CSF Aβ42 and Aβ42/Aβ40 Ratio in Alzheimer’s Disease and Frontotemporal Dementias
First Department of Neurology, School of Medicine, National and Kapodistrian University of Athens, Eginition Hospital, Vass. Sophias Ave. 74, 11528 Athens, Greece
Neurochemistry and Biological Markers Unit, First Department of Neurology, School of Medicine, National and Kapodistrian University of Athens, Eginition Hospital, Vass. Sophias Ave. 74, 11528 Athens, Greece
Second Department of Neurology, School of Medicine, National and Kapodistrian University of Athens, “Attikon” University General Hospital, Rimini 1, 12462 Athens, Greece
Abstract
Background: Alzheimer’s disease dementia (ADD) may manifest with atypical phenotypes, resembling behavioral variant frontotemporal dementia (bvFTD) and corticobasal syndrome (CBS), phenotypes which typically have an underlying frontotemporal lobar degeneration with tau proteinopathy (FTLD-tau), such as Pick’s disease, corticobasal degeneration (CBD), progressive supranuclear palsy (PSP), or FTLD with TDP-43 proteinopathy (FTLD-TDP). CSF biomarkers total and phosphorylated tau (τT and τP-181), and amyloid beta with 42 and 40 amino acids (Aβ42 and Aβ40) are biomarkers of AD pathology. The primary aim of this study was to compare the diagnostic accuracy of Aβ42 to Aβ42/Aβ40 ratio in: (a) differentiating ADD vs. frontotemporal dementias; (b) patients with AD pathology vs. non-AD pathologies; (c) compare biomarker ratios and composite markers to single CSF biomarkers in the differentiation of AD from FTD; Methods: In total, 263 subjects were included (ADD: n = 98; bvFTD: n = 49; PSP: n = 50; CBD: n = 45; controls: n = 21). CSF biomarkers were measured by commercially available ELISAs (EUROIMMUN). Multiple biomarker ratios (Aβ42/Aβ40; τT/τP-181; τT/Aβ42; τP-181/Aβ42) and composite markers (t-tau: τT/(Aβ42/Aβ40); p-tau: τP-181/(Aβ42/Aβ40) were calculated. ROC curve analysis was performed to compare AUCs of Aβ42 and Aβ42/Aβ40 ratio and relevant composite markers between ADD and FTD, as defined clinically. BIOMARKAPD/ABSI criteria (abnormal τT, τP-181 Aβ42, and Aβ42/Aβ40 ratio) were used to re-classify all patients into AD pathology vs. non-AD pathologies, and ROC curve analysis was repeated to compare Aβ42 and Aβ42/Aβ40; Results: Aβ42 did not differ from Aβ42/Aβ40 ratio in the differentiation of ADD from FTD (AUCs 0.752 and 0.788 respectively; p = 0.212). The τT/Aβ42 ratio provided maximal discrimination between ADD and FTD (AUC:0.893; sensitivity 88.8%, specificity 80%). BIOMARKAPD/ABSI criteria classified 60 patients as having AD pathology and 211 as non-AD. A total of 22 had discrepant results and were excluded. Aβ42/Aβ40 ratio was superior to Aβ42 in the differentiation of AD pathology from non-AD pathology (AUCs: 0.939 and 0.831, respectively; p < 0.001). In general, biomarker ratios and composite markers were superior to single CSF biomarkers in both analyses. Conclusions: Aβ42/Aβ40 ratio is superior to Aβ42 in identifying AD pathology, irrespective of the clinical phenotype. CSF biomarker ratios and composite markers provide higher diagnostic accuracy compared to single CSF biomarkers.
Untitled section
Keywords: cerebrospinal fluid, amyloid beta with 42 amino acids, tau proteins, Alzheimer’s disease, frontotemporal dementia
Article notes
Untitled section
Received 2023 Jan 10; Revised 2023 Feb 15; Accepted 2023 Feb 17; Collection date 2023 Feb.
1. Introduction
Over the past three decades, advances in cerebrospinal fluid (CSF) and positron emission tomography (PET) biomarkers have been the defining factor in evolving the conceptual framework of Alzheimer’s disease (AD) from a simple clinical entity, characterized by amnestic-predominant dementia, to a biological-clinical continuum, with diverse clinical manifestations [1]. The defining neuropathological lesions of AD are amyloid plaques (extracellular accumulation of pathologically misfolded β amyloid) and neurofibrillary tangles (consisting of hyper-phorphorylated tau protein), which result in neurodegeneration [2,3].
The AT(N) system has supported the classification of biomarkers of diverse modalities (CSF, PET, or MRI) into three groups: (A) for amyloidosis, (T) for tau-pathology, and (N) for neurodegeneration [4]. Within this framework, total tau protein (τΤ), phosphorylated tau protein at threonine 181 (τP-181), and amyloid beta with 42 amino acids (Aβ42) have been classified as markers of neurodegeneration, tau-pathology, and amyloidosis, respectively.
A decrease in CSF Aβ42 is characteristic of AD. However, due to significant inter-subject variability in Aβ42 levels, defining an Aβ42 cut-off with high diagnostic accuracy for discrimination between AD and non-AD pathologies has been problematic [5]. Moreover, Aβ42 measurement is particularly sensitive to alterations in pre-analytical factors [6,7]. Several studies have supported that the incorporation of CSF amyloid beta with 40 amino acids (Aβ40), which is a reflection of total CSF amyloid levels, by use of the Aβ42/Aβ40 ratio, is a better marker of the relatively selective decrease in Aβ42 in AD [8,9].
Most of the studies comparing the diagnostic accuracy of Ab42 to Aβ42/Aβ40 ratio have defined AD or other dementias by use of clinical criteria [10,11,12,13,14,15,16,17,18,19,20]. However, this approach is problematic, since AD may manifest with atypical non-amnestic presentations, including a language presentation (i.e., logopenic variant primary progressive aphasia), a visuospatial presentation (i.e., posterior cortical atrophy), a dysexecutive presentation (mimicking behavioral variant of frontotemporal dementia) and corticobasal syndrome [21,22]. Thus, the use of clinical criteria to define AD will result in the misclassification of AD patients as non-AD in cases of atypical manifestations and vice versa.
