Plasma amyloid-β biomarkers for predicting amyloid positivity in mild cognitive impairment: a prospective cohort study

doi:10.21203/rs.3.rs-4361497/v1

Download PDF

Research Article

Plasma amyloid-β biomarkers for predicting amyloid positivity in mild cognitive impairment: a prospective cohort study

https://doi.org/10.21203/rs.3.rs-4361497/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: This study aimed to determine the predictive accuracy of amyloid-β precursor protein 669-711/amyloid-β1-42, amyloid-β1-40/1-42, and their composite biomarkers for brain amyloid deposition or the development of Alzheimer’s disease (AD) dementia in community-dwelling older adults with mild cognitive impairment(MCI).

Methods: This was a prospective cohort study conducted from August 2015 to September 2019. Blood samples were collected from older adults aged ≥65 years with MCI without dementia at baseline. Cognitive function and amyloid deposition were evaluated annually. Positron emission tomography was used to evaluate amyloid deposition. Plasma amyloid-β biomarkers were analyzed using immunoprecipitation-mass spectrometry. The classification accuracy of plasma biomarkers for brain amyloid status was evaluated using receiver operating characteristic curve analysis. Associations of plasma biomarkers with clinical conversion to AD dementia were evaluated using Kaplan‒Meier curves. Associations of plasma biomarkers with body mass index, apolipoprotein E4 status, cortical amyloid uptake, medical history, and hepatic or renal dysfunction were examined using multiple linear regression analysis.

Results: The participants included 107 patients (57 [53.3%] females, median age: 76.00 [72.00–80.00] years). Plasma biomarkers correlated with cortical amyloid uptake. The composite biomarker had the best area under the curve (0.943) for predicting amyloid positivity. Participants with high levels of composite biomarkers at baseline had a greater rate of conversion to AD dementia. Apolipoprotein E4 status and hepatic disorders were significantly correlated with increased plasma amyloid biomarker levels.

Conclusions: Composite amyloid biomarkers may serve as a surrogate measure for amyloid deposition in a community-based cohort.

Trial Registration: UMIN000017442

amyloid positivity

mild cognitive impairment

plasma amyloid-β biomarkers

prospective cohort study

The prevalence rate of mild cognitive impairment (MCI) in Japanese adults aged 65 years and older is 17.0%, and 10–34% of adults with amnestic MCI develop Alzheimer’s disease (AD) dementia annually [1]. Recently, the Food and Drug Administration approved a new disease-modifying therapy targeting amyloid-β for adults with MCI or mild dementia due to AD [2, 3]. Therefore, accurate detection of amyloid pathology in individuals with MCI is crucial for enhancing the benefit of disease-modifying therapies in future clinical practice.

Although detecting amyloid-β levels in the cerebrospinal fluid and performing positron emission tomography (PET) are well-established methods for detecting brain amyloid deposition, PET is expensive and not widely available. Cerebrospinal fluid analysis further requires invasive lumbar puncture. Additionally, the amyloid positivity rate in community-based MCI cohorts was lower than that in memory clinic-based MCI cohorts [4]. Therefore, these methods are unsuitable as screening tools in clinical settings owing to their high costs. In the future, widely available and minimally invasive blood-based biomarkers for AD are required for prescreening amyloid-positive MCI patients [5, 6] and augmenting PET or cerebrospinal fluid analysis [7, 8]. Recent studies have shown that advanced technologies, such as mass spectrometry [9–11] and immunoassays [12, 13], can measure plasma amyloid-β with high precision and reproducibility. However, only a few studies have examined the predictive accuracy of plasma amyloid-β biomarkers in community-based cohorts [14–17]. Community-based cohorts are more diverse than clinical cohorts in terms of demographic characteristics, lifestyles, and comorbidities. The levels of amyloid-β precursor protein (APP) 669–711/amyloid-β1–42, amyloid-β1–40/1–42, and their composite biomarkers measured using immunoprecipitation-mass spectrometry have demonstrated the ability to distinguish between amyloid-positive and -negative adults in clinical settings [9]. Therefore, this study aimed to determine the predictive accuracy of these amyloid-β biomarkers for amyloid positivity on PET in individuals with MCI recruited from a community-based cohort. Our results suggest the general applicability of amyloid-β biomarkers from clinical to community settings. Moreover, we assessed the predictive ability of a baseline composite biomarker for clinical progression to AD dementia with 8 years of follow-up.

Participants

The Usuki study was a prospective cohort study in Usuki, Oita Prefecture, Japan, from August 2015 to September 2019 conducted to determine lifestyle risk factors for dementia or imaging biomarkers of AD [18]. This study included 118 adults with MCI aged ≥ 65 years. MCI was diagnosed according to a global rating of 0.5 on the Clinical Dementia Rating Scale. All participants underwent blood sampling at baseline and annual evaluations of cognitive function and amyloid PET at Oita University Hospital. Trained medical staff collected demographic information, including age, sex, years of education, body mass index (BMI), and medical history. Cognitive function was assessed using the Mini-Mental State Examination, the Japanese version of the Montreal Cognitive Assessment, and the Wechsler Memory Scale-Revised logical memory II test. Liver and renal functions were assessed by measuring alanine aminotransferase, aspartate aminotransferase, and γ-glutamyl transpeptidase levels and the estimated glomerular filtration rate, and plasma amyloid-β biomarker levels in blood samples were measured. Eleven plasma samples were excluded from the current study owing to analytic failure of amyloid-β biomarkers; therefore, the final study cohort included 107 participants who underwent assessment of amyloid-β biomarkers and PET. No participants were taking medication for dementia at baseline. Moreover, we collected follow-up data regarding dementia diagnosis, determined by a neurologist according to cognitive and clinical data or medication for AD from November to December 2023.

Immunoprecipitation-mass spectrometry

Blood samples were collected during morning hours following an overnight fast. After centrifugation (1,800 × g for 10 min at 4°C), the plasma was separated and stored at -80°C until use. Plasma levels of amyloid-β1–40, amyloid-β1–42, and APP669-711 were measured using immunoprecipitation-mass spectrometry [10]. The amyloid-β1–40/1–42 and APP669-711/amyloid-β1–42 ratios were generated by calculating the ratios of the normalized intensities of amyloid-β1–40 and APP669-711 to that of amyloid-β1–42, respectively.

