Demonstrating the value of medicines through health technology assessment (HTA) systems is becoming increasingly complex. Innovative therapies – such as immuno-oncology (IO) agents – are testing limits of methodological approaches in markets with established HTA systems. The objective of this study is to understand how requirements, approaches, and decision-making differ between select HTA agencies with a focus on specific PD-1/PD-L1 (programmed death receptor-1/programmed death-ligand 1) agents and cancer indications, and to describe how this variation impacts patient access. To achieve this objective, we conducted a detailed HTA dossier review for several recently launched IO products across Australia (AU), Canada (CA), France (FR), and the United Kingdom (UK).

Content experts reviewed HTA dossiers for pembrolizumab, nivolumab, and atezolizumab for non-small cell lung cancer (NSCLC) first-line monotherapy, NSCLC combination therapy, and adjuvant melanoma. A systematic analytic framework was developed to understand best-practice methodology across systems. Information on submitted data, patient/expert input, and access decisions were extracted; key themes were identified and refined through workshop discussion, and probed further through blinded primary research with eight individuals with current or recent experience of HTA systems.

We identified six major elements of variation impacting decision-making: evidentiary expectations for biomarkers, use/impact of patient-centered data; use/impact of real-world data, acceptance of surrogate endpoints, approaches for clinical data extrapolation, and accepted time horizons. Considerable variation in time to access was observed; for pembrolizumab (NSCLC first-line monotherapy), time from product registration to HTA decision ranged from 42 (CA) to 487 (AU) days; time from registration to listing ranged from 189 (CA) to 605 (AU) days.

Evaluated HTA systems demonstrate a large degree of variability in approaches to decision-making for novel IO medicines; resultant access decisions and time to access are also highly variable. Inconsistency between systems and duplication of effort when assessing similar clinical/economic data could be contributing to limited or delayed patient access; the relationship merits further exploration. Assessed HTA systems are currently undergoing process revisions but expert input suggests that this is not expected to reduce variation, and could further increase complexity. The influence of parallel scientific advice programs between HTA agencies and regulatory bodies in reducing variation must also be determined.