Evaluation of brachytherapy lung implant dose distributions from photon-emitting sources due to tissue heterogeneities

Yun Yang and Mark J. Rivard

Tufts University School of Medicine, Boston, Massachusetts 02111

(Received 26 May 2011; revised 28 August 2011; accepted for publication 31 August 2011; published 11 October 2011) Purpose: Photon-emitting brachytherapy sources are used for permanent implantation to treat lung cancer. However, the current brachytherapy dose calculation formalism assumes a homogeneous water medium without considering the inuence of radiation scatter or tissue heterogeneities. The purpose of this study was to determine the dosimetric effects of tissue heterogeneities for permanent lung brachytherapy. Methods: The MCNP5 v1.40 radiation transport code was used for Monte Carlo (MC) simulations. Point sources with energies of 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV were simulated to cover the range of pertinent brachytherapy energies and to glean dosimetric trends independent of specic radionuclide emissions. Source positions from postimplant CT scans of ve patient implants were used for source coordinates, with dose normalized to 200 Gy at the center of each implant. With the presence of brosis (around the implant), cortical bone, lung, and healthy tissues, dose distributions and PTVDVH were calculated using the MCNP *FMESH4 tally and the NIST mass-energy absorption coefcients. This process was repeated upon replacing all tissues with water. For all photon energies, 109 histories were simulated to achieve statistical errors (k 1) typically of 1%. Results: The mean PTV doses calculated using tissue heterogeneities for all ve patients changed (compared to dose to water) by only a few percent over the examined photon energy range, as did PTV dose at the implant center. The PTVV100 values were 81.2%, 90.0% (as normalized), 94.3%, 93.9%, 92.7%, and 92.2% for 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV source photons, respectively. Relative to water, the maximum bone doses were higher by factors of 3.7, 5.1, 5.2, 2.4, 1.2, and 1.0 The maximum lung doses were about 0.98, 0.94, 0.91, 0.94, 0.97, and 0.99. Relative to water, the maximum healthy tissue doses at the mediastinal position were higher by factors of 9.8, 2.2, 1.3, 1.1, 1.1, and 1.1. However, the maximum doses to these healthy tissues were only 3.1, 7.2, 11.3, 10.9, 9.0, and 8.1 Gy while maximum bone doses were 66, 177, 236, 106, 49, and 39 Gy, respectively. Similarly, maximum lung doses were 55, 66, 73, 74, 73, and 73 Gy, respectively. Conclusions: The current brachytherapy dose calculation formalism overestimates PTV dose and signicantly underestimates doses to bone and healthy tissue. Further investigation using specic brachytherapy source models and patient-based CT datasets as MC input may indicate whether the C observed trends can be generalized for low-energy lung brachytherapy dosimetry. V 2011 American Association of Physicists in Medicine. [DOI: 10.1118/1.3641872] Key words: lung, brachytherapy, Monte Carlo, dosimetry, permanent implants

I. INTRODUCTION Lung cancer is the leading cause of cancer death for both men and women in U.S.1 In 2010, an estimated 222,520 new cases of lung cancer and 157,300 deaths are expected.1 Both lobectomy and sublobar resection can be used to treat patients with stage I nonsmall cell lung cancer. Usually a course of radiation therapy will be provided after sublobar resection to reduce cancer recurrence. Available options of radiation therapy include external beam radiation therapy and brachytherapy, such as permanent 125I and 131Cs implant.26 Brachytherapy can be advantageous in comparison to external beam radiotherapy due to its dose conformity and positioning accuracy at the time of surgery. Using permanent 125I implants, cardiac toxicity, lung brosis, and loss of pulmonary function are observed infrequently.7 Permanent lung brachytherapy implants have also been shown
5857 Med. Phys. 38 (11), November 2011

to diminish the recurrence rates signicantly compared to the use of sublobar resection alone.2,7 Current brachytherapy dose calculations are based on the series of American Association of Physicists in Medicine (AAPM) TG-43 reports.810 This dosimetry formalism assumes a homogeneous water medium without considering the inuence of scatter or material heterogeneities. Brachytherapy sources for lung implants are generally characterized dosimetrically in treatment planning systems (TPS) as 1D point sources and the dose distribution is obtained for a liquid water phantom of a xed volume having full radiation scattering. While this TPS approach may be improved through using advanced brachytherapy dose calculation algorithms,11,12 no FDA-approved TPS are currently available that account for radiation scatter conditions and tissue heterogeneities (i.e., brosis around the implant or planning target volume (PTV), cortical bone, lung, and healthy tissues) for low-energy
C V 2011 Am. Assoc. Phys. Med.