Several studies have compared Aβ42 and Aβ42/Aβ40 ratio to amyloid-PET, in an attempt to investigate the optimal CSF amyloid marker [8,9,23,24,25]. Most of these studies conclude that the Aβ42/Aβ40 ratio results in higher concordance with amyloid-PET compared to Aβ42 [8,9,24,25], although a single study did not report a difference between the two markers [23]. However, most of these studies have only included healthy subjects or patients with mild cognitive impairment or dementia due to AD, without the inclusion of other dementias. To date, a study comparing CSF Aβ42 to Aβ42/Aβ40 in a cohort with neuropathological confirmation of clinical diagnoses is lacking.
The present study aimed to compare the predictive values of Aβ42 and Aβ42/Aβ40 ratio for an underlying AD pathology in a cohort of patients with diverse dementing disorders. For the purposes of this study, we selected to include patients with a clinical diagnosis of AD dementia (ADD) and various frontotemporal dementias (FTD), including behavioral variant FTD (bvFTD), progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD), as well as healthy subjects. We opted not to include patients with Lewy body dementia, due to the high incidence of co-occurrence of AD in these patients, rendering interpretation of CSF biomarkers problematic in the absence of biomarkers in other modalities (i.e., PET-CT).
Initially, the diagnostic accuracies of Aβ42 and Aβ42/Aβ40 as predictors of AD dementia were compared among study groups based on clinical diagnoses, in accordance with the methodology applied in most relevant studies. We then applied the BIOMARKAPD/ABSI criteria in all patients, irrespective of their clinical phenotype, and re-classified them as having an AD or a non-AD underlying pathology [26,27]. CSF Aβ42 and Aβ42/Aβ40 were subsequently re-applied in this setting, in order to compare their diagnostic accuracy for underlying AD pathology.
To further compare Aβ42 and Aβ42/Aβ40, we included several composite markers in our analyses, by replacement of Aβ42 with Aβ42/Aβ40 in the τT/Aβ42 and τP-181/Aβ42 ratios.
2. Materials and Methods
2.1. Patients
The medical files of all patients with available data on CSF biomarkers Aβ42, Aβ40, τT, and τP-181, who were admitted from 2011 to 2021 to the “Neurodegenerative Disorders and Epilepsy” Ward of our hospital, were retrospectively reviewed. For the purposes of this study, subjects were included if they fulfilled the established diagnostic criteria for the following diseases: (a) ADD [21]; (b) bvFTD [28]; (c) PSP [29], and (d) CBD [30]. For comparison reasons, a control group was included. This consisted of otherwise healthy subjects, with no comorbidities, undergoing knee or hip joint surgery or hernia repair under spinal anesthesia. These subjects had a negative history of cognitive or behavioral/psychiatric disorders and no clinical evidence of any major disease. All subjects had normal scores on neuropsychological testing (Mini Mental State Examination and Frontal Assessment Battery) [31,32].
2.2. CSF Sampling and Biomarker Measurements
All patients underwent lumbar puncture at 10–11 a.m., after overnight fasting, based on standard operating procedures in accordance to recommendations to standardize pre-analytical confounding factors in AD CSF biomarkers [33].
CSF biomarkers Aβ42, Aβ40, τT, and τP-181 were measured in duplicate with ELISA by commercially available kits (EUROIMMUN Beta-Amyloid (1–42) ELISA; EUROIMMUN Beta-Amyloid (1–40) ELISA; EUROIMMUN Total-Tau ELISA; EUROIMMUN pTau (181) ELISA respectively), according to manufacturer instructions.
Our laboratory implements both internal and external quality control measures to ensure the accuracy of measurements longitudinally. Specifically, for internal control a pooled CSF sample is used in every test run, resulting in an over >90% between-run precision. As for external control, we participate in “The Alzheimer’s Association’s QC program”, which provides additional external pooled CSF samples for validating results reliability regardless of the kit’s lot number.
Additionally, the following CSF biomarker ratios, which incorporate two CSF AD biomarkers were calculated: Aβ42 to Aβ40 (Aβ42/Aβ40), τT to τP-181 (τT/τP-181), τT το Aβ42 (τT/Aβ42) and τP-181 to Aβ42 (τP-181/Aβ42). All of these ratios have been previously applied in studies in an effort to increase the diagnostic accuracy of CSF biomarkers.
Lastly, in an effort to incorporate two CSF AD biomarkers in a single marker, the following composite markers were calculated:
- (a)Composite t-tau marker: τT/(Aβ42/Aβ40).
The τT/Aβ42 ratio has been previously applied as an AD neurochemical marker, based on the observed increase in τT and decrease in Aβ42 in patients with an underlying AD pathology. Several studies support that the Aβ42/Aβ40 ratio may provide improved diagnostic accuracy for amyloid pathology compared to Aβ42. To look into this hypothesis, we introduced this composite marker.
- (b)Composite p-tau marker: τP-181/(Aβ42/Aβ40).
The τP-181/Aβ42 ratio has been previously applied as an AD neurochemical marker, based on the observed increase in τP-181 and decrease in Aβ42 in patients with an underlying AD pathology. As mentioned previously, substituting Aβ42 with the Aβ42/Aβ40 ratio may provide improved diagnostic accuracy for amyloid pathology. Composite p-tau marker:
2.3. Ethical Considerations
All patients or their next of kin (in cases of compromised mental capacity) provided written informed consent for participation in this study. The study was approved by the Scientific and Ethics Committee of Eginition Hospital and was performed in accordance with the guidelines of the 1964 Declaration of Helsinki.
2.4. Statistical Analysis
The normality of distribution and homogeneity of variances were checked by Shapiro–Wilk’s and Levene’s tests, respectively. Comparison of clinical, neuropsychological, and CSF biomarker characteristics between study groups was performed by ANOVA (with Bonferroni correction for multiple comparisons) or Kruskal–Wallis test as appropriate.
