Abstract
Background:
Increasing evidence illustrates the value of plasma biomarkers of Alzheimer’s disease (AD) to screen for and identify dementia with Lewy bodies (DLB). However, confirmatory studies are needed to demonstrate the feasibility of these markers.
Objective:
To determine the feasibility of plasma tau phosphorylated at threonine 181 (p-tau181) and amyloid-β42 (Aβ42) as potential biomarkers to differentiate AD and DLB.
Methods:
We evaluated plasma samples from patients with DLB (n = 47) and AD (n = 55) and healthy controls (HCs, n = 30), using ELISAs to measure p-tau181 and Aβ42. Additionally, we examined neuropsychological assessment scores for participants. The plasma biomarkers were investigated for correlation with neuropsychological assessments and discriminant ability to identify DLB.
Results:
Plasma p-tau181 was significantly lower in DLB than in AD and HCs. Plasma Aβ42 was significantly higher in DLB than in AD but lower in DLB than in HCs. We found good correlations between plasma Aβ42 and neuropsychological scores in the whole cohort, while p-tau181 was associated with cognitive status in DLB. In the distinction between DLB and HCs, plasma p-tau181 and Aβ42 showed similar accuracy, while Aβ42 showed better accuracy than p-tau181 in discriminating DLB and AD.
Conclusion:
In a single-center clinical cohort, we confirmed the high diagnostic value of plasma p-tau181 and Aβ42 for distinguishing patients with DLB from HCs. Plasma Aβ42 improved the differential diagnosis of DLB from AD.
INTRODUCTION
Dementia with Lewy bodies (DLB) is a neurodegenerative disorder that affects cognition, movement, and behavior and is characterized by the presence of abnormal protein deposits called Lewy bodies [1]. DLB interferes with normal life and daily activities and is the second most common type of degenerative dementia in individuals over 65 years old, outranked only by Alzheimer’s disease (AD) [2]. A systematic review showed that DLB accounted for 4.2% of all diagnosed dementias in a community setting [3], and the true prevalence is likely to be much higher.
Currently, the diagnosis and management of DLB follows the guidelines of the fourth consensus report of the DLB Consortium [4]. DLB can be diagnosed according to clinical features and the presence of indicative biomarkers combined. However, the judgment of clinical features may lack reliability due to the patient’s own cognitive impairment or their family members’ insufficient understanding of the disease, and biomarkers detected with techniques such as PET, SPECT, and the cerebrospinal fluid α-synuclein real-time quaking-induced conversion assay (RT-QuIC) [4] are too costly to patients. With recent advances in detection techniques, plasma biomarkers are expected to support the use of DLB diagnostic biomarkers and hopefully to significantly improve clinical screening [5].
DLB shares several clinical and pathological features with AD. An autopsy report showed that nearly 50% of patients with DLB also presented neuropathological features of AD, namely, amyloid-β and tau neurofibrillary tangles [6], confirming the overlap in the pathology of these two diseases. Currently, research on DLB plasma biomarkers is not as in-depth as that of AD; hence, whether AD biomarkers are feasible in DLB diagnosis has aroused people’s interest. Plasma tau phosphorylated at threonine 181 (p-tau181) and amyloid-β42 (Aβ42) are promising noninvasive diagnostic and prognostic biomarkers for AD. Plasma p-tau181 was associated with positive tau protein in the brain and had high predictive accuracy for AD [7, 8]. Plasma Aβ has the potential to indicate the presence of cerebral Aβ through AD progression [9], which highlights the future application of these blood tests to present Aβ positivity.
In this regard, plasma p-tau181 and Aβ42 may be available candidates to screen for DLB, but studies on these markers in DLB are limited. The present study aimed to examine whether plasma p-tau181 and Aβ42 are promising markers for screening for DLB through comparison with the levels in AD patients. Furthermore, we analyzed the relationship between plasma p-tau181 and Aβ42 levels and the clinical symptoms of DLB patients. We sought to find potential plasma biomarkers with certain diagnostic value and differential diagnosis value for DLB screening.
METHODS
Participants
Participants included DLB patients, AD patients, and healthy controls (HCs). Patients were consecutively recruited from the Memory and Language Clinic of the Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, from August 2019 to May 2022. A total of 47 DLB patients, 55 AD patients, and 30 HCs were enrolled. The diagnosis was made by the consensus of a multidisciplinary team. Patients with a clinical diagnosis of probable DLB according to the consensus criteria of McKeith were enrolled [4]. AD patients were diagnosed according to the guidelines of the National Institute on Aging– Alzheimer’s Association (NIA-AA) working group, and the diagnoses were validated using amyloid PET and/or cerebrospinal fluid analysis [10]. HC participants were volunteers from the community and the spouses or caregivers of patients. All HCs were community-dwelling, cognitively and neurologically healthy individuals. Each participant underwent neuropsychological evaluations, laboratory tests, and head magnetic resonance imaging.
