo C s gamma-rays. Sinc majorite eSecondarth e th f yo y Standard Dosimetry Laboratory's (SSDLt equippeno 137e ar ) d withC sa 137 therapy source, the users will have to send the ionization chambers abroad for calibration. This paper proposes a calibration procedure to derive the Nx for a Farmer type ionization chambe interpolatioIry b , 192 r rfo n fro ^Ca m o gamma-rayx-rayV k 0 s 25 (HVL m d m san 5 =2. , effectivCu f o ex-ray V energke scalibratio1 ) y13 n factors [4]. 'This pape dedicates ri memore th do t Eugenif yo Cecatti. oR . 203 calibratioe Th n factors were determined usin same gth measuremente build-uth r fo p pca s with ^Co gamma-rays , 250 kV x-rays and Ir with the appropriated wall attenuation factors 192 taking into account. 2 METHODS . WalA l attenuation measurements The measurements were made using a NE Farmer thimble type chamber model 1975, with a 0.065 g.cnr graphite wall. Six build up caps were made of PMMA with thickness varying from 0.19 o 0.494t 2 6 g.cm- e thicknesTh . s varie wa sy addin b d g additional buildup caps. Measurements for 60Co 2gamma-rays and 250 kV x-rays (HVL = 2.5 mm Cu, effective energy = 131 keV) were done at the Laboratorio Nacional de Metrologia das RadiacOes lonizantes (SSDL/CNEN). A 10x10 cm field size and source to chamber distance of 100 cm for the Co 2 60 gamma-ray beam, and a 7 cm diameter circular field and 75 cm distance for the 250 kV x-ray beam were used192e ITh r . measurements were Hospitamade th t Vicentea o Sâ l Pauloe ed r Fo . the 192Ir source measurements Nucletroe th , n calibratio s usewa d g wit nji e e cente th hth f o r chamber positione fro m source c m th 0 1 et d a [5] . Figur showe1 attenuatioe sth n curvel al r sfo three energies. A combined (PMMA+graphite) buildup cap of 0.421 g.cnr is sufficient for electronic equilibriu 60m 192n i Cd Ioan r gamma-ray beams. 2 B. 192Ir Calibration Factor 192 The Ir calibratio obtaines i n r I factorinterpolatin) y db x (N , ^Ce gth o gamma-ray calibration factor and the 250 kV x-rays as originally proposed by Ezzel [2] and following the more rigorous proposal by Goetsch et al [4]. A combined build cap of 0.421 g.cnr2 was used for calibratio l threal n eni beamsinterpolate w s A . calibratioe eth n facto necessar192s r i fo It ri r o yt use the buildup cap and take into account its attenuation in the calibration process. The attenuation factors were determined fro slopee attenuatioe mth th f so n curve e Figurn si Th . 1 e correction factor calculates A^wa , y db - (slope) 1 Aw= (wall• thickness) The 192Ir calibration factor is determined by interpolating the ^Co gamma-rays and 250 kV x-rays calibration factors. Assuming that the calibration factor is linear with energy, it can be written that: where, kx.ray ond kc0 are interpolation factors given by: c kx.ray = /*_"/ * = 0-7989 Co X-ray and _ Elr-Ex_ray _ -F F Co~ •*-T1« *-* Y-~™> 204 1.02 0.98 4 0. 3 0. 2 0. 1 0. 0.0 0.5 0.6 Wall Thickness (g. cm'2) Figure 1. Attenuation curves for 250 kV (131 keV x-rays), 60Co and 192Ir gamma-rays in lucite caps coverin gFarmea r ionization chamber slopee least-squarTh .e sar e fit measureo st d data with differen thicknesp ca t s added. 192T Ejn Ex- ECd rexposure-weightee aOan yar d average gamma-rar I energ x-raV k d r y0 an fo 25 y, Co gamma-ray beams, respectively. (Nx )ithes ri n givey nb 'Co RESULTS The attenuation coefficients were found to be 0.023, 0.017 and 0.0016 cm^gr1 for 192Ir, and 250 kV x-rays, respectively. As one observes from Figure 1 a thickness of 0.28 g.cnr is 2 enough for electronic equilibrium inI ra gamma-ra 192 y but not for ^Co gamma-ray beam. A thickness of 0.421 g.cnr2 is necessary for 60Co gamma-ray beams, the attenuation correction factors bein than i g t case 0.9903, 0.992 d 0.999 8an x-ray V r k 192 fo 30 Irs ,25 60d Can o respectively. The 192Ir calibration factor is given by: ,),, = 0.8055-(ATr);r_ +0.1997 'Co The total thickness for buildup cap is 0.421 g.cnr . If one subtracts the graphite wall 2 thickness of the ionization chamber, a 0.356 g.cnr thick PMMA cap is needed. The combined 2 uncertainty associated to the (NjJx-ray and (Nx)co and the experimental procedures results in 0.4% overall uncertaint 192e th I rr calibratioyfo n factor, assuming thaionizatioe th t n chamber response is linear with energy in that region. 205 4 CONCLUSIONS A metho r determinindfo 192e gth I r calibration factothimbla r fo r e ionization chambes i r described using interpolation from a 60Co gamma-ray and 250 kV x-ray beams calibration factors. The method is valid for ionization chambers that responds linearly with energy. To use this method at other SSDL only the attenuation factor for the build up cap in the 250 kV x-ray beam have to be determined because it is the only beam energy that might change. attenuatioe Th 60n e I92 d C factorbuilduth e Ioan rn th i gamma-rar p sfo pca y beams determinen di this work can then be used for the derivation of the 192Ir calibration factor. Since the attenuation x-raV k y0 bea25 e msmalfactos i th r l fo r(les s than 0.1%) even this factor coul usee db d without compromising the accuracy of the method. A 0.421 g.cnr2 combined thickness build cap was used to calibrate the thimble ionization chamber. Since users normally hav ^Cea o gamma-rae futurye th wilbuilduth e e w n us ei l p pca Co buildup cap thickness (0.551 g.cnr ) for this method. Future plans include direct comparison wit60 h a Farmer ionization chamber with2 PMMA wall to minimize wall artifacts. This ionization chamber will be calibrated foCr sa 137 gamma-ray beam to allow direct comparison between this method and the AAPM proposed procedure. ACKNOWLEDGEMENTS woule W d lik thano et k Mrs. Maria Armanda Pinto Abrantes, medical physicist from Hospitao Sä l Vicente de Paulo and Mrs. Luzianete Amaral from SSDL/CNEN for their help with measurement theit sa r facilities. REFERENCES [1]. AAPM (American Association of Physicists in Medicine) Task Group 41. Remote Afterloading Technology. American Institut f Physicseo Eas5 Street5 33 ,4 t York w Ne , , USA, 1993. [2]. EZZEL G.A., Evaluation of Calibration Techniques for the MicroSelectron HDR, in Brachytherapy. Proceeding Internationae th f so l Selectron Users Meeting, 1989. [3]. SIBATA C.H., de ALMEIDA C.E. and SHIN K.H., Calibration of HDR and LDR brachytherapy sources usin a welg l type ionization chamber. AAPM Annual Meeting, Washington, DC 1992. [4]. GOETSCH S., ATTIX F.H., PEARSON D.W. and THOMADSEN B.R., Calibration of 192Ir high-dose rate afterloading systems Physd 462, .Me 18 , , 1991. [5]. FLYN A. and WORKMAN G., Calibration of a MicroSelectron HDR Ir source. Brit. J. 192 Radiol. , 73464 , , 1991. 206 QUALITY CONTRO IM92F LO , Cs-13 Ra-22D 7AN 6 SOURCES BRACHYTHERAPN I FOE RUS Y C.H. OYRZÜN CORTES, A.M. PALMA D., . PENALOZH . TOMICIM , AC. C Comisiôn Chileria de Energia Nuclear, Santiago, Chile Abstract In orde establiso t r a hcertai nd managin degrean e f reliabilitus o ef o ge th n i y radioactive materia n brachytherapyi l a minima, l quality contro o eact l h sourcs wa e implemented. The purpose was to estimate the degree of radioactive leak, resistance to the mechanic tractio stresd nan usef so . Throug physicaha l contro radioactive th f o l e materiaa ( l simple dip test and using a photon scanner or auto-radiography) the minimal conditions that guarante establishede ear safe eus . This informatio transmittes ni calibratioa o dt n laboratory for certificatio n exposuri n e rate and/or activity. Systemati f theso e e us ctests , enables discover radioactivf yo e material leakag sealfaulte o t th e .n si e du 1 INTRODUCTION During the last 10 years, sealed radioactive material has been used for radiotherapy especiall brachytherapyn yi sucn I . h practice, radioactive material with long half lives sucs ha Cs-137, Ra-226 have been use somr dfo e decades, lately short half life nuclides, particularly Ir-192, came into clinical use. Such sourcereactivatee b n sca nuclean di r reactors eithen I . r case radioactive th , e material mus purpose sealede th b t r f radioprotectioeFo o . f patientno s and personnel involved in the treatment, these seals must be tested. The radioactive material, normally Ra-226, Cs-137 or Ir-192 is in the form of spheres or cylinders and clustered inside a covering, sealed and mechanically labelled. This cylindrical covering is of Stainless Steel, Platinum and/or Platinum-Indium. For radiotherapy objectives a uniform dose distribution around a implant or therapy applicato primara s i r y goal. Thus radioactive th , e material mus uniformle b t y distributen di the isotope source. Simultaneousl e uniformit th ye sea s th cruciai l f o yo acquirt l a e symmetrical dose distribution outside the capsule (axial symmetry). This enables less complicated calculation f doso s e distribution e surroundinth o t s g tissues. Also froe mth radiation protection point of view the control of the seal in terms of leakage and symmetry is important. Nevertheless radioactive leaks from such sources have been detected lossee Th . s are due to mechanical loss of the covering material. A control applied on the covering would enable a determination of the degree of damage. The most meaningful tests turn out to be determination of; 1. Losses of radioactive material, degree Th radioactivf e. o 2 e homogeneity, inherene Th . 3 t resistanc sealine th f eo g materia physical-chemicao lt l coveringstrese th f so . A wholea s , this serie f testso s determin sourcf e shouli th r f o e s i suitabl ei d b e us o et considered as radioactive waste. 207 2 METHOD Loss of Radioactive Material determinatioe Th speee th whict df a n o h radioactive materia loss li t and/or radioactive contamination occurs accomplishes i , d through either gamma and/or beta spectrometrye Th . determination of the radioactive contamination, it is realised through chemical pureness tests in a base solution, and Gamma spectrometry and/or Beta emission analysis of the same solution. base Th e solutioHydroge Ethano% f o mixtur50 a % d s ni 50 an l nf eo Peroxide e Th . radioactive material is submerged in the base solution for a period not less than 76 hours at room temperature. After this time interval adjusted according to a radiological procedure previously established, the radioactive source is removed. The radioactive contaminants in the solutio thee nar n analysed. 226- a case R I, n f th Cs-13e o Ir-192d 7 an radioactiv e th , e contamination analysis si based on Gamma spectrometry using a high efficiency detector. As detector, a Nal(Tl) - crysta connecte s useds wa lwa t I . d throug amplifien ha pulsa d eran height analyser (PHAo )t a multi channel analyser with 4096 channels. Repeating the above sequence permits determination of external contamination due to: 1. Diffusion thoug coveringe hth . 2. Interaction with external radioactive materials, 3. Micro-fissures or structural damages of the covering. A rulesa estimate th , radioactivf eo e losse detectio e bases si th n do n limit eacr sfo h isotope, and the signal/noise ratio of the spectrometry system. Degree of Radioactive Homogeneity The degree of radioactive homogeneity, is determined through two complementary methods; 1. A longitudinal sweep with a semiconductor micro-detector, connected to a gamma spectrometry system. 2. Utilisation of radiographie plates to obtain autoradiographs. longitudinae Th Si(Lia f o l) e sweemini-detectous e basen pth i n do r mountea n do table that permits precision movements detectoe Th . moves ri d alon source gth measured ean s the event rate of each section of the radioactive material. The autoradiography is based on the use of radiographie plates put in a mammography cassette of the type used in the clinic. If the autoradiography shows evidence of low filtration zones through the covering, a contact autoradiograph is made in order to detect any possible bet Alfr ao a emission throug seae hth l (dependin radionuclide)e th f go . criterie Th a use determino dt lose coverinf eth s o g material bases i ,e studa th n df o y o average thicknes covee th f rso compared wit possible hth e damage zoneobvious i t I . s thae tth visual inspection of the material through an optical system is adequate. 208 Physical Chemical Resistance capacity physical-chemicae Th l resistance capacit determines yi orden di ensuro rt source eth e to overcome extrem violend ean t chemical reactions. Normall same yth e solutio thas na t used for determination of radioactive leak is used. The capacity of sustaining extreme mechanics actions such as beats and torsion was tested. It was shown that coverings of Platinum-Indium do not sustain more than five re- activating with neutron flows in a 5 MW reactor, conditioned by mechanical effort accomplished through metallic tweezers without protection coveringn I . f stainlesso s steels a , Ra-226e case th f havinth d e o s i an , gnormaa l manipulation history, surface damagee ar s show year5 1 n usef o so aftet .0 1 r 3 RESULTS The solid sources of Ra-226 with more than 10 years of use, show mechanical surface damages in about 15 and 30% of the total surface of the covering. This gives radioactive leaks 100o t 0 0 rang e 3 intimeth f detectioe o sth n limits recommendes i t I .Ra-22 e us o t 6 t dno for intracavitary implants in Chile. From a survey of 1000 Cs-137 needles surveyed, only 10 needles presented surface damage handlino t e sdu g with metallic barbed objects. Ir-19e Inth 2 wires activate firse th t y timedb 0,3a , % percent lac f radioactivko e material was shown. Wires reactivated three or four times present transverse fractures in the points of manipulation with metallic barbed objets. Ir-192 wires reactivated a third time are t use Chileno n dI . 4 CONCLUSIONS applicatioe Th minimaa f no l quality contro testiny lb radioactive gth e material usen di intracavitary implant, permits: a) To define and to guarantee the necessary technical specifications for the calculatio Dosimetre th f implantne o th f yo . b) To avoid radioactive contamination of the medical staff and the patient. maintaio T conditione n th ) accountinc e th f so storagd e gan th f eo same. An annual monitoring is recommendable. It can simultaneously accomplish a minimal quality control, base radioactiva n do e cleanlines coverine autoradiographyn th a f sd o gan , both accomplished through a clearly detailed radiological procedure. 209 IV. QUALITY ASSURANCE NETWOR RADIOTHERAPKN I Y QUALITY ASSURANCE NETWORK: THE EUROPEAN PILOT STUDY J.CHAVAUDRA 1111111111111111111 Servic Physiquee ed , XA9642863 Institut Gustave Roussa, Villejuif, France A. DUTREIX University Hospital, Leuven, Belgium S. DERREUMAUX . BRIDEA , R Servic Physiquee ed , Institut Gustave Roussa, Villejuif, France SCHUERER DE N EVA . N University Hospital, Leuven, Belgium Abstract Based on the IAEA/WHO experience in mailed dosimetry, a Quality Assurance (QA) Network, sponsored by the EC committee"Europe against Cancer", has been set up in 1991 for all European centres. Beside survesa f radiotherapyo y infrastructure projece th , t includes three measurement steps : primarily, a check of beam output and quality in reference conditions with mailea d TLD-procedure secona n i , d step mailee ,th d verificatio othef no r beam datdosd aan e calculation procedures with a multipurpose phantom, and finally in vivo dosimetry at the individual patient levels with mailed dosimeters. resulte Th s concernin beam2 g16 s fro centre5 m8 analysee sar 60Q8 d(5 X-ra)4 beam10 yd san beams). 27 beams present minor deviations (3 to 6 %) and 15 beams (4/58 ^Co beams and 11/104 X-ray beams) fro centre1 m1 s present majo e analysir deviationTh . %) s show6 > s( s that 17/27 minor deviation l majoal d rs an deviations have been detecte centren di s which have not benefited fro externan ma l check durin larglase 5 g1 th t f efiv o deviationst e ou years 4 1 n ;i , the measured dos smalles ei r tha statee nth d dose mosn I . t centres with major deviatione th , physicist havt necessare no eth d sdi y experienccalibratt no d di ed regularlean beamse yth 6 n I . dosimeteo n ther 1 s dosimetee 1 centre ef th wa o r t ro sou r availabl t beeno ns calibrateeha d recently centres3 n I . physicise t giv th ,explanationy no ean d di t conclusione Th . s concerning the second step (multipurpose phantom), outline the larger magnitude of the deviations for off axis points, oblique surface and the^use of wedge filters. 1. INTRODUCTION The Interest of Quality Assurance in Radiotherapy (medical and physics aspects) has been stresse decader dfo s wit numbeha f dedicatero d Meetings, papers, report protocolsd san . Nevertheless 198e ,th 9 world-wide investigation sponsore Internationae th y db l Atomic Energy Agency (IAEA) and the World Health Organization (WHO) revealed that about 15 % of all cancer patients treated with radiation receiv inadequatn ea esystematio t dos e edu c uncertaintie dose th en si deliver y (1). This percentag probabls ei y even much importana n i t number of Radiotherapy Departments not implementing Quality Assurance (QA) programmes. 213 In addition, it is well established that randomized clinical trials can improve the practice of Radiotherapy provided that suitable Quality Assurance Programme appliee sar ensuro dt e eth compliance of the treatments with the treatment regimes (2, 3). Due to the need for a large number of patients in order to improve the significance of the conclusions, to allow the evaluation time to remain reasonable, more and more multi-institution trials are launched, the main requirement being that any patient of any institution be treated according to the same protocol and receive the same dose. Consequently, in addition to the fact that both internal QA and external audits have prove, eve vere 7) b , ny6 o nt whe, usefu5 , n (4 lRadiotherap y Departments implement good internal Quality Assurance Programmes coherence th , e mus e ensureb t d betweee th n Radiotherapy Departments at larger scales, through Q.A. networks. At the present time in Europe, a few large centres have developped QA Programmes and have show benefite nth theif so r implementatio Countrienw , 12) (8,9Fe 11 ., A ,10 s have developped programme f Nationaso l Dosimetry Intercomparisons, often onl referencn yi e condition comprehensivo , 17)n t 16 ,bu , s15 (13 , 14 , e system comparabl CenterS U e th s o et for Radiological Physic availabls i Europeae th t ea n scale (18- 18-2)d 1an . At the International level, a few international bodies have developed mailed QA Programmes particularlyn I . , since 1967, IAEA/WHO have developpe dpostaa l dosimetry service (1,7, 19, 20) for developing countries, restricted to measurements at a reference point for ^Co radiation. This service has recently been extended to high energy X-ray beams with advancethw efe hela f po d centres. Since 1988 Europeae ,th n Organizatio Res,earcr nfo d han Treatment of Cancer (EORTC) has also developed a mailed service as a part of larger QA programmes to check beam outputs in reference conditions for the European Centres participatin clinican gi l trials (21,22) comparablA . purposn eo e servic alss eowa developey db the Institut Gustave-Roussy for a clinical trial on the use of the Etanidazole Radiosensitizer, including about 30 european centres, with a postal dosimetry service (reference conditions) associated wit purposn ho sitn eo e visit patiend san t treatment forms review (23). So, in Europe, only a few national or international dosimetry intercomparisons and on purpose action r clinicasfo l trial beed sha n performed whenrequese th numbe a t a f , o t f o r centre t involvesno clinican di l research attempn a , mads wa textrapolat o et expertise eth e acquire thesn di e firs standardizea t p studieu t se proceduro sA t dQ e which same coulth n e i d b all European Community (EC) countrie whicd san h would provide guideline d technicasan l back-u radiotherapl al o pt y Centres. The principles of this attempt are that the practical responsabilities would be given to national bodies as soon as procedures are established. An European coordinating and advisory function would still be maintained in order to ensure EC guidelines to remain coherent. Ai last, five radiotherapy centres from five European Countries (Belgian, France, Italy. e NetherlandTh Swedend san ) succède when proposin CommitteC E e gth e "Europe against cancer" to support the project of a European Network for Quality Assurance in Radiotherapy. . DEVELOPMEN2 NETWORE TH F TO K The Experimental European Network was implemented in 1991, with the following Structure : Coordinating Cemtre University Hospita Rafaëlt S l , Leuven. Belgium. (CC) Measuring Centre Institut Gustave-Roussy, Villejuif, France. (MC) 214 National Reference Centres (RC) Belgium Universit: y Hospita Rafaëlt S l , Leuven France : Centre Georges François Leclerc, Dijon Ital yUniversit: a degli Stud Firenzei id , Florence The Netherland Leeuwenhoekhuisn Va . A s: , Amsterdam. Sweden, University Hospital, Umea. 3. IMPLEMENTATION OF THE NETWORK ACTIONS (24) 3.1. Dosimetry checks They includ epreliminara y investigatio available th f no e infrastructur centere th d n ei s an three measurement steps : 3.7.7. Preliminary investigation initiae Inth l stage, befor measuremeny ean performeds ti questionnaira , e senth s e i y b t National Reference Centre to the local centres, regarding staff, radiotherapy and dosimetry equipment, simulation and treatment planning systems, in order to be able to find out possible correlations betwee e radiotherapth n y department structure e uncertaintiesth d an s e Th . questionnair identicas ei l countrieal r fo l orden si obtaio rt nstandardisea d data bases i t I . translated int locae oth l languag Nationae th y eb l Reference Centre. 3.7.2. First step : dosimetry check n referencei conditions A firsa t t step, beginnin outpue th bea d 1992n gi an t, m qualit f ^Co-rayo y beamd san high energy X-ray beams were checked in reference conditions with mailed TL Dosimeters placed in a small plastic holder, which were irradiated in a water tank ("simple phantom procedure"), following a procedure similar to the procedure used by the IAEA for the postai imercomparisons (fig. 1). 3.7.3. Second step : additional checks in a "multipurpose phantom" The uncertainty in the dose distribution due to beam dosimetry data other than in reference conditions is generally accepted to be much larger as additional errors originate in the estimation of beam flatness, wedge transmission and the calculation of dose distribution (25. . 