ESTRO 2020 Abstract book

S911 ESTRO 2020

Results Of the 60 patients, 27% reported xerostomia (grade>1) at 12 months. The highest median mAUC train / mAUC test were 0.93/0.83, 0.90/0.79 and 0.93/0.80 for the SL, DL and WP, respectively. As shown in Figure 2, the 50 combinations of predictors that were selected for the WP gland yielded a significantly higher median mAUC test compared to the DL (p<0.01). The selected combinations of the SL yielded higher median mAUC test compared to the WP (p<0.01).

Kingdom ; 8 Edinburgh Cancer Centre, Department of Oncology Physics / The University of Edinburgh- School of Engineering, Edinburgh, United Kingdom Purpose or Objective Several studies have investigated the relationship between the clinical symptoms of xerostomia and the radiotherapy dose received by subregions of the parotid glands (e.g. superficial lobe/stem cell region). The purpose of this study was to compare the predictive power of image biomarkers (IMB) calculated on the superficial lobe (SL), deep lobe (DL) and whole parotid (WP) gland for identifying late xerostomia in head and neck (H&N) cancer patients. Material and Methods All patients (N=60) received 30 fractions (fx) on a TomoTherapy HiArt System (Accuray, Sunnyvale, CA, USA) with daily mega-voltage CT (MVCT) images with CTCAE toxicity recorded at 12 months. As illustrated in Figure 1, the MVCT images (0.76 x 0.76 x 6 mm) were used to calculate by textural analysis, seventy-three IMBs consisting of first and higher order features on the SL, DL and WP gland. For each feature, their values for the period 1-20fx were used to perform linear regressions and their slopes considered as potential predictors. LASSO analysis was used (100 times) with 4-fold cross validation to select the best predictors (<9) for each region. To evaluate the robustness of each of the 100 combinations, the patients were split into four folds. The training set made up of 3 folds and the testing set, the remaining fold, were used to test the predictive power of each of the LASSO combinations using a logistic regression model. This was evaluated by the Area Under the Curve (AUC), resulting in 4 AUC test and 4 AUC train values which were averaged (mAUC test and mAUC train ). For robustness, this process was repeated 100 times for each combination of predictors and the median mAUC test and mAUC train calculated. The 50 combinations of predictors with the highest median mAUC were selected for the SL, DL and WP and their median mAUC test distributions compared using a t-test.

Conclusion The slopes of the IMBs extracted from the DL (from fraction 1-20) have a lower predictive power compared to those extracted from the WP gland. Despite the smaller number of voxels considered when analyzing the SL, the IMBs extracted from the SL have a higher predictive power compared to those from the WP. However, this requires further validation on a larger cohort and different selected periods. To obtain a greater understanding of the differences in predictive power of each of these subregions requires further information on the underlying biological processes. PO-1586 Prediction of late xerostomia with clinical, atlas based and deep learning contours L.V. Van Dijk 1,2 , C.D. Fuller 1 , C.S. Mayo 3 , S.Y. Lai 4 , A.S.R. Mohamed 1 , K.A. Hutcheson 4 1 MD Anderson Cancer Center, Radiation Oncology, Houston, USA ; 2 University Medical Center Groningen, Radiation Oncology, Groningen, The Netherlands ; 3 University of Michigan, Radiation Oncology, Ann Harbor, USA ; 4 MD Anderson Cancer Center, Head and Neck Surgery, Houston, USA Purpose or Objective Prediction of radiation-induced toxicities based big data is important in order to guide treatment, for example to select patients for proton therapy or guide dose optimization. Large datasets with adequate organ at risk (OAR) delineations are crucial for developing and validating toxicity prediction models. Unfortunately, large curated datasets are often not readily available, since delineated OARs are regularly missing or of inadequate quality. Multiple head and neck OAR auto-segmentation have been published in recent years, making dosimetric analysis on larger datasets feasible. Nevertheless, investigating the effect of auto-segmentation on model development and performance remains to be investigated. Our hypothesis is that toxicity prediction performance is similar using automated segmentation compared to clinical used contours. The purpose is to test the robustness of auto-segmented parotid gland contours for predicting moderate-to-severe radiation-induced xerostomia 12 months after radiotherapy (Xer 12m ). Material and Methods Clinically available, atlas-based (AB) and deep learning (DL) based auto-contours were obtained for 172 head and