We performed two sets of analyses. The initial analysis was based on the clinical diagnoses of the study subjects. Thus, Receiver Operating Characteristic (ROC) Curve analysis was performed to compare the diagnostic accuracy of all CSF biomarkers, biomarker ratios, and composite biomarkers in differentiating between patients with AD dementia vs. all other clinical groups. Area under the curve (AUC), 95% confidence interval of the AUC, cut-off point with optimal diagnostic accuracy (defined as maximal sensitivity and specificity), as well as specificity, sensitivity, and Youden Index (YI) of optimal cut-off points, were calculated.
In order to look into possible differences between AUCs of ROC curves of various biomarkers in the identification of AD dementia, the De Long method was applied. In an effort to compare the diagnostic accuracy of Aβ42 vs. Aβ42/Aβ40 ratio, the following comparisons of ROC curves were performed: (a) Aβ42 vs. Aβ42/Aβ40 ratio; (b) τT/Aβ42 vs. composite t-tau: (c) τP-181/Aβ42 vs. composite p-tau.
The second analysis aimed to investigate the diagnostic accuracy of single CSF biomarkers, biomarker ratios, and composite biomarkers in identifying AD pathology irrespective of the clinical phenotype. To this end, a two-step process was applied, as described elsewhere.
Initially, CSF biomarkers were transformed into binary variables (i.e., normal or abnormal), based on cut-off values of the Unit of Neurochemistry and Biomarkers (Aβ42 < 480 pg/mL; τT > 400 pg/mL; τP-181 > 60 pg/mL; Aβ42/Aβ40 < 0.094), as described previously [34]. Subjects with abnormal Aβ42, τT, τP-181, and Aβ42/Aβ40 ratio (A+T+N+) were considered to harbor AD pathology, based on the BIOMARKAPD/ABSI criteria [26,27]. Thus, all patients, irrespective of their clinical phenotype were classified into two categories: AD vs. nonAD.
We elected to apply the most stringent classification criterion (i.e., all three CSF biomarkers abnormal) for AD pathology identification in order to increase specificity, at the expense of sensitivity. For amyloid pathology (A) in particular, there were two available established CSF markers: Aβ42 and the Aβ42/Aβ40 ratio. In an effort to increase specificity, subjects with a decrease in both markers (Aβ42 and Aβ42/Aβ40) were considered A(+), and only subjects with normal values in both biomarkers were considered A(−). Of the 263 subjects, 22 (8.4%; AD dementia: 16 patients; CBD: 3 patients; PSP: 1 patients; controls: 2 subjects) had discrepant results in amyloid pathology identification based on Aβ42 and Aβ42/Aβ40 ratio and were excluded from this analysis.
Thus, this second analysis included 82 patients with AD pathology and 140 patients with non-AD pathology: 49 bvFTD patients, 42 CBD patients, and 49 PSP patients. The majority of the non-AD pathology patients are considered to have an underlying frontotemporal lobar degeneration (FTLD), with most PSP and CBD patients harboring an FTLD with tau proteinopathy (FTLD-tau) and most bvFTD patients harboring either an FTLD-tau or FTLD-TDP43 proteinopathy. Thus, in essence, the second analysis referred to a comparison of Aβ42 to Aβ42/Aβ40 ratio in the differentiation of AD pathology from FTLD. However, due to the lack of neuropathological confirmation, we elected to use the term non-AD pathology instead of FTLD, since the presumed underlying pathology is based on CSF biomarker profiles.
Following this classification of all subjects irrespective of their phenotype into the AD and nonAD groups, ROC curve analysis was applied to determine the discriminative power of CSF biomarkers, biomarker ratios, and composite markers for this differentiation. Cut-off points were determined based on the maximal combined sensitivity and specificity criterion. Area under the curve (AUC), 95% confidence interval of the AUC, YI, sensitivity, and specificity were also calculated.
In order to look into possible differences between AUCs of ROC curves of various biomarkers in AD pathology identification, the De Long method was applied. In an effort to compare the diagnostic accuracy of Aβ42 vs. Aβ42/Aβ40 ratio, the following comparisons of ROC curves were performed: (a) Aβ42 vs. Aβ42/Aβ40 ratio; (b) τT/Aβ42 vs. composite t-tau: (c) τP-181/Aβ42 vs. composite p-tau.
All analyses were performed by IBM SPSS Statistics® version 23.0.0.0 (SPSS Inc., Chicago, IL, USA, 2013). All graphs were designed using GraphPad Prism®, version 5.03 (GraphPad Software Inc., La Jolla, CA, USA, 2009).
3. Results
3.1. Clinical and Demographic Data
In total, 263 subjects were included (ADD: 98 patients; bvFTD: 49 patients; PSP: 50 patients; CBD: 45 patients; controls: 21 subjects). Study groups differed in age (p = 0.003), with the control group exhibiting the greatest age compared to other groups and sex distribution (p = 0.005). Control subjects performed significantly better in the MMSE and FAB tests compared to patient groups, as expected (p < 0.001) (Table 1).
