OBJECTIVES: Delayed diagnosis of Alzheimer's Disease (AD) impedes early intervention and complicates research into disease modifying treatments (DMTs). Polygenic risk scores (PRS) integrate multiple genetic variants associated with AD in a single score. Screening with PRS could enhance the likelihood of early AD diagnosis in patients with mild cognitive impairment (MCI). This study aimed to evaluate the cost-effectiveness of PRS integration in the diagnostic pathway (1) on identified cases and (2) in case a DMT becomes available and served as a proof-of-concept for early health economic evaluation of PRS as a screening tool.

METHODS: An open-source Markov model, simulating AD progression, was adapted and linked to a de-novo decision-tree model for the MCI diagnostic pathway. Flemish 65+ individuals with subjective cognitive impairment could undergo diagnosis-as-usual (DAU) or PRS assessment. The decision tree assessed the incremental cost per identified MCI case in Flanders. The endpoints were then used in the Markov model to evaluate incremental costs per quality-adjusted life year (QALY) gained if a DMT is available. A healthcare payer perspective was considered. Scenarios were validated with experts and leveraged relevant data sources.

RESULTS: In the base case PRS assessment initiated the diagnostic pathway, resulting in 304 additional MCI cases identified at an extra cost of 31.3 million € (i.e., €101 984 per identified case). Should a DMT for AD become available, the ICER is €284 061 per QALY gained.

CONCLUSIONS: DAU involves multiple steps, identifying relatively few MCI patients. The surplus of identified cases with PRS initiating the diagnostic pathway is small compared to the screening cost for the total group, making it currently unlikely to be cost-effective. The cost-effectiveness of PRS-screening heavily depends on the overall MCI diagnostic pathway and DMT availability. This study demonstrates the feasibility of an early health-economic evaluation of PRS as a screening tool for AD.