6) . 26 secona n Thai dwhys i t , step multipurposa , e phantom designed previousle th r yfo IAE groua y A b consultantf po uses i ) checo dt s28 (27k, various beam data (beam symmetry, wedge transmission) and dose calculation procedures (collimator output factor, surface obliquity correction, calculation of the dose at depth). The EC National Reference Centres have checke reliabilite dth procedure th f yo piloa n ei t study conducte cooperation di n wit IAEAe hth . After the feasibility study, the multipurpose phantom is now being used in the first Local Centres of the EC project The main features of the multipurpose phantom are displaid in fig.2. 3.7.4. Third step "postal n vivoi dosimetry". lasimplementede e Ib nth t o stept t ,ye proposes i t i , measuro dt dose eth e deliveret da the individual patient level with mailed dosimeters. 215 A. CALIBRATION CHECK: Co.60 • RX * Vertical beam 10 cm x TO cm at usuai SSD. Ï or 13 ~T, 2 Gy to point Pc (5 cm er 10 cm dependin beae th mn go quality) / B. BEAM QUALITY CHECK : RX j * Vertical beam m c 0 10 D SS t a m c 0 1 x 1m 0c /—w- poino t 1 y P tG 2 X Fig.l Simple phantom description and irradiation procedure 216 MULTIPURPOSE PHANTOM 11 point f measurementso s (LiF) Basic dosimetry data to be checked :: - Calibration and field size dependence - Central axis depth dose - Beam symmetry - Obliquity corrections - Wedge transmission - Summation of doses from multiple fields — Cheks doseof calculation procedures algorithms.and Fig.2 Multipurpose phantom descriptio checd nan k capabilities 217 3.2. Organization The role of the different partners is defined as follows : The Co-ordinating Centre prepares the questionnaires, information and data sheets. It supervises the organisation of the mailing and prepares the time schedule. It ensures the follow-u programme th f po analysed ean resultse sth . The Reference National Centres contac Locae th t l Centres (LC) that wiso ht participate and plan the timing and practical organisation. The NC assures the LC about the strict confidentiality of the questionnaires and of the results, which will be known only by the date Th a . wilNC l e alwayth d an s appeaC anonymoun M a e n ri reportth n i , y CC sswa and/or publications and could never be transferred to administrative or governmental bodies, without full written agreement from the Local Centres. The NC mails the dosimeters received from the MC, to the LC, together with the information and data sheets. It receives the irradiated dosimeters from the LC, together with the completed questionnaire and data sheets. It mails the dosimeters and a copy of the data sheets to the MC, receives the results from the MC and mails them with comments to the LC; It decides on further actio neededf ni . i The Measuring Centre is in charge of the study and optimisation of the methodology and the equipment in relation with the coordinating and National Reference Centres, and of all the measurements related to the Network checks. presene Th t networ representes ki floe th w n dcharo fact n I figf o . t ,.3 with respeco t Eastern Europe, a dedicated experimental network has been set-up in 1993, sponsored by the Belgian authorities (Ministry of the Flemmish Community), called EROPAQ. In this experimental network Universite ,th y Hospita Rafaëlt S l , Belgiud an actinms i C C bots ga e hth Thre. theMC e countrie presenue sar y connecte thio dt s network Czece th , h Republic, Poland and Hungary resulte .Th s presented below include onl firse yth t measurements performer dfo the Czech Republic befor EROPAe eth Q set-up. 4. METHODOLOGY 4.1. The Measuring Centre equipment and methodology (29) The basic MC methodology has benefited from the IAEA/WHO cooperation and the EQAG experience. At the present time, 162 photon beams (58 ^Co and 104 X-ray beams) from 85 centres have been checked, corresponding to more than 2000 TLD readings. 4.1.1. TL material The postal dosimetric checks are performed with LiF dosimeters (PTL 717 from the Desmarquest-CEC Company-France) s .enriche i Thi F sLi d , presentinwitLi 7 h w lo ga sensitivity to neutrons (high energy photon beams) and a low fading (less than 5 % per year at room temperature). Usepowdera s gooda a t dt ge allowi , precisiono t s su ordee th f n o ri , 1.5 % (1 s), for the dose measurements with one dosimeter (polyethylene tubes allowing 5 readings). 218 • Hôpital Si Ral'atl. Lcuvcn (J. Van Dam, J. Verstiaeic) COORDINATING CENTRE • Ccnlic Ci F l.cclcic, Dijon (J C Hoiïot, S. Nauciy) Hôpita t RafaëlS l , Leuven • In.siiiuio RiKli()l(>2>iCii Univoisiia, FloioiicoJL. Ci(>i)m_i, l7 Milano) E. VAN DER SCHUEREN, . DUTREIA X Nciiicilainls Cancel Institute, Amsiculam (11. Bailclink, B. Mijnhccr)! UnivetMiy Hospital UMEA (H. NystrOm) Medica . Angelukis(P l s Schooa u G PapavzsiliouPa • C. l, . KappasC , ) MEASURING CENTRE • Inu OncoloKic ti . a Fiancisco Genti - Lisbol dM a( e Vilhcna . GomeJ , s dd Silva) Institut Gustave-Roussy, Villejuif « I losp de la S.iiii.i Cm/, y San Pahln - Baicclona (1 Ci avert Partie, M Rihas Moi aies)] . CHAVAUDRAJ . BRIDIHRA , , S. DERREUMAUX C/ccli Katliolhciapy Instiitiic - Pi ague (J. Novoiny) Fig.3 Flow chan of the European Pilot Study network 4.1.2. Reading equipment The reading equipment include unito stw s: • a manual reader (SAPHYMO LDT 21) used exclusively for two years (classical heating system with preheating). automatin a • c reader (PCL Fimel), allowin sample0 g9 s loader reae b abou n di o s t 5 t4 presene minute th routin n t i a d te san timee us , using constant temperature ovens (30). 4.1.3. Calibration at the measuring centre The reproducibility, dose response linearity, fading, accuracy and energy dépendance of the TL dosimeters have been checked by the MC during the first year. The reliability of the mailed TLD measurements performed by the MC has been checked with the IAEA acting as an SSDL determinatioe mads i .Th C eM followin e dose th th t ef a nIAEo e gth A protocol (Report 277, 1987), with a reference dosimeter (calibration traceable to the French PSDL, the LPRI, Saclay, France, with an agreement better than 1 %). 4.2. Global postal checks procedure 4.2.1. First step 60co anj X-ray beams output in reference conditions and beam quality of X-ray beams. e dosimeterTh theid san r holder identicae ar s thoso t l e e IAE th th use r y Afo db international networ irradiatehave d b kan o et watea n di r phantom dosimetere .Th preparee sar d at Institut Gustave-Roussy (Villejuif), acting as the European Measuring Centre (MC) and sent ot each participant centre with the holders. protocoe Th participantle senth o t irradiatio e th r dosimetere sfo th f no similas si e th o rt protocol used by the IAEA. Two dosimeters are successively irradiated at 2 Gy to the required depth (usually 5 cm or 10 cm, depending on the beam quality), in a 10 cm x 10 cm vertical beamsource th t surfaca o ,e t e distancsource th t dosimetea o et er o (SSD) r distance (SDD) normally used in the centre. The SEFM Spanish protocol recommends a reference depth of 7 cm for the photon beams of nominal energy between 10 and 25 MV. The irradiations have f holdero t se sbeea wilt e adaptenb bu l performem c alloo dt 0 e irradiatio1 w th t a d f no recommendedosimeters i t I . deptcm a loca e 7 t sa f th h o l o d t centr measuro et outpue e eth th f to beam wit calibrateha d ionisation chamber just befor capsules irradiatioD e eth TL e r th Fo .f no the check of the beam quality of the X-ray beams, two dosimeters are irradiated simultaneously, one of them at 10 cm depth (measuring a dose DJO) and the other one at 20 cm depth (D20). in dosimetero tw e t righa Th e . ts ar angl cm 0 e10 fro f o m D eacSS hfieldn m a c t 0 a , 1 x m c 0 1 a other, to minimise the shadowing of the lower dosimeter by the upper one. A dose of 2 Gy has to be delivered to the upper dosimeter. The quality index (QI) of the beam, defined as the ratio tissue-phanto e facon i th f s ti evaluatem c 0 1 d m throus an ratio 0 2 measuret e sa h th d valuf eo D20/DJO- The readings of the TL dosimeters are carried out at the MC. The repeatibility on the smalA . l% readin e dosimetecorrectio 7 ordee on 0. th f f go n o ri r supralinearits i rnfo s yi applied. When a dosimeter is irradiated in an X-ray beam, a correction, depending on the quality index of the beam and of about 2.5 % for 20 to 30 MV is applied to the reading, to take into account the energy dependence of the dosimeter response. Because in the actual conditions of international postal dosimetry, delays of one to two months can occur between the irradiation and the reading of the dosimeters, a correction is applied for the fading and for unexpected irradiations during the travel. For this purpose, any dosimeter mailing is made using "additional do'simeters" : one is irradiated (2 Gy) at the MC and remains in the laboratory, another is traveld an C s witM mailee e hth th t irradiatea d y dosimetersG 2 t da unirradiatee wels on , a s a l d dosimeter, and the last one is irradiated at the MC (2 Gy) the day of the checking dosimeters return at the MC, before the readings. It is so possible to handle and overcome the fading, the consequences of unexpected irradiations and the possible reader response drift. 220 The accuracy of the dose calibration by the MC has been checked by intercomparisons wit EORTe hth C (Apri IAEle 1991th Ad (Marc)an Novembed han r 1992) dosimeterL .T s have X-raV M y 60C5 e beams4,12 e th d th o n 8i n an bea i , ,d man MC bee e nth t irradiatea y G 2 t da in reference conditions (water phantom, on axis, 5 or 10 cm depth, 10 cm x 10 cm field), and read by either the EORTC or the IAEA. The agreement was better than 1 % for the 60Co beam, and better than 1.6 % for each X-ray beam quality for both intercomparisons. postaL T C lsens checo procedureM T ha te C dosimeterkth M e IAEe th , th Ao st (November 1992) where they were irradiated at 2 Gy in reference conditions (water phantom, on axis, 5 cm depth, 10 cm x 10 cm field, 80 cm SSD) in a ^Co beam. The dosimeters were read at the MC, and the agreement between the absorbed dose in water determined by the MC anabsorbee dth d dos waten ei r IAEstatee th Ay bettes di b r . tha% n1 The dosimeter locae sene th slar o t centres , with instructions sheets explainino t w gho procee irradiatioe dosimetersL th T r e dfo th f no . Data sheet alse sar o maile eaco dt h centre whicspecifo t irradiatio e dosimeter s L date th hT ha yth f e o th f no s (fo estimatioe rth e th f no fading) dose th , e delivere dosimetere th o dt s (state dqualite dose)th d yan , X-raindee th f yxo beams determined by the local centre (stated quality index). The dose and the quality index stated by the centre are compared to the values measured by die MC. The stated quality index is determino t C alsM ocorrectioe e use eth th readin L energy y e applieT de b th b e r o th nt g fo o d t dependenc dosimetere th f eo s response participatine .Th g centr alss ei o aske reporo dt date tth e lase oth ft intercompariso qualitr no y audi whicbeen s ti ha nt h i involved anyf i , . A quadratic summatio uncertaintiee th quantitiee f th no l al n so enterine th n gi determination of the absorbed dose in water, with the MC TLD postal measurement procedure, give estimation sa finae th f l no uncertaint standar e ordere on th r n y,fo o d deviation5 1. = s f o , % for 60Co beams and s = 2 % for X-ray beams. resulte Nationae Th sene th sar o t l Reference Centre (NC) which reporlocae th lo t t centres. The results are also sent to the Coordinating Centre in Leuven for analysis. The NC conduct inquirn sa traco yt origie eth f largno e deviation checkD suggestd TL san .w ne sa Becaus limitee th f eo d resources secone ,th d checks have been performed mainl centree th r yfo s with major deviations. Dependin locae th perforn ln gsituationo ca C on-sitn mN a e th , e visir o t only discus methoe sth d applie dosr dfo e measurements wit locae hth l centre. Deviations lower than twic standare eth d deviatio justift furtheno y o yan nd r investigatio sincC N ee therth y neb higa s i h probabilit observee th procedurr e yth fo o dt deviatioe edu uncertaintye b o nt C N e .Th reports to the Coordinating Centre the results of the inquiries. A strict confidentiality is maintained all over the procedure and no details arc given in any publication, centreseithee countriecharacteristice th e th th n ro n n o ,o r so radiatioe th f so n units. 4.2.2 Second step : Evaluation of the multipurpose phantom (28) e procedurTh s beeeha n developpe t IGRda , usin gmultipurposa e phantom madt ea Houston by WF Hanson (27). 4.2.2.1. Preliminary evaluatio e multipurposth f o n e phantom R IG t a e geometr compositioe Th th d phantoe yan th f no m have been determined witT hC measurements. Attentio s beenha n reliabilit s paiit o dt y (suitabilit mailingr wateye fo th d r )an filling (bubbles require attention). The dosimeter positioning can be accurate (better than 2 mm). inhomogeneoue Th s structure (PVC, water, perspex lean correctiono d)ca t increas n a o t e esdu of the values of the measured dose by up to about 1.5 % with ^o beams and 3 % with 25 MV X-ray with respect to the values expected in an homogeneous water phantom. 221 4.2.2.2. Evaluation of the TL dosimetry procedure r eacFo h photon beam qualitydosimeterL T 3 actine R 1 , th IG sen e s sgar a y b t coordinating and measuring centre, to the participating centres : 11 to be irradiated in the phantom and 2 to detect respectively any unexpected irradiation and fading during transportation. In addition 2 TL dosimeters are kept at IGR to check the fading (see above). Guidelines explaining how to proceed for the irradiation of the TL dosimeters in the phantom, and data sheets in which each centre has to report information on the irradiation conditions and stated doses have been established and sent, with the TL dosimeters, to the participating centres. e conversioTh dosimeteL T f no r readings into absorbed dos wateo et s madi r s ea previously. The dose evaluated by IGR is then compared to the dose calculated by the centres accordin methoe gth d use clinican di l practice. • Step 2 A : Irradiation of the multipurpose phantom by European Reference Centres (Pilot StuUy) In orde multipurposchece o rt th reliabilite f ko th procedure e th us f e yo e th f phantoeo m for quality control by mail, a pilot study was performed during the period September 1991 - May IGRe 199th cooperation y 2i ,b n wit europeah6 h centres (Centr LeclerF eG c Dijon, France, UZ St Rafaël Leuven, Belgium, AVLZ Amsterdam, the Netherlands, Sahlgren Hospital Göteborg, Sweden, Universit Firenzei ad , Italy, Western General Hospital, Edinburgh, UK). In the study, the phantom was first sent by the IGR to the first centre and then directly froparticipatine mon g centenext e capsuleF th Li .o t rdocument d san s (instructio datd nan a sheets) were maile eaco dt frohR centrmIG sen d ean t back after irradiation. dosimeterL T e Th s irradiatio bees nha n performed accordin followine th o gt g procedure : (fig ) 4 . large on perpendicula) e• verticacm 5 1 phantoe x lth m beac o rt 5 m(1 surfac checo et k beae th m flatnes dossymmetrd d an san e , calculatio deptcm a depth m t 4 y c a f ho 2 1 . t na • one horizontal beam (8 cm x 8 cm) uiih a 30°. wedge filter directed on the oblique surface of the phantom. irradiatioe Ith n n procedur capsuleF usen Li ei set o f ,so consideree tw sar F Li : d irradiatee capsulear verticae 8 th o t y s1 db l bea mdosy G wit e2 delivereha deptm c 4 h t danda , on another hand irradiate, capsuleare 11 dto swit6 horizonta hthe l bea mdos Gy wit e2 at ha depthm thereforcapsulec e F 3 ar Li . 8 o t es 6 irradiate boty db h beams. At each measuring point of interest inside the phantom, the dose measured by LiF capsules is compared to that calculated by the centre according to the dose calculation procedure used in daily practice. : Irradiatio Ste• B p2 multipurpose th f no e phanto locay mb l Centres After having checke reliabilite dth procedure th f piloe yo th tn e i study witn e ru hth a , multipurpose phantom has been started in routine situations with the local centres of the EC project. Durin e periogth d 1992-1993, four local centres have particpate witn ru h 6 e th n di photon beam qualitie X-ra6®Co2 4 s ( d yan beams). 222 -Step2A- IRRADIATIONOFTHE MULTIPURPOSE PHANTOY MB THE PARTICIPATING CENTERS IRRADIATION PROCEDURE VERTICAL BEAM 1 5 cm x 1 5 cm at usual SSD i ) cm poin o t 4 ( y 3 tG 2 444 1 2, 4 - Uf capsules : 1 to 8 7 2) HORIZONTAL BEAM 6 _'- .,. 4 . " ' .' usuat a D m SS lc 8 x m c 8 30° wedge if available 2 Gy to point 10 (5cm) capsulef 1 -1 U o t 6 s: Fig.4 Multipurpose phantom Irradiation procedure accordin STEe th A o gPt 2 protocol 223 -STEP2B- 1RRADIATION PROCEDURE ) VERTICA1 L BEAM 15 cm x 15 cm at usual SSD 2 Gy to point 3 (4 cm) 444 1 2,d5 X 4 6 4 - L'rf A capsule8 o t A 1 s: 7 2} UNWEDGED HORIZONTAL BEAM 8 cm x 8 cm at usual SSD 2G poino yt (5cm0 1 t ) 6 4 8 - Lif capsules : 6B to 11B 4 3) WEDGED HORIZONTAL 8 cm x 8 cm at usual SSD 30° wedge if available poino t (5cmy 0 t1 G 2 ) -LifC capsule11 o t C s:6 F ig. 5 Mulüpurpose phantom Irradiation procedure according to the STEP 2 B protocol 224 Taking into account the difficulties in dose interpretation encountered in the pilot study, hane onon d wit combinatioe hth dosimeterbeamf L no T r sfo t deptsa hanothen o and, r hand, with the combined influence of the oblique surface and wedge filter, a new irradiation procedure has been set-up for the run. It consist f threso e different beams (vertical, unwedge wedged dan d horizontal beams) for which three different sets of TL dosimeters have to be irradiated (fig. 5). 5. PRELIMINARY RESULTS 5.1. Structure Radiotherapythe of Departments. questionnairee Th structure th Department e n th so f eo s show large variations between different departments from the same country and are in good agreement with the data published EORTe byth Researcr Cfo h Centres (21)examplen a s considee a ; w f i , physice rth s staffing quoted by 27 answers, it can be expected that the level of physics support in a department will be directly determine patiene th y db t loa carriee physiciste db th whico t y ds b hha . Overalle th , mean number of patients per physicst is 440/year. However, while there are five centres where each physicist is responsible for less thant 200 patients, there are also five centres where the workload is above 600 patients/year, and even two With over 1000 patients. It is evident that in such extreme diverse conditions of workload, there will be différents levels of involvment of the physicist in a number of controlling and preventive measures. There are systematic variations in some of the parameters from one country to another, but the total number of participating centres per country is still too small to draw valid general conclusions. 5.2. Dose measurements 5.2.1. Step one : Dosimetry checks in reference conditions (31) dateo T , cheks have been performe beam2 16 r s dfo beamo fro C centre5 m8 s 8 s(5 X-ra4 an0 1 d y beams). postae Th l measurements performed durin five gth e first month projece th f so t with research centres have presented, for the distribution of Dmeasured/Dstated> mean of 1.0 and a a standard deviation of 1.5 9? with 26 60 the 60c X-rad oan y beams, wherea centree sth presensB spectaculata r spreaDe /Dth f do m s 225 Centres)STEP(85 1 20 "A" centres n = 60 Cobal X-ra+ 0 6 t y beams m = 0.994 0.02= s 0 ' A = 0.083 N 10 ————— 1 nil || Ï7» IM" u") 00 = 0000 Measured dose/ Stated dose 30 "B" centres n = 102 Cobal X-ra+ 0 6 t y beams s = 0.74 20 -- - A0.10 - 2- 10 ä —————| | — _ra«nn i: ÎAME: • l|lnns — -o" / / n \o & <*•: in ec \ n i o —c\ . o ^ * r — n <-i r*- r^ ^ cc co oc o* o o — — o o*c c o" " o " o o" Measured dose/ Stated dose Fig.6 Distributions of the ratios Dm/Ds for the check of the beam output in reference conditions, considerin centre " l beamsal "B d tablr e o s an .Th " e g"A present e sth difference deviatione th n i s s observed between ^Co beam X-rad san y beams. numbee th s nmeani e beamsf standare ro th th s i ,s ,m d totadeviatioe th lA d nan spread 226 N 000000000000000000 o o o o o o o o o o o o o o o o o o OOOOOOOOO oooooo o o o OOOOOOOOO - O O O O O O O O O O o O O O O O OO 7 . g i F Distribution ratioe th f so QI beachece e m/QIth th mf r ko s fo quality , considering all centres and "A" or "B" centres. n is the number of beams, m is the mean, s the standard deviation and A the total spread 227 values table figurf .Th e o presente6 s additional data about this difference between centresA and centres B. The figure 7 is related to the determination of the Quality index (distribution of the ratios Qlm/QI^)- The analysis of the results does not show any significant difference between centres A and B. The only major deviation is nevertheless observed for a centre B. Among the 1 1 centres with deviations > 6 %, 5 have accepted to participate in a second check, resulting in an improvement, with 3 large deviations instead of 9. 5.2.2. Ste : Evaluatio p2 e multipurposth f no e phantom 5.2.2.1. Step 2 A : Irradiation of the multipurpose phantom by European Centres (Pilot Study) In this study, 22 photon beam qualities were checked, including 5 60co beams, 9 X- X-ra8 d . yraan beamyMV V 0 beamM 1 0 s> 1 s< The global results are presented in table 1. From these results, we can report the following comments : . fo centrebeamsl l ral al d dosdose e ratis,a an th th ef o ostatemeasuret o D dTL y db centre byth vers ei y clos unito et y f mino(1.004 o . Abou % % r 1 6 deviation)2 t2. wit = hs s majof o % onl d r observed 6 e deviationan y 1. ar ) ) (betwee% % 6 6 d . > s( an n3 . better results have been obtaine 60r Cdfo oon-axir thaX-rafo r d nfo syan points compare offo dt - axis points. Table 1 Multipurpose phantom experiments. Global results for the STEP 2 A. Rati dosf o e measure TLy db D / dos e state Institutioy b d: ~ n Dm Minor Major Points of Mea f ratioo n l al s r centrefo T=\~ s deviation» deviations U Otr.) Field meas. (IS > * (3%<-St%) (Tolal numb«) l energieAl s Cobalt-«) Low E High E All énergie« l energleAl i (KXOOMV) WX>10MV) Central Axli 1.000 1.013 0.995 0.999 5 0 Large (21) Central Axil 1.000 0.997 1.001 1.001 4 0 Wedged 124) (0.023) (0.019) (0.026) (0.025) (17%) lateral OttAiU'-ipli 1.005 1.019 1.001 1.000 10 3 (45) (0.031) (0.02S) (0.024) (0.041) (22%) (7%) Central Axis 1.010 1.011 1.002 1.017 5 1 Combined (22) (0.030) (0.04}) (0021) (0.0.13) (23%) (5%) field OffAjdiSpli 1.004 1.010 1.005 1.000 9 0 (.48) (0.025) 40.024) (0 025) (0.028) (19%) 22 BEAMS at 7 Institutions : Cobalt-60 to 25 MV X rays Cobalt-« beam5 ): s beamLo9 : wE M 4,5,s: 9 Vd 6an High E : 8 beams : 11,18 and 25 MV 228 . the best results have been obtained for on-axis 6®Co reference points for (s-value of . %) 9 1. o t 1.7% . the largest deviations have been observed for off-axis points in X-ray horizontal beams (s value = 3.3 % with a A value of 15 %) and also for points irradiated by the two beams eitheX-rar fo ^Cr d yrfo obeaman s (obliquity, wedge filter). 