ESTRO 2020 Abstract book

S875 ESTRO 2020

indicate a potentially increased cerebral radiosensitivity of the periventricular region (PVR) and a variable relative biological effectiveness (RBE) for proton therapy of the brain (Eulitz 2019, Harrabi 2019, Peeler 2018). The purpose of this study was to investigate the radiosensitivity of the PVR and the variability of clinical proton RBE using clinical data. Material and Methods All grade II and III glioma patients treated between 2014 and 2017 with (adjuvant) proton radio(chemo)therapy using the passive scattered technique to a total dose of 54- 60 Gy(RBE) were considered for analysis. Prospective follow-up included three-monthly contrast-enhanced T1- weighted MR imaging (fuMRI). Contrast enhancement (CE) diagnosed as treatment-related brain lesions (symptomatic or clinically silent) were traced back to the fuMRI of first appearance, delineated and deformably co- registered to the MRI and CT used for proton treatment planning (Fig 1A-D). Only lesions progressing over time, and/or those persisting for ≥ 6 months follow-up, and located outside the gross tumour volume were included in the analysis. A 4 mm expansion around the cerebral ventricles was contoured as PVR (Fig 1D). High-precision Monte-Carlo simulations (TOPAS) provided dose (D) and linear energy transfer (LET). CE spots were correlated voxel-wise using uni- and multivariable logistic regression analysis with the predictors D, LET and being in- or outside of the PVR. Additionally, area under the curve (AUC) and pseudo R-squared (pR 2 ) values were calculated for assessing the discrimination and prediction power of each model, respectively.

(N>35 000). Using transfer learning with images from the same anatomical site can yield remarkable performance in new tasks with as few as N=50. Surprisingly, we did not find any advantage for using images from other anatomical sites over networks that have been trained using ImageNet. This indicates that features learned may not be as general as currently believed, and performance decays rapidly even by just changing the anatomical site of the images. PO-1528 Predicting response to neoadjuvant chemoradiation in esophageal cancer using CT radiomic features I. Mallick 1 , V.H. Lal 1 , M. Arunsingh 1 , T. Bhattacharyya 1 , S. Chatterjee 1 , S. Chakraborty 1 , R. Achari 1 1 Tata Medical Center, Department of radiation Oncology, Kolkata, India Purpose or Objective To determine if CT based radiomic feature extraction can identify complete pathological responders from neoadjuvant chemoradiation (NACRT) in operable esophageal cancer. Material and Methods From a dataset of 41 patients with squamous cell carcinoma (SCC) of the esophagus who underwent NACRT with 41.4 Gy of radiation and concurrent weekly paclitaxel and carboplatin followed by esophagectomy we identified 20 patients with a complete pathological response (pT0) and 21 patients with a poor response (pT3). The tumor was delineated on the planning CT scan. DICOM files were read and radiomic features extracted using open source libraries (pydicom and PyRadiomics). Univariate feature selection was performed and several classifiers (logistic regression, k-nearest neighbors, random forest and support vector machines) were trained. The model was tested on a test data sub-set. Results Of the 107 radiomic parameters tested, using univariate feature extraction we identified eight radiomic parameters that were different in the two groups. These included first order features (kurtosis), gray level co- occurrence matrix features (inverse difference, inverse difference normalized, inverse difference moment normalized), gray level run length matrix features (runlength non-uniformity normalized, run percentage), neighbouring gray tone difference matrix features (contrast) and gray level difference matrix features (dependence non-uniformity normalized). The random forest classifier performed the best and had a precision of 0.67 for predicting pCR, an F1 score of 0.67 and area under curve of 0.78. The accuracy score was modest at 0.56. Conclusion In this exploratory dataset, CT based radiomic features had modest success at classifying complete pathological response. Larger datasets and multimodality imaging need to be used to improve the accuracy of prediction. PO-1529 Increased cerebral radiosensitivity of the periventricular region in proton therapy of gliomas A. Lühr 1 , F. Raschke 2 , C. Karpowitz 3 , F. Permatasari 1 , B. Lutz 4 , W. Enghardt 1 , M. Krause 1 , E. Troost 1 , J. Eulitz 1 1 OncoRay - National Center for Radiation Research in Oncology, Institute of Radiooncology – OncoRay, Dresden, Germany ; 2 Helmholtz-Zentrum Dresden - Rossendorf, Institute of Radiooncology – OncoRay, Dresden, Germany ; 3 Faculty of Medicine and University Hospital Carl Gustav Carus, Department of Radiotherapy and Radiation Oncology, Dresden, Germany ; 4 Helmholtz-Zentrum Dresden - Rossendorf, Institute for Radiation Physics, Dresden, Germany Purpose or Objective In neuro-radiooncology, a constant radiosensitivity is assumed within the brain. However, recent findings

- Results 39 patients fulfilling the inclusion criteria were included in the analysis and the median follow-up was 23 months. Median time between first fraction of radiotherapy and the earliest detection of a CE lesion was 19.3 months. The mean distances from the centre of the earliest detected lesion to the lateral cerebral ventricles and clinical target volume (CTV) were 2.3 mm and 2.6 mm, respectively (Fig 2). Voxels within lesions were significantly spatially correlated with D (AUC = 0.71), the PVR (0.72), and D×LET (0.75; p < 0.001, respectively). The multivariable combination of D and D×LET revealed an AUC and pR 2 value of 0.77 and 0.08, respectively. The model with D, PVR and D×LET as predictors resulted in the highest AUC and pR 2 values of 0.88 and 0.16, respectively, and the predicted probability for CE within an image voxel is illustrated in Fig 1F.