TESTICULAR DOSES IN IMAGE-GUIDED RADIOTHERAPY OF PROSTATE CANCER

Transcription

1 doi: /j.ijrobp Int. J. Radiation Oncology Biol. Phys., Vol. 82, No. 1, pp. e39 e47, 2012 Copyright Ó 2012 Elsevier Inc. Printed in the USA. All rights reserved /$ - see front matter CLINICAL INVESTIGATION Genitourinary Cancer TESTICULAR DOSES IN IMAGE-GUIDED RADIOTHERAPY OF PROSTATE CANCER JUN DENG, PH.D., ZHE CHEN, PH.D., JAMES B. YU, M.D., KENNETH B. ROBERTS, M.D., RICHARD E. PESCHEL, M.D., AND RAVINDER NATH, PH.D. Department of Therapeutic Radiology, Yale University, New Haven, CT Purpose: To investigate testicular doses contributed by kilovoltage cone-beam computed tomography (kvcbct) during image-guided radiotherapy (IGRT) of prostate cancer. Methods and Materials: An EGS4 Monte Carlo code was used to calculate three-dimensional dose distributions from kvcbct on 3 prostate cancer patients. Absorbed doses to various organs were compared between intensity-modulated radiotherapy (IMRT) treatments and kvcbct scans. The impact of CBCT scanning mode, kilovoltage peak energy (kvp), and CBCT field span on dose deposition to testes and other organs was investigated. Results: In comparison with one 10-MV IMRT treatment, a 125-kV half-fan CBCT scan delivered 3.4, 3.8, 4.1, and 5.7 cgy to the prostate, rectum, bladder, and femoral heads, respectively, accounting for 1.7%, 3.2%, 3.2%, and 8.4% of megavoltage photon dose contributions. However, the testes received 2.9 cgy from the same CBCT scan, a threefold increase as compared with 0.7 cgy received during IMRT. With the same kvp, full-fan mode deposited much less dose to organs than half-fan mode, ranging from 9% less for prostate to 69% less for testes, except for rectum, where full-fan mode delivered 34% more dose. As photon beam energy increased from 60 to 125 kv, kvcbct-contributed doses increased exponentially for all organs, irrespective of scanning mode. Reducing CBCT field span from 30 to 10 cm in the superior inferior direction cut testicular doses from 5.7 to 0.2 cgy in half-fan mode and from 1.5 to 0.1 cgy in full-fan mode. Conclusions: Compared with IMRT, kvcbct-contributed doses to the prostate, rectum, bladder, and femoral heads are clinically insignificant, whereas dose to the testes is threefold more. Full-fan CBCT usually deposits much less dose to organs (except for rectum) than half-fan mode in prostate patients. Kilovoltage CBCT contributed doses increase exponentially with photon beam energy. Reducing CBCT field significantly cuts doses to testes and other organs. Ó 2012 Elsevier Inc. Prostate cancer, Testicular doses, Image-guided radiotherapy, kvcbct, Monte Carlo simulation. INTRODUCTION Recently, kilovoltage cone-beam computed tomography (kvcbct) has been introduced into the radiotherapy community as a useful X-ray imaging tool for daily patient setup and tumor localization (1, 2). Consequently, our radiotherapy paradigms for prostate cancer have been shifting dramatically from three-dimensional conformal radiotherapy and intensity-modulated radiotherapy (IMRT) to most recent image-guided radiotherapy (IGRT). Imageguided radiotherapy of prostate cancer has been implemented with a variety of imaging techniques, such as kv radiographs and fluoroscopy, megavoltage (MV) electronic portal imaging, magnetic resonance imaging, and radiofrequency transponders (1, 2). One of the commonly used techniques usually involves daily kvcbct for patient setup verification before each fraction of radiation treatment using a kv X-ray imager attached to the linear accelerator. However, frequent use of kvcbct can potentially add a considerable amount of radiation dose to cancer patients who have already received a significant amount of radiation from MV photon treatments (3 11). Yet, this portion of kvcbct-contributed radiation dose has not been included appropriately in the calculation of total doses to the patients, owing largely to the lack of a dose calculation engine in commercial treatment planning systems for kv beams. Because the testes are one of the most radiosensitive tissues, with significant impairment of spermatogenesis occurring at doses as low as 10 cgy (12), dose deposition to the testes in radiotherapy of prostate cancer (13 18) and rectal cancer (19 22) has been a major concern for many years. Earlier studies suggested that clinically significant hypogonadism could occur at doses as low as 2 4 Gy (15, 22). Although testicular doses were found to be <2 Gy in most situations, recent frequent application of kvcbct in IGRT treatments of prostate cancer has increased the concern for testes sparing Reprint requests to: Jun Deng, Ph.D., Department of Therapeutic Radiology, Yale University, 15 York Street, LL508-Smilow, New Haven, CT Tel: (203) ; Fax: (203) ; jun.deng@yale.edu e39 Conflicts of interest: none. Received Oct 29, 2010, and in revised form Jan 24, Accepted for publication Jan 29, 2011.