5.2.2.2. Step 2 B: Irradiation of the multipurpose phantom by local centres The global results for this study are summarized in table II. The followinye leath o dt g comment conclusiond san s: As for the pilot study, the results are expressed, for each type of photon beam quality and X-ray), as the distribution of the ratio of the dose Dm measured by TLD (at IGR) to the dos centree s stateeD th y eacn db I . h case distributione th , s have been quantifiee th y b d following parameters. Dm/Ds mean value and estimation s of the standard deviation. For each beam considered (vertical, unwedged and wedged horizontal beams), the data relate on-axio dt 'off-axid san s points were examined separately. The majority of the distributions of the ratio Dm/Ds are symmetrical and have a mean close to the unity value : however, for one X-ray horizontal unwedged beam, unexpected very low valueratie th of so Dm/D s were observe 0.33)= d( analysi e resulte .Th th shows f so s ha n thasto o instabilitt n a tpd linairradiatiole e s th cf ha yo monito0 n22 aftee th r r f unitonlo 7 y7 s planned and the machine was not started again to end the irradiation. Disregarding these very low values, the results show that, for on-axis points, deviations of the same order than for those observed in the pilot study have been obtained for the vertical beam thar Fo .t points, slightly larger deviations have nevertheless been observed. However, for off-axis points, a large increase in the deviations has been observed for l beamsal . Compare on-axio dt s points estimatioe ,th standare th f no d deviatio increases ni y db a factoabouo t p , particularlu rt2 horizontar yfo . %) 5 l X-ra5. o t y4 beam4. = s s( In achieveshortn ru result e e ,th th routin n di f s o e conditions with local centres showa large increase of the deviations for the fields which are not perpendicular to the phantom surface and, for all fields, a large increase of the deviations for off-axis points. Although a rate of minor deviation comparabl) % betwee % 6 d 1 s(2 an ntha3 o e t t obtaine piloe th tn d i stud observes yi d locae foth r l centres rate majof ,th e o r deviationmucs i ) h% large6 > s( r (aboutha) e % nth 5 1 t pilot sutdy (= 1.5 %). The large deviations observed in non reference conditions for two of these local centres which have obtained correct result previoua n si s maile intercompariso- L dT referencn ni e conditions, point out the interest of performing such checks in order to verify on the one hand. the beam symmetry and the beam flatness and on the other hand, the accuracy of the algorithms used in routine for dose distribution calculations. 229 Tabl 2 e Multipurpose phantom experiments. Global resultSTEe . th B r Ps2 fo Rati dosf oo e measure dos/ D eTL statey db Institutioy db : m n D Minor Major Pointf so *Mean of ratios -prm— : all centres deviations deviations Field meas. as ) US UT. <-S6T.) (>67.) (Total number) AU énergie! Cobalt-«) RX All énergie« All energies i Central Axis 1.000 0.997 1.001 0 0 2 =• 4 cm (6) (0.016) (0.018) (0.018) Off Axis 1.008 0.997 1.013 4 4 Large (24) (0.036) (0.018) (O.«:) (17%) (17%) anterior Central Axis 0.999 0.995 1.001 0 0 deptn I h (6) (0.017) (0.014) (0.019) Oft Axis 0.991 0.980 0.996 3 0 m) (0.027) (0.020) (0.030) (25%) Centn Axil s 0.661 HQ397) 0.992 0.773 "(3.002) 0 1 Z = 5on (5) (0.300) (0.017) (0.011) (0-397) (0.025; (20%) Off Axis 0.863 'aaxw 0.994 0.777 >a.006> 3 2 Unwedged (10) (0.2S9) tO.OZS) (0.024) (OL337) (O.C23) (30%) (20%) lateral Central Axis 0.875 >fl.009J 1-012 0.783 VLOOSJ 2 1 deptn I h (5) (0301) (OJÛ3« (0.046) (OJS«) I0.M9; (40%) (20%) Of if Axis 0.867 »aw .a 1.007 0.774 i(lJ001) 0 5 (10) (0292) (0.04!) (0.045) (OJSS) (O.OS» (50%) Central Axis 0.974 0.980 0.970 2 0 Z = 5on (5) (0.023) (0.007) (0.031) (40%) Off Axis 0.971 0.979 0567 4 2 Wedged (10) (0.04«) (0.020) (0.062) (40%) Q0%) lateral Centra] Axis 0.993 1.005 0.985 0 0 In depth (5) (0.023) (0.03S) - (0.013) Off Axis 0581 0.984 0.980 5 1 no) (0.0«) (O.M7) (0.050) (50%) (10%) * : No correction applied for phantom composition and geometry. Withou: i t considerin beae mon vervaluee o (interruptiot w g th y e lo sdu irradiation)e th f no . 6 BEAMS at 4 Institutions : Cobalt-60 to 15 MV X rays Cobalt-60:2 beams RX : 4 beams : 4,6,10 and 15 MV 230 . 6 DISCUSSION 6.1. Step 1 : Dosimetry checks in reference conditions An important interest of this siud\ was to show that 11 "B" centres could present deviations > 6 % for the beam calibration, some being very large. We can notice first that these large deviations lea underdosago dt patiente th f eavoidino o s(s g alarming acute reactions) potentially inducin glowea r cure irradiatio e rateth f i ,dosimetere th f no representativs si e th l eo irradiation ol" the patients. Secondly, the physicists were supposed to check the beam output before irradiating the TLD's. From a detailed analysis, submitted to publication (31), it appears that, after on-site visits in most centres with major deviations, the physicists did not have the necessary i calibratexperiencno d di ed regularlean ihcir. beame 11 centre6 yth sf n o I st sou dosimetero n s dosimetere wa th r so s availabl t beeno nd calibrateeha d recentl centres3 n I y e th , physicist t giv explanationy no e an d sdi . 6.2. Step 2 : Multipurpose phantom It is interesting to quote that, with centres B, the large deviations observed in non reference conditions appeare centredn i s having obtained correcs i t i , t resultstee So . th p 1 r sfo very important, in such a network, not to limit the investigations to the reference conditions It seems worthwhile to go further in the evaluation of such a phantom, an improved version being under développement (WH Hanson). The procedure could also be improved (already, from step 2. A to step 2.B a better tracing of the reasons for deviations is obtained). The problem is the number ol dosimeters involved, making the large scale use of this phantom expensive 7. CONCLUSION This pilot study has shown the feasibility of the experimental european network, with its organisation relying on Coordinating, Measuring and Reference National Centres. This experience can allow each european country to plan the set-up of a national network connected europeae toth n network. From the scientific point of view, the accurancy of the dose checks is compatible with the action levels considered in most instances of the radiotherapy littérature. The discovery of numerous major deviations outlines the need for any European radiotherapy department to be able to benefit from a QA network and support the initial sucr applicatiofo projech a C E e tth o nt Further improvements are expected from the pilot study, regarding the accuracy, the optimisation of the procedures, the addition of electron beams checking and in vivo measurements. The final outcome will rely on the capability of each country to develop its national network. ACKNOWLEDGMENTS This work was supported by grants from the Commission "Europe Against Cancer" an dcooperatioa n wit IAEe authore hth ATh gratefue physicistse ar th o t l radiotherapistd san s having actively participated in the set-up of the project, and in the initial experiments and checks, especially : - In Belgium : J Van Dam and J. Verstraete Franc Hono. n -I NaudS C . d J e: an t y 231 - In Italy : L. Cionini and F. Milano Sweden I - . NyströmH n: . SvenssonH , . ZackrissonB , , K.A. Johansso . HansoU d nan n Netherlande th n I - . BartelinH . sMijnhee: B d kan r - In the UK, Scotland : D.I. Thwaites IAEe . Svenssoth H n A I : . Nett - P d enan REFERENCES [I] SVENSSO , HANSONH. d ZSDANSKNan O.Pe IAEA/WH. Th . YK. L OT Dosimetry Service for Radiotherapy Centres 1969-1987. SSDL Newsl. 28: 3-21 (1989). [2] GOLDEN , CUNDIFFR. , , J.H., GRANT, W.H., SHALEK reviee A . th f wRJ ,o activitie AAPe th f sMo radiological physics cente interinstitutionan ri l trials involing radiation therapy. Cancer 29: 1468-1472 (1972). [3] JOHANSSON, K.A., HANSON, W.F., HORIOT, J.C. Worksho EORTe th f po C Radiotherapy Grou qualitn po y assuranc cooperativn ei e tria radiotherapyf lo :A recommendation for EORTC Cooperative Groups. Radiother. Oncol. 11: 201-203, (1986). (4] WHO, Quality Assurance in Radiotherapy, A monograph, WHO, Geneva, (1988). RAHMEB [5] . (ed)A , Accuracy requirement qualitd san y assuranc externaf eo l beam therapy with photon electronsd san . Acta Oncol , Suppl15 . (1988)1 . . [6] DUTREIX A. When and how can-we improve precision in radiotherapy ? Radiother. Oncol. 2: 275-292 (1984). [7] SVENSSON H. Quality assurance in radiation therapy : physical aspects. Int. J. Radiât. Oncol., Biol. Phys , Suppl10 . 59-6: 1 . 5 (1984) [8] LEUNENS G., VAN DAM J., DUTREIX A. and VAN DER SCHUEREN E. Quality assurance in radiotherapy by in vivo dosimetry. 1. Entrance dose measurements, a reliable procedure. Radiother. Oncol : 141-117 . 51990L . |9| LEUNENS. G. VERSTRAETE J., VAN DEN BOGAERT W.. VAN DAM J., SCHUERER DE N DUTREI VA . Human NE d an . XA s error datn si a transfer during the preparation and delivery of radiation treatment affecting the final result. "Garbage in, garbage out". Radiother. Oncol 217-222: .23 , 1992. [10] LEUNENS , VERSTRAETE, J. G. ,BOGAERT N DAMN ,DE VA N , VA W. , , J. , DUTREIX, A. and VAN DEN SHUEREN, E. Humans errors in daya transfer during preparatioe th deliverd nan radiatiof yo n treatment affectin finae gth l result. "Garbage in, garbage out". Radiother. Oncol. 23: 217-222, (1992). [II] ALETTd NOE s mesurean . Le L A . P Ivivn i s o systématiques.A 0 propo80 e d s contrôles. In : "Proceedings of the XXVIth Congress SFPH Vittel, 124-136, 1987. [12] RÜDEN B.I. Evaluation of 'he clinical use of TLD. Acta Radiol. 15: 447-464,1976. [13] , HOORNAERTVYNCKIEJ. M BOUILLEd DA an . N RS A VA . , RA M.T. , dosimetric quality audit of photon beams by the Belgians Hospital Physiscists Association. Radiother. Oncol 37-4: .28 3 (1993). [14] JOHANSSON K.A., MATTSON L.O. and SVENSSON H. Dosimetric intercomparison at the Scandinavian radiation therapy centres. Acta Radiol. 21: 1-10, 1982. 232 [15] THWAITES D.I., WILLIAMS J.R., AIRD E.G., KLEVENHAGEN S.C. and WILLIAMS P.C dosimetri.A c intercompariso megavoltagf no e photon beamK U n si radiotherapy centres. Phys. Med. Biol 445-461: 37 . , 1992. [16] WITTKÄMPER F.W., MIJNHEER B.J. and VAN KLEFFENS H.J. Dose intercompariso radiotherape th t na y centre Netherlandse th n si Photo. 1 . n beams under reference condition prostatr fo d esan cancer treatment. Radiother. Oncol : 33-449 . . 1987. [17] HUNTLEY, R.B., BERANETTEd an . P . ,Dosimetr P , y intercomparisor fo n Australia "organise e Australiath y b d n Radiation Laboratory"n i . Pape0 7 ° n r Measurement Assurance in Dosimetry, "Proceedings of an IAEA International Symposium", Vienna, (1993). [18-1] SAMULSK . DUBUQUT I E G.L., CACAK R.K.. COURLA , DEWERSG. D L.A.,HILK , OHUMPHRIESR. , L.J., JONE , MASTERSOSD. N M.E., MILLER D.W..STOVA WOCHO& . LM S J.F. Radiation therapy dosimetry reviewe th y sb centers for radiological physics. Int. Rad. Oncol. Bio]. Phys. 7: 373-383 (1981). [18-2] KIRBY, T.H., HANSON W.F., Gastof R.J. u C.H ,Ch d SHALEK .an . R.J. Mailable TLD System for photon and electron therapy beams Int. J. Radiât. Oncol. Biol. Phys. 12: 261-265, (1986). [19] BOYD, A.W., EISENLOHR, H.H. and GIRZIKOWSKY R. Postal dose intercompariso f 60Cno o dosimetry using thermoluminescence dosimeters among- secondary standard dosimetry laboratories. Med. Phys : 112-1139 . , 1982. |2()j SVENSSON, H;, HANSON, O.P., ZSDANSKY, K. The IAEA/WHO TL dosimetry servic radiotherapr efo y centres 1969-1987. Acta Oncol : 416-4629 . 7 (1990). [21 ] HOR1OT J.C., JOHANSSON K.A., GONZALEZ D.G., VAN DER SCHUEREN - E., VAN DEN BOGAERT W. and NOTTER W. Quality assurance control in the EORTC Cooperative Group of Radiotherapy. 1. Assessment of radiotherapy staff and equipment. Radiother. Oncol.. 6: 275-284, 1986. |22] HANSSON , JOHANSSOU. , N K.A. Quality audi f radiotherapo t y with EORTC mailed in water TL-dosimetry. Radiother. Oncol. 20: 191-196, (1991). [23] BRIDIER A., CHAVAUDRA J., NGUYEN J.. KERAUDY K., BEAUVAIS H.. LHALLABI A., CHASSAGNE D., ESCHWEGE F. Results of the dosimetric studies programmA performeQ e e e framEtanidazolth th th f f n o ei o de 2508R (S e ) radiotherapy multicenter study. IAEA-TEC-DOC 734. Radiation dose in Radiotherapy from prescription to delivery Feb 1994, PP 209-223. [24] DUTREIX A., VAN DER SCHUEREN E., DERREUMAUX S., CHAVAUDRA J. Preliminary result qualita f so y assurance networ radiotherapr kfo y centres Europen i . Radiotherap Oncologyd yan . 1993 , November? 2 vol , 29 . 97-103p p , . [25] American Associatio Physicistf no Medicinen si , Repor . Physica13 t l aspect qualitf so y assuranc radiotherapyn ei ; American Institut Physicsf eo Yorkw ,Ne , 1984. [26] Société Française de Physiciens d'Hôpital (Editor). Publication 4. Quality Control of electron accelerator medicar sfo l use, SFPH, Institut Curie, Paris, 1989. [27] HANSON W.F. A simple geometrical phantom to be used in the IAEA network. In : International Semina r Europefo r e Middlth , e EasAfricd an tn Radiotherap o a y Dosimetr yRadiatio; n Dos Radiotherapn ei y from Prescriptio Deliveryno t . Leuve- n16 , Septembe22 r 1991, IAEA Techn. Doc. 734, Feb.1994 311-316P ,P . 233 [28] BRIDIER, A., DERREUMAUX, S., CHAVAUDRA, J., DUTREIX, A., SVENSSON . ChecH , multipurposa f ko e phanto frame useIAEe a th b f n mdeo o i t A quality assurance network. 2nd Biennal Meeting on Physics in Clinical Radiotherapy, ESTRO. Prague 28-3 1993y 0Ma , Abstrac . 117p t . [29] S. DERREUMAUX, J. CHAVAUDRA, A. BRIDIER, V. ROSSETTI and A. DUTREIX. Quality Assurance Network for external Radiotherapy in the E.C. Methodology and technical aspects of the dose measurements. Submitted to Radiotherapy and Oncology in 1994. [30] BARTHE , MARINELLOJ. , , G.,, POLLACK PORTALd automatian w . J , Ne . G , c fast reader for powder or sintered pellets used in medical physics. Protection Dosimetry 34: 261-263 (1990). |31] DUTREIX, A., DERREUMAUX, S., CHAVAUDRA, J. and VAN DER SCHUEREN, E. Quality control of Radiotherapy Centres in Europe. Beam calibration, Submitted to Radiother. Oncol. in 1994. [32] KIRBY, T.H., HANSON, W.F., JOHSON, D.A. Uncertainly analysis of absored dose calculations from thermoluminescence dosimeters. Med. Phys. 19: 1427-1433. (1992). 234 V. QUALITY ASSURANCE PROGRAMME IN RADIOTHERAPY si. MINIMUM REQUIREMENT PROGRAA Q A N SMO N I XA9642864 RADIATION ONCOLOGY PR. ALMOND J.G. Brown Cancer Center, University of Louisville, Louisville, Kentucky, USA Abstract Apriln I , 1994 Americae th , n Associatio f nPhysicisto Medicinn si e publisheda "Comprehensive QA for radiation oncology:" a report of the AAPM Radiation Therapy Committee. This is a comprehensive QA program which is likely to become the standard for such program Unitee th n si d States prograe Th . m stresse interdisciplinare sth y naturA Q f eo in radiation oncology involving the radiation oncologists, the radiotherapy technologists (radiographers), dosimetrists acceleratod an , r engineers medicae wels th a , s a l l physicists. This paper describes a comprehensive quality assurance program with the main emphasis on the quality assuranc radiation ei n therapy usin glineaa r accelerator papee Th r. deals with lineaa Qr Afo r accelerato treatmenr fo simulatod A an r Q td planninran g computers. Nexe th t treatment planning process and QA for individual patients is described. The main features of this report, which should apply to QA programs in any country, emphasizes the responsibilities of the medical physicist 1. Introduction The role of quality assurance in radiation oncology has received increasing attention in the last few years and its importance is now fully recognized in maintaining consistent accuracy of the absorbed dose delivered to patients undergoing radiation therapy. Sources of erro derivn rca e from deficiencie tumon si r localization, patient immobilization, field placement, daily patient set-up, dose calibration and calculation as well as from equipment related problems. Quality assurance in radiation oncology may be defined as those procedures that ensure a consistent and safe fulfillment of the dose prescription to the target volume, with minimal dose to normal tissues and minimal exposure to personnel and the public (AAPM 1984). [1] prograA Q e mradiation i Th n therapy, therefore, cover widsa e rang areasf eo , often involving several medical disciplines and the medical institutions management Coordination, therefore criticas i , l among radiation oncology physicists, dosimetrists, accelerator engineers, radiation oncologists, radiotherapy technologists (radiographers), other medical disciplines and management In many institutions the medical physicist is best placed to oversee such a program. The aim of a QA program for radiation therapy is an ongoing evaluation of the functional performance characteristics of the associated equipment and calculations, because these characteristics influence bot geometricae hth dosimetrid an l c accurac appliee th f yo d dose. There are two main parts of such a program: (i) periodic QA measurements and 237 evaluation and (il) regular preventative maintenance. The medical physicist should be responsible for making sure that both parts of the program are carried out three Th e main area sourcer sfo f inaccuracso dosn yi e deliver identifiee b n yca : das a) Physical dosimetry, i.e., the commissioning and calibration of treatment machines and sources. b) Clinical dosimetry, i.e., the delineation of the target volume and critical structures, acquisitio f patienno t specific factor dosd san e distribution calculations. c) Patient treatment, i.e., the setup of the patient and the recording of the treatment and final verificatio accurace th delivere e f nth o f yo d dose. Any QA program will be based upon a complete determination of baseline values at the time of acceptance and commissioning of the equipment Data for any machine must be measure should assumee b dan t dno d identica dato t l a from similar equipment Most manufacturers provide in written form, their acceptance test procedures which list the mechanical and radiation parameters which will provide the benchmark for the equipment Commissioning provide detailee sth d information abou equipmente th t , e.g., table f beaso m data. This data obtaine eacr dfo h piec equipmenf eo tbenchmare addth o st k data. Once eth acceptance tests, commissioning and calibrations have been completed a QA program must commence to insure that the accuracy of the treatments is maintained, i.e., the goal of such a program is to assure that the performance characteristics established during commissioning shows no serious deviations. 2. The QA Program in Radiation Therapy radiatioa Q An i n therapy department cover widsa e rang activitief eo e th d san treatment proces viewee b n mann sdca i y differen purposee t th ways r Fo thif .so s discussion three main areas have been identified. They are: a. External Beam Treatments . b Brachytherapy Treatments . c Measurement Equipment d. Clinical Aspects of the Treatments These area e shosar Tabldiscussinwn n i I . eI topie g th f thico s seminar "Radiation Dos Radiotherapn ei y from Prescriptio Deliveryo threne t th f e o aspectl "al s mentioned above are of importance to quality assurance. Because it is not possible to cover this whole subject in a single paper the quality assurance progra lineaa r mfo r accelerator throug completioe hth patiena f no t treatment will presentee b somn di e detai althougd an l specifice hth s wil differene b l other fo t r areae sth general approach will be quite similar. Listed in Table II are some general references which describe quality assurance program l aspectal r sradiatiofo f so n oncology. 238 Table I ___QA Progra Radiatioa mn i n Oncology Department A. External Beam Treatments 1. External Beam Radiation Equipment a) Superficial and Orthovoltage Machines b) Cobalt 60 Units c) Medical Electron Accelerators (i) X-rays (ii) Electron Beams 2. Simulators and CT Scanners a) Simulators b) CT Scanners . 3 Treatment Planning Computers ) a Treatment Planning Program c) Test Procedures . 4 External Beam Treatment Planning a) Treatment Planning Process ) b Treatment Plannin Individuar gfo l Patients B. Brachy therapy . 1 Sealed Sources ) a Calibration b) Wipe Tests 2. Unsealed Sources a) Calibration b) Contamination . 3 Source Inventories a) Long Lived Sources b) Short Lived Sources 4. Brachytherapy Equipment ) a Applicators b) Seed Inserters . 5 Remote Afterloaders a) Calibration of Source b) Source Position 6. Treatment Planning and Dosimetry a) Source Localization b) Dose Calculation c) Delivery of Treatment . C Measurement Equipment 1. Standard and Field Instruments for Absorbed Dose . 2 Relative Dose Detectors 3. Patient Monitoring Detectors . 4 Beam Scanning Equipment . 5 Accessories Clinicaf o A Q l Aspect . D s 1. Chart Review . 2 Film Review 3. Beam Modifying Devices 239 ______Table n______Physical Aspect Qualitf so y Assuranc n Radiatioei n Therapy AAPM Report No. 13 Task Group 24 (New York, NY - American Association of Physicists in Medicine 1984) (AAPM 1984) [2] Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Grou 1990 p4 4 (AAPM ] 1994[1 ) Report No 2 Radiation Control and Quality Assurance in Radiation Oncology: A Suggested Protocol. The American College of Medical Physics 1986 (ACMP 1986) [3] Joh . HortonnL : "Handboo f Radiatioko n Therapy Physics" 1987, Prentice Hall, Inc., New Jersey (Horton 1987) [4] Commissionin Qualitd gan y Assuranc Lineaf eo r Accelerators Institute Th , e of Physical Sciences in Medicine, Report No. 54, 1988 (IPSM 1988) [5] Quality Assurance in Radiotherapy Physics Proceedings of an American Colleg f Medicaeo l Physics Symposium 1991y ,Ma . Medical Physics Publishing (ACMP 1991) [6] PrograA Q A mRadiation i . 