This study excluded the following participants: 1) patients with a current or past history of any intracranial pathology (stroke, space-occupying lesions, previous intracranial surgery); 2) patients with other physical or psychiatric conditions that may contribute to cognitive impairment; 3) patients with severe primary diseases of the circulatory, respiratory, endocrine, digestive, or hematopoietic systems; 4) patients with a family history of AD in their immediate relatives; and 5) patients using the following drugs in the following timeframes: antibiotics within three months or steroid hormones for a long period. We did not include patients who had developed parkinsonism at least one year before the onset of dementia, as this presentation is typically classified as Parkinson’s disease dementia. This study was approved by the Ethics Committee of Xuanwu Hospital, Capital Medical University. All enrolled patients signed informed consent forms.
Clinical examination
The diagnostic team consisted of three doctors from Xuanwu Hospital of Capital Medical University. Standardized assessments, including medical history evaluation, informant-based history-taking, physical and neurological exams, and neuropsychological testing, were conducted by trained examiners who were not informed of the purpose of the study or the patient’s diagnosis. Cognitive function was assessed using a standardized neuropsychological kit that included the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA). Complications, including hypertension, hyperlipidemia, diabetes mellitus, and coronary heart disease, were clearly diagnosed through a thorough review of the past medical history. The diagnosis of rapid-eye-movement sleep behavior disorder (RBD) was primarily based on the description of clinical symptoms combined with polysomnography. Polysomnography was performed in 22 out of 47 patients with DLB to assess for the presence of RBD.
Plasma analysis
For each participant, blood for plasma analysis was collected in tubes containing EDTA and centrifuged at 3000 rpm for 15 min at 4°C. The aliquots were immediately frozen at – 80°C and stored until assayed. Plasma samples were assayed for p-tau181 and Aβ42 using an enzyme-linked immunosorbent assay kit from ANQUN (Shenzhen, China). The technicians who performed the analysis did not have access to clinical data. Testers return plasma to room temperature for testing before use. The measured concentration and the relative deviation between the concentration and the quality control product were within±15%, and the test results were valid. The logarithm of the calibration product concentration was taken as the horizontal coordinate, and the logarithm of the OD value was taken as the vertical coordinate to carry out linear fitting and draw the standard curve. The corresponding concentration can be found on the standard curve by the OD value of the sample to be tested.
Statistical analysis
The distributions of p-tau181 and Aβ42 were skewed, so a log 10 transformation was used to improve the distributions. Statistical analysis was performed using IBM SPSS Statistics for Windows, version 18.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA). Outliers with plasma biomarker values above or below 1.5 interquartile range (IQR) of the mean were excluded from the study analysis (Aβ42, n = 1 DLB; n = 3 AD).
Numerical variables are expressed as the means±standard deviations (SD), and categorical data are expressed as numbers (proportion %). The significance level for all tests was set to p < 0.05. Group differences in numerical data were analyzed using two-tailed Student’s t test, Welch’s t test or the Mann-Whitney U test. Group differences in categorical data were analyzed using the χ2 test. Linear regression analysis was performed to assess the correlation between plasma biomarkers and neuropsychological test scores and clinical features. Receiver operating characteristic (ROC) curve analysis was used to determine the accuracy of both plasma markers in clinical diagnosis. The area under the ROC curve (AUC) was computed from binary logistic regression, and the predicted values were used for ROC analyses.
RESULTS
Demographics
A total of 132 participants were selected, including 47 DLB patients, 55 AD patients, and 30 HCs. The participant demographic data and neuropsychological performance are summarized in Table 1. As expected, the sex distribution was different between groups, with significantly more male participants in DLB (61.7%) than in AD (40.0%, p = 0.03). The distribution of groups also varied according to age such that DLB patients were significantly older than AD patients (71.7±6.7 versus 64.8±8.9 years, respectively, p < 0.0001) and HCs (64.5±8.3 years, p < 0.0001). The ages of AD and HCs were not significantly different. The years of disease duration were similar between DLB and AD (3.8±2.7 versus 3.2±1.9 years, respectively, p = 0.39). According to the neuropsychological test results, patients with DLB and AD had significantly lower scores on the MMSE and MoCA than HCs, and the scores of patients with DLB were not significantly different from those of patients with AD.
