Real World Issues With Real World Data When Balancing Populations of Cancer Patients Who Have Histories, Treatment Sequences, and Biomarkers

PURPOSE: The utility of comparative-effectiveness evidence can be called in to question when real world data (RWD) cohorts differ in their treatment sequences before and/or after the index line of interest. Panellists will discuss methods available for balancing which can assist in preparing submissions for regulatory and HTA decision-making, providing case studies to illustrate the issues.

DESCRIPTION: Dr. Adamson will open with an audience polling question to assess the variability in audience opinion of the optimal study designs to overcome challenges in balancing real world populations when treatment sequences and biomarkers differ.

Dr Patel will share a recent case study in CAR-T and discuss the challenges of using RWD to support the launch. He will highlight the technical solutions that were difficult to implement and computationally expensive, but critical in the acceptance of RWD i.e., multiple imputation combined with propensity scoring, alongside the conduct of a simulation study to justify the choice of methods.

Dr Vioix will talk about how differences in subsequent lines of therapy can confound comparisons (and thus acceptability to payers), demonstrated in a non-small-cell lung carcinoma (NSCLC) case study. The issues relating to the use of RWD in a biomarker defined population will also be highlighted as adding difficulty to cross-dataset comparisons.

Dr Hatswell brings a focus on methodology, highlighting practical obstacles to the use of RWD – but how these may be (to an extent) be overcome to generate clinically meaningful evidence . These include data access issues, incompatible data classifications, and the uncertain place of indirect comparisons in the research pathway. His focus will be on providing the best possible estimates to inform healthcare decisions.

Dr Adamson will transition the talks in to a discussion by challenging assumptions in the case studies (leaning on her experience with Flatiron) - and inviting audience participation.