The composite biomarker was computed by averaging the normalized scores of amyloid-β1–40/1–42 and APP669-711/amyloid-β1–42, as previously reported [9].

PET

¹¹C-Pittsburgh compound-B-PET was conducted with a Biograph mCT PET/CT scanner (Siemens, Erlangen, Germany). A 20-min static PET image was acquired 50 min after an intravenous bolus of 543 ± 57 MBq ¹¹C-Pittsburgh compound-B was injected with a saline flush. The standardized uptake value ratio was calculated for the evaluation of ¹¹C-Pittsburgh compound-B uptake using the following methods. Statistical Parametric Mapping Version 8 (Wellcome Trust Center for Neuroimaging, London, UK) implemented in MATLAB 7.9.0. (R2009b; MathWorks, Natick, MA) was used for spatial normalization of PET images to a customized PET template in the Montreal Neurological Institute reference space. The standardized uptake value ratio for ¹¹C-Pittsburgh compound-B-PET was calculated as the ratio of the voxel number-weighted average of the mean uptake in the frontal, temporoparietal, and posterior cingulate cortices to that in the cerebellar cortex. The global mean standardized uptake value ratio combined single mean values for all regions. ¹¹C-Pittsburgh compound-B PET positivity was defined according to the global cortical standardized uptake value ratio of ≥1.2 [19].

Apolipoprotein E isoform

A human apolipoprotein E4/panapolipoprotein E (ApoE) enzyme-linked immunosorbent assay (ELISA) kit (MBL Co., Woburn, MA) was used for ApoE phenotyping, similar to in a previous study [18]. This kit can identify individuals with a ratio of ApoE4 to ApoE of ≥ 0.3 to have at least one APOE ε4 allele.

Statistical analysis

All participants were classified into amyloid-β-negative (n = 71) and amyloid-β-positive (n = 36) subgroups according to the standardized uptake value ratio cutoff of ≥ 1.2. Sex, ApoE4 status, and medical history were compared using the χ² test, and age, education level, BMI, Mini-Mental State Examination score, Japanese version of the Montreal Cognitive Assessment score, Wechsler Memory Scale-Revised logical memory II test score, cortical ¹¹C-Pittsburgh compound-B uptake values, aspartate aminotransferase, alanine aminotransferase, and γ-glutamyl transpeptidase levels, estimated glomerular filtration rate, and plasma amyloid-β biomarkers were compared using the Mann–Whitney U test between subgroups. Correlations between plasma amyloid-β biomarkers and cortical ¹¹C-Pittsburgh compound-B uptake were assessed using Spearman’s correlation coefficients. Moreover, we conducted a voxel wise linear regression analysis using Statistical Parametric Mapping 8 to determine the spatial association between plasma amyloid-β biomarkers and brain amyloid deposition.

Logistic regression with receiver operating characteristic curve analysis

The accuracy of plasma-β biomarkers for predicting amyloid positivity and AD conversion status was assessed using area under the receiver operating characteristic curve values within a binary logistic regression model. Amyloid positivity was indicated by amyloid positivity on PET. AD conversion status categorized participants based on whether they did or did not convert to dementia. In both analyses, the plasma biomarker level served as an independent variable. The other models included plasma biomarkers with age, sex, and ApoE4 status to examine the influence of additional covariates. There were no variables with collinearity or multicollinearity. The areas under the curves were compared using the DeLong test. The best cutoff value for discriminating between the amyloid-positive and amyloid-negative subgroups was determined as the value with the greatest sensitivity and specificity.

Reweighting for 60% prevalence of amyloid positivity

We estimated the negative predictive value and positive predictive value by assuming that the prevalence of amyloid positivity ranged from 33.6–60%. Based on the original dataset with a 33.6% prevalence of amyloid positivity, we calculated the following values.

LN_33.6% = Number of amyloid-negative patients in the low group (true negative)/Total number of amyloid-negative patients

LP_33.6% = Number of amyloid-positive patients in the low group (false negative)/Total number of amyloid-positive patients

IN_33.6% = Number of amyloid-negative patients in the intermediate group/Total number of amyloid-negative patients

IP_33.6% = Number of amyloid-positive patients in the intermediate group/Total number of amyloid-positive patients

HN_33.6% = Number of amyloid-negative patients in the high group (false positive)/Total number of amyloid-negative patients

HP_33.6% = Number of amyloid-positive patients in the high group (true positive)/Total number of amyloid-positive patients

These calculated values (LN_33.6%, LP_33.6%, IN_33.6%, IP_33.6%, HN_33.6% and HP_33.6%) are theoretically constant but rates of amyloid-negative patients and amyloid-positive patients in the population change dependent on differing prevalence rates in the population. Therefore, we simulated the following values for the 60% prevalence of amyloid positivity using the calculated values.

The negative predictive value (proportion of amyloid-negative patients in the low group to patients in the low group) was (1 - prevalence) * LN_33.6%/((1 - prevalence)*LN_33.6% + prevalence *LP_33.6%).

The positive predictive value (number of amyloid-positive patients in the high group/total number of patients in the high group) = prevalence * HP_33.6%/((1 - prevalence)*HN_33.6% + prevalence *HP_33.6%).

Number of amyloid-negative patients in the intermediate group/Total number of patients in the intermediate group = (1 - prevalence) * LN_33.6%/((1 - prevalence)*IN_33.6% + prevalence *IP_33.6%)

Effects of simulated bias on sensitivity and specificity

We added different bias percentages to the measured values of plasma APP669-711/amyloid-β1–42 and amyloid-β1–40/1–42. Using APP669-711/amyloid-β1–42 and amyloid-β1–40/1–42 with bias, we evaluated the sensitivity and specificity at dual cutoff values of ≥ 90.0% for sensitivity and specificity, respectively.

Multiple linear regression analysis

Multiple linear regression was used to determine the associations of plasma amyloid-β biomarkers with BMI; ApoE4 status; cortical ¹¹C-Pittsburgh compound-B uptake; medical history; aspartate aminotransferase, alanine aminotransferase, and γ-glutamyl transpeptidase levels; and estimated glomerular filtration rate, controlling for age and sex. The plasma amyloid-β biomarkers were z-scored relative to the entire sample to compare coefficients. A P-value of < 0.05 was considered to indicate statistical significance.