Clinical and demographic characteristics of DLB, AD, and HC participants
AD, Alzheimer’s disease; DLB, dementia with Lewy bodies; HC, healthy control; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment. The sample size, age, sex, plasma biomarker levels, and neuropsychological scores of each group are reported. A log10 transformation was used for p-tau181 and Aβ42. Continuous variables are expressed as the mean (SD). The number of male patients is reported as a percentage. p1 values indicate differences between the DLB and AD groups, and p2 values indicate differences between the DLB and HC groups. *p < 0.05; ***p < 0.001; ****p < 0.0001.
Clinical features and complications
For these 47 DLB patients, the mean disease duration was 3.8 years (range from 1 to 10). All DLB patients developed cognitive impairment, with a mean course of disease of 2.7 years and a mean age of symptom onset of 69.1 years. Forty-five patients presented with memory loss, and 2 presented with location orientation decline. Forty-four patients (93.6%) developed Parkinson’s-like symptoms, with a mean symptom duration of 2.0 years and a mean age of symptom onset of 69.8 years. Thirty patients (63.8%) developed RBD; the mean duration was 4.8 years, and the mean onset age was 67.1 years. Thirty-nine patients (83.0%) developed visual hallucinations, with a mean symptom duration of 1.8 years and a mean onset age of 69.0 years. Moreover, 29 patients (61.7%) developed autonomic symptoms, including constipation, frequent urination, and postural hypotension; 17 patients (36.1%) had personality changes; and 5 patients (10.6%) had anosmia. Regarding complications of DLB, 23 patients (48.9%) had hypertension, 10 patients (21.3%) had hyperlipidemia, 13 patients (27.7%) had diabetes mellitus, and 10 patients (21.3%) had coronary heart disease (Supplementary Table 1).
Plasma p-tau181 and Aβ42 levels in the study cohort
In the whole cohort, age and sex showed no effect on either plasma p-tau181 or Aβ42 values. Years of education contributed to higher Aβ42 (r = 0.20, p = 0.03) but had no effect on p-tau181. In AD and DLB patients, plasma p-tau181 and Aβ42 were not correlated with disease duration (Supplementary Table 2).
The comparisons of plasma p-tau181 and Aβ42 concentrations between different groups are shown in Fig. 1. Plasma p-tau181 was significantly lower in DLB than in AD and HCs (Fig. 1 and Table 1, DLB 1.2±0.2 pg/ml; AD 1.3±0.2 pg/ml; HC 1.4±0.1 pg/ml; DLB versus AD: p = 0.04; DLB versus HC: p = 0.003). Plasma Aβ42 was significantly higher in DLB than in AD but lower in DLB than in HCs (Fig. 1 and Table 1, DLB 2.0±0.2 pg/ml; AD 1.7±0.2 pg/ml; HC 2.2±0.1 pg/ml; DLB versus AD: p < 0.0001; DLB versus HC: p < 0.001). After controlling for age and sex, the concentration of plasma p-tau181 was significantly associated with Aβ42 in the whole cohort (r = 0.51, p < 0.0001).
Associations of plasma p-tau181 and Aβ42 concentrations with neuropsychological scores and clinical features
Because age and sex are potential confounding factors, we performed an analysis of correlation with age and sex as covariables. For neuropsychological scales, decreased Aβ42 was associated with lower MMSE (r = 0.33, p = 0.0002) and MoCA (r = 0.37, p < 0.0001) scores in the whole cohort. We found no significant correlation between plasma p-tau181 and neuropsychological assessment scores among the three groups (Supplementary Table 3). In the DLB cohort, on the contrary, only increased p-tau181 showed a significant association with lower MMSE scores (r = –0.38, p = 0.02) (Table 2).
The clinical symptoms of DLB patients, including cognitive decline, Parkinson’s-like symptoms, RBD, and recurrent visual hallucinations, showed no significant correlation with plasma p-tau181 and Aβ42 levels. According to clinical complications, hypertension, hyperlipidemia, and diabetes showed no significant correlation with plasma biomarkers with age and sex as controlled variables. Coronary heart disease was correlated with p-tau181 (r = 0.36, p = 0.02), but the correlation was not significant after controlling for MMSE scores (Supplementary Table 4).
