Abstract Book

S754

ESTRO 37

Figure 1 : CT images of A) a tumor without cavitation and B) a tumor with cavitation.

Table 1 : Performance of the radiomic models predicting semantic features for the training- and validation dataset. Semantic feature c-index (training) c-index (validation) Gender 0.604 0.549 Cavitation 0.806 0.668 Pleural Retraction 0.681 0.550 Emphysema 0.554 0.463 GGO 0.598 0.456 Air Bronchogram 0.790 0.632 Pleural Thickening 0.477 0.620 Necrosis 0.703 0.586 Satellite Nodules 0.624 0.524 Suspect Nodules 0.616 0.450 Fibrosis 0.653 0.524 Pleural Effusion 0.784 0.735 Pleural Contact 0.749 0.616 Conclusion It is possible to describe some common semantic features by a combination of radiomic features. Since our analysis only looked at the primary tumor, model performance of lung specific features can potentially be increased by calculating radiomic features for the whole lung. Patients with lung abnormalities generally have a worse prognosis, models containing semantic features as well as radiomic features have potentially a better prognostic value. EP-1381 Role of Metabolic Tumor Volume and Total Lesion Glycolysis on FDG-PET/CT in NSCLC treated with SBRT J. Cabrera 1 , J. Infante 2 , C. Cruz 1 , M. Moreno 2 , M. Gonzalez 1 , J. Rayo 2 , P. Simon 1 , B. Ortiz 1 , J. Muñoz 1 1 Universitary Hospital Infanta Cristina, Radiation Oncology., Badajoz, Spain 2 Universitary Hospital Infanta Cristina, Nuclear Medicine, Badajoz, Spain Purpose or Objective Maximum Standardized Uptake Value (SUV max) is the most widely applied parameter on FDG-PET/CT for staging and monitoring treatment response in NSCLC. Recent work suggest that volumetric parameters Metabolic Tumor Volume (MTV) and Total Lesion Glycolysis (TLG) can be used as prognostics factors in surgical and non-surgical series of patients with early stage NSCLC. The aim of this study was to determine whether MTV and TLG are associated with outcomes in patients treated with SBRT. Material and Methods Retrospective study of 42 patients treated with SBRT for NSCLC Stage T1 – T2 (AJCC 8 th ed.) between June 2008 to February 2016. Mean age 75.3 ± 9.2 years. Men: 34, women: 8. SBRT dose BED 10 mean was 108.6 Gy ± 16.7. Biopsy proven cancer: 22, non-biopsy: 20. Central tumors: 7, peripheral tumors: 35. The volume within the GTV with SUV greater than or equal to a given SUV threshold x was MTV X TLG was defined as MTV X x SUVmean. MTV and TLG at several levels were registered. Prognostic factors for overall survival (OS) local control (LC) and disease free survival (DFS) were analyzed using Cox’s proportional hazards model and survival curves were calculated using the Kaplan-Meier method. P values < 0.05 were considered statatistically significant.

Material and Methods On thoracic CT scans of 235 NSCLC patients (center A, training), experienced radiologists scored 13 semantic features (see table 1) for all patients. Because we assume that tumors show characteristics related to features in the whole lung we extracted 1417 features from the primary tumor, using in-house developed radiomics software. For validation, a dataset of 113 NSCLC patients from another center (center B) was used. Spearman’s correlation coefficient (≥0.85) was used to determine radiomic features that were strongly correlated. Per correlation pair, the feature with the highest average correlation was excluded to remove redundant information. Next, an ANOVA analysis was performed to select the significant radiomic features per semantic feature and model performance of a logistic regression was tested by calculating the c-index. The logistic regression model was validated in the dataset of center B. Results In total 130 radiomics were selected after removal of redundance and used for the logistic model to predict semantic features. The semantic feature models consisted of 1-15 radiomic features. Model performance is shown in table 1.