Amyloid PET Quantitation and Centiloid Thresholds in the Diagnosis of Alzheimer Disease: An Individual Participant Data Meta-Analysis.

JAMAOct 6, 2026

Ganna Blazhenets, David N Soleimani-Meigooni, Konstantinos Chiotis et al.

✦ AI-curated · Sources linked

In 30 seconds

This individual participant data meta-analysis examined amyloid PET scans from 49,227 participants across 53 studies to establish Centiloid positivity cutoffs for Alzheimer disease diagnosis. The analysis identified a single cutoff of 18 Centiloids for positivity and a double-cutoff approach indicating high confidence for negative scans below 11 Centiloids and positive scans above 26 Centiloids.

Key findings

  • A single cutoff for positivity was established at 18 Centiloids (95% CI, 16-19).
  • Negative scans were confidently interpreted below 11 Centiloids (95% CI, 9-13).
  • Positive scans were confidently interpreted above 26 Centiloids (95% CI, 24-28).
  • Centiloids showed high predictive value for visual positivity with a Cohen κ of 0.86 (95% CI, 0.83-0.89).

Why it matters

Establishing standardized Centiloid cutoffs enhances the accuracy of amyloid PET interpretations, which is crucial for diagnosing Alzheimer disease and determining eligibility for amyloid-targeting therapies.

Source

Published in JAMA. This summary was written by xxcode from the publication's abstract and metadata. It is not peer reviewed and is not a substitute for the original article. For clinical decisions, review the original publication.

ShareTelegramLinkedIn

AI-generated summaries may contain errors or omissions. Verify clinically important information with the original publication.

More in Neurology

Want this personalized?

Stop searching the literature. Choose what you follow and xxcode will build your personalized medical digest.

  • Your specialties
  • Your filters and thresholds
  • Clinical Research + Case Reports, tuned separately
  • Automatic weekly delivery
  • Audio and text

Prefer listening? Personalized audio digests are available with Pro.

Stay updated for free

Get the 3 most interesting publications in one specialty each week.

Weekly email. Unsubscribe anytime.