Cortical Amyloid Burden Relates to Basal Forebrain Volume in Subjective Cognitive Decline

Abstract

Background:

Atrophy of cholinergic basal forebrain (BF) nuclei is a frequent finding in magnetic resonance imaging (MRI) volumetry studies that examined patients with prodromal or clinical Alzheimer’s disease (AD), but less clear for individuals in earlier stages of the clinical AD continuum.

Objective:

To examine BF volume reductions in subjective cognitive decline (SCD) participants with AD pathologic changes.

Methods:

The present study compared MRI-based BF volume measurements in age- and sex-matched samples of N = 24 amyloid-positive and N = 24 amyloid-negative SCD individuals, based on binary visual ratings of Florbetaben positron emission tomography (PET) measurements. Additionally, we assessed associations of BF volume with cortical amyloid burden, based on semiquantitative Centiloid (CL) analyses.

Results:

Group differences approached significance for BF total volume (p = 0.061) and the Ch4 subregion (p = 0.059) only, showing the expected relative volume reductions for the amyloid-positive subgroup. There were also significant inverse correlations between BF volumes and CL values, which again were most robust for BF total volume and the Ch4 subregion.

Conclusions:

The results are consistent with the hypothesis that amyloid-positive SCD individuals, which are considered to represent a transitional stage on the clinical AD continuum, already show incipient alterations of BF integrity. The negative association with a continuous measure of cortical amyloid burden also suggests that this may reflect an incremental process. Yet, further research is needed to evaluate whether BF changes already emerge at “grey zone” levels of amyloid accumulation, before amyloidosis is reliably detected by PET visual readings.

Keywords

Acetylcholine Alzheimer’s disease amyloid atrophy basal forebrain cognitive decline

INTRODUCTION

Alzheimer’s disease (AD), the most frequent cause of dementia, affects more than 45 million people worldwide [1]. Beyond extracellular deposition of amyloid-β peptide (Aβ) [2] and intracellular accumulation of cytoskeleton-associated tau protein [3], AD pathogenesis is linked with neurodegenerative changes, including early neurofibrillary degeneration of cholinergic basal forebrain (BF) neurons (chBFN) which eventually results in widespread cholinergic denervation of the cortical mantle [4, 5]. In humans, magnetic resonance imaging (MRI) based measurements of BF atrophy are used as in vivo surrogate markers for cholinergic degeneration [6], indicating that BF gray matter (GM) volume declines in mild cognitive impairment (MCI) and dementia stages of AD, especially in the posterior aspects of the Nucleus basalis of Meynert (NBM, Ch4) [7 –12].

BF volume reductions may already emerge during the preclinical AD stage (i.e., preceding MCI). There is previous evidence showing negative correlations between cortical amyloid load (as measured by positron emission tomography [PET]) and BF (especially NBM) volume among amyloid-positive cognitively normal individuals [13, 14] that could support previous animal research suggesting neurotoxic effects of excessive Aβ formation [15]. Since cortical Aβ pathology develops over several years (or even decades) before clinical symptoms appear [16], Aβ burden may need to reach a certain threshold (or duration) before incipient atrophic BF changes become detectable. Subjective cognitive decline (SCD) individuals who report a self-experienced persistent decline in cognitive capacity despite normal performance on standardized cognitive tests [17] may be of special interest here. They represent a clinical at-risk population for later MCI and AD dementia [18], especially in combination with AD-specific biomarkers like cerebrospinal fluid (CSF) and PET Aβ positivity or presence of Apolipoprotein E (APOE) ɛ4 alleles (‘SCDplus’: [19]): AD-biomarker positive SCD corresponds to stage 2 of the clinical AD continuum in the novel NIA-AA (National Institute on Aging and Alzheimer’s Association) working group framework [20], and may present the earliest clinical manifestations of AD [21].

Previous studies [22, 23] reported significant BF volume reductions in SCD individuals relative to controls but did not have AD-specific biomarker information available which confirmed that these effects were driven by SCD participants with actual AD pathology. While an earlier analysis in CSF-stratified samples [24] indicated posterior Ch4 volume reduction in Aβ-positive SCD (SCD_Aβ+), as compared to Aβ-negative SCD (SCD_Aβ–), and both Aβ-positive and Aβ-negative normal controls, subsequent analyses in an extended sample, but with different methodology, could not find complementary posterior BF volume reductions in SCD_Aβ+, relative to Aβ-negative normal controls [25]. Only amyloid PET allows the direct measurement of cortical Aβ burden [26]: In fact, a substantial proportion of individuals may already show pathological CSF amyloid levels, but not yet classify as amyloid-positive in amyloid PET, according to conventional visual reading or semiquantitative thresholds derived from clinical populations [27]. Meanwhile, studies examining BF volume of SCD patients based on established cortical amyloid burden in PET imaging are rare [28].

On the other hand, there is growing interest in “grey zone” levels of PET cortical amyloid burden [29], i.e., semiquantitative PET measurements that fall between noise level and conventional thresholds for established amyloid positivity but may already trigger biologically meaningful downstream effects [27, 30]. Whether such an incipient effect of lower-level cortical amyloid burden is also relevant for BF volume changes in preclinical stages of AD, which seems possible as they are thought to start earlier than in other AD-critical brain regions [10], remains to be determined.

We aimed to extend the current knowledge of neurodegenerative effects in the cholinergic BF system associated with cerebral amyloid deposition in preclinical AD stages. We hypothesized that SCD individuals with positive amyloid PET scans (SCD_Aβ+) according to visual qualitative ratings, who represent stage 2 in the NIA-AA AD continuum, show reduced BF nuclei volumes, as compared to SCD with negative amyloid PET (SCD_Aβ–) who served as a clinical reference sample (“stage 0” in the NIA-AA AD framework). Based on the literature, we hypothesized to primarily observe volume reductions in the Ch4 subregion. We furthermore expected the degree of cortical amyloid load in semiquantitative assessments to be inversely related with BF nuclei volumes in SCD, thus, linking incremental accumulation of cortical amyloid pathology with incipient changes in cholinergic BF system integrity.

METHODS

Participants

Forty-eight SCD participants were included, matched for amyloid status, age, and sex (see also: [31]). This sample was obtained from the PET sub-cohort of the German Center for Neurodegenerative Diseases Longitudinal Cognitive Impairment and Dementia Study (DELCODE), an ongoing observational longitudinal memory clinic-based multicenter study focusing on SCD in the context of AD [32], and included SCD participants from 7 of the 10 participating DELCODE sites. These SCD participants were recruited via referrals/self-referrals by the university-based memory clinics of these sites, focusing on individuals seeking medical help because of their SCD [33]. They were included into DELCODE if reporting subjective experience of decline in cognitive functioning that caused concern (as expressed to the physician of the recruiting memory center), having an onset within the last six months to 5 years, while still performing above –1.5 standard deviations of the age-, sex-, and education-adjusted mean on all subtests of the Consortium to Establish a Registry for Alzheimer’s Disease battery [34] performed in clinical routine (i.e., before entry into DELCODE). Further mandatory inclusion criteria were ≥60 years of age, fluent German language skills, capacity to provide informed consent, and presence of an informant (e.g., close relative). Main exclusion criteria for DELCODE were past or present neurological disorders; significant medical diseases; any other detectable cause of memory impairment; and current and lifetime psychiatric disorders. Additionally, PET exams only had regulatory approval for the SCD participants from the DELCODE main study and were not allowed for individuals who had radiation exposure for therapeutic or research purposes within the last 10 years.

The selection of the present subsample was based on qualitative visual readings of [¹⁸F]Florbetaben (FBB, Neuraceq©: Life Molecular Imaging, Berlin) PET scans (see below). From the ongoing data collection, the first available 24 SCD_Aβ+ individuals with significant cerebral amyloid depositions were selected for analysis, and compared with 24 case-wise selected, age- and sex-matched SCD_Aβ–participants showing no cerebral amyloid depositions. Note that a cognitively normal amyloid PET - negative reference group was not available for regulatory reasons (see also: Discussion).

DELCODE and its PET substudy were conducted in accordance with the currently effective version of the Declaration of Helsinki. The PET study protocol was approved by the institutional review boards of all participating sites, coordinated by the ethics committee of the medical faculty of the University of Bonn (local registration number: 221/13). The PET substudy was additionally approved by the federal radiation protection authority (Bundesamt für Strahlenschutz). All procedures were performed in accordance with the relevant guidelines and regulations. All participants provided written informed consent.