| ADD n = 98 | bvFTD n = 49 | PSP n = 50 | CBD n = 45 | Controls n = 21 | p Value | |
|---|---|---|---|---|---|---|
| Demographic/clinical data | ||||||
| Gender (m/f) | 38/60 | 28/21 | 28/28 | 21/24 | 3/18 | 0.005 * |
| Age (y) | 65 (57–72) | 63 (58–70) | 65.5 (60–69) | 65 (60–74) | 74 (69–76) | 0.003 † |
| Age at onset (y) | 58.5 (53–70) | 61 (55–66) | 62 (57–67) | 62.5 (57.5–71) | NA | 0.413 † |
| Disease duration (m) | 36 (24–48) | 30 (24–48) | 30 (18–48) | 24 (18–48) | NA | 0.593 † |
| Education (y) | 12 (6–16) | 12 (9–15) | 12 (6–15) | 9 (6–15.5) | 14.5 (12–17) | 0.720 † |
| MMSE | 17 (13–21) | 23 (16–28) | 27 (23–29) | 22 (17–26) | 29 (29–30) | <0.001 † |
| FAB | 9 (5–11) | 9 (6–14) | 11 (8–14) | 9 (6–12) | 15 (14–16) | <0.001 † |
| BIOMARKAPD/ABSI Classification (n/%) | ||||||
| AD | 46 (47%) | 0 (0%) | 1 (2%) | 12 (27%) | 1 (5%) | |
| non-AD | 36 (37%) | 49 (100%) | 48 (96%) | 30 (67%) | 48 (86%) | |
| discrepancy | 16 (16%) | 0 (0%) | 1 (2%) | 3 (6%) | 2 (9%) | |
| CSF biomarkers | ||||||
| Aβ42 | 421 (274) | 697 (342) | 616 (268) | 577 (333) | 846 (311) | <0.001 ‡ |
| Aβ40 | 6215 (2795) | 5630 (2648) | 4900 (2545) | 5975 (3083) | 8221 (2875) | <0.001 ‡ |
| τT | 577 (255) | 305 (123) | 233 (120) | 384 (176) | 304 (102) | <0.001 ‡ |
| τP-181 | 115 (81–151) | 31 (25–39) | 30 (21–39) | 73 (34–103) | 42 (29–55) | <0.001 † |
| CSF biomarker ratios | ||||||
| Aβ42/Aβ40 | 0.075 (0.042) | 0.132 (0.049) | 0.142 (0.068) | 0.105 (0.062) | 0.106 (0.030) | <0.001 ‡ |
| τP-181/τT | 0.217 (0.183–0.263) | 0.120 (0.084–0.164) | 0.137 (0.106–0.176) | 0.186 (0.124–0.252) | 0.144 (0.126–0.153) | <0.001 † |
| τT/Aβ42 | 1.50 (0.98–2.22) | 0.44 (0.32–0.69) | 0.38 (0.28–0.49) | 0.63 (0.40–1.29) | 0.32 (0.27–0.46) | <0.001 † |
| τP-181/Aβ42 | 0.312 (0.223–0.514) | 0.049 (0.035–0.072) | 0.051 (0.037–0.063) | 0.098 (0.054–0.305) | 0.047 (0.036–0.068) | <0.001 † |
| CSF composite markers | ||||||
| τT/(Aβ42/Aβ40)/1000 | 8.34 (4.67–13.18) | 2.25 (1.53–3.41) | 1.75 (1.08–2.45) | 4.01 (2.14–7.18) | 2.67 (1.93–3.98) | <0.001 † |
| τP-181/(Aβ42/Aβ40)/1000 | 1.92 (0.91–2.98) | 0.26 (0.17–0.37) | 0.22 (0.13–0.36) | 0.73 (0.30–1.70) | 0.34 (0.27–0.52) | <0.001 † |
3.2. Comparison of CSF Biomarkers, Biomarker Ratios, and Composite Markers with ANOVA
Study groups differed significantly in all CSF biomarkers, biomarker ratios, and composite markers, with ADD patients exhibiting higher values in τT, τP-181, τP-181/τT ratio, τT/Aβ42 ratio, τP-181/Aβ42 ratio, and in all composite markers. Additionally, ADD patients exhibited the lowest values of Aβ42 and Aβ42/Aβ40 ratio as expected p < 0.001 for all comparisons) (Table 1, Figure 1).
3.3. ROC Curve Analysis of CSF Biomarkers for the Differentiation between Alzheimer’s Disease Dementia vs. All Other Clinical Groups
The composite t-tau and p-tau markers provided comparable high diagnostic accuracies (AUCs: 0.868 and 0.869, respectively). Regarding CSF biomarker ratios, the τT/Aβ42 and τP-181/Aβ42 ratios resulted in higher AUCs (0.893 and 0.878, respectively) compared to Aβ42/Aβ40 and τP-181/τT (0.788 and 0.755, respectively). Among single CSF biomarkers, τP-181 provided maximal AUC, followed by τT and Aβ42 (0.886, 0.840, and 0.752, respectively) (Table 2) (Figure 2a).
| AUC | 95% CI | p-Value | YI | Cut-Off | Sens | Spec | |
|---|---|---|---|---|---|---|---|
| CSF biomarkers | |||||||
| Aβ42 | 0.752 | 0.695–0.803 | <0.0001 | 0.411 | <408.5 | 65.3 | 75.8 |
| τΤ | 0.840 | 0.790–0.882 | <0.0001 | 0.518 | >455.4 | 63.3 | 88.5 |
| τP-181 | 0.886 | 0.841–0.922 | <0.0001 | 0.684 | >60.3 | 87.8 | 80.6 |
| CSF biomarker ratios | |||||||
| Aβ42/Aβ40 | 0.788 | 0.734–0.836 | <0.0001 | 0.481 | <0.096 | 76.5 | 71.5 |
| τP-181/τT | 0.755 | 0.698–0.806 | <0.0001 | 0.482 | >0.181 | 75.5 | 72.7 |
| τT/Aβ42 | 0.893 | 0.850–0.928 | <0.0001 | 0.688 | >0.75 | 88.8 | 80.0 |
| τP-181/Aβ42 | 0.878 | 0.832–0.915 | <0.0001 | 0.714 | >0.122 | 87.8 | 83.6 |
| CSF composite markers | |||||||
| τT/(Aβ42/Aβ40)/1000 | 0.868 | 0.821–0.907 | <0.0001 | 0.617 | >3.96 | 84.7 | 77.0 |
| τP-181/(Aβ42/Aβ40)/1000 | 0.869 | 0.822–0.908 | <0.0001 | 0.643 | >0.55 | 86.7 | 77.6 |
4. Discussion
The primary aim of this study was to compare the predictive value of Aβ42/Aβ40 ratio to Aβ42 in identifying: (a) Alzheimer’s disease dementia (ADD) from other clinical phenotypes and (b) Alzheimer’ disease pathology from non-AD pathologies. Regarding the clinical distinction of ADD from other phenotypes, the Aβ42/Aβ40 ratio and Aβ42 provided comparable diagnostic accuracy in our cohort. In agreement with our study, most relevant studies in the literature relying on clinical criteria for AD definition support that the Aβ42/Aβ40 ratio provides greater AUC values compared to Aβ42 in diagnosing ADD [12,13,14,15,16,17,18,19,20]. Few studies, however, have conflicting results, supporting the superiority of Aβ42 over theAβ42/Aβ40 ratio or no difference between the two markers [10,11]. Importantly, contrary to our study, these studies do not include statistical comparisons of AUC values and rely on numerical differences between AUCs.
