ESTRO 2021 Abstract Book

S1629

ESTRO 2021

PO-1908 Inverse planning in Gamma Knife radiotherapy for benign tumors M. Spaniol 1 , S. Mai 1 , T. Zakrzewski 1 , M. Ehmann 1 , F. Stieler 1 1 University medical center Mannheim, university of Heidelberg, Department of radiation oncology, Mannheim, Germany Purpose or Objective Intracranial radiotherapy using Leksell Gamma Knife® (LGK, Elekta Instrument AB, Sweden) is a well-established treatment option in radiation therapy. With the introduction of LGK Lightning (Elekta Instrument AB, Sweden), the treatment planning options in Leksell Gamma Plan® (LGP, Elekta Instrument AB, Sweden) allow real inverse planning instead of optimized forward planning. In this study, we investigated the benefits of inverse planning in single session and fractionated Gamma Knife radiotherapy in terms of plan quality and efficiency for meningiomas (men) and vestibular schwannomas (vs). The prescription doses ranged from 11 Gy to 50 Gy for vs (1–27 fractions) and 12 Gy to 52.2 Gy for men (1–26 fractions). Materials and Methods Twenty-three patient treatment plans (n=12 men, n=11 vs) that were created using forward planning with automated shot optimization in LGP and treated with LGK Icon™ at our institution were selected retrospectively as patient cohort. For each forward plan, an inverse (inv) and inverse full coverage (fc) treatment plan was generated using LGK Lightning. Plan quality (coverage, selectivity, gradient index, maximum dose to the target “D max target” and organs at risk “D max OAR”) and efficiency (beam-on-time “BOT”) characteristics were compared between the clinical forward and inverse (inv and fc) plans. Additionally, the medians were reported and the significance of the results was assessed using a paired-sample Wilcoxon test (significance level <0.05). Results For the inv plans, a significant dose reduction has been achieved for 87% of the OARs, while maintaining the target coverage. In the forward plans, the median D max OAR was 14.9 Gy (men) and 14.1 Gy (vs) and was reduced to 12.25 Gy (men, p=0.002) and 12.95 Gy (vs, p=0.007) in the inv plans. For the fc plans, a reduction was observed in 53% of the OARs where the median D max OAR was lowered to 14.4 Gy (men, p=0.192) and 13.35 Gy (vs, p=0.985). Additionally, the fc plans resulted in significantly higher target coverage. The median coverage rose from 0.98 (men) and 0.97 (vs) to 1.0 (men, p=0.005) and 0.99 (vs, p=0.006). The BOT was significantly shortened for 87% of the inv and fc plans. The median BOT was reduced for men from 22.5 min to 18.5 min (inv, p=0.007) and 19 min (fc, p=0.017) and for vs from 19.3 min to 15 min (inv, p=0.008) and 9.8 min (fc, p=0.01). Conclusion The results of this study show that inverse planning with LGK Lightning can reduce the BOT and the dose to the OARs, while maintaining the target coverage for meningiomas and vestibular schwannomas. PO-1909 Improving plan quality and reducing workload for whole breast irradiation: a semi-automatic approach L. Marrazzo 1 , S. Calusi 2 , I. Meattini 2,3 , D. Raspanti 4 , C. Arilli 1 , M. Casati 1 , A. Compagnucci 1 , C. Talamonti 1,2 , M. Zani 1 , I. Desideri 2 , L. Livi 2,3 , S. Pallotta 1,2 1 Careggi University Hospital, Medical Physics Unit, Firenze, Italy; 2 University of Florence, Department of Experimental and Clinical Biomedical Sciences “Mario Serio”, Firenze, Italy; 3 Careggi University Hospital, Radiation Oncology Unit, Firenze, Italy; 4 Tema Sinergie S.p.A., Tema Sinergie S.p.A., Faenza , Italy Purpose or Objective Breast radiotherapy represents at least 25-30% of treatments in a Radiation Oncology Department. Thus, automation is likely to play an important role in this field and to have a large impact on workload and quality. The aim of this work was implementing and introducing in the clinical practice a semi-automatic planning technique for whole breast irradiation (WBI) with 2 tangential IMRT fields. The produced dose distribution was compared with clinical non-IMRT plans and evaluated for robustness toward setup errors. Materials and Methods For left-sided breast cancer patients, imaged and treated in Deep Inspiration Breath Hold, a treatment technique (beam parameters and optimization goals) was generated using the Auto-Planning (AP) module of the Pinnacle3 (Philips) treatment planning system. Ten patients were used for AP technique assessment, while 20 further patients, who had previously been planned with 3DCRT (wedged tangential fields), were used for the validation. The technique was then extended to right- sided patients. The choice of the beam angles (gantry and collimator) was automatically performed through an in-house script and, if needed, manually adjusted. AP was run with no further manual intervention. A home-made script was used for applying a 2 cm skin flash margin after optimization. Dose distributions were evaluated in terms of dosimetric plan parameters. Statistical significance of differences was evaluated using paired two-sided Wilcoxon signed-rank test. Plan robustness toward setup errors was assessed by inducing shifts to the isocenter in latero-lateral and antero-posterior directions. Results A statistically significant improvement in target coverage was observed for autoIMRT plans compared to 3DCRT (from 95±2% to 98±2%, p=0.004). Dose homogeneity to the target was also largely improved, with a PTV V 105% changing from 11±5% (3DCRT) to 3±3% (p=0.008). No statistically significant differences were observed in OARs doses. Heart D mean was 1.1±0.6 Gy and 1.1±0.8 Gy (p=0.9) for autoIMRT and 3DCRT, respectively, while ipsilateral lung D mean was 4.3±0.8 Gy and 4.4±1.1 Gy (p=0.8). Box plots are shown in figure 1; the comparison between two dose distributions is shown in figure 2. AutoIMRT plans were also shown to be more robust toward setup errors, with a maximum loss in PTV coverage of 1% compared to 6% in 3DCRT plans. This is due to the fact that in manual 3DCRT plans the field is tightened to the target, while IMRT applies an automatic margin. Planning time was reduced from an average of about 40 minutes for manual to about 5 minutes for autoIMRT.