0094-2405/2011/38(11)/5857/6/$30.00

5857

5858

Y. Yang and M. J. Rivard: Tissue heterogeneity-corrected dose distributions for brachytherapy lung implants

5858

brachytherapy sources.13 Until such TPS are available, it is of interest to estimate the dosimetric inuence of realistic conditions present in brachytherapy implants. While lung tissue heterogeneities have been investigated, only high-energy sources such as 192Ir and 137Cs have been examined.1416 Therefore, research on accurate brachytherapy dose calculations for low-energy lung implants should be performed. II. MATERIALS AND METHODS Half of the chest volume was simulated as 15 20 20 cm3, with 15 cm in the X-axis (i.e., lateral left=right), 20 cm in the Y-axis (i.e., posterior=anterior), and 20 cm in the Z-axis (i.e., inferior=superior) directions. Tissue compositions were taken from ICRU 44,17 with healthy (soft) tissues (q 1.06 g cm3) of 5 cm thickness in all three-axes except the X direction (3.0 cm). To maximize the dosimetric effect of tissue heterogeneities on healthy tissue dose calculations, a small (7 10 10 cm3) lung volume composed of normal lung tissue (q 0.20 g cm3) was examined. The brotic tissue (PTV margin) created following surgical sublobar resection was simulated as lung tissue of unity density (q 1.00 g cm3) within the healthy lung. The rib was simulated as a 1.5 cm diameter right cylinder, comprised of cortical bone (q 1.92 g cm3) and a 1.0 cm inner diameter marrow core (q 1.06 g cm3). These dimensions were included in the Monte Carlo (MC) simulations (Fig. 1). Based on our institutional experience, these dimensions and tissue assignments were considered as representative for patients receiving lung brachytherapy implants. All patients were implanted with two strands of ten 125I seeds spaced 1 cm center-to-center except for patient #3 having one strand. Source positions from postimplant CT scans were used for source coordinates. While these linear strands had collapsed upon clinical implantation into a chaotic geometrical distribution, position of each individual source was readily observed upon adjusting the TPS window-and-level to discern seeds versus surgical clips. The cell transformation card (TRCL) (Ref. 18) is a tool within the MC program

that makes it possible to dene only once the surfaces that bound several cells identical in size and shape, but located at different places in the geometry such as each monoenergetic point source to represent the brachytherapy source positions from the clinical implants. The MCNP5 radiation transport code (version 1.40) was used for all simulations with photoatomic cross-sections based on EPDL97.1820 The photon-emitting brachytherapy sources were simulated as point sources. Photon energies E of 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV were chosen to cover the range of pertinent brachytherapy energies and to glean dosimetric trends independent of specic radionuclide emissions. The 0.02 MeV source roughly correlated with 103Pd, 0.03 MeV with 125 I and 131Cs, 0.1 MeV with 169Yb, and 0.4 MeV with 192Ir. The brotic tissue was simulated to have nominal dimensions of 3 4 7 cm3. Additional simulations of 2 4 7 and 4 4 7 cm3 volumes were also performed to study the dose sensitivity on the examined volume. Fibrotic tissue sizes of 2 4 7, 3 4 7, and 4 4 7 cm3 corresponded to PTV sizes of 0.4 2 5, 1 2 5, and 2 2 5 cm3, respectively. The ratio of the mean PTV, maximum bone, maximum lung, and maximum tissue doses to water as a function of source photon energy with different brotic lesion sizes were obtained and graphed. Dose distributions were calculated throughout the PTV with (0.1 cm)3 voxels, and dose distributions were imported into a spreadsheet to calculate the PTVDVH. Dose was normalized to 200 Gy at the PTV center (coordinate system origin) to approximate the clinical prescriptive scheme of prescribing 100 Gy at a distance of 0.7 cm from the implant.4 For PTV, absorbed dose were calculated in the X-Y, Y-Z, and Z-X planes at Z 0, X 0, and Y 0, respectively. For bone, lung, and healthy tissues, absorbed dose were calculated in the Y-Z planes at X 2.05, X 1.55, and X 5.55 cm, respectively. The planar resolution was (0.5 cm)2. Collision kerma to tissue in tissue was calculated to approximate absorbed dose using the photon energy uence estimator (*FMESH4 tally) and the NIST massenergy absorption coefcients len=q.21 Only photon scoring was performed since dosimeric effects of electrons were assumed to be negligible for this study.22 This entire simulation process was repeated by replacing all tissues with water to approximate results from the TG-43 dose calculation formalism as available in conventional brachytherapy TPS. For each MC simulation, 109 photon histories were simulated. These calculations took three days each on a 3 GHz PC running the Windows XP operating system. Typical statistical uncertainties (k 1) for all photon energies were 0.3% at the PTV center and 3% in the far corner of the healthy tissues. III. RESULTS In general, the results indicated that variations in source coordinates did not signicantly inuence the PTV, bone, lung, and healthy tissue.