2 e40 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 because of enhanced doses from kv beams. In fact, our recent study indicated that kvcbct-contributed testicular dose has increased to 1.2 Gy during a typical IGRT treatment of prostate cancer, up by 330% compared with regular IMRT without kvcbct, yet this portion of dose was neglected in the radiotherapy design and decision making (11). Hence, the goal of this work was to apply a Monte Carlo method to systematically investigate excessive dose to the testes and other organs due to kvcbct for prostate cancer patients undergoing IGRT. Specifically, we investigated On-Board Imager (OBI; Varian Medical Systems, Palo Alto, CA) kvcbct. In this work, an EGS4/BEAM user code, MCSIM, was used for Monte Carlo treatment planning of IMRT plans for prostate cancer treated with five-field 10-MV photon beams. We also calculated three-dimensional dose distributions of a 364 arc from kvcbct scanned with pelvis protocol and a 204 arc scanned with high-quality head protocol using Monte Carlo. The impact of such clinical parameters as scanning mode, kv peak energy, and scanning field on testicular doses was also investigated. Finally, a dosimetric comparison between whole-pelvis irradiation, prostate-only IMRT, and IGRT was conducted. METHODS AND MATERIALS IGRT treatments of prostate cancer Three male adult patients diagnosed with prostate carcinoma and treated with IGRT were selected in this study. The prostate radiotherapy at Yale included a primary irradiation of 66.6 Gy (1.8 Gy 37 fractions) to the planning target volume (PTV) followed by a boost of 12.6 Gy (1.8 Gy 7 fractions) using five-field IMRT and 10-MV photons. The prostate gland and all the critical structures, such as rectum, bladder, femoral heads, and testes, were contoured and/or reviewed by the physicians consistently. Planning target volumes for primary and boost plans were generated on an Eclipse treatment planning system according to the following margin specifications: 0.8 cm all around the prostate gland except for 0.6 cm between rectum prostate interface for primary plan, and 0.6 cm all around for boost plan. No seminal vesicles were included in the PTV volume. Before each fraction, a half-fan kvcbct scan in pelvis protocol was performed for patient setup. For the purpose of this study, both half-fan scan in pelvis protocol and full-fan scan in high-quality head protocol were simulated for comparison. The contours of PTVs and critical structures in kvcbct scans were identical to those defined in IMRT treatment plans. Monte Carlo treatment planning and patient dose calculations The Monte Carlo code EGS4/BEAM was used to simulate the 10- MV photon beams from a Varian Trilogy (Varian Medical Systems) (23, 24). On the basis of our previous study, we have constructed a three-source model to represent 10-MV photon beams emanating from the linear accelerator treatment head for this work: an extended annular source for the target, a planar ring source for the primary collimator, and a planar annular source for the flattening filter (25). Likewise, we have also characterized kv photon beams from the Varian Trilogy OBI system in both half-fan and full-fan modes using EGS4/BEAM code. On the basis of beam characteristics, it was determined that a four-source model could be used to best represent the four most important components of OBI in producing kv photons, that is, the target, the exit window, the blades, and the bowtie filter (11). We developed eight individual four-source models in this study to represent the eight clinically viable configurations (i.e., half-fan pelvis