Importantly, the Aβ42/Aβ40 ratio provided significantly greater diagnostic accuracy compared to Aβ42 in the differentiation of AD pathology from non-AD pathologies. Likewise, Aβ42/Aβ40-derived composite markers produced greater AUCs το Aβ42-derived ratios, although these differences did not reach statistical significance. Due to the inclusion in the present study of phenotypes that typically have an underlying FTLD pathology, either with tau proteinopathy (FTLD-tau) or TDP-43 (FTLD-TDP), this finding signifies the importance of the Aβ42/Aβ40 ratio in differentiating between AD pathology and FTLD. However, this finding needs verification by studies with both CSF biomarker and neuropathological data available. This finding is particularly important from a clinical and research perspective because it highlights the importance of the Aβ42/Aβ40 ratio in the in vivo identification of amyloid pathology in patients with diverse clinical phenotypes. Μost studies comparing CSF biomarkers to amyloid-PET support this finding [8,9,24,25], although a study reported conflicting results, depending on the assay used [23].
A third finding of our study is the significant disparity between clinical diagnosis and underlying pathology, highlighting the significant clinical heterogeneity of AD [21,22] as well as the significant pathological heterogeneity of phenotypes such as CBS [35,36]. Multiple clinical—pathological studies have supported this clinical—pathological overlap [37,38,39,40]. However, this disparity was particularly high for ADD patients in our cohort (47% of ADD patients had a CSF AD profile, 16% had discrepant results and 37% had non-AD pathologies) and exceedingly low for bvFTD (all patients had non-AD CSF profile).
This finding can be attributed to the criteria we applied for the definition of the CSF AD biochemical profile. We elected for the purposes of this study to define the AD CSF profile according to the BIOMARKAPD/ABSI criteria, which require all three major biomarker groups (amyloid, tau, and neurodegeneration) to be abnormal. Additionally, we elected to define as A(+) patients with abnormal values in both Aβ42 and Aβ42/Aβ40 ratio. These stringent criteria were applied to enhance the specificity of AD classification at the expense of sensitivity, in the absence of other biomarkers such as amyloid PET, and with no neuropathological data available. Thus, the classification of patients in AD or non-AD pathologies do not reflect the routine clinical practice, wherein a single amyloid CSF biomarker is sufficient to characterize the (A) status in the ATN system, and patients with A(+)T(+)N(−) are classified as AD. To support this hypothesis, when applying the ATN criteria based on Aβ42/Aβ40 ratio, 72% of patients were classified as AD pathology (data not shown).
In CSF biomarker studies looking into inherent differences among neurodegenerative disorders, the significant clinical—pathological overlap in neurodegenerative disorders can be overcome if a two-step algorithm is applied, as proposed previously [41,42]. As a first step, all patients, irrespective of their clinical phenotype, should initially be classified as AD or non-AD based on a biomarker taxonomic system (e.g., ATN or BIOMARKAPD/ABSI). This will assist in identifying atypical presentations of AD (e.g., PPA, behavioral-frontal dementia, CBS, etc.) and avoid misclassification of these subjects. The second step would involve the direct comparison of biomarker levels among different disorders.
An interesting finding was the improved diagnostic accuracy of CSF biomarker ratios and composite markers compared to single CSF biomarkers. The τT/Aβ42 ratio and the composite markers (t-tau and p-tau composite marker) provided excellent discriminative power for AD pathology identification. Although these ratios and composite markers lack inherent biological meaning, the incorporation of two biomarkers (in the case of CSF biomarker ratios) or three biomarkers (in the case of composite markers) in a single continuous variable for each patient increases discriminative power compared to single CSF biomarkers. To this extent, CSF biomarker ratios have been previously implemented and have yielded excellent concordance with neuropathological data in FTD [43,44]. Additionally, CSF ratios were superior to single CSF biomarkers in the differential diagnosis of AD from bvFTD [43,45,46,47].
An additional theoretical advantage of using CSF biomarker ratios and composite markers instead of single CSF biomarkers is that they may assist in decreasing the inter-subject variability in pre-analytical variables, such as type of sampling and storage tubes, incubation time, number of freeze/thaw cycles, type of pipettes used, etc. This has previously been proven for the Aβ42/Aβ40 ratio [6,7].
The present study has certain limitations. Firstly, our cohort, as is the case with all relevant studies in the literature, lacks neuropathological confirmation, which is the golden standard for diagnosis in neurodegenerative disorders. For this reason, we only included well-characterized patients based on the most recently established diagnostic criteria with adequate follow-up. A second limitation is the absence of a biomarker of a different modality (e.g., amyloid PET), to use as a reference. However, PET-CT can only provide information on a single axis of the ATN triad. For this reason, we selected to apply the BIOMARKAPD/ABSI criteria, to classify patients into AD or non-AD pathologies. Additionally, in order to minimize the effect of cyclical error in comparing Aβ42 το Aβ42/Aβ40, we used all four major AD CSF biomarkers for AD classification (Aβ42, Aβ42/Aβ40 ratio, τT, and τP-181).
Our study, in accordance with the literature, supports the improved diagnostic accuracy of the Aβ42/Aβ40 ratio compared to Aβ42 in identifying AD pathology. Moreover, we highlighted the exceptionally high diagnostic accuracy of composite markers. Further studies, with neuropathological confirmation, are needed, to establish the correlation between CSF biomarkers and neuropathological characteristics.
Acknowledgments
We would like to thank the patients and their relatives for their contribution to this research project.
Institutional Review Board Statement
The study was approved by the Scientific and Ethics Committee of Eginition Hospital (protocol code: 244, date of approval: 5 July 2010) and was performed in accordance with the guidelines of the 1964 Declaration of Helsinki.
