Prediction of future Alzheimer's disease dementia using plasma phospho-tau combined with other accessible measures.

TLDR: Plasma P-tau in combination with brief cognitive tests and APOE genotyping may greatly improve the diagnostic prediction of AD dementia and facilitate recruitment for AD trials.

Abstract: A combination of plasma phospho-tau (P-tau) and other accessible biomarkers might provide accurate prediction about the risk of developing Alzheimer’s disease (AD) dementia. We examined this in participants with subjective cognitive decline and mild cognitive impairment from the BioFINDER (n = 340) and Alzheimer’s Disease Neuroimaging Initiative (ADNI) (n = 543) studies. Plasma P-tau, plasma Aβ42/Aβ40, plasma neurofilament light, APOE genotype, brief cognitive tests and an AD-specific magnetic resonance imaging measure were examined using progression to AD as outcome. Within 4 years, plasma P-tau217 predicted AD accurately (area under the curve (AUC) = 0.83) in BioFINDER. Combining plasma P-tau217, memory, executive function and APOE produced higher accuracy (AUC = 0.91, P < 0.001). In ADNI, this model had similar AUC (0.90) using plasma P-tau181 instead of P-tau217. The model was implemented online for prediction of the individual probability of progressing to AD. Within 2 and 6 years, similar models had AUCs of 0.90–0.91 in both cohorts. Using cerebrospinal fluid P-tau, Aβ42/Aβ40 and neurofilament light instead of plasma biomarkers did not improve the accuracy significantly. The clinical predictions by memory clinic physicians had significantly lower accuracy (4-year AUC = 0.71). In summary, plasma P-tau, in combination with brief cognitive tests and APOE genotyping, might greatly improve the diagnostic prediction of AD and facilitate recruitment for AD trials. Plasma P-tau, in combination with clinical measures, predicts future Alzheimer’s disease dementia in two independent cohorts with high accuracy and is superior to the clinical diagnostic predictions of specialists.

Figures

Figure 2. Comparison with the clinical prediction for predicting AD dementia within 4 years. A, Models from Fig. 1 compared to the clinical diagnostic prediction of memory clinic physicians at the baseline visit (blinded to fluid biomarker data). B, ROC curves of the different models. Note that comparison with the clinical prediction was performed on a subsample where the clinical prediction was available (n=285), hence the slightly different AUCs (and 95% CIs) compared with those shown in Fig. 1. Abbreviations: AD, Alzheimer’s disease; AIC, Akaike Information Criterion; APOE, Apolipoprotein E genotype (number of ε4 alleles); AUC, Area under the ROC curve; Exec. function, Executive Function; MRI, Cortical thickness of a temporal AD-specific region; ROC, Receiver Operating Characteristic

Figure 4. A. Cross-validation of the logistic regression model using plasma P-tau status instead of continuous plasma P-tau levels. Model coefficient were established in BioFINDER (AUC 0.89) and tested in ADNI (AUC 0.86). Z-scores have been inverted so that higher scores equal poorer results. B. Implementation of the logistic regression model at https://brainapps.shinyapps.io/PredictionADdementia/ where one can enter the raw cognitive test scores that constitute the z-

Table 1. Baseline characteristics for the BioFINDER and ADNI cohorts

Figure 1. Model selection process and performance for predicting AD dementia within 4 years. A, The model selection process. Best Model Fit shows the data-driven model selection with the lowest AIC (i.e. the best model fit). The parsimonious model shows the model that had a similar model fit (ΔAIC <2) with as few significant predictors as possible. In subsequent models, modalities were removed in a step-wise procedure. Model specifications including comparisons between all models are shown in Supplementary Table 1. B, ROC curves of the different models. Abbreviations: AD, Alzheimer’s disease; AIC, Akaike Information Criterion; APOE, Apolipoprotein E genotype (number of

Figure 3. Model selection and performance in ADNI for predicting AD dementia within 4 years with comparisons using models selected in BioFINDER. A, Model selection process. Best Model Fit shows the data-driven model selection with the lowest AIC in the ADNI cohorts. This was compared to the models selected in BioFINDER and applied in ADNI. Model specifications are shown in Supplementary Table 8. B, ROC curves of the different models. Abbreviations: AD, Alzheimer’s disease; AIC, Akaike Information Criterion; APOE, Apolipoprotein E genotype (number of ε4 alleles); AUC, Area under the ROC curve; Exec. function, Executive Function; MRI, Cortical thickness of a temporal AD-specific region; ROC, Receiver Operating Characteristic

Full text

Prediction of future Alzheimer’s disease dementia
using plasma phospho-tau combined with other
accessible measures
Sebastian Palmqvist (  sebastian.palmqvist@med.lu.se )
Lund University
Pontus Tideman
Lund University
Nicholas Cullen
Lund University
Henrik Zetterberg
University of Gothenburg https://orcid.org/0000-0003-3930-4354
Kaj Blennow
University of Gothenburg https://orcid.org/0000-0002-1890-4193
Jeffrey Dage
Eli Lilly and Company https://orcid.org/0000-0002-2192-1699
Erik Stomrud
Lunds University
Shorena Janelidze
Lund University https://orcid.org/0000-0003-2869-8378
Niklas Mattsson-Carlgren
Lund University
Oskar Hansson
Lund University https://orcid.org/0000-0001-8467-7286
Article
Keywords: Biomarkers, Subjective Cognitive Decline, Mild Cognitive Impairment, Logistic Regression
Models, Memory, Executive Function, APOE Genotyping
Posted Date: January 5th, 2021
DOI: https://doi.org/10.21203/rs.3.rs-134146/v1
License:   This work is licensed under a Creative Commons Attribution 4.0 International License. 
Read Full License

Version of Record: A version of this preprint was published at Nature Medicine on May 24th, 2021. See
the published version at
https://doi.org/10.1038/s41591-021-01348-z.

