127MO A practical toolkit with recommendations for analysing and visualising patient-reported outcomes in early phase dose-finding oncology trials: OPTIMISE-AR

Background
Patient-reported outcomes (PROs) are increasingly recognised for their role in tolerability assessment within dose-finding oncology trials (DFOTs). However, PRO analysis and reporting within DFOTs is often unclear and inconsistent. To address this challenge, the OPTIMISE-AR project (incOrporating PaTIent-reported outcoMes In doSE-finding trials–Analysis Recommendations) was established to develop toolkits to support the statistical analysis and visualisation of PRO data within DFOT publications.

Methods
International, cross-sector statistical analysis and data visualisation working groups, comprising 17 multidisciplinary experts were established to identify analytical approaches for PRO data that address key PRO research objectives for DFOTs as per established OPTIMISE-Research Objective Recommendations (OPTIMISE-ROR). Drawing on evidenced-based approaches for PROs and recommendations for reporting tolerability data in other trial settings, the working groups evaluated established statistical analysis and data visualisation methods, subsequently adapting them to ensure suitability for the DFOT context.

Results
Analysis recommendations are provided for ordinal, continuous, and binary PRO data that inform PRO endpoints addressing key PRO objectives in DFOTs, including tolerability assessment across dose levels and timepoints as well as the integration of PROs within formal interim and final decision-making. Approaches are informed by existing literature. Statistical analysis methods are presented alongside contextual strengths and limitations for the DFOT setting to support trialists in selecting methods aligned with their trial objectives, and exemplar data visualisations facilitate clear interpretation of PRO data. The toolkit also outlines how PROs can contribute to statistical inference for recommended phase II dose selection.

Conclusions
As PROs play an ever-increasing role in tolerability assessment, this work promotes appropriate analysis and data visualisation of PRO data, facilitating robust, patient-centred tolerability conclusions and supporting the broader development of tolerable and effective treatments.