Informed Consent Statement
Written informed consent for participation in this study was provided by all patients or their next of kin in cases of compromised mental capacity.
Data Availability Statement
The data presented in this study are available on request from the corresponding author.
Conflicts of Interest
G.P.P. received fees from Biogen International as a consultant of the advisory board. The other authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding Statement
This research received no external funding.
Footnotes
Footnote Group
References
Untitled section
References
- 1.Jack C.R., Jr., Albert M.S., Knopman D.S., McKhann G.M., Sperling R.A., Carrillo M.C., Thies B., Phelps C.H. Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7:257–262. doi: 10.1016/j.jalz.2011.03.004.
- 2.Glenner G.G., Wong C.W. Alzheimer’s disease: Initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem. Biophys. Res. Commun. 1984;120:885–890. doi: 10.1016/S0006-291X(84)80190-4.
- 3.Williams D.R. Tauopathies: Classification and clinical update on neurodegenerative diseases associated with microtubule-associated protein tau. Intern. Med. J. 2006;36:652–660. doi: 10.1111/j.1445-5994.2006.01153.x.
- 4.Jack C.R., Jr., Bennett D.A., Blennow K., Carrillo M.C., Dunn B., Haeberlein S.B., Holtzman D.M., Jagust W., Jessen F., Karlawish J., et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimer Dement. 2018;14:535–562. doi: 10.1016/j.jalz.2018.02.018.
- 5.Hansson O., Lehmann S., Otto M., Zetterberg H., Lewczuk P. Advantages and disadvantages of the use of the CSF Amyloid beta (Abeta) 42/40 ratio in the diagnosis of Alzheimer’s Disease. Alzheimer’s Res. Ther. 2019;11:34. doi: 10.1186/s13195-019-0485-0.
- 6.Willemse E., van Uffelen K., Brix B., Engelborghs S., Vanderstichele H., Teunissen C. How to handle adsorption of cerebrospinal fluid amyloid beta (1-42) in laboratory practice? Identifying problematic handlings and resolving the issue by use of the Abeta(42)/Abeta(40) ratio. Alzheimer’s Dement. 2017;13:885–892. doi: 10.1016/j.jalz.2017.01.010.
- 7.Gervaise-Henry C., Watfa G., Albuisson E., Kolodziej A., Dousset B., Olivier J.-L., Jonveaux T.R., Malaplate-Armand C. Cerebrospinal Fluid Aβ42/Aβ40 as a Means to Limiting Tube- and Storage-Dependent Pre-Analytical Variability in Clinical Setting. J. Alzheimer’s Dis. 2017;57:437–445. doi: 10.3233/JAD-160865.
- 8.Amft M., Ortner M., Eichenlaub U., Goldhardt O., Diehl-Schmid J., Hedderich D.M., Yakushev I., Grimmer T. The cerebrospinal fluid biomarker ratio Abeta42/40 identifies amyloid positron emission tomography positivity better than Abeta42 alone in a heterogeneous memory clinic cohort. Alzheimer’s Res. Ther. 2022;14:60. doi: 10.1186/s13195-022-01003-w.
- 9.Niemantsverdriet E., Ottoy J., Somers C., De Roeck E., Struyfs H., Soetewey F., Verhaeghe J., Bossche T.V.D., Van Mossevelde S., Goeman J., et al. The Cerebrospinal Fluid Abeta1-42/Abeta1-40 Ratio Improves Concordance with Amyloid-PET for Diagnosing Alzheimer’s Disease in a Clinical Setting. J. Alzheimer’s Dis. 2017;60:561–576. doi: 10.3233/JAD-170327.
- 10.Hertze J., Minthon L., Zetterberg H., Vanmechelen E., Blennow K., Hansson O. Evaluation of CSF biomarkers as predictors of Alzheimer’s disease: A clinical follow-up study of 4.7 years. J. Alzheimer’s Dis. 2010;21:1119–1128. doi: 10.3233/JAD-2010-100207.
- 11.Slaets S., Le Bastard N., Martin J.-J., Sleegers K., Van Broeckhoven C., De Deyn P.P., Engelborghs S. Cerebrospinal fluid Abeta1–40 improves differential dementia diagnosis in patients with intermediate P-tau181P levels. J. Alzheimer’s Dis. 2013;36:759–767. doi: 10.3233/JAD-130107.
- 12.Dumurgier J., Schraen S., Gabelle A., Vercruysse O., Bombois S., Laplanche J.-L., Peoc’H K., Sablonnière B., Kastanenka K.V., Delaby C., et al. Cerebrospinal fluid amyloid-beta 42/40 ratio in clinical setting of memory centers: A multicentric study. Alzheimer’s Res. Ther. 2015;7:30. doi: 10.1186/s13195-015-0114-5.
- 13.Janelidze S., Zetterberg H., Mattsson N., Palmqvist S., Vanderstichele H., Lindberg O., Westen D., Stomrud E., Minthon L., Blennow K., et al. CSF Abeta42/Abeta40 and Abeta42/Abeta38 ratios: Better diagnostic markers of Alzheimer disease. Ann. Clin. Transl. Neurol. 2016;3:154–165. doi: 10.1002/acn3.274.
- 14.Lehmann S., Delaby C., Boursier G., Catteau C., Ginestet N., Tiers L., Maceski A., Navucet S., Paquet C., Dumurgier J., et al. Relevance of Abeta42/40 Ratio for Detection of Alzheimer Disease Pathology in Clinical Routine: The PLM(R) Scale. Front. Aging Neurosci. 2018;10:138. doi: 10.3389/fnagi.2018.00138.
- 15.Shoji M., Matsubara E., Kanai M., Watanabe M., Nakamura T., Tomidokoro Y., Shizuka M., Wakabayashi K., Igeta Y., Ikeda Y., et al. Combination assay of CSF tau, A beta 1–40 and A beta 1–42(43) as a biochemical marker of Alzheimer’s disease. J. Neurol. Sci. 1998;158:134–140. doi: 10.1016/S0022-510X(98)00122-1.