Kaplan‒Meier curves

Kaplan‒Meier curves were generated to analyze the time to AD dementia progression in the three groups categorized by the plasma composite biomarker level, and the overall difference between the groups was calculated using the log-rank test. Cox proportional hazards regression analysis was performed to investigate the hazard ratio for the conversion from MCI to AD dementia with adjustment for the Japanese version of the Montreal Cognitive Assessment score. All statistical analyses were conducted using SPSS 25.0 (IBM, Armonk, NY) and R 4.2.3 (R Foundation, Vienna, Austria).

Ethics

This prospective study received ethical approval from the Institutional Ethics Committee of Oita University Hospital (Approval No. UMIN000017442). Written informed consent was obtained from all participants.

Clinical and demographic characteristics

The median age was 76.00 years (interquartile range 72.00–80.00 years), 46.7% were males, and 15.9% were ApoE4 carriers (Table 1). The amyloid-β-positive subgroup was older (P = 0.001), more likely to be female (P = 0.048), had a greater incidence of heart diseases (P = 0.028), had greater ¹¹C-Pittsburgh compound-B uptake (P < 0.001), and had lower scores on the Japanese version of the Montreal Cognitive Assessment (P = 0.006) and Wechsler Memory Scale-Revised II test (P < 0.001) than the amyloid-β-negative subgroup. Moreover, the amyloid-β-positive subgroup had significantly greater plasma levels of the composite biomarker APP669-711/amyloid-β1–42 and amyloid-β1–40/1–42 than the amyloid-β-negative subgroup (P < 0.001 for each, Cliff’s d effect size = -0.886, 95% confidence interval [CI], -0.944 to -0.771 for the composite biomarker, Cliff’s d = -0.824, 95% CI, -0.950 to -0.468 for APP669-711/amyloid-β1–42, Cliff’s d = -0.748, 95% CI, -0.855 to -0.581 for amyloid-β1–40/1–42).

Table 1

Clinical and demographic characteristics of all participants
Characteristic	All (n = 107)	Amyloid-negative (n = 71)	Amyloid-positive (n = 36)
Characteristic	Median (IQR)/N (%)	Median (IQR)/N (%)	Median (IQR)/N (%)	P-value
Age, years	76.00 (72.00, 80.00)	74.00 (69.00, 79.00)	78.50 (75.75, 81.00)	0.001
Male	50 (46.7%)	38 (53.5%)	12 (33.3%)	0.048
Education level, years	12.00 (9.00, 12.00)	12.00 (9.00, 12.00)	12.00 (9.75, 12.00)	0.333
BMI, kg/m²	23.27 (21.39, 24.89)	23.86 (21.43, 24.96)	22.82 (20.56, 24.79)	0.228
MMSE score	26.00 (25.00, 27.00)	26.00 (25.00, 27.00)	26.00 (24.00, 27.00)	0.498
MoCA-J score	22.00 (19.00, 25.00)	23.00 (19.00, 25.50)	20.00 (18.00, 22.25)	0.006
WMS-R Ⅱ score	6.00 (2.00, 11.00)	8.00 (3.50, 13.00)	4.00 (0.00, 6.25)	< 0.001
ApoE status	17 (15.9%)	8 (11.3%)	9 (25.0%)	0.066
PiB uptake	0.94 (0.83, 1.36)	0.85 (0.81, 0.94)	1.69 (1.37, 2.25)	< 0.001
Hypertension	59 (55.1%)	37 (52.1%)	22 (61.1%)	0.497
Diabetes mellitus	22 (20.6%)	15 (21.1%)	7 (19.4%)	> 0.99
Hypercholesterolemia	34 (31.8%)	19 (26.8%)	15 (41.7%)	0.179
Stroke	7 (6.5%)	3 (4.2%)	4 (11.1%)	0.343
Heart diseases	19 (17.8%)	8 (11.3%)	11 (30.6%)	0.028
Thyroid diseases	10 (9.3%)	4 (5.6%)	6 (16.7%)	0.133
Hepatic disorders	2 (1.9%)	1 (1.4%)	1 (2.8%)	> 0.99
Malignant tumors	8 (7.5%)	5 (7.0%)	3 (8.3%)	> 0.99
AST	22.60 (19.40, 26.05)	22.70 (19.50, 27.15)	22.45 (19.25, 25.03)	0.700
ALT	16.90 (14.00, 21.65)	17.50 (14.35, 21.90)	15.20 (13.33, 21.05)	0.136
γ-GTP	20.70 (14.60, 29.80)	22.20 (14.60, 31.90)	17.60 (13.78, 26.55)	0.379
eGFR	65.70 (56.65, 75.35)	66.30 (58.55, 76.15)	60.05 (53.05, 74.12)	0.150
Composite biomarker	0.45 (–0.14, 1.20)	–0.0037 (–0.31, 0.49)	1.53 (1.08, 1.78)	< 0.001
APP669-711/amyloid-β1–42	1.00 (0.88, 1.19)	0.93 (0.84, 1.05)	1.26 (1.11, 1.39)	< 0.001
Amyloid-β1–40/1–42	23.06 (20.63, 25.82)	21.65 (20.00, 23.52)	25.90 (25.19, 27.56)	< 0.001
ALT, alanine aminotransferase; ApoE, apolipoprotein E; APP, amyloid-β precursor protein; AST, aspartate aminotransferase; BMI, body mass index; eGFR, estimated glomerular filtration rate; γ-GTP, γ-glutamyl transpeptidase; IQR, interquartile range; MMSE, Mini-Mental State Examination; MoCA-J, Japanese version of Montreal Cognitive Assessment; PiB, ¹¹C-Pittsburgh compound-B; WMS-R II, Wechsler Memory Scale-Revised logical memory II test. P < 0.05 indicated significance.