3 n Therapy Usin Lineaga r Accelerator The QA program for radiation therapy treatments with a linear accelerator will involve other equipment and areas and these are listed in Table III. Table III Q Medicaf Ao l Linear Accelerators f Simulatoro A Q s Q f MeasuremenAo t Equipment Treatment Planning Computers External Beam Treatment Planning: General Individual Patients In-Vivo Dosimetry Chart Review In developing a QA program it is important to use measurement techniques which are simple, rapid, and reproducible and which can determine parameter changes smaller than the toleranc r actioeo n leve whicd an l h will minimiz tese eth t time. f Medicao A Q le LineaTh r . Accelerator4 s The most significant factors affecting the ability to deliver the correct tumor dose include: exact dose calibration, accurately determined depth dose, off-axis dose characteristics, wedge and block factors, etc., that will be obtained during commissioning of a linear accelerator additionn I . , certain other data wil obtainee b l d during acceptance tesr fo t such machines, including measurements of the mechanical, radiation-beam and radiation protection specifications goo A prograA . dQ m will, therefore, monitor eac thesf ho e 240 parameters in order to ensure that the machine is performing within specifications and that the dose is being accurately delivered. The protocol for these QA tests should recommend the equipmen usede b frequenc e o t th , f measuremento y techniquee th , followede b o st d an , suggested performance criteria. Obviously the equipment and techniques should be a simple as possible in order to reduce the time, effort, and cost of such tests. The frequency is chosen depending upon the likelihood of a change and the effect upon treatment if a change should unexpectedly occur. Finally, actio notificatiod nan n levels mus establishede b t . T*MclV Q Medicaf Ao l Accelerator» Frequency rr.^™ Tokruxe* Daa, Dotamctry 3« X-ny output <&a*'w»cvfy Ekcoon cnttp« cc.oia.ocy* Kleebauucij LnraJaliag Uaent 2inm DlaftaaOOC n>d.C..IOr (ODI) atun Sftfeiy DoorB»dHock PXUKDCQ.Î AudtovouaJ monitor FuKtk«..! MoodJy Doamctiy X-ny oatpoi coaaomcy* 2% ElCCtRD OBtfHat COKtaVKy* 2% 2« X-»y oenaT.,1 *xi* AjaimiUy pa.r..inttgj (TOD. TAR) caottaoc/ 2» X-ny betuD tbcoe.« ooon.tf.cy 2ft 3** X-ny «ad ctcctrai gynmrtry 3* Sttfciy IxBcriocka. EjDerxncy off *wiict.e» Pancoonal Wedge, elecoroo cooe io.crioclu PincDooal Mcduiuc..! Cbccla LicbaAutiaVUoa field cwncidenoe aide• n *o * 1 r o m 2m GaaDiry/caUiixu.x.r angk a.tiic..tcin lde( Wcdjcp«.«» 2 mm (or 2% change «a tranaaiaeioa factor) Tny pouboct 2mm Ai^icuor potioaa 2mm Fkld •!•* atidJCMon 2mm CRMavbur ooacT...« 2mmdumcter TirtumcE« coocb poai&oa indicaion 2 mm/I dec. T-tfrhint of »edge», «ociffi« tny PuActioBal Jaw cyn.OK.ry* 2 mm Fkld IltJat aaatCaaatiTy FimcrJoml Annual Docmcny X*nyjkl«ctroa ou.fm c*tibraIJoa coculincy 2« Fkl t dependeflcdva *-nf co y oaipA cautaacy 2% ODtpnl laaCtOr COaaaCaaOC r ClOCtTOfa y B «ppatC..!«.- 2« Ceatnl tuii« pVaUDCacr caoABacy (FDD. TAR) 2% Off-aUM fkctor cacHt..ocy 2% Truanaitaioa fcctor '"TBrTTiry far •! tretatrocnt 4>ccea.»ocka 2« Wedcc QMrffMfiftffrffi otctor couCaaDCy* 2« MoailOf clambcT Incmty 1« X-ny ODD«! oooftuicy vi cmtiy aogk TO. Ekctroo oœpui «oaBKicy w SMCry «..de 2« Off-ajit* Etctor coaniac gMn» yv y cac^c 2« Arcoodc Mfrjvspcci. S-vfcty Lacilock* FoUo* mÄmfacniren m pRxedare« Puflctiooal Mcctauuc«! QieckB CoUimuar rooboD •wonacr 2mmdlan»ncr 2mmdiamecr CoUCfa lOfaatkaa. kP^tMCT CMncidccKc of ccU.TBB.ry. graay. cottcf * wita«c h boocotcr 2mn!.um^ Coiacide..cc of cwbatooa and twcbu..«! uoceiacr 2nMndiamcKr Table top .M 2mm VemCaat D^vd Tac tolerances littcd m me ubka abould be mtcrpreled to mean mat if a paramcajr entier: (1) czcccda the tabulated value (c^, me mcaawed iaoccnter ander gaavy rotation exceed* 2mm diameter); or (2) DM me change a die parameter cjtcceda die nommai value •Whicheve pealcra r . Should ala checkee ob d after chang litt ea » fide tcarce. Ta- aymmerry a dt fined at djfcrcooe m daaace of oeb jiw fiom «a wxxatx. *««" — •'IT-' •— — -- — «—•— — «-" •— — ' •«-T-«- * ) ' • 241 measurementA Q e Th s recommende medicar dfo l accelerator gives si Tabln ni V eI grouped by the frequency of such measurements and is taken from the latest AAPM report on "Comprehensiv radiatior fo A eQ n oncology" (AAP Mbeae 1994)Th m ] parameter[1 . d san tolerancee th baseline giveth e s ar d eacr nan efo h measurement shoul those db e established during acceptance and commissioning. It should be pointed out that some flexibility should appliee b thio dt s table, especially with regarfrequencye th o dt . Some monthle yb testy sma done weekly, for example a policy for the precise frequency at which each test is done should developee b t eacda h institution. of-SimulatorA Q e Th s . 5 Since simulator designee sar reproduco dt geometrie eth c condition radiatioe th f so n therapy equipment (BJR 1989) [5] they are subject to the same mechanical checks as the accelerators additionn I . simulatoe th , r shoul checkee db imagr dfo e quality accordino gt guideline r diagnostisfo c radiography units (AAPM 1984 [2], ACMP 1986 [3]). TableV summarizes the QA tests for simulators. Table V QA of Simulators Frequency Procedure Tolerance* Daily Localizing lasers 2 mm Distance indicator (ODD 2 mm Monthly Field size indicator 2 mm Gantry/collimalor angle indicators Ideg Cross-hair centering diametem m 2 r Focal spot-axis indicator 2 mm Fluoroscopic image quality Baseline Emergency/collision avoidance Functional Light/radiation field coincidence % 1 r o m m 2 Film processor sensitometry Baseline Annual Mechanical Checks Collimator rotation isocenter 2 mm diameter Gantry rotation isocenter diametem m 2 r Couch rotation isocenter diametem m 2 r Coincidence of collimator, gantry, couch axes and isocenter 2 mm diameter Table top sag 2 mm Vertical trave f conco l h 2 mm Radiographie Checks Exposure rate Baseline Tabl exposurp eto e with fluoroscopy Baseline calibratios mA d an nP Kv Baseline High and low contrast resolution Baseline The tolerances mean that the parameter exceeds the tabulated value (e.g., the measured isocenter under gantry rotation exceeds 2 mm diameter). f Measuremeno A Q . 6 t Equipment e measuremenTh t equipmen equalls i t y importan radiatioe th s a t n treatment equipment and should be part of the QA program. The recommended QA tests, frequency, and tolerance limits are given in Table VI including tests for automatic beam scanners. 242 Table VI QA of measurement equipment. I, Initial me for each mode uted or foBowing malfunction and repairs; E, each use (measurement sequence ongoinr )o g evaluation; B, eac l appropriata h x hatcbo r ho e energy (dosimeter element position should alse ob considered); D, documented and correction applied or noted to report of measurement; M, monthly. Instrument tjpe Test Frequency Tolerance* Local Standard* ADCL calibration 2/ D Lmcanty 2y- OS* Ventmg 2/ D Extra-camcral signal (stem effect) ! 1 0.5* Leakage E 01* Redundancy check E 2« Recombination I D Collecting r*^*'*'«i E D Field ' "•""•"»"*• Local std. companion 2y 1% Lsxanty 2y D îy D Extra-cameral «goal 2y D Leakage E 0.1« Recombination I D Colkcnng potential E D Output Check Local ttd companson M 1« Relative dose T1D CalibraQon E D Lmeanry I D Film Dose icsponae B D Dcnsitomcler lioeanry iy D Processor untfonmry/repiodacibtliry E D o Cbambclo r Lmeanry iy D Ex&a-camcnl siBial l 1* Diode« cncnjy ottpr nV'n i. I D Excm-omeral signal I D Lneanry I D Positioning Accoracy E 2 T"" Kyseicsts E 2mm Automated Scannen Mcchuucal I 2 mm Posibotal accuracy E 1 mm CoUcclmg potential of detector E D Detector Imeanty I 03« Exm-cameral signal I OJ« Detector teatage E 0.5* Accuracy of data analysis I 1« Accuracy of pnntouta I 1 mm Accessories I 0.1 deg/C Barometer Calibranoo 3mo 1 mm/Hg Linear rale Calibration I OJ* "Percen deviatioe t th value parametee ± th e f sn ar o respec n nominale rwa th o distancet d an , referree sisocemee ar ib o dt nominar re l SSD. "Local standard instrumen calibratioa s ha t n directly traceabl NISo et should Tan reservee db calibratior dfo f radiationo n beams., field """TffBcntt. ^fxJ unercoQipmsoaav. Two years required by NRC Without a redundancy program, dus may be inadequate, with a redundancy program, dosaneiry systems mairaain calibration beto significantlr nfo y longer pcnot ornf bo e *With a radionudide (e.g., SR-90) or chamber aaercompartsoo. Redundanc dosn yi e calibration equipmen necessars i t assuro yt e that instrumente sar holding their calibration. This can be established by comparing the response of the measurements equipment with an appropriate long-lived radioactive source (Strontium-90). If access to a Strontium-90 check source is unavailable, then at least two independent dosimetry systems should be maintained. A Cobalt 60 teletherapy machine can be used as part of a redundant measuring system but with care. If only one dosimetry system is available, a redundant system should be formed with a dosimetry system at another institution, with quarterly intercomparisons. . 7 Treatment Planning Computers The treatment planning computer is a critical component of the entire treatment process. Computer usee b calculate o dt y sma d patient dose distributions, monitor unita r sfo 243 given prescribed dose and fixed point dose calculations for irregular fields, etc. All such systems should undergo rigorous acceptance testin commissionind gan prograA Q a md gan implemented. Complete documentation by the manufacturer should include the methods for obtainin beae gth m data librar othed yan r data necessar implemeno yt systeme th t A . complete descriptio physicae th f no l model dosr sfo e calculations with expected accuracd yan limitations shoul providee db d along with complete input-outpu operatind an t g instructions. Various test procedures need to be described to be carried out initially by the manufacturer an userde th the y . nb Tests should als done ob e after program modification pars a f o d t an s an ongoin programA gQ . Tabl I list recommendee eVI sth treatmenr fo A dQ t planning system monitod san r unit calculations. Table VII QTreatmenr Afo t Planning System Monitod san r Unit Calculations Frequency Test Tolerance* Commissionin followind gan g Understand algorithm Functional software update Single field or source isodose 2%' or 2 mm» distributions MU calculations 2% Test cases m m 2 2 r %o systeO I/ m 1 mm Daily I/O devices 1 mm Monthly Checksum No change Subset of reference QA test set (when c mm 2 2 r %o checksum availablet sno ) systeO I/ m 1 mm Annual MU calculations 2% Reference QA test set d mm 2 2 %r o I/O system 1 mm *% difference between calculatio computee th f no r treatment planning syste measuremend man r independen(o t t calculation). "In the region of high dose gradients the distance between isodose lines is more appropriate than % difference. In addition, less accurac obtainee b y f singlo yma d een nea sourcese rth . These limit scompariso e refeth o t r f dosno e calculation t commissioninsa same th eo g t calculation s subsequently. These limits refer to comparison of calculations with measurement in a water tank. . 8 External Beam Treatment Planning In this section, QA for the treatment planning process is discussed, followed by a discussion of QA for individual patients. QA in treatment planning may refer to two distinct processes. (1) Nongraphical planning is often used for single or parallel opposed fields. In this approach monitoe th , r units (minutes prescribee th r )fo d centrae dospointa th o en t o l axi usualls si y calculated using central axis depth dose, tissue phantom ratio r tissuso e maximum ratios bead an ,m output calibration tables. Furthermore fiele th , d apertures, which define the treatment volume, are usually designed on radiographs obtained during simulation; (2) Traditional graphical plannin uses g i manr dfo y patients thin I .s method targea , t volume is defined from CT or orthogonal simulation films, and the patient's contour is either obtained using a mechanical device (e.g., lead solder wire) or from CT. The field arrangements are designed and dose distributions calculated on one or a limited number of axial cross sections using a computerized treatment planning system. The radiation oncologist then prescribes the dose to a point or an isodose curve, and the field apertures are usually defined, as in procedur , froe1 m simulation films. 244 8.1. General Treatment Planning Process Treatment planning is a process that begins with patient data acquisition and continues through graphical planning, plan implementatio treatmend nan t verification t entailI . s interactions betwee radiatioe nth n oncology physicists, dosimetrists, radiation oncologists, residents radiatiod an , larga f o en e numbetherapistsus e th f softward ro an , e programd san hardware devices for graphical -treatment planning. Each step of the complex treatment planning process involve numbesa f issueo r s relevan qualito t t y assurance procese Th . s si represented schematically in Table VIII. Table VIII Treatment Planning Process Process Related QA Procedures Positionin immobilizatiod gan n Port films. Laser alignment Simulation Simulato includinA rQ g image qualit mechanicad yan l integrity Patient data acquisition (CT. MRI. manual contouring) CT, MRI QA including image quality and mechanical integrity (Accurac f mechanicayo l contouring) Data transfe treatmeno rt t planning system entire th f eo datA aQ transfer process, including digitizers, digital data transfer, etc. Definition f targeso t volumes Peer review, e.g., new patient planning conference, chart rounds. Aperture design Independent chec f deliverko y (e.g., port films peed an ,r review) Computation of dose distributions Machine data from commissioning and QA of treatment machines. Accuracy and QA of treatment planning system. Plan evaluation Peer revie f planwo , e.g., during chart rounds. Independent chec radiatioy kb n oncology physicist. Prescription Written, signed datedd an , . Computatio f monitono r units Treatment planning syste . IndependenmQA t check within 48 hours Production of blocks, beam modifiers blocr fo kA cuttinQ compensatod gan r systems. Port film review. Plan implementation Review of set-up by treatment planning team. Chart review. PatienA Q t Treatment plan review. Chart review after new or modified field, weekly chart review, port film review. In-vivo dosimetry for unusual fields, critical organ doses (e.g., gonadal dose). Status check, followup. 245 2 8. Individual Treatment Planning Process All parameter treatmene th n si t plan shoul verifiee db d during first set-u thao y ps an t ambiguities or problems can be corrected immediately. Special care should be taken to assure tha l beaal t m modifying devices (blocks, wedges, compensators correctle ar ) y positioned. Although errors in block fabrication and mounting are often discovered during the review of port films, wedg r compensatoeo r misalignmen mucs i t h more insidious remaiy ma d n an , throughout the course of treatment if not discovered during initial patient setup. A check of th ephysicise setuth y pb t will minimize error undetectee b s y thao t ma te ddu misunderstanding of the physical concepts and details of QA recommendations for individual patient gives si Tabln ni . eDC TablX eI Summary of QA Recommendations for Individual Patients Procedure Recommendations Monitor unit (minutes) calculation Reviewed prior to treatment by an authorized t perforno individua d mdi o initiawh l l calculation wher o ,possiblt nno e (e.g., emergency treatment), then fractio d prio3r o rt n dose bees th o eha rf n o befor delivered% 10 e , whichever occurs first Graphical treatment plan review 1. Reviewed prior to treatment, or when not possible, then prior to 3rd fraction or before dose bees 10th e%ha f n o delivered, whichever occurs first. 2. Reviewed by a radiation oncology physicist who did not formulate treatment plan. Where only one physicist and that person performed the plan, then reviewe anothey db r authorized individual. 3. Review includes calculated monitor units, input-outpu plad nan t quality. . Independen4 t calculatio f dos npointo a t ea : Compar eacr efo h fiel d- wit independenn ha t calculatio f dosnpoino a o et t usine gth calculated monitor unitprescribee th s- d dan calculated dose. 5. If these differ by more than 5%, then the discrepancy shoul resolvee db d before continuing treatment. Plan set-up Radiation oncologist present at first set-up for major changes in treatment. Beam (portal) films—curative and high morbidity risk palliative patients Initial films reviewed by radiation oncologist prio firso rt t treatment additionn I . , ongoing portal films (the standard is weekly) also reviewe radiatioe th y db n oncologist Beam (portal) films—palliative patients Films reviewed prio secono rt d fraction. In-vivo dosimetry l institution1Al . s should have accesO TL o st or other in-vivo dosimetry systems. 2. Shoul usee db measurdo t e dos criticao et l structures (e.g.. lens, gonads). 3. May be used to record dose for unusual treatment conditions. 246 . 9 In-Vivo Dosimetry In-vivo dosimetr usee b identifo dn t yca y major deviation delivere th n si f treatmenyo t an verifo d t documen d yan dose th tcriticao et l structures. Institutions should have acceso st other o D r in-vivTL o systems. Thermoluminescent dosimetry (TLD oftes )i n used because device th smals ei relativel d an l y eas calibrateo yt , while diodes hav advantage eth f eo instantaneous readout These in-vivo systems can have relatively large uncertainties, which should be assessed before using them. While in-vivo systems are useful for individual patient measurements, they should not substitute for an adequate QA program. . 10 Chart Review A procedure for checking the patient charts for the technical parameters of treatment shoul developede db . Belo wgives i outlin n parameterne a th f e o checke e b o st verifiedd dan : 10.1 Review of New or Modified Treatment Fields firse Th charte tasth f tk o revie identifo t ws changei y treatmene yan th w n si ne r o t treatment fields sinc previoue eth s review followine Th . g specific chare areath f t so shoul d be reviewed: a) Treatment Prescription ) b Simulator Instructions -c) Isodose Distribution, Special Dose Calculation Measurement (MinuteU M s ) Calculatedd ) ) e In-Vivo Measurements f) Daily Record 10.2 Weekly Chart Review In additio initiae th o nlt chart chec kweekla y review should take place: a) Review of Previous Fields ) b Cumulative Dose 10.3 Review at Completion of Treatment As a final review before the chart is placed in a file, the following items should be checked: a) Prescribed dose delivered ) b Chart properly documented accordino gt , department policy c) Treatment summary included 11. The Role of the Radiation Oncology Physicist Radiation oncology physicist primarile sar professionalld yan y engagee th n di evaluation, delivery, and optimization of radiation therapy. A major responsibility of the radiation oncology physicist is to provide a high standard of clinical physics and supervision. rolee responsibilitied Th san radiatioe th f so n oncolog youtlinee physicisar A Q d n belowi t . 247 11.1 Specifications of Therapy Equipment The radiation oncology physicist should help define the specifications for the purchase of treatment unit(s), therapy simulator(s), therapy imaging systems (e.g., CT Scanner, on-line portal imaging systems), and treatment planning system. The radiation oncology physicist is involved in the design of the facility and must assure that all radiation safety requirements are met. 11.2 Acceptance Testing, CommissioninA Q d gan The radiation physicis responsibls i t acceptancr efo e testing, commissioning, calibration periodid f therapo an , A cQ y equipment particularn I . physicise th , t must certify thatherape th t y unit plannind san g system performine sar g accordin specificationso gt , generate beam data, and outline written QA procedures which include tests to be performed, tolerances frequencd an ,testse th f .yo 11.3 Measuremen Analysid tan Beaf so m Data Important component commissionine th f so g phase includ t onl generatioe eno y th f no beam dat eacr afo h modalit energyd yan alst bu o, evaluatios it qualite date d th th af an f n o yo appropriateness for treating different disease sites. Such evaluation may lead to the initiation of further measurements and refinements for different treatment techniques. 11.4 Tabulatio f Beano m Dat Clinicar afo e Us l It is the responsibility of the radiation oncology physicist to assure that the beam and source dat correctle aar y entered inttreatmene oth t planning syste thad m printean a t d copf yo the beam data is tabulated in a form that is usable by the radiation therapist, dosimetrists, and radiation oncologists. 11.5 Calibratio f Radiationo n Oncology Equipmen, t primare Onth f eo y responsibilitie radiatioe th f so n oncology physicis assuro t s i t e that l treatmenal t machine radiatiod san n source correctle sar y calibrated accordin accepteo gt d protocols. 11.6 Establishmen Dosf o t e Calculation Procedures A major responsibilit radiatioe th f yo n oncology physicis establiso t s i t dose hth e calculation procedures thausee ar t d throughou department—ane th t assuro dt e their accuracy. 11.7 Establishment of Treatment Planning and Treatment Procedures Along with the radiation oncologist and other members of the treatment planning team, the radiation oncology physicist is responsible for establishing treatment planning and treatment procedures. This includes both the technical aspects of the process (e.g., how block performede cuttinb floo e t proceduref th ws g o i d an ) s entaile procese th n di s (e.g., when different steps in the process of planning are to be performed). 