Neurodegeneration Original research Plasma phosphorylated tau 181 predicts amyloid status and conversion to dementia stage dependent on renal function Sylvain Lehmann ‍ ‍,1 Susanna Schraen-­Maschke,2 Jean-­Sébastien Vidal ‍ ‍,3 Constance Delaby ‍ ‍,1,4 Frédéric Blanc,5 Claire Paquet,6 Bernadette Allinquant,7 Stéphanie Bombois,2,8 Audrey Gabelle,9 Olivier Hanon ‍ ‍3 ► Additional supplemental ABSTRACT material is published online Objectives Plasma P-­tau181 is an increasingly WHAT IS ALREADY KNOWN ON THIS TOPIC only. To view, please visit ⇒ The clinical use of plasma phosphorylated tau the journal online (http://​dx.​ established diagnostic marker for Alzheimer’s disease doi.​org/​10.​1136/​jnnp-​2022-​ (AD). Further validation in prospective cohorts is still 181 (P-­tau181) for Alzheimer’s disease is being 330540). needed, as well as the study of confounding factors that considered but further validation and study of could influence its blood level. confounding factors are still needed. For numbered affiliations see Methods This study is ancillary to the prospective WHAT THIS STUDY ADDS end of article. multicentre Biomarker of AmyLoid pepTide and ⇒ In our large prospective cohort, P-­tau181 Correspondence to AlZheimer’s diseAse Risk cohort that enrolled predicts brain amyloidopathy and conversion Professor Sylvain Lehmann, participants with mild cognitive impairment (MCI) who to dementia in patients with mild cognitive LBPC-­PPC, Montpellier were examined for conversion to dementia for up to Universite d’Excellence, impairment, but renal function significantly 3 years. Plasma Ptau-­181 was measured using the Montpellier 34295, France; ​ alters plasma levels and thus may induce ultrasensitive Quanterix HD-­X assay. sylvain.​lehmann@​umontpellier.​ diagnostic errors if not taken into account. fr Results Among 476 MCI participants, 67% were amyloid positive (Aβ+) at baseline and 30% developed HOW THIS STUDY MIGHT AFFECT RESEARCH, Received 10 October 2022 dementia. Plasma P-­tau181 was higher in the Aβ+ PRACTICE OR POLICY Accepted 15 December 2022 population (3.9 (SD 1.4) vs 2.6 (SD 1.4) pg/mL) and in ⇒ Our study suggests that measurement of Published Online First 3 April 2023 MCI that converted to dementia (3.8 (SD 1.5) vs 2.9 creatinine or estimation of glomerular filtration (SD 1.4) pg/mL). The addition of plasma P-­tau181 to a rate, which are easy and standardisable ways logistic regression model combining age, sex, APOEε4 to provide information on renal function, will status and Mini Mental State Examination improved contribute to optimal interpretation of plasma predictive performance (areas under the curve 0.691– P-­tau181 results in routine clinical practice. 0.744 for conversion and 0.786–0.849 for Aβ+). The Kaplan-­Meier curve of conversion to dementia, according to the tertiles of plasma P-­tau181, revealed a significant that modifiable risk factors account for 40% of predictive value (Log rank p<0.0001) with an HR of 3.8 AD.1 Furthermore, many potential treatments are (95% CI 2.5 to 5.8). In addition, patients with plasma in the final stages of study.2 Early diagnosis will be P-­Tau(181) ≤2.32 pg/mL had a conversion rate of less the key to further understanding of AD and success- than 20% over a 3-­year period. Using a linear regression fully treating it. approach, chronic kidney disease, creatinine and The development of biological biomarkers, estimated glomerular filtration rate were independently linked to amyloid and tau pathology, has greatly associated with plasma P-­tau181 concentrations. contributed to the understanding of the presymp- Conclusions Plasma P-­tau181 effectively detects Aβ+ tomatic and postsymptomatic AD ‘continuum’, in status and conversion to dementia, confirming the value which, thanks to genetic forms, it has been shown of this blood biomarker for the management of AD. that the disease was present from a biological point However, renal function significantly modifies its levels of view decades before its clinical appearance.3 This ► http://​​dx.​​doi.​​org/​​10.​​1136/​ jnnp-​2022-​330619 and may thus induce diagnostic errors if not taken into opens up therapeutic perspectives and allows us to account. use these biomarkers in early diagnosis and even in risk assessment. Their relevance has already been proven in cerebrospinal fluid (CSF) justifying their © Author(s) (or their implementation in clinical routine.4 5 Furthermore, employer(s)) 2023. Re-­use permitted under CC BY-­NC. No INTRODUCTION the blood is now amenable to diagnostic assays, commercial re-­use. See rights Alzheimer’s disease (AD) accounts for 60%–70% of thanks to ultrasensitive techniques, including mass and permissions. Published dementia and is thus a major public health problem spectrometry. by BMJ. and socioeconomic burden that is only increasing The first analytes used in blood were amyloid To cite: Lehmann S, Schraen-­ with the ageing of the population. For a long time, peptides. Their levels in serum can detect amyloid Maschke S, Vidal J-­S, et al. J diagnostic efforts have been minimal, due to the positive (Aβ+) patients, as defined by amyloid PET Neurol Neurosurg Psychiatry lack of preventive measures or curative treatment. or CSF analysis, and predict evolution of patients 2023;94:411–419. However, in recent years, it has been demonstrated within the AD continuum, including conversion of Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 411 Neurodegeneration patients, with mild cognitive impairment (MCI), to dementia.6 Germany). Baseline blood samples were used to measure fasting While the detection of total tau in blood was not so discrimi- glycaemia, cholesterol (total, high-­density lipoproteins (HDL), nating,7 detection of its phosphorylated forms at position thre- low-­ density lipoprotein (LDL)), prealbumin, albumin, creat- onine 181 was a real breakthrough.8 P-­tau181 concentration is inine.16 Estimated glomerular filtration rate (eGFR), based on predictive of amyloid or tau PET findings and is significantly creatinine, age and sex, was computed using the chronic kidney higher in AD and MCI compared with subjects without cognitive disease (CKD)-­ Epidemiology Collaboration (CKD-­ EPI) equa- impairment or compared with other causes of dementia.8–10 Its tion18. CSF biomarkers were measured in a single centralised prognostic value has also been established in some longitudinal laboratory using commercially available Innotest assays for tau studies11–13 and its overall performance allow to consider a clin- and phosphorylated tau at position T181 (P-­tau181) or Euro- ical application. To reach this milestone, we will need additional immun for amyloid peptides Aβ1–42 and Aβ1–40. Positive prospective data, in vitro diagnostic certified kits and sufficient amyloid status (Aβ+) was defined, as