protocol or full-fan high-quality head protocol with 60, 80, 100, or 125 kv peak energy [kvp]). An EGS4/BEAM user code, MCSIM, was used for patient dose calculations with multiple source models as beam input. The dose calculation accuracy of MCSIM has been benchmarked against EGS4/BEAM/DOSXYZ and verified by extensive phantom measurements (26, 27). Specifically, in simulating IMRT treatments, each dynamic multileaf collimeter (MLC) field was simulated by changing particle weighting factors using an intensity distribution reconstructed from the MLC leaf sequence file for that field. When simulating the half-fan kvcbct in a pelvis protocol (variable kv, 80 ma, 12 ms, 364 arc, half-bowtie filter), a series of coplanar fields around the gantry rotation axis was simulated to mimic the 364 arc. Similarly, the full-fan mode with default settings of a high-quality head protocol (variable kv, 80 ma, 25 ms, 204 arc, full-bowtie filter) was also simulated. In MCSIM patient dose calculations, the energy cut-offs for electrons (ECUT) and for photons (PCUT), and the energy thresholds for d-ray production (AE) and for bremsstrahlung production (AP) were set as ECUT = AE = 521 kev and PCUT = AP = 10 kev for both 10-MV and kv photon beams. A statistical uncertainty (1s) of 0.5% and 2% has been achieved for 10-MV and kv beam dose calculations, respectively. To convert Monte Carlo simulations into absolute doses, for 10-MV photons, Monte Carlo simulation was performed to a calibration point in a water phantom at 100-cm source-to-surface distance (SSD) for a 10 cm 10 cm field using MCSIM with three-source model as beam input. The calibration point was at 2.4-cm depth along the central axis where linear accelerator beam output has been tuned to be 1 cgy per monitor unit, per TG-51 protocol (28). The reciprocal of the dose to the calibration point gave a unique conversion factor to this three-source model. For kv photons, absorbed dose was first measured at isocenter of an acrylic ball phantom of 5-cm diameter per TG-61 protocol with an EXRADIN A12 ionization chamber (Standard Imaging, Middleton, WI) for a 125-kV CBCT acquisition in pelvis protocol (29). Monte Carlo simulation was then performed to a chamber volume inside the ball phantom with the same beam setup using a corresponding four-source model. The ratio of the two yielded a conversion factor, which was unique to the clinical setup and beam configuration. We repeated the process for eight different scanning scenarios (i.e., half-fan or full-fan mode with 60, 80, 100, or 125 kvp) and obtained eight unique conversion factors. With all those conversion factors, all the Monte Carlo simulations in patient CT geometry can be converted into absolute doses to water. Dose measurements of MV and kv photon beams To validate Monte Carlo simulations, we have first measured central axis percent depth doses (PDD) and dose profiles for both 10-MV and 125-kV photons with an IBA blue phantom water scanning system (IBA Dosimetry America, Bartlett, TN). Specifically, for 10-MV photons, the PDDs were measured with a CC-13 ionization chamber while the dose profiles were scanned with a photon diode. The field size was 10 cm 10 cm at 100-cm SSD, and the three depths for profile measurements were 2.4, 10, and 20 cm. For 125-kV photons, both PDD and dose profiles were scanned with two CC-13 ionization chambers with one as field detector and the other as reference detector. Default blade settings in OBI were used (i.e., X1 = 8.3 cm, X2 = 24.9 cm, Y1 = Y2 = 11.8 cm for half-fan mode; and X1 = X2 = 15.1 cm and Y1 = Y2 = 11.8 cm for full-fan mode at 100-cm SSD, respectively).