FIG. 1. Lung geometry for the Monte Carlo simulation, depicting an anteriorposterior view with the superior direction oriented at top. The brotic lesion is 3 4 7 cm3 in the lateral, anterior, and superior directions, respectively. Medical Physics, Vol. 38, No. 11, November 2011

III.A. PTV

Compared to dose to water in water, the heterogeneitycorrected dose to tissue in tissue at the coordinate system

5859

Y. Yang and M. J. Rivard: Tissue heterogeneity-corrected dose distributions for brachytherapy lung implants

5859

origin changed by 1.7, 2.6, 0.7, 0.8, 0.6, and 0.2% for 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV source photons, respectively. The mean PTV doses [Fig. 2(a)] calculated using tissue heterogeneities for all ve patients changed by 1.5, 0.9, 2.4, 3.8, 2.6, and 1.7% when compared to dose to water for the aforementioned photon energies. Plotting PTVDVH for the six source photon energies showed superior dose coverage for 0.05 MeV (Fig. 3). The V100 values were 81.2%, 90.0% (as normalized), 94.3%, 93.9%, 92.7%, and 92.2% for 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV, respectively. The regions of high dose exceeding V600 on the Fig. 3 X-axis are artifacts due to the point sources and (0.1 cm)3 voxels and are not expected upon simulating realistic brachytherapy sources.23 Differences in brotic tissue size surrounding the PTV did not substantially (within 3%) affect heterogeneitycorrected mean doses PTV or maximum dose sto bone, lung, and tissue as shown in Fig. 4.
III.B. Bone

FIG. 3. PTVDVH for patient #2 with source photon energies of 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV. The V100 values for these photon energies were 81.2%, 90.0% (as normalized), 94.3%, 93.9%, 92.7%, and 92.2%, respectively.

Relative to dose to water in water, the maximum bone doses [Fig. 2(b)] were 3.7, 5.1, 5.2, 2.4, 1.2, and 1.0 for the aforementioned photon energies. In general, differences between bone and water dose increased as photon energy increased, and reached a maximum at E $ 0.05 MeV. For E > 0.05 MeV, differences between bone and water dose

decreased as photon energy increased. For the mean of all ve patients when accounting for tissue heterogeneities, the maximum bone doses were 66, 177, 236, 106, 49, and 39 Gy at each of the aforementioned photon energies. Similarly, the mean bone doses were 8.1, 35.5, 70.4, 38.2, 16.8, and 12.6 Gy. Given that signicant likelihood for bone fracture can occur at doses exceeding 59 Gy,24 there is concern that

FIG. 2. Ratio of the mean PTV, maximum bone, maximum lung, and maximum tissue doses to water as a function of photon energy. The curves are meant to guide the eye, with data points obtained for 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV photons. Medical Physics, Vol. 38, No. 11, November 2011

5860

Y. Yang and M. J. Rivard: Tissue heterogeneity-corrected dose distributions for brachytherapy lung implants

5860

FIG. 4. Ratio of the mean PTV, maximum bone, maximum lung, and maximum tissue doses to water as a function of source photon energy for patient #1 with different brotic lesion sizes. The curves are meant to guide the eye, with data points obtained for 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV source photons.

