Every Centiloid, from Everywhere, All at Once

Abstract Background Ten years after the original publication, the Centiloid framework is now broadly used to harmonize amyloid‐PET quantification, facilitate data sharing and comparison across cohorts, and even assist visual interpretation in clinical settings. We evaluated the global implementation of Centiloids by comparing their distribution and corresponding positivity thresholds across cohorts. Methods We gathered data from publicly available cohorts and reached out to investigators across the world to collect cross‐sectional Centiloids, demographic and clinical information, and visual read data. Gaussian mixture models (GMM, k=2) were fitted to Centiloid values for each cohort and cutoffs were calculated as mean + 2SD of the lower Gaussian. When visual reads were available, we determined Centiloid cutoffs that maximized correspondence with visual reads (Cohen's kappa). Data was combined across cohorts using random effects meta‐analyses. Results As of January 2025, we included 37 cohorts ( n = 41,678 participants) with heterogeneous pipelines, radiotracers, and clinical and demographic characteristics (Table‐1). The low Gaussian peaks ranged from ‐9 to 10CL; the meta‐analysis identified a common peak at 1CL. The second peaks were more heterogeneous (range=38‐102CL; meta‐analysis outcome=64CL). Across cohorts, the proportion of cognitively unimpaired versus impaired participants impacted the position of both peaks, with better separation in cohorts enriched in impaired individuals (Figure 1C). The meta‐analysis indicated a GMM‐based cutoff of 19CL (95%CI: 16‐21CL, Figure 1B); subgroup analyses showed no evidence of significant effect between single versus multicenter settings (17 versus 20CL, p = 0.30), MRI‐based or PET‐only processing (18 versus 19CL, p = 0.88), and no evidence of difference across radiotracers (Flutemetamol: 16CL; PIB: 17CL, Flutafuranol: 18CL, Florbetaben: 19CL, Florbetapir: 20CL, p = 0.78). In a subset of 29,496 participants with visual reads available, binary visual reads corresponded well to Centiloids (common kappa=0.86, Figure 2A). The visual read‐based cutoff of 24CL (95%CI: 21‐27CL, Figure 2B) maximized correspondence between visual read and quantification and was slightly higher than the GMM‐based cutoff. All meta‐analysis models showed high non‐random heterogeneity (I 2 >80%) across studies, suggesting non‐random differences in peaks and cutoffs. Conclusions Meta‐analysis‐based cutoffs align well with thresholds from the existing literature. High heterogeneity among studies underscores the need to investigate contributing factors, raising concerns about applying common cutoffs.

Authors

Institutions

Publication Details

Journal
Alzheimer s & Dementia
Published
2025-12-01
DOI
https://doi.org/10.1002/alz70856_107124
Primary Topic
Dementia and Cognitive Impairment Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Every Centiloid, from Everywhere, All at Once

Val J. Lowe, Heather M. Snyder, Xiaoxie Mao, D. Vaillancourt et al.
Alzheimer s & Dementia
Dementia and Cognitive Impairment Research
article

Every Centiloid, from Everywhere, All at Once

Val J. Lowe, Heather M. Snyder, Xiaoxie Mao, D. Vaillancourt, Theresa Harrison, Yuna Gu, Valentina Garibotto, Rikio Katou, William Coath, Cally Xiao, Débora Elisa Peretti, Chia‐Hsuin Chang, Lise Colmant, Ganna Blazhenets, Gill Farrar, Pan Sun, Christina B. Young, K Ishii, Wei‐En Wang, David Soleimani‐Meigooni, Hsin‐I. Chang, Mahnaz Shekari, M R Yun, Oskar H. Hansson, Juan M. Fortea, Shu‐Hua Huang, Sang Won Seo, María Franquesa‐Mullerat, Tengfei Guo, Yeojun Yoon, Yishu Chao, David M Cash, Elizabeth C. Mormino, Sangwon Lee, Sid E. O'Bryant, Florence Mézenge, Liana G. Apostolova, Andrzej Sokolowski, Qi Huang, Colin L. Masters, William J. Jagust, Lisa Quenon, Wiesje M. van der Flier, Christopher G Schwarz, Robert A. Koeppe, Jonathan M Schott, Akinori Nakamura, Bruno Dubois, Bernard J Hanseeuw, AMYPAD and LEADS Consortia, Maria C. Carrillo, Lyduine E. Collij, Fang Xie, Susan M. Landau, Raquel C Gardner, Brian J Lopresti, Breton M. Asken, Pierrick Bourgeat, Stéphanie Bombois, Alfonso Fajardo Valdez, Gemma Salvadó, Brigitte Landeau, Meredith N. Braskie, Kaitlin B Casaletto, Ruben Smith, Christopher C. Rowe, Arthur W. Toga, Giovanni B Frisoni, Andrew March, Michael S. W. Weiner, Joel H Kramer, Gael Chételat, Ioannis Pappas, Reisa A. Sperling, Gil D. Rabinovici, Renaud La Joie, Victor L. Villemagne, Brad C. Dickerson, Daniel Schwartz, Elsmarieke van de Giessen, Prashanthi Vemuri
article en

