Presentation (CTI)

OBJECTIVES: Activated phosphoinositide-3-kinase delta syndrome (APDS) is an ultra-rare, complex condition with limited available data, leading to uncertainty for cost-effectiveness modelling and payer decision-making. This study employed a robust methodological framework to collect expert judgement, informing the cost-effectiveness model (CEM) for the NICE evaluation of leniolisib. The study aimed to address broad evidence gaps and uncertainties and validate assumptions underpinning the CEM through a modified structured expert elicitation (SEE) exercise and qualitative and quantitative surveys.
METHODS: Ten APDS clinical experts were recruited from the UK, Europe and Canada. Exercise 1 (Ex1) was a modified SEE exercise that followed best practices of including training slides and an Evidence Dossier. It comprised two surveys which elicited experts’ estimates of APDS manifestation occurrence and survival with standard of care (SoC) and leniolisib, thus deriving hazard ratios and associated uncertainty for the impact of leniolisib vs SoC on manifestation incidence and survival. In Ex2, APDS-treating clinicians completed EQ-5D-5L to generate utilities for APDS health-state vignettes. Ex3 and Ex4 used qualitative and quantitative surveys to validate key assumptions and gather key numerical inputs, respectively, to support the CEM. In Ex5, six APDS UK clinical experts evaluated post-leniolisib discontinuation risks of manifestation recurrence and SoC requirements in response to HTA agency questions.
RESULTS: This study generated robust, expert-driven data to validate assumptions and address evidence gaps in the CEM: Ex1 informed comparative efficacy, Ex2 validated the utilities used and Ex3 and Ex4 refined assumptions and informed resource use estimates. Together, this evidence reduced uncertainty in decision-making and contributed to NICE’s positive recommendation for leniolisib.
CONCLUSIONS: When conventional evidence sources are limited, as in ultra-rare conditions like APDS, expert-informed evidence generation can support HTA. This study demonstrates how multi-method approaches can address evidence gaps, validate model assumptions and inform decision-making, ultimately facilitating patient access in rare conditions.