Kidney dysfunction and anemia systematically elevate plasma phosphorylated tau 217 levels, altering diagnostic classification when uniform cutoffs are applied, according to a recent study.
In a large multicenter cohort, subgroup-specific plasma phosphorylated tau 217 cutoffs improved diagnostic accuracy and reduced downstream costs compared with a standard single cutoff, with the greatest gains observed in patients with advanced chronic kidney disease. In obesity, the double-cutoff strategy showed greater diagnostic accuracy and cost efficiency than alternative approaches but also generated a substantial proportion of intermediate results requiring confirmatory imaging.
Researchers analyzed data from the Korea Registries to Overcome Dementia and Accelerate Dementia cohort, which included cognitively unimpaired participants, patients with mild cognitive impairment, and patients with dementia of Alzheimer type. All participants underwent amyloid positron emission tomography (PET), standardized clinical evaluation, and plasma p-tau217 testing. The cross-sectional study was conducted between 2016 and 2023, with analyses completed in 2025, and included 2,571 participants evaluated using the University of Gothenburg (UGOT) Simoa assay, 1,578 using the Roche Elecsys assay, and 304 assessed with a mass spectrometry–based p-tau217 to nonphosphorylated tau217 ratio.
Amyloid-β positivity was defined using a Centiloid threshold of 25.5 on PET. Three diagnostic strategies were compared: a standard single cutoff derived from the overall cohort, subgroup-specific optimal cutoffs adjusted for kidney function, body mass index, and anemia, and a double-cutoff strategy designed to reduce false-positive and false-negative classifications. Plasma p-tau217 measurements were performed blinded to clinical data. Diagnostic performance was assessed using receiver operating characteristic analyses, with accuracy, sensitivity, specificity, and modeled cost implications evaluated across biological subgroups.
Plasma p-tau217 concentrations were inversely associated with estimated glomerular filtration rate, body mass index, and hemoglobin, independent of amyloid burden, on both the UGOT and Roche platforms. In contrast, the normalized p-tau217 ratio showed minimal modulation by these systemic factors. Application of subgroup-specific optimal cutoffs substantially improved diagnostic accuracy in chronic kidney disease, increasing accuracy from 0.65 to 0.83 in the UGOT cohort, with the largest gains observed in participants with an estimated glomerular filtration rate below 45 mL/min/1.73 m². Accuracy in patients with anemia also improved, rising from 0.80 to 0.86, with consistent results observed in the Roche cohort.
The double-cutoff strategy further reduced false classifications but yielded intermediate results in up to 39% of cases, necessitating confirmatory PET and increasing downstream costs. Cost analyses showed that optimal cutoffs provided a more favorable balance of diagnostic accuracy and economic efficiency in chronic kidney disease and anemia, whereas the double-cutoff strategy remained superior in obesity, supporting a biologically informed approach to plasma p-tau217 implementation.
Several limitations were noted. The cohort consisted exclusively of East Asian participants, limiting generalizability to more diverse populations. Not all comorbidities that may influence plasma p-tau217 levels were evaluated. In addition, cost-effectiveness estimates were based on modeled assumptions that may vary across health care systems. External validation in multinational cohorts is needed to confirm the broader applicability of biologically adjusted plasma p-tau217 cutoffs.
“Collectively, these findings indicate that kidney dysfunction and anemia elevate baseline plasma p-tau217 levels, necessitating modest upward adjustment of diagnostic thresholds to preserve precision for Aβ positivity,” noted lead researcher Jihwan Yun, MD, of the Department of Neurology, Kyung Hee University Hospital, Kyung Hee University College of Medicine, Seoul, Republic of Korea, and colleagues.
Disclosures can be found in the study.
Source: JAMA Neurology