- 16.Lewczuk P., Esselmann H., Otto M., Maler J.M., Henkel A.W., Henkel M.K., Eikenberg O., Antz C., Krause W.-R., Reulbach U., et al. Neurochemical diagnosis of Alzheimer’s dementia by CSF Abeta42, Abeta42/Abeta40 ratio and total tau. Neurobiol. Aging. 2004;25:273–281. doi: 10.1016/S0197-4580(03)00086-1.
- 17.Gabelle A., Roche S., Gény C., Bennys K., Labauge P., Tholance Y., Quadrio I., Tiers L., Gor B., Boulanghien J., et al. Decreased sAbetaPPbeta, Abeta38, and Abeta40 cerebrospinal fluid levels in frontotemporal dementia. J. Alzheimer’s Dis. 2011;26:553–563. doi: 10.3233/JAD-2011-110515.
- 18.Nutu M., Zetterberg H., Londos E., Minthon L., Nägga K., Blennow K., Hansson O., Öhrfelt A. Evaluation of the cerebrospinal fluid amyloid-beta1-42/amyloid-beta1-40 ratio measured by alpha-LISA to distinguish Alzheimer’s disease from other dementia disorders. Dement. Geriatr. Cogn. Disord. 2013;36:99–110. doi: 10.1159/000353442.
- 19.Struyfs H., Van Broeck B., Timmers M., Fransen E., Sleegers K., Van Broeckhoven C., De Deyn P.P., Streffer J.R., Mercken M., Engelborghs S. Diagnostic Accuracy of Cerebrospinal Fluid Amyloid-beta Isoforms for Early and Differential Dementia Diagnosis. J. Alzheimer’s Dis. 2015;45:813–822. doi: 10.3233/JAD-141986.
- 20.Bousiges O., Cretin B., Lavaux T., Philippi N., Jung B., Hezard S., Heitz C., Demuynck C., Gabel A., Martin-Hunyadi C., et al. Diagnostic Value of Cerebrospinal Fluid Biomarkers (Phospho-Tau181, total-Tau, Abeta42, and Abeta40) in Prodromal Stage of Alzheimer’s Disease and Dementia with Lewy Bodies. J. Alzheimer’s Dis. 2016;51:1069–1083. doi: 10.3233/JAD-150731.
- 21.McKhann G.M., Knopman D.S., Chertkow H., Hyman B.T., Jack C.R., Jr., Kawas C.H., Klunk W.E., Koroshetz W.J., Manly J.J., Mayeux R., et al. The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. J. Alzheimers Assoc. 2011;7:263–269. doi: 10.1016/j.jalz.2011.03.005.
- 22.Dubois B., Feldman H.H., Jacova C., Hampel H., Molinuevo J.L., Blennow K., DeKosky S.T., Gauthier S., Selkoe D., Bateman R., et al. Advancing research diagnostic criteria for Alzheimer’s disease: The IWG-2 criteria. Lancet Neurol. 2014;13:614–629. doi: 10.1016/S1474-4422(14)70090-0.
- 23.Palmqvist S., Zetterberg H., Mattsson N., Johansson P., For the Alzheimer’s Disease Neuroimaging Initiative. Minthon L., Blennow K., Olsson M., For the Swedish BioFINDER study group. Hansson O. Detailed comparison of amyloid PET and CSF biomarkers for identifying early Alzheimer disease. Neurolongy. 2015;85:1240–1249. doi: 10.1212/WNL.0000000000001991.
- 24.Leuzy A., Chiotis K., Hasselbalch S.G., Rinne J.O., de Mendonça A., Otto M., Lleó A., Castelo-Branco M., Santana I., Johansson J., et al. Pittsburgh compound B imaging and cerebrospinal fluid amyloid-beta in a multicentre European memory clinic study. Brain. 2016;139:2540–2553. doi: 10.1093/brain/aww160.
- 25.Janelidze S., Pannee J., Mikulskis A., Chiao P., Zetterberg H., Blennow K., Hansson O. Concordance Between Different Amyloid Immunoassays and Visual Amyloid Positron Emission Tomographic Assessment. JAMA Neurol. 2017;74:1492–1501. doi: 10.1001/jamaneurol.2017.2814.
- 26.Simonsen A.H., Herukka S.-K., Andreasen N., Baldeiras I., Bjerke M., Blennow K., Engelborghs S., Frisoni G.B., Gabryelewicz T., Galluzzi S., et al. Recommendations for CSF AD biomarkers in the diagnostic evaluation of dementia. Alzheimer’s Dement. 2016;13:274–284. doi: 10.1016/j.jalz.2016.09.008.
- 27.Molinuevo J.L., Blennow K., Dubois B., Engelborghs S., Lewczuk P., Perret-Liaudet A., Teunissen C.E., Parnetti L. The clinical use of cerebrospinal fluid biomarker testing for Alzheimer’s disease diagnosis: A consensus paper from the Alzheimer’s Biomarkers Standardization Initiative. Alzheimer’s Dement. 2014;10:808–817. doi: 10.1016/j.jalz.2014.03.003.
- 28.Rascovsky K., Hodges J.R., Knopman D., Mendez M.F., Kramer J.H., Neuhaus J., Van Swieten J.C., Seelaar H., Dopper E.G.P., Onyike C.U., et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 2011;134:2456–2477. doi: 10.1093/brain/awr179.
- 29.Höglinger G.U., Respondek G., Stamelou M., Kurz C., Josephs K.A., Lang A.E., Mollenhauer B., Müller U., Nilsson C., Whitwell J.L., et al. Clinical diagnosis of progressive supranuclear palsy: The movement disorder society criteria. Mov. Disord. 2017;32:853–864. doi: 10.1002/mds.26987.
- 30.Armstrong M.J., Litvan I., Lang A.E., Bak T.H., Bhatia K.P., Borroni B., Boxer A.L., Dickson D.W., Grossman M., Hallett M., et al. Criteria for the diagnosis of corticobasal degeneration. Neurology. 2013;80:496–503. doi: 10.1212/WNL.0b013e31827f0fd1.