the inability of conventional brachytherapy TPS to account for material heterogeneities may prevent the radiation oncologist from being concerned with bone dose or correlating clinical outcomes to realistic doses. Future clinical studies should correlate the incidence of rib fracture or costocondritis with accurate lung brachytherapy dosimetry.
III.C. Lung

Relative to dose to water in water, the maximum lung doses [Fig. 2(c)] were 0.98, 0.94, 0.91, 0.94, 0.97, and 0.99. for the aforementioned photon energies. Lung doses were always lower than water within $10% for the aforementioned photon energies. For the mean of all ve patients when accounting for tissue heterogeneities, the maximum lung doses were 55.1, 65.9, 72.9, 73.7, 73.2, and 73.1 Gy at each of the aforementioned photon energies. Similarly, the mean lung doses were 9.4, 17.2, 23.2, 23.3, 21.8, and 21.2 Gy, respectively. These doses obtained using the MC calculations that account for tissue heterogeneities are in agreement with the observed low pneumonitis rates following 125I lung brachytherapy.
III.D. Healthy tissue

implant at X 5.5 cm) was always higher than the corresponding dose to water, and increased on average (for the ve patients) by factors of 9.8, 2.2, 1.3, 1.1, 1.1, and 1.1 [Fig. 2(d)] for the 0.02, 0.03, 0.05, 0.1, 0.2, and 0.4 MeV source photons, respectively. Especially for the E < 0.05 MeV results, these dose ratios exceeded unity primarily due to the higher attenuation of water than lung tissue over 4 cm. However, the maximum healthy tissue doses were only 3.1, 7.2, 11.3, 10.9, 9.0, and 8.1 Gy for the aforementioned photon energies, respectively. The mean healthy tissue doses were only 0.6, 2.1, 4.5, 4.7, 3.8, and 3.4 Gy. However, different lesion positions will cause different mediastinal doses. IV. DISCUSSION The maximum tissue doses and mean PTV doses seemed most clinically-relevant. Maximum tissue doses will correlate more so with toxicities than mean tissue doses, while mean PTV doses reect target coverageheterogeneous dose distributions are a hallmark of brachytherapy implants. While all calculated doses were dependent on the simulated geometry, the lesion (PTV) size did not signicantly affect the tissue doses or the mean PTV doses as shown in Fig. 2. For cortical bone, the ratio of len=q for tissue to water shown in Fig. 5 is qualitatively similar to that in Fig. 2(b). The simulation results peak at 0.05 MeV with bone dose 5.2

The maximum healthy tissue dose (approximating mediastinal dose, positioned at Z Y 0 opposite the lung from the
Medical Physics, Vol. 38, No. 11, November 2011

5861

Y. Yang and M. J. Rivard: Tissue heterogeneity-corrected dose distributions for brachytherapy lung implants

5861

patients. The 0.02 MeV source was the least conformal, and differences between 0.030.4 MeV were minimal. Given concerns for radiation safety and desire to minimize critical structure doses, it seems that the dose conformity available from low-energy photon-emitting brachytherapy sources near 0.05 MeV (such as 125I and 131Cs) may be most desirable for lung brachytherapy. V. CONCLUSION Using a simplied lung implant geometry with unencapsulated mono-energetic photon-emitting point sources, the origin dose and mean PTV dose differed by only a few percent in comparison to the current brachytherapy dose calculation formalism (TG-43) based on dose to water in water. However, the TG-43 formalism signicantly underestimates bone and healthy tissue doses by factors of three or more. Further investigation using specic source models and patient-based CT datasets as MC input for tissue may indicate whether the observed energy trends can be generalized for lung brachytherapy implants, and how advanced dosimetry methods may help develop new metrics for correlating patient outcomes with accurate dosimetry. ACKNOWLEDGMENTS The authors thank colleagues at Tufts Medical Center for their input on this manuscript. Nolan Gagne assisted running some of the MC input les, while he, Christopher Melhus, and Kathryn Huber assisted by reviewing preliminary versions of this manuscript.
1 2

FIG. 5. Ratio of len=q for tissues to water as a function of photon energy (from the NIST website) (Ref. 23).