Background characteristics

General sample characterization is detailed elsewhere [32]. Briefly, DELCODE study participants complete regular neuropsychological testing during yearly study visits, which includes the Mini-Mental State Examination (MMSE) [35], as well as a modified Alzheimer’s Disease Assessment Scale-Cognitive-Plus (ADAS-Cog-13) [36] battery (which incorporates the learning and delayed recall of a 10-item word list), an oral version of the Symbol Digit Modalities Test (SDMT) [37] and the Trail Making Test, Parts A and B [38]. Apolipoprotein (APOE) genotyping was performed with a commercially available TaqMan® SNP Genotyping Assay (ThermoFisher Scientific).

[¹⁸F]Florbetaben (FBB) PET acquisition and analysis

PET data were acquired on PET/CT (or PET/MRI) scanners at the participating sites. Dynamic listmode data acquisition was started approximately 90 min (M = 90.9±3.6 min) after intravenous FBB tracer application of max. 300 MBq (M = 282±9 MBq), for a total acquisition time of 20 min. Additionally, low-dose CT (PET/MRI: 3D Dixon-VIBE sequences) were collected to calculate attenuation correction maps. The obtained emission data were subsequently reconstructed into 4×5 min time frames. Iterative reconstruction included decay, random, scatter, dead time, normalization, and attenuation correction according to the established PET brain protocols at the local sites (Supplementary Table 1).

FBB PET analyses

Visual reads of individual FBB PET scans were conducted by two experienced readers (HB, FCG), resulting in consensus ratings of either amyloid positivity (SCD_Aβ+) or amyloid negativity (SCD_Aβ–). Following the established manufacturer guidelines [39], they evaluated whether significant cortical tracer binding (relative to neighboring white matter [WM]) was visible across the majority of slices in any of four predefined regions of interest (frontal, lateral temporal, parietal, and posterior cingulate/precuneus), with individuals being rated amyloid-positive (SCD_Aβ+) if significant cortical tracer binding was observed in at least one of the critical regions.

To examine potential linear relationships with the severity of global cortical amyloid load, we conducted secondary semiquantitative Centiloid (CL) analyses [40] which were implemented in the PNEURO Maximum Probability Atlas pipeline of PMOD 4.2 (PMOD Technologies LLC, Zurich, Switzerland). Before CL analysis, the original PET time series were preprocessed to derive image data with uniform geometry and ∼8 mm full-width-half-maximum image resolution (similar to: [41]). Technical details (including CL calibration benchmarks) can be found in the Supplementary Material.

To explore convergence with visual reading classifications, CL levels were also classified post hoc as negative (CL ≤ 12), intermediate (12 <CL ≤ 30) and established amyloid pathology (CL >30), following the suggestions by Salvadó et al. [29].

MRI acquisition and analysis

MRI data acquisition was performed on 3T Siemens scanners at the participating sites, using harmonized imaging protocols across scanners. For the present analysis, T1-weighted magnetization-prepared rapid gradient echo sequences were analyzed (TR: 2500 ms, TE: 4.37 ms, flip angle: 7°, IT: 1100 ms, GRAPPA = 2, 256*256 matrix, field of view: 256*256 mm², slice thickness: 1 mm, 192 sagittal sections, no gap).

MRI data processing and BF volume calculation

The MRI data were processed using the Computational Anatomy Toolbox (CAT12, r1742: https://neuro-jena.github.io/cat [42]) in MATLAB R2019b (Natick, MA: MathWorks Inc.). After segmenting the T1-weighted data into GM, WM, and CSF, the individual GM partitions were normalized to IXI555-MNI152-template space applying modulation for both linear and non-linear normalization. Non-linear normalization was performed using the Diffeomorphic Anatomical Registration Using Exponentiated Lie Algebra algorithm [43], as suggested in Grothe et al. [6]. No additional smoothing of the GM maps was performed.

Based on these modulated GM maps in MNI standard space, individual BF volumes were obtained using a cytoarchitectonic BF atlas, by summing up the modulated GM voxel values intersecting with corresponding atlas masks for specific BF subregions. This atlas was derived from combined post-mortem MRI and histological data of an autopsied brain [9]. It follows the Mesulam et al. [44] chBFN nomenclature. For this study we analyzed the medial septal nucleus (Ch1), the vertical nucleus of the diagonal band of Broca (Ch2), its horizontal limb (Ch3), and NBM (Ch4). Due to the small sizes, Ch1 and Ch2 were combined into one volume (Ch1/2). For all BF structures the volumes of both hemispheres were combined. The total BF volume (BF_total) was calculated by summing up the volume of all BF subregions, while also including a small compartment of interstitial voxels that is represented in the atlas. Figure 1 presents a schematic outline of the masks (superimposed on the IXI555-MNI152-template, displayed in MRIcroGL [45]: https://www.nitrc.org/projects/mricrogl/).

Fig. 1

Coronar display of the cholinergic basal forebrain (Ch) nuclei subregions analyzed in the present study. Due to their small size, the medial septal nucleus and vertical nucleus of the diagonal band of Broca were combined in a single volume (Ch1/2, in red). Moreover, the horizontal limb of the diagonal band of Broca (Ch3, in blue), and the nucleus basalis Meynert (Ch4, in green) were analyzed. Since the interstitial areas (Ch_inter, in yellow) only covered a small number of scattered voxels, they were only included in the calculation of total basal forebrain volume. The regional masks are superimposed on the IXI555-MNI152-template using MRIcroGL, with the numbers representing the y-coordinates of the respective slices. The relative positioning of the field of view is represented by the magenta-colored rectangle in the sagittal midline view (lower right corner).

To control for potential non-specific regional effects of spatial normalization, volume estimates of adjacent brain regions were derived with the same method, using a caudate mask from the Hammers Atlas [46], and a subgenual anterior cingulate mask (Brodman area 25) from the Julich-Brain Cytoarchitectonic Atlas [47], as implemented in CAT12. Moreover, to examine potential atrophy in other brain regions which show early changes in AD, bilateral entorhinal [10] and hippocampal volumes [13] were calculated, again using masks generated from the Julich-Brain Atlas.

Finally, total intracranial volume (TIV) was calculated to account for head size differences in statistical analysis [48], by summing up the volumes of the GM, WM, and CSF partitions, as provided by CAT12.

Statistical analysis

All statistical models and tests were conducted using SPSS 23 (IBM Corp., Armonk, NY).

For demographic variables, group differences were assessed using two-sample t-tests, Mann-Whitney U-tests or chi-square tests (p < 0.05), as necessary. Analyses of group differences of volume measures were performed using an ANCOVA with TIV as a primary covariate. Additional models also included APOE ɛ4 carrier status. The limited sample size did not allow an explicit modelling of the single MR and PET scanners as additional nuisance variables.

Since the original sample construction was based on amyloid positivity in visual-qualitative PET ratings, this binary classification into SCD_Aβ+ and SCD_Aβ–was the primary independent variable for the presented group analyses. Given that this was not completely consistent with CL-based classification (see below), we tested the robustness of our observations by conducting supplementary analyses (A) for the reduced sample of consistent cases, and (B) alternative analyses of variance which were solely based on the abovementioned three-level CL classification [29].

To explore linear relationships with global cortical Aβ load, partial correlations between CL values and the volume measures were calculated (based on residuals, after regression with TIV and age). Again, APOE ɛ4 status was added in extended models.

Finally, to check the potential functional relevance of observed differences, we explored whether the CL values and volume measures, respectively, showed significant relationships with cognitive functioning.

The omnibus significance level was set to p < 0.05.

RESULTS

Group comparisons for background characteristics

Both SCD groups did not differ statistically with respect to age, gender distribution, or cognitive performance level (Table 1). There was a significantly larger proportion of APOE ɛ4 carriers in the SCD_Aβ+. Since participants were grouped according to cortical amyloid deposition as belonging to SCD_Aβ+ or SCD_Aβ–, the mean CL values for global cortical amyloid burden were significantly larger for the SCD_Aβ+, as compared to the SCD_Aβ–group. Notably, a smaller subset of 9 participants (primarily from the SCD_Aβ–group) fell into the “intermediate” category according to the CL-based classification. Visual and CL-based classifications were generally consistent (Spearman correlation r_s = 0.835, p < 0.001), except for three discordant cases (Table 1).