Correlations between plasma amyloid-β biomarkers and ¹¹C-Pittsburgh compound-B uptake

Among the three plasma amyloid-β biomarkers, the composite biomarker showed the strongest correlation with the ¹¹C-Pittsburgh compound-B standardized uptake value ratio (ρ = 0.754, 95% CI, 0.649 to 0.829, P < 0.001; Supplemental Fig. 1). Moreover, Statistical Parametric Mapping analysis revealed a significant correlation between plasma amyloid-β biomarkers and ¹¹C-Pittsburgh compound-B uptake, mainly in the frontal and parietotemporal cortices and posterior cingulate gyrus (Supplemental Fig. 2).

[insert Table 1 here]

Predictive accuracy of plasma amyloid-β biomarkers for amyloid PET

Models predicting amyloid positivity on PET based on plasma amyloid-β biomarker levels had areas under the curves of 0.943 (95% CI, 0.901 to 0.985) for the composite biomarker, 0.912 (95% CI, 0.857 to 0.968) for APP669-711/amyloid-β1–42, and 0.874 (95% CI, 0.806 to 0.942) for amyloid-β1–40/1–42 (Fig. 1). Combining the composite biomarker with ApoE, age, and sex slightly increased the area under the curve from 0.943 to 0.970 (95% CI, 0.945 to 0.995) but did not significantly increase the area under the curve (P = 0.082, Fig. 2).

Performance of plasma amyloid-β biomarkers for classifying brain amyloid status

To enhance the accuracy of plasma amyloid-β biomarkers in classifying brain amyloid status, two cutoff values were set: a lower cutoff at a sensitivity ≥ 90.0% and an upper cutoff at a specificity ≥ 90.0%. Low-, intermediate-, and high-probability groups were categorized by the dual cutoffs on three amyloid-β biomarkers (Table 2). According to the composite biomarker, a lower proportion of adults (10.3%) belonged to the intermediate group. The composite biomarker had a 95.1% (95% CI, 86.3–99.0%) negative predictive value and an 80.0% (95% CI, 63.1–91.6%) positive predictive value for amyloid positivity on PET. Increasing the prevalence of amyloid positivity on PET to 60% led to a decreased negative predictive value (86.7%) and an increased positive predictive value (92.2%). We further evaluated the performance of the composite biomarker using other dual cutoff values (Table 3).

Table 2

The performance of amyloid biomarkers using other dual cutoff criteria
Biomarker level	All, No. (%)	Amyloid PET status, 33.6% prevalence, %		Reweighted for 60% prevalence, %
Biomarker level		Amyloid-negative	Amyloid-positive	Amyloid-negative	Amyloid-positive
Sensitivity/Specificity ≥ 90.0%
Composite biomarker
Low (< 0.661)	61 (57.0)	95.1	4.9	86.7	13.3
Intermediate (0.661–1.029)	11 (10.3)	54.5	45.5	28.9	71.1
High (≥ 1.029)	35 (32.7)	20.0	80.0	7.8	92.2
APP669-711/amyloid-β1–42
Low (< 0.995)	52 (48.6)	94.2	5.8	84.7	15.3
Intermediate (0.995–1.149)	24 (22.4)	62.5	37.5	36.0	64.0
High (≥ 1.149)	31 (29.0)	22.6	77.4	9.0	91.0
Amyloid-β1–40/1–42
Low (< 24.067)	59 (55.1)	94.9	5.1	86.3	13.7
Intermediate (24.067–25.792)	20 (18.7)	40.0	60.0	18.4	81.6
High (≥ 25.792)	28 (26.2)	25.0	75.0	10.1	89.9

APP, amyloid-β precursor protein; PET, positron emission tomography

Table 3

Performance of composite biomarker using other dual cutoff criteria
Composite biomarker level	All, No. (%)	Amyloid PET status, 33.6% prevalence, %		Reweighted for 60% prevalence, %
Composite biomarker level		Amyloid negative	Amyloid positive	Amyloid negative	Amyloid positive
NPV/PPV ≥ 95.0%
Low (< 0.661)	61 (57.0)	95.1	4.9	86.7	13.3
Intermediate (0.661–1.406)	24 (22.4)	50.0	50.0	25.3	74.7
High (≥ 1.406)	22 (20.6)	4.5	95.5	1.6	98.4
NPV/PPV ≥ 90.0%
Low (< 0.959)	70 (65.4)	90.0	10.0	75.3	24.7
Intermediate (0.959–1.187)	10 (9.3)	60.0	40.0	33.6	66.4
High (≥ 1.187)	27 (25.2)	7.4	92.6	2.6	97.4

NPV, negative predictive value; PET, positron emission tomography; PPV, positive predictive value

With the cutoff set at a negative predictive value/positive predictive value of ≥ 95.0%, the negative predictive value and positive predictive value were 95.1% (95% CI, 86.3–99.0%) and 95.5% (95% CI, 77.2–99.9%), respectively, whereas more participants were classified as having an intermediate probability (22.4%). A cutoff of a negative predictive value/positive predictive value of ≥ 90.0% decreased the percentage of adults classified into the intermediate group (9.3%). We simulated changes in sensitivity and specificity at lower and upper cutoffs by adding different bias percentages to APP669-711/amyloid-β1–42 and amyloid-β1–40/1–42 for a robustness assessment. The sensitivity and specificity were less affected by bias for APP669-711/amyloid-β1–42 than for amyloid-β1–40/1–42 (Supplemental Fig. 3).

Associations between plasma amyloid-β biomarkers and multiple comorbidities

Linear regression analysis adjusted for age and sex revealed that ApoE4 status was associated with increased plasma levels of the composite biomarker APP669-711/amyloid-β1–42 and amyloid-β1–40/1–42. Moreover, hepatic disorders were associated with increased composite biomarker and APP669-711/amyloid-β1–42 levels, although there were only two patients with hepatic disorders in this study (Fig. 3). One patient had increased levels of APP669-711/amyloid-β1–42, which resulted in increased levels of composite biomarkers (Supplemental Fig. 4). Patients with higher levels of APP669-711/amyloid-β1–42 had relatively greater levels of APP669-711 than those with amyloid-β1–42 and amyloid-β1–40, who had hepatic encephalopathy. No comorbidity effects beyond the PET uptake level were observed.