previously, when the CSF information on preanalytical stability.14 15 It will equally be para- Aβ1–42/Aβ1–40 ratio was below 0.1.19 mount to identify any confounding factors that may alter clinical Plasma P-­tau181 was determined using a commercial P-­tau181 performance in routine use.14 assay kit (Quanterix, USA) based on ultrasensitive Simoa tech- In this work, we used the prospective multicentre Biomarker nology20 on an HD-­ X analytical platform. All samples were of AmyLoid pepTide and AlZheimer’s diseAse Risk (BALT- fourfold diluted with the provided dilution buffer to minimise AZAR) cohort16 to confirm the ability of plasma P-­tau181 to matrix effects. After dilution, the lowest limit of detection was detect brain amyloidopathy and also to predict the conversion 0.019 pg/mL and the limit of quantification was of 0.085 pg/ of MCI patients to dementia stage. As plasma biomarkers are mL. Quality controls with low (QC 1 with mean concentration soon to be implemented in routine practice, we also investigated of 3.82 pg/mL) or high (QC 2–52.4 pg/mL) P-­tau181 known potential confounding factors that must be considered to inter- concentration were provided in the kits. Inter-­assay variation for pret the results adequately. QC 1 and QC 2 was low, with coefficient of variation (CV) of 7% and 5%, respectively. We also used two serum pools (average P-­tau181 of 4.47 pg/mL and 2.81 pg/mL) as internal QCs run at MATERIALS AND METHODS the beginning and end of each sample plate. These had low inter-­ Study population assay CV of 3% and 6%, respectively. The BALTAZAR study is a multicentre prospective cohort study (​ ClinicalTrials.​ gov Identifier #NCT01315639) that enrolled patients with MCI or AD, according to a previously described Statistical analyses protocol.16 All participants had clinical, neuropsycholog- General characteristics were analysed in the whole MCI sample, ical, brain MRI and biological assessments (see below). Right according to MCI subtype (aMCI and naMCI), conversion to and left hippocampal volume was obtain for each participant dementia and to plasma P-­tau181 tertile. Categorical variables using automatic segmentation of the hippocampus. The hippo- are presented as percentages and counts (% (N)); continuous campal volume was normalised using the following calculation: variables, as mean and SD (M (SD)), or median (25–75th percen- hippocampal volume/total brain volume×mean total brain tile), and comparisons were assessed by χ2 tests, t-­tests, Mann-­ volume. CSF samples were collected only in accepting partici- Whitney U test and analysis of variance (ANOVA, Kruskal-­Wallis pants. APOE was genotyped in a single centralised laboratory. test). The relationship between conversion and plasma P-­tau181 MCI subjects were selected according to the Petersen’ criteria17 was assessed using regression models with age, sex and baseline and then they were dichotomised into amnestic (aMCI) and presence of APOE ε4 allele as covariables. non-­amnestic (naMCI) phenotypes, based on the presence of Kaplan-­Meier curves were drawn for conversion according to memory impairment on the free and cued selective reminding plasma P-­tau181tertile and overall differences between tertiles test related to age, sex and educational level. The characteristic was calculated by log rank test. We also examined how plasma of the aMCI/naMCI population is fully described elsewhere.16 P-­tau181 improved dementia risk prediction using logistic Patients had visits every 6 months for 3 years. MCI participants regression with age, sex, APOE ε4 and MMSE score at baseline were reassessed for conversion to dementia at each visit by the and by calculating continuous net reclassification improvement clinician.6 The progression from MCI to dementia was defined (NRI).21 Receiving operator characteristic (ROC) curves, using by evaluating the following parameters: (1) decline in cognitive conversion as a dependent variable, were also used to compute function (measured by changes from the baseline in scores of the for different factors. The corresponding areas under the curve Mini Mental State Examination (MMSE)), (2) disability in activ- (AUCs) were compared using the Delong method.22 Logistic ities of daily living (ADL) (instrumental ADL (IADL) >1) and regression model (enter model), Kaplan Meier and ROC curves (3) clinical dementia rating sum of boxes (>1). The conversions were generated with MedCalc (V.20.111) software. In all anal- were reviewed by an adjudication committee. yses, the two-­sided α-level of 0.05 was used for significance In this study, we analysed 476 available baseline plasma testing. samples from patients with MCI diagnosis (365 aMCI and 111 naMCI). RESULTS Characteristics of the MCI participants at baseline Biological biomarker measurements Of the 539 MCI participants enrolled in the BALTAZAR study, To minimise preanalytical and analytical problems, identical 63 were excluded due to missing data or absence of plasma collection tubes were used across centres to collect plasma (EDTA P-­ tau181 biomarkers. In this study, we analysed, 476 MCI BD Vacutainer K2E, ref 367 525, Becton Dickinson, USA) and participants (mean age 77.7 (SD 5.5) years, 61.4% women) with for CSF (10 mL polypropylene tube, ref 62.610.201, Sarstedt, 365 aMCI (77%) and 111 naMCI (23%) at baseline (table 1, Germany). Blood and CSF samples were collected on the same online supplemental table 1). Average MMSE score was 26.4 day. All aliquots were stored in the same low-­binding Eppen- (SD 2.5) and 39.8% (n=185) were APOE ε4 carriers. During dorf LoBind microtubes (Eppendorf, ref 022431064, Hamburg, the clinical follow-­up period of 6 to 36 months, 30% (n=144) of 412 Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 Neurodegeneration Table 1 Characteristics in the whole MCI population and between MCI participants who converted, or not, to dementia within 3 years All MCI MCI non-­converters MCI converters N=476 N=332 N=144 P value P$ Patient characteristics  Age (years) 77.7 (5.5) 77.3 (5.4) 78.5 (5.7) 0.048 0.003  Women (%) 292 (61.4) 205 (61.7) 60.4 0.78 0.67  BMI (kg/m2) 25 (3.8) 25.1 (3.8) 24.8 (3.8) 0.39 0.95  MMSE (/30) 26.4 (2.5) 26.7 (2.5) 25.6 (2.5) <0.0001 0.0002  1 or 2 APOE4 alleles 185 (39.8) 105 (31.6) 80 (55.5) <0.0001 <0.0001  Hippocampal volume (R+L) (cm3) 4.55 (1.12) 4.79 (1.05) 4.01 (1.09) <0.0001 <0.0001 CSF biomarkers*  Aβ1–40 (pg/mL) 7434 (2241) 7458 (2298) 7389 (2143) 0.83 0.64  Aβ1–42 (pg/mL) 766 (384) 857 (398) 593 (288) <0.0001 <0.0001  Aβ1–42/Aβ1–40 0.104 (0.045) 0.116 (0.045) 0.082 (0.036) <0.0001 <0.0001  Tau (pg/mL) 578 (254) 376.1 (185.2) 547 (223.4) <0.0001 <0.0001  p-­tau181 (pg/mL) 76.7 (30.2) 58.58 (24.47) 80.3 (33.58) <0.0001 <0.0001 Blood biomarkers  Fasting glycaemia (mmol/L) 5.37 (1.19) 5.34 (1.21) 5.45 (1.14) 