Table 1

Demographic and amyloid characteristics

	SCD	SCD	Test Statistics	p
	(Amyloid+)	(Amyloid –)	(df)
Age (y)	74.5±4.4	74.0±4.1	t(46) = –0.406	0.687
Sex (f/m)	9/15	9/15	χ²(1) = 0.0	1.000
Years of education	14.4±2.8	15.0±3.4	t(46) = 0.609	0.546
MMSE	29.1±0.9 ^*	29.2±1.0 ^*	U(46) = 247.5	0.687
ADAScog-13 - Total	8.4±3.7	7.3±3.3	t(46) = –1.12	0.269
ADAScog Word List Learning: Items Trial 1–3	20.3±4.2	21.6±3.5	t(46) = 1.127	0.266
ADAScog Word List Delayed recall: Items	6.6±1.9	7.4±1.7	t(46) = 1.442	0.156
SDMT: Correct Responses	41.5±6	44.9±10.3	t(36.9) = 1.405	0.167
TMT-A: Seconds	48.8±18.5	45.0±14.1	t(46) = 0.817	0.418
TMT-B: Seconds	96.8±28.5	100.1±30.6 *	t(45) = 0.382	0.704
APOE (ɛ4/non-4)	17/6 ^*	6/17 ^*	χ²(1) = 10.5	0.001
Total intracranial volume	1454±112 ml	1414±155 ml	t(46) = –1.01	0.318
Centiloid: Values	80.0±31.8	7.9±14.4	t(32.02) = –10.1	<0.001
Centiloid: Classified
CL ≤ 12: amyloid-negative	1	15
12 < CL ≤ 30: intermediate	1	8
CL >30: established amyloid-positive	22	1

^*Denotes 1 missing values. Centiloid classification according to cut-off values suggested by Salvadó et al. [29]. APOE, Apolipoprotein E; CL, Centiloid value; U, Mann-Whitney U statistic; MMSE, Mini-Mental State Examination; ADAScog-13, Alzheimer’s Disease Assessment Scale-cognitive part; SCD, subjective cognitive decline; SDMT, Symbol Digit Modalities Test, oral version; TMT-A/ TMT-B, Trail Making Test Part A and B, respectively. Data shown as means±standard deviations, or number of cases, respectively.

Group comparisons for volumetric data

The descriptive data for the volumetric analyses are presented in Table 2. The ANCOVA for BF total volume showed that the effect of the factor ‘group’ approached significance (F_1,45 = 3.705; p = 0.061; partial η² = 0.08) and was driven by lower volumes for the SCD_Aβ+ group. The same trend was found for Ch4 subvolume (F_1,45 = 3.768; p = 0.059; partial η² = 0.08). Adding APOE status as an additional covariate (available for 46 participants only) reduced the observed trends slightly, both for BF total volume (F_1,42 = 2.952; p = 0.093; partial η² = 0.07) and Ch4 volume (F_1,42 = 2.947; p = 0.093; partial η² = 0.07). Meanwhile, no comparable trends were found for Ch1/2 (F_1,45 = 2.21; p = 0.144; partial η² = 0.047) and Ch3 (F_1,45 = 2.019; p = 0.162; partial η² = 0.043). Similarly, no effects were also found for the caudate (F_1,45 = 0.032; p = 0.859; partial η² = 0.001) and subgenual anterior cingulate (F_1,45 = 1.254; p = 0.269; partial η² = 0.03) control regions. Finally, there were no significant group effects for entorhinal (F_1,45 = 0.085; p = 0.772; partial η² = 0.002) and hippocampal volume (F_1,45 = 1.867; p = 0.179; partial η² = 0.04).

Table 2

Regional volumes (raw and adjusted for total intracranial volume for amyloid-positive and amyloid-negative participants

	SCD_Aβ+				SCD_Aβ–
	Raw		Adjusted		Raw		Adjusted
ROI	Mean	SEmean	Mean	SEmean	Mean	SEmean	Mean	SEmean	p
BF total	600.7	10.7	597.1	8.6	616.9	8.8	620.5	8.6	0.061
Ch1/2	88.57	1.9	87.5	1.3	89.31	2.0	90.4	1.34	0.144
Ch3	148.45	3.0	147.6	2.4	151.48	2.2	152.3	2.35	0.162
Ch4	356.42	6.4	354.8	5.6	368.6	5.5	370.2	5.61	0.059
Entorhinal	1307.3	33.8	1300	33.8	1306.6	35.8	1314	33.8	0.772
Hippocampus	8620.6	183.1	8564.1	161.1	8820.6	171.0	8877.1	161.1	0.179
Caudate	6184.5	134.9	6128.7	124.8	6104.4	155.4	6160.3	124.8	0.859
Subgenual ACC	498.5	9.6	491	8.3	496.7	16.1	504.2	8.3	0.269

Listed are the raw and adjusted values (estimated marginal means after correcting for total intracranial volume) in mm³. ACC, anterior cingulate cortex; BF, cholinergic basal forebrain; Ch1/2, medial septal nucleus + vertical nucleus of the diagonal band of Broca; Ch3, horizontal limb of the diagonal band of Broca; Ch4, nucleus basalis of Meynert; SCD_Aβ+/SCD_Aβ–, amyloid-positive / amyloid-negative subjective cognitive decline participants; SEmean, standard error of the mean.

Excluding the three individuals with discordant visual-qualitative and semiquantitative classifications actually improved the observed effects slightly: The ‘group’ effect for BF total volume became significant (F_1,42 = 5.1; p = 0.030; partial η² = 0.11), and still approached significance after adding APOE status (F_1,39 = 3.97; p = 0.053; partial η² = 0.09). The same was observed for Ch4 (F_1,42 = 4.6; p = 0.038; partial η² = 0.1), also after adding APOE status (F_1,39 = 3.6; p = 0.066; partial η² = 0.08). An additional group effect for Ch3 emerged (F_1,42 = 4.2; p = 0.047; partial η² = 0.09) but did not survive after adding APOE status into the model (F_1,39 = 1.6; p = 0.21; partial η² = 0.04). None of the other volume measures showed significant (or trend-level, p < 0.1) group differences (Supplementary Table 2).

Group comparisons based on the three-group CL-based classification provided similar findings. The newly assembled groups did not differ significantly in their background characteristics (except for CL and APOE, as in the main analysis: Supplementary Table 3). For the volumetric data (Supplementary Table 4), group had a significant effect on BF total volume (F_2,44 = 3.4; p = 0.042; partial η² = 0.134), showing the expected linear trend, although only the post hoc contrast between the CL < 12 and CL > 30 subgroups was significant (p = 0.043, Sidak-corrected). Adding APOE status changed this only slightly (F_2,41 = 3.3; p = 0.046; partial η² = 0.139), though attenuating the post hoc contrast (p = 0.061). A similar group effect was found for Ch4, but only on trend-level (F_2,44 = 2.9; p = 0.068; partial η² = 0.12), also after adding APOE status (F_2,41 = 2.5; p = 0.092; partial η² = 0.11). In contrast, a group effect for Ch3 emerged (F_2,44 = 3.5; p = 0.040; partial η² = 0.136), again with a significant contrast between the CL < 12 and the CL > 30 subgroup only (p = 0.038, Sidak-corrected). These effects were slightly attenuated after adding APOE status (F_2,41 = 2.9; p = 0.068; partial η² = 0.11). For the Ch1/2 region, as well as the other brain regions, no significant (or even trend-level, p < 0.1) group effects were observed.

Linear relationships between BF volumes and cortical amyloid burden

Moderate-size, inverse partial correlations between CL values and BF_total and Ch4 volume were observed across the whole sample (Fig. 2) which were also significant when additionally partializing out APOE status (Table 3). Moreover, a significant inverse correlation for Ch3, and trend-level correlation with Ch1/2 volume emerged, but both were reduced to trend level after additionally controlling for APOE status (Table 3). Meanwhile, no-to-weak partial correlations with CL values were observed for the entorhinal and hippocampal AD reference regions (Table 3). Finally, only non-significant weak correlations were found for the caudate (r = –0.103, p = 0.488) and subgenual anterior cingulate (r = –0.135, p = 0.362) control regions.

Fig. 2

Scatterplots of the significant negative partial correlations between global cortical amyloid burden (expressed in Centiloid) and basal forebrain volumes. Shown are the residuals after removing variance related to TIV and age, including linear regression curve and 95% confidence interval. A) Basal forebrain –total volume. B) Ch4 volume. Black/filled data points indicate amyloid-positive (SCD_Aβ+) patients, white/open data points mark amyloid-negative (SCD_Aβ–) patients. Additionally, data points are split according to APOE ɛ4 status for further inspection, with triangles representing carriers, and circles representing non-carriers. a.u., arbitrary units; BF, basal forebrain; Ch4, nucleus basalis of Meynert.

