Steve Axelrod

Calculated and measured brachytherapy dosimetry parameters in water for the Xoft Axxent X‐Ray Source: An electronic brachytherapy sourcea)

A new x-ray source, the model S700 Axxent™ X-Ray Source (Source), has been developed by Xoft Inc. for electronic brachytherapy. Unlike brachytherapy sources containing radionuclides, this Source may be turned on and off at will and may be operated at variable currents and voltages to change the dose rate and penetration properties. The in-water dosimetry parameters for this electronic brachytherapy source have been determined from measurements and calculations at 40, 45, and 50kV settings. Monte Carlo simulations of radiation transport utilized the MCNP5 code and the EPDL97-based mcplib04 cross-section library. Inter-tube consistency was assessed for 20 different Sources, measured with a PTW 34013 ionization chamber. As the Source is intended to be used for a maximum of ten treatment fractions, tube stability was also assessed. Photon spectra were measured using a high-purity germanium (HPGe) detector, and calculated using MCNP. Parameters used in the two-dimensional (2D) brachytherapy dosimetry formalism were determined. While the Source was characterized as a point due to the small anode size, <1mm, use of the one-dimensional (1D) brachytherapy dosimetry formalism is not recommended due to polar anisotropy. Consequently, 1D brachytherapy dosimetry parameters were not sought. Calculated point-source model radial dose functions at gP(5) were 0.20, 0.24, and 0.29 for the 40, 45, and 50kV voltage settings, respectively. For 1<r<7cm, measured point-source model radial dose functions were typically within 4% of calculated results. Calculated values for F(r,θ) for all operating voltages were within 15% of unity along the distal end (θ=0°), and ranged from F(1cm,160°)=0.2 to F(15cm,175°)=0.4 towards the catheter proximal end. For all three operating voltages using the PTW chamber, measured dependence of output as a function of azimuthal angle, ψ, was typically on average ±3% for 0°⩽ψ⩽360°. Excluding an energy response function, measurements of normalized photon energy spectra were made for three operating voltages, and were typically within 2% agreement with the normalized Monte Carlo calculated spectra. In general, the model S700 Source exhibited depth dose behavior similar to low-energy photon-emitting low dose rate sources I125 and Pd103, yet with capability for variable and much higher dose rates and subsequently adjustable penetration capabilities. This paper presents the calculated and measured in-water brachytherapy dosimetry parameters for the model S700 Source at the aforementioned three operating voltages.

SU-FF-T-58: Dosimetric Study of a New Surface Applicator for the Xoft Axxent System

Purpose: To determine the dosimetricproperties of a new surface applicator designed for the Xoft 50 kVp x‐ray system Method and Materials: A 35mm conical stainless steel applicator with Al flattening filter interfaces to the Xoft 50 kVp x‐ray source to provide therapeutic radiation to surfaces such as skin. Measurements of the dose profile at the surface and at depths to 20 mm were performed using a PTW ionization chamber in a water phantom. The chamber was controlled by a stepper motor and linear stage, and data was read into a controlling computer. In addition, film data was taken both parallel and perpendicular to the surface.Results: Nominal dose profiles and depth dose characteristics were determined from an average of ten sources. Dose profiles across at least 80% of the applicator width were flat to within ± 10% for all sources, and to within 5% on average. Depth dose characteristics are very similar to those published for HDR based surface applicators, with the percent depth dose at 5 and 10 mm at 58% and 36%. Such data can be used for treatment planning purposes. The variation seen among sources determines an error band representative of what will be experienced in clinical use. Conclusion: The FDA approved surface applicator delivers a dose profile that is flat to within ± 10% over a span of 80% of the defined width of the device. It can be used for treatment of skin lesions in lightly shielded rooms due to the low energies employed. There are also intriguing possibilities for use intra‐operatively, since the system can be used in an operating room. Conflict of Interest: Research supported by Xoft, Inc.