Abstract

Abstract Background Ten years after the original publication, the Centiloid framework is now broadly used to harmonize amyloid‐PET quantification, facilitate data sharing and comparison across cohorts, and even assist visual interpretation in clinical settings. We evaluated the global implementation of Centiloids by comparing their distribution and corresponding positivity thresholds across cohorts. Methods We gathered data from publicly available cohorts and reached out to investigators across the world to collect cross‐sectional Centiloids, demographic and clinical information, and visual read data. Gaussian mixture models (GMM, k=2) were fitted to Centiloid values for each cohort and cutoffs were calculated as mean + 2SD of the lower Gaussian. When visual reads were available, we determined Centiloid cutoffs that maximized correspondence with visual reads (Cohen's kappa). Data was combined across cohorts using random effects meta‐analyses. Results As of January 2025, we included 37 cohorts ( n = 41,678 participants) with heterogeneous pipelines, radiotracers, and clinical and demographic characteristics (Table‐1). The low Gaussian peaks ranged from ‐9 to 10CL; the meta‐analysis identified a common peak at 1CL. The second peaks were more heterogeneous (range=38‐102CL; meta‐analysis outcome=64CL). Across cohorts, the proportion of cognitively unimpaired versus impaired participants impacted the position of both peaks, with better separation in cohorts enriched in impaired individuals (Figure 1C). The meta‐analysis indicated a GMM‐based cutoff of 19CL (95%CI: 16‐21CL, Figure 1B); subgroup analyses showed no evidence of significant effect between single versus multicenter settings (17 versus 20CL, p = 0.30), MRI‐based or PET‐only processing (18 versus 19CL, p = 0.88), and no evidence of difference across radiotracers (Flutemetamol: 16CL; PIB: 17CL, Flutafuranol: 18CL, Florbetaben: 19CL, Florbetapir: 20CL, p = 0.78). In a subset of 29,496 participants with visual reads available, binary visual reads corresponded well to Centiloids (common kappa=0.86, Figure 2A). The visual read‐based cutoff of 24CL (95%CI: 21‐27CL, Figure 2B) maximized correspondence between visual read and quantification and was slightly higher than the GMM‐based cutoff. All meta‐analysis models showed high non‐random heterogeneity (I 2 >80%) across studies, suggesting non‐random differences in peaks and cutoffs. Conclusions Meta‐analysis‐based cutoffs align well with thresholds from the existing literature. High heterogeneity among studies underscores the need to investigate contributing factors, raising concerns about applying common cutoffs.

Alzheimer s & DementiaVol. 21(S2)
American College of Radiology (US), University of Geneva (CH), University of Southern California (US), Universitat Autònoma de Barcelona (ES), University of North Texas (US), Alzheimer's Association (US), Cliniques Universitaires Saint-Luc (BE), Harvard University (US), National Sun Yat-sen University (TW), University of North Texas Health Science Center (US), The University of Melbourne (AU), University of Pittsburgh (US), Chang Gung University (TW), University of California, San Francisco (US), Lund University (SE), General Electric (Spain) (ES), Yonsei University (KR), Peking University (CN), Florey Institute of Neuroscience and Mental Health (AU), University of Michigan (US), Sheba Medical Center (IL), University of Florida Health (US), Buckinghamshire Healthcare NHS Trust (GB), Hospital de Sant Pau (ES), Samsung Medical Center (KR), University of Florida (US), Sorbonne Université (FR), Biomedical Research Networking Center on Neurodegenerative Diseases (ES), Massachusetts General Hospital (US), Kaohsiung Chang Gung Memorial Hospital (TW), University Memory and Aging Center (US), Alzheimer’s Disease Neuroimaging Initiative (US), Normandie Université (FR), University Hospital of Geneva (CH), Pasqual Maragall Foundation (ES), Institut de Myologie (FR), Pitié-Salpêtrière Hospital (FR), Mayo Clinic in Arizona (US), UK Dementia Research Institute (GB), Barcelonaβeta Brain Research Center (ES), Peking University Shenzhen Hospital (CN), Skåne University Hospital (SE), Tokyo Metropolitan Institute of Gerontology (JP), National Hospital for Neurology and Neurosurgery (GB), Amsterdam University Medical Centers (NL), Huashan Hospital (CN), National Center for Geriatrics and Gerontology (JP), Indiana University School of Medicine, University College London (GB), McGill University (CA), Indiana University – Purdue University Indianapolis (US), Vrije Universiteit Amsterdam (NL), Amsterdam UMC Location Vrije Universiteit Amsterdam (NL), Kindai University (JP), University of California, Berkeley (US), Stanford University (US), Université de Caen Normandie (FR)
Openalex Percentile: Top 29%
Dementia and Cognitive Impairment Research
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.