Table 3

Associations between regional volumes and cortical amyloid burden

		Correlation: Centiloid * regional volumes
		BF total	Ch1/2	Ch3	Ch4	Entorhinal	Hippocampus
	Residualized for	r (p)	r (p)	r (p)	r (p)	r (p)	r (p)
Whole sample	TIV + Age	–0.369 (0.010)	–0.258 (0.077)	–0.303 (0.037)	–0.371 (0.009)	–0.071 (0.632)	–0.190 (0.197)
	TIV + Age + A PO E	–0.345 (0.019)	–0.290 (0.051)	–0.242 (0.106)	–0.338 (0.022)	–0.058 (0.700)	–0.227 (0.129)
SCD_Aβ+	TIV + Age	–0.274 (0.195)	–0.161 (0.464)	–0.188 (0.379)	–0.307 (0.144)	–0.040 (0.855)	–0.017 (0.935)
	TIV + Age + AP O E	–0.373 (0.080)	–0.131 (0.550)	–0.155 (0.480)	–0.430 (0.041)	–0.150 (0.494)	–0.176 (0.422)
SCD_Aβ–	TIV + Age	–0.267 (0.207)	–0.064 (0.766)	–0.460 (0.024)	–0.176 (0.410)	–0.054 (0.807)	–0.034 (0.873)
	TIV + Age + AP O E	–0.185 (0.399)	–0.303 (0.160)	–0.244 (0.263)	–0.089 (0.687)	–0.060 (0.786)	0.002 (0.994)

Correlations between regional volume and Centiloid (residuals after controlling for TIV + age or TIV + age + APOE status, respectively), for the whole sample and subgroups. APOE, Apolipoprotein E; BF, cholinergic basal forebrain; Ch1/2, medial septal nucleus + vertical nucleus of the diagonal band of Broca; Ch3, horizontal limb of the diagonal band of Broca; Ch4, nucleus basalis of Meynert; SCD_Aβ+/SCD_Aβ–, Amyloid-positive / amyloid-negative subjective cognitive decline participants; TIV, total intracranial volume.

Looking at each two subgroups separately, the general result patterns were broadly consistent with the whole group analysis, showing moderate-strength correlations for BF total and Ch4 preferentially in the amyloid-positive subgroup, while correlations for Ch1/2 and Ch3 were less stable, and no-to-weak effects were found in the entorhinal and hippocampal AD reference regions (Table 3). Yet, most of these correlation coefficients did not reach statistical significance in the substantially reduced samples.

While not subject to the present research question, note the additional observation that BF total and Ch4, and also Ch3 showed positive partial correlations with entorhinal and hippocampal volume, respectively, which were mainly driven by the amyloid-positive subgroup, while the complementary coefficients in the amyloid-negative subgroup did not reach significance (Supplementary Table 5).

Linear relationships between BF volumes, amyloid burden, and cognitive functioning

Both the CL measure for global cortical amyloid burden and the BF, entorhinal and hippocampal volumes mostly showed non-significant, none-to-weak correlations with the cognitive test measures, with the exception of the SDMT correct responses, which showed moderate positive correlations with BF volumes (significant for BF total and Ch4, trend for Ch1/2 and Ch3: Supplementary Table 5). Yet, follow-up correlation analyses in the subgroups indicated that this was primarily the case for the amyloid-negative subgroup, not for the amyloid-positive subgroup.

DISCUSSION

The present study aimed to examine whether SCD individuals with significant cortical amyloid burden show detectable BF volume reductions. This would be predicted for individuals in NIA-AA stage 2 of the clinical AD continuum [20], the transitional phase between asymptomatic AD pathologic changes and the MCI stage of AD, where BF atrophy is already established, especially for its Ch4 subregion [9]. Group comparisons of BF volume between amyloid-positive and amyloid-negative SCD participants, as defined by binary visual FBB PET reading, i.e., the clinical gold standard for amyloid PET interpretation, were generally consistent with this prediction, although the observed BF volume reductions were modest, and approached statistical significance only for the BF total and Ch4 volume. Complementing the binary classification based on visual reading, especially BF total and Ch4 volume showed inverse linear relationships with continuous CL scores of global cortical amyloid burden, i.e., the higher the cerebral amyloid deposition, the lower the volume of these cholinergic basal forebrain nuclei within this SCD cohort. This correlation was still present when accounting for the APOE status. The latter observations are consistent with the idea that incipient cortical amyloid accumulation exerts a gradual effect on BF integrity during the late preclinical phase of AD, although the directionality of the observed associations cannot be clarified in the present study design.

The current data extend observations made in independent cohorts of SCD subjects [22, 23] by additionally stratifying the SCD participants according to biomarker status (FBB PET-based cortical Aβ), and generally support early alterations in the BF system in SCD individuals who fall into stage 2 of the clinical AD continuum [20]. They may predate detectable structural changes in other AD-sensitive areas, given that there were no comparable trends for entorhinal and hippocampal reference regions. An early involvement of BF degeneration in AD pathogenesis is consistent with earlier longitudinal data which suggest that accelerated NBM atrophy over two years was already detectable in Aβ-positive healthy elderly (based on CSF analyses), and seemed to precede entorhinal degeneration [10]. In this context, it is interesting that positive correlations between BF and entorhinal and hippocampal volumes (i.e., individuals with smaller BF volumes also showed smaller medial temporal volumes) were primarily observed for our amyloid-positive SCD subgroup. This may indicate that incipient changes also emerge in these AD-related regions but were not closely linked with amyloid status. While similar to a previous study, which observed homologous correlations in SCD, but not in normal control participants [23], this finding needs further corroboration, given that previous morphometric studies did not report complementary structural covariations in SCD participants [49].

At first sight, these and the current findings seem to differ from previous studies in cognitively normal populations that observed no group differences in BF volumes between Aβ-positive and Aβ-negative participants [13 , 50], or a study that even reported increased BF volumes (based on manual delineation) in cognitively normal elderly who subsequently developed clinical AD symptoms, yet did not provide biomarker information [51]. BF volume reductions may not appear with detectable amyloidosis per se (e.g., in CSF or plasma: [25, 50]), but depend on a certain threshold or duration of cortical amyloid accumulation, as directly measured with PET. The fact that our SCD_Aβ+ participants did not only show significant cerebral amyloid burden, but also had the additional clinical experience of subjective cognitive decline, may reflect a more advanced stage of AD pathogenesis than asymptomatic Aβ+ controls. Thus, while previous studies focused on the asymptomatic stage 1 of the clinical AD continuum (according to the novel NIA-AA framework: [20]), our SCD_Aβ+ participants would be considered to have progressed to stage 2.

Considering that the statistical group comparisons based on visual amyloid PET readings only approached statistical significance, it is important that an alternative semiquantitative analysis (generating continuous CL measurements of cortical amyloid burden) provided complementary, and actually even more significant results than the original binary classification, for BF total and Ch4 volume. In general, the finding confirms previous observations of negative linear associations between quantitative Aβ burden and BF measures, both across the whole clinical AD continuum [52], and in cognitively intact populations [13, 14]. While similar trends were also observed within the single subgroups (especially among the amyloid-positive participants), it must be noted that they were not statistically significant and need to be substantiated in better-powered samples.

It is clear that cortical amyloid accumulation is an incremental process, and recent studies have shown increasing interest in “subthreshold” or “grey zone” levels of amyloid burden that do not yet fulfill clinically validated cut-off points of amyloid positivity but may already show incipient effects on cognitive and brain function [27 , 30]. Adopting a tripartite scheme where 12<CL ≤ 30 denotes such intermediate levels of cortical amyloid burden [29], mainly amyloid-negative SCD participants (according to visual rating) fell into this range. Yet, the number of these intermediate cases was too small to test systematically whether subthreshold CL levels of global cortical amyloid burden are sufficient to be linked with measurable BF volume reductions: Exploratory analyses generally confirmed group differences for BF total, Ch4 and additionally Ch3 volume, but post hoc contrasts indicated that only the difference between the larger amyloid-negative and established amyloid-positive subgroups reached significance, whereas the small group with intermediate CL values was nominally in-between, but not significantly different from any of the other groups.

The inverse linear relationship between CL amyloid burden and BF volumes suggests a link between two core features of AD already in the preclinical stage. Thus, one may infer that the level of amyloid accumulation is directly or indirectly related to cholinergic dysfunction. In general, it must be acknowledged that our cross-sectional analyses do not allow to deduce causal relations: If a direct causal link exists, animal studies present evidence for both directions, such that the amount of cerebral (cortical) amyloid deposition affects the (subcortical) cholinergic system in a remote fashion, and vice versa, i.e., the increase of cortical amyloid burden can be a consequence of cholinergic denervation on the cortical level [5 , 54]. More importantly, one has to note that MRI-based volumetry methods like in the present study do not provide direct measures of BF cholinergic function, even though there is first human evidence for direct correlations in clinical populations [55, 56]. In vivo measurements of cholinergic transmission using PET have become available, but mainly focused on prodromal [56 –58] or clinical dementia patients [55], where cholinergic denervation is already advanced. Incipient changes of cholinergic function in preclinical stages of the AD continuum may be more difficult to detect and will probably require substantially larger samples. Meanwhile, there is preliminary evidence that healthy individuals with intact ‘short-latency sensory afferent inhibition’ responses to transcranial magnetic stimulation also show larger BF volumes than non-responders [59]: While this electrophysiological measure is an indirect marker for central cholinergic signal transmission, such methods may provide a less invasive method for further exploring cholinergic dysfunction during the preclinical AD stage.