SU‐FF‐T‐380: Radiological Dependence of Electronic Brachytherapy Simulation On Input Parameters

Purpose: In comparison to 125I or 192Ir, characterization of dose rate distributions from electronic brachytherapy is subject to the additional challenge of unforeseen photon energy spectra. Towards simulating photon energy spectra and resultant dose rate distribution, Monte Carlo investigators first generate electrons which bombard the x‐ray tube anode and subsequently create photons via bremsstrahlung. Modeling techniques for this endeavor are largely unexplored. Therefore, sensitivities of spectra and dose rate distributions were assessed through varying modeling parameters for the Xoft Axxent x‐ray source. Materials & Methods: MCNP5 was used to simulate photonspectra and dose rate distributions, with comparisons to experimental measurements (PTW model 34013 chamber in liquid water) for 1<r⩽7 cm and 0°⩽θ⩽150° with simulations covering 0.3 ⩽r⩽ 15 cm and all available angles. The following source modeling parameters were evaluated for impact on in‐water spectra and dose:electron beam radius (R), electron beam annularity (R′) like a doughnut, and anode film thickness (t). Since simulations of electron:photon transport are inefficient in comparison to Monte Carlo modeling of radionuclides, MCNP variance reduction techniques such as cell importances (IMP), electron cutoff energies (PHYS:E), high‐energy biasing of bremsstrahlung spectrum (BBREM), and bremsstrahlung photon multiplicity (BNUM) were assessed. Results: Due to the complex anode shape, F(r,θ) was highly‐dependent on R, varying a factor of 2 when changing R from 0 to 0.084 cm. This effect was more pronounced when varying R′ due to less radial volume averaging. Through comparison with experimental measurements, the optimal electron beam shape had the largest spot size which could fit within the anode and no annularity; it was a uniform pencil beam. Altering MCNP variance reduction techniques did not significantly alter results, but greatly hastened simulation efficiency. Conflict of Interest: Research was sponsored in part by Xoft, Inc.

SU‐GG‐T‐36: Film Based Treatment Plan Validation for a New Vaginal Applicator Using the Xoft Axxent™ 50 KVp Miniature X‐Ray Source

Purpose: Compare delivered to planned dose for the Xoft Axxent™ vaginal applicator and 50 kVp x‐ray source using radiochromic film. Method and Materials: A 25mm diameter vaginal applicator (FDA clearance pending) was used to deliver a simulated treatment in a water phantom. The treatment was planned with Varian BrachyVision™, using the Xoft 50 kVp source TG‐43 parameters. The prescription dose was 7 Gy at 5mm from the applicator surface. The applicator and a 5″ square of GAFChromic EBT film were held in a Solid Water™ frame in a water phantom. The film plane was parallel to the long axis. The exposed film was scanned and processed to create a calibrated dose profile. The BrachyVision isodose‐line plot was transformed into an image with identical size and pixel density to the film then combined with the film image to create a new image with dose exposure values only along the planned isodose contours. These contours were analyzed to determine the variation in actual delivered dose along them. Results: Visual comparison of isodose contours and film image showed qualitatively good agreement of the delivered treatment with the plan. Further image processing quantified the agreement. An ad hoc film calibration was employed to estimate dose values along planned isodose contours, with emphasis on the prescription dose of 7 Gy. Thus absolute dose values averaged along a given contour were only approximately correct but the more germane variation of dose along each contour was found to be less than 8% (2 sigma) for dose contours from 1.75 to 8.75 Gy. Conclusion: Dose measured by film exposure in a plane parallel to the applicator axis was found to be constant along plan isodose contours with SD less than 8% (2 sigma). Conflict of Interest: Research supported by Xoft, Inc.

SU‐FF‐T‐311: Calibration of Lot 47207‐01I GAFCHROMIC EBT Film Using the Xoft Axxent X‐Ray Source