times larger than water while the NIST data peak at 0.03 MeV with bone dose about seven times higher than water. This difference is attributed to the higher penetration for increasing photon energies in our simulations and subsequent reduction in bone=water dose ratio compared to the thin-layer assumption for len=q. By not considering tissue heterogeneities, signicant errors in bone dose calculations are caused by the current TG-43 based brachytherapy dosimetry formalism. Analyzing dose to soft tissues, differences between tissue and water resulted mainly from medium attenuation differences. While maximum healthy tissue doses at 0.02 and 0.03 MeV were 9.8 and 2.2 times higher than water dose, respectively, these dose points were positioned at large distances where doses were low. Dose differences between tissue and water were always larger than the statistical uncertainties (k 1). For 0.4 MeV source photons, doses to PTV, bone, lung, and tissue were only a few percent lower than the corresponding doses to water. These agreed with literature results for 192Ir.15 Unlike some studies in which only a single source was involved and only several points at a certain distance were investigated,16 the multiple sources and volume of interest approach in this study provided more clinicallyrelevant results for lung brachytherapy implants. Another study limitation is simulating brachytherapy sources as point sources. This likely caused the high-dose regions in Fig. 3. TPS using the TG-43 formalism do not account for intersource attenuation as similarly inherent to this MC study of unencapsulated point sources. Also, clinically-available low-energy lung brachytherapy seeds, such as 103Pd, 125I, and 131Cs, have an active length < 0.5 cm and thus can be approximated with the point-source approximation given the spatial scale of this study. By ignoring dose from electrons in the MC simulations, tissue and water dose may be underestimated. However, the range of primary and secondary electrons could not penetrate the PTVeven for 0.4 MeV source photons.25 From PTVDVH results, the 0.05 MeV source was the most conformal and had the steepest curve for all ve
Medical Physics, Vol. 38, No. 11, November 2011

American Cancer Society, Cancer Facts and Figures, 2010. R. Santos, A. Colonias, D. Parda, M. Trombetta, R. H. Maley, R. Macherey, S. Bartley, T. Santucci, R. J. Keenan, and R. J. Landreneau, Comparison between sublobar resection and 125Iodine brachytherapy after sublobar resection in high-risk patients with Stage I non-small-cell lung cancer, Surgery 134, 691697 (2003). 3 G. Voynov, D. E. Heron, C. J. Lin, S. Burton, A. Chen, A. Quinn, R. Santos, A. Colonias, and R. J. Landreneau, Intraoperative 125I Vicryl mesh brachytherapy after sublobar resection for high-risk stage I nonsmall cell lung cancer, Brachytherapy 4, 278285 (2005). 4 W. Lee, B. D. T. Daly, T. A. DiPetrillo, D. M. Morelli, A. C. Neuschatz, J. Morr, and M. J. Rivard, Limited resection for non-small cell lung cancer: Observed local control with implantation of I-125 brachytherapy seeds, Ann. Thorac. Surg. 75, 237243 (2003). 5 B. Parashar, A. G. Wernicke, A. Pavese, P. Singh, S. Trichter, A. Sabbas, D. I. Kutler, W. Kuhel, J. L. Port, P. C. Lee, D. Nori, and K. S. C. Chao, Cesium-131 permanent seed brachytherapy: Dosimetric evaluation and radiation exposure to surgeons, radiation oncologists, and staff, Brachytherapy (in press). 6 S. Thomas, H. Odau, R. K. Das, S. M. Bentzen, R. R. Patel, and D. Khuntia, Feasibility of Cs-131 for intraoperative brachytherapy for patients undergoing sublobar resections for early stage non-small cell lung cancer (NSCLC): A dosimetric comparison with I-125 and Cs-131, Brachytherapy 6, 108 (2007). 7 M. Johnson, A. Colonias, D. Parda, M. Trombetta, O. Gayou, B. Reitz, and M. Miften, Dosimetric and technical aspects of intraoperative I-125 brachytherapy for stage I non-small cell lung cancer, Phys. Med. Biol. 52, 12371245, (2007). 8 R. Nath, L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni, Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43, Med. Phys. 22, 209234 (1995).

