What does a Centiloid score mean for individual prognosis after a positive amyloid PET?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
A Centiloid (CL) score is a standardized, continuous measure of brain amyloid-β burden derived from PET scans, anchored so that 0 CL represents young controls and 100 CL represents typical mild‑AD dementia levels [1] [2]. For an individual with a positive amyloid PET, the absolute CL value conveys where that person sits on the continuum from no amyloid to high amyloid and correlates with group-level risk of future decline, but prognostic certainty for any single patient depends on thresholds, local variability, and other biomarkers [1] [3] [4].
1. What the number means: an anchored continuum, not a prophecy
A Centiloid score quantifies amyloid plaque burden on a single, harmonized scale so values can be compared across tracers and centers—0 CL is average young amyloid‑negative controls and 100 CL is average mild‑AD dementia—but it is a measure of pathology, not a direct deterministic prediction of future clinical course [1] [2]. Studies show CL increases with clinical stage (subjective cognitive decline < MCI < dementia), so higher CL correlates with higher group risk of progression [5].
2. Practical thresholds used in prognosis and decision‑making
Researchers and consortia propose ranges rather than a single magic number: AMYPAD recommends CL <10 as confidently amyloid‑negative, 10–30 as an intermediate/evolving range, 30–60 as established amyloid pathology, and >60 as high amyloid burden often associated with tau positivity [3]. Other studies set positivity thresholds around 12–30 CL for scan positivity and have identified specific cut points—e.g., 26 CL was optimal to predict progression to dementia over six years in a memory‑clinic cohort [6] [7] [8].
3. Prognostic strength and limits in individual patients
Centiloids improve risk stratification at the group level—higher CL associates with greater likelihood of cognitive decline or neuropathologic evidence of AD—but thresholds vary by outcome and study: neuropathology comparisons produced optimal thresholds ranging roughly from ~20 CL to >45 CL depending on the metric and post‑mortem adjustment [9] [10] [11]. Importantly, quantitative CL values near common cutoffs carry uncertainty: methodological work found 95% confidence intervals around cutoffs (for example ±3.95 CL around 24 CL), meaning values close to thresholds should be interpreted cautiously for an individual [4] [12].
4. Sources of variability that affect individual prognosis
Centiloid values can be influenced by tracer, acquisition and processing pipeline, reference region choice, brain atrophy, and site harmonization; stress‑testing work created dozens of pipelines showing measurable variability and recommending the standard whole‑cerebellum pipeline for robustness [4]. Clinical samples, sample composition, and statistical choices also explain the spread of proposed cutoffs across studies, so a single CL value may map differently onto neuropathology or clinical risk depending on context [12] [13].
5. How clinicians and trials use Centiloids—and the implicit agendas
In trials and some clinical settings Centiloids are used for patient selection, confirming amyloid for anti‑amyloid therapy eligibility, and monitoring amyloid reduction after treatment, which incentivizes precise quantification [14] [3]. That utility creates an implicit agenda toward standardizing cutoffs for access to disease‑modifying drugs, but regulatory or reimbursement thresholds may not fully reflect the measurement uncertainty or the need to integrate tau, CSF, MRI, genetics and clinical profile for individual prognosis [14] [13].
6. Bottom line for an individual after a positive amyloid PET
A positive amyloid PET with a Centiloid well above the established positive range (e.g., >30 CL and especially >60 CL) indicates substantial amyloid pathology and higher group‑level risk of progression, whereas values in the 10–30 CL “intermediate” band or within a few CL of common cutoffs require caution because of measurement variability and contextual modifiers; therefore CL should be combined with clinical status and other biomarkers when estimating an individual prognosis [3] [4] [12]. If a patient’s CL falls near a threshold, expert review, repeat imaging or complementary biomarkers (tau PET, CSF, MRI) are the appropriate next steps rather than relying on the single number alone [5] [13].