Centiloid cut-off values for optimal agreement between PET and CSF core AD biomarkers

Area of Science:

• Neurology
• Radiology
• Biomarker Discovery

Background:

• The Centiloid scale standardizes amyloid PET imaging measurements.
• Reference cut-offs are crucial for multicenter research and clinical trials in Alzheimer's disease (AD).
• This study aimed to derive Centiloid thresholds that agree with core AD cerebrospinal fluid (CSF) biomarkers.

Purpose of the Study:

• To establish reliable Centiloid cut-offs for amyloid PET imaging.
• To maximize agreement between Centiloid values and established CSF AD biomarkers.
• To validate these cut-offs across independent cohorts.

Main Methods:

• 516 participants from ALFA+ Study and ADNI underwent amyloid PET scans.
• CSF biomarkers (Aβ42, tTau, pTau) were measured using Elecsys® tests.
• Receiver operating characteristic (ROC) analysis identified Centiloid cut-offs maximizing agreement with CSF biomarkers.

Main Results:

• Optimal Centiloid cut-offs ranged from 25-35, except for CSF Aβ42 at 12 CL.
• Tau/Aβ42 ratios showed higher agreement with PET than Aβ42 alone.
• Cut-off robustness was confirmed in independent cohorts and with varying CSF cut-offs.

Conclusions:

• A Centiloid cut-off of 12 CL aligns with previous findings for early AD pathology.
• Two key inflection points identified: ~12 CL for subtle pathology and ~30 CL for established pathology.
• Deriving robust, generalizable cut-offs requires cohorts with diverse intermediate pathology levels.