ESTRO 2020 Abstract book

S983 ESTRO 2020

SUVmax appears to correlate with a metabolic response. These results need to be confirmed in larger studies. PO-1691 Apparent diffusion coefficient changes in weekly MRI during radiotherapy in head and neck cancer S.P. Ng 1 , C. Cardenas 2 , H. Bahig 3 , B. Elgohari 2 , A. Moreno 2 , S. Shah 2 , A. Garden 2 , J. Phan 2 , G.B. Gunn 2 , S. Frank 2 , D. Rosenthal 2 , W. Morrison 2 , J. Wang 2 , C. Fuller 2 1 Peter MacCallum Cancer Centre, Radiation Oncology, Melbourne, Australia ; 2 The University of Texas MD Anderson Cancer Center, Radiation Oncology, Houston, USA ; 3 Centre hospitalier de l'Université de Montréal, Radiation Oncology, Montreal, Canada Purpose or Objective To quantify change in apparent diffusion coefficient (ADC) of primary tumor on weekly magnetic resonance imaging (MRI) obtained during radiotherapy in patients with mucosal head and neck squamous cell carcinoma. Material and Methods Patients with localized mucosal head and neck squamous cell carcinomas undergoing definitive radiotherapy were enrolled on the prospective IRB approved PREDICT-HN study. Pre-treatment and weekly in-treatment MRIs were obtained in radiotherapy treatment position. T1-weighted, T2-weighted and DWI sequences were obtained on 1.5T Siemens MRI. ADC maps were generated. Primary gross tumor volume (GTV) was contoured on T2-weighted MRI. GTV volume and ADC parameters were recorded. Patient, tumor and treatment characteristics were documented. As GTV decreases over time and measurements of ADC may be inaccurate, ADC values of pre-treatment GTV area across the serial MRI were used for analysis. Longitudinal changes in ADC values were analysed using mixed models. Results A total of 36 patients completed the study. The median age was 58.5 years (range: 41 – 81 years) and 33 were males. The predominant primary site of disease was tonsil (42%), followed by base of tongue (28%). Twenty- three (64%) had p16/ HPV-positive disease. The median dose/ fractionation delivered was 6996 cGy in 33 fractions. Twenty-three (64%) received volumetric modulated arc therapy (VMAT) and 13 had intensity modulated proton therapy (IMPT). Thirty patients (83%) received concurrent chemotherapy. The median pre-treatment GTV volume was 14.1cc (range: 1.3 – 44.9cc). Overall, there was significant decrease in GTV volume in week 4 of treatment. The median ADC for pre-treatment GTV was 92 (range 16 – 1796). Throughout treatment, there was a significant rise across 25 th – 75 th percentiles of ADC values (p<0.0001). At 2 to 3 months post-treatment evaluation, 34 patients had complete response and 2 had partial response at the primary site on imaging. For the patient who had partial response, there was an initial rise in ADC values in weeks 1 to 4 but in weeks 5 and 6, the ADC values decreased to similar values as weeks 2 and 3. Conclusion Preliminary results from this study showed that ADC changes significantly during radiotherapy. Further follow up is required to determine if changes in ADC during radiotherapy predicts for subsequent local failures. PO-1692 A novel approach to proton radiography and CT using a large-volume scintillator detector system C. Darne 1 , D. Robertson 2 , F. Alsanea 1 , T. Pan 3 , D. Grosshans 4 , S. Beddar 1 1 The University of Texas MD Anderson Cancer Center, Radiation Physics, Houston, USA ; 2 Mayo Clinic, Radiation Oncology, Phoenix, USA ; 3 The University of Texas MD Anderson Cancer Center, Imaging Physics, Houston, USA ; 4 The University of Texas MD Anderson Cancer Center, Radiation Oncology, Houston, USA

regions; 3 ) No Response (NR) - SUVpost was the same as or greater than SUVpre . Morphological: relied in RECIST 1.1 The variables age, type of tumor, dose-fractionation scheme were taken into account with SUVmax and TZ to estimate a possible correlation with local control (LC), In- field progression (IFP) and distant progression (DP). The differences between SUVmax and TZ at 5 and 8 months were calculated ( table 1 ). The possible changes during the evaluation and the final results were analyzed by means of a Kruskal Wallis test as a possible predictor of three clinical results: LC, IFP and DP. Results Thirty patients met the criteria, with a median follow-up of 17.2 months (range: 12-31). 86.7% were male, 40% were between 61-70 years old. 80% were primary tumors, 86.7% were peripheral lesions. The most commonly used fractionation scheme was 5 x10 Gy (76.7%) ( table 2 ). The median gross tumor volumen was 1.91 ± 1.17 (range 0.50- 6.70), pre treatment SUVmax was 7.71 ± 4.37( range 1.00 - 17.2). LC rate 94%. No statistical correlation was observed between the pre-SBRT variables (SUV max and TZ) and clinical parameters, including dose-fractionation scheme ( p =0.080). SUVmax was significantly related to metabolic response ( p = 0.039), but not for clinical outcomes ( p =0.879). TZ shows no correlation for morphological and clinical outcome ( p = 0.329).

Conclusion According to these results, the differences between TZ and SUVmax at 5 and 8 months after treatment, couldn’t be used as early predictors of LC, IFP and DP, although

Purpose or Objective