# Optimizing PET spatial extent measures to detect the earliest amyloid-beta and tau accumulation and associated cognitive decline in preclinical Alzheimer's disease

> **NIH NIH K01** · MASSACHUSETTS GENERAL HOSPITAL · 2024 · $128,520

## Abstract

Successful prevention of Alzheimer’s disease (AD) dementia may rely on intervention at the earliest possible
point in the AD pathological cascade, hypothesized to be the accumulation of amyloid-beta (Aβ). The standard
PET approach for Aβ detection is based on widespread cortical Aβ burden and often fails to capture early Aβ.
The proposed K01 project will develop optimized spatial extent measures that allow for robust identification of
individuals in an earlier stage of amyloidosis than previously studied. This approach should provide a more
dynamic biomarker of A localization for investigating early A‘s role in the AD pathological cascade and has
the potential to serve as a more sensitive outcome in future prevention trials. Aim 1 will use longitudinal data
from clinically normal (CN) individuals from the Harvard Aging Brain Study (HABS) to 1) develop an optimized
spatial extent metric based on a scientifically rigorous investigation of how different possible methods for
computing spatial extent respond to simulated noise, 2) validate the sensitivity and specificity of the optimal
spatial extent approach (EXT) to predict future A accumulation over 3-11 years, and 3) implement EXT in an
independent sample of standard PET- screen fails from the AHEAD 3-45 Study to assess the alignment
between plasma Aβ+ and EXT+ for detecting the earliest Aβ deposits. Aim 2 will use EXT to evaluate the
earliest changes in tau using plasma biomarkers (ptau217) and flortaucipir-PET and assess whether EXT may
provide a better predictive marker of those at risk for tau proliferation than standard cortical Aβ-PET
approaches. Aim 3 will evaluate whether the improved quantification of the earliest A deposits with EXT
allows for more sensitive detection of long-term and immediate cognitive changes associated with emerging
A. These findings will help provide a framework for future prevention trials to intervene earlier in the disease
process than has previously been attempted. Furthermore, the EXT approach will open a wide range of
potential future directions to study how emerging Aβ relates to other important factors in AD pathogenesis (i.e.
inflammation, synaptic function) to establish an independent research program. To help Dr. Michelle Farrell
achieve these aims, a multidisciplinary mentorship team has been assembled from the Harvard Medical School
community to complement didactic coursework in PET physics and pharmacokinetics, plasma biomarkers,
advanced statistical analysis, and clinical trials. Dr. Reisa Sperling will serve as the primary mentor overseeing
research and career progress and providing training in clinical trial design and conduct to maximize the
translational potential of the planned research. Dr. Keith Johnson and Dr. Julie Price will serve as co-mentors
to provide comprehensive training in PET research. The mentorship team will be rounded out by two research
advisors (Dr. Rob Rissman: plasma biomarkers, Dr. Brian Healy: statistics) and two contr...

## Key facts

- **NIH application ID:** 10914233
- **Project number:** 5K01AG083062-02
- **Recipient organization:** MASSACHUSETTS GENERAL HOSPITAL
- **Principal Investigator:** Michelle Elizabeth Farrell
- **Activity code:** K01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $128,520
- **Award type:** 5
- **Project period:** 2023-09-01 → 2028-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10914233

## Citation

> US National Institutes of Health, RePORTER application 10914233, Optimizing PET spatial extent measures to detect the earliest amyloid-beta and tau accumulation and associated cognitive decline in preclinical Alzheimer's disease (5K01AG083062-02). Retrieved via AI Analytics 2026-05-25 from https://api.ai-analytics.org/grant/nih/10914233. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*