Determination of transit dose profile for a 192Ir HDR sourceG. P. Fonseca, R. A. Rubo, R. A. Minamisawa, G. R. dos Santos, P. C. G. Antunes, and H. Yoriyaz Citation: Medical Physics 40, 051717 (2013); doi: 10.1118/1.4802731 View online: http://dx.doi.org/10.1118/1.4802731 View Table of Contents: http://scitation.aip.org/content/aapm/journal/medphys/40/5?ver=pdfcov Published by the American Association of Physicists in Medicine

Determination of transit dose profile for a 192Ir HDR sourceG. P. FonsecaInstituto de Pesquisas Energéticas e Nucleares – IPEN-CNEN/SP, São Paulo 05508-000, Brazil andDepartment of Radiation Oncology (MAASTRO), GROW School for Oncology and Developmental Biology,Maastricht University Medical Center, Maastricht 6201 BN, The Netherlands

R. A. RuboHospital das Clínicas da Universidade de São Paulo – HC/FMUSP, São Paulo 05403-900, Brazil

R. A. MinamisawaLaboratory for Micro- and Nanotechnology Paul Scherrer Institut, 5232 Villigen PSI, Switzerland

G. R. dos SantosHospital das Clínicas da Universidade de São Paulo – HC/FMUSP, São Paulo 05403-900, Brazil

P. C. G. Antunes and H. Yoriyaza)

Instituto de Pesquisas Energéticas e Nucleares – IPEN-CNEN/SP, São Paulo 05508-000, Brazil

(Received 14 September 2012; revised 8 April 2013; accepted for publication 8 April 2013;published 30 April 2013)

Purpose: Several studies have reported methodologies to calculate and correct the transit dose com-ponent of the moving radiation source for high dose rate (HDR) brachytherapy planning systems.However, most of these works employ the average source speed, which varies significantly with themeasurement technique used, and does not represent a realistic speed profile, therefore, providingan inaccurate dose determination. In this work, the authors quantified the transit dose component ofa HDR unit based on the measurement of the instantaneous source speed to produce more accuratedose values.Methods: The Nucletron microSelectron-HDR Ir-192 source was characterized considering the TaskGroup 43 (TG-43U1) specifications. The transit dose component was considered through the calcu-lation of the dose distribution using a Monte Carlo particle transport code, MCNP5, for each sourceposition and correcting it by the source speed. The instantaneous source speed measurements wereperformed in a previous work using two optical fibers connected to a photomultiplier and an oscil-loscope. Calculated doses were validated by comparing relative dose profiles with those obtainedexperimentally using radiochromic films.Results: TG-43U1 source parameters were calculated to validate the Monte Carlo simulations. Theseagreed with the literature, with differences below 1% for the majority of the points. Calculated doseprofiles without transit dose were also validated by comparison with ONCENTRA R© Brachy v. 3.3dose values, yielding differences within 1.5%. Dose profiles obtained with MCNP5 corrected usingthe instantaneous source speed profile showed differences near dwell positions of up to 800% incomparison to values corrected using the average source speed, but they are in good agreement withthe experimental data, showing a maximum discrepancy of approximately 3% of the maximum dose.Near a dwell position the transit dose is about 22% of the dwell dose delivered by the source dwelling1 s and reached 104.0 cGy per irradiation in a hypothetical clinical case studied in this work.Conclusions: The present work demonstrated that the transit dose correction based on average sourcespeed fails to accurately correct the dose, indicating that the correct speed profile should be consid-ered. The impact on total dose due to the transit dose correction near the dwell positions is significantand should be considered more carefully in treatments with high dose rate, several catheters, multipledwell positions, small dwell times, and several fractions. © 2013 American Association of Physicistsin Medicine. [http://dx.doi.org/10.1118/1.4802731]

Key words: Ir-192, Brachytherapy, Transit dose, Monte Carlo

I. INTRODUCTION

Accurate determination of absorbed radiation dose in tissuessurrounding radioactive implants is of major importance forbrachytherapy treatments. To ensure the optimum treatmentconditions, dosimetry measurements are usually performedin water equivalent phantoms which can be associated withcomputational simulations, besides source dwell position ver-

ification and air-kerma strength measurements. These qual-ity assurance measurements are typically performed to deter-mine dosimetric parameters required by standard protocols.However, almost no hospitals perform measurements to deter-mine the instantaneous source speed, which can significantlyimpact the dose delivered to the patient.1, 2 This is particu-larly important in high dose rate (HDR) remote afterloadingunits containing a cable-driven radiation source, where the