3 Testicular doses during prostate IGRT d J. DENG et al. e41 Furthermore, we measured point doses at the center of a rectilinear solid water phantom (25 cm 25 cm 23 cm) using a Capintec PR-05P mini-chamber (Capintec, Ramsey, NJ) for the five-field 10-MV IMRT plans used for radiotherapy of three prostate cancer patients. For kv photon beams, we also measured absolute doses in the same way with an A12 ionization chamber for the eight scanning scenarios being investigated in this study. kvcbct scanning mode, beam energy, and scanning field To study the impact of CBCT scanning mode, photon beam energy, and scanning field on testicular dose during IGRT treatments of prostate cancer, we performed a series of Monte Carlo simulations on patient CT anatomy with variable clinical parameters while keeping the rest of the parameters intact. Those clinical parameters being investigated included half-fan vs. full-fan mode, peak photon energy of 60, 80, 100, or 125 kv, and scanning field along the superior inferior direction ranging from 10 to 30 cm. Testicular doses in whole-pelvis irradiation, prostate-only IMRT, and IGRT To evaluate testicular doses in three treatment modalities of prostate cancer, Monte Carlo simulations were performed on patient CT anatomy with beam arrangements, MLC leaf sequence patterns, and monitor unit settings identical to the actual radiotherapy treatments. Specifically, whole-pelvis irradiation was done with a fourfield box technique with photon jaws defining 14 cm 15.4 cm field size and MLC blocking four corners. RESULTS Validation of Monte Carlo simulations As shown in Figs. 1a and 1c, the central axis PDD and dose profiles in water at three depths (i.e., 2.4, 10, and 20 cm) of a 10-MV photon beam calculated by Monte Carlo were compared with measurements for a 10 cm 10 cm field at 100-cm SSD. The maximum differences in measured and calculated PDDs were <1.1% from the d max of 2.4 cm to the depth of 25 cm and <2% between the depth of 0.3 cm and d max. The dose profile comparisons indicated a better than 1.1%/1 mm agreement between the two. A similar comparison between Monte Carlo simulations and measurements in water for 125-kV photon beams is also shown in Figs. 1b and 1d. The PDD comparison showed a <2.3% differences from 0.1 cm down to 20 cm, whereas the dose profile comparisons along the x axis at depths of 1, 5, and 10 cm yielded a better than 3%/1 mm agreement between the two. The overall agreement between the measurements and Monte Carlo simulations has verified the accuracy of our Monte Carlo Fig. 1. Comparison of Monte Carlo simulations with the measurements for central axis percent depth doses of (a) 10-MV photons and (b) 125-kV photons at 100-cm source-to-skin distance, and for dose profiles along x axis of (c) 10-MV photons at depths of 2.4, 10, and 20 cm and (d) 125-kV photons at depths of 1, 5, and 10 cm. Error bars representing 1s standard deviation were approximately 0.5% for Monte Carlo simulations and hardly visible in the graphs.

4 e42 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Table 1. Comparison between ionization chamber measurements and Monte Carlo simulations in solid water phantoms for 10-MV five-field IMRT prostate plans and kvcbct scans in half-fan and full-fan modes with 60, 80, 100, and 125 kvp IMRT kvcbct half-fan pelvis protocol kvcbct full-fan high-quality head protocol Parameter Pt 1 Pt 2 Pt 3 60 kv 80 kv 100 kv 125 kv 60 kv 80 kv 100 kv 125 kv Measurements (cgy) Monte Carlo (cgy) (MC-Mea)/Mea (100%) Abbreviations: IMRT = intensity-modulated radiotherapy; kvcbct = kilovoltage cone-beam computed tomography; kvp = kilovoltage peak energy; Pt = patient; MC = Monte Carlo; Mea = measurement. All doses in table are in cgy per fraction for IMRT or per scan for kvcbct. simulations using multiple source models for both 10-MV and 125-kV photon beams. Table 1 lists the absolute dose comparison between Monte Carlo simulations and ionization chamber measurements in solid water phantom for both IMRT plans and kvcbct scans. Overall, the Monte Carlo predicted absorbed doses at isocenter of the phantom were within 1% of the measurements for the three IMRT plans and within 3.1% of the measurements for kvcbct scans, with kvp ranging from 60 to 125 kv in either half-fan or full-fan mode. The excellent agreement (approximately 1%) between the measurements and Monte Carlo simulations for 10-MV IMRT plans verified our Monte Carlo multiple source modeling, MLC leaf sequencing, and absolute dose calculation accuracy. Considering the statistical uncertainty of Monte Carlo simulations (approximately 2%) and measurement uncertainty (approximately 2%) for kv photon beams, the differences of up to 3.1% between the two were clinically acceptable. On the basis of this validation study, we have demonstrated the accuracy of both MVand kv photon beam dose calculations in patient CT anatomy with Monte Carlo multiple source models. Dosimetric comparison of 10-MV IMRT vs. 125-kV CBCT in prostate IGRT The comparison of Monte Carlo calculated dose distributions for a 10-MV five-field IMRT plan and a 125-kV CBCT scan in half-fan mode is shown in Fig. 2, with the corresponding dose volume histograms compared in Fig. 3. Shown in Fig. 2a is a typical IMRT plan: a highly conformal dose distribution around the prostate gland with adjacent critical structures largely spared. For adult men, because the testes were on average 12 cm apart inferior to the prostate, the testes Fig. 2. Comparison of Monte Carlo calculated dose distributions of a five-field 10-MV intensity-modulated radiotherapy treatment (a, b) and a 125-kV half-fan cone-beam computed tomography scan in pelvis protocol (c, d) on a prostate cancer patient. The arrow indicates the 23.6-cm field span along the superior inferior direction with respect to the isocenter.

