HAI Book 2025 - Flipbook - Page 484
Schultz, Stephanie
Improving history-based prediction of biomarker and clinical progression of
autosomal dominant Alzheimer9s disease using mutation-level analysis of
Aβ production
Stephanie Schultz1, Yiwen Rao1, Courtney Maa1, Jean-Pierre Bellier1, Lei Liu1, Dennis Selkoe1,
Eric McDade2, Richard Perrin2, Brian Gordon2, John Morris2, Jorge Llibre-Guerra2, Celeste
Karch2, Tammie Benzinger2, Cahngjie Xiong2, Johannes Levin3, Mathias Jucker11, Takeshi
Ikeuchi5, Nick Fox6, Peter Schofield7, Edward Huey4, Ralph Martins9, Gregory Day8, Pedro
Rosa-Neto10, Randall Bateman2, Jasmeer Chhatwal1
1
Harvard Medical School, Cambridge, MA, US
Washington University in St. Louis, St. Louis, MO, US
3
German Center for Neurodegenerative Diseases, Munich, DE
4
Butler Hospital, Providence, RI, US
5
Brain Research Institute, Niigata University, Niigata, JP
6
Dementia Research Centre, UCL, London, GB
7
Neuroscience Research Australia, Sydney, AU
8
Mayo Clinic, Jacksonville, FL, US
9
Edith Cowan University, Perth, AU
10
McGill University, Montreal, QC, CA
11
University of Tuebingen, Tuebingen, DE
2
Introduction: In autosomal dominant Alzheimer9s disease (ADAD) mutation carriers, age at symptom onset (AAO) is
similar across families, enabling prediction of carriers9 AAO from family history. However, comprehensive family
histories are not available for all carriers. Furthermore, history-derived AAO, though informative, does not fully
capture the clinical and biomarker heterogeneity seen in ADAD. We hypothesized that mutation-level measures of
altered A´ production may augment history-derived AAO in predicting disease course. Leveraging a cell-culture
model that assesses A´ production by individual mutations, we summarized mutation-level A´ production.
Notably, mutation-level A´ production measurements are derived outside the context of family polygenic factors,
age, and demographics, thereby potentially providing unique information regarding ADAD. Here, we examined
whether this A´ production composite can improve the prediction of ADAD clinical and biomarker progression
using data from the Dominantly Inherited Alzheimer9s Network study (DIAN).
Methods: 161 PSEN1 variants were co-transfected with APP into HEK293 cells deficient in PSEN1/2, and A´37/40/42 production was quantified for each variant. The composite measure of A´ production was derived outof-sample and then applied to 56 PSEN1 variants with available clinical and biomarker data from DIAN (Figure 1,
n=190 carriers).
Results: Mutation-level A´ production was strongly associated with PiB-PET, grey-matter volume, and clinical
measures, above and beyond APOE4, age, sex, and history-derived AAO, and improved the prediction of all clinical
and biomarker measures assessed (Tables 1+2). History-derived AAO and the A´ production composite were
independently associated with amyloid-PET and grey-matter, and the inclusion of the A´ production composite
particularly improved the prediction of subcortical pathologic change (Figure 2).
Conclusions: Incorporation of a mutation-level cellular A´ production measure improves prediction regarding the
clinical and biomarker course of ADAD. Incorporating A´ production measures may greatly improve our
understanding of inter-individual heterogeneity in ADAD and improve the design and execution of ADAD clinical
trials.
HAI2025 - 484