051717-1 Med. Phys. 40 (5), May 2013 © 2013 Am. Assoc. Phys. Med. 051717-10094-2405/2013/40(5)/051717/8/$30.00

051717-2 Fonseca et al.: 192Ir HDR transit dose profile 051717-2

determination of the source positioning is the main uncer-tainty for calculation of the delivered dose.3 Another impor-tant uncertainty component stems from the determination ofthe transit dose, which, in turn, depends on the transit speedof the radiation source.4 Most of the studies aimed at deter-mining the transit speed of the source were able to measureonly the average source speed, which does not represent therealistic motion of the source, e.g., neglecting the low speedof the source approaching a dwell position may lead to in-accurate determination of the transit dose.5–7 So far, tran-sit dose corrections are implemented inaccurately into HDRbrachytherapy treatment planning systems (TPS) due to a lackof quality assurance tests able to accurately measure the tran-sit speed profile of the moving radiation source.

In the present work we report the accurate determination ofthe transit dose component for a Nucletron Ir-192 brachyther-apy remote afterloader unit using measurements of the in-stantaneous speed profile of the radioactive source. Qualityassurance tests of the transit speed of the source were per-formed using a high performance optical fiber based radia-tion detector.4 The computational simulations were based ona Monte Carlo (MC) code which provides a more accuratedose distribution, not limited to homogeneous water media.8

To the best of our knowledge, this is the first determinationof the real transit dose profile of a brachytherapy unit. Fur-thermore, we compare our results with those obtained usingthe commercial planning system ONCENTRA R© Brachy ver-sion 3.3 (Ref. 9) and with experimental measurements usingradiochromic film dosimeters.

II. MATERIALS AND METHODS

II.A. Measurements of the instantaneoussource speed

Recently, we have reported real time measurements of theinstantaneous speed profile of a HDR Ir-192 brachytherapysource using a high resolution optical fiber based detector.4

The device consists of two parallel optical fibers that are per-pendicularly positioned on top of a treatment catheter, andconnected to a photomultiplier and an oscilloscope. When thesource moves directly under each optical fiber, radiation in-duced Cerenkov light is generated and measured in the oscil-loscope in the form of two intensity peaks as a function oftime. Since the distance between the optical fibers is knownand approximately equivalent to the source length (l = 0.4cm), the extracted velocity corresponds to the instantaneousspeed vins. The vins profile is finally obtained by positioningthe optical fibers at several positions along the catheter andrepeating a specific source motion/dwell situation for each de-tector position. Using this device, we measured the vins profilefor different program situations of clinical relevance, deter-mining that the radioactive source follows a uniformly accel-erated linear motion, with an acceleration of α = (113 ± 2)cm/s2, and maximum source speed of vmax. = (52 ± 1) cm/s,which were used to obtain all instantaneous speed values, vins,through classical kinematics. Note that Nucletron provides anaverage source speed of 〈v〉 = 50 cm/s (Ref. 9) which is ap-

proximately ten times higher than the average speed for aninterdwell position distance of 0.5 cm, obtained consideringthe instantaneous source speed.

II.B. Monte Carlo calculations

Although the TG-43U1 formalism employed byONCENTRA R© could be used to calculate the transit dose inwater, a MC method was adopted since this methodology canbe implemented in a TPS considering heterogeneous media,which has particular relevance in brachytherapy treatments.8

The simulations were performed using a general-purposeMC code, MCNP5 (Monte Carlo N-Particle Transport code,version 5), an Ir-192 National Nuclear Data Center (NNDC)spectrum, and transporting photons with an energy cut-offof 1 keV, using the MCPLIB04 photon cross section libraryin Mode P which uses the thick target bremsstrahlung (TTB)model where secondary electrons are generated in the direc-tion of the incident photon and are immediately annihilated,after generating bremsstrahlung.10

Dose distributions were calculated using the track lengthcell energy deposition MCNP5 tally (F6), which consists ofthe track length tally multiplied by a reaction rate convolvedwith an energy-dependent heating function (in MeV/g).10

II.B.1. Absolute dwell dose profile

Here, the dwell dose profile is defined as the spatialdose distribution delivered only during dwelling, therefore,not considering the transit dose component. The accuracyof the MCNP5 simulation for dwell dose profile calcula-tions has been verified by comparing results provided by theONCENTRA R© planning system and also through the TG-43U1 dose-calculation formalism.11, 12