Purpose: To determine the absolute dose delivered for a given exposed optical density of GAFCHROMIC® EBT film over a range of 0.25 to 14 Gy and over a range of distances from 1cm to 4cm in water for a Xoft Axxent® Model S700 X‐ray Source. Method and Materials: GAFCHROMIC film is used for dose delivery validation for the Xoft x‐ray source and associated applicators used to deliver prescription dose distributions. Custom Gammex RMI 457 Solid Water™ film fixtures were designed to locate both a PTW34013 ionization chamber and a film coupon at precise distances from the x‐ray source. The fixtures complete integral rotations around the x‐ray source during exposures to minimize azimuthal effects. Three sets of 12 exposures were performed for the dose range at each distance. Film coupons were scanned with an Epson® Expression® 10000XL scanner and the maximum pixel value was determined for each coupon corresponding to the calibrated ionization chambermeasured dose. Results: The measured maximum pixel values for each data set were fit to a 5th order polynomial.Calibration coefficients were determined at different distances from the x‐ray source to investigate spectral differences. All fits were well behaved with residuals within ±4%. Differences between the fitted dose as a function of distance were small, indicating little sensitivity of the film to the range of source spectra. This calibration was applied to Xoft Axxent Vaginal Applicator validation study films, producing good agreement with the isodose contours predicted by BrachyVision™ treatment planningsoftware.Conclusion: A precise calibration of GAFCHROMIC EBT film was performed for the Xoft x‐ray source using a PTW 34013 ionization chamber. It showed no significant difference in the measured optical density for a given measured dose over a distance range of 1cm to 4cm. Conflict of Interest: Research sponsored by Xoft, Inc.

