Radiotherapy quality control in cervical cancer

Resum

Introduction - E-learning programmes allow e ective medical training for geographically dispersed participants. Few studies aim primarily to validate these programmes objectively. Considering its logistical aptitudes, an ODW using FALCON was used to homogenise delineation practice in cervical cancer as quality control among European centres participating in a multicentre prospective trial on locally advanced cervical cancer (LACC). Two identical ODW were performed, to evaluate the delineations of RAIDs centres on a ctitious patient. Aims Primary endpoints: 1. Interobserver validation of the ODW in LACC by using qualitative and quantitative assessments of the improvement of clinician contours between clinicians (interobserver variability). 2. Intraobserver validation of the ODW in LACC by assessing qualitatively and quantitatively the improvement of the clinician contours for each clinician (intraobserver variability). Secondary endpoints: 1. Evaluation of the teaching methodology of the ODW as re ected by the participant satisfaction questionnaires. 2. Analysis of clinician contouring on magnetic resonance imaging (MRI) for European centres which use MRI for brachytherapy planning versus those which do not. Methods - Two ODW included 46 clinicians from 14 RAIDs centres. A technical collaboration was established with European Society for Radiotherapy & Oncology (ESTRO). Training was performed by an expert in the eld, CHM. Through 3 live online presentations, the contouring platform FALCON EduCaseTM, the delineation guidelines and the participant contours were presented. Participants were asked to complete baseline contouring (C1), guideline contouring (C2) and nal contouring (C3) for external beam radiotherapy (EBRT) and brachytherapy (BT) for LACC. Interobserver and intraobserver variability was quantitatively evaluated (DICE index) and analysed by a linear mixed model. Intraobserver variability was qualitatively evaluated by the McNemar test. Results - Nine participants contoured for EBRT and BT for C1 - C3. Thirty-two clinicians submitted any contour. Concerning interobserver quantitative comparisons for EBRT, signi cant improvement was observed for C2 vs. C1 for bowel, nodal elective volume (CTV-node), GTV-p, uterus and vagina, at least 20 mm below GTV-p (CTV-p) and radiologically pathological lymph nodes to boost (GTV-node), versus a signi cant detriment for GTV-node (C3 vs. C1; C2), CTV-p (C3 vs. C2) and bowel (C3 vs. C2), showing overall an improvement in C2 vs. C1, versus a detriment in C3 vs. C2 for two target volumes and one organ at risk. In the BT treatment there was signi cant improvement for C2 vs. C1 for bladder, gross tumor volume (GTV), high risk CTV (HR-CTV) and intermediate risk CTV (IR-CTV), versus a signi cant detriment for bladder (C3 vs. C2), thus a general improvement in C2 vs. C1, with only a detriment in C3 vs. C2 for bladder. As for intraobserver quantitative comparisons, a signi cant improvement was observed for contouring a region of interest between C2 vs. C1, C3 vs. C1 and C3 vs. C2 for EBRT and between C2 vs. C1 for BT. Notable intraobserver qualitative comparisons were found for BT, a signi cant improvement towards the right and posterolateral right directions for target volumes in C2 vs. C1, which became signi cantly worse in these directions in C3 vs. C1 and C2. The average result of the Organisation and Content items in the satisfaction questionnaire for the 20 ODW participants who responded (scale 1-5, 1=poor, 5=excellent) was 4.36 (range 3.95-4.60). With regard to the imaging technique used for BT planning, centres using magnetic resonance imaging (MRI) did significantly better in the BT case for HR-CTV than centres using other techniques (C2 vs. C1: p-value < 0.005; C3 vs. C1: p-value = 0.02). Conclusions - ODW allow to train, initially harmonise contouring and assess the experience of centres before patient inclusion conveniently and eficiently in the context of a multicentre trial.