In general, the functional significance of the imaging findings remains to be determined, since correlational relationships between BF volumes and neuropsychological test performance (which were previously observed in larger samples [60]) could not be replicated. While there was some evidence for a positive association between Symbol Digit Modalities performance and BF volume measures, it remains inconclusive because it was not significant for the amyloid-positive subgroup.

The present analyses utilized amyloid-negative SCD participants as a reference group for the amyloid-positive SCD participants because no PET exams from cognitively normal participants were available. Using SCD individuals without current evidence for significant cortical amyloid pathology as a reference is necessarily suboptimal because we cannot exclude that the SCD symptoms of these memory clinic patients were caused by undetected non-AD pathologies which were not sufficiently screened out by the general study inclusion and exclusion criteria (which only considered clinically established neurological, psychiatric, and medical disorders). In fact, recent studies also show BF volume reductions in other neurodegenerative disorders (e.g., frontotemporal dementia [61], Lewy body dementia [62]). But even if the SCD symptoms of the amyloid-negative participants would reflect the prodromal effects of such pathologies, this would have reduced the chance to find group differences relative to the amyloid-positive subgroup and add unexplained variance in the correlational analyses.

While matched for sex and age, the two SCD samples also differed significantly regarding their relative proportion of APOE ɛ4 carriers, with an increased ratio of positive cases especially in the amyloid-positive subgroup. This is not unexpected given that APOE ɛ4 carriers have a higher risk of developing amyloid positivity [63, 64], but makes it difficult to disentangle direct and indirect (via increased amyloid accumulation) contributions of APOE ɛ4 to the observed BF volume differences. For example, APOE ɛ4-related disruptions of lipid metabolism in chBFN cells may impair the maintenance of their large axonal membranes [65, 66]. A cross-sectional study [67] found that a negative effect of APOE ɛ4 status on BF volume was present, but interacted with disease status, i.e., did not emerge for normal controls and SCD, but mainly in late MCI and AD dementia participants. Schmitz et al. [66] observed accelerated longitudinal decline of posterior BF (NBM) volume already in cognitively normal APOE ɛ4 carriers (but see: [28]), despite similar CSF Aβ_1 - 42 and pTau levels as non-carriers, suggesting detrimental effects that were independent from amyloid load or tau pathology. It should be noted, however, that the inverse correlation between CL amyloid burden and BF total and Ch4 volumes in our study remained significant after additionally partializing out APOE status.

A related limitation is the lack of complementary tau information, which was not systematically available for the examined participants: There is a long-established link between cholinergic BF neuron loss and neurofibrillary changes [68], suggesting that neurotoxic effects may depend on an interplay between Aβ and tau-related processes [15]. A recent study [50] observed that posterior BF (NBM) atrophy in cognitively normal elderly was not associated with abnormal CSF amyloid, but pathological p-tau levels: Yet, their participants with isolated amyloidosis (but no tauopathy) were substantially younger than the respective samples with pathological tau levels, leaving the possibility that these individuals were in an earlier stage of AD pathogenesis. Another study [69] provided evidence for stronger baseline and longitudinal BF degeneration in individuals with pathologically high CSF p-tau/Aβ ratios, both in analyses pooling across diagnostic categories, and in cognitively normal individuals more specifically (see also: [66]). Finally, higher plasma total tau (t-tau), but not amyloid PET positivity was found to predict longitudinal BF volume decline over 24 months in a sample of SCD individuals [28]. Whether amyloid accumulation is already sufficient for triggering BF atrophy, or whether additional tauopathy has to come into play (i.e., the presence of preclinical AD in the strict sense) is an aspect that should be taken into account in future studies examining SCD populations. While most current evidence in SCD populations comes from CSF or plasma measurements, the ability of tau PET to measure the spatial distribution patterns of cortical tau depositions may provide valuable additional information to further differentiate between AD–and non-AD-related tauopathies [70].

There are some general limitations that need to be acknowledged, including the restricted size of the age- and sex-matched samples and, hence, statistical power (which hindered in-depth statistical control of technical nuisance factors, e.g., potential scanner differences), the relatively high educational level in both groups, and also the observation that the amyloid classifications based on visual readings (i.e., the clinical gold standard) are not completely consistent with the CL-based classifications of cortical amyloid burden in single cases (Table 1). Previous studies indicate that the correspondence between binary FBB visual reading and classification according to global standardized uptake value ratio (SUVR) cut-offs is high (>85%), but not perfect [71]. Apart from classification problems with marginal values close to cut-off values, it should be noted these two approaches prioritize different aspects (visual rating: contrast between cortical and neighboring white matter uptake versus CL and other SUVR measures of cortical amyloid burden: ratio between uptake in cerebral and cerebellar regions of interest), which can produce divergent findings, especially in equivocal cases. In certain cases, visual scoring may actually be more sensitive if amyloid burden is restricted to circumscribed (e.g., unilateral) regions.

Conclusions

The present findings provide further evidence for the idea that SCD individuals with cortical amyloid pathology reflect a transitional stage of the clinical AD continuum, by suggesting amyloid-related volume reductions in BF structures which are known to present atrophic changes in the subsequent MCI stage of AD. The negative association with a continuous measure of cortical amyloid burden also suggests that this could be a gradual process: BF atrophy may already start while cortical amyloid burden is still on a subthreshold level (with regard to established visual and semiquantitative cut-offs for PET amyloid positivity). Yet, additional research with a targeted sampling of SCD individuals within this “grey zone” of amyloid accumulation, and longitudinal follow-up is needed to further explore this idea.

Footnotes

ACKNOWLEDGMENTS

The authors would like to thank the following institutions: Max Delbrück Centrum für Molekulare Medizin in der Helmholtz-Gemeinschaft (MDC), Berlin; Center for Cognitive Neuroscience Berlin (CCNB), Freie Universität Berlin; Bernstein Center for Computational Neuroscience Berlin; Klinik und Poliklinik für Nuklearmedizin, Universitätsklinikum Bonn; Klinik und Poliklinik für Nuklearmedizin, Universitätsklinikum Köln; Klinik und Poliklinik für Nuklearmedizin, Klinikum der Universität München; Institut für Klinische Radiologie, Klinikum der Universität München; Klinik und Poliklinik für Nuklearmedizin, Universitätsmedizin Rostock; MR-Forschungszentrum, Universitätsklinikum Tübingen; Nuklearmedizin und Klinische Molekulare Bildgebung, Universitätsklinikum Tübingen.

Most importantly, we thank the DELCODE participants and their families.

DELCODE STUDY GROUP

Slawek Altenstein, Holger Amthauer, Abdelmajid Bader, Juan Carlos Baldermann, Miriam Barkhoff, Henning Boecker, Frederic Brosseron, Martina Buchmann, Katharina Buerger, Cihan Catak, Arda Can Cetindag, Lisa Coloma Andrews, Nicoleta Carmen Cosma, Marcel Daamen, Martin Dichgans, Dominik Diesing, Britta Dölle, Alexander Drzezga, Martin Dyrba, Marie Ehrlich, Tanja Engels, Birgit Ertl-Wagner, Claus Escher, Jennifer Faber, Friederike Fenski, Klaus Fließbach, Christiana Franke, Silka Dawn Freiesleben, Daniela Frimmer, Ingo Frommann, Manuel Fuentes, Nasim Roshan Ghiasi, Marcus Grobe-Einsler, Katja Hardenacke, Dietmar Hauser, Tanja Heger, Guido Hennes, Gabi Herrmann, Petra Hinderer, Brigitte Huber, Nicole Hujer, Enise Irem Incesoy, Heike Janecek-Meyer, Daniel Janowitz, Frank Jessen, Lorraine Jost, Christian Kainz, Pascal Kalbhen, Ingo Kilimann, Okka Kimmich, Xenia Kobeleva, Barbara Kofler, Max Kreuzer, Elke Kuder-Buletta, Catharina Lange, Chris Lappe, Christoph Laske, Esther Lau, Katja Lindner, Andrea Lohse, Hannah Lützerath, Franziska Maier, Benjamin Marquardt, Anja Martikke, Cornelia McCormick, Herlind Megges, Dix Meiberth, Lisa Miebach, Carolin Miklitz, Anna Müller, Claudia Müller, Matthias Munk, Christian Mychajliw, Demet Oender, Oliver Peters, Snjezana Petzler, Henrike Pfaff, Alexandra Polcher, Lukas Preis, Josef Priller, Veronika Purrer, Heike Raum, Axel Rominger, Sandra Röske, Ayda Rostamzadeh, Petr Sabik, Lena Sannemann, Ann-Katrin Schild, Jennifer Schmid, Monika Schmidt, Christine Schneider, Anja Schneider, Heike Schulz, Sarah Schwarzenböck, Anna Seegerer, Surjo Soekadar, Susanne Sorgalla, Annika Spottke, Eike Jakob Spruth, Julia Stephan, Simone Stockter, Patricia Sulzer, Stefan Teipel, Manuela Thelen, Theresia Trunk, Maike Tscheuschler, Franziska Uhle, Irene Villar Munoz, Ina Vogt, Michael Wagner, Marc-Andre Weber, Steffen Wolfsgruber, Yilmaz Sagik, Philip Zeyen, Adelgunde Zollver.