Box-and-whisker plot showing plasma p-tau181 and plasma Aβ42 levels in DLB, AD, and HC participants. Data are shown as medians with 25th– 75th percentiles (boxes), minimum– maximum values (whiskers), and individual values (dots). p-tau181: tau phosphorylated at threonine 181; Aβ42, amyloid-β peptide 42; DLB, dementia with Lewy bodies; AD, Alzheimer’s disease; HC, healthy controls; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
Association between plasma biomarkers and neuropsychological scores in the DLB cohort after adjusting for age and sex
DLB, dementia with Lewy bodies; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment. *p < 0.05.
Diagnostic accuracy of plasma p-tau181 and Aβ42
In the discrimination between the DLB and HC groups, ROC curve analysis revealed high accuracy for both plasma p-tau181 and Aβ42 with a similar diagnostic performance (Fig. 2). Plasma p-tau181 and Aβ42 had a significant discriminatory ability between patients with DLB and HCs, with AUC values of 0.7762 for p-tau181 and 0.7431 for Aβ42. Furthermore, p-tau181 and Aβ42 were found to have a significant discriminatory ability between the disease groups. Plasma Aβ42 showed overall higher accuracy in distinguishing DLB from AD (AUC 0.8892) than p-tau181 (AUC 0.5969).
DISCUSSION
Here, we examined the concentrations of plasma p-tau181 and Aβ42 in 132 participants. We found that the concentration of plasma p-tau181 was significantly lower in DLB than in AD and HCs. Plasma Aβ42 was significantly higher in DLB than in AD but lower in DLB than in HCs. This suggests that alterations in plasma concentrations of p-tau181 and Aβ42 may occur not only in patients with AD but also in patients with DLB and possibly in patients with all neurodegenerative diseases. The clinical symptoms and complications of DLB patients did not show a significant correlation with plasma biomarker levels. Although decreased Aβ42 indicated worse cognitive status in the whole cohort, this relationship was not significant in DLB patients. Furthermore, we analyzed the diagnostic accuracy of plasma p-tau181 and Aβ42 for DLB. Most specifically, plasma levels of p-tau181 and Aβ42 improved the diagnosis of DLB, and Aβ42 had a better ability to distinguish DLB from AD according to the AUC. The results of this study confirmed and expanded the diagnostic feasibility of plasma p-tau181 and Aβ42, which showed a different ability to differentiate DLB patients from healthy controls and AD patients. Our study measured both p-tau181 and Aβ42 concentrations in DLB patients and compared these concentrations with those of AD patients, thus potentially extending the scope of plasma biomarker use to all neurodegenerative diseases.
Blood-based markers have greater utility for population screening, with the advantages of convenience, noninvasiveness, and cost-effectiveness. While the present research has been predominantly focused on AD, proteins that were identified as potential biomarkers for AD are likely to be relevant to other neurodegenerative disorders, such as DLB. Evidence indicates that plasma neurofilament light chain, an unspecific marker of axonal injury as well as a promising biomarker for AD, could be a marker to predict DLB progression [11].
Diagnostic accuracy of plasma p-tau181 and Aβ42. A) DLB versus HC; B) DLB versus AD.
It is generally accepted that the presence of AD pathology is associated with clinical symptoms of DLB and rates of disease progression [12, 13]. Previous studies suggested that plasma p-tau181 and Aβ42 may be available candidates to screen for DLB [14]. In this study, we found that plasma Aβ42 was higher in DLB than in AD but lower than in HCs. Aβ42 levels were correlated with MMSE scores over the whole cohort, which verified that plasma Aβ42 might be a viable biomarker for AD, whereas Aβ42 levels were not correlated with neuropsychological scores in DLB. Lin et al. suggested that Aβ42 did not correlate with MMSE scores, which aligned with our findings [14]; they found that patients with DLB had Aβ42 levels that were comparable to those of controls. However, a multicenter study showed that cerebral Aβ positivity in patients with DLB was associated with MMSE scores [15]. In summary, although cerebral Aβ was a good predictor of cognitive impairment in patients with DLB, we found that plasma Aβ42 was negatively associated with global cognitive function. Aβ42 showed good performance in discriminating DLB from HCs (AUC 0.7431) and AD (AUC 0.8892). The origin of Aβ42 in plasma is still controversial. Previous experiments found that the plasma Aβ42 concentration was consistent with the cerebrospinal fluid Aβ42 concentration [16–19], while recent studies suggested that plasma Aβ42 is mainly derived from peripheral tissues and does not reflect the transformation and metabolism of Aβ in the brain [20].