The validation procedure was performed assuming the di-mensions and material composition of the source providedby Nucletron, except for the steel cable density which wasdetermined experimentally,13 and consists of the followingsteps: (a) the air-kerma strength was obtained inside a vac-uum sphere using a cylindrical cell positioned at 1 m fromthe source, following the methodology described by Borg andRogers;14 (b) the dose-rate constant was obtained consideringthe dose in water at 1 cm from the source and the air-kermastrength; (c) the anisotropy, the radial dose function, and thedose rate per air-kerma strength were obtained using a spheri-cal phantom with 30 cm diameter as used by Daskalov et al.13

II.B.2. Studied cases

The transit dose effect was evaluated considering a (40× 40 × 40) cm3 water phantom, adopting its center as theorigin of coordinates and the last dwell position (catheter tip)at 6 cm from the origin, which is the same reference adoptedin the experimental procedure.

The scoring axis is aligned at two distances parallel to thecatheter: (1) 2.5 mm, which was set to be higher than theminimum distance to establish electronic equilibrium15 and(2) 5.0 mm. The studied cases are: (a) source travelling from

Medical Physics, Vol. 40, No. 5, May 2013

051717-3 Fonseca et al.: 192Ir HDR transit dose profile 051717-3

the safe and dwelling at position x0 = 6 cm, (b) source trav-elling from the safe, dwelling at positions x1 = 1 cm and x0

= 6 cm (interdwell distance d = x0 − x1 = 5 cm), and trav-elling back to the safe, (c) source travelling from the safe,dwelling at positions x1 = −4 cm and x0 = 6 cm (d = 10 cm),and travelling back to the safe, and (d) source travelling fromthe safe, dwelling at multiple positions between x1 = −4 cmand x0 = 6 cm in steps of 2 cm, and travelling back to the safe.These cases were simulated with 2 × 109 particles, resultingin a Type A component of uncertainty less than 1% and 3.5%(k = 1), for the cases at 2.5 and 5.0 mm, respectively.

Besides the aforementioned cases, a hypotheticalprostate case with 10 catheters was studied by comparingONCENTRA R© and MCNP5 to verify the transit dose effectin a clinical case using a template for prostate treatmentwhich has been set without considering the transit dose effectto define the position of the catheters or the prescribed dose.Each catheter has 11 dwell positions with 0.5 cm spacing. Thedwell times were defined automatically using ONCENTRA R©

geometrical optimization on volume, considering a 100%isodose volume of 121.17 cm3, a prescribed dose of1000 cGy, and an air-kerma strength of 4.107 × 104 U.Dose distributions were analyzed in 14 axes parallel to thecatheters at the same depth as the dwell positions.

ONCENTRA R© provides a broad range of source dwelltimes from 0.1 to 999.9 s, uses TG-43U1 parameters equiv-alent to those obtained by Daskalov et al.,13 presents a highspatial resolution, and considers the transit dose by subtract-ing the transit time (up to 0.1 s), obtained using an averagespeed of 〈v〉 = 50 cm/s, from the source dwelling time.9 Dueto the high average speed adopted and due to the no apparenteffect of the transit dose corrections in the dose distributionsobtained, results from ONCENTRA R© were considered with-out transit dose correction.

II.C. Dynamic dose calculation

The transit dose profile was obtained in a 22 cm long ar-ray positioned concentrically in the middle of the phantom,composed of 440 water sphere targets with a volume of ap-proximately 10−3 mm3 and separated by a 0.5 mm lateral dis-tance. A transit dose profile was simulated one time, with onesource aligned within the middle of the phantom, and then

the result obtained was shifted in intervals of 0.5 mm repre-senting the movement of the source and avoids having to per-form hundreds of simulations. Figure 1 shows the simulationgeometry.

The total transit dose per air-kerma strength (Gy U−1) ineach target sphere, Dtr, is obtained through the frame-by-frame integration of the dose delivered by the radioactivesource while moving through two consecutive positions, �S(0.05 cm), as given by the following equation:

Dtr =∑n−1

i=1

(Di

vi+ Di+1

vi+1

)

2· �S, (1)

where n is the last source position, Di (Gy U−1 s−1) is thecalculated dose per air-kerma strength delivered in one targetsphere due to the source at position i, Di + 1 is the calculateddose per air-kerma strength delivered in one target sphere dueto the source at a consecutive position i + 1, vi (cm/s) is theinstantaneous speed at position i, and vi+1 is the instantaneousspeed at a consecutive position i + 1.