FUNDING

The study was funded by the German Center for Neurodegenerative Diseases (Deutsches Zentrum für Neurodegenerative Erkrankungen, DZNE: reference number BN012). The reported visual analyses of the Florbetaben PET exams were subject of a research agreement with Life Molecular Imaging.

CONFLICT OF INTEREST

Alexander Drzezga reports research support from Siemens Healthineers, Life Molecular Imaging, GE Healthcare, AVID Radiopharmaceuticals, Sofie, Eisai, and Novartis/AAA; speaker honorary/advisory boards for Siemens Healthineers, Sanofi, GE Healthcare, Biogen, Novo Nordisk, Invicro, Novartis/AAA, Bayer Vital; stock in Siemens Healthineers, Lantheus Holding; and patents pending for 18F-PSMA7 (PSMA PET imaging tracer).

Bernd J. Krause reports travel funding from AAA/Novartis; speaker honorary from AAA/Novartis, Bayer, Janssen; Advisory boards including Terumo, Rotop, AAA/Novartis, PSI CRO, ITM, Bayer, and Janssen; and third party funding from AAA/Novartis.

Michel J. Grothe is supported by the “Miguel Servet” program [CP19/00031] and a research grant [PI20/00613] of the Instituto de Salud Carlos III-Fondo Europeo de Desarrollo Regional (ISCIII-FEDER).

Ralph Buchert, Stefan J. Teipel, and Ayda Rostamzadeh are Editorial Board Members of this journal but were not involved in the peer-review process nor had access to any information regarding its peer-review.

The other authors declare no relevant competing interests.

DATA AVAILABILITY

The datasets generated and/or analyzed during the current study are not publicly available due to restrictions of patient consent regarding data distribution and use, but can be applied from the DELCODE study consortium on reasonable request.

The supplementary material is available in the electronic version of this article: .

References

GBD 2016 Dementia Collaborators (2019) Global, regional, and national burden of Alzheimer’s disease and otherdementias, 1990-2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 18, 88–106.

Jamasbi

, Wade

, Separovic

, Hossain

(2016) Amyloid beta (Abeta) peptide and factors that play important roles in Alzheimer’s disease. Curr Med Chem 23, 884–892.

Goedert

(1993) Tau protein and the neurofibrillary pathology of Alzheimer’s disease. Trends Neurosci 16, 460–465.

Schliebs

, Arendt

(2011) The cholinergic system in aging and neuronal degeneration. Behav Brain Res 221, 555–563.

Hampel

, Mesulam

, Cuello

, Farlow

, Giacobini

, Grossberg

, Khachaturian

, Vergallo

, Cavedo

, Snyder

, Khachaturian

(2018) The cholinergic system in the pathophysiology and treatment of Alzheimer’s disease. Brain 141, 1917–1933.

Grothe

, Kilimann

, Grinberg

, Heinsen

, Teipel

(2018) In vivo volumetry of the cholinergic basal forebrain. In Biomarkers for Preclinical Alzheimer’s Disease, Perneczky R, ed. Springer New York, New York, NY, pp. 213–232.

Teipel

, Meindl

, Grinberg

, Grothe

, Cantero

, Reiser

, Moller

, Heinsen

, Hampel

(2011) The cholinergic system in mild cognitive impairment and Alzheimer’s disease: An in vivo MRI and DTI study. Hum Brain Mapp 32, 1349–1362.

Grothe

, Zaborszky

, Atienza

, Gil-Neciga

, Rodriguez-Romero

, Teipel

, Amunts

, Suarez-Gonzalez

, Cantero

(2010) Reduction of basal forebrain cholinergic system parallels cognitive impairment in patients at high risk of developing Alzheimer’s disease. Cereb Cortex 20, 1685–1695.

Kilimann

, Grothe

, Heinsen

, Alho

, Grinberg

, Amaro

Jr. , Dos Santos

, da Silva

, Mitchell

, Frisoni

, Bokde

, Fellgiebel

, Filippi

, Hampel

, Kloppel

, Teipel

(2014) Subregional basal forebrain atrophy in Alzheimer’s disease: A multicenter study. J Alzheimers Dis 40, 687–700.

10.

Schmitz

, Nathan Spreng

, Alzheimer’s Disease Neuroimaging Initiative (2016) Basal forebrain degeneration precedes and predicts the cortical spread of Alzheimer’s pathology. Nat Commun 7, 13249.

11.

Cantero

, Zaborszky

, Atienza

(2017) Volume loss of the nucleus basalis of Meynert is associated with atrophy of innervated regions in mild cognitive impairment. Cereb Cortex 27, 3881–3889.

12.

Grothe

, Heinsen

, Teipel

(2012) Atrophy of the cholinergic basal forebrain over the adult age range and in early stages of Alzheimer’s disease. Biol Psychiatry 71, 805–813.

13.

Grothe

, Ewers

, Krause

, Heinsen

, Teipel

, Alzheimer’s Disease Neuroimaging Initiative (2014) Basal forebrain atrophy and cortical amyloid deposition in nondemented elderly subjects. Alzheimers Dement 10, S344–353.

14.

Kerbler

, Fripp

, Rowe

, Villemagne

, Salvado

, Rose

, Coulson

, Alzheimer’s Disease Neuroimaging Initiative (2015) Basal forebrain atrophy correlates with amyloid beta burden in Alzheimer’s disease. Neuroimage Clin 7, 105–113.

15.

Pepeu

, Giovannini

(2017) The fate of the brain cholinergic neurons in neurodegenerative diseases. Brain Res 1670, 173–184.

16.

Sperling

, Mormino

, Johnson

(2014) The evolution of preclinical Alzheimer’s disease: Implications for prevention trials. Neuron 84, 608–622.

17.

Molinuevo

, Rabin

, Amariglio

, Buckley

, Dubois

, Ellis

, Ewers

, Hampel

, Kloppel

, Rami

, Reisberg

, Saykin

, Sikkes

, Smart

, Snitz

, Sperling

, van der Flier

, Wagner

, Jessen

, Subjective Cognitive Decline Initiative Working Group (2017) Implementation of subjective cognitive decline criteria in research studies. Alzheimers Dement 13, 296–311.

18.

Pike

, Cavuoto

, Li

, Wright

, Kinsella

(2022) Subjective cognitive decline: Level of risk for future dementia and mild cognitive impairment, a meta-analysis of longitudinal studies. Neuropsychol Rev 32, 703–735.

19.

Jessen

, Amariglio

, van Boxtel

, Breteler

, Ceccaldi

, Chetelat

, Dubois

, Dufouil

, Ellis

, van der Flier

, Glodzik

, van Harten

, de Leon

, McHugh

, Mielke

, Molinuevo

, Mosconi

, Osorio

, Perrotin

, Petersen

, Rabin

, Rami

, Reisberg

, Rentz

, Sachdev

, de la Sayette

, Saykin

, Scheltens

, Shulman

, Slavin

, Sperling

, Stewart

, Uspenskaya

, Vellas

, Visser

, Wagner

, Subjective Cognitive Decline Initiative Working Group (2014) A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 10, 844–852.

20.

Jack

Jr. , Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

, Contributors (2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14, 535–562.

21.

Jessen

, Wolfsgruber

, Kleineindam

, Spottke

, Altenstein

, Bartels

, Berger

, Brosseron

, Daamen

, Dichgans

, Dobisch

, Ewers

, Fenski

, Fliessbach

, Freiesleben

, Glanz

, Görß

, Gürsel

, Janowitz

, Kilimann

, Kobeleva

, Lohse

, Maier

, Metzger

, Munk

, Preis

, Sanzenbacher

, Spruth

, Rauchmann

, Vukovich

, Yakupov

, Weyrauch

A-S

, Ziegler

, Schmid

, Laske

, Perneczky

, Schneider

, Wiltfang

, Teipel

, Bürger

, Priller

, Peters

, Ramirez

, Boecker

, Heneka

, Wagner

, Düzel

(2023) Subjective cognitive decline and stage 2 of Alzheimer disease in patients from memory centers. Alzheimers Dement 19, 487–497.

