BACKGROUND: Alzheimer’s Disease (AD) is the leading cause of neurocognitive disorder (NCD) worldwide. Currently, its diagnosis relies on identifying amyloid plaques through expensive or invasive techniques, such as amyloid PET imaging and cerebrospinal fluid (CSF) biomarker analysis. A significant challenge in laboratory medicine is providing a blood-based marker for AD diagnosis to reduce the reliance on invasive procedures like lumbar punctures and the high costs associated with PET scans. Recently, phosphorylated tau at threonine 217 (ptau217) has emerged as an early and highly accurate blood biomarker for detecting typical AD changes. AIM OF THE STUDY: This thesis evaluated the diagnostic performance of the blood-based ptau217 assay in a real-world cohort of NCD subjects in a second-level care setting. The study compared results with current diagnostic gold standards to identify the optimal diagnostic cut-off and evaluate a dual cut-off strategy to minimize diagnostic uncertainty. MATERIALS AND METHODS: Sixty-five patients with NCD referred to the Padua University Hospital were enrolled. Clinical and neuropsychological data, along with biological samples (blood and CSF), were collected for each subject. The ptau217 concentrations were measured using the fully automated Lumipulse G1200 platform (Fujirebio). Diagnostic performances were evaluated through ROC curve analysis, using the Youden index to identify the single optimal cut-off. A dual cut-off model was designed to enhance sensitivity and specificity. For nine patients, ptau217 was measured in both K2-EDTA plasma and serum samples. RESULTS: Mean ptau217 levels were significantly higher in groups positive for amyloid pathology (PET A+ and CSF A+) and in patients with a clinical diagnosis of AD compared to controls (p < 0.001). ROC curve analysis showed high diagnostic accuracy, with an AUC of 0.91 relative to amyloid PET and 0.92 relative to CSF biomarkers. The best single cut-off was identified at 0.241 ng/L, with a diagnostic accuracy between 83% and 86%. The adoption of a dual cut-off model (approximately 0.16 ng/L to rule out pathology and 0.31 ng/L to confirm it) achieved a specificity of 96-97% for diagnostic confirmation, reducing the need for invasive tests to only those cases falling within the intermediate zone of uncertainty. Ten patients showed discordant results between PET and ptau217. No significant association was observed between plasma ptau217 and serum creatinine, and the plasma-serum comparison showed a strong correlation (ρ = 0.92; p = 0.001). CONCLUSIONS: In this real-world cohort, blood-based ptau217 demonstrated robust diagnostic performance and substantial concordance with both amyloid PET imaging and CSF biomarkers. These data support its clinical utility as a non-invasive, cost-effective, and easily implementable tool in the diagnostic pathway for patients with NCD. In complex clinical settings where multiple comorbidities can influence interpretation, the application of a dual-threshold strategy for ptau217 may further improve diagnostic accuracy in the most challenging cases.
BACKGROUND: La malattia di Alzheimer (Alzheimer’s Disease, AD) rappresenta la principale causa di disturbo neurocognitivo (DNC) nel mondo. La diagnosi si basa attualmente sull’identificazione del deposito di placche amiloidi mediante metodiche costose o invasive, quali la PET con tracciante per amiloide e l’analisi dei biomarcatori nel liquido cerebrospinale (CSF). Una delle principali sfide della Medicina di laboratorio è individuare biomarcatori ematici affidabili per la diagnosi di AD, riducendo il ricorso a procedure invasive e i costi diagnostici. Tra questi, la tau fosforilata in treonina 217 (ptau217) si è dimostrata un biomarcatore precoce e accurato delle alterazioni tipiche della malattia. SCOPO DELLO STUDIO: Valutare le performance diagnostiche della ptau217 ematica in una coorte real-world di soggetti sintomatici afferenti a un contesto assistenziale di secondo livello, confrontandone le concentrazioni con la PET amiloide e i biomarcatori nel CSF. Lo studio mira a identificare il cut-off ottimale per l’individuazione dei soggetti affetti e a valutare una strategia diagnostica basata su un doppio cut-off, per ridurre l’incertezza diagnostica. MATERIALI E METODI: Sono stati arruolati 65 pazienti con DNC presso l’Azienda Ospedale-Università di Padova. Per ciascun soggetto sono stati raccolti dati clinici, neuropsicologici e campioni di sangue e CSF. La concentrazione di ptau217 è stata determinata su plasma K2-EDTA e siero mediante piattaforma automatizzata Lumipulse G1200 (Fujirebio). Le performance diagnostiche del biomarcatore sono state valutate in confronto alla diagnostica standard mediante analisi delle curve ROC, calcolando il singolo miglior cut-off mediante indice di Youden. Un modello diagnostico a doppio cut-off è stato sviluppato per massimizzare sensibilità e specificità. Per 9 pazienti è stato eseguito un confronto diretto dei livelli di ptau217 misurati su plasma e