II.D. Dose measurements

Gafchromic EBT2 radiochromic films (lot: A08151101A)were positioned between (30 × 30) cm2 PTW RW3 solid wa-ter plates composed of C8H8 + 2.1% TiO2 with a densityof 1.045 g/cm3 (Ref. 16), at a depth of 0.1 cm, with threeplates with a thickness of 1 cm each positioned below thefilm. The 2.5 mm thick catheter was positioned onto the sur-face plate and inside of an oil wax of dimensions (14.0 × 7.0× 0.5) cm3 manufactured by Epoxiglass (Diadema-SP,Brazil), and concentrically positioned in the middle of thephantom, which was adopted as the origin of the coordinatesystem. The catheter was aligned parallel to the longest waxdirection with its tip, which is the last dwell position presentedin each case, at 1.0 cm from the border of the wax and con-sequently 9.0 cm from the closest border of the phantom. Theirradiated films were digitized using an Epson scanner model10000 XL with a resolution of 720 × 720 pixels, and ana-lyzed with IMAGEJ (Ref. 17) using the green channel and thelinearization equation proposed by Devic et al.18 MC simula-tions were performed considering a (40 × 40 × 40) cm3 wa-ter phantom and also considering the experimental geometryand the solid water composition to assure that the differences

FIG. 1. MCNP5 geometry model representing the source in three different positions and water sphere targets inside a water phantom for transit dose profilecalculation.

Medical Physics, Vol. 40, No. 5, May 2013

051717-4 Fonseca et al.: 192Ir HDR transit dose profile 051717-4

observed in the experimental profiles are due to the transitdose effect and not due to the phantom size.

III. RESULTS AND DISCUSSION

III.A. Monte Carlo validation

The air-kerma strength per activity (Sk/A) obtained inthis work is 9.79 × 10−8 U Bq−1 differing by 0.7% fromthe value obtained by Borg and Rogers,14 9.72 × 10−8

U Bq−1. The dose-rate constant value obtained, (1.109± 0.011) cGy h−1 U−1, is in good agreement with the valueobtained by Daskalov et al.,13 (1.108 ± 0.001) cGy h−1

U−1, notwithstanding the differences in dose rate at severalpoints around the source, which vary by ±1% for the ma-jority of the points. Besides the aforementioned comparisons,ONCENTRA R© and MCNP5 dwell dose profile results werealso compared in the studied clinical case, presenting a goodagreement as expected, since the TG-43U1 formalism em-ployed on ONCENTRA R© is also based on MC simulations.

III.B. Studied cases

In this subsection, we investigate the impact of the transitdose on the total dose profile using both the instantaneous vins

and the average 〈v〉 source speeds. We define here the totaldose profile, Dtot, as the sum of two components: the dwelldose, Ddw, and the transit dose, Dtr.

Figures 2(a) and 2(b) show the results obtained for thefour aforementioned cases, with the score axis positionedat 2.5 mm from the catheter, consisting of the instanta-neous speed profile of the source (first row), the respectiveabsolute total dose distribution, the dwell dose distribu-tion and the transit dose distribution, normalized to the air-kerma strength (second row), and the ratios Dtot

ins/Ddw andDtot

avg/Ddw to show the impact of using instantaneous speedon the transit and total dose (third row). The average speedvalue adopted in each case is presented in the speed pro-file, considering that the transit dose profiles were obtainedthrough the arrival and return average speed of 24.9 cm/s,which was obtained considering the instantaneous speed pro-file over the last 11 cm before the first dwell position and theinterdwell average speed for each studied case.

The dwell time is tdw = 1 s for all cases and was definedconsidering that the transit dose effect is more prominent incases with short dwell times, which are easier to validate ex-perimentally and that 1 s is the lowest reproducible dwelltime. Experiments using dwell times less than 1 s showedhigher variation and could not be reproduced. Some experi-ments were also performed using dwell times of 2 s and haveshown similar results. Figures 2(a)(a2-b2) and 2(b)(c2-d2)show that the transit dose using the instantaneous speed pro-file, Dtr

ins, significantly impacts the total dose profile, Dtotins,

near to the dwell positions xn, while the impact of the tran-sit dose using the average speed, Dtr

avg, is less significant onthe total dose, Dtot

avg. This is a consequence of the sourceslowing down near the dwell positions, which is taken intoaccount when considering instantaneous source speed. Thisresult demonstrates that Dtr

avg fails to accurately correct the

FIG. 2. (a) Results obtained from studied cases a and b: the first row (a1-b1) presents the instantaneous source speed, the second (a2-b2) presents theabsolute dose profiles, and the third (a3-b3) presents the ratio of dose profiles.(b) Results obtained from studied cases c and d: the first row (c1-d1) presentsthe instantaneous source speed, the second (c2-d2) presents the absolute doseprofiles, and the third (c3-d3) presents the ratio of dose profiles.