Recent studies have shown that p-tau181 can highly accurately distinguish between cerebral Aβ-positive and Aβ-negative AD patients as well as DLB patients and is closely related to Aβ and tau pathophysiology [21–24]. We found that the plasma concentration of p-tau181 was significantly associated with the plasma concentration of Aβ42 in our cohort, which strengthened the relationship between DLB and AD pathology. Our results suggested that the plasma p-tau181 was lower in DLB than in HCs, contrary to previous findings that the plasma p-tau181 concentration in DLB patients was significantly higher than that in HCs [14, 26]. This unexpected result may be attributed to the presence of some form of preclinical neurodegenerative disease in our healthy participants, leading to elevated levels of plasma p-tau181. Another possibility is that there are individual differences in protein synthesis, degradation, or clearance, which could lead to higher levels of plasma p-tau181 in some HCs. In addition, this is a small-scale study, and it is possible that our sample size may not be representative of the larger population. In the comparison between DLB and AD, we found p-tau181 was lower in DLB than in AD, which was consistent with previous studies [25, 26]. We also found a significant association between higher plasma p-tau181 and lower MMSE scores in DLB. Plasma p-tau181 has been demonstrated to be strongly associated with cognitive impairment in AD [27], and a similar correlation was found in our DLB cohort. A recent study also suggests that high plasma p-tau181 might be a potential biomarker to predict rapid cognitive decline in patients with DLB over time [25]. Furthermore, previous studies failed to observe a significant association between plasma p-tau181 and longitudinal MMSE scores of non-demented patients with Parkinson’s disease [28, 29], another type of Lewy body disease, indicating that cognitive impairment in patients with DLB might be due to AD co-pathology. Plasma p-tau181 significantly discriminated DLB from HCs (AUC 0.7762) but only modestly discriminated between DLB and AD (AUC 0.5969), which is in keeping with previous work [26, 30].
The strength of the study is the use of blood tests for the study of DLB participants, which bridges the gap in the field of DLB blood-based biomarkers. Ample evidence has shown that plasma p-tau181 and Aβ42 are potential blood biomarkers for AD; however, few studies have focused on their implications in DLB, a neurodegenerative dementia that shares part of the same pathology as AD. We not only analyzed the biomarker levels in DLB patients and healthy participants but also compared biomarker levels between DLB patients and AD patients, and these biomarkers may be used in differential diagnosis. Our results may extend the differential screening ability of Aβ42 and p-tau181 to other neurodegenerative diseases. This study also has some limitations. First, owing to its retrospective design, there were missing data for some variables, including longitudinal data, years of education, and neuropsychological assessments. Given the relatively low number of participants, additional studies should investigate a larger population. Second, the present study did not assess α-syn seeding activity by RT-QuIC in DLB patients, and we lacked data on genetic variants. There is evidence indicating that apolipoprotein ɛ4 carrier status might influence the diagnostic utility of plasma biomarkers [31]; therefore, further studies should be combined with genetic tests to refine the conclusion. Third, the DLB group was older than the other groups and included more men than the AD group; differences in baseline characteristics may have influenced the results. Finally, the only neuropsychological scales we used were the MMSE and the MoCA; we did not administer the Clinical Dementia Rating Scale, the Neuropsychiatric Inventory, or specific scales for DLB to stratify disease severity.
Overall, our study suggests that plasma p-tau181 and Aβ42 represent promising markers for DLB with high sensitivity and specificity. Aβ42 is more suitable than p-tau181 in the differential diagnosis of DLB from AD. Further research should explore the ability of other emerging blood-based pathological biomarkers, such as p-tau217, p-tau231, and Aβ42/40. The association between plasma biomarkers and neuropathology may be a fruitful direction for future studies.
Footnotes
ACKNOWLEDGMENTS
The authors thank all the participants for their participation.
Plasma p-tau181 and Aβ42 were measured by Beijing Huanuo Aomei Gene BioTech Co., Ltd.
FUNDING
This project has received funding from the National Key R&D Programme of China (2017YFE0118800), the Young Elite Scientists Sponsorship Program by CAST (2021QNRC001), Beijing Hospitals Authority Innovation Studio of Young Staff Funding Support (202118), the Beijing Nova Program (Z211100002121051), and the National Natural Science Foundation of China (82201568).
CONFLICT OF INTEREST
The authors have no conflicts of interest to report. Qi Qin is an Editorial Board Member of this journal but was not involved in the peer-review process for this article and did not have access to any information regarding its peer review.
DATA AVAILABILITY
The data supporting the findings of this study are available from the corresponding author on request. The data are not publicly available due to privacy concerns and ethical restrictions.