0.34 0.20  Triglycerides (mmol/L) 1.21 (0.59) 1.2 (0.58) 1.24 (0.6) 0.54 0.82  Cholesterol total (mmol/L) 5.5 (1.16) 5.48 (1.2) 5.53 (1.07) 0.67 0.91  Cholesterol HDL (mmol/L) 1.74 (0.52) 1.75 (0.53) 1.72 (0.48) 0.66 0.98  Cholesterol LDL (mmol/L) 3.2 (1) 3.19 (1.01) 3.24 (0.98) 0.65 0.90  Prealbumin (mg/dl) 27.6 (5.4) 28.1 (6.2) 27.7 (5.8) 0.53 0.52  Albumin (g/L) 40.3 (3.9) 40.5 (3.5) 39.8 (4.5) 0.09 0.11  Creatinine (μmol/L) 78.2 (21.3) 79.4 (22.2) 73.7 (13.6) 0.10 0.82  eGFR (mL/min/1.73 m2) 76.9 (14.7) 77.0 (14.8) 76.7 (14.7) 0.85 0.80  Plasma P-­tau181(pg/mL) 3.19 (1.49) 2.9 (1.4) 3.8 (1.5) <0.0001 <0.0001 P: Comparison between the three groups, by ANOVA or χ2; P$: comparison between the three groups by linear regression adjusted for age, sex and the presence of the APOE ε4 allele; % (number) were used to describe categorical variable, mean±SD for continuous variables. *CSF biomarkers were available in 140 and 74 MCI non-­converters and converters, respectively. ANOVA, analysis of variance; APOE, apolipoprotein E; BMI, body mass index; CSF, cerebrospinal fluid; eGFR, estimated glomerular filtration rate; MCI, mild cognitive impairment; MMSE, Mini Mental State Examination; R+L, right+left. the MCI participants developed dementia, on average 14.6 (SD Almost half of the MCI participants had a lumbar puncture and 8.2) months after the baseline visit and in 95% of the cases, they 117 of them were Aβ+ and 97 were Aβ−. At baseline, Aβ+ converted to clinically probable AD.6 patients were older, had a lower MMSE score and were more often APOE ε4 carriers. However, unlike MCI converters, Aβ+ Comparison between MCI converters and non-converters: patients did not have a lower hippocampal volume (table 2). CSF P-tau predicts conversion to AD Aβ1–42, Tau and p-­tau181 values were also highly differential At baseline, the MCI converters to dementia were older, had a between Aβ+ and Aβ− patients. Plasma P-­tau181 was signifi- lower MMSE score, were more often APOE ε4 carriers and had cantly higher on average by 50% in Aβ+ than in Aβ− (3.9 more severe hippocampal atrophy (table 1). As lumbar puncture (SD 1.4) vs 2.6 (SD 1.4) pg/mL, p<0.0001). This difference was optional in our cohort, CSF values were available in 214 remained significant after adjustment for age, sex and APOE ε4 subjects. CSF Aβ1–40, Aβ1–42, tau and p-­tau181 values were status (table 2). highly differential between MCI that converted or not. We next turned our attention to the efficacy of markers in the P-tau181 improves predictive power of age, sex, APOEε4 blood, as these samples were available from all participants. We status and MMSE for MCI conversion and amyloid status have previously described differential levels of plasma amyloid detection biomarkers in the BALTAZAR cohort.6 Here we found plasma Using a logistic regression approach with conversion as a P-­tau181 increased significantly, on average by 30%, between dependent variable and age, sex, APOEε4 status and MMSE MCI non-­converters at 2.9 (SD 1.4) pg/mL and converters at as independent variables, it was possible to predict conversion 3.8 (SD 1.5) pg/mL (p<0.0001). Plasma P-­tau remained signifi- (p<0.0001) with an AUC of the model fit of 0.691 (95% CI cant after adjustment for age, sex and APOE ε4 status (table 1). 0.655 to 0.741). The addition of plasma P-­tau181 resulted in No difference was observed between MCI converters and non-­ a significant increase of the AUC to 0.744 (95% CI 0.702 to converters for metabolic or renal function blood biomarkers: 0.784) (online supplemental table 3). The added value of plasma fasting blood glucose (glycaemia), triglycerides, cholesterol P-­tau181 was further documented by computing the NRI of the (total, HDL, LDL), prealbumin, albumin and creatinine or eGFR. two models. This revealed a 12.8% improvement in patient clas- sification between MCI converters and non-­converters due to Comparison between Aβ+ and Aβ− patients: tau predicts plasma P-­tau181. Since blood biomarkers are intended to replace amyloid status CSF biomarkers, we compared the respective values of plasma The amyloid status, Aβ+ corresponding to the (A+) ATN classi- and CSF P-­tau181 for amyloid status detection in the subcohort fication,3 was defined based on the CSF Aβ1–42/Aβ1–40 ratio.23 where patients had undergone a lumbar puncture. The addition Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 413 Neurodegeneration Table 2 Characteristics in the whole population and between Aβ− and + patients All Aβ− Aβ+ N=214 N=97 N=117 P value P value$ Patient characteristics  Age (years) 77.4 (5.6) 76.6 (5.1) 78 (5.9) 0.073 0.0122  Women (%) 127 (59.3) 53 (54.6) 74 (64.9) 0.20 0.36  BMI (kg/m2) 24.7 (3.7) 25.4 (3.7) 24.2 (3.6) 0.024 0.26  MMSE (/30) 26.4 (2.4) 27.1 (2) 25.8 (2.5) <0.0001 0.0007  One or 2 APOE4 alleles 78 (36.4) 15 (15.5) 63 (57.3) <0.0001 <0.0001  Hippocampal volume (R+L) (cm3) 4.56 (1.09) 4.64 (1.23) 4.5 (0.96) 0.41 0.91 CSF biomarkers  Aβ1–40 (pg/mL) 7434 (2241) 7421 (1896) 7446 (2499) 0.93 0.62  Aβ1–42 (pg/mL) 766 (385) 1095 (287) 494 (196) <0.0001 <0.0001  Aβ1–42/Aβ1–40 0.104 (0.045) 0.149 (0.022) 0.068 (0.018) <0.0001 <0.0001  Tau (pg/mL) 433 (213) 320 (133) 533 (222) <0.0001 <0.0001  p-­tau181 (pg/mL) 66.4 (30.2) 51.3 (15.1) 79 (33.8) <0.0001 <0.0001 Blood biomarkers  Fasting glycaemia (mmol/L) 5.4 (1.1) 5.4 (1.1) 5.4 (1.2) 0.99 0.71  Triglycerides (mmol/L) 1.17 (0.6) 1.18 (0.48) 1.15 (0.69) 0.70 0.62  Cholesterol total (mmol/L) 5.5 (1.2) 5.5 (1.3) 5.5 (1.1) 0.78 0.29  Cholesterol LDL (mmol/L) 1.7 (0.5) 1.8 (0.5) 1.7 (0.5) 0.58 0.81  Cholesterol HDL (mmol/L) 3.2 (1) 3.3 (1.1) 3.2 (0.9) 0.92 0.21  Prealbumin (mg/dL) 28.4 (6.4) 28.5 (7.6) 28.3 (5.1) 0.77 0.77  Albumin (g/L) 40.1 (4.3) 39.9 (3.6) 40.2 (4.9) 0.60 0.38  Creatinine (μmol/L) 79.0 (23.0) 80.1 (24.9) 78.2 (21.5) 0.54 0.76  eGFR (mL/min/1.73 m2) 76.9 (14.7) 76.9 (15.5) 77.5 (15.5) 0.62 0.45  Plasma P-­tau181(pg/mL) 3.3 (1.5) 2.6 (1.4) 3.9 (1.4) <0.0001 <0.0001 P: Comparison between the three groups, by ANOVA or χ2); P$: comparison between the three groups by linear regression adjusted for age, sex and the presence of the APOE ε4 allele; % (number) were used to describe categorical variables, mean±SD for continuous variables. HDL, high-­density lipoproteins; LDL, low-­density lipoproteins. ANOVA, analysis of variance; APOE, apolipoprotein E; Aβ+, amyloid positive; BMI, body mass index; CSF, cerebrospinal fluid; eGFR, estimated glomerular filtration rate; MCI, mild cognitive impairment; MMSE, Mini Mental State Examination; R+L, right+left. in the model of plasma or CSF P-­tau181 resulted in a significant was further documented by plotting the Kaplan-­ Meier curve increase of the AUC from 0.786 (95% CI 0.723 to 0.84) for of conversion to dementia according to the tertiles (figure 1A). age, sex, APOEε4 