Comparison of investigator-delineated gross tumour volumes and quality assurance in pancreatic cancer: Analysis of the on-trial cases for the SCALOP trial.

Abstract

We performed a retrospective central review of tumour outlines in patients undergoing radiotherapy in the SCALOP trial. The planning CT scans were reviewed retrospectively by a central review team, and the accuracy of investigators' GTV (iGTV) and PTV (iPTV) was compared to the trials team-defined gold standard (gsGTV and gsPTV) using the Jaccard Conformity Index (JCI) and Geographical Miss Index (GMI). The prognostic value of JCI and GMI was also assessed. The RT plans were also reviewed against protocol-defined constraints. 60 patients with diagnostic-quality planning scans were included. The median whole volume JCI for GTV was 0.64 (IQR: 0.43-0.82), and the median GMI was 0.11 (IQR: 0.05-0.22). For PTVs, the median JCI and GMI were 0.80 (IQR: 0.71-0.88) and 0.04 (IQR: 0.02-0.12) respectively. Tumour was completely missed in 1 patient, and⩾50% of the tumour was missed in 3. Patients with JCI for GTV⩾0.7 had 7.12 (95% CIs: 1.83-27.67, p=0.005) higher odds of progressing by 9months in multivariate analysis. Major deviations in RT planning were noted in 4.5% of cases. Radiotherapy workshops and real-time central review of contours are required in RT trials of pancreatic cancer.