M.S.U. Siddiqui

TH‐A‐137‐10: Comprehensive Investigation of Dose Calculation for Lung Stereotactic Ablative Radiation (SABR): Effects of Tumor Size and Location and Clinical Recommendations

PURPOSE To determine the influence of tumor size and location on dose calculation accuracy of 6 different algorithms in treatment of lung cancer patients using SABR. METHODS The study consisted of 135 patients with non-small-cell lung cancer previously treated using SABR treatment plans created using a 1-D pencil beam algorithm (1-D equivalent-path-length (1D-EPL) in iPlan). The dose regimen was 12Gy in 4 fractions. Dose was recomputed using six clinically commissioned dose calculation algorithms; 3-D pencil beam (3D equivalent-path-length (3D-EPL) in Eclipse), anisotropic analytical algorithm (convolution/superposition type, AAA in Eclipse), collapsed cone convolution (convolution/superposition type CCC in Pinnacle), AcurosXB (Eclipse), and Monte Carlo (MC iPlan). Location of the lung tumors was categorized as follows: island-type peripheral tumors surrounded by lung tissues (lung-island; N=39), tumors attached to the chest-wall (lung-wall; N=44), and tumors located in the central area (lung-central; N=52). Average irradiated field size (FS) was subdivided into three groups, between 3-5, 5-7, and 7-10 cm. RESULTS For lung-island tumors, D95 values relative to the 1D-EPL (with which the patients were treated) were 81.6±4.4%, 81.4±5.8%, 78.4±6.6%, and 81.4±5.8% for AAA, CCC, AcurosXB, and MC algorithms respectively. For lung-wall tumors, dose values for AAA, CCC, AcurosXB, and MC were 86.8±4.9%, 87.4±5.6%, 84.7±6.6%, and 86.9±5.7%, while respective values for lung-central tumors dose values were 86.1±5.9%, 86.5±6.6%, 85.0±7.0%, and 86.7±6.2%. Considering the effects of both tumor size and location, the lowest D95 values (relative to 1D-EPL) were observed for the smallest lung-island tumors (3-5 cm): 80.2±4.3%, 80.0±6.0%, 76.6±6.9%, and 79.7±5.9% for AAA, CCC, AcurosXB, and MC algorithms, respectively. The smallest dose difference was observed for the largest tumors (7-10 cm) for lung-central tumors. CONCLUSION For small lung tumors located peripherally, treated using SABR, EPL-based algorithms must be used with extreme caution. Under these circumstances even convolution/superposition and MC-based algorithms show larger variation in PTV dose.

WE‐E‐141‐09: FMEA and Fault Tree Ananlysis Applied to Vendor Customer Technical Bulletins

PURPOSE Over the past 3 years our institution has received 44 Customer Technical Bulletins (CTBs) from vendors that supply us with hardware of software used for patient care. We have ranked the failure modes presented to our institution in these CTBs using Failure Mode and Effects Analysis (FMEA) tools as described in the AAPM report TG-100. METHODS FMEA applied to the failure modes comes up with a Risk Probability Number based on the severity, probability of occurrence and probability it will go undetected. In addition to FMEA, the adverse effects on our patient population are examined by looking at the number of people affected. For example, CTBs applying to dose calculations would affect 100% of our patients while CTBs applying to IMRT calculations will affect approximately 50% or our patient population. RESULTS Of the 44 CTBs received from vendors we identified nineteen with Risk Priority Numbers (RPN) greater than 200 and four of those had RPN greater than 300. Of the four with RPN greater than 300, two dealt with incorrect dose calculations due to incorrect commissioning of the treatment planning system, one with the CT datasets getting flipped and the last with isocenter being transferred incorrectly. Each of these failure pathways would have resulted in a systematic error that would have affected a large population of patients. For the failure modes with RPN greater than 200 the average RPN was reduced from 275 to 70 after modifying policy, checklists and improving clinical flow. CONCLUSION Analysis of vendor CTBs and review by a departments QA committee is an essential part of any QA program. By applying FMEA and analyzing the fault trees discussed in these bulletins, we were able to reduce the RPN from an average of 275 to 70.

MO‐D‐105‐07: Results of Applying FMEA and Fault Tree Analysis to the Online Incident Reporting Database

PURPOSE As part of the continuous quality of care improvement, an internal online database for processing reported incidents was established. In 3 years, 710 incidents were reported by 5 clinics. We have ranked the fault trees of the reported incidents using the AAPM report TG-100 Failure Mode and Effects Analysis (FMEA) tools Methods: Risk Probability Number (RPN) generated as a Result of applying FMEA is based on the severity, probability of occurrence and probability of going undetected. The reports were sorted in two categories. Potentially affecting dose delivery, e.g. incorrect setup instructions; and deviations from an established workflow defined by policies and procedures (P&P), e.g., incorrect naming of the fields. In addition to FMEA, the impact of new, as well as periodic reviews of P&P by staff members is assessed Results: Of 710 reports 676 were analyzed, 374 were variation in the workflow not directly affecting quality of care, 302 were potentially affecting dose delivery. 19 of 302 had dosimetric impact; however, due to low occurrence only 4 instances, related to bolus placement, reached the RPN above 200. Review of current P&P reduced the RPN from 270 to 9. Periodic review, introduction of the new or revising the existing P&P had a dual effect: drop in dose-affecting incidents and increased reporting of process deviations Conclusion: Analysis of reported incidents and review by the departmental QA committee is an essential part of any QA program. By defining the fault trees and applying FMEA to the reported incidents, we were able to reduce the RPN from an average of 150 for dose related incidents to 9, and for process variations from 295 to 28 on average. Event-triggered revising of P&Ps and periodic review with staff of the existing P&P is an effective tool in incident reduction.