Baseline plasma amyloid-β biomarkers and conversion to AD dementia

Among 107 adults with MCI, 28 progressed to AD dementia within 7 years and 61 remained stable with MCI at the 7-year follow-up. Others converted to non-AD dementia or dropped out. The composite biomarker (Fig. 4A), APP669-711/amyloid-β1–42 (Fig. 4B), and amyloid-β1–40/1–42 (Fig. 4C) were significantly more abundant in the MCI conversion to AD (MCI conversion) group than in the stable MCI group (P < 0.001, Cliff’s d=-0.720, 95% CI, -0.866 to -0.462). Receiver operating characteristic analysis for the composite biomarker demonstrated a high area under the curve (0.860, 95% CI 0.778 to 0.943; Fig. 4D), sensitivity of 0.857 (95% confidence interval 0.714 to 0.964), and specificity of 0.754 (95% CI 0.639 to 0.853) at a cutoff of 0.661 for discriminating between MCI conversion and stable MCI.

The relationship between the composite biomarker and the time to AD dementia conversion was further analyzed using Kaplan‒Meier curves (Fig. 5). The risk of AD conversion significantly differed among the low-, intermediate-, and high-risk groups when the dual cutoff was set at 90.0% or greater sensitivity/specificity (P < 0.001, log-rank test). The scores on the Japanese version of the Montreal Cognitive Assessment were significantly different among the three groups (P = 0.026; Supplemental Table 1). Cox regression analysis adjusted for the Japanese version of the Montreal Cognitive Assessment scores revealed that the risk of AD dementia in the high- and intermediate-risk groups was greater than that in the low-risk group (hazard ratio 6.51, 95% CI 1.73 to 24.51, P = 0.006 for intermediate risk; hazard ratio 10.33, 95% CI 3.46 to 30.80, P < 0.001 for high risk).

This study provides several novel and interesting insights into the usefulness of plasma biomarkers for predicting brain amyloid burden at the MCI stage. First, the area under the curve values for predicting amyloid positivity on PET were 0.943 for the composite biomarker, 0.912 for APP669-711/amyloid-β1–42, and 0.874 for amyloid-β1–40/1–42 in a community-based cohort. Second, ApoE4 status and hepatic disorders influenced the plasma levels of amyloid-β. Third, higher plasma composite biomarker levels were associated with a high risk of AD. These results suggest the potential for a composite biomarker to replace PET and cerebrospinal fluid tests.

Several studies have examined the predictive accuracy of plasma amyloid-β biomarkers measured using immunoprecipitation-mass spectrometry to detect brain amyloid deposition based on PET or cerebrospinal fluid analysis. Plasma amyloid-β42/40 could predict amyloid positivity on PET, with areas under the curves ranging from 0.793 to 0.935 in cognitively healthy adults and adults with MCI or AD [9, 10, 20–24] and with areas under the curves ranging from 0.752 to 0.88 in cognitively healthy adults [25–28]. A recent review of plasma amyloid biomarkers reported that amyloid-β42/40 had a weighted average area under the curve of 0.834 using amyloid PET as a reference standard [29]. In particular, in head-to-head comparative studies, Shimadzu and Washington University showed that plasma amyloid-β42/40 could accurately predict amyloid positivity on PET [30]. The areas under the curves of plasma amyloid-β42/40 and composite biomarkers were 0.935 and 0.954, respectively, for identifying amyloid positivity on PET in healthy adults and adults with MCI or AD from the Australian Imaging Biomarker and Lifestyle Study of Aging cohorts and the Japanese National Center for Geriatrics and Gerontology [9]. Our results showed that the areas under the curves for amyloid-β1–40/1–42 and the composite biomarker were 0.874 and 0.943, respectively, for identifying amyloid positivity on PET in a community-based cohort. The areas under the curves of cerebrospinal fluid amyloid-β1–42/1–40 and cerebrospinal fluid amyloid-β1–42/P-tau181 were 0.94 and 0.95, respectively, in 288 individuals selected from the Amsterdam Dementia Cohort [31], with a sensitivity/specificity of 94%/84% and 91%/86%, respectively, in 77 individuals selected from BioFINDER [32]. The areas under the curves in our study are similar to those of cerebrospinal fluid amyloid-β1–42/1–40 [31] and cerebrospinal fluid amyloid-β1–42/P-tau181 [32], which the Pharmaceuticals and Medical Devices Agency in Japan and the Food and Drug Administration have already approved. Previous studies have proposed a two-step workflow in which plasma biomarkers are screened for amyloid-β, with additional confirmatory testing for uncertain cases [9]. In this study, when thresholds were set to satisfy negative predictive values/positive predictive values of 90% and 95%, similar to those of PET [33–35], the intermediate ranges were 9.3% and 22.4%, respectively. Therefore, we suggest that the Shimadzu composite biomarker can predict amyloid positivity on PET with high performance in community-dwelling adults before dementia onset. Assuming that the cost of immunoprecipitation-mass spectrometry is the same as that previously reported [36] and that the cost ratio of PET to plasma biomarker analysis is 8× or 4×, plasma biomarkers may have a cost benefit for the Japan Universal Health Insurance System. Moreover, the use of a composite biomarker combined with age and ApoE4 status slightly improved the accuracy of detecting brain amyloid deposition, consistent with previous findings [23, 28, 30].

Longitudinal analysis showed that adults with higher baseline levels of composite biomarkers had a greater rate of conversion to AD dementia than those with lower levels. The usefulness of plasma amyloid biomarkers for predicting the development of dementia has been examined using ELISA or single-molecule arrays over a follow-up period of < 5 years [37–41]. Only a few studies have reported the association of amyloid-β-42/40 or composite biomarkers with cognitive function or the longitudinal association of plasma amyloid-β-42 or amyloid-β-38 measured with ELISA or single-molecule arrays with the risk of developing AD over 14 years of follow-up [42–45]. Our findings revealed an association between higher plasma composite biomarker levels and a greater risk of developing AD dementia over 8 years of follow-up in a community-based MCI cohort.

Several factors may affect plasma amyloid-β biomarkers, influencing the interpretation of results or the development of reference ranges because patient populations are heterogeneous [46]. This study showed that ApoE4 status and hepatic disorders were associated with increased plasma amyloid-β biomarker levels. Previous studies have reported that age, ApoE4 status, ischemic heart disease, hypertension, diabetes, hepatic disorders, and chronic kidney disease affect plasma amyloid-β-42/40 levels [16, 47–49]. Conversely, sex and BMI are not associated with plasma amyloid-β-42/40 [7, 15, 16]. Our results are consistent with those of previous studies showing that hepatic disorders are associated with elevated levels of plasma amyloid biomarkers through reduced clearance of amyloid-β [47, 48].