Dosimetric measurements of a new electronic brachytherapy surface applicator

Poster presented at the Annual American Brachytherapy Society Meeting, May 31-June 2, 2009, Toronto, Canada ION CHAMBER RESULTS PURPOSE The development of surface applicators for use with the Axxent® eBx system included an aluminum flattening filter, to create as uniform a dose distribution as possible. The filter also served to harden the beam to provide a depth dose relationship which is similar to that obtained with HDR systems. Height of the cone(s) were selected to provide dose delivery rates similar to HDR based systems. Owing to the low penetration of 50 kVp radiation in higher Z materials, a wall thickness of 1 mm stainless steel in the cone provides full shielding. This results in lower weight compared to similar units designed for use with seeds such as Ir-192. The channel is compatible with many commercially available fixation systems. Dosimetric measurements of these applicators are key to providing guidance for planning treatments. In particular, flattening filters have been incorporated to reduce the variation of dose across the treated area, as occurs with traditional Leipzig style applicators. The measurements reported here show the degree to which flatness or uniformity has been achieved, the variation in dose with depth, and the absolute dose rates in Gray per minute. Measurements of dose rate, dose profiles and percent depth dose (PDD) were made using two techniques, ion chamber and radiochromic film based, for the 35 mm applicator. Laboratory measurements were made using standard Xoft Axxent® sources running at the nominal 50 kVp, 300 μA operating point used in most indications of the system. Data was taken with a 35 mm diameter applicator, the first of four applicator sizes: 10, 20, 35 and 50 mm. Film measurements (type EBT) were made in a water phantom at several distances from the surface, in an orientation parallel to the surface. Film was also exposed in the perpendicular orientation, along the central axis of the applicator. Ion Chamber: Absolute dose rate was measured in a water phantom using a PTW 34013 ion chamber calibrated to dose in water. The ion chamber was encased on a Solid WaterTM jacket to make it watertight. It was mounted on a computer-controlled linear stage, allowing transverse scanning for measurement of dose profiles. The applicator was mounted on a linear stage as well, to set the distance from the ion chamber (Figure 3). Measurements were made at distances of 2, 5, 10 and 15 mm from the face of the applicator to the ion chamber. The applicator comes with disposable sterile end caps to ensure proper positioning of the treated tissue. To prevent water from entering the cone volume, the end cap window was watertight bonded to the applicator cone. This apparatus thus provides dose profiles, absolute dose and depth dose information. Radiochromic film Radiochromic film was used to supplement the ion chamber measurements and to emulate what will likely be used in clinical settings for routine QA. Film provides high spatial resolution but must be carefully calibrated to the radiation quality in use. Such calibrations were performed at several distances from the source in water to investigate the effect of a varying spectrum. No dependence on distance was found over the range from 1 to 4 cm. Accuracy of the calibration fits in terms of residual error was below 2% (2 σ) above 0.5 Gy. METHODS A new set of cone-style applicators has been developed for use in treating surfaces in conjunction with the Axxent® 50 kVp X-ray system. The applicator described here has a diameter of 35 mm and is one of a set that spans a range from 10 to 50 mm. The cone walls are approximately 1 mm thick stainless steel, which provides adequate shielding for the 50 kVp radiation while keeping overall weight low. Ion chamber data show good flatness at all depths, with slow changes in the overall shape with depth, attributed to geometry and the hardening effect of the flattening filter. Scans made with the ion chamber provide reliable depthdose and absolute dose rate data. A fit to the dose rate results at each depth allows prediction of percent depth dose values at various distances from the surface. Absolute dose rate, normalized to the nominal Axxent® 50 kVp source strength of 110,000 U, is approximately 1.45 Gy/min at the surface. At 5mm depth the dose rate is over 60% of that at the surface. SUMMARY SURFACE APPLICATOR Funding provided by Purpose. Xoft, Inc., is developing a set of surface applicators for use in treating skin lesions and other indications where a superficial dose is appropriate. The Xoft Axxent system provides hardened 50 kVp x-ray radiation at high dose rates and can be used as an alternative to HDR sources such as Ir-192, with the advantage that the treatment room can be lightly shielded and staff can be in the room with the patient during treatment. Dosimetric measurements of these applicators are key to providing guidance for planning treatments. In particular, flattening filters have been incorporated to reduce the variation of dose across the treated area, as occurs with traditional Leipzig style applicators. The measurements reported here show the degree to which flatness or uniformity has been achieved, the variation in dose with depth, and the absolute dose rates in Gray per minute. Materials and Methods. Laboratory measurements were made using standard Xoft Axxent sources running at the nominal 50 kVp, 300 μA operating point used in most indications of the system. Data was taken with a 35 mm diameter applicator, the first of four anticipated applicator sizes: 10, 20, 35 and 50 mm. Film measurements (type EBT) were made in a water phantom at several distances from the surface in an orientation parallel to the surface. Film was also exposed in the perpendicular orientation, along the central axis of the applicator. Film provides excellent spatial information, but at these energies, with the spectrum hardening with depth, absolute dose calibration is challenging. Therefore the film measurements were augmented with miniature ionization chamber (PTW type 34013) readings. These were done in a water phantom with a stepper motor controlling the position of the chamber. Results. Ion chamber data show good flatness at all depths, with slow changes in the overall shape with depth, attributed to geometry and the hardening effect of the flattening filter. Scans made with the ion chamber at distances of 2, 5, 10 and 15 mm from the surface provide reliable depth-dose and absolute dose rate data. Conclusions. Dose rate at the surface was approximately 1.45 Gy/min with a nominal strength source. A fit to the dose rate results at each depth allows prediction of percent depth dose values at various distances from the surface: 62%, 38% and 24% at 5, 10 and 15 mm depth, respectively. Flatness or uniformity of the dose across the treatment area can be characterized in several ways. In one manner, the standard deviation of dose within the 80% of the 35 mm diameter (± 14 mm) measured 1.0, 1.9, 3.0 and 3.1% at depths of 2, 5, 10 and 15 mm, respectively. Xoft has developed a set of surface applicators for use with the Axxent® electronic brachytherapy (eBx) system. The applicators are conically shaped, with a linear channel for introduction of the source catheter, and are made of stainless steel (Figure 1). The set consists of 10, 20, 35 and 50 mm diameter versions. These surface applicators were designed for use in treating superficial sites on the head and neck as well as the extremities, especially hands and legs. Superficial radiation therapy is often used with older patients. 10 million actinic keratoses and 1 million basal cell carcinoma cases occur each year. Radiation oncologists currently see a small percentage of these patients secondary to patterns of referral, inconvenience of treatment, and inefficiency of equipment utilization The Axxent® eBx system allows lesions up to 5-10 mm deep to be treated without risk of excessive skin dose and can cover small areas up to nearly 5 cm diameter. The FDA approved Xoft’s 510(k) application in February of 2009. Figure 1. Surface Applicator Data with a scanning PTW 34103 ion chamber was taken for 10 sources, using a single applicator. Data was also taken with a single source and 3 different applicators, and no significant differences were observed. These measurements were made as part of a formal Design Verification protocol. The average profiles of the 10 sources at distances of 2, 5, 10 and 15 mm in water are shown in Figure 4. Vertical lines delineate the 80% width boundaries (at ± 14 mm). Data was taken at 1 mm intervals. Note that the ion chamber has an active diameter of 3 mm, so sharp edges will be convolved with the response function and thus smeared. Table 1 shows maximum and minimum values in the central 80% region for the average profile, in percentage of the average dose rate. Also shown is the standard deviation (σ) across the profile, representing the RMS deviation from flat response. Table 1 Distance Max Min σ 2 101.3% 98.1% 1.0% 5 102.3% 97.0% 1.9% 10 102.9% 93.7% 3.0% 15 102.7% 92.9% 3.1% Individual source runs naturally show greater variation than the average profile, attributable to small differences in source spatial characteristics. Figure 5 shows individual profiles at 2 mm distance for the 10 sources, normalized to 1.0 over the central 80% region. Comparable amounts of variation are seen at distances of 5, 10 and 15 mm. Figure 3. Water Tank Apparatus. The figure shows a water tank with a PTW model 34013 ion chamber attached to a computer-controlled, precision stepper motor. Stage and motor assembly are at top. The ion chamber is encased in solid water (brown in the figure) and scans from left to right. The applicator under test enters from the rear; its depth is controlled by a micrometer stage. 2 mm 5 mm 10 mm 15 mm Figure 4. Average Dose Rate vs. Position Position, mm N o rm a li z e d r e a d in g s Source 1 2 3 4 5 6 7 8