TH-C-141-11: Evaluation of MR Images as the Planning and Reference Dataset for Daily CBCT-Based IGRT of the Prostate

PURPOSE An important question rarely discussed in the CT vs. MRI simulation debate is whether MRI reference images provide an adequate image-set to use with daily localization such as cone-beam CT (CBCT). This study compares clinical couch shifts based on daily CBCT images to shifts measured from MR images as the reference dataset for prostate IMRT treatment. METHODS Eight patients undergoing a pilot study had MR imaging along with CT simulation with the intent of evaluating a MR-simulation process. Patients had T1, T2 and bTFE (balanced Turbo Field Echo) sequences. The remainder of the treatment planning process continued using traditional procedures using CT. Therapists used only the CT scan as a reference for localization. Retrospectively, an observer measured shifts between daily CBCT images and MR reference images. RESULTS The differences in shift positions for the cohort between therapists and the observer are -0.16cm ± 0.25cm (AP), 0.04cm ± 0.19cm (SI), and - 0.01cm ± 0.14cm (LR). The mean group error for the therapists and the observer were less than 2 mm in all directions. Based on these shifts, the calculated margins for the therapists would be 0.87cm (AP), 0.65cm (SI), and 0.71cm (LR) and for the observer would be 1.1cm (AP), 0.66cm (SI), and 0.70cm (LR). For SI and LR directions both sets of margins are very close to one another. An outlier impacted the AP margin difference by 2.3mm and should be investigated further. This initial analysis suggests that each modality can be considered clinically sufficient for daily localization. CONCLUSION The results of this study suggest that MR reference image-sets can be used for daily image-guided localization of prostate cancers with at least the same accuracy as current methods. MR simulation provides substantial soft-tissue contrast and can improve tissue targeting in radiation oncology, as a Result further investigation is warranted.

SU‐E‐T‐651: Radiobiological Effect of Target Volume in SBRT of Lung Tumor: Comparison of Treatment Planning Algorithms Between Pencil Beam Algorithm and Monte Carlo Method

Purpose: To use equivalent uniform dose (EUD) and tumorcontrol probability (TCP) to retrospectively analyze the radiobiological effect of target volumes in patients with NSCLC planned and treated with Stereotactic Body Radiotherapy(SBRT). Methods: Eighty‐three stage I–IIlungcancer patients with 86 lesions treated with SBRT were retrospectively analyzed. For each patient, a Pencil Beam (PB) algorithm‐based treatment plan with a dose regimen of 12 Gy/fraction in 4 fractions was generated. To overcome the known uncertainties of conventional PB algorithm in lungtissue,Monte Carlo(MC) treatment plans were also created in the iPlan (BrainLab) system using the same monitor units derived from the PB‐based plan. Niemierkos EUD and TCP (Poisson model) were computed using different surviving fraction (SF) parameters for each dose calculation algorithm. The radiobiological effects of target volume were analyzed by correlating EUD and TCP with PTV volumes. Results: Mean PTV volume was 39.31 +/− 28.96 cc. The mean PB EUDs were 50.61, 50.60, 50.57 and 50.55 Gy for SF parameter values of 0.36, 0.34, 0.3 and 0.28, compared with MC EUDs 43.97, 43.84, 43.56 and 43.41 Gy. The mean PB TCP values were up to 18% higher than the mean MC TCP. EUDs calculated using both PB and MC were not sensitive to SF parameters, whereas they were for TCP calculation. Overall, MC EUDs were more sensitive to PTV volumes than PB EUDs, measured by Pearson correlation 0.46 vs. 0.07. Larger PTV volume decreased PB TCP values while this was not the case for MC TCP Conclusions: This work demonstrates encouraging evidence that radiobiological effect of target volume and dose calculation algorithm selection is significant in EUD and TCP estimations. Further studies confirming this relationship and relating to treatment outcomes are warranted. Work supported in part by NIH R01 CA106770

SU‐GG‐T‐515: The Influence of Longitudinal CT Resolution on Target Delineation and Treatment Planning for Stereotactic Radiosurgery

Purpose: To investigate the effect of longitudinal CT resolution on target delineation and treatment planning in stereotactic radiosurgery(SRS) to determine optimal acquisition parameters for SRS simulation. Method and Materials:CTimages for 6 SRSbrain patients were acquired and image sets were retrospectively reconstructed with 1, 2 and 3mm slice thicknesses. In total, 9 lesions were contoured 3 times each to quantify intra‐observer variability. The volumes contoured on the 2mm scan were the designated reference consistent with current clinical protocol. Variation in GTV with different CT slice thicknesses was evaluated. Additionally, treatment plans were created and optimized using the 2mm scans. Superimposing these plans on the 1mm and 3mm image sets provided a basis for dosimetric comparison—specifically in terms of target coverage. Results: All of the 3mm GTVs were larger than the 2mm GTVs. However, this consistent overestimation in the GTV was not observed at 1mm, suggesting that the 1mm and 2mm scans produced comparable GTVs. The average % difference in GTV volume when compared to the 2mm scans was 4.6% higher for the 3mm versus 1mm scans. Increasing CT slice thickness from 1 to 3mm caused an increase in the intra‐observer variation; volume standard deviations increased up to 51%. The intra‐observer variation also increased at larger tumor volumes. Average standard deviations (SD) in GTV extent were 0.14, 0.11 and 0.14cm in the sup‐inf, ant‐post, and lat dimensions, respectively, implying no directional dependence in target delineation. Dosimetrically, variations in GTV resulted up to a 19% loss in PTV dose coverage. Conclusion: This study indicates that longitudinal CT resolution causes considerable variation in GTV definition which can potentially compromise SRS treatment plan quality. Therefore, the SRS simulation process may benefit from improvement in scanning parameters in order to yield the most accurate information on the CT dataset.