Limitations

This study has some limitations. The number of participants recruited in this study was small. Moreover, the prevalence of adults with amyloid positivity on PET was relatively small. Further studies, including more large sample size and sampling techniques, are needed to establish the usefulness of plasma amyloid biomarkers. The standardization of preanalytical, analytical, and standard references was required to use plasma amyloid biomarkers.

A composite biomarker may be suitable for routine clinical practice as a screening tool for adults at a risk of AD dementia. Amyloid-β biomarkers may accurately indicate brain amyloid deposition although blood collection in real-world situations is not standardized. Our results suggest that a composite biomarker may be a suitable surrogate for PET positivity or amyloid-β-42 levels in the cerebrospinal fluid.

AD, Alzheimer’s disease; ApoE, apolipoprotein E4; APP, amyloid-β precursor protein; BMI, body mass index; CI, confidence interval; ELISA, enzyme-linked immunosorbent assay; MCI, mild cognitive impairment; PET, positron emission tomography

Ethics approval and consent to participate: This prospective study received ethical approval from the Institutional Ethics Committee of Oita University Hospital (Approval No. UMIN000017442). Written informed consent was obtained from all participants.

Consent for publication: Not applicable

Availability of data and materials: The data cannot be shared publicly owing to ethical restrictions. The consent form signed by the participants states that their data are provided in an anonymized form only to research institutions. The raw data utilized in this study contain sensitive and personally identifiable information that could negatively affect the privacy of the study participants. However, the data supporting the results of this study are available upon ethical approval by the local ethics committee of Oita University Hospital. The ethics committee of Oita University Hospital can be contacted to obtain these data ([email protected]).

Competing interests: The authors declare that they have no competing interests. N. Kaneko, TT, SH, and SI are employed at Shimadzu Corporation.

Funding: This research was supported by the Japan Agency for Medical Research and Development (grant number 18he1402003). The funders had no role in the design or execution of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Author contributions: N. Kimura, N. Kaneko, and EM conceived of and designed the study. TA, N. Kimura, N. Kaneko, T. Masuda, YT, KY, AN, T. Mizukami, HI, and TI acquired, analyzed, and interpreted the data. TA, N. Kimura, N. Kaneko, and TI drafted the manuscript. All authors critically revised the manuscript for important intellectual content. N. Kaneko, TI, and SH conducted the statistical analysis. EM obtained funding for the study. N. Kimura, N. Kaneko, and TT provided administrative, technical, or material support. EM supervised the study.

Acknowledgements: We gratefully acknowledge the assistance of Usuki city employees for their efforts in recruiting adults. We thank Suzuki Co., Ltd., for their assistance with data collection and HCL Technologies Confidential and Fusa Matsuzaki for database construction and data analysis. The authors are sincerely grateful to all the participants who enrolled in this study. The authors would like to thank Editage (www.editage.com) for English language editing.