5862
9

Y. Yang and M. J. Rivard: Tissue heterogeneity-corrected dose distributions for brachytherapy lung implants
18

5862

M. 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, Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations, Med. Phys. 31, 633674 (2004). 10 M. J. Rivard, W. M. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, A. S. Meigooni, C. S. Melhus, M. G. Mitch, R. Nath, and J. F. Williamson, Supplement to the 2004 update of the AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations, Med. Phys. 34, 21872205 (2007). 11 M. J. Rivard, J. L. M. Venselaar, and L. Beaulieu, The evolution of brachytherapy treatment planning, Med. Phys. 36, 21362153 (2009). 12 M. J. Rivard, L. Beaulieu, and F. Mourtada, Necessary enhancements to commissioning techniques of brachytherapy treatment planning systems that use model-based dose calculation algorithms, Med. Phys. 37, 26452658 (2010). 13 B. R. Thomadsen, J. F. Williamson, M. J. Rivard, and A. S. Meigooni, Anniversary Paper: Past and current issues, and trends in brachytherapy physics, Med. Phys. 35, 47084723 (2008). 14 I. Das, S. Both, and C. Cheng, Inhomogeneity correction in brachytherapy, Med. Phys. 33, 2112 (abstract) (2006). 15 R. M. Chandola, S. Tiwari, M. K. Kowar, and V. Choudhary, Effect of inhomogeneities and source position on dose distribution of nucletron high dose rate Ir-192 brachytherapy source by Monte Carlo simulation, J. Cancer Res. Ther. 6, 5457 (2010). 16 G. Gialousis, A. Dimitriadis, and E. Yakoumakis, Monte Carlo estimation of dose difference in lung from Ir-192 brachytherapy due to tissue inhomogeneity, International Conference on Radiation Protectection in Medicine, S3.P1 eposter (2010), http://www.rpm2010.org/posters/S3.P1. Gialousis_50.pdf. Last accessed September 2011. 17 International Commission on Radiation Units and Measurements, Tissue substitutes in radiation dosimetry and measurement, ICRU Report No. 44 (ICRU, Bethesda, MD, 1989).

X-5 Monte Carlo Team, MCNP: A general Monte Carlo N-particle transport code, Version 5, Los Alamos National Laboratory, Los Alamos, NM, 2003. 19 M. White, Photoatomic data library MCPLIB04: A new photoatomic library based on data from ENDF=B-VI release 8, Los Alamos National Laboratory Memorandum No. LAUR-03-1019, 2003. 20 E. Cullen, J. H. Hubbell, and L. Kissel, EPDL97: The Evaluated Photon Data Library, 97 Version, Lawrence Livermore National Laboratory Report No. UCRL-50400, 1997. 21 J. H. Hubbell and S. M. Seltzer, Tables of x-ray mass attenuation coefcients and mass energy-absorption coefcients from 1 keV to 20 MeV for elements Z 1 to 92 and 48 additional substances of dosimetric interest, Ionizing Radiation Division, Physics Laboratory, The National Institute of Standards and Technology, NISTIR 5632, http://www.nist.gov/pml/data/ xraycoef/index.cfm. Last accessed September 2011. 22 F. Ballester, D. Granero, J. Perez-Calatayud, C. S. Melhus, and M. J. Rivard, Evaluation of high-energy brachytherapy source electronic disequilibrium and dose from emitted electrons, Med. Phys. 36, 42504256 (2009). 23 M. J. Rivard, Renements to the geometry factor used in the AAPM Task Group Report No. 43 necessary for brachytherapy dosimetry calculations, Med. Phys. 26, 2445250 (1999). 24 C. I. Dickie, A. L. Parent, A. M. Grifn, S. Fung, P. W. Chung, C. N. Catton, P. C. Ferguson, J. S. Wunder, R. S. Bell, M. B. Sharpe, and B. OSullivan, Bone fractures following external beam radiotherapy and limb-preservation surgery for lower extremity soft tissue sarcoma: relationship to irradiated bone length, volume, tumor location and dose, Int. J. Radiat. Oncol., Biol., Phys. 75, 11191124 (2009). 25 M. J. Rivard, D. Granero, J. Perez-Calatayud, and F. Ballester, Inuence of photon energy spectra from brachytherapy sources on Monte Carlo simulations of kerma and dose rates in water and air, Med. Phys. 37, 869876 (2010).

Medical Physics, Vol. 38, No. 11, November 2011