status and MMSE to 0.849 (95% CI 0.792 to A very significant overall difference was observed (Log rank 0.895) and 0.857 (95% CI 0.801 to 0.902), respectively (P(dif- p<0.0001) and the HR between the first and the third tertile ference)=0.00075 and 0.0002). AUCs obtained by the addition was 3.8 (95% CI 2.5 to 5.8). of plasma or CSF P-­tau181 for conversion or amyloid status detection (0.750 and 0.752, respectively) were not different Impact of comorbidities and covariates on P-tau181 (P(difference)=0.81). Importantly, when creatinine or eGFR concentration and diagnostic performance were associated with P-­tau181 using logistic regression, they did The relationship between plasma P-­ tau181 concentration not give in better performance models. and comorbidities, demographic factors and biological infor- mation collected at baseline in the BALTAZAR cohort was Association of plasma P-tau181 with other biomarkers and investigated using a linear regression approach. We identified cohort characteristics APOE status, creatinine and eGFR as strongly connected to The relationships and correlations between plasma P-­ tau181 plasma P-­tau181 (figure 1B). To a lesser degree, age and BMI concentration and the other biomarkers and cohort charac- also affected plasma P-­tau181levels. Among comorbidities we teristics were analysed after splitting the population by tertile tested, chronic kidney disease (CKD) appeared strongly linked (table 3). Age, body mass index (BMI) and APO ε4 were signifi- to P-­tau181 levels. All these results remained very similar after cantly different between tertile. Patient conversion rate was also adjustment for age, sex and APOE ε4 status. To further eval- clearly correlated with plasma P-­tau181, with values of 16.4%, uate the impact of these covariates, we plotted the correlation 26.1% and 47.8% in the first, second and third tertile, respec- between creatinine, eGFR, BMI and age in the population tively. Distribution of Aβ+ patients was also greatly increased stratified by amyloid status (figure 2A–D). The correlation along with tertile, a relationship that was confirmed by the in the BALTAZAR subpopulation with lumbar puncture high correlation observed between plasma P-­tau181 and CSF (n=214) remained significant only for eGFR and creatinine Aβ1–42/Aβ1–40 use to define the Aβ+ status (correlation (online supplemental table 2). We observed higher values of coefficient=−0.4428, p<0.0001). eGFR decreased (p=0.017) plasma P-­tau181 in the Aβ+ population. P-­tau181 levels were while creatinine very significantly increased (p<0.0001) in the also correlated with low and high values of eGFR and creat- higher plasma P-­tau181 tertiles. All these results remained signif- inine. To confirm this observation, we stratified the popula- icant after adjustment for age, sex and APOE ε4 status (table 3). tion based on tertile of creatinine or eGFR levels (figure 2E,F, The relationship between plasma P-­tau181 and MCI conversion table 4). The ANOVA confirmed that the mean level of plasma 414 Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 Neurodegeneration Table 3 Characteristics in the different P-­tau181 tertiles First tertile Second tertile Third tertile N=158 N=157 N=161 P value P value$ Plasma P-­tau181  Age (years) 76.9 (5.4) 77.9 (5.1) 78.3 (5.9) 0.06 0.0009  Women (%) 101 (63.9) 96 (61.1) 95 (59.0) 0.66 0.32  BMI (kg/m2) 25.8 (3.7) 24.9 (3.9) 24.4 (3.7) 0.004 0.01  MMSE (/30) 26.7 (2.4) 26.4 (2.5) 26.1 (2.7) 0.16 0.19  1 or 2 APOE4 alleles (%) 38 (17.7) 63 (40.1) 84 (52.2) <0.0001 <0.0001  Hippocampal volume (R+L) (cm3) 4.75 (1.13) 4.38 (1.16) 4.5 (1.04) 0.02 0.28  Aβ+ status (%) 13 (19.1) 40 (62.5) 64 (78.0) <0.0001 <0.0001  Conversion MCI (%) 26 (16.4) 41 (26.1) 77 (47.8) <0.0001 <0.0001 Blood biomarkers  Fasting glycaemia (mmol/L) 5.46 (1.28) 5.4 (1.1) 5.26 (1.18) 0.32 0.12  Triglycerides (mmol/L) 1.2 (0.6) 1.3 (0.7) 1.2 (0.5) 0.26 0.23  Cholesterol (mmol/L) 5.52 (1.19) 5.48 (1.23) 5.5 (1.06) 0.94 0.82  Cholesterol HDL (mmol/L) 1.72 (0.47) 1.76 (0.56) 1.75 (0.52) 0.85 0.29  Cholesterol LDL (mmol/L) 3.25 (1.03) 3.15 (1.07) 3.21 (0.9) 0.69 0.59  Prealbumin (mg/dL) 27.5 (6.2) 28 (6.7) 28.4 (5.4) 0.47 0.17  Albumin (g/L) 40.4 (3.7) 40.2 (4.8) 40.3 (3.1) 0.87 0.90  Creatinine (μmol/L) 73.5 (16.5) 78.6 (19) 82.1 (26.1) 0.002 0.0005  eGFR (mL/min/1.73 m2) 80.3 (13.5) 76 (14.1) 74.4 (15.9) 0.02 0.01  Plasma P-­tau181(pg/mL) 1.75 (0.38) 2.93 (0.36) 4.86 (1.19) NA NA P: Comparison between the three groups, by ANOVA or χ2); P$: comparison between the three groups with linear regression adjusted for age, sex and the presence of the APOE ε4 allele; % (number) were used to describe categorical variables, mean±SD for continuous variables. ANOVA, analysis of variance; APOE, apolipoprotein E; BMI, body mass index; eGFR, estimated glomerular filtration rate; MCI, mild cognitive impairment; MMSE, Mini Mental State Examination; NA, not available; R+L, right+left. P-­tau181 was globally differential among both creatinine or have 30% higher levels of plasma P-­tau181 independently of eGFR tertiles (p<0.001). Finally, to evaluate the impact on age, sex or APOE ε4. Importantly, 48% of MCI participants P-­tau181 on the detection of amyloid status, we computed the among the highest tertile of plasma P-­tau181 (>3.61) converted ROC curves and determined best cutpoints and corresponding to dementia and thus had a fourfold higher risk. In addition, performance for Aβ+ (table 4). patients in the first P-­tau(181) tertile (ie, with a value ≤2.32 pg/ DISCUSSION mL) have a conversion rate of 19.8% over a 3-­year period. It is Here, we present results from a large-­scale multicentre prospec- likely that combining P-­tau(181) with other blood biomarkers tive longitudinal cohort of clinically defined MCI participants, such as plasma amyloid peptides could improve this predic- referred to memory centre, with a follow-­up of 3 years. Our tion. This information is valuable for patient management and principal finding is that patients who convert to dementia for using therapeutic strategies to prevent progression. In this Figure 1 (A): Kaplan-­Meier curve of conversion to dementia according to the tertiles of plasma P-­tau181 in MCI subjects. (B): Associations between multiple factors and plasma P-­tau181 concentrations. Forest plots of associations between demographic, comorbidities (in blue) and biological variables and plasma P-­tau181, using linear regression. Means and 95% CIs are provided. Z-­scores are used to compare the factors between them. APOE, apolipoprotein E; BMI, body mass index; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; MCI, mild cognitive impairment; MMSE, Mini Mental State Examination; TIA, transient ischaemic attack; HDL, high-­density lipoproteins; LDL, low-­density lipoproteins. Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 415 Neurodegeneration Figure 2 (A–D): Correlation between plasma P-­tau181 and creatinine, eGFR, age and BMI in the amyloid negative and positive populations. (E, F): Levels of plasma P-­tau181 in the amyloid negative and positive populations, by tertiles of creatinine or eGFR. Concentrations of plasma P-­tau181 were significantly different between amyloid negative and positive patients in all cases, as tested using a Mann-­Whitney U test (p<0.05). Aβ+, amyloid positive; BMI, body mass index; eGFR, estimated glomerular filtration rate (unit: mL/min/1.73 m2). MCI population, plasma P-­tau181 also predicted amyloid status P-­tau181 accurately predicted future dementia and Aβ+ status (based on the CSF Aβ1–42/Aβ1–40 ratio), with Aβ+ patients (as defined by PET). As well as validating this previous study, having 50% higher P-­tau181 levels than their Aβ− counterparts. our study has the added value of the biological data collected One important element that raises the interest of using in the BALTAZAR cohort. These include metabolic blood this plasma biomarker in the future is its added value above biomarkers: fasting glycaemia, triglycerides, cholesterol (total, that of just using a combination of age, sex, APOEε4 status HDL, LDL), prealbumin, albumin, creatinine and eGFR, which and MMSE. It is noteworthy that adding plasma P-­ tau181 can be used to monitor diabetes, cardiovascular risk, nutritional significantly improved the detection of both Aβ+ patients and status or kidney function. None of these factors were differen- MCI converters. Even more striking is that this added value tial, either in comparing MCI converters to non-­converters, or of plasma P-­tau181 was equivalent to that of CSF P-­tau181. when comparing Aβ+ and Aβ− patients. However, when we This finding will impact future clinical use of the approach, investigated factors influencing P-­tau181 level, by comparing as it might avoid the need for lumbar puncture. The capacity tertiles or through a linear regression method, we first iden- of plasma P-­tau181 to detect Aβ+ patients as well as AD and tified age and BMI as cofounding factors. These two factors MCI when compared with control and to other diseases has have previously been associated with P-­tau181, as well as with been described.8 9 11–13 24–27 However, the only previous other other blood biomarkers like neurofilaments.28 Age increases large study focusing on MCI conversion was that of Karikari both plasma and CSF values of neurodegenerative biomarkers et al11 who observed that baseline concentrations of plasma like total tau29 yet to be determined reasons. For BMI, a likely 416 Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 Neurodegeneration Table 4 Performance of plasma P-­tau181 for Aβ+ detection with regard to renal function (creatinine and eGFR) Creatinine <68 Creatinine 69–82 Creatinine ≥82 Population with CSF biomarkers available Total First tertile Second tertile Third tertile Creatinine (μmol/L) 74 (71–77) 61 (57–64) 73 (71–80) 95 (84–111) Plasma P-­tau 181(pg/mL) 3.0 (2.8–3.4) 2.3 (1.8–3.3) 2.9 (2.5–3.9) 3.7 (2.7–5.2) Plasma P-­tau 181 in Aβ− 2.2 (1.9–2.4) 1.9 (1.5–2.4) 2.3 (1.6–2.7) 2.9 (1.9–4.0) Plasma P-­tau 181 in Aβ+ 3.7 (2.8–4.5) 3.1 (2.2–4.3) 3.6 (2.9–4.0) 4.2 (3.6–5.4) AUC Plasma P-­tau 181 Aβ+ 0.783 0.763 0.872 0.733 Plasma P-­tau 181 cutpoints (pg/mL) 2.77 2.31 2.77 3.52 Sensitivity (%) 80.9 75.0 89.5 80.0 Specificity (%) 69.6 73.0 80.0 70.0 Population with CSF biomarkers available Total eGFR <74 eGFR 74–86 eGFR >86 first tertile second tertile third tertile eGFR (mL/min/1.73 m2) 79 (68–89) 64 (54–69) 79 (76–84) 91 (89–93) Plasma P-­tau 181(pg/mL) 3.0 (2.1–3.9) 3.5 (2.4–4.2) 3.1 (2.1–3.9) 2.7 (1.8–3.9) Plasma P-­tau 181 in Aβ− 2.2 (1.7–3.0) 2.7 (1.7–3.4) 2.0 (1.7–2.6) 2.0 (1.6–2.9) Plasma P-­tau 181 in Aβ+ 3.7 (2.8–4.5) 3.8 (3.1–5.4) 3.8 (2.8–4.5) 3.3 (2.7–4.2) AUC Plasma P-­tau 181 Aβ+ 0.783 0.763 0.872 0.733 Plasma P-­tau 181 cutpoints (pg/mL) 2.77 2.81 2.71 2.44 Sensitivity (%) 80.9 86.4 84.2 78.8 Specificity (%) 69.6 60.0 77.8 68.6 Values of creatinine and plasma P-­tau181 are express as median (25–75th percentile). Cutpoints correspond to the best Youden index on the ROC curves. AUC, area under the curve; Aβ+, amyloid positive; CSF, cerebrospinal fluid; eGFR, estimated glomerular filtration rate; ROC, receiving operator characteristic. relevant factor is the dilution of neuronal biomarkers in the developed dementia which in 95% of cases was represented by blood volume. probable AD. Amyloid status was available in only a part of the Among the comorbidities that were associated with P-­tau181, population, since the BALTAZAR study focused on conversion, CKD was the most differential. This association with CKD has and it was defined using CSF biomarkers rather than with PET already been reported in recent studies.30 31 However, in the amyloid. BALTAZAR cohort, we have access to the clinical chemistry The main strengths of the study lie in the large sample size of profile realised on the same plasma sample used for P-­tau181 MCI participants that are well described, the controlled preana- measurement. We thus noted that P-­ tau181 correlates with lytical conditions, the centralised plasma P-­tau181 analyses and markers of kidney function: creatinine and eGFR. Adding the availability of clinical chemistry analyte measurement real- these parameters improved the ability of P-­ tau181 to detect ised in the same sample tube. Aβ+ patients, whereas adding age and BMI did not. Strikingly P-­tau181 and creatinine stratify together irrespective of other CONCLUSION variables. Namely, in situations with increased creatinine (≥82 This study of our well-­ characterised population confirms µmol/L) or low eGFR (<74 mL/min/1.73 m2), indicating a the clinical relevance of plasma P-­tau181 for the detection of moderate impaired kidney function, levels of P-­ tau181 were amyloid status, which is important for risk assessment, patient increased in both Aβ+ and Aβ− patients, as well as in MCI management and inclusion in clinical trials. We also demon- patients converting or not to dementia. strate the strong predictive value of this blood biomarker for A major suggestion of our study is tailoring the clinical the prognosis of MCI patients, thus addressing an important cutpoints of P-­tau181 to renal function. We advocate minimising medical need in memory centres. The question remains as to the the false detection of a pathological situation in patients by use of blood biomarkers as a screening tool in patients without always combining plasma P-­tau181 with an assessment of renal cognitive impairment who have risk factors and may benefit function, for example, through creatinine measurement and most from preventive strategies, and/or as triage