su siero. RISULTATI: Le concentrazioni plasmatiche di ptau217 sono risultate significativamente più elevate nei soggetti positivi alla patologia amiloidea secondo PET e CSF e con diagnosi clinica di AD rispetto ai soggetti non AD (p < 0,001). L’analisi ROC ha evidenziato un’elevata accuratezza diagnostica del biomarcatore, con AUC pari a 0,91 rispetto a PET amiloide e diagnosi clinica e a 0,92 rispetto a CSF. Il miglior cut-off singolo è stato identificato a 0,241 ng/L, con accuratezza diagnostica compresa tra 83% e 86%. L’applicazione di un modello a doppio cut-off (0,16 ng/L per escludere la patologia e 0,31 ng/L per confermarla) ha raggiunto una specificità del 96-97%, limitando gli accertamenti invasivi ai soli casi della zona intermedia di incertezza. Abbiamo ottenuto 10 casi discordanti tra PET e ptau217. Non è stata osservata alcuna associazione significativa tra ptau217 plasmatica e la creatinina sierica (rho = −0,13; p = 0,355), mentre il confronto tra le misurazioni ottenute su plasma e siero ha evidenziato una forte correlazione (ρ = 0,92; p = 0,001). CONCLUSIONI: In questa coorte real-world, la ptau217 ematica ha mostrato elevate performance diagnostiche e una buona concordanza con PET amiloide e biomarcatori del CSF, supportando il suo potenziale impiego come strumento non invasivo, economicamente sostenibile e facilmente implementabile nella pratica clinica per la valutazione dei pazienti con DNC. Nei casi di maggiore complessità diagnostica, una strategia a doppio cut-off potrebbe contribuire a migliorare l’accuratezza diagnostica.
Valutazione delle performance diagnostiche del test ptau217 ematico in soggetti con disturbo neurocognitivo.
PENELLO, VALENTINA
2025/2026
Abstract
BACKGROUND: Alzheimer’s Disease (AD) is the leading cause of neurocognitive disorder (NCD) worldwide. Currently, its diagnosis relies on identifying amyloid plaques through expensive or invasive techniques, such as amyloid PET imaging and cerebrospinal fluid (CSF) biomarker analysis. A significant challenge in laboratory medicine is providing a blood-based marker for AD diagnosis to reduce the reliance on invasive procedures like lumbar punctures and the high costs associated with PET scans. Recently, phosphorylated tau at threonine 217 (ptau217) has emerged as an early and highly accurate blood biomarker for detecting typical AD changes. AIM OF THE STUDY: This thesis evaluated the diagnostic performance of the blood-based ptau217 assay in a real-world cohort of NCD subjects in a second-level care setting. The study compared results with current diagnostic gold standards to identify the optimal diagnostic cut-off and evaluate a dual cut-off strategy to minimize diagnostic uncertainty. MATERIALS AND METHODS: Sixty-five patients with NCD referred to the Padua University Hospital were enrolled. Clinical and neuropsychological data, along with biological samples (blood and CSF), were collected for each subject. The ptau217 concentrations were measured using the fully automated Lumipulse G1200 platform (Fujirebio). Diagnostic performances were evaluated through ROC curve analysis, using the Youden index to identify the single optimal cut-off. A dual cut-off model was designed to enhance sensitivity and specificity. For nine patients, ptau217 was measured in both K2-EDTA plasma and serum samples. RESULTS: Mean ptau217 levels were significantly higher in groups positive for amyloid pathology (PET A+ and CSF A+) and in patients with a clinical diagnosis of AD compared to controls (p < 0.001). ROC curve analysis showed high diagnostic accuracy, with an AUC of 0.91 relative to amyloid PET and 0.92 relative to CSF biomarkers. The best single cut-off was identified at 0.241 ng/L, with a diagnostic accuracy between 83% and 86%. The adoption of a dual cut-off model (approximately 0.16 ng/L to rule out pathology and 0.31 ng/L to confirm it) achieved a specificity of 96-97% for diagnostic confirmation, reducing the need for invasive tests to only those cases falling within the intermediate zone of uncertainty. Ten patients showed discordant results between PET and ptau217. No significant association was observed between plasma ptau217 and serum creatinine, and the plasma-serum comparison showed a strong correlation (ρ = 0.92; p = 0.001). CONCLUSIONS: In this real-world cohort, blood-based ptau217 demonstrated robust diagnostic performance and substantial concordance with both amyloid PET imaging and CSF biomarkers. These data support its clinical utility as a non-invasive, cost-effective, and easily implementable tool in the diagnostic pathway for patients with NCD. In complex clinical settings where multiple comorbidities can influence interpretation, the application of a dual-threshold strategy for ptau217 may further improve diagnostic accuracy in the most challenging cases.| File | Dimensione | Formato | |
|---|---|---|---|
|
Penello_Valentina.pdf
Accesso riservato
Dimensione
999.96 kB
Formato
Adobe PDF
|
999.96 kB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/109265