Dtot profile in the vicinity of dwell positions, which are theregions of interest for a clinical treatment. At distances farenough from the dwell position Dtr

avg ≈ Dtrins since vins ap-

proaches 〈v〉.Taking the particular case near the dwell position of

xn = 6 cm, Fig. 2(a)(a2), the dwell and transit dose us-ing ν ins are Ddw(xn) = 0.152 Gy h−1 U−1 and Dtr

ins(xn)

Medical Physics, Vol. 40, No. 5, May 2013

051717-5 Fonseca et al.: 192Ir HDR transit dose profile 051717-5

TABLE I. Transit dose comparison between experimental results obtainedwith alanine (Ref. 21) and a result obtained with MCNP5 at the position 8 cm(this work). The dwell dose obtained with MCNP5 at the compared point is3.54 × 10−13 cGy Bq−1.

Transit dose(10−11 cGy Bq−1)

MCNP5 0.80 ± 0.01Alanine-calibration curve obtained with a Linac 0.97 ± 0.19Alanine-calibration curve obtained with anIr-192 source

1.00 ± 0.32

= 0.033 Gy h−1 U−1, respectively. Therefore, the impactdue to the transit dose correction near the dwell position isDtot

ins(xn)/Ddw(xn) = [Ddw(xn) + Dtrins(xn)]/Ddw(xn) ∼ [tdw

* 0.152 Gy h−1 U−1 + 0.033 Gy h−1 U−1]/[tdw * 0.152 Gyh−1 U−1] ∼ [1 + (0.22/tdw)]. Although the transit dose ineach dwell point varies with the interdwell distance, the transitdose component is approximately 22% of the dwell dose for adwell time of 1 s corresponding to an effective transit time of0.22 s, which is on the same order as (0.4 ± 0.1) s obtained byAde.19 The difference between the results could be attributedto the experimental uncertainty and also to differences in thesource speed profiles of each piece of equipment.

The impact of the transit dose is slightly higher in the dwellpositions near the end position of the catheter since the sourcereturns back to the safe passing once again through these po-sitions with lower speed at these points, as can be observedin the return source speed profiles shown in Figs. 2(a) and2(b). Such impact is, thus, particularly relevant in treatmentswith high dose rate, several catheters, multiple dwell posi-tions, and/or short dwell times since treatments may employdwell times as short as 0.1 s at specific hot spots.20

Considering that the absolute transit dose is independent ofthe dwell time and corresponds to 22% of the dose delivereddue to a 1 s dwell time, for a particular case correspondingto the shortest allowed dwell time of 0.1 s the transit dose

would be approximately 220% of the delivered dose. How-ever, in practical cases the delivered dose in each position iscomposed of the contribution of several dwell positions withdwell times significantly higher than 0.1 s for the majority ofthe positions, which explains why even with some short dwelltimes the relation between the transit dose and the dwell doseis almost constant for dose distributions which are approxi-mately homogeneous.

Figures 2(a)(a3-b3) and 2(b)(c3-d3) show the dose ra-tio profiles where one can observe the dose impact,Dtot

ins/Ddw and Dtotavg/Ddw. The impact in the dose valley re-

gions between interdwell positions (x = d/2) is more compli-cated to estimate analytically since it depends not only on thespeed profile in that region but also on the significant overlapof dose peaks from the dwell positions for small d. Smaller dvalues result in higher impacts due to the low source speed.

The result obtained with the scoring axis positioned 0.5cm from the source axis and at 8 cm from the dwell positionwas compared to the value obtained by Calcina et al.21 andpresented in Table I, where the MCNP5 uncertainty consistsof a Type A (k = 1) component of uncertainty.