tests in patients GFR estimation. This recommendation should be confirmed for with early symptoms for whom future investigations, including other P-­tau isoforms (P-­tau217, P-­tau231) measured by immuno- imaging and spinal tap, are being considered. Finally, we iden- assay25 or mass spectrometry.32 We cannot exclude at this stage tified and quantified the impact of renal function, assessed by that altered renal function may also contribute in some way to creatinine levels and GFR estimation, on P-­tau181 blood levels. the progression of AD.33 Indeed, this hypothesis is supported by These measures are an easy and standardisable way to provide the difference in creatine level between naMCI and aMCI popu- essential information about kidney function and thus to optimise lation. To understand the relationship between renal function interpretation of results in routine clinical practice. and P-­tau levels, i’s clearance by the kidney will therefore have to be studied in more detail. Of note, only very small amounts of Author affiliations this biomarker were detected in the free form or associated with 1 LBPC-­PPC, Université de Montpellier, INM INSERM, IRMB CHU de Montpellier, exosome in urine.34 35 Montpellier, France 2 The present study has some limitations. To increase the likeli- Université de Lille, Inserm UMRS1172, CHU Lille, Lille, France 3 Université Paris Cité, EA 4468, Hopital Broca, Geriatric department, Memory hood of conversion to AD we excluded participants with Lewy Resource and Research Centre of Paris-­Broca-­Ile de France, APHP, Paris, France body, Parkinson, frontotemporal or vascular MCI disorders. 4 Hospital de la Santa Creu i Sant Pau - Biomedical Research Institute Sant Pau, Therefore, 77% of subjects had aMCI and 30% of participants Universitat Autònoma de Barcelona, Barcelona, Spain Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 417 Neurodegeneration 5 Neuropsychology Unit and Geriatric Day Hospital, University Hospital of Strasbourg, [23] CHU de Limoges, Memory Research Resource Center of Limoges, F-­87000 Strasbourg, France Limoges, France. [24] Université de Rouen Normandie, CHU Charles Nicolle, Memory 6 Université Paris Cité, GHU APHP Nord Lariboisière Fernand Widal, Centre de Resource and Research Centre of Haute-­Normandie, Inserm U1079, IRIB, F-­76000 Neurologie Cognitive, Paris, France Rouen, France. [25] CHU de Besançon, Memory Resource and Research Centre 7 UMR-­S 1266, Université Paris Cité, Institute of Psychiatry and Neurosciences, of BesançonFranche-­Comté, F-­25000 Besançon, France. [26] Université de Reims Inserm, Paris, France Champagne Ardenne, EA 3797, CHU de Reims, Memory Resource and Research 8 Assistance Publique-­Hôpitaux de Paris (AP-­HP), Département de Neurologie, Centre Centre of Champagne Ardenne, F-­51100 Reims, France. [27] CHU d’Angers, Memory des Maladies Cognitives et Comportementales, GH Pitié-Salpêtrière, Paris, France Resource and Research Centre of Angers, F-­49000 Angers, France. [28] CHRU de 9 Université de Montpellier, CHRU Montpellier, Memory Research and Resources Tours, Memory Resource and Research Centre of Tours, F-­37000 Tours, France. [29] center, department of Neurology, Inserm INM, Montpellier, France Université de Strasbourg, CHRU de Strasbourg, Memory Resource and Research Centre of Strasbourg/Colmar, Inserm U-­118, F-­67000 Strasbourg, France. [30] Univ Twitter Sylvain Lehmann @sylvain_lehmann Paris Est Creteil, EA 4391 Excitabilité Nerveuse et Thérapeutique, CHU Henri Mondor, Department of Neurology, F-­94000 Créteil, France. [31] Fondation ophtalmologique Collaborators BALTAZAR study groupOlivier Hanon [1], Frédéric Blanc [2], Yasmina Adolphe de Rothschild, Department of Neurology, F-­75019 Paris, France. [32] Boudali [1], Audrey Gabelle [3], Jacques Touchon [3], Marie-­Laure Seux [1], Hermine CHU d’Amiens-­Picardie, Department of Gerontology, F-­80000 Amiens, France. Lenoir [1], Catherine Bayle [1], Stéphanie Bombois [4], Christine Delmaire [4], Xavier [33] Sorbonne Universités, APHP, Hôpitaux Universitaires Pitie-­Salpêtrière-­Charles Delbeuck [5], Florence Moulin [1], Emmanuelle Duron [6], Florence Latour [7], Foix, Memory Resource and Research Centre, Centre des Maladies Cognitives et Matthieu Plichart [1], Sophie Pichierri [8], Galdric Orvoën [1], Evelyne Galbrun [9], Comportementales IM2A, Inserm UMR 8256, F-­75013 Paris, France. [34] Université Giovanni Castelnovo [10], Lisette Volpe-­Gillot [11], Florien Labourée [1], Pascaline François-­Rabelais de Tours, CHRU de Tours, Memory Resource and Research Centre Cassagnaud [12], Claire Paquet [13], Françoise Lala [14], Bruno Vellas [14], Julien of Tours, Inserm CIC 1415, F-­37000 Tours, France. [35] Sorbonne Universités, Dumurgier [13], Anne-­Sophie Rigaud [1], Christine Perret-­Guillaume [15], Eliana APHP, Hôpitaux Universitaires Pitie-­Salpêtrière-­Charles Foix, Memory Resource Alonso [16], Foucaud du Boisgueheneuc [17], Laurence Hugonot-­Diener [1], Adeline and Research Centre, Centre des Maladies Cognitives et Comportementales Rollin-­Sillaire [12], Olivier Martinaud [18], Clémence Boully [1], Yann Spivac [19], IM2A, F-­75013 Paris, France. [36] Université Bourgogne Franche-­Comté, EA 481 Agnès Devendeville [20], Joël Belmin [21], Philippe-­Henri Robert [22], Thierry Neuroscience, IFR 133, CHU de Besançon, Memory Resource and Research Centre Dantoine [23], Laure Caillard [1], David Wallon [24], Didier Hannequin [18], Nathalie of BesançonFranche-­Comté, F-­25000 Besançon, France. [37] Université François-­ Sastre [14], Sophie Haffen [25], Anna Kearney-­Schwartz [15], Jean-­Luc Novella Rabelais de Tours, CHRU de Tours, Memory Resource and Research Centre of [26], Vincent Deramecourt [12], Valérie Chauvire [27], Gabiel Abitbol [1], Nathalie Tours, Inserm U930 Imagerie et Cerveau, F-­37000 Tours, France. [38] Université de Schwald [19], Caroline Hommet [28], François Sellal [29], Marie-­Ange Cariot [16], Limoges, EA 6310 HAVAE, CHU de Limoges, Memory Research Resource Center of Mohamed Abdellaoui [30], Sarah Benisty [31], Salim Gherabli [1], Pierre Anthony Limoges, F-­87000 Limoges, France. [39] Université Paris-­Saclay, Neurospin, CEA, ​ [29], Frédéric Bloch [32], Nathalie Charasz [1], Sophie Chauvelier [1], Jean-­Yves cati-­​neuroimaging.