248 11.8 Treatment Planning physicise Th t should perfor overser mo determinatioe eth radiatiof no n dose distribution patientn si s undergoing treatment (i.e., computerized treatment plannin r direcgo t radiation measurements). This includes consultation with the radiation oncologist and the evaluation and optimization of radiation therapy for specific patients. 11.9 Establishmen ProcedureA Q f o t s physicise Th ensurn ca t e thapoliciee th t procedured san s contain proper elementf so good radiation oncology practice, delivery of treatment, radiation safety, quality control, and regulatory compliance (AAPM, 1987). Moreover radiatioe th , n oncology physicist should perform a yearly review of the appropriate sections of the policies and procedures manual. 11.10 Supervisio Therapf no y Equipment Maintenance Regular maintenance of the treatment machines is required and can be overseen by the radiation oncology physicist. Whil supervisine eth g radiation oncology physicist no y ma t perform the actual machine maintenance, he or she is responsible for releasing a treatment machine into clinical service after maintenance, and for documenting that any alteration causemaintenance th y db repaid ean r schedule doe affect acceleratoe sno th t r performancr eo calibration. 12. Discussion The treatment of a patient with radiation is a complex undertaking with multiple steps. An erro miscalculatior ro stey an p effectn ni overale th s l treatment aim qualite Th . y assuranc sucr efo undertakinn ha , thereforegis , complex, involving mane y th step d san medical radiation physicist is often the one person qualified to oversee the total program. Although this paper describes in some detail the various steps for treatments done with medical linear accelerators, essentiall same yth e steps mus treatmentr carriee b tfo t dou s with Cobalt 60 teletherapy, High Dose Rate brachytherapy systems, brachytherapy with sealed source therape th r so y applicatio unsealef no d sources. More e detailfoune b th n n di sca reference particulan i d san AAPe rth M Task grou Report0 p4 ] [1 . Also, even though the QA program has been written as though various computers are available, e.g., for treatment planning or dose monitor calculations, the process applies also for hand calculations thesn I . e case hane sth d calculations shoul checkee db n a y db independent calculation. REFERENCES ] [1 AAPM (1994). "Comprehensiv Radiatior fo A eQ n Oncology: Repor f AAPo t M Radiation Therapy Committee Task Group 40," Med. Phys. 21:581, 1994. ] [2 AAPM (1984). "Physical aspect f qualitso y assuranc radiation ei n therapy." American Association of Physicists in Medicine Report Series No. 13 (American Institute of Physics, New York). 249 [3] ACMP (1986). "Radiation control and quality assurance in radiation oncology: a suggested protocol," American College of Medical Physics Rep. Ser. No. 2 (American College of Medical Physics). [4] BJR (1989). "Treatment simulators," British Journal of Radiology Supplement No. 23. [5] Horton, J. (1987). "Handbook of Radiation Therapy Physics," Prentice-Hall, Inc., Englewood, 1987. [6] ACMP (1991). "Quality Assurance in Radiotherapy Physics Proceedings of an American College of Medical Physics Symposium May 1991," Ed. George Starkschall and John Horton. Medical Physics Publishing, Madison 1991. ] [7 IOPSM (1988). "Commissionin Qualitd gan y Assuranc Lineaf o e r Accelerators, Report No. 54." 250 ACCURAC RADIOSURGERYYN I INFLUENCE :TH COLLIMATOF EO R DIAMETERS AND ARC WEIGHTS ON THE DOSE DISTRIBUTION FOR SINGLE TARGET M.S. PLAZAS XA9642865 National Universit f Colombiayo , Colombia D. LEFKOPOULUS, M. SCHLIENGER Unité de Radiophysique, Service de Radiothérapie, Hôpital Tenon L. MERIENNE Service de Neurochirurgie, Hôpital Sainte Anne Paris, France Abstract The dosimetric characteristics of mini-beams and dose distributions in beams used for radio surgery defer substantially from beams use common di n radiotherapyf o m ai e Th . radio surgery is to deliver a high dose to the lesion in one single fraction, while minimisin e dosgth e delivere e surroundinth o t d g normal brain tissue. This typf o e irradiatio s performeni d wit hnumbea f continuouo r s arcs locate varioun di s corneal (patient sitting r sagitao ) l (patien supina n i t e position) inclined planes usin ga linea r accelerator. A treatment planning system should take into account a large number of irradiation parameters such as the collimator diameter, number of arcs, their angular positions, lengt d e weigharcse analyseth an h W f . o te influenc th d f collimatoo e r X-rayV usin M stereo-tactim d 5 m san g 1 diameter 0 2 o rangt e 6 th cf n se i o irradiatio f no ellipsoidal inclined arterio venous malformations (AVMs) wit hsingla e isocenter. Special arc weights were used to obtain an optimised dose distribution with 13 arcs distributed over an angular sector of 120° x!30 °. In the two studies made we used 3 dimensional dosimetric calculations resulte Th . s wer treatmene useth r dfo f patiento t enabled san d choice optimae th th f eo l irradiation configuratio eacr nfo h patient. I Introduction dosimetrie Th c characteristic f minibeamso dosd san e distributions use r Radiosurgerdfo vere yar y different froe mth radiation beams use clinican di l Radiotherapy. f Radiosurgero Thm eai deliveo t s yi higra h single lesio e dosth on n o ei t fraction, while minimizin dose gth e delivered surroundine toth g normal brain tissue. This typ f irradiatioeo performes ni d wit hnumbea continouf ro s arcs situaten di different coronal (sitting position r sagittao ) l (supine position) inclined planes usin glineaa r accelerator treatmenA . t planning system should take into account a large number of irradiation parameters such as the collimator diameter, number of arcs, their angular positions and angular lengths and the weight of the arcs, in order to obtain dose distributions adapted to each individual treatment plan stereotacticallo .T y irradiate ellipsoidal inclined arteriovenous malformations (AVMs) with a single isocenter needs the special arc weights to obtain a optimise dose distribution. 251 We analyse influence dth diametef eo r collimator c weight X-rayar V usin m d M rang- e m 5 0 an th s( g1 0 en 2 si froo t m6 1.0) using 13 arcs distributed over an angular sector of 120° x!30° (Fig. 1) II Material and Methods A. Material The studies were realized with: -Saturn (GE-CGRMeV3 e4 ) linear accelerato X-rayV M )5 witr(1 h isocentric technique (ASD=100 cm), -Talairach stereotactic fixation system, -a special movable seat (O.Betti hol o patientt pe dth , -3D dosimetric planning system (VaxStation 3200 and array processor Numerix-332 using ARTEMIS_3D planning program) with a dose matrix of 100mm x 100mm x 100mm, -collimator diamete 6,8,10,12,14,16. : r mm 0 2 d an 8 1 , B. The 3D treatment planning In Radiosurgery, the treatment planning is a fundamentally three-dimensional task. The ARTEMIS_3D treatment planning systeemphasin a s mha thren so e dimensional computation graphicsd s an bee s ha nt I . develope Tenot da n Hospitas i t i d lan based on the general formulation proposed by Siddon 1. The Tenon methodology the minimal therapeutic dose is delivered at the periphery of the lesion. It corresponds usually to the 60%-70% isodoses. The dose contribution from each arc is simulated by stationary beams separated by 10° increments, after which the computer system retrieve necessare sth y percentage depth dos dose eth edatd profileaan calculated san dose sth e distribution on a 64 x 64 x 64 mm or 100 x 100 x 100 mm 3D dose matrix. The system allows qualitative evaluation, such as isodose surface displays and quantitative evaluation such as dose-volume analysis. . ParamentersjQTjhC e evaluajiojLof dose distribution differene Th t mean usee sb demonstrattha o dt n tca qualit e treatmene eth th f yo dimensionalte planon e sar dimensionao tw , l and three dimensional. Taking into accoun importance th t functionae th f eo l structures surroundin lesione gth propose ,w e in these studie dose-volume sth e analysis, becaus planae eth r dose distribution (coronal sagittal transversad ,an l planeso )d not give a complete evaluation. Dose-volume histograms are a useful feature of treatment plans and complement the information coming from planar dose distributions. To check the treatment planning quality by assessing the integral dose potential variations insid outsidd ean lesione eth employede ,w : (a) The reference isodose surface volume (VRI) which represents the lesion volume. The reference isodose volume represents the therapeutic pan of the irradiation. (b) The volumetric dose fall-off(VFO)Ri_iL% is uie difference between the reference isodose (RI) and an arbitrary inferior limit isodose (IL) volumes. In these studies the RI is the 70% isodose and the inferior limit is the 10%.isodose. It represent possible sth e damag neurologicae th o et l healthy structures. (c) The ratio of the volumetric dose fall-off to the reference isodose volume (VFIR)Ri_rjL% expressed as the "multiplying factorhealthe th f "o y irradiated tissues volume compare lesioe th o ndt volume representt I . volume sth f healtheo y tissues which should be irradiated in order to achieve the planned lesion treatment. It is a measurement of the radiosurgical treatment quality outsid lesione eth shoule On . d selec smallee th t r VFIR irradiation configuration whe nchoica e between differeno thetw t plans wit same hth e reference isodos emadee volumb o t .s eha 252 Metho. D d presene W studie2 t s analysin influence g th collimato e th f eo weighc dose r ar diamete th e n distributionth to d ran . 1. Collimator diameter influence Th f eigheo t differen diametern i t collimatorm m 0 )2 oveo t angulan 6 ra s( r irradiation secto f 130ro ° usin3 g1 arcs spaced by steps of 10° (total angulation of 120°) weight2.Arc This stud s base referenccentee e considereywa s clinicath th n d o f n wa i ro e M eb l o stereotacticasesdAV t e Th . c frame (Ant=0 mm, Right=0 mm, Sup=0). We have studied the relationship between the AVM long axis and the superior-inferior direction angle ß and the weight of the arcs. The irradiation space is represented geometrically in the same conditions to collimator diameters study (120° x 130°). 0* -10- I FIG 1. Irradiation parameter a centrall r fo s y situated anterior- superior/ posterior-inferior ellipsoidal inclined AVM: weighte3 (a1 ) d arc degree0 1 s s separated (b) Angula degreer0 legt13 , f eacho sc har 253 To obtain inclined isodoses encompassing the ellipdcal shaped AVMs witb different inclination angles on the sagittal angiogram, each arc should be differently weighted. We defined as "weighting vector" the vector having as elements the w; weightc ar . s 1) (Fig. Ill Results 1. Diamenter colimator -The 70% reference volume isodose ( VRI) increases when increasing tbe collimator diameter (Fig. 2) -The increase of VFO70-10% is proportional to collimator diameter (Fig. 3) while tbe VFIR70-io% is inversely proportional (Fig. 4). volu • 10 Bo -12 0 3 -fl t I -t l Collator di«Mt«r (••) ne 2. ric i. 70% R«f«r«nc« iaodose volime (target volune) Volusetric fill-oft (70«-10%) ou ta id« th« target. V.F.I.R 3.O Collaator di«Bot*r (BB) Til* ratio of volumetric fall-off (VFO) by r«f*r«tc* isodoaa volvnM (70%) 254 Weightc Ar . 2 We studieweightsc ar e dth , with seven linear weighting vectors (LWV LWV7o 1t ) havin gsimpla e shape e (FigW . .5) 0.1choso t extrems ,0 a e1. e value thesr sfo e veaors parametee .Th r representing vector thesV e LW thei s si r angleot(Fig. 5). -The minimal inclination isodoses;<=0o and the maximal inclination isodosesoL=38° in the sagittal planes after calculation of the dose distributions are presents in Fig.6 «-0 •C-28- «-32- 1=08' »6 Angle ^f arctherapo y plvw. FIG 5. Linear weighting vector« LWV a« f une it ion of their oc angle. (From LHVl: cj-0' to LWV7:«-38') (A) (B) rie «. Relative dose distributions in the : (a) coronal and (b) sagittal planes, corresponding to the linear «eigting vectors (A) LNV1: t -o'• , /»-o ) LWV?(B r : «.-38*. A-2S*. 255 These dose distributions permit the calculation of the realtionship between the isodose inclination angle /2>and the LWV angled (Hg.7). -Figur showe8 volumee sth s enclose 10%e th isodose y % differen,de b 30% 90 th r d sfo ,an 50% t LWV% ,70 volume e .Th s encompasse isodose e remaith % y 90 d b no st relativelfro% m30 y constantisodos% 10 e e th , volume increases slightls ya the angle increases. -The VFO70-iO% remains unchanged and the VFC>30-10% presents a variation of 10.8%, passing from 39.7cm3 (LWV1) to 44.3 (LWV70cm ) (Fig. .9) -The relationships betwee deliveree nth d doses insid outsidd ean lesioe eth n (VFIR70-io% changt )no e significatively 18,3 (LWVl)tol9,8(LWV7). 30% * 40 ' 30 * 20 10* Linear weighting vector* angles (<*) FIG 7. Relationship between the isodoses Inclination angle ft and the linear weighting vector», angle ot. 50 -, 4. 0 • um,• o*. . o- 01MV2, •7*. .2- EÜLMV3, .5- 0LWV4, -«• DLWVS, «28-, - 12" 20 . • LWV6, -32». -20- BLWVT, «38-. .25' % 10 % 30 % 50 70% FIG B. The 90%, 50%, 30% and 10% isodoae volusea for 13 arctherapi« corresponding to the seven linear weighting vectors «AW. 256 50-1 70%-10% 30%-10% 70%-30% leodooo volume PIC 9. Volumetric dos« fall - off« 70%-1O%, 30t-10% and 70%-30% and the 70% reference ioodode voluaa for 13 arctherapie« corresponding to th« nevan weighting vectors (LWV). IV Conclusion 1. Diameter Collinwtor -The reference volume (target volume VRl7Q%) increases when increasing the collimator diameter. -The VFIR7Q.io% shows that the doses outside the lesion decreases when increasing the diameter collimators. 2. Arc Weight -The applicatio lineae th f rno weighting vectors produce inclined isodoses characterize theiy db r inclination angteyg'. This last angle must coincid lon M superior-inferioe geth axi AV wit d e shan th r stereotactic direction angle. Studyin gfamila f yo linear weighting vectors we have found the relationship existing between the^Cand ^angles. This result (Fig. 7) constitutes an "a priori knowledge" which can be used in a dose distribution optimization procedure permitting the reduction of the treatment plan preparation. -The dose distributions resulting from the application of the different LWV show that the isodoses volumes remain constant isodos% 10 excepe e th volumr fo t e which increases slightl angleoe th s ya ( increases. -The dose delivered 10 the healthy tissues (Fig. 9) VFO70-30% remains unchanged whereas the VFO30-10% increases for the angles superio 28°o rt . Clinical experience shows that most AVMs can be classified in categories according to their shape and topopgraphy. To the inclined AVMs, we have found a relationship between the isodose inclination angleß, and the weights of the arcs. 1 Siddon R. "Solution to treatment planning problems using coordinate transformations", Med.Phys. 8, 766-774, 1981. 2 Lefkopoulos D, Schlienger M, Plazas M.C, Merienne L, Dominique C, Laugier A. "3-D dosimetry for radiosurgery of complex AVMs" Proc 10th Conf. on the Use of Computers in Radiation Therapy ICCR, Lucknow, India, Nov. 11-14, edite . HukkS P.Sy d b .u an lyer (Lucknow, 1990) 254-257 Bett3 Derechinsk, iO Irradiatio" . iV n stéréotaxique multifaisceaux", Neurochirurgi , 295-298e29 , 1983. 4 Lefkopoulo , SchliengeD s , Toubou dosimetrrM D "3 . E l radiosurger r yfo y methodolog SALe th f yTo grou comr pfo - plex arteriovenous malformations", submitted to Radiotherapy and Oncology, 1992. 257 5 Podgorsak E.B, Pike B, Olivier A, Pla M, Soubami L. "Radiosurgery with high energy photons beams : A comparison among techniques". Int. J. Radiât. Oncol. Biol. Phys. 16, 857-865, 1989. . Phillip6 s M.H, Frankel K.A, Lyman J.T, Fabrikant J.I, Levy R.P. "Compariso f differenno t radiation type irradiad san - tion geometries in stereotactic radiosurgery". Int. J. Radiât. Oncol. Biol. Phys. 18, 211-220, 1990. 7. Graham J.D, Nahum A.E, Brada M. "A comparison of techniques for stereotatic radiotherapy by linear accelerator based on 3-dimensional dose distributions", Radiotherapy Oncol. 22,29-35, 1991. . Schel8 l M.C, Smit , LarsohV , FlickingeA n u D.AW , r J.C. "Evaluatio radiosurgerf no y techniques with cumulative dose volume histogram linac-basen si d stereotactic external beam irradiation". . IntRadiatioJ . n Oncol. Biol. Phys , 132520 . - 1330, 1990. 9. Plazas M.C, Lefkopoulos D, Ozsahim M, Schlienger M, Laugier A. "The influence of two irradiation parameters on the dose distributions for single target AVMs", J. Int. Fédérât Med. Biomed. Engin. 29 (2), 1086 (1991). 10. Plazas M.C, Lefkopoulos D, Schlienger M. "The influence of arc weights on the dose distribution for single target ra- diosurgery". Med. Phys (5)0 2 ., 1485-1490, Sept/Oct, 1993 258 DOSIMETR BREASF YO T CANCER XA9642866 G. RAMIREZ C., I RESTREPO, C.A. AGUIRRE Hospital Universitario del Valle Call, Colombia Abstract The systemic therapy of breast cancer has also changed profoundly during the last 60 years, and in this time the integration of treatment modalities involve a major area of investigation.(l). The dosimetr breasf yo t cancer presents different complications whic rangn hca e fro Physician'e mth s handlin neoplasi simple e th th f go o t e aspectp au physicaf so l simulation, contour designs, radiation fields, irregular surface computed an s r programs containing mathematical equations which differ little or largely with the reality of the radiation distribution into the volume to be irradiated. We have studied the problem using two types of measurements to determine how the radiation distributio irregulan i s ni r surfaces designind an , easien ga rusee b skil do t l with each patientn i , orde optimizo rt treatmene eth t with respec simulatioe th o t verificatiod nan n process. 1 MATERIAL AND METHOD: To measure the distribution of radiation in a volume of water we used a Wollhofer Phantom, to which was added an acrylic surface with a curvature similar to the Costal wall of a patient who has had a radical mastectomy performed. Moving the ionization chamber detector by means of a program that generates rectangle trajectories of different longitudes, like an integral step by step way, we got the isodose curves. These curves generated are the most exact distribution to a real measure within an irregular volume, just like the volume to be irradiated in the practice. At the same time we initiated the film densitometer calibration of a Kodak X-OMAT V verification mora fil e ordemn ei us preciso rt e dosimetr eacn yo h patient startee W . d constructing witx hwa (densit 0.9y= 4 gr./cc) ste,a p wedg generato et characteristiea filme th ,f o whic ) D chd curvan H e( irradiates wa d locatin stee gth p wedge ove irradiatefild e mth ran d wit hCobala t sourc6 d ean MeV, x-ray. A Theratron-780 machine with a cobalt source of 2 cm of diameter and a Siemens Mevatron 6740 linear accelerator were utilize irradiato dt watee eth r phanto filmse constructewelme s th a s W .a l d a wax phantom with the same form of a chest wall. The films were cut to form the figure of a chest wal orden li irradiato rt e specificall tangentiao tw e yth l breast fields onee whicth se thahar t present the major problems of coincidence. Risk for tumor recurrence is a prerequisite to designing a radio- therapeutic, where the structures potentially at risk include the breast, the chest wall, the ipsilateral axillary, supraclavircula internae th d lan r mammary lymph nodes (2)additionallyd an , , thae th t target tumor areas stays withi proposee nth isodos% d80 e line. Equipment's use measuro dt radiatioe eth n wer Campintee eth c 192X type WK92.WA10, wite hth ionization chamber PR-05 type 1C-10 serial 477 (3), (4), (5). tangentiar fo Radiatiom c 2 l1 fieldx n m sbeam c witwithoud 6 f han so t beam modifiers, like eth breast cone manufactured by Theratronics, 45 and 60 degrees for the Mevatron 6740, to get the isodose ove irregulae rth r surface entrance were used. 259 These fields can take one of the three different forms. First, one can use standard tangential fields patiente th lind seconthae f eo mi Th t . e ente dexith d techniqut deee an a rt us po t tangentia s ei l fields that enter and exit contra laterally across the mid line of the patient. The third technique matches shallow tangentfacn e en a interna o st l mammary (IM) lymph nodes orden I . visualizo rt e whethe e deeth r p tangential fields encompass these eobtain scath T pointsca C n ni na e on , treatment position with marker entrance th n si exid ean t point manualld san y connect these points with a line to see if the fields are deep enough. Alternatively, on a simulator film the tangential fields shoul seee db pasno t s throug sternume hth . deef o The pe us tangentia l field drawbackss sha . Foremost among increasethee th ms i d amounf o t lung tha includes i t d withi tangentiae nth l field thes directesa e yar d more deeply int patiene oth n i t order to treat the lymph nodes. Pneumonitis is a direct function of the amount of lung irradiated, sincd (6)difficuls i an , t e lympi M treaI o t te h tth node mosn si t patients without subtendin t leasga t 3.5 cm into the lung, the risk of radiation Pneumonitis from such treatment scales rapidly (7). irradiate e filme b o Th t s d were place slicesfor"sandwich"e x a df th m wa o n betwee i ,o tw e n,th which are used to represent the patient's tissues, and irradiated with two tangential fields. The films were developed under controlled conditions of darkroom and film processors, they were read wit filha m densitometer Macbeth TD-528 conneco t , pointe tth s wit same hth e optic density, that represent the levels of radiation and generate the isodose of the different field combinations studied by us. 2 CONCLUSIONS: direct/indirece Th t measure levele th distributiof f so o radiatiof no irregulan i r e surfaceth s i s sa breast and Costal walls, permits verification that the proposed dose or dosimetry, complies with wha proposes ti Medicae th y db l Docto treatmene th r ro t protocols. With respec measuresr visualizou n o ca t e w ,e that accordin classificatioe th o gt n histologicaf o l the disease (C. A,), and the target areas (tumor bed and internal mammary lymph nodes), as well as healthe th y tissues that shoul preservee db diminisd dan morbidite hth patiene th f yo t (lungs, heart and spinal cord), the best options are presented in the distribution given when the treatment equipment is a 6 MeV Mevatron. mose th t t importanBu t thin foune gw thin di confidentialite s worth s kwa measuree th f yo s wite hth films since no differences larger than 2% were found in the whole range of measures of the film, surfacclose th eo welt s e a deeps a l . Also important is the facility of how to construct the wax phantom of a patient's contours and verifying if the entrance angles of the fields, as well as the modifier implements (filter wedges, cone and bolus), presents the best real option of the treatment proposed in patients to which a radical or modified mastectomy has been effected. 