ESTRO 38 Abstract book

S1044 ESTRO 38

potential to stratify for cell line-specific radiation sensitivity. Material and Methods The study was performed on immuno-deficient nude mice with tumors from different HNSCC cell lines grown in the hind leg (n=46) that had been measured on a 7T-PET/MR scanner (Bruker Biospec). The imaging protocol included anatomical T2w MRI and DW-MRI using 8 b-values between 0 and 800 s/mm^2. Based on previously published tumor control doses 50% (TCD50) [1], the cell lines were grouped into three categories of radiation sensitivity: high (H), medium (M), and low (L) (cf. Table 1). The GTV was defined manually on T2w MRI for each mouse by an experienced radiation oncologist . ADC maps were derived from the DW-MRI images by a mono-exponential fit. Two methods were used to define ADC-based variables for stratification according to radiosensitivity: mean ADC (M1) and volume fraction of ADC voxels below a pre- determined threshold (M2). For M2, the threshold was determined on a sub-group of the total cohort (n=30) based on optimal separation of the groups H-M and M-L. Final analysis of stratification potential using the determined threshold was performed on all samples. Results The optimal threshold derived for M2 was ADC = 450E-6 mm^2/s. The results of the two ADC-based variables for each sample are shown in Figure 1. M1 could stratify groups M and L (p = 7.83E-5, U-test), but not groups H and M (p = 0.17). M2 allowed to stratify all three groups (H-M: p=9.82E-5, M-L: p=7.83E-5). UTSCC-45 seems to have higher volume fractions in M2 compared to other high- sensitive cell groups (cf. Figure 1, M2). Possibly this is due to the fact that this cell line is HPV positive in contrast to all other cell lines. Conclusion ADC was found to be a good biomarker for stratification of GTVs into different radiosensitivity groups in small-animal tumors. Particularly, the fractional volume of low ADC (M2) outperformed mean ADC (M1). These findings are being translated into a currently running clinical study to explore the performance of fractional volumes of low ADC against the more commonly used mean ADC. [1] Yaromina et al., Radiother Oncol 96:116 (2010) Acknowledgements: Research has received funding from the European Research Council (ERC).

seems to relate to local response in high dose regions.

Conclusion This framework uses multimodal deformable anatomical image registration to relate functional imaging changes to treatment information (dose and LET). This may allow in vivo assessment of RBE dependence on LET for proton RT in larger patient cohorts. EP-1920 ADC mean versus fractional volume to predict radiation sensitivity of HNSCC xenografted in nude mice S. Leibfarth 1 , R.M. Winter 1 , S. Böke 1,2 , P. Mena-Romano 1 , E.C. Sezgin 2 , M. Krüger 3 , B. Pichler 3,4 , D. Zips 2,4 , D. Thorwarth 1,4 1 Section for Biomedical Physics- Department of Radiation Oncology, University Hospital Tübingen, Tübingen, Germany ; 2 Department of Radiation Oncology, University Hospital Tübingen, Tübingen, Germany ; 3 Department of Preclinical Imaging and Radiopharmacy- Werner Siemens Imaging Center, Eberhard Karls University Tübingen, Tübingen, Germany ; 4 German Cancer Consortium DKTK- partner site Tübingen, and German Cancer Research Center DKFZ, Heidelberg, Germany Purpose or Objective Pre-therapy diffusion weighted MR imaging (DW-MRI) has been associated with clinical RT outcome in various studies. In most studies the mean apparent diffusion coefficient (ADC) in the gross tumor volume (GTV) has been investigated for patient stratification. To further explore the potential of pre-therapeutic ADC, ADC maps obtained in pre-clinical head and neck squamous cell carcinomas (HNSCC) tumor models were studied for their

Table 1 : Characteristics of the small animal cohort. TCD50 values were taken from [1].