The other compared positions have also presented similardose value conformity between simulated and experimentalresults, except at 1 cm from the dwell position, where the cor-rection time applied by ONCENTRA R© to reduce the transitdose effect could present a more significant effect, especiallyconsidering that the dwell times employed in the experimentsperformed by Calcina et al. are about 0.1 s. The agreementbetween the results was expected, even considering that theresults obtained by Calcina et al. were obtained using the mi-croSelectron v1 (classic) source model while this work usedmicroselectron v2, since these models are not significantlydifferent, and the source maximum speed of 52 cm/s is thesame in both works, yielding a similar speed profile.21

The validation of the transit dose correction was per-formed by comparing the MCNP5 relative total dose profilesusing ν ins with those measured using radiochromic filmdosimeters, as shown in Figs. 3(a) and 3(b). The relative

FIG. 3. Relative dose profiles obtained with ONCENTRA R© (Ddw), MCNP5 (Dtotins), and from film measurements for two different cases: (a) dwell position

at 6 cm and (b) dwell positions at −4 and 6 cm. Both cases have a dwell time of 1 s. The right vertical axis presents the ratio of the measured dose to the totaldose and the dwell dose.

Medical Physics, Vol. 40, No. 5, May 2013

051717-6 Fonseca et al.: 192Ir HDR transit dose profile 051717-6

TABLE II. Uncertainty components obtained from experimental results andfrom Aldelaijan et al. (Ref. 22). The methodology proposed by Devic et al.(Ref. 18) eliminates calibration curve fits and reference dose uncertainty val-ues, reducing the experimental uncertainty. Uncertainty values were multi-plied by

√2 since the results consist of the ratio between two points.

Uncertainty components Type A (%) Type B (%)

Source-to-film positioning 0.5Scanner homogeneity 0.2Scanner reproducibility 0.1Net OD measurementreproducibility

1.1

Mean of the experimental pointsstandard deviation—case a

2.1

Mean of the experimental pointsstandard deviation—case b

3.5

Total uncertainty (k = 1)multiplied by

√2—case a

3.4

Total uncertainty (k = 1)multiplied by

√2—case b

5.2

dose profiles were normalized to the dose at the first dwellposition because the differences in the peak width and inthe dose between dwell positions can be easily visualized.Since no significant differences have been observed in the rel-ative profiles obtained considering the real geometry and theprofiles simulated in the middle of a homogeneous phantom,the latter was adopted to reproduce the configuration used byONCENTRA R©. Two cases aforementioned (cases a and b)were investigated. For both cases, the MCNP5 simulationscorrected by the ν ins have shown good agreement with theexperimental measurements with a maximum difference ofapproximately 3% of the maximum dose, while the resultsobtained without the transit dose do not agree with the rela-tive profiles obtained experimentally, presenting differencesof up to 7% of the maximum dose.

Except for differences induced by the transit dose profile,some differences shown here could be partially due to ex-perimental uncertainty, which has been obtained by summing

in quadrature the mean standard deviation of three adjacentrelative profiles obtained for each case and the uncertaintycomponents obtained from the literature, as described inTable II.22

The group mean of the standard deviations of each casewas obtained by averaging the standard deviation of each ex-perimental point, which was obtained through three relativeprofiles. The standard deviation of each point is shown inFig. 3 presenting mean values of 2.12 ± 0.02% and 3.5± 2.4% (Type A, k = 1) for cases a and b, respectively.

In Fig. 3, case b presented more noise effects in low doseregions, so these uncertainties could be reduced by applyingfilters, averaging over two or more pixels. However, severalfilters which were tested changed the peak widths, which isundesirable since that is the region where all the relative pro-files can be distinguished.

III.C. Clinical case

All dwell dose profiles obtained with MCNP5 show differ-ences of less than 1.5% when compared with ONCENTRA R©,which with transit dose correction applied will reduce thedwell times by 0.01 s, which represents 0.3% of the short-est dwell time. This effect is not significant when comparedto differences of up to 5.1%, which are obtained when consid-ering the instantaneous source speed, as can be seen in Fig. 4.This figure presents dose distributions only for two axes, sincethe other axis profiles are similar due to the symmetry of thegeometry adopted.

In Sec. III.B it was stated that the transit dose is approxi-mately 22% of the dwell dose considering a dwell time of 1 s.This result can be utilized with an approximation for morecomplex cases through the average dwell time, for example,considering 〈tdw〉 = [(

∑Ni tdwi

)/N], where N is the number ofdwell positions. In this case, the average dwell time will be5.1 s, so that the fraction of transit dose is 5.1 times lowerthan 22% or 4.3% of the prescribed dose. This is a reasonableapproximation, especially when dwell positions are spaced

FIG. 4. Dwell dose profiles obtained with MCNP5 and ONCENTRA R©, transit and total dose profiles obtained with MCNP5 considering the instantaneoussource speed for a hypothetical prostate case using 10 catheters with a treatment length of 5 cm, totaling 110 dwell positions and 563.2 s of treatment time. (a)Score axis at the center; (b) score axis 1.12 cm from the center.