5 Testicular doses during prostate IGRT d J. DENG et al. e43 8.4% more dose to prostate, rectum, bladder, and femoral heads, respectively, while inducing a 300% increase in testicular dose, as compared with IMRT. Because of enhanced photoelectric effect in bone for kv X-rays compared with MV photons, the kvcbct-contributed doses were most significant at femoral heads and sacrum (approximately 4.5% of prescribed dose). High doses were also observed on anterior and posterior (AP) sides of patient anatomy (approximately 3% of prescribed dose) owing to the 364 arc, with slightly higher doses on left lateral than on right lateral, consistent with earlier measurements (4, 5) and Monte Carlo simulations (6, 8, 10). Impact of kvcbct scanning mode on testicular dose The most important difference between half-fan and fullfan CBCT scans came from the fact that whereas the half-fan mode involved a 364 arc rotating around patient anatomy, the full-fan mode delivered a 204 arc between patient s left posterior and right posterior in patient supine position. Compared in Figs. 3b and 3c are dose volume histograms of various organs due to half-fan and full-fan CBCTs scanned at 125 kvp, respectively, with mean organ doses shown in Fig. 4a and organ dose ratios shown in Fig. 4b. Overall, full-fan CBCT deposited less dose to almost all the organs than half-fan CBCT, ranging from 9% less for prostate to 69% less for testes. The large sparing of testes in full-fan mode was primarily caused by their anterior and superficial location. The only exception was rectum, where full-fan mode delivered 34% more dose owing to its posterior location. On the basis of our results, a compromise has to be made in choosing kvcbct scanning mode when both rectum and testes needed to be spared simultaneously. Fig. 3. Comparison of dose volume histograms of (a) a five-field 10-MV intensity-modulated radiotherapy treatment, (b) a 125-kV half-fan cone-beam computed tomography (CBCT) scan, and (c) a 125-kV full-fan cone-beam computed tomography scan on a prostate cancer patient. essentially lay 8 cm outside of the field and underneath photon jaws. Because photon jaws blocked almost all of the primary photons, the testicular dose from direct transmission and indirect scatter was reduced significantly, as shown in Fig. 2b and Table 2. The 125-kV CBCT, on the other hand, delivered significantly more dose to the testes, as shown in Figs. 2c and 2d, owing to large field span (23.6 cm) along the superior inferior direction defined in default half-fan mode. According to Table 2, the 125-kV half-fan CBCT added 1.7%, 3.2%, 3.2%, and Impact of kvcbct beam energy on testicular dose An early study indicated that kvcbct-contributed doses were essentially linearly dependent on the setting of milliampere-second (mas) (5). In this study, we demonstrated that kvcbct-contributed doses increased dramatically as kvp increased from 60 to 125 kv for all the organs, regardless of scanning mode, as shown in Fig. 5. On the basis of mean absorbed doses listed in Table 2, our further investigation revealed that kvcbct-contributed mean doses followed an exponential relationship with kvp for all the organs. Specifically, testicular doses were least dependent on kvp as compared with doses to other organs in both half-fan and full-fan modes. Impact of kvcbct scanning field on testicular dose The fact that testicular doses depended heavily on field size has been reported in early studies in which MV photon beams were used in radiotherapy of prostate and rectal cancers (13, 20). In this work, we have manually set a variety of locations for Y blades (responsible for CBCT field span along the superior inferior direction) in our Monte Carlo simulations and calculated dose depositions accordingly. As shown in Fig. 6a, increasing the scanning field from 10 to 30 cm caused significant dose deposition to various organs from half-fan kvcbct, ranging from 82% increase for