22.

Scheef

, Grothe

, Koppara

, Daamen

, Boecker

, Biersack

, Schild

, Wagner

, Teipel

, Jessen

(2019) Subregional volume reduction of the cholinergic forebrain in subjective cognitive decline (SCD). Neuroimage Clin 21, 101612.

23.

Chen

, Wu

, Li

, Sun

, Chen

, Lu

, Zhang

, Liu

, Qing

, Nedelska

, Hort

, Zhang

(2021) Basal forebrain atrophy is associated with allocentric navigation deficits in subjective cognitive decline. Front Aging Neurosci 13, 596025.

24.

, Grothe

, Daamen

, Wolfsgruber

, Brosseron

, Buerger

, Catak

, Dobisch

, Incesoy

, Fliessbach

, Franke

, Heneka

, Kilimann

, Kleineidam

, Laske

, Metzger

, Munk

, Peters

, Preis

, Scheef

(2020) Basal forebrain atrophy along the Alzheimer’s disease spectrum and its relevance for subjective cognitive decline [Preprint]. ResearchSquare, rs.3.rs-20438/v20431.

25.

Herdick

, Dyrba

, Fritz

H-CJ

, Altenstein

, Ballarini

, Brosseron

, Buerger

, Can Cetindag

, Dechent

, Dobisch

, Duezel

, Ertl-Wagner

, Fliessbach

, Dawn Freiesleben

, Frommann

, Glanz

, Dylan Haynes

, Heneka

, Janowitz

, Kilimann

, Laske

, Metzger

, Munk

, Peters

, Priller

, Roy

, Scheffler

, Schneider

, Spottke

, Jakob Spruth

, Tscheuschler

, Vukovich

, Wiltfang

, Jessen

, Teipel

, Grothe

(2020) Multimodal MRI analysis of basal forebrain structure and function across the Alzheimer’s disease spectrum. Neuroimage Clin 28, 102495.

26.

Laforce

Jr. , Soucy

, Sellami

, Dallaire-Theroux

, Brunet

, Bergeron

, Miller

, Ossenkoppele

(2018) Molecular imaging in dementia: Past, present, and future. Alzheimers Dement 14, 1522–1552.

27.

Milà-Alomà

, Salvadó

, Shekari

, Grau-Rivera

, Sala-Vila

, Sánchez-Benavides

, Arenaza-Urquijo

, González-de-Echávarri

, Simon

, Kollmorgen

, Zetterberg

, Blennow

, Gispert

, Suárez-Calvet

, Molinuevo

(2021) Comparative analysis of different definitions of amyloid-βpositivity to detect early downstream pathophysiological alterations in preclinical Alzheimer. J PrevAlzheimers Dis 8, 68–77.

28.

Cavedo

, Lista

, Houot

, Vergallo

, Grothe

, Teipel

, Zetterberg

, Blennow

, Habert

M-O

, Potier

, Dubois

, Hampel

(2020) Plasma tau correlates with basal forebrain atrophy rates in people at risk for Alzheimer disease. Neurology 94, e30.

29.

Salvadó

, Molinuevo

, Brugulat-Serrat

, Falcon

, Grau-Rivera

, Suárez-Calvet

, Pavia

, Niñerola-Baizán

, Perissinotti

, Lomeña

, Minguillon

, Fauria

, Zetterberg

, Blennow

, Gispert

, Camí

, Cacciaglia

, Crous-Bou

, Deulofeu

, Dominguez

, Gotsens

, Gramunt

, Hernandez

, Huesa

, Huguet

, León

, Marne

, Menchón

, Milà

, Operto

, Pascual

, Polo

, Pradas

, Sala-Vila

, Sánchez-Benavides

, Segundo

, Soteras

, Tenas

, Vilanova

, Vilor-Tejedor

, for theAlzheimer’s Disease Neuroimaging Initiative, for the ALFA Study (2019) Centiloid cut-off values for optimal agreement between PETand CSF core AD biomarkers. Alzheimers Res Ther 11, 27.

30.

Guo

, Landau

, Jagust

(2020) Detecting earlier stages of amyloid deposition using PET in cognitively normal elderly adults. Neurology 94, e1512.

31.

, Daamen

, Scheef

, Gaertner

, Buchert

, Buchmann

, Buerger

, Catak

, Dobisch

, Drzezga

, Ertl-Wagner

, Essler

, Fliessbach

, Haynes

, Incesoy

, Kilimann

, Krause

, Lange

, Laske

, Priller

, Ramirez

, Reimold

, Rominger

, Roy

, Scheffler

, Maurer

, Schneider

, Spottke

, Spruth

, Teipel

, Tscheuschler

, Wagner

, Wolfsgruber

, Düzel

, Jessen

, Peters

, Boecker

, the DELCODE Study Group (2021) Abnormal regional and global connectivity measures in subjective cognitive decline depending on cerebral amyloid status. J Alzheimers Dis 79, 493–509.

32.

Jessen

, Spottke

, Boecker

, Brosseron

, Buerger

, Catak

, Fliessbach

, Franke

, Fuentes

, Heneka

, Janowitz

, Kilimann

, Laske

, Menne

, Nestor

, Peters

, Priller

, Pross

, Ramirez

, Schneider

, Speck

, Spruth

, Teipel

, Vukovich

, Westerteicher

, Wiltfang

, Wolfsgruber

, Wagner

, Duzel

(2018) Design and first baseline data of the DZNE multicenter observational study on predementia Alzheimer’s disease (DELCODE). Alzheimers Res Ther 10, 15.

33.

Jessen

, Amariglio

, Buckley

, van der Flier

, Han

, Molinuevo

, Rabin

, Rentz

, Rodriguez-Gomez

, Saykin

, Sikkes

SAM

, Smart

, Wolfsgruber

, Wagner

(2020) The characterisation of subjective cognitive decline. Lancet Neurol 19, 271–278.

34.

Morris

, Heyman

, Mohs

, Hughes

, van Belle

, Fillenbaum

(1989) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 39, 1159–1165.

35.

Folstein

, Folstein

, Mchugh

(1975) “Mini-mental state”. a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

36.

Mohs

, Knopman

, Petersen

, Ferris

, Ernesto

, Grundman

, Sano

, Bieliauskas

, Geldmacher

, Clark

, Thal

(1997) Development of cognitive instruments for use in clinical trials of antidementia drugs: Additions to the Alzheimer’s Disease Assessment Scale that broaden its scope. The Alzheimer’s Disease Cooperative Study. Alzheimer Dis Assoc Disord 11(Suppl 2), S13–21.

37.

Smith

(1982) Symbol digit modality test (SDMT): Manual (revised), Western Psychological Services, Los Angeles.

38.

Reitan

(1958) Validity of the Trail Making Test as an indicator of organic brain damage. Percept Mot Skills 8, 271–276.

39.

Seibyl

, Catafau

, Barthel

, Ishii

, Rowe

, Leverenz

, Ghetti

, Ironside

, Takao

, Akatsu

, Murayama

, Bullich

, Mueller

, Koglin

, Schulz-Schaeffer

, Hoffmann

, Sabbagh

, Stephens

, Sabri

(2016) Impact of training method on the robustness of the visual assessment of 18F-Florbetaben pet scans: Results from a phase-3 study. J Nucl Med 57, 900–906.

40.

Klunk

, Koeppe

, Price

, Benzinger

, Devous

Sr. , Jagust

, Johnson

, Mathis

, Minhas

, Pontecorvo

, Rowe

, Skovronsky

, Mintun

(2015) The Centiloid Project: Standardizing quantitative amyloid plaque estimation by PET. Alzheimers Dement 11, 1-15e11–14.

41.

Jagust

, Landau

, Koeppe

, Reiman

, Chen

, Mathis

, Price

, Foster

, Wang

(2015) The Alzheimer’s disease neuroimaging initiative 2 PET core: 2015. Alzheimers Dement 11, 757–771.

42.

Gaser

, Dahnke

, Thompson

, Kurth

, Luders

, Alzheimer’s Disease Neuroimaging Initiative (2022) CAT – A Computational Anatomy Toolbox for the analysis of structural MRI data. bioRxiv, 2022.2006.2011.495736.

43.

Ashburner

(2007) A fast diffeomorphic image registration algorithm. Neuroimage 38, 95–113.