SU‐GG‐T‐549: Dosimetry of An X‐Ray Endocavitary Proctoscope Adapted for Use with the Axxent® Electronic Brachytherapy System

Purpose: To analyze the dose rate and uniformity at the aperture of an Electro‐Surgical Instrument (ESI). Proctoscope when used with an Axxent® Model S700 X‐ray Source. Method and Materials: The ESI proctoscope was originally designed for use with a Philips RT‐50 x‐ray contact therapy unit. Because these units are no longer supported, an alternative x‐ray source is being sought which allows for duplication of the Papillon technique for treating anal‐rectal lesions. A source holder was fabricated to position the Axxent model S700 X‐ray Source collinearly on axis within a lead‐lined proctoscope possessing a 24 mm inner diameter aperture. This mechanism allows the distance from the proctoscope aperture to the source to be adjusted from 1 to 10 cm and allows a filter to be placed at the distal end of the source to harden the beam and/or flatten the dose distribution. The dose rate was measured at the proctoscope aperture using a PTW 34013 Soft X‐ray Chamber set into the surface of a Gammex RMI 457 solid water phantom. GAFChromic EBT radiochromic film was used for profile measurements. Results: For operation at 50 kVp and 0.30 mA beam current with a 0.5 mm thick Al filter and source‐to‐aperture distance of 3.4 cm, the dose rate was 1.1 Gy per minute. Profile evaluation indicated a dose uniformity of 5%, 8% and 15% across 80%, 93% and 98% of the aperture diameter, respectively, without use of a flattening filter. Conclusion: Measurements of dose rate and uniformity at the aperture of a 24 mm diameter ESI proctoscope indicate that the Axxent S700 x‐ray source may be a suitable alternative to the Philips RT‐50 Endocavitary Unit for treatment of anal‐rectal lesions. Conflict of interest: Research sponsored by Xoft, Inc.

SU‐FF‐T‐383: Stability of the Xoft Axxent® X‐Ray Source During Simulated APBI

Purpose: To evaluate the x‐ray output stability of the Xoft Axxent® Electronic Brachytherapy System while delivering fractionated doses to a phantom. Method and Materials: A balloon applicator was inflated in a 4.4 cm diameter spherical simulated lumpectomy cavity within an acrylic “breast” on a supine female full‐body phantom. Radiation treatments for 5 “patients” were delivered in 10 fractions BID for 5 days using one x‐ray source per patient. Per the usual procedure, the air kerma strength for each source was measured prior to each fraction with a calibrated well chamber then dose delivery time was automatically adjusted to account for this source strength. During each fraction, exposure rate was monitored using a calibrated Victoreen Model 451B ion chamber survey meter positioned approximately 10 cm below the phantom. Readings were downloaded at 1 second intervals to a spreadsheet. These data were then analyzed for stability and reproducibility. Results: Fifty fractions were delivered during the five days of simulated treatment. Average treatment time for each fraction was 11 minutes so each source had cumulative operating time of approximately 120 minutes including turn‐on and calibration time. Exposure rate measurements increased with decreased source‐to‐meter distance and with less intervening absorbing material. Exposure rates ranged from 0.1 to 0.55 R/h. The standard deviations from average exposure rates varied from 0.5% to 2.9% with an average over 50 fractions of 0.9%. Conclusions: The Axxent® Electronic Brachytherapy System performed well in five simulated APBI treatments. X‐ray source output and system stability were demonstrated to be within 3% for all treatment fractions. Research sponsored by Xoft, Inc.