Ninomiya T, Nakaji S, Maeda T, Yamada M, Mimura M, Nakashima K, et al. Study design and baseline characteristics of a population-based prospective cohort study of dementia in Japan: the Japan Prospective Studies Collaboration for Aging and Dementia (JPSC-AD). Environ Health Prev Med. 2020;25:64.
Budd Haeberlein S, Aisen PS, Barkhof F, Chalkias S, Chen T, Cohen S, et al. Two randomized Phase 3 studies of aducanumab in early Alzheimer’s disease. J Prev Alzheimers Dis. 2022;9:197-210.
van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, et al. Lecanemab in early Alzheimer’s disease. N Engl J Med. 2023;388:9-21.
Yabuuchi K, Kimura N, Masuda T, Matsubara E. Comparison of brain amyloid deposition and cortical glucose metabolism between clinic- and community-based cohort. J Alzheimers Dis. 2023;95:299-306.
Palmqvist S, Janelidze S, Stomrud E, Zetterberg H, Karl J, Zink K, et al. Performance of fully automated plasma assays as screening tests for Alzheimer disease-related β-amyloid status. JAMA Neurol. 2019;76:1060-9.
Hansson O, Edelmayer RM, Boxer AL, Carrillo MC, Mielke MM, Rabinovici GD, et al. The Alzheimer’s Association appropriate use recommendations for blood biomarkers in Alzheimer’s disease. Alzheimers Dement. 2022;18:2669-86.
Teunissen CE, Verberk IMW, Thijssen EH, Vermunt L, Hansson O, Zetterberg H, et al. Blood-based biomarkers for Alzheimer’s disease: towards clinical implementation. Lancet Neurol. 2022;21:66-77.
Brum WS, Cullen NC, Janelidze S, Ashton NJ, Zimmer ER, Therriault J, et al. A two-step workflow based on plasma p-tau217 to screen for amyloid β positivity with further confirmatory testing only in uncertain cases. Nat Aging. 2023;3:1079-90.
Nakamura A, Kaneko N, Villemagne VL, Kato T, Doecke J, Doré V, et al. High performance plasma amyloid-β biomarkers for Alzheimer’s disease. Nature. 2018;554:249-54.
Li Y, Schindler SE, Bollinger JG, Ovod V, Mawuenyega KG, Weiner MW, et al. Validation of plasma amyloid-β 42/40 for detecting Alzheimer disease amyloid plaques. Neurology. 2022;98:e688-99.
Monane M, Johnson KG, Snider BJ, Turner RS, Drake JD, Maraganore DM, et al. A blood biomarker test for brain amyloid impacts the clinical evaluation of cognitive impairment. Ann Clin Transl Neurol. 2023;10:1738-48.
Palmqvist S, Stomrud E, Cullen N, Janelidze S, Manuilova E, Jethwa A, et al. An accurate fully automated panel of plasma biomarkers for Alzheimer’s disease. Alzheimers Dement. 2023;19:1204-15.
Janelidze S, Bali D, Ashton NJ, Barthélemy NR, Vanbrabant J, Stoops E, et al. Head-to-head comparison of 10 plasma phospho-tau assays in prodromal Alzheimer’s disease. Brain. 2023;146:1592-601.
Karikari TK, Pascoal TA, Ashton NJ, Janelidze S, Benedet AL, Rodriguez JL, et al. Blood phosphorylated tau 181 as a biomarker for Alzheimer’s disease: a diagnostic performance and prediction modelling study using data from four prospective cohorts. Lancet Neurol. 2020;19:422-33.
Brickman AM, Manly JJ, Honig LS, Sanchez D, Reyes-Dumeyer D, Lantigua RA, et al. Plasma p-tau181, p-tau217, and other blood-based Alzheimer’s disease biomarkers in a multi-ethnic, community study. Alzheimers Dement. 2021;17:1353-64.
Syrjanen JA, Campbell MR, Algeciras-Schimnich A, Vemuri P, Graff-Radford J, Machulda MM, et al. Associations of amyloid and neurodegeneration plasma biomarkers with comorbidities. Alzheimers Dement. 2022;18:1128-40.
Mielke MM, Dage JL, Frank RD, Algeciras-Schimnich A, Knopman DS, Lowe VJ, et al. Performance of plasma phosphorylated tau 181 and 217 in the community. Nat Med. 2022;28:1398-405.
Kimura N, Aso Y, Yabuuchi K, Ishibashi M, Hori D, Sasaki Y, et al. Association of modifiable lifestyle factors with cortical amyloid burden and cerebral glucose metabolism in older adults with mild cognitive impairment. JAMA Netw Open. 2020;3:e205719.
Okada Y, Kato T, Iwata K, Kimura Y, Nakamura A, Hattori H, et al. Evaluation of PiB visual interpretation with CSF Aβ and longitudinal SUVR in J-ADNI study. Ann Nucl Med. 2020;34:108-18.
Rabe C, Bittner T, Jethwa A, Suridjan I, Manuilova E, Friesenhahn M, et al. Clinical performance and robustness evaluation of plasma amyloid-β42/40 prescreening. Alzheimers Dement. 2023;19:1393-402.
Hu Y, Kirmess KM, Meyer MR, Rabinovici GD, Gatsonis C, Siegel BA, et al. Assessment of a plasma amyloid probability score to estimate amyloid positron emission tomography findings among adults with cognitive impairment. JAMA Netw Open. 2022;5:e228392.
West T, Kirmess KM, Meyer MR, Holubasch MS, Knapik SS, Hu Y, et al. A blood-based diagnostic test incorporating plasma Aβ42/40 ratio, ApoE proteotype, and age accurately identifies brain amyloid status: findings from a multi cohort validity analysis. Mol Neurodegener. 2021;16:30.
Jang H, Kim JS, Lee HJ, Kim C-H, Na DL, Kim HJ, et al. Performance of the plasma Aβ42/Aβ40 ratio, measured with a novel HPLC-MS/MS method, as a biomarker of amyloid PET status in a DPUK-Korean cohort. Alzheimers Res Ther. 2021;13:179.
Ovod V, Ramsey KN, Mawuenyega KG, Bollinger JG, Hicks T, Schneider T, et al. Amyloid β concentrations and stable isotope labeling kinetics of human plasma specific to central nervous system amyloidosis. Alzheimers Dement. 2017;13:841-9.
Janelidze S, Palmqvist S, Leuzy A, Stomrud E, Verberk IMW, Zetterberg H, et al. Detecting amyloid positivity in early Alzheimer’s disease using combinations of plasma Aβ42/Aβ40 and p-tau. Alzheimers Dement. 2022;18:283-93.
Benedet AL, Brum WS, Hansson O, Alzheimer’s Disease Neuroimaging Initiative, Karikari TK, Zimmer ER, et al. The accuracy and robustness of plasma biomarker models for amyloid PET positivity. Alzheimers Res Ther. 2022;14:26.
Tosun D, Veitch D, Aisen P, Jack Jr CR, Jagust WJ, Petersen RC, et al. Detection of β-amyloid positivity in Alzheimer’s Disease Neuroimaging Initiative participants with demographics, cognition, MRI and plasma biomarkers. Brain Commun. 2021;3:fcab008.
Schindler SE, Bollinger JG, Ovod V, Mawuenyega KG, Li Y, Gordon BA, et al. High-precision plasma β-amyloid 42/40 predicts current and future brain amyloidosis. Neurology. 2019;93:e1647-59.
Brand AL, Lawler PE, Bollinger JG, Li Y, Schindler SE, Li M, et al. The performance of plasma amyloid beta measurements in identifying amyloid plaques in Alzheimer’s disease: a literature review. Alzheimers Res Ther. 2022;14:195.
Janelidze S, Teunissen CE, Zetterberg H, Allué JA, Sarasa L, Eichenlaub U, et al. Head-to-head comparison of 8 plasma amyloid-β 42/40 assays in Alzheimer disease. JAMA Neurol. 2021;78:1375-82.
Willemse EAJ, Tijms BM, van Berckel BNM, Le Bastard N, van der Flier WM, Scheltens P, et al. Comparing CSF amyloid-beta biomarker ratios for two automated immunoassays, Elecsys and Lumipulse, with amyloid PET status. Alzheimers Dement (Amst). 2021;13:e12182.
Leuzy A, Mattsson-Carlgren N, Cullen NC, Stomrud E, Palmqvist S, La Joie R, et al. Robustness of CSF Aβ42/40 and Aβ42/P-tau181 measured using fully automated immunoassays to detect AD-related outcomes. Alzheimers Dement. 2023;19:2994-3004.
Clark CM, Pontecorvo MJ, Beach TG, Bedell BJ, Coleman RE, Doraiswamy PM, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-β plaques: a prospective cohort study. Lancet Neurol. 2012;11:669-78.
Sabri O, Sabbagh MN, Seibyl J, Barthel H, Akatsu H, Ouchi Y, et al. Florbetaben PET imaging to detect amyloid beta plaques in Alzheimer’s disease: phase 3 study. Alzheimers Dement. 2015;11:964-74.
Salloway S, Gamez JE, Singh U, Sadowsky CH, Villena T, Sabbagh MN, et al. Performance of [18F]flutemetamol amyloid imaging against the neuritic plaque component of CERAD and the current (2012) NIA-AA recommendations for the neuropathologic diagnosis of Alzheimer’s disease. Alzheimers Dement (Amst). 2017;9:25-34.
Cullen NC, Janelidze S, Stomrud E, Bateman RJ, Palmqvist S, Hansson O, et al. Plasma amyloid-β42/40 and apolipoprotein E for amyloid PET pre-screening in secondary prevention trials of Alzheimer’s disease. Brain Commun. 2023;5:fcad015.
Verberk IMW, Slot RE, Verfaillie SCJ, Heijst H, Prins ND, van Berckel BNM, et al. Plasma amyloid as prescreener for the earliest Alzheimer pathological changes. Ann Neurol. 2018;84:648-58.
Pérez-Grijalba V, Romero J, Pesini P, Sarasa L, Monleón I, San-José I, et al. Plasma Aβ42/40 ratio detects early stages of Alzheimer’s disease and correlates with CSF and neuroimaging biomarkers in the AB255 study. J Prev Alzheimers Dis. 2019;6:34-41.
Stockmann J, Verberk IMW, Timmesfeld N, Denz R, Budde B, Lange-Leifhelm J, et al. Amyloid-β misfolding as a plasma biomarker indicates risk for future clinical Alzheimer’s disease in individuals with subjective cognitive decline. Alzheimers Res Ther. 2020;12:169.
Verberk IMW, Hendriksen HMA, van Harten AC, Wesselman LMP, Verfaillie SCJ, van den Bosch KA, et al. Plasma amyloid is associated with the rate of cognitive decline in cognitively normal elderly: the SCIENCe project. Neurobiol Aging. 2020;89:99-107.
Cullen NC, Leuzy A, Janelidze S, Palmqvist S, Svenningsson AL, Stomrud E, et al. Plasma biomarkers of Alzheimer’s disease improve prediction of cognitive decline in cognitively unimpaired elderly populations. Nat Commun. 2021;12:3555.
Giudici KV, de Souto Barreto P, Guyonnet S, Li Y, Bateman RJ, Vellas B, et al. Assessment of plasma amyloid-β42/40 and cognitive decline among community-dwelling older adults. JAMA Netw Open. 2020;3:e2028634.
Lim YY, Maruff P, Kaneko N, Doecke J, Fowler C, Villemagne VL, et al. Plasma amyloid-β biomarker associated with cognitive decline in preclinical Alzheimer’s disease. J Alzheimers Dis. 2020;77:1057-65.
Hilal S, Wolters FJ, Verbeek MM, Vanderstichele H, Ikram MK, Stoops E, et al. Plasma amyloid-β levels, cerebral atrophy and risk of dementia: a population-based study. Alzheimers Res Ther. 2018;10:63.
de Wolf F, Ghanbari M, Licher S, McRae-McKee K, Gras L, Weverling GJ, et al. Plasma tau, neurofilament light chain and amyloid-β levels and risk of dementia; a population-based cohort study. Brain. 2020;143:1220-32.
Pais MV, Forlenza OV, Diniz BS. Plasma biomarkers of Alzheimer’s disease: a review of available assays, recent developments, and implications for clinical practice. J Alzheimers Dis Rep. 2023;7:355-80.
Wang Y-R, Wang Q-H, Zhang T, Liu Y-H, Yao X-Q, Zeng F, et al. Associations between hepatic functions and plasma amyloid-beta levels-implications for the capacity of liver in peripheral amyloid-beta clearance. Mol Neurobiol. 2017;54:2338-44.
Lyu H, Ye Y, Lui VCH, Wu W, Chung PHY, Wongn KKY, et al. Plasma amyloid-beta levels correlated with impaired hepatic functions: an adjuvant biomarker for the diagnosis of biliary atresia. Front Surg. 2022;9:931637.
Janelidze S, Stomrud E, Palmqvist S, Zetterberg H, van Westen D, Jeromin A, et al. Plasma β-amyloid in Alzheimer’s disease and vascular disease. Sci Rep. 2016;6:26801.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Plasma amyloid-β biomarkers for predicting amyloid positivity in mild cognitive impairment: a prospective cohort study

