ESTRO 2021 Abstract Book

S289

ESTRO 2021

Conclusion In this study we showed that DCE and IVIM derived parameters generally increase in the tumor and prostate during radiation treatment. The changes were highly correlated on the group level. This shows potential to use IVIM for monitoring changes in perfusion characteristics during treatment in prostate cancer patients.

OC-0394 Early prediction of tumour changes in NSCLC patients during radiotherapy L.M. Amugongo 1 , E. Vasquez Osorio 1 , A. Green 2 , D. Cobben 3 , M. van Herk 4 , A. Mcwilliam 4

1 University of Manchester, Division of Cancer Sciences, Manchester, United Kingdom; 2 University of Manchester, Division of Cancer Science, manchester, United Kingdom; 3 The Clatterbridge Cancer Centre NHS Foundation Trust, Lung Cancer, Liverpool, United Kingdom; 4 University of Manchester, Division of Cancer Science, Manchester, United Kingdom Purpose or Objective Given the heterogeneity of tumour response to radiotherapy (RT) in lung cancer and the potential of adaptive RT (ART), there is a clinical need to identify cases with large tumour changes early on during treatment. Early identification will improve the feasibility of ART for many patients, enabling planning of beneficial adaptations and allowing allocation of resources. We developed an automatic method to predict tumour volume and shape at week 3 and 4 of RT, using cone-beam computed tomography (CBCT) scans acquired up to week 2, allowing identification of large tumour changes Materials and Methods This retrospective study included 201 advanced NSCLC patients treated with 55 Gy in 20 fractions. CBCTs were rigidly registered to the planning CT (pCT). Intensity values were extracted in each voxel of the PTV across all CBCT images from days 1, 2, 3, 7, and 14 (image guidance protocol). Four regression models were fitted to voxel intensity; applying linear, Gaussian, quadratic, and cubic methods. These models predicted the intensity value for each voxel at the end of weeks 3 and 4, and the tumour volume found by thresholding. Finally, the sensitivity and specificity to predict 30% change in volume were calculated for each model. Results Fig 1 shows a patient CBCT series visualised in the axial, sagittal and coronal planes, predicted using the linear model. Quantitatively, the gross tumour volume has shrunk from 41.79 cm 3 (tumour volume measured on pCT at the start of treatment) to 5.24 cm 3 (at week 3). The predicted tumour volume at the end of week 3 of RT for this patient was 4.65, 15.37, 4.46 and 25.84 cm 3 for the linear, Gaussian, quadratic and cubic models respectively. The sensitivity and specificity for predicting 30% shrinkage at week 3 for all models is provided in Table 1. Overall, the linear model performed best at predicting those patients with large shrinkage that will benefit from RT adaptation. 21% and 23% of patients in our cohort with more than 30% tumour volume reduction were identified at weeks 3 and 4 respectively that could benefit from treatment adaptation.