Medical Physics, Vol. 40, No. 5, May 2013

051717-7 Fonseca et al.: 192Ir HDR transit dose profile 051717-7

equidistantly and the dose distribution is approximately ho-mogeneous. This approach does not eliminate the necessityof transit dose calculations in real cases, especially wheredose distributions cannot be homogeneous to preserve someorgans.

The average of the maximum transit dose using the instan-taneous source speed correction was obtained by averagingthe maximum value of transit dose in each score axis, result-ing in a value of 80.2 cGy. The transit dose behavior is sim-ilar in all axes analyzed, with higher variation at the centerof the central axis, where the maximum transit dose is 104.0cGy. This value cannot be directly compared with those inthe literature5–7, 9 since the results do not represent the samesituation, although the results presented so far in the litera-ture are approximately a few cGy per fraction, which is upto one order of magnitude lower than the results obtained inthis work. These differences are due to the source speed andcan be verified by calculating the average speed through theinstantaneous speed profile, which can be compared with thevalues in the literature, e.g., using the instantaneous sourcespeed profile, the average speed for an interdwell distance of0.5 cm was calculated to be 3.8 cm/s, which is smaller than thevalues of 33.3 cm/s measured by Sahoo,6 27.1 cm/s measuredby Houdek et al.,7 25.5 cm/s measured by Bastin et al.,20 and50 cm/s described in the ONCENTRA R© userguide.9 Thesedifferences result in an underestimated transit dose, e.g., themaximum transit dose obtained in this work for a hypo-thetical clinical case (104.0 cGy) is 2.77 times higher thanthe value considered as a maximum value, 37.5 cGy, by theONCENTRA R© userguide.9 Using 50 cm/s, the average tran-sit dose is reduced to 8.1 cGy and the maximum transit doseis reduced to 10.4 cGy, which is about the same order of mag-nitude as the literature results.

IV. CONCLUSION

Transit dose values calculated based on instantaneoussource speed measurements have shown significant differ-ences compared to values obtained using average sourcespeed, which underestimates the dose. These differences aremore pronounced near the dwell positions when the sourcespeed decreases. Differences have also been confirmed withdose measurements using radiochromic films. Transit doseprofiles could reach up to a few hundred cGy per applicationin treatments with several catheters, a high activity source,and fractioned applications. Even when considering higheraverage speed, the transit dose could be significant,7, 20 whichhighlights the importance of considering the transit dose withthe correct speed profile, which is slower than that previouslyadopted. Using an average dwell time it is possible to estimatethe transit dose component in some cases, however, transitdose profiles vary considerably due to the dwell positions andneed to be adequately calculated to obtain an accurate dose fortreatment planning. The instantaneous speed measurements4

can be performed directly by the hospital or by the sourcemanufacturer, providing a speed profile which can be used tocalculate the transit dose using methodologies described inthe literature,20 MC simulations, or the TG-43U1 formalism.

ACKNOWLEDGMENTS

The authors would like to thank Professor Frank Verhae-gen for his valuable comments. This work was supported byFundação de Amparo à Pesquisa do Estado de São Paulo(FAPESP), Grant Nos. 2011/01913-4, 2011/23765-7, and2011/22778-8.

a)Author to whom correspondence should be addressed. Electronic mail:hyoriyaz@ipen.br; Telephone: +55 (11) 3133-9482; Fax: +55 (11) 3133-9423.

1International Commission on Radiation Units and Measurements, “Doseand volume specification for reporting intracavitary therapy in gynecol-ogy,” ICRU Report No. 38 (ICRU, Bethesda, MI, 1985).

2International Commission on Radiation Units and Measurements, “Doseand volume specification for reporting interstitial therapy,” ICRU ReportNo. 58 (ICRU, Bethesda, MI, 1997).

3I. Gerardy, J. Ródenas, and S. Gallardo, “Uncertainty analysis in MCNP5calculations for brachytherapy treatment,” Appl. Radiat. Isot. 69, 1108–1111 (2011).

4R. A. Minamisawa, R. Rubo, R. M. Seraide, J. R. O. Rocha, andA. Almeida, “Direct measurement of instantaneous source speed for aHDR brachytherapy unit using an optical fiber based detector,” Med. Phys.37(10), 5407–5411 (2010).