​com, CATI Multicenter Neuroimaging Platform, F-­91190 Gif-­sur-­ Gaubert [1], Guillaume Sacco [22], Olivier Guerin [22], Jacques Boddaert [33], Yvette, France. [40] LBPC-­PPC, Université de Montpellier, INM INSERM, IRMB CHU Marc Paccalin [17], Marie-­Anne Mackowiak [12], Marie-­Thérèse Rabus [9], Valérie de Montpellier, Montpellier, France. [41] Université de Paris, Inserm UMR-­S 894, Gissot [34], Athanase Benetos [15], Candice Picard [20], Céline Guillemaud [35], F-­75014 Paris, France. [42] Univ. Lille, Inserm, CHU Lille, U1172-­LilNCog, LiCEND, Gilles Berrut [8], Claire Gervais [22], Jaques Hugon [13], Jean-­Marc Michel [29], LabEx DISTALZ, F-­59000 Lille, France. JeanPhilippe David [19], Marion Paulin [12], Pierre-­Jean Ousset [14], Pierre Vandel [36], Sylvie Pariel [21], Vincent Camus [37], Anne Chawakilian [1], Léna Kermanac’h Contributors SL, J-­SV and OH take responsibility for the integrity of the data [1], Anne-­Cécile Troussiere [12], Cécile Adam [23], Diane Dupuy [20], Elena Paillaud and the accuracy of the data analysis. Concept and design: OH, SB, AG, SS-­M [16], Hélène Briault [9], Isabelle Saulnier [38], Karl Mondon [37], Marie-­Agnès and SL. Acquisition, analysis, or interpretation of data: All authors. Drafting of the Picat [23], Marie Laurent [16], Olivier Godefroy [20], Rezki Daheb [16], Stéphanie manuscript: SL. Critical revision of the manuscript for important intellectual content: Libercier [29], Djamila Krabchi [1], Marie Chupin [39], Luc Buée [42], JeanSébastien All authors.Statistical analysis: SL and J-­SV. Obtained funding: OH, SB, AG, SS-­M and Vidal [1], Edouard Chaussade [1], Sylvain Lehmann [40], Bernadette Allinquant [41], SL. All authors had full access to the data and contributed to revision and editing of Susanna Schraen-­Maschke [42]. Affiliations of the BALATAZAR group[1] Université the manuscript. SL is responsible for the overall content as the guarantor. de Paris, EA 4468, APHP, Hopital Broca, Memory Resource and Research Centre of Funding The French ministry of Health (Programme Hospitalier de Recherche de Paris-­Broca-­Ile de France, F-­75013 Paris, France. [2] Université de Strasbourg, Clinique), Grant/Award Numbers:PHRC2009/01-­04,PHRC-­13-­0404; The Foundation CHRU de Strasbourg, Memory Resource and Research Centre of Strasbourg/ Plan Alzheimer; Fondation pour la Recherche Médicale (FRM); The Gerontopôle d’Ile Colmar, French National Centre for Scientific Research (CNRS), ICube Laboratory de France (GEROND’IF). None of the funding bodies had any role in study design, in and Fédération de Médecine Translationnelle de Strasbourg (FMTS), Team Imagerie the collection, analysis, and interpretation of data, in the writing of the report or in Multimodale Intégrative en Santé (IMIS)/Neurocrypto, F-­67000 Strasbourg, France. the decision to submit the paper for publication. [3] Université de Montpellier, CHRU Montpellier, Memory Research and Resources Competing interests None declared. center, department of Neurology, Inserm INM NeuroPEPs team, excellence center of neurodegenerative disorders, F-­34000 Montpellier, France. [4] Univ. Lille, CHU Lille, Patient consent for publication Not applicable. Memory Resource and Research Centre of Lille, Department of Neurology, Inserm Ethics approval This study involves human participants and was approved by The U1171, Degenerative and Vascular Cognitive Disorders, F-­59000 Lille, France. [5] study was approved by the Paris EthicsCommittee (CPP Ile de France IV Saint Louis ​ Univ. Lille, Inserm U1171 Degenerative and Vascular Cognitive Disorders, F-­59000 Hospital).​ClinicalTrials.​gov Identifier#NCT01315639 Participants gave informed Lille, France. [6] Université Paris-­Saclay, APHP, Hôpital Paul Brousse, département consent to participate in the study before taking part. de gériatrie, UVSQ, Inserm, CESP, Team MOODS, F-­94800 Villejuif, France. [7] Centre Hospitalier de la Côte Basque, Department of Gerontology, F-­64100 Bayonne, Provenance and peer review Not commissioned; externally peer reviewed. France. [8] Université de Nantes, EA 4334 Movement-­Interactions-­Performance, Data availability statement Data are available upon reasonable request. CHU Nantes, Memory Research Resource Center of Nantes, F-­44000 Nantes, France. Data and informed consent form are available upon request after publication [9] Sorbonne Universités, AP-­HP, Centre Hospitalier Émile-­Roux, Department of (APHP, Paris). Requests will be considered by each study investigator based on the Gérontology 2, F-­94450 Limeil-­Brévannes, France. [10] CHU de Nimes, Neurology information provided by the requester regarding the study and analysis plan. If the Department, F-­30000 Nimes, France. [11] Hôpital Léopold Bellan, Memory Clinic, use is appropriate, a data sharing agreement will be put in place before distributing F-­75014 Paris, France. [12] Univ. Lille, CHU de Lille, Memory Resource and Research a fully de-­identified version of the dataset, including the data dictionary used for Centre of Lille, Department of Neurology, F-­59000 Lille, France. [13] Université analysis with individual participant data. de Paris, APHP, Groupe Hospitalier Saint Louis-­LariboisièreFernand Widal, Memory Resource and Research Centre of Paris Nord-­Ile de France, F-­75010 Paris, France. Supplemental material This content has been supplied by the author(s). It [14] Université de Toulouse III, CHU La Grave-­Casselardit, Memory Resource and has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have Research Centre of Midi-­Pyrénées, F-­31300 Toulouse, France. [15] Université de been peer-­reviewed. Any opinions or recommendations discussed are solely those Lorraine, CHRU de Nancy, Memory Resource and Research Centre of Lorraine, of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and F-­54000 Nancy, France. [16] AP-­HP, HEGP, Service de Gériatrie, F-­75015, Paris, responsibility arising from any reliance placed on the content. Where the content France. [17] CHU de Poitiers, Memory Resource and Research Centre of Poitiers, includes any translated material, BMJ does not warrant the accuracy and reliability F-­86000 Poitiers, France. [18] CHU Charles Nicolle, Memory Resource and Research of the translations (including but not limited to local regulations, clinical guidelines, Centre of HauteNormandie, F-­76000 Rouen, France. [19] APHP, Centre Hospitalier terminology, drug names and drug dosages), and is not responsible for any error Émile-­Roux, Department of Gérontology 1, F-­94450 Limeil-­Brévannes, France. [20] and/or omissions arising from translation and adaptation or otherwise. CHU d’Amiens-­Picardie, Memory Resource and Research Centre of Amiens Picardie, Open access This is an open access article distributed in accordance with the F-­80000 Amiens, France. [21] APHP, Hôpitaux Universitaires Pitie-­Salpêtrière-­ Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which Charles Foix, Service de Gériatrie Ambulatoire, F-­75013 Paris, France. [22] CHU permits others to distribute, remix, adapt, build upon this work non-­commercially, de Nice, Memory Research Resource Center of PACA Est, F-­06100 Nice, France. and license their derivative works on different terms, provided the original work is 418 Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 Neurodegeneration properly cited, appropriate credit is given, any changes made indicated, and the use 16 Hanon O, Vidal Jean‐Sébastien, Lehmann S, et al. Plasma amyloid levels within the is non-­commercial. See: http://creativecommons.org/licenses/by-nc/4.0/. Alzheimer’s process and correlations with central biomarkers. Alzheimer's & Dementia 2018;14:858–68. ORCID iDs 17 Petersen RC, Smith GE, Waring SC, et al. Mild cognitive impairment: clinical Sylvain Lehmann http://orcid.org/0000-0001-6117-562X characterization and outcome. Arch Neurol 1999;56:303–8. Jean-­Sébastien Vidal http://orcid.org/0000-0001-6770-0720 18 Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular Constance Delaby http://orcid.org/0000-0002-8606-6814 filtration rate. Ann Intern Med 2009;150:604–12. Olivier Hanon http://orcid.org/0000-0002-4697-122X 19 Lehmann S, Delaby C, Boursier G, et al. Relevance of Aβ42/40 ratio for detection of Alzheimer disease pathology in clinical routine: the PLMR scale. Front Aging Neurosci 2018;10:138. 