SU-DD-A1-05: Turn-On Dose and Transit Time Adjustments in Treatment Planning for the Axxent® Electronic Brachytherapy System

Purpose: To analyze the dosimetric impact of x‐ray source turn‐on time and inter‐dwell position transit times for application of the Axxent® Electronic Brachytherapy System to APBI. Materials and Methods: At the first dwell position, the treatment timer starts after the source has ramped‐up to full operating voltage and beam current (50 kV, 300 μA), so planned dose‐delivery time does not account for a small “turn‐on” dose. Radial dose functions were calculated with MCNP5 for operating voltages from 20 to 50 kVp. Turn‐on dose was estimated by temporally averaging these distributions using the voltage and beam current ramp profiles. For subsequent dwell positions, the timer starts when the source begins moving to the next dwell position so elapsed time includes the transit time. (The source remains on during the time between dwell positions, typically 0.7 seconds for a 0.5 cm step). Dose contribution during transit was estimated using Varian BrachyVision™ by subtracting the transit time from the second and subsequent dwell positions, then adding extra dwell positions at midpoints between original positions with times equal to the transit time. Results: The composite turn‐on dose profile from Monte Carlo results was equivalent to 2 seconds of additional time at the first dwell position with source operation at 50 kVp. This corresponds to < 0.5% of a typical treatment time. Whether or not transit time is accounted for, the planned doses at prescription points 1 cm outside of a typical balloon agree to within an average of 0.1% with a standard deviation of 0.2%. Conclusions: Turn‐on dose may be approximated in treatment planning by adding 2 seconds to the first dwell time. Dose during source transit may be ignored when using a balloon applicator for APBI. Research sponsored by Xoft, Inc.

SU-FF-T-110: Benchmarking MCNP Low-Energy Bremsstrahlung Modeling for Electronic Brachytherapy Simulations

Purpose: Electronic brachytherapy (eBx) sources have been used clinically for over a decade; however, dosimetric characterization methods using measurements or calculations are not well‐established. Monte Carlo methods for simulating electron transport, and subsequently photon production, have not been benchmarked to the same degree as for photon‐emitting HDR 192Ir or LDR 125I brachytherapysources.Materials & Methods: Towards better understanding the capabilities of MCNP5 to simulate radiationtransport for the Xoft Axxent eBx source, this study presents a comparison of calculated MCNP5 results obtained using coupled electron:photon transport with measured bremsstrahlung spectra from the literature. Given the electron energy and target material, MCNP5 bremsstrahlung modeling accounts for photon energy, angle, and probability based on the cross‐sections and angular distributions from NIST (Seltzer and Berger, 1985). The Axxent eBx source currently operates at 50 kV with electrons bombarding a ∼ 1 μm thick high/low Z target. Pertinent high/low Z comparisons for thin targets, defined as materials thin enough to produce negligible electron absorption in the target, were available from Motz and Placious (1958) using 50 kV on 5 nm Au and 63 nm Al, from Cosslett and Dyson (1957) using 10 kV on 25 nm Au, and from Doffin and Kuhlenkampff (1957) using 34 kV on 25 nm Al. Results: Comparisons of calculations and experimental data indicate that bremsstrahlung angular peak, relativistically shifted forward, agreed within a few degrees with measurements in the literature. However, the overall simulated distribution exhibited angularly invariant regions in the forward direction, attributed to MCNP low‐energy physics simplifications of the NIST dataset. Given that the brachytherapy target is ∼ 50 times thicker, with resultant smearing of the energy/angular distributions, the practical impact of this effect is under investigation, and complementary EGSnrc simulations are in progress. Conflict of Interest: This research was sponsored in part by Xoft, Inc.