J. Dolan

Monte Carlo and experimental dosimetry of an brachytherapy seed

We have performed a comprehensive dosimetric characterization of the Oncura™ model 6711 I125 seed using both experimental [LiF thermoluminscent dosimetry (TLD)] and theoretical (Monte Carlo photon transport) methods. In addition to determining the dosimetric parameters of the 6711, this report quantified: (1) the angular dependence of LiF TLD energy response functions for both point and volume detectors in water, poly(methylmethacrylate), and solid water media; and (2) the contribution of underlying geometric uncertainties to the overall uncertainty of Monte Carlo derived dosimetric parameters according to the National Institute of Standards and Technology Report 1297 methodology. The theoretical value for the dose rate constant in water was 0.942cGyU-1h-1±1.76% [combined standard uncertainty (CSU) with coverage factor k=1] and the experimental value was 0.971cGyU-1h-1±6.1%. Agreement between experimental and theoretical radial dose function values was well within the k=1 CSU, while agreement between experimental and theoretical anisotropy function values was within the k=1 CSU only after incorporating the use of polar angle-dependent energy response functions. The angular dependence of the relative energy response was found to have a complex and significant dependence on measurement medium and internal geometry of the source.

SU‐E‐T‐692: Validation of Permanent Interstitial I‐125 Implant Pre‐Plans Against An Experimentally Derived Nomogram Can Potentially Reduce Variability of Day 0 Post‐Implant Dosimetry

PURPOSE To evaluate whether standardization of total pre-plan activity against a nomogram reduces variability of post-implant dosimetry for I-125 permanent interstitial prostate implants (PIPI). METHODS Between January 2007 and December 2011, 843 patients underwent TRUS guided PIPI (MPD 108Gy) in combination with external irradiation. The implants and all contouring for pre-and post-plans were performed by a single physician. Post-implant dosimetry was determined based on CTs taken on the day of implant. Implants were grouped into the following categories based on their post-implant dosimetry: 1.Excellent (90%< D90 <120% and V150>70%); 2.Hot (D90>120% and V150>70%); 3. Cold (D90<80%); and 4.Other. A known-effective activity versus volume nomogram was derived from a fit of the data from the excellent implants. The actual activities used in the hot and cold implant groups were evaluated for deviations from the predictions of the known-effective nomogram. The potential impact of imposing a tolerance window was investigated. RESULTS Of all implants, 347, 81, 37 and 378 were excellent, hot, cold and other, respectively. The activity versus volume data of the excellent implants were well fit with a power law (R^ 2=0.972). The average and standard deviation of the differences between actual implanted activity and the prediction of the known-effective nomogram was 3.8% +/- 4.0 and -0.6% +/- 5.0 for the hot and cold groups, respectively. The implanted activity agreed with the known-effective nomogram prediction within +/-4% for 514 plans. For these plans, 44%, 7.0% and 3.9% were excellent, hot and cold, respectively. Outside this tolerance window, the corresponding values were 36%, 14% and 5.2%. CONCLUSION Imposing a quality assurance requirement that total pre-plan activity agree with the prediction of a known-effective nomogram should improve homogeneity of post-implant dosimetric parameters by increasing the frequency of excellent implants and reducing the frequency of hot implants.

SU‐E‐T‐697: Correlation Between Day 0 Post‐Implant Dosimetry Parameters and Differences in Pre‐and Post‐Implant Prostate Volume‐How Much Can Post‐Plan Quality Be Improved Through Better Assessment of Volume for Permanent Interstitial I‐125 Prostate Implants?

PURPOSE To evaluate the relationship between the difference in pre-and post-implant prostate volume and post-implant dosimetric parameters for pre-planned permanent interstitial prostate implants. METHODS Between January 2007 and December 2011, 964 patients underwent TRUS guided interstitial prostate I-125 brachytherapy either as monotherapy (MPD 144Gy) or in combination with external irradiation (MPD 108Gy). All procedures and contouring for pre-and post-planning was performed by a single physician. Post-implant dosimetric analysis was performed on CTs taken on the day of implant. Differences in prostate volume as determined by TRUS and CT were evaluated by computing a parameter, δvol, defined as the TRUS prostate volume (VTrus) minus the CT prostate volume (VCT) divided by VTrus. Implants were grouped into the following categories based on their post-implant dosimetry: 1.Hot (D90>120% and V150>70%); 2. Cold (D90<80%) and; 3.) Standard. Differences in δvol across groups were evaluated. D90, V100 and V150 were plotted against their corresponding δvol values. RESULTS Overall, the average and standard deviation of δvol, D90, V100 and V150 was 0.0%+/-17.6, 101%+/-12.9, 89.3%+/- 7.0, and 47%+/-12.8, respectively. For the standard, hot, and cold groups, the average and standard deviation of δvol was 0.3%+/-15.5, - 18.0%+/-9.3 and 26.6%+/-24.0, respectively. Plots of the dosimetric parameters versus δvol show wide dispersion, for example R^ 2 = 0.422 for the D90 plot, but they strongly suggest a downward slope, highlighting the fact that cases with a large difference between VTrus and VCT are more likely to have nonstandard dosimetry than cases with a small difference in these volumes. It is noted that there are cases with δvol close to 0 that have nonstandard post-plan dosimetry. CONCLUSION Our work demonstrates that limiting variation in δvol may reduce but not eliminate the occurrence of nonstandard post-plans.

