Box 5 | Expert opinion on prospective biomarkers

• Non-invasive and easy-access biological setups remain as potential proxies for blood-based biomarkers in the detection of cognitive impairment. Future studies should explore whether a combination of non-invasive biomarkers is superior to single biomarkers in the early detection of AD

• Digital biomarker technologies have a role in improving access to AD screening, diagnosis, and monitoring of disease progression. They can be used at home and in tandem with other physical activity tracking or monitoring devices. While cost effective, they have a high logistical burden in that they would need to be validated, standardized, and adapted to various language and cultural requirements

• AI-supported continuous data and algorithms have value in monitoring disease progression and helping to predict AD dementia. The post-processing of imaging data and quantitative aspect of AI may help to identify subtle changes and atrophies in the brain that are typical of AD

• Novel biomarkers for neuroinflammation may be useful for the differential diagnosis and staging of AD, but their utility in clinical practice remains inconclusive

• The APOE ɛ4 genetic risk test, in combination with other blood-based biomarkers, may have value in the screening of AD and in predicting treatment safety with DMTs

• Although polygenic risk score models may be promising in the detection of AD, different risk scores may need to be developed for different ethnic groups in the region

• EEG with special analysis algorithms may be useful for screening, diagnosis, and progression tracking in AD, but there is currently no recommendation for its routine use in clinical practice

• The clinical applicability of advanced imaging modalities remains limited in the Asian region because of prohibitive costs and complex follow-up assessments

AD, Alzheimer’s disease; AI, artificial intelligence; APOE ɛ4, apolipoprotein E ɛ4 allele; DMT, disease-modifying therapy; EEG, electroencephalography