44.

Mesulam

, Mufson

, Levey

, Wainer

(1983) Cholinergic innervation of cortex by the basal forebrain: Cytochemistry and cortical connections of the septal area diagonal band nuclei, nucleus 1145 basalis (substantia innominata), and hypothalamus in the rhesus monkey. J Comp Neurol 214, 170–197.

45.

Rorden

, Brett

(2000) Stereotaxic display of brain lesions. Behav Neurol 12, 191–200.

46.

Hammers

, Allom

, Koepp

, Free

, Myers

, Lemieux

, Mitchell

, Brooks

, Duncan

(2003) Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp 19, 224–247.

47.

Amunts

, Mohlberg

, Bludau

, Zilles

(2020) Julich-Brain: A 3D probabilistic atlas of the human brain’s cytoarchitecture. Science 369, 988.

48.

Malone

, Leung

, Clegg

, Barnes

, Whitwell

, Ashburner

, Fox

, Ridgway

(2015) Accurate automatic estimation of total intracranial volume: A nuisance variable with less nuisance. Neuroimage 104, 366–372.

49.

, Zhao

, He

, Wang

, Lu

, Li

, Kang

, Han

, Li

(2021) Divergent connectivity changes in gray matter structural covariance networks in subjective cognitive decline, amnestic mild cognitive impairment, and Alzheimer’s disease. Front Aging Neurosci 13, 686598.

50.

Cantero

, Atienza

, Lage

, Zaborszky

, Vilaplana

, Lopez-Garcia

, Pozueta

, Rodriguez-Rodriguez

, Blesa

, Alcolea

, Lleo

, Sanchez-Juan

, Fortea

, Alzheimer’s Disease Neuroimaging Initiative (2020) Atrophy of basal forebrain initiates with tau pathology in individuals at risk for Alzheimer’s disease. Cereb Cortex 30, 2083–2098.

51.

Butler

, Harvey

, Deshpande

, Tanzi

, Li

, Tsui

, Silver

, Fischer

, Wang

, Chen

, Rusinek

, Pirraglia

, Osorio

, Glodzik

, de Leon

(2018) Basal forebrain septal nuclei are enlarged in healthy subjects prior to the development of Alzheimer’s disease. Neurobiol Aging 65, 201–205.

52.

Jethwa

, Dhillon

, Meng

, Auer

(2019) Are linear measurements of the nucleus basalis of Meynert suitable as a diagnostic biomarker in mild cognitive impairment and Alzheimer disease? Am J Neuroradiol 40, 2039.

53.

Boncristiano

, Calhoun

, Kelly

, Pfeifer

, Bondolfi

, Stalder

, Phinney

, Abramowski

, Sturchler-Pierrat

, Enz

, Sommer

, Staufenbiel

, Jucker

(2002) Cholinergic changes in the APP23 transgenic mouse model of cerebral amyloidosis. J Neurosci 22, 3234–3243.

54.

Ramos-Rodriguez

, Pacheco-Herrero

, Thyssen

, Murillo-Carretero

, Berrocoso

, Spires-Jones

, Bacskai

, Garcia-Alloza

(2013) Rapid β-amyloid deposition and cognitive impairment after cholinergic denervation in APP/PS1 mice. J Neuropathol Exp Neurol 72, 272–285.

55.

Aghourian

, Aumont

, Grothe

, Soucy

J-P

, Rosa-Neto

, Bedard

M-A

(2021) FEOBV-PET to quantify corticalcholinergic denervation in AD: Relationship to basal forebrain volumetry. J Neuroimaging 31, 1077–1081.

56.

Xia

, Eeles

, Fripp

, Pinsker

, Thomas

, Latter

, Doré

, Fazlollahi

, Bourgeat

, Villemagne

, Coulson

, Rose

(2022) Reduced cortical cholinergic innervation measured using [18F]-FEOBV PET imagingcorrelates with cognitive decline in mild cognitive impairment. Neuroimage Clin 34, 102992.

57.

Schmitz

, Mur

, Aghourian

, Bedard

, Spreng

, Alzheimer’s Disease Neuroimaging Initiative (2018) Longitudinal Alzheimer’s degeneration reflects the spatial topography of cholinergic basal forebrain projections. Cell Rep 24, 38–46.

58.

Sabri

, Kendziorra

, Wolf

, Gertz

, Brust

(2008) Acetylcholine receptors in dementia and mild cognitive impairment. Eur J Nucl Med Mol Imaging 35(Suppl 1), S30–S45.

59.

Peter

, Mayer

, Kammer

, Minkova

, Lahr

, Klöppel

, Grothe

, Orth

(2021) The relationship between cholinergic system brain structure and function in healthy adults and patients with mild cognitive impairment. Sci Rep 11, 16080.

60.

Lammers

, Borchers

, Feinkohl

, Hendrikse

, Kant

IMJ

, Kozma

, Pischon

, Slooter

AJC

, Spies

, van Montfort

SJT

, Zacharias

, Zaborszky

, Winterer

(2018) Basal forebrain cholinergic system volume is associated with general cognitive ability in the elderly. Neuropsychologia 119, 145–156.

61.

Convery

, Neason

, Cash

, Cardoso

, Modat

, Ourselin

, Warren

, Rohrer

, Bocchetta

(2020) Basal forebrain atrophy in frontotemporal dementia. Neuroimage Clin 26, 102210.

62.

Grothe

, Schuster

, Bauer

, Heinsen

, Prudlo

, Teipel

(2014) Atrophy of the cholinergic basal forebrain in dementia with Lewy bodies and Alzheimer’s disease dementia. J Neurol 261, 1939–1948.

63.

Jansen

, Janssen

, Tijms

, Vos

SJB

, Ossenkoppele

, Visser

, Amyloid Biomarker Study Group (2022) Prevalence estimates of amyloid abnormality across the Alzheimer disease clinical spectrum. JAMA Neurol 79, 228–243.

64.

Yamazaki

, Zhao

, Caulfield

, Liu

C-C

, Bu

(2019) Apolipoprotein E and Alzheimer disease: Pathobiology and targeting strategies. Nat Rev Neurol 15, 501–518.

65.

Poirier

(1994) Apolipoprotein E in animal models of CNS injury and in Alzheimer’s disease. Trends Neurosci 17, 525–530.

66.

Schmitz

, Soreq

, Poirier

, Spreng

(2020) Longitudinal basal forebrain degeneration interacts with TREM2/C3 biomarkers of inflammation in pre-symptomatic Alzheimer’s disease. J Neurosci 40, 1931–1942.

67.

Lai

Y-LL

, Chen

, Lee

T-W

, Tso

C-W

, Lin

H-H

, Kuo

L-W

, Chen

C-Y

, Liu

H-S

, for the Alzheimer Disease Neuroimaging Initiative (2021) The effect of the APOE-ɛ4 allele on the cholinergic circuitry for subjects with different levels of cognitive impairment. Front Neurol 12, 651388.

68.

Mesulam

, Shaw

, Mash

, Weintraub

(2004) Cholinergic nucleus basalis tauopathy emerges early in the aging-MCI-AD continuum. Ann Neurol 55, 815–828.

69.

Fernández-Cabello

, Kronbichler

, Van Dijk

KRA

, Goodman

, Spreng

, Schmitz

, Alzheimer’s Disease Neuroimaging Initiative (2020) Basal forebrain volume reliably predicts the cortical spread of Alzheimer’s degeneration. Brain 143, 993–1009.

70.

Ossenkoppele

, Hansson

(2021) Towards clinical application of tau PET tracers for diagnosing dementia due to Alzheimer’s disease. Alzheimers Dement 17, 1998–2008.

71.

Bullich

, Seibyl

, Catafau

, Jovalekic

, Koglin

, Barthel

, Sabri

, De Santi

(2017) Optimized classification of 18F-Florbetaben PET scans as positive and negative using an SUVR quantitative approach and comparison to visual assessment. Neuroimage Clin 15, 325–332.

Cortical Amyloid Burden Relates to Basal Forebrain Volume in Subjective Cognitive Decline

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

INTRODUCTION

METHODS

Participants

Background characteristics

[18F]Florbetaben (FBB) PET acquisition and analysis

FBB PET analyses

MRI acquisition and analysis

MRI data processing and BF volume calculation

Statistical analysis

RESULTS

Group comparisons for background characteristics

Group comparisons for volumetric data

Linear relationships between BF volumes and cortical amyloid burden

Linear relationships between BF volumes, amyloid burden, and cognitive functioning

DISCUSSION

Conclusions

Footnotes

ACKNOWLEDGMENTS

DELCODE STUDY GROUP

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References

[¹⁸F]Florbetaben (FBB) PET acquisition and analysis