ESTRO 38 Abstract book

S437 ESTRO 38

Material and Methods This study included a cohort of 494 patients treated primarily with radiotherapy between 1978 and 2015 with a curative intent. We analyzed five clinical variables, i.e. FIGO stage (IB - IIB), age, tumor diameter (<4cm and > 4cm), histologic type (squamous or adenocarcinoma), treatment (Radiotherapy (XRT), XRT and Chemotherapy, XRT and Hyperthermia). The outcome of interest is OS defined as the time between first treatment and death or last follow-up. Univariate analyses consists of Kaplan- Meier plots and log-rank test. The final multivariate Cox proportional hazard regression model used to build the nomogram included only variables with a statistical significance (p<0.05). The nomogram’s performance was assessed by comparing the predicted values of the surviving and non-surviving patients to the observed values in a bootstrapped-internal calibration plots and the concordance index (C-index) where a C-index = 1 indicates a perfect prediction and C-index = 0.5 is comparable to a random guess. Results We excluded 64 patients with incomplete data reducing the cohort to 430 patients. The mean age in this cohort was 60 (29 - 93) years; Table 1 shows the variable distribution for this study cohort, including the univariate and multivariate statistical significance. The 5 and 10-year overall survival rate since first treatment was approximately 82% and 80% with a median survival period of 90 and 101 months respectively. The model had a C- index of 0.64 with a standard error of 0.024. Figure 1 shows the nomogram representation of the model and the bootstrapped calibration plot. The calibration plot shows an agreement between the models’ predicted values and the observed values since most of the points are close to the diagonal line, which indicates perfect prediction.

Conclusion We developed and internally validated a nomogram to predict 5 and 10 - year overall survival for patients with cervical cancer who were treated with radiotherapy, optionally combined with hyperthermia or chemotherapy.

Poster: Clinical track: Prostate

PO-0834 Virtual imaging for patient information on radiotherapy planning and delivery for prostate cancer. J. Sulé-Suso 1 , J. Bisson 1 , S. Jassal 1 , M. Martínez 2 , N. Huxley 1 , C. Ellis 1 , D. Chambers 1 , K. Fields 1 , C. O'Donovan 1 , C. Edwards 2 , S. Vengalil 1 , R. Bhana 1 1 University Hospital of North Midlands, Oncology Department, Stoke on Trent, United Kingdom ; 2 University Hospital of North Midlands, Physics- Oncology department, Stoke on Trent, United Kingdom Purpose or Objective To assess whether provision of information on RT planning and delivery with a virtual reality (VR) system (VERT) improves, apart from patient’s satisfaction 1 , patient’s compliance to RT (holding better their water prior to each RT session), and reduces side effects. Material and Methods This was a randomised study where patients were allocated to group 1 (information on RT planning and delivery using the VERT system was given prior to starting RT) or group 2 (information on RT planning and delivery using the VERT system was given after the last day of RT). Ninety-two patients with prostate cancer receiving radical RT were included in the study. The study was approved by the local ethics committee. For each study patient, their planning CT Scan images and RT plan were uploaded onto the VERT system using the Digital Imaging and Communication in Medicine (DICOM) standard. Patients and relatives were shown using VERT and on a one-to-one basis with a radiographer, a standard room where RT is given, a linear accelerator, and how RT is planned and delivered using their own planning CT Scans. Emphasis was put on the area to be treated and the organs around it. Bladder volumes were calculated from the planning CT Scan (prior to RT) and then on days 1, 2, 3 and then weekly