OC-0177: Effects of calcification in post-planification dosimetry for low-dose rate prostate seed brachytherapy treatment
Radiotherapy and Oncology
2 nd ESTRO Forum 2013 S69 prostate volume and volume that received 100% of the prescribed isodose, respectively) achieved in glands ≥ 60 cc were not significantly different compared to those with glands < 60 cc (p≥ 0.2). Nonetheless, biochemical control of disease in patients with larger CTV was significantly higher (91% vs 78% at 6 years; p = 0.004). In uni and multivariate analysis, CTV was a significant predictor for risk of biochemical relapse. This was not at the expense of an increase in
... of an increase in either moderate (p = 0.6) or severe (p = 0.3) late genitourinary toxicity. Usage of hormonal therapy was 17% lower in the large gland group (p = 0.01). There was no significant difference in FFbR or in the incidence of late genitourinary or rectal morbidity between the three fractionation schedules. Conclusions: Prostate gland size does not affect dosimetric parameters in HDR-BT assessed by D90 and V100. In patients with larger prostate glands a significantly higher biochemical control of disease has been observed with no difference in late toxicity. This improvement cannot be attributed to differences in dosimetry. Gland size should not be considered when selecting patients for HDR-BT. OC-0177 Effects of calcification in post-planification dosimetry for low-dose rate prostate seed brachytherapy treatment C. Purpose/Objective: Even though calcifications are observed in clinical practice, the actual dose calculation algorithm (TG-43) overrides them with water material and density. However, it is a known fact that at low energy in LDR seed implants the cross section depends strongly on the effective atomic number, increasing the probability of important dose discrepancies around the calcified regions. It is hypothesized that significant effects on the dose metrics such as the dose volume histogram (DVH), target volume and urethra are not taken into accounts. This study investigates these effects on delivered dose from calcifications in post-implant dosimetry. Materials and Methods: 42 patients with visible calcifications in the prostate were identified. Threshold segmentation is used to distinguish normal prostate tissue from calcification. Monte Carlo (MC) simulations are made with the Geant4-based ALGEBRA software  in 1mm 3 voxels. Materials are assigned based on TG-186 recommendations and the sources are explicitly modeled. Breast calcification definition is used to define the prostate calcification with density taken from the CT numbers. We identified 6 different dose calculation scenarios for our study: MC in water with (1) and without (2) seeds, 2) MC in a water prostate with densities from the CT but with calcification areas with (3) and without seeds (4) and finally full tissue heterogeneities with (5) and without seeds (6). Results: Relative to the TG-43 algorithm, we measure large differences, up to -27 (-28)% on the D90 (V100) with full tissue heterogeneities and seeds modeled, and up to -25 (-22)% when only the calcificified areas and seeds are taken into account. The negative sign indicated a decrease in dose relative to TG-43. Calcifications, inter-seed attenuation and tissue heterogeneities can lead to interplay phenomena. Hot-spot regions as defined by V200 are the most affected, with difference to TG-43 up to -74%. The urethra also shows significant dose discrepancies depending on the location of the calcifications. Urethra D90 differences of -17% relative to full MC calculations are observed, -10% considering calcification only. For all cases under investigation, the average decrease of D90 is -15% for the prostate and -8% for the urethra. Conclusions: Taking into account tissue heterogeneities, composition, and calcifications composition produces the highest change relative to TG-43. Based on the patient cohort, calcifications have a measurable impact on the target volume dose parameters and also may bring a significant decrease in the delivered dose to the urethra. The real composition of the prostate calcification remains to be determined to establish a better understanding ofthe dosimetric consequences.  Afsharpour, et al, 'ALGEBRA: ALgorithm for the Heterogeneous Dosimetry Based on GEANT4 for BRAchytherapy. The Christie NHS Foundation Trust, Clinical Oncology, Manchester, United Kingdom Purpose/Objective: MRI provides the gold standard for tumour delineation in gynaecological brachytherapy. While traditionally CT has been used for applicator reconstruction and then fused with MRI, in recent years the use of MRI only for the reconstruction of applicators, with or without MRI compatible line markers, has become more widespread. The purpose of this study is to evaluate the interoperator differences in DVH parameters for OAR and HR-CTV when reconstructing titanium applicators with a fixed geometry, based only on T2 MRI images. Line markers are not currently commercially available for this applicator type. GEC-ESTRO recommendations regarding HR-CTV were followed and identical library plans used, prescribing to point A. Materials and Methods: Three patient plans using an identical OAR and HR-CTV structure set were used to retrospectively reconstruct a titanium Rotterdam type applicator using a 3D library model. Eight regular planners (radiographers, medical physicists and dosimetrists) independently reconstructed applicators and set an origin point based upon the applicator model. Identical library plans were loaded onto each applicator and the variations in DVH parameters for both HR-CTV and OAR assessed. Results: Co-ordinate positions of point A were used as a surrogate for applicator model alignment to the applicator image. Vector differences between eight regular planners were assessed by comparing the average co-ordinates of point A between all of the plans. The vector differences between the co-ordinates, compared against the average co-ordinates, show that the mean absolute difference between planners is 1.0mm (SD 0.4mm). All values were within 2mm, which shows good agreement between planners. The average percentage absolute-differences in DVH parameters arising from these variations, when compared against the average value for each, can be seen in Table 1 . Table 1 These results show that the average absolute differences are very low (<=3.3%), with the highest difference being for the HR-CTV D100%, which is known to be less reproducible than the D98% for HR-CTV.