5T. P. Y. Wong, W. Fernando, P. N. Johnston, and I. F. Bubb, “Transit dose ofan Ir-192 high dose rate brachytherapy stepping source,” Phys. Med. Biol.46, 323–331 (2001).

6N. Sahoo, “Measurement of transit time of a remote after-loading high doserate brachytherapy source,” Med. Phys. 28(8), 1786–1790 (2001).

7P. V. Houdek, J. G. Schwade, X. Wu, V. Pisciotta, J. A. Fiedler, C. F. Serago,A. M. Markoe, A. A. Abitbol, A. A. Lewin, P. G. Braunschweiger, andM. D. Sklar, “Dose determination in high dose-rate brachytherapy,” Int. J.Radiat. Oncol., Biol., Phys. 24, 795–801 (1992).

8L. Beaulieu, A. C. Tedgren, J. F. Carrier, S. D. Davis, F. Mourtada, M. J. Ri-vard, R. M. Thomson, and F. Verhaegen, “Report of the Task Group 186 onmodel-based dose calculation methods in brachytherapy beyond the TG-43formalism: Current status and recommendations for clinical implementa-tion,” Med. Phys. 39(10), 6208–6236 (2012).

9Planning system manual, ONCENTRA R© MasterPlan v3.3 Physics and Al-gorithms.

10J. F. Briesmeister, “MCNP: A general Monte Carlo N-particle transportcode, version 5,” Los Alamos Scientific Laboratory, Los Alamos, NewMexico, 2008.

11R. Nath, L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson,and A. S. Meigooni, “AAPM Technical Report 51: Dosimetry of in-terstitial brachytherapy sources: Recommendations of the AAPM Radia-tion Therapy Committee Task Group No. 43,” Med. Phys. 22, 209–234(1995).

12M. J. Rivard, B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq,G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson, “Technical Re-port 84 - Update of AAPM Task Group No. 43 Report: A revised AAPMprotocol for brachytherapy dose calculations,” Med. Phys. 31, 633–674(2004).

13G. M. Daskalov, E. Löffler, and J. F. Williamson, “Monte Carlo-aideddosimetry of a new high dose-rate brachytherapy source,” Med. Phys. 25,2200–2208 (1998).

14J. Borg and D. W. O. Rogers, Monte Carlo Calculations of Photon Spectrain Air from 192Ir Sources (Institute for National Measurement Standards,Ottawa, Canada, 1999).

15R. E. P. Taylor and D. W. O. Rogers, “EGSnrc Monte Carlo calculateddosimetry parameters for 192Ir and 169Yb brachytherapy sources,” Med.Phys. 35(11), 4933–4944 (2008).

16R. F. Hill, S. Brown, and C. Baldock, “Evaluation of the water equivalenceof solid phantoms using gamma ray transmission measurements,” Radiat.Meas. 43, 1258–1264 (2008).

17IMAGEJ. Last accessed 26 April 2012 (available URL: http://rsb.info.nih.gov/ij/).

18S. Devic, N. Tomic, S. Aldelaijan, F. DeBlois, J. Seuntjens, M. F. Chan,and D. Lewis, “Linearization of dose-response curve of the radiochromicfilm dosimetry system,” Med. Phys. 39(8), 4850–4857 (2012).

Medical Physics, Vol. 40, No. 5, May 2013

051717-8 Fonseca et al.: 192Ir HDR transit dose profile 051717-8

19N. Ade, “Transit dosimetry in 192Ir high dose rate brachytherapy,” M.S.thesis, University of the Witwatersrand, Johannesburg, South Africa,2009.

20K. T. Bastin, M. B. Podgorsak, and B. R. Thomadsen, “The transit dosecomponent of high dose rate brachytherapy: Direct measurements andclinical implications,” Int. J. Radiat. Oncol., Biol., Phys. 26(4), 695–702(1993).

21C. S. G. Calcina, A. Almeida, J. R. O. Rocha, F. C. Abrego, and O. Baffa,“Ir-192 HDR transit dose and radial dose function determination using ala-nine/EPR dosimetry,” Phys. Med. Biol. 50, 1109–1117 (2005).

22S. Aldelaijan, H. Mohammed, N. Tomic, L. Liang, F. DeBlois, A. Sar-fehnia, W. Abdel-Rahman, J. Seuntjens, and S. Devic, “Radiochromic filmdosimetry of HDR 192Ir source radiation fields,” Med. Phys. 38(11), 6074–6083 (2011).

Medical Physics, Vol. 40, No. 5, May 2013