SU‐E‐T‐368: A Comparison of Small MU Sub‐Field Dosimetry for Step‐And‐Shoot IMRT Fields on Varian IX and Truebeam Machines

Purpose: To compare measured and calculated dose distributions for step‐and‐shoot IMRT fields comprised of sub‐fields with small MUs on Varian IX and Truebeam machines. Methods: The Eclipse treatment planning system (v 8.9.9) was used to create step‐and‐shoot IMRT fields comprised of a sequence of rectangular sub‐fields of decreasing size. MU settings of 1, 2 and 50 MU per sub‐field and energies of 15x and 6x were investigated on both an IX (console v7.8.05) and a Truebeam (console v1.5.10.2). Dose rates of 300 MU/min and 600 MU/min were investigated. All fields were delivered with the sub‐fields merged into a single field and with them unmerged as separate fields. The delivered dose distributions were measured and compared to the calculated distributions using the Mapcheck QA system(v 4.01.01). Results: For unmerged fields, the measured and calculated dose distributions showed similar agreement for both machines. For the merged 1 and 2 MU per sub‐field fields on IX, a previously described ‘overshoot’ phenomenon was readily apparent. In this phenomenon, a slight discrepancy in the beam off timing between sub‐fields seems to introduce a systematic error in dose delivery. This phenomenon was not evident for the merged 50 MU per sub‐field field on the IX or for any merged fields on the Truebeam. The magnitude of the overshoot phenomenon on the IX was found to increase with increasing dose rate. Conclusions: The Truebeam appears to be able to deliver low MU step‐and‐ shoot IMRT fields more accurately than the IX. On the Truebeam, neither field merging nor dose rate appear to affect the accuracy of step‐and‐shoot delivery. On the IX, merging fields introduces a small systematic error that may suggest the need for limiting dose rate and minimum sub‐field MU in certain applications.

SU‐FF‐T‐238: Establishing An Efficient EPID‐Based IMRT QA Protocol Using Gamma Analysis Coupled with DVH

Purpose: This study is to establish an efficient EPID‐based IMRT QA protocol using gamma analysis coupled with DVH. Method and Materials: An EPID is configured for 6‐ and 10 MV X‐ray beams for the IMRT portal dosimetry. The EPID system is assessed for dosimetric accuracy using several open fields spanning 3cm × 3cm to 38cm × 10cm with dMLC patterns. To build a QA decision tree, retrospective analysis of over 100 dMLC fields is undertaken to examine correlation between passing rates and field sizes. The fields are divided into 2 categories according to their field sizes. A statistical analysis is done using Statistical Process Control to determine passing criteria. For the failed fields from the gamma analysis, fractional dose contribution to PTV or critical structures is considered for the second criteria. The dose difference between the planned and the acquired images and DVHs from initial and a modified plan excluding the failed field is utilized. Results: From the system accuracy study, the dose profile matches within 2.7% up to 30cm × 10cm and the gamma score varies with field size. In the retrospective analysis, to make 2% pass rate, gamma score of 0.992 and 0.944 are chosen for the small, rm<18.3cm, and large field groups. For 10MV beam, 7% pass rate is chosen with gamma scores, 0.969 and 0.872 for rm<15cm and the larger field group, respectively. For failed fields, calculated dose difference is less than 5%, which is the second criteria. A field which fails both criteria needs IMRT re‐optimization. Conclusion: The gamma thresholds chosen are based on our statistical findings. The proposed IMRT QA decision tree utilizes gamma scores and overall dose distribution of the individual fields to plan. This is a useful tool to standardize the patient‐specific QA.