260 REFERENCES: 1. Allen Recht : EditorialMD , , Conservative Managemen f Breaso t t Cancer. Seminarn i s Radiatio , 1992n2 Oncologo N , , W.B2 l yVo . Saunders Company 2. Allen S. Lichter, Benedick A. Fraass and Bethany Yanke, TREATMENT TECHNIQUES IN THE CONSERVATIVE MANAGEMENT OF BREAST CANCER, Seminars in Radiation Oncology Vol 2, No 2, April 1992. 3. CAPINTEC EXPOSURE/EXPOSURE-RATE METER, Model 192, Operation Manual Montvale, New Jersey 1976 . 4 Ministeri Industrie od Energiaay . CLASFICACIO INSTRUMENTOS LO E ND E SD METROLOGIA DE RADIACIONESIONIZANTES. Espana 1985 . 5 Arata Suzuki, Marci . SuzukiaN , CAPINTEC IONIZATION CHAMBER PROBR EFO EXPOSURE MEASUREMENTS. Owner's Manual Montvale Jerseyw ,Ne , 1976. 6. Schoeppel SL, Fraass B A, Lichter AS: Radiation Pneumonitis after Treatment with Tangential Breast Fields: A Dose Volume Analysis. Proc Eur Soc Therap Radiol Oncol, Pag , 1988.41 . 7. Fraas AsB , Roberso , Lichte n ContralateraPL e : Dos th AS ro e t l PrimarBreaso t e eDu y Breast Irradiation, Int J. Radiât Oncol Biol Phys 11:485-497, 1984. KrXY PAOSÎSJ I i-v?» r- v» l 261 PLATON V2.0 & BRA Vl.O SYSTEM FOR TELETHERAPY AND BRACHYTHERAPY XA9642867 C. AKTES, G. COSCIA Argentina A. LUONGO Institute de Oncologia, Depto Radioterapia, Ministeri Salue od d Publica, Montevideo, Uruguay Abstract locallA y designed fully automatic Radiation Field Analyser (RFA constructeds )wa systee Th . ms i controlled by a PC and includes a graphic system ionisation chambers and an electrometer. The syste mcapabls i readinf eo g doses instantl poiny an tyn i insid wateea r phanto providd man e graphs dose th ef o distributions (isodose curvestherapeutiy an f )o c unit informatioe th , automaticalls ni y stored in the PC and can be transferred to Treatment Planning Systems (TPSs) such as the PLATON and BRA developed in Latin America. INTRODUCTION. During the last decade, there has been a vast projection of Radiotherapy on the general strategy of cancer treatment. Thidimensiow sne radiatioe th f no generalisatioe nth treatmeno t e du s f i tno computers, planning systems and to the modern advances in Brachytherapy (new sources and equipment). As we proceed towards the 21st century, the goal is to optimise the quality of treatments improving their design, developing computerised systems for medical techniques. In this setting we worked har manr dfo y years tryin develoo gt desigproduce e pth th f nintroducino e tar tha e w t g todan yi this meeting: Platon V2.0 & BRA Vl.O. This product has been created in Latin America, more precisel Argentinyn i Uruguayn i d aan . Platon V2.0 replaces version 1.0, giving additional ways of treatment and a new interface with the user. This new system is completely graphic and operates by the use of a personal computer not demanding special hardware, the maintenance and availability of the equipment is easier than with other ones because they are standard, it can be configured for a single operator or may be part of a multiple workstation network design to optimise the effectiveness of the work group within the department. The integration of all the patient database is possible and readily accessible, h is designe Spanishdn i , Portugues Englisr e o doesn't i d han t require previous experienc usere th r .efo same Ath t e tim providet ei greasa t variet therapeutif yo c modalities thatoole ar tf calculatio so n and analysis suc irregulas ha r tangentia rotationad an l l fields, beam' viewe sey , three dimensional calculations, irradiated volume calculation representatiod san regionf no interestf so . 263 2 MATERIAL AND METHODS thay sa t n Platoca e n W V2.0 combines clinically oriented functionality with proven treatment planning methodologies, providin professionae gth l wit hvera y fast, friendl simple d us an y o et decision support system .systef Througo e mus e controhth l files operatine th , g characteristicy sma be adapted to the department's work load and professional criteria to maximise the usefulness in documenting and evaluating alternative treatment planning approaches. Wit representatioe hth thie se s treatmene n systeth ca f ne o m w t plan characteristicse , witth l hal . functiop beae u t Th mse n reflect evere clinicay sth yda l realitie specifiee l beamb al y s- y sma db their numeric parameters and/or positioned on the contour slice visually. Otherwise, using a beam's eye view, collimator rotation, couch (isocenter) position and block placement over several contour slices may be interactively adjusted. A treatment plan may have up to 5 separate patient contour slices defined and each contour slice ma contoury5 havo t p eu s with different densities. Also we can see isodose maps with the use of the different fields and weights. The estimation of irradiated volumes and integral doses are also in this system. In a histogram way we can see the confined volume in each isodose curve, the median dose and the integral dose. Distance and angle measurements, critical organ dose monitoring, isocenter localisation, dose normalisation, zoom, mantle fields and irregular fields, are able to be visualised on screen. BRA 1.0 is the system that allows planning of brachytherapy treatments. Its design includes a variet f ycalculatioo evaluatiod nan extremels i n t i tool d san y eas useo yt . This system calculates dose distributio thren i e dimension radioactivr sfo e seeds, line r ribbonso interstitian si intrad an l - cavitary brachytherapy. Planes, cubes, cylinders automaticalle coneb d y an , sma y constructea y db few simple commands combinatioy an d an , f sourcno e type s possiblesi proue ar f thie do sW . product and we can assure that our country is able to compete in the same line of the most developed countries. Perhaps we can find similar or even better equipment, but our achievement is to have obtained a great design on less than a tenth of the cost of any other system on the market with the same quality standards of the most sophisticated and expensive ones. ultimatr Ou e purpos t onl producr no sel o yt s obtaie o i ou lt t greatese nbu tth t credibility fror mou reae th l t technologicacolleaguesge o t s a o s , l exchanges which will alloachievo t s wu greateea r growth woule .W d lik proposeo t internationae eth l organisation idee buyinsf th ao g pricethiw slo d system and to give it to the professional teams of the countries who might be in greatest need of them. 264 3 CONCLUSIONS In Latin America there are more than 200 tele therapy machines using gamma sources. For example internationan a f i , l project distribute mentioninge system0 ar kin e 10 se th d w f so , installs them, trains people for their correct use and if this investment is below the cost of one high energy linear accelerator, the impact in the change of radiotherapy treatment quality would be very important. And it could be easily financed by an international organisation. In this way we will consciously increase the quality of radiotherapy around the world, working all together, integrally universaa n i , l project remembering tharelevanr ou t t purpos e curee w th f s I .ei can understand that we are different but that we are all branches of the same tree, it will be the greatest achievement of this meeting and a indicative point of professional and personal growth. 265 THERMOLUMINESCENCE DOSIMETRY APPLIED TO QUALITY ASSURANCE IN RADIOTHERAPY, BRACHYTHERAPY RADIODIAGNOSTID AN C l iiiiimi i i XA9642868 G. MARINELLO Département inter-hospitalie cancérologiee d r , Service de radiothérapie. Centre hospitalo-universitaire, Créteil, France Abstract Thermoluminescence (TL) dosimetry is very interesting for in vivo measurements because TL detectors have the advantages of being very sensitive under a very small volume and do not need to be connecte electrometen a o dt r wit unwieldn ha y cable e principl e methoTh .th f o ed being briefly recalled, criteria of choice of a TL material according to the applications to be performed are given. It is shown that to be used for in vivo measurements, TL material should have he same response at roo patiend man t temperature equivalene b d san sofo t t t tissue, lung energe r boneo sth r ysfo ranges encountered in practice. Theoritical data are provided in order to facilitate the user's choice. The different heating processes (linear or isothermal heating kinetics, hot gas, etc.) and light detection systems of TL readers are also presented. TL manual and automatic readers commercially available in 1994, and the emission temperature and wavelength the dosimetric peaks of usual TL materials are presented in two tables, respectively. The principae nth l propertie dosimeterL T f so viv n usei e r b odo fo s t measurement theid san r practical consequence e summarizedar s : signal stability after irradiation, intrinsic precision, sensitivity, response with dose, dose-rate, mas energyd san last .A t some example f applicationso s sa differen s totaa t l bodd skian yn irradiations, brachytherapy, diagnostic radiolog d qualitan y y assurance purpose givene sar . . INTRODUCTIO1 N Thermoluminescence dosimetry [1,2 bees ]ha n developed considerably ove pase th r t twenty years e commerciath , l availabilit f reliablo y e detector materiale commercializatioth d an s f o n automatic readout systems bein gdecisiva e factor widA . e materialchoicL T e forf th eo f mo n i s powder, microrods, pellets, etc allow the dosimetry to be adapted to applications as different as radiation protection, radiotherapy, curietherapy, diagnostic radiology and quality assurance purposes suc calibratios ha f treatmeno n t unit d radioactivan s e sources, verificatio f computeo n r programs, validatio protocolw ne f no s before clinical use vivn i , o dosimetry eithe r particulafo r r technique deteco t r o s t error individuan si l patients. For in vivo measurements (§ 6) TL dosimeters are competitive with other detectors [3, 4] and have the advantages of being very sensitive under a very small volume, to be tissue-equivalent and not to have to be connected to an electrometer with an unwieldy cable. As for time required for readout, it can be considerably decreased by a good choice of the equipment and a good methodology [4 ]. 2. PRINCIPLE OF THE METHOD Thermoluminescence dosimetry (TLD )bases i d upo e abilitth n f imperfecto y s crystalo t s absor stord energbe an eth f ionizinyo g radiation, which upon heatin s re-emittegi e forth f mo n di electromagnetic radiation, mainlvisible th n yi e wavelength lighe Th . t emitte s thedi n detected an d correlate absorbee th o dt d dos materialeL receiveT e th .y db Many general theoritical models have been postulate t o w explait stildbu no l, it n difficulties arose when specific dosimetric materials are considered [1, 2]. One of the possible 267 BEFORE IRRADIATION READOUT IRRADIATION (HEATING) CONDUCTION BAND CO ce o FORBIDDEN GAP TRAP mCO t- UGHT t- co VALENCE o BAND cc z m ELECTRONS Fig. l : A possible mechanism for thermoluminescence. After [4). developee b mechanisy ma referiny L db bane T r th m dfo o g t e theormulti-atomith f o y c crystalline structure (Fig. 1). Energy states in a crystal can be represented with energy increasing upwards along the ordinal e: irradiatio n produces free electron holesd electrone san Th . free traveo sar t e l through conductioe solie th th n di n banshora r dfo t time ultimalle .b The y yma y either trappe t defectda s (i.e at a metastable energy state) or fall back into the valence band and recombine either radiatively (fluorescence radiativeln no r o ) y with holescapturee b r o , t luminescenda t centers already activated by holes as a result of the irradiation, and deactivate the center with the emission of light. Under the heating effect the electron trapped at the metastable energy states are given sufficient thermal energy to escape from the trap into the conduction band again, where they are free to travel and have three possibles fates, as before : - they are either be retrapped at defects falr o l- int valence oth e ban recombind dan e radiativel radiativeln no r yo y with holes, - or recombine radiatively at a hole-activated luminescent center. The light emitted by the last process is thermoluminescence (TL). Heating and light collection are performed in a readout system called reader (§ 4). The TL signal as a function of temperature is of complex nature and is called a thermoluminescence spectra or a glow curve. It consists of different TL peaks, each peak corresponding to a different energy state in crystale materiath L dependd T e an , th l n (naturso annealind ean g proceduresirradiatioe th n o d n)an sources (Fig. 2) . After readout the TL material it is either entirely in its original state, and in this case it is just ready for re-use, or it requires a special heating treatments called annealing in order to restore it to its original state. When these treatments are not performed the sensitivity and background of TL dosimeter e considerablar s y altere d theian dr dosimetric propertie t remaino o nd s constantr Fo . instance CaSO requirF ^ CaFLi d e2 an annealin g procedures depending upo fore th whicnmn i h they have been manufactured [5]. On the contrary Li2&4P7 either doped with Mn or Cu can be used and re- used many times without thermal treatments between the successive irradiations and readouts. MATERIA3L . T CHOIC E TH F ELO Most commonly used TL detectors are obtained by doping phosphors such as lithium fluoride (LiF), lithium borate (LÏ2B407), calcium sulphate (CaSO4 d calciuan ) m fluoride (CaF2) with impureties called activators: e.g. LiF:Mg-T lithius i i m fluoride doped with magnesiu titaniumd man , u is lithium borate doped with copper, etc. All TL materials are available either in the 268 ~ 1.2 T Neutroe-lrradlited TLD-600 Fig Thermoluminescenc: 2 . e spectr 600F dependt Li I . f aprevioue o th n so s annealing proceduren o d san irradiatioe th n sources :neutro d X-rayan ] s[A n beams [B]. After [6]. form of powder or of solid dosimeters. The solid dosimeters may be made entirely of phosphors as single crystal r polycrystallino s e extrusions (extruded rods, sintered pellet r s chips)o a s r o , homogeneous composit phosphoe th f eo r powde .somd ran e binding material t shoulI . e notedb d that characteristice th pure th ef s o phospho r dosimeter considerable b y sma y different from e thosth f eo composite. To be used for in vivo measurements TL material should : hav- ehiga h sensitivity unde verra y small volume - have the same response at room and patient temperatures - be equivalent to soft tissue, lungs or bones for the considered beam and in the energy range encountered in radiotherap r radiodiagnosticyo . 269 The use of theoritical data such as Tables 1 or 2 for photon beams, or Fig. 3 for electron beams, allow gooa s materiaL d T usechoic e e b th do lt f accordineo applicatione th o g t practicn i t Bu . e influence th surroundine th f eo g material (build-u patient)d shape e an th th p sizf e d eo pca th ,e an TL dosimeters havtakee b o et n into accoun energr fo t y correctio 5.7)§ n( . ELECTRON ENERGY(MeV) Q75 0.1 Q2 0,5 1 10 20 50 10C Fig : mase Rati3 . th f soo radiative stoppin materiaL gmase T poweth e so th t l radiativ n ri e stopping powe dosimetern L wateri T r rfo s irradiated with electron beams. Curves have been obtained from I.C.R.U. N° 35 Report data [7]. materialTablL T : e1 s equivalen sofo t t tissu photor efo n beams. Photo-electric Compton effect pair production P TL effect PHOSPHORS Zeff e"/g Zeff g/cm3 LiF (Mg/Ti) 8.2 0.83 7.50 2.64 LiF (MgJi,Na) 82 0.83 7.50 2.64 Li2B4U7:Mn 7.4 0.87 6.90 2.30 Li2B4U7:Cu 7.4 0.87 6.90 2.30 Soft tissue 7.42 1* 6.60 1.04 Air 7.64 0.90 7.36 0.0013 270 materialL TablT : e2 s equivalen bono t e for photon beams Photo-electric Compton effect pair productiop n TL effect PHOSPHORS Zeff e'/g Z eff g/cm3 CaSO4 : Mn 15.3 0.90 2.61 CaSO4 : Dy 15.3 0.90 2.61 CaF2:Mn 16.3 0.88 3.18 CaFy D 2: 16.3 0.88 3.18 Bone 14 0.94 10 1.01 to 1.60 4. HOW TO READ OUT AN IRRADIATED TL MATERIAL ? principle Th readerL T f e o shows si Fign i . The.4 y mainly consis heatina ligha f o td t an g detection systems. 4.1. Heating system Different heating systems are encountered in TL readers commercially available (Table 3). All of them offer the possibility to heat the TL dosimeter at two different temperatures: the preheating temperature use cleao dt r unstabl readoute peakth d san temperature use colleco dt e th t information from dosimetric peaks. The readout chamber should be continuously flushed with nitrogen gas in order to reduce spurious phenomena [1 ] and then decrease the background. The metallic support containing TL material may be heated by an electric current, or contact with ovens, or a hot finger moved by a lift mechanism. Generally the temperature of the support is measured by a thermocouple in close contact with it. Then heating kinetics are either linear or isothermal materia.L WheT linears e i t th progressivelns i i l , y heated until preheatin d readougan t temperatures. When it is isothermal the TL material is quasi-instantaneously heated by isothermal ovens to both these temperatures and generally readouts take less than 10 seconds. In any case dose contacts between TL dosimeter, support and heating system are necessary to obtain a good reproducibility (§ 5.2). Soilne no-contact procedures, such as heating by hot nitrogen gas or air, optical infrared heating usin intensn ga e light pulse fro halogen ma n lam heatinr po alsn lasea e ca oy gb ) b r bea10 , m[9 used. In these cases the heating kinetics are particular. Readers which are designed for the readout of a great number of dosimeters in a short time have generally an isothermal heating kinetics [11, 12 ] or heat the TL dosimeters in hot gas. 4.2. Light detection system Generally the luminous flux emitted by the TL dosimeter is collected and guided into the PM with a light guide leading to one or several filters placed in front of the PM window (Fig. 4). These filters have to be adapted to the spectral response of the PM and to the wavelength of the light emitted by the TL material to be readout (Table 4). The response of the P.M. depends on the compositio photocathode th f spectrae no th n o ld transmissioean tube th ef no window mosn I . t readers photocathodes are bialkalis with a peak sensitivity around 400 nm, in good agreement with the blue respectivelym n 8 t wit36 no r Li2B4O emissiod ht o an LJ2B4OF bu ) (emissiou Li m n f C nyo : 0 740 t na 271 glow curve nitrogen dosetneter oven thermocouple heating cycle controller T'C Sionai . Principl 4 readerL FigT . a f e.o After [8]. Tabl : Som e3 reader L eT s commercially availabl 1994n ei . Afte] 4 [ r Manufacturer country model heating detectors manual / system automatic Panasonic Japan UD-513A hot gas rods + P. G manual Harshaw USA 3500 heated support S.D. +P. manual Harshaw USA 5500 hot gas S.D.- automatic Teledyne USA 310 heated support S.D. + P. manual Isotopes Alnor Finland DOSACUS hot gas S.D. automatic or manual Fimel France LTM heated support S.D. + P. manual Fimel France P.C.L isothermal S.D. + P. automatic ovens • P. : powder soli: S.Dd. detector P.G. : powder within glass 272 doped with manganese (emissio nm)0 60 . t na Nevertheles s adequate filters correctely chosen ca n improve considerably the response. A good reader should allow a quick interchange of the associated filter ordeadaptabln e i sb o t r differeno et materialsL X t . Tabl Emissio: : e4 n temperature wavelengtd san mose th f t hstablo e peaks use dosimetrn di f o y different TL materials. WAVELENGTH (nm) EMISSION 368 380 400 450 478 480 500 600 TEMPERATURE and and (°C) 571 577 80-90 CaS04:Mn 200 CaS04Sm 200-240 CaF2:Dy 210-220 LiF Li2B407:Mn 220-250 CaS04:Tm CaS04:Dy 240-270 Li2B407:Cu 260 CaF2:Nat 300 CaF2:Mn Depending upoe readou type th nf th eo t system used signae th , l proportionae th o t l light emissio eithes n i integrato n a ro t amplifie d fe r d (d.cdan operation regimen) r converteo , d into sealea o t rd pulse(pulsfe d san e counting regimen) [8]. Irrespectiv regimee th f eo voltage ne th th f eo P.M. tube has to be correctly stabilised in order to get a good reproducibility of measurements (§ 5.2). Results which correlatehave b o et absorbeo dt d dos eithee ear stored r an rea y t db dou the operator, or safeguarded on the hard disk of a computer and most often printed. In many readers glow curves are also displayed during dose measurements so as to provide a maximum amount of information. . DOSIMETRI5 C PROPERTIES The principal properties of TL dosimeters to be used for in vivo measurements and their practical consequences are summarized. For more detailed informations the reader could refer to genera ldosimetr bookL T n o s "Methodso yt r [1,2vivon i o ] r fo dosimetry externaln i therapy" [4]. 5.1. Signal stability after irradiation An important consideration in the choice of a TL dosimeter is the stability of the signal. In particular it is necessary to assess whether the charges trapped during the irradiation have t beeno n lost befor readoue eth unwantey tb d exposur heao et t (thermal fading), light (optical fading) or any other factor (anomalous fading). This is expressed by a decrease of the TL dosimeter response dependin delae th yn g o separatin g irradiatio readoutd nan . 