ESTRO 2021 Abstract Book

S673

ESTRO 2021

Table 1.

Conclusion ERI confirms its role as valid predictor for pCR, also when compared with respect to delta radiomic features. The combination of ERI with delta radiomic features may slightly improve the discriminative performances of the resulting model, although the advantage does not reach statistical significance, probably due to sample dimension. Studies with larger cohorts of patients are ongoing to better identify the delta radiomic role in pCR prediction. PD-0838 Rectal cancer volume delineation between morphological and functional images: CT, T2/DWI- MRI, PET-CT C. Rosa 1,2 , F.C. Di Guglielmo 3 , L. Gasparini 3 , L. Caravatta 3 , M. Di Tommaso 3 , G. Martino 4 , P. Castaldi 4 , D. Genovesi 3,1 1 G. D'Annunzio University, Department of Neuroscience, Imaging and Clinical Sciences, Chieti, Italy; 2 SS. Annunziata Hospital, Department of Radiation Oncology , Chieti, Italy; 3 SS. Annunziata Hospital, Department of Radiation Oncology, Chieti, Italy; 4 SS. Annunziata Hospital, Nuclear Medicine Department, Chieti, Italy Purpose or Objective Nowadays, neoadjuvant chemoradiotherapy followed by surgery is a standard treatment for locally advanced rectal cancer (LARC) patients, with pathological complete response (pCR) representing the main predictive factor of clinical outcomes. Dose escalation strategies, with advanced radiotherapy techniques, could increase pCR, without compromising toxicity. T2 weighted Magnetic Resonance Imaging (MRI) represents the gold standard imaging modality for LARC staging. A potential role of functional imaging, as Diffusion-weighted MRI (DWI) and positron emission tomography (PET-CT), is now under investigation for treatment intensification approaches. Since an accurate GTV delineation is mandatory in a dose intensification strategy, the aim of this study was to evaluate the inter-observer agreement between two radiation oncologists, and the difference in volumes of Gross Tumor Volume (GTV) delineation in simul-CT, T2-MRI, DWI-MRI and PET-CT in LARC patients. Materials and Methods Twenty-four LARC patients (20 males and 4 females, mean age: 69.25 years; range: 40–88 years) were retrospectively evaluated by two radiation oncologists, blinded and independently from each other. The GTVs were manually delineated on simul-CT (CT GTV ), T2-weighted MRI (T2 GTV ), echo planar b1000 DWI (DWI GTV ) and PET-CT (PET GTV ). Observers agreement was assessed using Dice index (ranging from 0 as no spatial overlap, to 1 as spatial overlap) and Wilcoxon test. A p-value < 0.05 was considered statistically significant. Results Figure 1 reported an example of GTVs delineated on simul-CT, PET-CT, T2-weighted MRI and echo planar b1000 DWI. Table 1 reported mean and standard deviation GTV obtained for each imaging modalities for both radiation oncologists. Mean Dice index was 0.85 for CT GTV , 0.84 for T2 GTV, 0.82 for DWI GTV and 0.89 for PET GTV , representative of an almost perfect agreement between observers. Significantly larger volumes delineations were obtained for CT GTV respect to T2 GTV (p= <0.001), for T2 GTV respect to DWI GTV (p= <0.001), and for PET GTV respect to DWI GTV (p= <0.001), according to Wilcoxon test.