Conclusions: Monaco VMARS was comparable to LGK SRS in terms of PGI and PCI for targets larger than 1 cc. Beam delivery time for Monaco VMARS was significantly shorter than that for LGK SRS. Monaco VMARS has a potential to be an alternative option for treatment of multiple brain metastases. PD-0452 Knowledge-based treatment planning of volumetric arc modulated radiotherapy for cervical cancer Purpose/Objective: There is a growing trend towards algorithms which improve the efficiency and

more »

... ncy of treatment plan generation. One such implementation is the Varian RapidPlan™ knowledge-based treatment planning module which uses a library of clinically acceptable treatment plans to generate a model that can predict optimum DVH optimisation objectives on a patient-specific basis. The aim of this study was to evaluate the suitability of using RapidPlan to create acceptable VMAT plans for cervical cancer cases against those that were previously considered clinically acceptable. Materials and Methods: The RapidPlan model was generated, according to manufacturer guidance, based on 37 clinical patients treated for Cervical Cancer using VMAT. The model included all patients with varying clinical and physiological indications, i.e. all intact uteri and post-hysterectomy cases were included, as well as those where the inguinal nodes were clinically included. The two different prescription levels used in our centre were included; 45Gy/25 fractions and 50.4Gy/28 fractions. The model required minor adjustments after generation; to fix the PTV optimisation priorities and the normal tissue objective. VMAT plans were independently re-generated for 5 different cases using RapidPlan who had not been included in the model configuration. The cases included post-hysterectomy (45Gy/25#, n=2) and intact-uteri (50.4Gy/28#, n=3) cases. A single optimisation was used to assess the quality of the model. VMAT plans consisted of one isocentre and two coplanar 360° arcs. Doses to the rectum, bowel, bladder, femoral heads and bone marrow were evaluated. The 95% conformity index (CI 95% ) was calculated for the PTVs, as well as D 98% and D 2% ; representing the nearmin and near-max doses respectively. The total number of monitor units required for the two techniques was compared. Results: Table 1 shows a summary of the average parameter values (± s.d.) for the PTV, OARs, and MU; p-values are based on the Wilcoxon signed-rank test. In all cases, RapidPlan indicated on average comparable or better sparing of the OARs, with the most significant difference being for the femoral heads, bladder D 1cm3 and bone marrow V 10Gy . PTV coverage was more conformal for RapidPlan as indicated by the CI 95% being on average 0.09 lower than the original plans (p=0.01). The D 98% (p=0.19) and D 2% (p<0.01) were marginally better for the original plans. The MUs for RapidPlan were on average 57 MU higher than the original plans (p=0.14), which is due to increased modulation as indicated by the better dose conformity.