Status:

Version 1

Abstract

Figures

BACKGROUND

MATERIALS AND METHODS

Participants

Immunoprecipitation-mass spectrometry

PET

Apolipoprotein E isoform

Statistical analysis

Logistic regression with receiver operating characteristic curve analysis

Reweighting for 60% prevalence of amyloid positivity

Effects of simulated bias on sensitivity and specificity

Multiple linear regression analysis

Kaplan‒Meier curves

Ethics

RESULTS

Clinical and demographic characteristics

Correlations between plasma amyloid-β biomarkers and ¹¹C-Pittsburgh compound-B uptake

Predictive accuracy of plasma amyloid-β biomarkers for amyloid PET

Performance of plasma amyloid-β biomarkers for classifying brain amyloid status

Associations between plasma amyloid-β biomarkers and multiple comorbidities

Baseline plasma amyloid-β biomarkers and conversion to AD dementia

DISCUSSION

Limitations

CONCLUSIONS

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Plasma amyloid-β biomarkers for predicting amyloid positivity in mild cognitive impairment: a prospective cohort study

Status:

Version 1

Abstract

Figures

BACKGROUND

MATERIALS AND METHODS

Participants

Immunoprecipitation-mass spectrometry

PET

Apolipoprotein E isoform

Statistical analysis

Logistic regression with receiver operating characteristic curve analysis

Reweighting for 60% prevalence of amyloid positivity

Effects of simulated bias on sensitivity and specificity

Multiple linear regression analysis

Kaplan‒Meier curves

Ethics

RESULTS

Clinical and demographic characteristics

Correlations between plasma amyloid-β biomarkers and 11C-Pittsburgh compound-B uptake

Predictive accuracy of plasma amyloid-β biomarkers for amyloid PET

Performance of plasma amyloid-β biomarkers for classifying brain amyloid status

Associations between plasma amyloid-β biomarkers and multiple comorbidities

Baseline plasma amyloid-β biomarkers and conversion to AD dementia

DISCUSSION

Limitations

CONCLUSIONS

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Correlations between plasma amyloid-β biomarkers and ¹¹C-Pittsburgh compound-B uptake