Centiloid recommendations for clinical context-of-use from the AMYPAD consortium

Abstract

Amyloid-PET quantification through the tracer-independent Centiloid (CL) scale has emerged as an essential tool for the accurate measurement of amyloid-β (Aβ) pathology in Alzheimer's disease (AD) patients. The AMYPAD consortium set out to integrate existing literature and recent work from the consortium to provide clinical context-of-use recommendations for the CL scale. Compared to histopathology, visual reads, and cerebrospinal fluid, CL quantification accurately reflects the amount of AD pathology. With high certainty, a CL value below 10 excludes the presence of Aβ pathology, while a value above 30 corresponds well with pathological amounts. Values falling in between these two cutoffs ("intermediate range") are related to an increased risk of disease progression. Together, CL quantification is a valuable adjunct to visual assessments of amyloid-PET images. An abnormal amyloid biomarker assessment is a key criterion to determine eligibility for anti-amyloid disease-modifying therapies, and amyloid-PET quantification can add further value by precisely monitoring amyloid clearance, and hence guiding patient management decisions. HIGHLIGHTS: Centiloid (CL) quantification robustly reflects of the amount of Aβ pathology. CL < 10/CL > 30 reflects Aβ-negativity/positivity thresholds with high certainty. CL quantification is a valuable adjunct to visual assessments of amyloid-PET. CL quantification can support trial design and treatment management. CL quantification could support the identification of early or emerging Aβ pathology.

Keywords: Amyloid‐β; Centiloid quantification; clinical trials; positron emission tomography.

FIGURE 1

Overview of Centiloid implementation and available software in research and clinical settings. PiB, Pittsburgh compound B.

FIGURE 2

Centiloid publications in PubMed (search performed on the 01/01/24).

FIGURE 3

AMYPAD participants categorized by cognitive stage and baseline CL. Bar plots illustrate the distribution of CL groups across two studies, (A) DPMS and (B) PNHS, from the AMYPAD consortium. Subjects were categorized into four groups based on their baseline amyloid burden: <10 CL (no amyloid pathology), 10 to 30 CL (“intermediate range,” evolving amyloid pathology), 30 to 60 CL (established amyloid pathology with increased prevalence of tau positivity), and >60 CL (established amyloid pathology with high certainty of tau‐PET positivity). AMYPAD DPMS and PNHS datasets are publicly available (https://amypad.eu/data/). AMYPAD, Amyloid Imaging to Prevent Alzheimer's Disease; CDR, Clinical Dementia Rating; CL, Centiloid; DPMS, Diagnostic and Patient Management Study; MCI, mild cognitive impairment; PNHS, Prognostic and Natural History Study; SCD+, subjective cognitive decline plus.

FIGURE 4

Overview of Centiloid scale interpretation. An illustrative figure summarizing the main studies regarding CL‐based cut points to reliability exclude (<10 CL) and include (>30 CL) Aβ‐pathology at the individual level. In addition, the 95% confidence interval across the scale is provided and detailed cutoffs in the “intermediate range” are highlighted. Aβ, amyloid‐beta; CL, Centiloid; CI, confidence interval; CSF, cerebrospinal fluid; GAAIN, Global Alzheimer's Association Interactive Network; PET, positron emission tomography; SPM, Statistical Parametric Mapping.

FIGURE 5

Amyloid‐β removal profiles for Phase‐III trials of aducanumab, donanemab, gantenerumab, and lecanemab as measured by Centiloid (CL). Sample sizes in the treatment arms varied for the individual trials and were at the last visit: N = 614 for donanemab, N = 210 for lecanemab, N = 50 (Graduate 1) and N = 41 (Graduate 2) for gantenerumab, and N = 109 (Emerge high‐dose) and N = 112 (Engage high‐dose) for aducanumab. Data points as reported in Sims et al. (2023), van Dyck et al. (2022), Bateman et al. (2023), and Budd Haeberlein et al. (2022). Placebo groups are not plotted and showed constant or slightly increasing amyloid levels over time, as expected. The dotted line represents 24.1 CL, the cutoff for amyloid‐negativity as defined within the GRADUATE 1 and 2 trials and implemented in the head‐to‐head studies of donanemab versus Aduhelm.