20 Rissin DM, Walt DR. Digital concentration readout of single enzyme molecules using REFERENCES femtoliter arrays and Poisson statistics. Nano Lett 2006;6:520–3. 1 Lee M, Whitsel E, Avery C, et al. Variation in population attributable fraction of 21 Pencina MJ, D’Agostino RB, D’Agostino RB, et al. Evaluating the added predictive dementia associated with potentially modifiable risk factors by race and ethnicity in ability of a new marker: from area under the ROC curve to reclassification and the US. JAMA Netw Open 2022;5:e2219672. beyond. Stat Med 2008;27:157–72. 2 Piton M, Hirtz C, Desmetz C, et al. Alzheimer’s disease: advances in drug development. 22 DeLong ER, DeLong DM, Clarke-­Pearson DL. Comparing the areas under two or J Alzheimers Dis 2018;65:3–13. more correlated receiver operating characteristic curves: a nonparametric approach. 3 Jack CR, Bennett DA, Blennow K, et al. NIA‐AA Research Framework: Toward a Biometrics 1988;44:837–45. biological definition of Alzheimer’s disease. Alzheimer's & Dementia 2018;14:535–62. 23 Delaby C, Alcolea D, Hirtz C, et al. Blood amyloid and tau biomarkers as predictors of 4 Dumurgier J, Schraen S, Gabelle A, et al. Cerebrospinal fluid amyloid-β 42/40 ratio cerebrospinal fluid profiles. J Neural Transm 2022;129:231–7. in clinical setting of memory centers: a multicentric study. Alzheimers Res Ther 24 Thijssen EH, La Joie R, Strom A, et al. Plasma phosphorylated tau 217 and 2015;7:30. phosphorylated tau 181 as biomarkers in Alzheimer’s disease and frontotemporal 5 Gabelle A, Dumurgier J, Vercruysse O, et al. Impact of the 2008-­2012 French lobar degeneration: a retrospective diagnostic performance study. The Lancet Alzheimer plan on the use of cerebrospinal fluid biomarkers in research memory Neurology 2021;20:739–52. center: the PLM study. J Alzheimers Dis 2013;34:297–305. 25 Bayoumy S, Verberk IMW, den Dulk B, et al. Clinical and analytical comparison of 6 Hanon O, Vidal Jean‐Sébastien, Lehmann S, et al. Plasma amyloid beta predicts six Simoa assays for plasma p-­tau isoforms p-­Tau181, P-­tau217, and P-­tau231. conversion to dementia in subjects with mild cognitive impairment: the BALTAZAR Alzheimers Res Ther 2021;13:198. study. Alzheimer's & Dementia 2022 (published Online First: 2022/02/22). 26 Thijssen EH, La Joie R, Wolf A, et al. Diagnostic value of plasma phosphorylated 7 Zetterberg H, Wilson D, Andreasson U, et al. Plasma tau levels in Alzheimer’s disease. tau181 in Alzheimer’s disease and frontotemporal lobar degeneration. Nat Med Alzheimers Res Ther 2013;5:9. 2020;26:387–97. 8 Mielke MM, Hagen CE, Xu J, et al. Plasma phospho‐tau181 increases with Alzheimer’s 27 Benussi A, Karikari TK, Ashton N, et al. Diagnostic and prognostic value of serum NfL disease clinical severity and is associated with tau‐ and amyloid‐positron emission and p-­Tau 181 in frontotemporal lobar degeneration. J Neurol Neurosurg Psychiatry tomography. Alzheimer's & Dementia 2018;14:989–97. 2020;91:960–7. 9 Karikari TK, Pascoal TA, Ashton NJ, et al. Blood phosphorylated tau 181 as a 28 Benkert P, Meier S, Schaedelin S, et al. Serum neurofilament light chain for individual biomarker for Alzheimer’s disease: a diagnostic performance and prediction prognostication of disease activity in people with multiple sclerosis: a retrospective modelling study using data from four prospective cohorts. The Lancet Neurology modelling and validation study. The Lancet Neurology 2022;21:246–57. 2020;19:422–33. 29 Blomberg M, Jensen M, Basun H, et al. Cerebrospinal fluid tau levels increase with 10 McGrath ER, Beiser AS, O’Donnell A, et al. Blood phosphorylated tau 181 as a age in healthy individuals. Dement Geriatr Cogn Disord 2001;12:127–32. biomarker for amyloid burden on brain PET in cognitively healthy adults. J Alzheimers 30 Syrjanen JA, Campbell MR, Algeciras‐Schimnich A, et al. Associations of amyloid and Dis 2022;87:1517–26. neurodegeneration plasma biomarkers with comorbidities. Alzheimer's & Dementia 11 Karikari TK, Benedet AL, Ashton NJ, et al. Diagnostic performance and prediction 2022;18:1128–40. of clinical progression of plasma phospho-­tau181 in the Alzheimer’s Disease 31 Mielke MM, Dage JL, Frank RD, et al. Performance of plasma phosphorylated tau 181 Neuroimaging Initiative. Mol Psychiatry 2021;26:429–42. and 217 in the community. Nat Med 2022;28:1398–405. 12 Clark C, Lewczuk P, Kornhuber J, et al. Plasma neurofilament light and phosphorylated 32 Janelidze S, Bali D, Ashton NJ, et al. Head-­To-­Head comparison of 10 plasma tau 181 as biomarkers of Alzheimer’s disease pathology and clinical disease phospho-­tau assays in prodromal Alzheimer’s disease. Brain 2022. doi:10.1093/brain/ progression. Alzheimers Res Ther 2021;13:65. awac333. [Epub ahead of print: 10 Sep 2022]. 13 Chen S-­D, Huang Y-­Y, Shen X-­N, et al. Longitudinal plasma phosphorylated tau 181 33 Singh-­Manoux A, Oumarou-­Ibrahim A, Machado-­Fragua MD, et al. Association tracks disease progression in Alzheimer’s disease. Transl Psychiatry 2021;11:356. between kidney function and incidence of dementia: 10-­year follow-­up of the 14 Hansson O, Edelmayer RM, Boxer AL, et al. The Alzheimer’s Association Appropriate Whitehall II cohort study. Age Ageing 2022;51. Use Recommendations for Blood Biomarkers in Alzheimer’s Disease. Alzheimer's & 34 Chan H-­N, Xu D, Ho S-­L, et al. Ultra-­Sensitive detection of protein biomarkers for Dementia 2022;18. diagnosis of Alzheimer’s disease. Chem Sci 2017;8:4012–8. 15 Teunissen CE, Verberk IMW, Thijssen EH, et al. Blood-­based biomarkers for Alzheimer’s 35 Sun R, Wang H, Shi Y, et al. A Pilot Study of Urinary Exosomes in Alzheimer’s Disease. disease: towards clinical implementation. The Lancet Neurology 2022;21:66–77. Neurodegener Dis 2020;19:184–91. Lehmann S, et al. J Neurol Neurosurg Psychiatry 2023;94:411–419. doi:10.1136/jnnp-2022-330540 419