- 31.Folstein M.F., Robins L.N., Helzer J.E. The Mini-Mental State Examination. Arch. Gen. Psychiatry. 1983;40:812. doi: 10.1001/archpsyc.1983.01790060110016.
- 32.Dubois B., Slachevsky A., Litvan I., Pillon B. The FAB: A frontal assessment battery at bedside. Neurology. 2000;55:1621–1626. doi: 10.1212/WNL.55.11.1621.
- 33.del Campo M., Mollenhauer B., Bertolotto A., Engelborghs S., Hampel H., Simonsen A.H., Kapaki E., Kruse N., Le Bastard N., Lehmann S., et al. Recommendations to standardize preanalytical confounding factors in Alzheimer’s and Parkinson’s disease cerebrospinal fluid biomarkers: An update. Biomark. Med. 2012;6:419–430. doi: 10.2217/bmm.12.46.
- 34.Constantinides V.C., Paraskevas G.P., Boufidou F., Bourbouli M., Stefanis L., Kapaki E. Cerebrospinal fluid biomarker profiling in corticobasal degeneration: Application of the AT(N) and other classification systems. Park. Relat. Disord. 2020;82:44–49. doi: 10.1016/j.parkreldis.2020.11.016.
- 35.Constantinides V.C., Paraskevas G.P., Efthymiopoulou E., Stefanis L., Kapaki E. Clinical, neuropsychological and imaging characteristics of Alzheimer’s disease patients presenting as corticobasal syndrome. J. Neurol. Sci. 2019;398:142–147. doi: 10.1016/j.jns.2019.01.046.
- 36.Constantinides V.C., Paraskevas G.P., Paraskevas P.G., Stefanis L., Kapaki E. Corticobasal degeneration and corticobasal syndrome: A review. Clin. Park. Relat. Disord. 2019;1:66–71. doi: 10.1016/j.prdoa.2019.08.005.
- 37.Perry D.C., Brown J.A., Possin K.L., Datta S., Trujillo A., Radke A., Karydas A., Kornak J., Sias A.C., Rabinovici G.D., et al. Clinicopathological correlations in behavioural variant frontotemporal dementia. Brain. 2017;140:3329–3345. doi: 10.1093/brain/awx254.
- 38.Wakabayashi K., Takahashi H. Pathological heterogeneity in progressive supranuclear palsy and corticobasal degeneration. Neuropathology. 2004;24:79–86. doi: 10.1111/j.1440-1789.2003.00543.x.
- 39.Ling H., O’Sullivan S.S., Holton J.L., Revesz T., Massey L.A., Williams D.R., Paviour D.C., Lees A.J. Does corticobasal degeneration exist? A clinicopathological re-evaluation. Brain. 2010;133:2045–2057. doi: 10.1093/brain/awq123.
- 40.Lee S.E., Rabinovici G.D., Mayo M.C., Wilson S.M., Seeley W.W., DeArmond S.J., Huang E., Trojanowski J.Q., Ba M.E.G., Jang J.Y., et al. Clinicopathological correlations in corticobasal degeneration. Ann. Neurol. 2011;70:327–340. doi: 10.1002/ana.22424.
- 41.Constantinides V.C., Paraskevas G.P., Emmanouilidou E., Petropoulou O., Bougea A., Vekrellis K., Evdokimidis I., Stamboulis E., Kapaki E. CSF biomarkers beta-amyloid, tau proteins and a-synuclein in the differential diagnosis of Parkinson-plus syndromes. J. Neurol. Sci. 2017;382:91–95. doi: 10.1016/j.jns.2017.09.039.
- 42.Lleó A., Irwin D.J., Illán-Gala I., McMillan C.T., Wolk D.A., Lee E.B., Van Deerlin V.M., Shaw L.M., Trojanowski J.Q., Grossman M. A 2-Step Cerebrospinal Algorithm for the Selection of Frontotemporal Lobar Degeneration Subtypes. JAMA Neurol. 2018;75:738–745. doi: 10.1001/jamaneurol.2018.0118.
- 43.Irwin D.J., McMillan C.T., Toledo J.B., Arnold S.E., Shaw L.M., Wang L.-S., Van Deerlin V., Lee V.M.-Y., Trojanowski J.Q., Grossman M. Comparison of cerebrospinal fluid levels of tau and Abeta 1–42 in Alzheimer disease and frontotemporal degeneration using 2 analytical platforms. Arch. Neurol. 2012;69:1018–1025. doi: 10.1001/archneurol.2012.26.
- 44.Bian H., Van Swieten J.C., Leight S., Massimo L., Wood E., Forman M., Moore P., de Koning I., Clark C.M., Rosso S., et al. CSF biomarkers in frontotemporal lobar degeneration with known pathology. Neurology. 2008;70:1827–1835. doi: 10.1212/01.wnl.0000311445.21321.fc.
- 45.Kapaki E., Paraskevas G.P., Papageorgiou S.G., Bonakis A., Kalfakis N., Zalonis I., Vassilopoulos D. Diagnostic value of CSF biomarker profile in frontotemporal lobar degeneration. Alzheimer Dis. Assoc. Disord. 2008;22:47–53. doi: 10.1097/WAD.0b013e3181610fea.
- 46.Rivero-Santana A., Ferreira D., Perestelo-Pérez L., Westman E., Wahlund L.-O., Sarría A., Serrano-Aguilar P. Cerebrospinal Fluid Biomarkers for the Differential Diagnosis between Alzheimer’s Disease and Frontotemporal Lobar Degeneration: Systematic Review, HSROC Analysis, and Confounding Factors. J. Alzheimer’s Dis. 2016;55:625–644. doi: 10.3233/JAD-160366.
- 47.Casoli T., Paolini S., Fabbietti P., Fattoretti P., Paciaroni L., Fabi K., Gobbi B., Galeazzi R., Rossi R., Lattanzio F., et al. Cerebrospinal fluid biomarkers and cognitive status in differential diagnosis of frontotemporal dementia and Alzheimer’s disease. J. Int. Med. Res. 2019;47:4968–4980. doi: 10.1177/0300060519860951.
Associated Data
Data Availability Statement
The data presented in this study are available on request from the corresponding author.