Background: Quality assurance (QA) in randomised controlled trials is essential to ensure reliability and accuracy of the full trial dataset.
Patients with melanoma brain metastases (MBM) have high rates of morbidity and mortality. A randomised phase III trial (NCT01503827) has been in progress to assess whole brain radiotherapy (WBRT) following local treatment of MBM. Whole Brain Radiotherapy in Melanoma Trial (ANZMTG 01.07) is an investigator initiated trial coordinated by Australia and New Zealand Melanoma Trials Group (ANZMTG) with funding from Cancer Australia Priority-driven Collaborative Cancer Research Scheme (Project ID Number 1084046). This study assesses the outcomes of QA activities conducted for the WBRTMel study.
Methods: The WBRTMel trial’s primary objective is to compare distant in-brain failure of patients with 1-3 locally treated MBM randomised to either WBRT or observation. Recruitment was conducted from 30 sites in 3 countries between 2007 and 2017, achieving accrual of 215 participants in total. QA activities conducted for this trial were coordinated centrally by Australia and New Zealand Melanoma Trials Group (ANZMTG). QA activities included review of 100% of all data received centrally. Further QA occurred including 100% review of histopathology reports to ensure confirmed melanoma in all cases, and radiological review of a proportion of all Magnetic Resonance Imaging (MRI) scans from each centre (the first five and 20% of the subsequent MRI scans). Quality of Life (QOL) and health economics data have been reviewed. Neurocognitive data were reviewed and scored by three experts. The first five from each site, then 20% of Radiotherapy treatment plans have been reviewed by Trans-Tasman Radiation Oncology Group (TROG).
Results: Radiotherapy post-treatment reviews reported 1 major (0.2%) and 3 minor (0.6%) variations to protocol-defined treatment, showing a high level of protocol compliance. Hippocampal-sparing radiotherapy has shown good compliance with the landmark RTOG0933 constraints, with 100% of patients having compliance for at least one hippocampus and 82% of patients with compliance for both hippocampi. Pathology report review was completed by a central pathologist and shows high levels of concordance. QOL, Health economic, and neurocognitive data show a high level of concordance between patient responses and entered data, with high completion rates (>90%) for patients participating in each component of the protocol.
Conclusions: ANZMTG processes and the involvement of specialist groups such as Cooperative Trials Group for Neuro-oncology (COGNO) and TROG have ensured a high level of quality and complete data in this trial. Additionally, a multidisciplinary Trial Management Committee has provided ongoing consultation to support the protocol to oversee data quality and ensure the reliability and accuracy of the dataset, in preparation for analysis and presentation of the primary endpoint in 2019.