273 An appropriate preheating allow e eliminatioth s f thao n e signat th par f o lt (low temperature peaks) which presents an important thermal fading, and then reduces considerably thermal fading for most TL materials. In practice thermal fading should be evaluated on each individual reader with the TL material which is intended to be used. It should be approximately of r 1month% pe e differen th r less o r , fo , t preparation wheF Li nf o scorrec t readou d annealinan t g conditions are reached [1, 2, 13]. For l^I^O/ it varies from 0,5 to 1% per week depending upon the doping [1, 14]. When a long delay separates irradiation from readout, a fading correction may be necessary. Optical fading can be avoided by manipulating the dosimeters in a room illuminated with incandescent ligh wrappind an t g the opaqumn i e container r envelopesso , when viv n i use r odfo dosimetry in treatment rooms illuminated with fluorescent light. Anomalous fading is much more difficult to detect than either thermal or optical fading becaus t generallei y occurs much more slowly. Possibl t beeyye nt becausno s f thisha eo t i , demonstrated to be a problem for in vivo dosimetry. 5.2. Intrinsic precision Intrinsic precision is the reproducibility of a given TL material associated with a given readout system. It is very dependent on the quality of the TL material used, reader characteristics, the way in which the preheating and heating cycle have been defined, the purity of the nitrogen gas circulatin readoue th n gi t chamber,etc evaluatee b n ca t randomly .I d b y takin sample0 g 1 powde L T f so r dosemeterr o e samth f eirradiatiny o b batct d sou han g e samtheth meo t dose. After readoutd an , annealing procedure when necessary operatioe th , repeates ni d several times. When readout parameters have been optimised, a standard deviation of ±2% or less, can routinel obtainee yb d with either manua r automatio l c reader f gooso d quality associated with reliabl materialL eT . ] 6 1 s , [1215 , 5.3. Sensitivity 5.3.1. Solid dosimeters: identification Some variations in sensitivity within a batch of TL dosimeters is unavoidable. Two methods can be used to limit the effect of these variations : e methoon - d consist f irradiatino s e dosemeterth l e samal gth en i s geometrical conditions, to read them out and to attribute to each of them a sensitivity factor Sj equal to R, //? where Rj is the TL readout from dosimeter number i and R the mean of all values of Rj. Sensitivity factors should be checked periodically to take into account a possible loss of material occuring when TL dosimeter handlet no e sar d carefully; - another method giving a similar accuracy consists of dividing the TL dosimeters into sensitivity groups without identifying them individually (e.g. group f dosimeterso s wit hresponsa e variation less than + 1 or + 2% from the group mean) and to increase the number of dosimeters used foi each point of measurement. When an automatic reader is available, such a method is very suitable because readouts take a very short time. As the distribution of sensitivities within a group can vary wit hsame timth r e ereasonfo aboves sa t shouli , checkee db frequenca t da y which depends upoe nth accuracy require measurementsr dfo . 5.3.2. Powder: response variation with mass If optimum accurac obtainee b o t ys i d whe powderL nT usede s ar quantit e ,th powdef yo r and the readout conditions must be accurately defined and corrections made when necessary. The response variations with mass of TL material should be established for the readout conditions used in practice because they depend upon the heating kinetics. For most TL materials the signal is 274 proportiona e masth so t lwhe n the e reat witar you d h linear heating kinetics: eithe a linear r correction should be made with samples of different weight, or samples of equal weight should be used. i Some TL material such as 1^26407:01 have a response which can be considered as independent of mass, in a certain range of mass, when they are read out with automatic readers using isothermal kinetics [12]. In this case it is not necessary to weigh the powder, a simple volumetric measurement being enough (Fig.5). Fig Volumetri: 5 . c measurement consist completelf so y fillin calibratea g d hol powdereL T wit e hth . 5.4. Response variation with dose In practic recommendes i t ei dosimeterL T e regioe us th o d t n si n where their responss ei proportional to the dose received (linear region). As an indication the dose ranges corresponding to the linear zone vary from 5 x 10'5 to 1 Gy for LiF [1], 10"4 to 3 Gy for Li2B4O7:Mn [1] and 5 x 10'4 to 120 Gy for Li2B4O7:Cu [14]. Whe dosimeterL n T linea e t useth no n di re sregionar correctioa , n shoul appliee db o dt the signal from a curve established with the TL material as well as the reader used, (and not from a published curve because the readout parameters may have an influence on its shape). This curve should be checked periodically. TL dosimeter ssublineae shoule useth b n t di dno r region approaching saturationt I . should also be noted that supralinearity and saturation dose can both be affected by bad heating conditions, by previous exposures to irradiation and by thermal treatments (§ 2). 5.5. Influence of dose-rate TL dosimeters are to a large extent dose-rate independent. As shown by Tochilin [17] Goldsteid an r pe Li2B4U7:Md ny an G [18F 3 Li ] 10 independene d ar nan y G dose-rat e 5 th 4 f o o tt p eu pulse of 0.1 ms, respectively. This property implies in practice that no correction for dose-rate is necessar vivn i r oy fo measurements . Eve extreme nth e high dose-rates produce scannen di d electron beamcaust speciay no eo an sd l difficulty. 275 5.6. Influenc temperaturf eo e As the temperatures required to get the light signal out of the TL crystal is high compare rooo dt r patienmo t temperature response th ,dosimeter L T f eo indepedens si f temperaturto e variations in the range concerned by in vivo dosimetry. However care should be taken not to store the dosimeters close to a heat source. 5.7. Influence of energy 5.7.1. High energy photon beams Except for superficial measurements, TL dosimeters should be surrounded by a suitable build-up cap corresponding to the energy and geometrical irradiation conditions considered to insure electronic equilibrium [4]. When the build-up cap is made of tissue-equivalent material, it is theoretically possible to evaluate the absorbed dose in TL dosimeters and associated build-up cap irradiated with photon beams, knowing the relative variation of mass energy absorption coefficient between the TL material considered and water as a function of photon energy. In practice, due to the influence of the surrounding material (build-up cap and patient), the size and the shape of the TL dosimeter modify expectee ma s th y r cenpe dt w result[19 , 21]fe heatine a 20 , Th .y sb g conditiony sma also modify slightly the results [Ij. So the most reliable method consists of comparing directly the TL dosimeter d associatean s d e a calibrateusebuild-ub o t do t p dpca ionisation chambe e energth r y response of which is well known, to irradiate both the detectors in the same beam as those used for patient treatments, using the experimental conditions shown in Fig. 6 and to compare their responses. Becaus e sloth wf eo variatio f responsno e versu e energth sn i yenergF rangLi f Uvfiofyjyo d ean considered calibratioe th , n factors obtained with this metho then l patient dal usee ca n b r dfo s treated e samith n e beam r patiento , s treate n photoi d n e beamsamth f eo s energy, irrespective th f eo geometrical conditions of the irradiation (field size, SSD, presence or not of compensating filters, etc). It should als notee ob 52840 L d d thatypF 7an Li te6 respon sloo dt w neutrona vi s reactions with ^Li and l^B. As X-rays of very high energy are sometimes contaminated by neutrons, a particular attention has to be taken when in vivo measurements are performed with X-ray beams of energy greater than 12 MV. The best solution consists of using LiF enriched in ''Li which is not sensitive neutronso t . 5.7.2. Low energy photon beams For photon energies below 300 keV, TL dosimeters should be very thin and applied without build-up capalss i t o.I preferabl lithiue us o et m borate instea fortiora LiF f dd o ,an i otheL T r material becaus variatioe eth responsf no e with energ less yi s importan thin I t . s(Table cas2) & e s1 theoretical dat f responso a e versus e useenergb dn yca ever dosimeteyL T tim e f smalo th e s i rl dimensions [22]. For very low photon energies (below about 50 keV), theoretical curves or any other theoretical data ,use e musb t d tno directl y becaus shape edimensionth d ean detectoe th f sinduco n rca e considerable response variations withi dosimetee nth r volum , 23]e[1 . Moreove e nature th rth f o e activator may also yield too large differences in the response of TL materials in this energy range [24]. The only solution consist f comparinso g directl dosimeter L response T y th e a th calibrate f o eo t s d ionization chamber usin energw lo gmethoa e yth range showo e dt similae e on Figth e ,n Du i th . .6 o rt effective point of measurement of the chamber, which should be adapted to these low X-ray energies, thes i n situate sam e dosimeterL th T t eda e leve th s a l. 5.7.3. Electron beams Theoreticall possibls i t yi evaluato et absorbee eth d dosimeter L dosT n ei s irradiated with electron beams knowing the variation of the ratio of the mass collision stopping power of the TL material to that of tissue or water as a function of energy (Fig.3). This variation is less than 2% and 5% for LiF and Li2B4 276 detector build-up cm ion chamber ion chamber detector + ? entrance = build-up [A] Fig Determinatio: 6 . entrance th f no exid ean t dose calibratio dosimeternL factorX a f o s , Fcad i/ean n Fcal,ex ' respectively. The calibrated ionisation chamber is put at maximum depth dma* (for the entrance dose calibration factor [A]), or at a distance dmax from the exit side of the field for the exit dose calibration [B] and the detector is positioned at the entrance or exit surface, respectively. After [4]. 5.8. Directional effect No correction for directional effect is necessary except if the dosimeter container and associated build-up cap have an asymmetric shape; even for the tangential irradiation of the breast or thoracie th c wal directionao n l l dependenc detectof eo r respons observeds ei . 6. PRACTICAL APPLICATIONS Many papers have been published showin usefulnese gth dosimetrL T f so vivn i r oyfo dosimetry and quality assurance purposes. Its interest is still increasing with the apparition of modern automated readers, especially dedicated to medical applications, which can read out about 50 dosimeter minutes5 1 n si . Some example applicationf so presentee sar d below. 6.1. Calibration of treatment units and radioactives sources Mailed TL dosimeters (generally LiF) are used by several national or international institutions (I.A.EA., OMS, EORTC, etc) to check beam calibration between different radiotherapy centers. Tl dosimeters are sent to the different centers, most often with a joined special holder (or phantom). After being irradiated by the user in water, in reference conditions, they are sent back to the official institution for readout. Dose delivered by each center is then compared to the dose actually received by the TL dosimeters. It has been proven that frequent quality audit by mailed dosimetry can improve treatment quality [26]. 277 Tl dosimetr n als e useca yob o chect d k calibratio f radioactivo n e sources usen i d brachytherapy. 6.2. Verification of computer programs It consists of comparing doses calculated by the Treatment Planning System (TPS) with measurements in phantoms of different shape and composition. The method can be used either for external beam therapy [27] or brachytherapy [28, 29]. Regions in which discrepancies are observed can be pointed out, and then the algorithms of dose calculation improved. 6.3. Dose mesurements in situations where the dose calculations may be inaccurate or impossible An example of such situation is the evaluation of the dose delivered to the axilla during brachytherapy for carcinoma of the breast using indium 192. The position of the radioactive material with respect to the axillary zone differing with the patient seated or lying, only in vivo measurement able syielar o ecorrect a o dt t dose evaluation [30]. In external beam therapy, due to the limited size of the detectors, and therefore their excellent spatial resolution, another typical application of TL dosimetry is the exploration of zones of high dose gradient suc s junctioha n zones, penumbra region,etc. 6.4. Check of correct delivery of external irradiation A viv n possibli f oo dosimetrm eai checo t targes ye i kth t dos orden ei verifo t r y correct delivery of irradiation. Except when the target is the skin [31, 32] or when detectors can be introduced in natural cavities such as oesophagal tube, rectum, vagina, etc. this is impossible. Nethertheless targe deducete b dos n eca d from entranc exid ean t dose measurements performed at the patient's skin provided certain precautions are taken [4]. Then it is essential to check each beam contributing to the target dose individually, at least at the first treatment session, in order to identif possible yth e cause correcerroro f t so d san t particulatheme th n I . dosimetryr L casT f eo , that implie nee e detectorf so th chang o dt t se e esth after each irradiation beam. Once the quality of the irradiation delivered individually by each beam has been checked at the first treatment session, some users may wish to check also the reproducibility of the treatment during the following sessions. In order to save time, they often prefer to leave the same in vivo detectors on the patient's skin for the full treatment session including all beams. In this case, it shoul verifiee db d that entranc exid ean t dose f eacso ht influence beano e mar contributiony db s from other beams [4]. Such in vivo measurements can be used to detect errors in individual patients [33], to evaluat qualite eth usuaf yo speciar o l l treatment techniques (total body irradiation before bone graft marrow) or to estimate the global accuracy of a department [34). 6.6. Diagnostic radiology Different authors have used TL dosimetry for determining doses to patients during radiological procedures. Few difficulties are mentioned when lithium borate dosimeters are used for measuring gona maximud dan m skidosee th n o s[35]t . REFERENCES KEEVEe M [1] R S.W.S Thermoluminescenc: solidsf eo . University Press (New York) (1985). [2] Me KINLAY A.F. : Thermoluminescence Dosimetry - Medical Physics Handbrooks N°5. Adam Hilger, Ldt (1981). [3] AUKETcomparisoA : . J . TR f semiconuctoo n d thermoluminescenan r t dosimeter n vivi r ofo s dosimetry Britise Th . h Journa Radiologyf o l , 947-9564 , 2 (1991) 278 [4] VAN DAM J. , MARINELLO G. : Methods for in vivo dosimetry in external radiotherapy. ESTRO BOOKLET PHYSICN SO RADIOTHERAPYN SI , 1991 ° 4N , f [5] DRISCOLL C.M.H., BARTHE J.R., OBERHOFER M., BUSUOLI G., HICKMAN C. : Annealing procedure r commonlfo s y used radiothermoluminescent materials. Radiation Protection Dosimetry , 17-31 , 2 (1986). [6] TOWNSEND P.D., AHMED K., CHANDLER P.J., Me KEEVER S.W.S. , WITLOW H.J. : Radiât. (1983)5 24 . .72 Eff. [7] INTERNATIONAL COMMISSIO RADIATION NO N UNIT MEASUREMENTD SAN S: Radiation dosimetry: electron beams with initial energies between 1 and 50 MeV. ICRU Report Washingto, N5 °3 n D.C. (1984). [8] BARTH , PORTAJ. E . TechnologiG L s lecteurde e e dosimètred s s thermoluminescents. Radioprotection 25, 135-156 (1990). [9] GASIOT J., BRAUNLICH P., PILLARD J.P. : Laser heating in thermoluminescence dosimetry. J. Appl. Phys. 53, 5200-5209 (1982). [10] KEARFOTT K.J., GRUPEN-SHEMANSKY M.E .: Positiona l radiotherapy beam dosimetry using a laser heated thermoluminescent plate. Med. Phys. 17, 429-435. [11] BARTHE J. , MARINELLO G., POLLACK J. , PORTAL G. : New automatic fast reader for powder or sintered pellets. Proceedings of 9th Int. Conf. on Solid State Dosimetry. Radiation Protection Dosimetry, 34, 261-263 (1990). [12] MARINELLO G., BARTHE }., POLLACK J. PORTAL G. : "PCL", a new automatic fast reader suitabl vivn i r oe fo dosimetry . Radiother. Oncol , 63-625 . 6 (1992). [13] PORTA : Preparatio . G L d propertiean n f principao s productsL T l n "ApplieI . d Thermo- luminescence Dosimetry " by Oberhofer M. and ScharmannA. Adam Hilger Ldt, Bristol (1981). [14] VISOCEKA LORRAI, SR. , MARINELL . NS Evaluatio: . OG preparatioa f no Li2B40f r no fo u C 7: thermoluminescence dosimetry. NUCL. SCI, 22: 61-66 (1985). [15] FEIS : Einflus . T H s Regenerier de s Auswerteverfahrend un - s supralinearda f au s e Verhalten von LiF-Thermoluminiszenszdosimetern. Strahlenther. Onkol. 164, 223-227 (1988). [16] KIRBY T.H., HANSON W.F. , JOHNSTON D.A. : Uncertainty analysis of absorbed dose calculations from thermoluminescence dosimeters. Med. Phys.19, 1427-1433 (1992). [17] TOCHILIN E. , GOLDSTEIN N. : Dose-rate and spectral measurements from a pulsed X-ray generator. Health Physics 12, 1705 (1966). [18] GOLDSTEI Dose-rat: . NN e dependenc lithiuf eo m fluorid exposurer efo s above 15,00 pulser pe 0R . Health Phys. 22, 90-92 (1972). [19] ALMOND P., Me CRAY K. : The energy response of LiF, CaF2, and Li2B4C>7:Mn to high energy radiations. Phys. Med. Biol , 335-3415 . 2 (1970). [20] CHAVAUDRA J., MARINELLO G., BRULE A.M. and NGUYEN J. : Utilisation pratique du borate de lithium en dosimétrie par thermoluminescence. J. Radiol. Electrol. 57, 435-445 (1976). [21 BAGN] comprehensivA : . EF responsL eT studF Li hig eo f t y o h energy photon electronsd san . Radiology 123, 753-760(1977) 279 [22] JAYACHANDRAN : The response of thermoluminescent dosimetric lithium borates equivalent to air, water and soft tissue and of LiF TLD-100 to low energy X-rays. Phys. Med. Biol. 15, 325-334 (1970). 23] BASSI P., BUSUOLI G., RIMONDI O. : Calculated energy dependence of some RTL and RPL detectors. Int Appl. .J . Radiât. Isot 291-30, .27 5 (1976). [24] WALL B.F., DRISCOLL C.M.H., STRONG J.C., FISCHER E.S. : The suitability of different preparation thermoluminescenf so t lithium borat medicar efo l dosimetry. Phys. Med. Biol, .27 1023-1034 (1982). [25] HOLT J.G., EDELSTEIN G.R., CLARK T.E. : Energy dependence of the response of lithium fluoride TLD rod hign si h energy electron fields. Phys. Med. Biol 559-57, .20 0 (1975). [26] HANSON W.F., STOVALL M., KENNEDY P. : Rewiew of dose intercomparison at reference point Radiatio" n I . n dos radiotherapn ei y from prescriptio delivery"no t , I.A.E.A. -TECDOC-734, 121-237, (1994) [27] MASTERSON M.E., BARES , CHUTG. I C.S., DOPPKE : K.P.Interinstitutionaal & , l experience inverification of external photon dose calculations. Int. J. Radiation Oncol. Biol. Phys. 21, 37-58 (1991). [28] MEERTENS H.: A comparison of dose calculations at points around an intracavitary cervix applicato Radiotherapr. Oncologd yan y 15,199-206 (1989). [29] JOELSO RÜDE, NL N B.I., COST , DUTREIAA. , ROSENWALXA. D J.C.: Determinatio dosf no e distribution in the pelvis by by measurement and by computer in gynecologic therapy. Acta Radiologica 11,289-304 (1972). [30] MARINELLO G., RAYNAL M., BRULE A.M., PŒRQUIN B. : Utilisation du LiF en dosimétrie clinique - Application à la mesure de la dose délivrée à la région axillaire par 1 '192jr dans l'endocuriethérapi cancers ede seinu sd Radiol. .J . Electrol , 794-79.56 6 (1975). [31] CONTENTO G., MALISAN M.R., PADOVANIR. : Response of thermoluminescence dosemeters to betskid aan n dose assessment Phys. Med. Biol 661-67, .29 8 (1984). [32] MARINELL BARRE, O BOURGEOIG. E L , TC S J.P.: Lithium borate disc skir sfo n measurements: Applicatio totao nt l body superficial electron beam therapy. Nuclear Instrument Methodd san s 175,198-200 (1980). [33] BASCUAS J.L., CHAVAUDR VAUTHIE, AJ. DUTREI, RG. Intérê: X. J mesures tde vivon s"i " systématiques en radiothérapie. J. Radiol. Electrol. 58, 701-708 (1977). [34] LEUNEN r SCHUEREde DUTREI, N J. , VERSTRAETSM G. VA : , DA N XE. A. N VA , EJ. Experience of in vivo dosimetry investigations in Leuven. In "Radiation dose in radiotherapy from prescription to delivery", IAEA Report TECDOC-734, 283-289 (1994). [35] LANGMEAD W.A., WALL B.F.: An assessment of lithium borate thermoluminescent dosimeters for the measurementof doses to patients in diagnostic radiology. British J. of Radiol. 49, 956-962 (1976). 