ESTRO 2021 Abstract Book

S1607

ESTRO 2021

readings. Results

We observed that the target coverage (V95%) was significantly higher in SeB plan as compared to SIB IMRT1and SIB IMRT 2( 99.2 vs 97.7 vs 97.9 respectively, p-0.000). However, the doses to OARs were significantly higher in SeB as compared to SIB plans. The mean kidney doses were higher in SeB compared to both SIB plans (Left Kidney: 11.1 vs 10.7 vs 10.9 Gy , p- 0.0026. Right Kidney: 11.5 vs 10.9 vs 10.9 Gy, p-0.000). The V40(cc) and V30 (cc) of bowel was significantly higher in SeB and SIB IMRT II as compared to SIB IMRT 1and (V40: 355 vs 321 vs 271 cc, p-0.000, V30: 687 vs 650 vs 635 cc). The bladder D40 (%) was higher in SeB and SIB IMRT 2 as compared to SIB (45.5 vs 44.6 vs 43 Gy, p-0.005). Rectum V30 was higher in SeB and SIB IMRT2 as compared to SIB 1(80.3 vs 79.3 vs 79.8 Gy, p-0.007). There was no difference in the doses to femoral heads between the plans was observed. Conclusion The target coverage is higher in SeB plans with modest increase in OAR doses volume paratmers especially for bowel bag.This was observed despite that nodal shrinkage was considered and a second CT was acquired for nodal boost in SeB. Simultaneous integrated boost technique provides a therapeutic gain as compared to sequential plans. PO-1887 Auto-segmentation and knowledge-based planning for radiotherapeutic palliation of spine metastases E. Jones 1 , C. Napoleone-Filho 1 , C. Sisodia 1 , M. Couper 2 , V. Harris 3 , A. Greener 2 1 Guy's and St Thomas' NHS Foundation Trust, Medical Physics, London, United Kingdom; 2 Formerly Guy's and St Thomas' NHS Foundation Trust, Medical Physics, London, United Kingdom; 3 Guy's and St Thomas' NHS Foundation Trust, Clinical Haematology and Oncology, London, United Kingdom Purpose or Objective Palliative treatment of spine metastases should be delivered in a timely manner, achieve durable symptom control with little toxicity and minimal patient inconvenience. As such the standard of care is typically simple, single or parallel opposed beams to deliver a high dose of radiation to the target area. The practicability of using VMAT rather than conventional planning for spine metastases patients is limited due to the complexity and labour-intensive nature of VMAT treatment planning, treatment delivery and quality assurance. Here we present the feasibility of a comprehensive ‘automated’ treatment planning workflow for the radiotherapeutic palliation of thoracic spine metastases using atlas-based auto- segmentation (ABAS) and knowledge-based planning (KBP) solutions. Materials and Methods 16 patients had the thoracic vertebrae and OAR (oesophagus, heart, kidneys, lungs, spinal cord and cauda equina) delineated. These formed the ABAS solution for application to subsequent patients using Eclipse SmartSegmentation. VMAT plans were produced for 20 patients (each with 2 CTVs incorporating T6-8 and T10-12) to produce a 40 plan Eclipse RapidPlan KBP model. ABAS and KBP were subsequently applied to a further 8 patients (16 CTVs) and evaluated to determine the feasibility of these ‘automated’ solutions for generating VMAT plans for this cohort of patients. Results The mean time to perform ABAS of T1-T12 and OAR was 1.6 mins and 1.8 mins respectively. Table 1 shows similarity metrics of DICE, volume difference and centre of mass shift compared to the manually modified structures.

Table 1: Similarity metrics for 8 patient ABAS. All ABAS structures required manual modification prior to RapidPlan generation. ABAS of CTVs performed well in some patients, achieving a maximum DICE = 0.92 for T10, but scored DICE = 0 for some patients due to misidentification of individual vertebrae. ABAS of OAR was variable, performing well for lungs and heart, DICE = 0.99 and 0.89 respectively and adequately for spinal cord and cauda equina, DICE = 0.72 and 0.69 respectively. But performed less well for kidneys and oesophagus, DICE = 0.57 and 0.1 respectively. RapidPlan KBP plans were generated. The mean dose to CTV1 = 99.4% (min 92.6%, max 105.1%), the mean dose to CTV2 = 98.0% (min 86.6%, max 104.6%). OAR in the vicinity of CTV1 received max doses: spinal cord 98.8%, lungs 101.7%, heart 91.4% and oesophagus 59.2%, those in the vicinity of CTV2 received max doses: spinal cord 96.7%, lungs 99.2% and kidneys 76.5%. Conformity and homogeneity were superior, with significant reduction of dose to the superficial tissues, compared to simple radiotherapy. The mean time to generate the plan was 7.1 mins.