CODAGLIO 2.0: Combining clinical trial datasets in glioma patients to answer clinically relevant questions about patients’ health related quality of life

The CODAGLIO 2.0 project, which started in October 2022, aims to update the original CODAGLIO database to a version 2.0 by adding new randomized controlled trials (RCTs) conducted in glioma (i.e., primary brain tumor) patients as well as the addition of clinical variables, including tumor molecular data and physician-reported adverse events data, to already existing trials in the original database. The project is a collaboration between the Quality of Life Group, the Brain Tumor Group, and the Quality of Life Department. We have included 17 RCTs conducted in about 5931 patients in the database. With this large database, we will be able to answer five clinically relevant questions, such as: ‘1) Is there a difference in HRQoL patterns between glioma patients with different molecular profiles (WHO 2021 criteria)?’, ‘2) Are HRQoL patterns in glioma patients related to sociodemographic characteristics (e.g., age or sex)?’, ‘3) Can we identify symptom(s) (clusters) that are predictive of radiological tumor progression in glioma patients?’, ‘4) Are physician-reported adverse events translated into a deterioration in HRQoL over time in glioma?’, and ‘5) Can HRQoL during the course of disease be accurately predicted using machine-learning models?’.

• Since the start of the project, we have worked on obtaining approval from the principal investigators or pharmaceutical companies to include the data collected in clinical trials in our database
• We also finalized the data sharing process as well as setting up the statistical analysis plans for all five clinically relevant questions
• We have completed harmonization of different RCT datasets in the CODAGLIO database as well as the data cleaning process
• Furthermore, we have completed a detailed and coherent data analysis for research questions 1, 2, and 3.
• Finally, the draft manuscripts for research questions 1, 2, and 3 have been finished.
• While data analysis for research question 4 is ongoing.

The manuscript for research question 1 has been published by The Oncologist. The manuscript for research question 2 is under peer review by the journal Neuro-Oncology. The manuscript for research question 3 is under co-authors’ review. The researcher is currently performing a data analysis for research question 4.

The thesis introduction is under review by supervisors

In the remaining time of the project, we will finalize statistical analyses for questions 4. Write a thesis general discussion. However, further design methodological evaluation is needed regarding the feasibility of research question 5 in the light of the current available dataset.

Patients with a glioma, the most prevalent primary malignant brain tumor, demonstrate a high symptom burden including motor weakness, cognitive dysfunction, seizures, etc. Although patients are treated with surgery, radiotherapy, and chemotherapy, current treatment options are not curative. Therefore, the quality of survival is for these patients at least as important as the duration of survival. Collection of enough quality of survival data is crucial but challenging in patients with glioma, as glioma is a relatively rare disease, only a few trials included health-related quality of life (HRQOL) assessments, and patients’ compliance rates rapidly decline over time as a result of tumor progression, neurocognitive impairments, and treatment toxicity. To increase statistical power, combining different datasets is important in order to answer clinically relevant questions. For example, we will identify multiple concurrent symptoms that will be predictive of tumor progression. Also, we will use different models to assess clinical and sociodemographic variables that are related to worse HRQoL over time. Such information may help physicians in the early identification of subgroups of glioma patients prone to worse HRQoL as well as be relevant in clinical decision-making over offering timely and effective personalized supportive care. Effective supportive care strategies may eventually result in an improved HRQOL.