280 SUMMARY AND CONCLUSIONS ACCURACY REQUIREMENTS IN RADIATION THERAPY Tumours can be thought of as "parallel" tissues (all the functional subunits i.e. clonogens act independentl eacf yo h other). Therefore their viability remain onlf sclonogenii e yon c cell survives. There are often as many as 107 to 109 clonogens in tumours at time of diagnosis. A possible explanation for the shallow dose-response curves observed clinically for tumours is that a relatively small number radio-resistanf o t clonogen hypoxiaso t (e.ge du . ) dominat responsee eth muce ;th h larger numbef ro relatively radiosensitive clonogens are then practically irrelevant. An additional explanation for the shallower slope inter-patiens i t variation radiosensitivityn si . Both these explanations will appl moso yt t tumour type possiblt sbu varyinyn i g proportions. Tumour cur trules i statisticaya l phenomenon unles dose sth e delivered result tumoun si r control probabilities (TCP 0.90.10)> < relativ e r 0o .Th e standard deviatio eacr 0.5e fo h th 0t % n a leve 25 s i l patient. Normal tissue wels s a organ s a l rist sa k cannot generall seee y b parralel s na l tissues. Thee yar much more complex than tumour thein si r respons radiatioeo t n with many more level increasinf so g sevent damagef yo . Broadly speakin distinco g thertw e tear classe organf so s with regar theio dt r response to partial irradiation: "parallel" e.g. lung and liver and "series" e.g. the gut and the spinal cord. The former exhibit a large "volume effect", the latter a small one. In a group of patients the cure and damage probabilities may be correlated i.e. cells in the tumour anorgane dparticularlth e b t y rissa kma y radiosensitiv r radioresistanteo ; thiimplications sha r sfo therapeutic strategies as there is now evidence that such correlation may involve as many as 80% of patient somr sfo e type tumoursf so . assayf Piloo e tcellulaf sus studieo e th f rs o respons radiatioeo t n woul highle db y desirabl orden ei exploro rt potentiae eth individualisatior lfo dosf no e prescription; normal-tissue assays will probably be more useful than tumour assays as the former determines the tolerance dose. Further, the most resistant tumour clonogen is very hard if not impossible, to assay. Unless the tumour is very heterogeneous (in clonogen population, etc.) inhomogeneities in the target-volume dose distribution always result in a decrease in TCP but this decrease is small if the heterogeneitie dose th en s i distributio smale ncurvear P l (anTC ,slope e i.edgammth e th f .th e o a value, is not too large); consequently the mean dose, D, is generally the best single parameter to characterize distributioe th n generall s (i.ei preferre.D e b yo t d over D,^.). Ther larga es i e sprea clinicalln di y observed gamma value tumoursr sfo most ,bu thef n i to e mar the range 2-5; the few high values e.g. larynx are probably due to the absence of hypoxia. The data on normal-tissue gamma values is of very poor quality but in general it is likely that such gammas are larger than for tumours, especially if the number of subunits is large. The much quoted "5%" figure from ICRU report 24 for accuracy requirements of the prescribed dose to the target volume (e.g. Dre£) is probably a 2-sigma value. Thus the one-sigma number is 2.5%. This is not realistic at the present time, however. A 2-3 % requirement is consistent with current radiobiological models whe meae nth n dos usees i references da . Physicists should continu strivo et o et achieve as low uncertainties as possible in the dose delivered to the tumour. Only in this way will reliable clinical data ever be generated on which to base future improvements in clinical outcome, i.e. better estimate gammvaluesf o s o D d a.an Currently ther unacceptabln a es i e lac uniformitf ko y between different clinics (even withie non country) in the way that target doses are prescribed and reported, some preferring Dref, others D, yet others D,^. For tumour response, D is generally the most relevant single measure of the target-volume dose distribution and physicists should endeavour to calculate this even in the absence of a 3D treatment planning system; this can often be done quite accurately by manual methods as heterogeneities in the dose distribution usually hav smalea l influenc large-volumn eo e averages goos i t I .d practic alseo t o 281 report the one-sigma of the dose distribution in the target volume (and organs at risk) as well as The above suggestions regarding dose reporting strictly applies only to high-dose radiotherapy given with curative intent, i.e. not to palliative treatments. EQUIPMENT REQUIREMENTS The need to learn from accidents which have occurred in the past was recognised, and the analysi a numbe f o s f reporteo r d accidents indicated that ther s oftei e a roon t causd an e contributing factor radiotherapn I s analysie yth s shows that human erro muca s ri h more important source of accidents than equipment failures and this underscores the continuing need for adequate trainin qualificationd gan personnef so addition i l adequato nt e procedure supervisiond san e Th . identificatio f initiatinno g event contributind san g factor valuabla s si e exercis testinr efo e gth vulnerabilit radiotherape th f yo y syste disciplinel al my b s involve processe th n di : Radiotherapist, Physicist, technologist, maintenance engineer, administrator, manufacture regulatord an r y authority. This exercise should include scenario unusuaf so l events which could caus accidentn ea wels a , s la case histories of events which were successfully handled and did not result in an accident. Regarding equipment, in addition to the traditional approach of spécifications and standards, some concepts with origin in a safety culture philosophy, such as "defence in depth" should be considered durin desige gth n stage positive Th . e respons nationaf eo l professiona scientifid an l c organisations in preparing quality assurance protocols, was reported. Also the attention given to design considerations, procedures, calibration, quality assurance, investigatin accidentf go d san adequate record e revisioth n si f Basi no c Safety Standard r Protectiofo s n Against Ionising Radiation and the Safety of Radiation Sources was mentioned. In an attempt to view the phenomenon in perspective, it was observed that in a single radiotherapy room thousands of patient set-ups are performed annually. In addition due to the nature of radiotherapy when considering misadministratio intendee th f no d dose, there mamarginaa e yb "greyr lo " area between uncertainty and what might be considered an accident. In reviewing the recommendations which IAEA, PAHO, and WHO have made concerning radiotherapy equipment and facilities since the 1960s, the participants recognised that in view of globae th l situation existingw sno , mos thesf o t e recommendation stile sar l applicabl somn ei e places in the world. It was observed that in 1994 in many places that reliable electrical power (voltage, current & frequency), water supplies (quantity, and quality) and environmental modification systems (temperature, humidity, and dust control) are unavailable, or only available in a few large cities. The availability of adequate maintenance services, and spare parts and/or the funds to provide them is severely limited in many locations. The perquisites remain unchanged for the successful use of any complex equipment, Co-60 or accelerator; Expert personnel, excellent infrastructure, good communications, large patient load, excellent dosimetry, efficient organisations and supporting structure. Taking into accoun existine tth g situatio modificationd nan s whice har feasibl sustainabld ean economi e vien ei th f w o socia d can l situation, priority shoul givee db no t acquirin kine gequipmenth f do t whic moshs i t likelable b functio o et yo t loca e th n lni environmen t (climate, staff, supporting services, operating resources) where it is to be used. r externaFo l beam radiotherapy concepe essentiath n , a f equipmen o f o tt se l provido t t ea satisfactory level of quality, for example a reliable source of megavoltage photons (Cobalt-60 or energw lo a y accelerator, depending local conditions) simpla , reliabld ean e treatment simulatod ran simpla e treatment planning syste proposeds mwa . Regarding the projections for the future, the World Health Organisation's estimate that currently there are approximately nine million cancer cases per year, world-wide, and that by year 2015 there will be 15 million new cancer cases annually, with about two thirds of these in developing countries was mentioned. Consequently, the need to provide simpler, more reliable and 282 less costly equipment (considering purchase, operating, source replacemen case Co-60th f eo n ti , maintenance and decommissioning costs plus operating characteristics such as "up-time") was recognised. Furthermore, it was noted that the cost involved in disposal of spent radionuclide sources discourages owners from proper remova storaged an l accidentd an , s like that onen si Ciudada Juarez, Mexic Gioaniad oan , Brazil occur. Various component factord san s thainvolvee ar t deliverinn di prescribee gth d e dosth o et planning target volume (PTV, Re; ICRU report No 50), were identified and discussed: Patient marking, patient positioning (from simulation through treatment), treatment, mobility of organs, repeatability and verification. Even though considerable effort is made to carefully reproduce the patient positioning from simulator to treatment unit, and to maintain the correct positioning during treatment, considerable deviations between the prescribed and the delivered dose, estimated to be ordee inth severaf ro l percent occuy onle ma ,patienr o ydu t t positioning variations. Thus larga , e portion of the allowable uncertainty (perhaps even exceeding some estimates of the overall acceptable uncertainty) may be caused by patient positioning deviations. The need for simulation was emphasised and it was observed that the number of simulators in use is often very insufficient for the number of treatment units. The need for X-ray films for treatment planning and verification was indicated and it was observed the fluoroscopy alone is not adequate e nee r multidisiplinarTh .dfo y team work including radiotherapists, physiciand san technicians, working closely togethe recogniseds rwa role rulef eregulationo Th .d san creatinn si g a situation where quality and quality assurance are accorded the required priority was demonstrated, for example training criteria and standards, equipment standards, quantity and qualifications of staff and mandatory participation's in quality assurance activities. Details concerning requirement Membee on n si r State (Argentina providee )ar Annen di . Thesx1 e could serve as a good model for other interested Member States. DOSIMETRY PROCEDURES Absolute dosimetry procedures continue to be based on ionization chamber measurements using NK-Na D formalism statue Th thf .s o e IAEA Cod Practicef eo , TRS-277 solia s e d,ha th groun o t e ddu good agreement found in dosimetry intercomparisons with different Primary Standard Dosimetry Laboratories. These mainly refer to ""Co gamma-ray beams, but deviations not larger than 2% have also been obtained for higher photon beam qualities and electrons. IAE s continueAha progras dit f disseminatinmo g ionometric standard achievind san g high accurac radiotherapyn i y dosimetry same th t e A tim. e efforts have been addressed toward scontinuoua s critical analysis of new developments in the field to take into account their possible influence on TRS- futur7 27 e updates. Accordingl Codw Practicf ne e yo calibratioplane-parallef a e o th e r eus fo d nan l ionization chambers is under development which will incorporate most of the advances in radiotherapy dosimetry since the publication of TRS-277 in 1987. Some of these advances have been treated in detail in this session, such as proposals for corrections to take into account the bremsstrahlung contamination in clinical electron beams. The measurement of relative dose distributions will also be improved by the implementation of almost tissue equivalent plastic scintillator detectors which, hopefully, will be able to overcome some of the limitations of detectors in current use. New trends in absolute dosimetry have also been included in this session. Efforts have been addressed to the use of clinical high-energy photon beams to verify calibration factors in terms of absorbed dos wateeo t r (NJ. importanThin a s i t difference compare verificationdo t s done previously in non-clinical photon beams, suc thoss ha e produce acceleratory db mosn si t Primary Standard Dosimetry Laboratories (PSDL's). The results presented in this seminar are encouraging and two aspects should be stressed. First is that the comparison of experimental and calculated photon beam quality factors (kg) gives support to the data in TRS-277 at high-energy photon beams. Second, that prio implementino rt high-energt a w gN y photon beams PSDL's shoul t effort reaf dpu o le s us int e oth clinical beams for possible direct Nw calibrations. 283 BRACHYTHERAPY Considering; increasine Th • g numbe Ir-19f o machinerR probleme worl2e th HD th d n dsan i s associated wits hit dosimetric aspectneee uniformizatior th dfo d san source th f no e calibration methods. • The increasing acceptance of Ir-192 wires for interstitial brachytherapy and problems associated with its production and dosimetry. potentiae Th • l mismanagemen patienf laco to t training f ke o tdu . proposedIs i t . promoto T I. workinea g group reviewin available gth e informatio proposd nan e guidelinew ho n so to approach these questions; u. To prepare guidelines for the SSDLs assisting Ihe users with appropriate dosimetric methodology in order to assure Ihe necessary metrological coherence in each country. These documents should also contemplate. Radiation protectio emergencd nan y procedures, quality assurance, staffing, training, a review of available clinical data from LDR & HDR treatments and cost-benefit analysi f technologthiw o s ne s y especialln i t i y f o wit e hus regarde th o t s developing countries. QUALITY ASSURANCE NETWOR RADIOTHERAPKN I Y Since 1966 IAEA/WHe ,th O Postal Dosimetry Programm hospitalr efo proves sha usefulnese nth s and the reliability of external audits, not only for developing countries, but also for countries where quality assurance programmes are applied. Large deviations have been observed in most areas of the worl repeated dan d checks resulte substantiadn i l improvements. The European Pilot Study on postal dosimetry service, which was set up in 1991 to explore the feasibility of enforcing the capabilities of such remote postal assistance through additional checks including not only the calibration and the beam quality at the beam axis, but also off-axis doses, the treatment planning system as well as patient set-up through in-vivo measurements and port films. In particular piloe ,th t study showed that special device s"multipurposee sucth s ha " phanto usefue b n mlca and should be made available to other networks when completely evaluated. The discussions following the presentations pointed out a number of important points. I. Depending on the conditions existing in a given country, remote assistance may not be sufficient. Local conditions, regulation stafr so f migh evet tno n allow check performee b o st d locally reasone .Th n sca be lac physicistf ko qualifiey an r so d staff ablmeasurementse carreo th t t you , insufficient trainine th f go physicist lace equipmentf th k o r so . This might require on-sit firse eth t visit e actiob o st n insteaa f do follow-up visit aftescorew lo r s wit probe hth e service (educational interest). u. The need for more qualified physicists in many countries has also been stressed, as well as a reasonable rati dosimetristsf oo , technologist engineersd san . . m Difficulties were also identified regardin recognitioe gth somy nb e radiation oncologist role th e f so and the need for physicists in clinical trials, especially for the physics part of the Q. A. and more generally in some radiotherapy departments. It has been suggested that improvements can be expected from the participatio physiciste th trainine f radiatioe nth o th r f sfo go n oncologist medican si l physicso t , reciprocal contribution scientific/medican si l meetings, and, more generally commoo t , n effortr fo s cooperation. . IV With respec neee morr th dfo o tt e accurate radiotherapy emphasises wa t ,i d tha mann ti y countries a severe lack excists of devices for accurate tumour definition and localization as well as for immobilization and imaging. 284 . V Quality audits should includ onlt eno y external beam therapy alst ,bu o brachytherapy particulan i , r some verification and calibration, software. The development of high dose rate brachytherapy machines stresses the need for such QA systems Finally, in general there was a feeling that in many developing countries the lack of physicists in combination with the low recognition of this professional group from Radiation Oncologists is a main obstacle for introducing proper QA programmes and thus, improving clinical dosimetry. The lack of recognition might be understandable given the perspective that these physicians have worked with radiation therapy for many years without a single medical physicist in the country. Primary goals must therefor creato t e eb e resource educatior sfo thif no s typ specialistf eo o t d san make clinicians aware of the value of such staff. The IAEA could play an important role in helping creating the resources for education and in combination with WHO the physicians could be informe necessite th n do medicaf yo l physicist clinie th cn si leftCL.V-5 28 I X LIST OF PARTICIPANTS ARGENTINA Bruna, A.E. Faculta Matematice dd a Astronomi aFisicy a Medina Allende-Hay Torra l e ad e Ciudad Universitaria 5000 Cordoba Brunetto, M.G. Centro Medico Privado Dean Funes Dean Funes 2869 5000 Cordoba Salvoe D . F , Institute Lavalle Lavall9 e25 8000 Bahia Bianca Feld, D.B. Comision Naciona Energie d l a Atomica Radiacion-Gcia. de Aplicacioness Division Dosimetria Avda. Libertador 8250 1429 Buenos Aires Hisano, J.O. Centro de Radioterapia Crespo 953 2000 Rosaho-Sante aF Mac Donnell, D.J. Centr Unidae od Terapie dd a Radiant Rosane ed o S.A. Necochea 1327 2000 Rosario-Santa Fe Meoli, P.J. Grupo Colisiones Atomicas Institute de Fisica Rosano Av. Pellegrini 250 2000 Rosario-Santa Fe Migues. V , Hospita Pediatrie d l a "Jua . GarrahamnP " Terepia Radiante Combate de los Pozos 1818 1245 Buenos Aires Negri, H.A. Diagona . 334No 34 7 l La Plata (1900) Nelli, F.E. Garibaldi 405 Mendoza C.P. 5500 Olivera, G.H. Grupo Colisiones Atomicas Institut Fisice ed a Rosano Av. Pellegrin0 25 i 2000 Rosano Pereira Duarte, S.E. Diagona l Ombde l u 3915 Mar del Plata (7600) 287 Reyes, L.E. Avda. Santa Fe 5978 5978 Carapachay (1605) Buenos Aires Rodriguez, H.R. Calle 508 e/16 y 18 1897 Gönnet. Provincia de Buenos Aires Sacc, R.A. Boulevard Zaballa 2947 3000 Sante aF Saravi, M. Comision Nacional de Energia Atomica Centre Regiona Referencie d l a para Disimetria Av. del Libertador 8250 1429 Buenos Aires Singer, E. Mevaterapia Medical Centre Tte. Oral. Peron 1198 Buenos Aires Vêlez, G.R. Hospita Roqun Sa l e Deprtament Radioterapie od a Rosario de Santa Fe 374 5000 Cordoba Venencia, C.D. Institute Privad Radioterapie od a Obispo Oro 131 5000 Cordoba AUSTRIA Ortiz-Lopez. P , Divisio Nucleaf no r Safety International Atomic Energy Agency Wagramerstr. 5 P.O.Bo0 x10 A-1400 Vienna BRAZIL Costa, J.M.L. Rua Pedro J. 19 AP.811 - Centre Rio de Janeiro CEP 20060-050 de Almeida. C , Laboratori Cienciae od s Radiologicas Universidade do Estdo do Rio de Janeiro (UERJ) Rua Sao Francisco Savier, 524 Maracena 2055P CE , 0 Rio de Janeiro Diz Rodriguez de Biaggio, ABACC M.F. Av. Rio Branco 123 5-andar Janeire d o oRi Ferreira, M.L. Institute Nacional do Cancer Praca Cruz Vermelha 23 3-andar-Radioterapia Ri Janeire od oCentr- e 288 Henn, K.C. RuaPedro I, 19 Apto. 819 Ri Janeire od oCentr- e Maréchal, M.H. Institut Radioprotecae ed Dosimetrioe aLNMR- I Comissao Naciona Energie d l a Nuclear Av. Salvador Allende 37750-CEP:22780-160 Ri Janeire od o Morais. E , Institute Naciona Canceo d l r Praca Cruz Vermelha3 2 , CEP 20060-050 Ri Janeire od oCentr- e Peregrine, A.A. Universida Janeire d Estado o dd oRi o od Rua San Francisco Javier 524 Maracana Rio de Janeiro Rodriguez, L.N. LNMRI-IRD C.P. 37750 22780-160 Rio de Janeiro Russo, D.M.B. Servicio de Radioterapia Estereotaxica Hospital Beneficiencia Portugesa . MaestrR o Cardim 769-BlocoV 3 Subsolo - Parai so 01323.00 Paulo 1Sa o Salzberg, L. Av. Guadalupe 375/1011 Jardim Lindoia Porto Alegre-Rio de Janeiro Torres de Fanas, J. Av. Salvador Allende, SN Barra da Dijuca Rio de Janeiro Viamello, E.A. . MoyseR s Santiago 245 Eldorado Juiz de Foro-M6 CEP 36046200 CHILE Oyarzun Cortes, C.H. Laboratori Metrologie od Radiacionee ad s lonizantes Comision Chilena de Energia Nuclear Amunategui 9 Casilla 188D Santiago Schwartzmann. L , Pucis Ve o Norte 1314 Santiago 289 COLOMBIA Diaz Molina, F. Call 9-8+ e1 5 27466 Bogota Plazas M.C. National University of Colombia Facultad de Ciencias Ciudad Universitaria Santa Fe de Bogota R am irez, G. Hospital Universitano del Valle . 36-0CallNo . 8 e5a Call CUBA del Risco Reyna, L. National Control Centr Medicar efo l Divices Habana CZECH REPUBLIC Novotny, J. Radiotherapy Institute Ministry of Public Health Natruhlarc0 e10 CS-180 00 Prague 8 ECUADOR Alum Alvaredoe ad , M.D. Sociedad de Lucha Contra el Cancer (SOLCA) Nucle Cuence od a Abril e ParaisAved de 2 .l 1 oy Apartado N-860 Quito Gutierrez, P.N. SOLCA, Fisica Medica Avda P.J. Menendez Casilla Correo 5255-3623 Guayaquil FINLAND Aaltonen, P. Finish Centr Radiatior efo Nuclead nan r Safety (STUK) P.O.Bo8 x26 SF-00101 Helsinki FRANCE Chavaudra, J. Service de Physique Institut Gustave Roussy Rue Camille Desmoulins F-94805 Villejuif Cedex Marinello, G. Department Inter-hospitalie Cancérologie rd e Servic Radiotherapie ed e Centre Hospitalo-Universitaire 51, Av. Maréchal de Lattre de Tassigny F-94010 Creteil 290 GERMANY Quast, U. Division of Clinical Radiation Physics Department of Radiotherapy D-45122 Essen Regulla, D.R. GSF-Research Centre D-85758 Neuherberg Roos. M , Physikalisch-Technische Bundesanstalt Bundesalle0 e10 D-38116 Braunschweig ITALY Laitano. F , ENEA-Casaccia Départiraento Amiente Institute Nazional Metrologia delle Radiazioni lonizzanti Viale Anguillares1 e30 1-00060 S.Mariaoi Galeria Rom a JAMAICA Buchanan, R.E. Dept Medicaf o . l Physics Radiotherapo c/ y Department Cornwall Regional Hospital Montegy oBa Mt. Salem PERU Garcia-Angulo, A.H. Hospital Nacional Norte Institute Peruan Segundae od d Social (IPSS) Servicio de Radioterapia Apartado 75 Chiclayo SOUTH AFRICA Murphy, E.M. Atomic Energy Corporatio f Soutno h Africa Ltd. Dept. of Applied Radiation Technology P.O-Box 582 Pretoria Strydom, WJ. Depratment of Medical Physics Medunsa Medunsa 0204 SWEDEN Andreo. P , Radiophysics Dept. Lunds Hospital Lund S-225 18 291 Brahme, A. Radiation Physics Department Karolinska Institute P.O-Box 60211 S-104 01 Stockholm SWITZERLAND Hanson, O.P. Radiation Medicine World Health Organization CH-1211 Genev7 a2 UK Klevenhage, S.C. Medical Physics Department The Roya London Hospital Whitechapel Road LondoB 1B I nE Nahum. A , Joint Dept. of Physics The Royal Marsden Hospital Downs Road Sutton SM2 5PT URUGUAY Apardian. R , Medical University University Hospital Joanico 3265 Luongo, A.I. Institue de Oncologia Depto Radioterapia Ministerio de Salud Publica Soriano 1171 Montevideo USA Almond, P.R. Dept Radiatiof o . n Oncology J.G. Brown Cancer Center Universit f Lousvillyo e Louisville KY 40292 Borras, C. PANO HSD/RA- D Pan American Health Organization 525 23rd Str. N.W. Washington 20037-289C D , 5 VENEZUELA de Padilla, M.C. Institute Vénézolane de Investigaciones Cientifîcas Laboratori Calibracioe od n Dosimetrica Apartado 21827 Caracas 1020-A Gutt, F. IVIC-LSCD Carretera Panam Km 11 Caracas 292 RECENT IAEA PUBLICATIONS ON RADIATION DOSIMETRY 1987 Absorbed Dose Determination in Photon and Electron Beams: An International Code of Practice (Technical Report- s7 Serie27 . sNo STI/DOC/10/277) 1988 Dosimetry in Radiotherapy (Proceedings Series - STI/PUB/760) 1994 Measurement Assurance in Dosimetry (Proceedings Series - STI/PUB/930) 1994 Radiation Dose Radiotherapyn i from Prescription Deliveryo t (IAEA- TECDOC-734) 1994 Calibration of Dosimeters Used in Radiotherapy (Technical Reports Series No. 374) 293