ESTRO 38 Abstract book

S650 ESTRO 38

Conclusion In conclusion our prospective clinical data point out the N/Dg trend during RT or CMT in HN cancer patients. We provided a quantitative and qualitative assessment of N/Dg, potentially useful for future comparison. EP-1175 Use of different RBE-models in carbon ion RT - saving OAR constraints from being lost in translation J.E. Dale 1,2 , S. Molinelli 3 , V. Vitolo 3 , B. Vischioni 3 , M. Bonora 3 , G. Magro 3 , A. Mairani 3,4 , A. Hasegawa 5 , O. Dahl 2 , F. Valvo 3 , P. Fossati 3,6 1 Haukeland University Hospital, Department of Oncology, Bergen, Norway ; 2 University of Bergen, Department of Clinical Science- Faculty of Medicine, Bergen, Norway ; 3 National Center for Oncological Hadrontherapy, Clinical department, Pavia, Italy ; 4 Heidelberg Ion-Beam Therapy Center, Medical physics unit, Heidelberg, Germany ; 5 Osaka Heavy Ion Therapy Center, Clinical department, Osaka, Japan ; 6 MedAustron, Carbon ion radiotherapy unit, Wiener Neustad, Austria Purpose or Objective In order to optimize carbon ion radiotherapy (CIRT) there is a need to validate dose constraints for important organs at risk. For the optic nerve (ON), constraints have been validated by the National Institute of Radiobiological Sciences (NIRS), Japan, in which the Relative Biological Effectiveness (RBE) of the CIRT has been predicted by the mixed beam model (RBE NIRS ). These constraints are not immediately useful for centers where the Local effect model (RBE LEM ) is used, because comparative studies show that RBE LEM can predict a 60% higher RBE in the entrance region of the beam, and 5-15% higher RBE in the spread- out Bragg peak, relative to RBE NIRS . At the National Center for Oncological Hadrontherapy (CNAO), Italy, current dose constraints for ONs comply with the NIRS constraints: D1% < 40 Gy(RBE) and D20% < 28Gy(RBE), although RBE LEM is used in treatment plan optimization. This is a conservative approach, possibly resulting in excessive underdosage to tumors close to the optic pathways. The aim of this work is to improve CNAO’s ON dose constraints by analyzing institutional toxicity and by relating the results to RBE NIRS . Material and Methods A total of 65 optic nerves from 38 patients treated at CNAO with CIRT to the head and neck region in the period 2013- 14 were analyzed. The physical dose of the patients’ treatment plans was reproduced. Subsequently, both the RBE LEM and RBE NIRS was applied as RBE model, thus relating CNAO clinical toxicity to the NIRS constraints in RBE NIRS - weighted dose. Prediction of doses to selected volumes of the ONs giving x% probability of toxicity (TDx%) was derived using logistic regression. Results Median follow-up time was 47 (26-67) months. No toxicity occurred in the 56 ONs in which the current constraints were obeyed. Three ONs developed visual decline at doses D1%≥71 Gy(RBE LEM )/68 Gy(RBE NIRS ) and D20%≥68 Gy(RBE LEM )/62 Gy(RBE NIRS ). Table 1 presents patient characteristics and results of NTCP modelling. TD50 was comparable to the NIRS publication, while TD5 was substantially higher in our results, probably due to a scarcity of ONs receiving mid- to high doses relative to the ONs in NIRS publication. Figure 1 presents DVHs of all ONs in both RBE LEM - and RBE NIRS -weighted doses and demonstrates that potential new CNAO D1% and D20% constraints should be maximum 49 and 40 Gy(RBE LEM ), substantially lower than the predicted TD5 doses, in order to comply with the respective NIRS constraints of 40 and 28 Gy(RBE NIRS ).

For detailed CiTAS score results see Tab.2 . Mean Hypo- Ageusia score was 1.14(±0.4) at Bs progressively increasing during RT and reaching the maximum values during the VII week at 2.82(±1.4) and decreasing after RT down to 2.13(±1.23) at 1 month from the end of treatment. The same could be observed for discomfort score which was 1.14(±0.47) at Bs, increased up to 1,74(±0.98) at week VII and decreased at 1.27(±0.45)1 month after treatment. The phantogeusia/parageusia score, which was 1.16(±0.60) at Bs, increased to 1.92(±1.08) at week VII and decreased to 1.36(±0.7) at week I after RT. A similar pattern was observed for general taste alterations score: 0.17(±0.43) at Bs, 2.35(±1.04) at week VII of RT and 1.73(±0.87) after 1 month from RT end.