6 e44 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Table 2. Mean absorbed doses to the organs compared between PO-IMRT and kvcbct scans at half-fan and full-fan modes Organ PO-IMRT (10 MV) kvcbct half-fan pelvis protocol kvcbct full-fan high-quality head protocol 60 kv 80 kv 100 kv 125 kv 60 kv 80 kv 100 kv 125 kv Prostate Rectum Bladder Testes Left femoral head Right femoral head Abbreviations: PO-IMRT = prostate-only intensity-modulated radiotherapy; kvcbct = kilovoltage cone-beam computed tomography. All doses in table are in cgy per fraction for PO-IMRT or per scan for kvcbct. rectum and bladder to 3200% increase for testes. A similar relationship is also observed in Fig. 6b for full-fan kvcbct, for which the dose increase ranged from 73% for rectum to 1600% for testes. Further investigation also indicated that whereas CBCT-contributed doses to other organs increased with CBCT field and gradually reached a plateau at a field size of 30 cm, the testicular doses increased exponentially with field size, regardless of scanning mode. Dosimetric comparison of WPI vs. PO-IMRT vs. IGRT Whole-pelvis irradiation (WPI), prostate-only IMRT (PO- IMRT), and IGRT are compared in Table 3 in terms of dose depositions to various organs. Generally speaking, in comparison with IMRT, half-fan kvcbct added only 1.7%, 3.2%, 3.2%, and 8.4% more doses to prostate, rectum, bladder, and femoral heads, respectively, but it contributed 400% more dose to the testes. On the other hand, WPI delivered much more dose to testes, rectum, bladder, and femoral heads as compared with PO-IMRT and IGRT, indicating excellent normal tissue sparing achieved in IMRT. DISCUSSION According to this study, kvcbct contributed significantly more dose to testes compared with PO-IMRT, whereas doses to the rest of the organs at risk were minimal and clinically insignificant. This was mainly due to two reasons: (1) kvcbct scanned with a large superior inferior field whose inferior border touched the superior portion of testes, whereas PO-IMRT focused high doses to PTV with photon jaws and MLC leaves heavily blocking MV photon beams outside of the field, and (2) testes superficial location made them generally accessible to kv photons with large scattering power. During a regular course of prostate IGRT (79.2 Gy/44 fractions), the kvcbct-contributed testicular dose would be approximately 1.3 Gy, whereas rectum, bladder, and femoral heads would receive 1.7, 1.8, and 2.5 Gy, respectively. Although 1 3 Gy doses were added, the total doses for rectum, bladder, and femoral heads were only 54, 57, and 32 Gy, respectively, still within tolerances. Yet, because hypogonadism can be clinically appreciated at doses of 2 4 Gy, the extra 1.3 Gy from kvcbct would create a much higher risk for testicular malfunction in the long term. The testes are very sensitive to radiation, spermatogenesis being more easily affected than androgen production (12). Testicular doses as low as 10 cgy can cause morphologic and quantitative changes to spermatogonia, whereas a total dose to the testicles exceeding 200 cgy may lead to permanent infertility (12). Various studies have reported that testicular doses in the range of 1.5 to 5.5 Gy can result in a significant increase in serum follicle-stimulating hormone Fig. 4. Comparison of (a) kilovoltage cone-beam computed tomography (kv CBCT)-contributed mean organ doses, and (b) organ dose ratios of full-fan mode to half-fan mode. Minimum, mean, and maximum ratios have been shown for each organ.