1
Prediction of future Alzheimer’s disease dementia using
plasma phospho-tau combined with other accessible
measures
Sebastian Palmqvist, MD, PhD
1,2*
, Pontus Tideman, MSc
1,2
, Nicholas Cullen, MSc
1
, Henrik
Zetterberg, MD, PhD
3,4,5,6
, Kaj Blennow, MD, PhD
3,4
, for the Alzheimer’s Disease
Neuroimaging Initiative, Jeffery L. Dage, PhD
7
, Erik Stomrud, MD, PhD
1,2
, Shorena
Janelidze, PhD
1
,
Niklas Mattsson-Carlgren, MD, PhD
1,8,9
,
Oskar Hansson MD, PhD
1,2*
1
Clinical Memory Research Unit, Department of Clinical Sciences, Lund University, Sweden.
2
Memory Clinic, Skåne University Hospital, Sweden.
3
Department of Psychiatry and Neurochemistry, the Sahlgrenska Academy at the University
of Gothenburg, Mölndal, Sweden.
4
Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden.
5
Department of Neurodegenerative Disease, UCL Institute of Neurology, Queen Square,
London, United Kingdom.
6
UK Dementia Research Institute at UCL, London, United Kingdom.
7
Eli Lilly and Company, Indianapolis IN 46285 U.S.A.
8
Department of Neurology, Skåne University Hospital, Sweden.
9
Wallenberg Center for Molecular Medicine, Lund University, Lund, Sweden
* Corresponding authors
Oskar Hansson
e-mail:
oskar.hansson@med.lu.se
Sebastian Palmqvist
e-mail: sebastian.palmqvist@med.lu.se
Both at
Memory Clinic
Skåne University Hospital
205 02 Malmö
Sweden

2
One sentence summary
Plasma P-tau in combination with brief and accessible measures predict future AD dementia
in two independent cohorts with high accuracy that is clearly superior to specialists’ clinical
diagnostic prediction.
ABSTRACT
A combination of plasma phospho-tau (P-tau) and other accessible biomarkers may provide
accurate prediction about the risk of developing Alzheimer’s disease (AD) dementia. We
examined this in patients with subjective cognitive decline (SCD) and mild cognitive
impairment (MCI) who were consecutively recruited in the BioFINDER study (n=340). The
results were validated in SCD/MCI participants in the ADNI study (n=543). Plasma P-tau,
plasma Aβ42/Aβ40, plasma neurofilament light, APOE genotype, brief cognitive tests and
MRI (cortical thickness in AD-specific regions) were examined as predictors of progression
to AD dementia primarily within 4 years. The accuracy was determined using the area under
the ROC curve (AUC) from logistic regression models. Within 4 years, plasma P-tau217
predicted AD dementia accurately (AUC 0.83) in BioFINDER. A model of plasma P-tau217,
memory, executive function, and APOE had higher accuracy (AUC 0.91, p<0.001). In ADNI,
this model produced a similar AUC (0.90) using plasma P-tau181 instead of P-tau217. A
cross-validated version of this model was implemented online for prediction of the individual
probability of progressing to AD dementia. Within 2 and 6 years, parsimonious models
performed similar in both cohorts (AUCs 0.90-0.91). Using cerebrospinal fluid P-tau,
Aβ42/Aβ40 and neurofilament light instead of plasma biomarkers did not improve the
accuracy. The clinical prediction by memory clinic physicians had significantly lower
accuracy than all models (4-year AUC 0.71). In summary, plasma P-tau in combination with
brief cognitive tests and APOE genotyping may greatly improve the diagnostic prediction of
AD dementia and facilitate recruitment for AD trials.

3
INTRODUCTION
Correctly determining if a patient with subtle cognitive symptoms, such as memory decline,
suffers from prodromal or preclinical Alzheimer’s disease (AD) and will progress to AD
dementia within the near future remains a challenge for clinicians. The task is nonetheless of
utmost importance for a timely referral to a memory clinic, a correct and early AD diagnosis,
initiation of symptomatic treatment, planning for the future, and hopefully soon for initiating
disease-modifying treatments. Although there have been impressive developments in
biomarkers for AD and progression to AD dementia, such as cerebrospinal fluid analysis of β-
amyloid (Aβ42
1
or the ratio of Aβ42/Aβ40
2
), phosphorylated tau (P-tau)
3,4
and neurofilament
light (NfL)
5
, as well as Aβ-PET
6,7
and tau-PET
8,9
, the invasive nature, high cost, and limited
availability restrict their use to a limited number of highly specialized centers. A possible
turning point has emerged with the recent development of blood-based biomarkers, making it
possible to measure NfL
10,11
, Aβ42/Aβ40
2,12,13
, and P-tau (either phosphorylated at threonine
181 or 217)
14-16
in plasma.
Plasma P-tau181 and P-tau217 in particular have shown especially high diagnostic
performance for discriminating AD dementia from other neurodegenerative diseases.
14-16
Plasma P-tau has also recently been shown to be suitable for individualized prediction of
cognitive decline in individuals with mild cognitive impairment (MCI).
17
In the clinical work-
up of patients with cognitive complaints, however, it is unlikely that plasma P-tau (or any
other biomarker) will achieve the highest potential predictive accuracy on its own due to the
multifactorial nature of AD etiology and its heterogenous clinical presentation. There is
therefore now a need to identify which other measures plasma P-tau should be combined with
to produce the most accurate prediction of future AD and establish an optimal diagnostic
algorithm of non-invasive, cost-effective and easily available methods for early diagnosis of
AD.