7 Testicular doses during prostate IGRT d J. DENG et al. e45 Fig. 5. Kilovoltage cone-beam computed tomography (CBCT)-contributed mean organ doses as a function of X-ray peak energy for (a) half-fan mode in pelvis protocol, and (b) full-fan mode in high-quality head protocol. and luteinizing hormone and a significant decrease in testosterone levels, with more prominent impact on men older than 70 years (15, 16, 19 22). These results strongly suggest that prominent and permanent testicular damage is sustained during radiotherapy of prostate and rectum cancers, frequently severe enough to cause hypogonadism (15, 16, 19 22). Although kvcbct-contributed doses were relatively small compared with the large amount of radiation doses from MV treatments, an attempt should be made to avoid any unwanted doses and keep the organ doses as low as reasonably achievable when applying image guidance clinically. According to this study, it is clear that by taking some simple steps, large amounts of unnecessary dose from kvcbct can be avoided. These steps include (1) minimizing CBCT field span and maximizing distances from organs at risk to field border, (2) reducing kvp if possible, and (3) choosing the appropriate scanning mode to reduce radiation exposure to organs at risk. Our recommendations based on this study of the Varian OBI system are largely applicable to other CBCT systems, such as the Elekta X-ray Volumetric Imager system (Elekta, Stockholm, Sweden), except that the absolute dose depositions to various organs due to the Varian OBI are significantly larger than those from Elekta X-ray Volumetric Imager, usually by a factor of 2 or more (3, 5, 6). In light of recent concerns regarding testicular doses (17, 18), our investigation indicated that WPI via four-field 10-MV photons deposited approximately 1% of prescription dose to testes, as compared with 0.4% in PO-IMRT and 2% in IGRT, which were lower than 3 16% of prescription dose to testes measured on a phantom with MV photons (13, 14, 16, 19, 20). The main reason for the large differences was the inferior field border. As shown in Fig. 6, the testicular doses increased exponentially with CBCT field, ranging from 8 cgy in a course of 79.2 Gy for testes 7 cm outside of field to 250 cgy when testes were 3 cm in the field. By setting the inferior field border at the level of ischial tuberosity in WPI, the testes were more than 5 cm outside of the treatment field and therefore largely spared, which was consistent with the Fig. 6. Kilovoltage cone-beam computed tomography (CBCT)-contributed mean organ doses as a function of CBCT field span for (a) half-fan mode in pelvis protocol, and (b) full-fan mode in high-quality head protocol.

8 e46 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Table 3. Mean absorbed doses to the organs compared between WPI, PO-IMRT, and IGRT (PO-IMRT plus daily kvcbct in half-fan pelvis protocol) Organ WPI (10 MV) PO-IMRT (10 MV) IGRT (PO-IMRT + kvcbct half-fan pelvis protocol) 10 MV + 60 kv 10 MV + 80 kv 10 MV kv 10 MV kv Prostate Rectum Bladder Testes Left femoral head Right femoral head Abbreviations: WPI = whole-pelvis irradiation; PO-IMRT = prostate-only intensity-modulated radiotherapy; IGRT = image-guided radiotherapy; kvcbct = kilovoltage cone-beam computed tomography. All doses in table are in cgy per fraction. observation of Roach et al. (30) in their recent response to an earlier work by King and Kapp (17). Another reason came from the fact that MV photons used in earlier studies produced much more neutrons and secondary particles than 10-MV beams, therefore adding more scatter doses to the testes. CONCLUSIONS In general, kvcbct deposits higher doses on the anterior and posterior sides of patient anatomy, owing to smaller dimension in the anterior posterior direction compared with lateral, and highest at femoral heads and sacrum. Full-fan CBCT usually deposits much less dose to organs (except for rectum) than half-fan mode in prostate patients. Kilovoltage CBCT contributed doses increase exponentially with photon beam energy for both half-fan and full-fan modes. Reducing the CBCT field significantly cuts doses to testes and other organs. For prostate cancer patients, kvcbct likely has a more significant impact on the testes than other critical structures. REFERENCES 1. Jaffray DA, Siewerdsen JH, Wong JW, et al. Flat-panel conebeam computed tomography for image-guided radiation therapy. Int J Radiat Oncol Biol Phys 2002;53: Dawson LA, Sharpe MB. Image-guided radiotherapy: Rationale, benefits, and limitations. Lancet Oncol 2006;7: Islam MK, Purdie TG, Norrlinger BD, et al. Patient dose from kilovoltage cone beam computed tomography imaging in radiation therapy. Med Phys 2006;33: Wen N, Guan H, Hammoud R, et al. Dose delivered from Varian s CBCT to patients receiving IMRT for prostate cancer. Phys Med Biol 2007;52: Song WY, Kamath S, Ozawa S, et al. A dose comparison study between XVI and OBI CBCT systems. Med Phys 2008;35: Chow JC, Leung MK, Islam MK, et al. Evaluation of the effect of patient dose from cone beam computed tomography on prostate IMRT using Monte Carlo simulation. Med Phys 2008;35: Ding GX, Duggan DM, Coffey CW. Characteristics of kilovoltage x-ray beams used for cone-beam computed tomography in radiation therapy. Phys Med Biol 2007;52: Ding GX, Duggan DM, Coffey CW. Accurate patient dosimetry of kilovoltage cone-beam CT in radiation therapy. Med Phys 2008;35: Ding GX, Pawlowski JM, Coffey CW. A correction-based dose calculation algorithm for kilovoltage x rays. Med Phys 2008; 35: Ding GX, Coffey CW. Radiation dose from kilovoltage cone beam computed tomography in an image-guided radiotherapy procedure. Int J Radiat Oncol Biol Phys 2009;73: Deng J, Chen Z, Nath R. Impact of kilo-voltage cone beam computed tomography on image-guided radiotherapy of prostate cancer. IFMBE Proc 2009;25/I: Howell S, Shalet S. Gonadal damage from chemotherapy and radiotherapy. Endocrinol Metab Clin North Am 1998;27: Amies CJ, Mameghan H, Rose A, et al. Testicular doses in definitive radiation therapy for localized prostate cancer. Int J Radiat Oncol Biol Phys 1995;32: Zagars GK, Pollack A. Serum testosterone levels after external beam radiation for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 1997;39: Daniell HW, Clark JC, Pereira SE, et al. Hypogonadism following prostate-bed radiation therapy for prostate carcinoma. Cancer 2001;91: Boehmer D, Badakhshi H, Kuschke W, et al. Testicular dose in prostate cancer radiotherapy: Impact on impairment of fertility and hormonal function. Strahlenther Onkol 2005;181: King CR, Kapp DS. To treat pelvic nodes or not: Could the greater testicular scatter dose from whole pelvic fields confound results of prostate cancer trials? J Clin Oncol 2009;27: King CR, Maxim PG, Hsu A, et al. Incidental testicular irradiation from prostate IMRT: It all adds up. Int J Radiat Oncol Biol Phys 2010;77: Dueland S, Guren MG, Olsen DR, et al. Radiation therapy induced changes in male sex hormone levels in rectal cancer patients. Radiother Oncol 2003;68: Hermann RM, Henkel K, Christiansen H, et al. Testicular dose and hormonal changes after radiotherapy of rectal cancer. Radiother Oncol 2005;75: Mazonakis M, Damilakis J, Varveris H, et al. Radiation dose to testes and risk of infertility from radiotherapy for rectal cancer. Oncol Rep 2006;15: Bruheim K, Svartberg J, Carlsen E, et al. Radiotherapy for rectal cancer is associated with reduced serum testosterone and increased FSH and LH. Int J Radiat Oncol Biol Phys 2008;70:

9 Testicular doses during prostate IGRT d J. DENG et al. e Nelson W, Hirayama H, Rogers DW. The EGS4 code system. Stanford Linear Accelerator Center Report SLAC-265. Stanford, CA: Stanford University; Rogers DW, Faddegon BA, Ding GX, et al. BEAM: A Monte Carlo code to simulate radiotherapy treatment units. Med Phys 1995;22: Deng J, Jiang SB, Kapur A, et al. Photon beam characterization and modeling for Monte Carlo treatment planning. Phys Med Biol 2000;45: Ma CM, Li JS, Pawlicki T, et al. A Monte Carlo dose calculation tool for radiotherapy treatment planning. Phys Med Biol 2002;47: Li JS, Pawlicki T, Deng J, et al. Validation of a Monte Carlo dose calculation tool for radiotherapy treatment planning. Phys Med Biol 2000;45: Almond PR, Biggs PJ, Coursey BM, et al. AAPM s TG-51 protocol for clinical reference dosimetry of high-energy photon and electron beams. Med Phys 1999;26: Ma CM, Coffey CW, DeWerd LA, et al. AAPM protocol for kv x-ray beam dosimetry in radiotherapy and radiobiology. Med Phys 2001;28: Roach M 3rd, Lawton CA, Chen J. To treat pelvic nodes or not? Greater testicular scatter does not explain the results of randomized trials. J Clin Oncol 2010;28:e450 e452.