Impaired Glymphatic Flow on Diffusion Tensor MRI as a Marker of Neurodegeneration in Alzheimer’s Disease: Correlation with Gray Matter Volume Loss and Cognitive Decline Independent of Cerebral Amyloid Deposition

Abstract

Background:

Impaired glymphatic flow on the Alzheimer’s disease (AD) spectrum may be evaluated using diffusion tensor image analysis along the perivascular space (DTI-ALPS).

Objective:

We aimed to validate impaired glymphatic flow and explore its association with gray matter volume, cognitive status, and cerebral amyloid deposition on the AD spectrum.

Methods:

80 participants (mean age, 76.9±8.5 years; 57 women) with AD (n = 65) and cognitively normal (CN) (n = 15) who underwent 3T brain MRI including DTI and/or amyloid PET were included. After adjusting for age, sex, apolipoprotein E status, and burden of white matter hyperintensities, the ALPS-index was compared according to the AD spectrum. The association between the ALPS-index and gray matter volume, cognitive status, and quantitative amyloid from PET was assessed.

Results:

The ALPS-index in the AD was significantly lower (mean, 1.476; 95% CI, 1.395–1.556) than in the CN (1.784;1.615–1.952; p = 0.026). Volumes of the entorhinal cortex, hippocampus, temporal pole, and primary motor cortex showed significant associations with the ALPS-index (all, p < 0.05). There was a positive correlation between the ALPS-index and MMSE score (partial r = 0.435; p < 0.001), but there was no significant correlation between the ALPS-index and amyloid SUVRs (all, p > 0.05).

Conclusions:

Decreased glymphatic flow measured by DTI-ALPS in AD may serve as a marker of neurodegeneration correlating with structural atrophy and cognitive decline.

Keywords

Alzheimer’s disease amyloid PET diffusion tensor glymphatic function volumetry

1 INTRODUCTION

Recent evidence suggests that the glymphatic system may be responsible for clearance of protein aggregates from the brain that underlie neurodegenerative diseases [1, 2]. The accumulation of amyloid-β plaques as well as neurofibrillary tangles of hyperphosphorylated tau protein is implicated in Alzheimer’s disease (AD) supported by histopathological and positron emission tomography (PET) studies, and is correlated with cognitive decline [3, 4]. Previous studies have shown that the net clearance of amyloid-β plaques can be affected by the glymphatic system, and perivascular localization of aquaporin-4, which is known to play an important role in the glymphatic system, is significantly associated with AD [1 , 5]. This suggests that glymphatic dysfunction may engage in the pathologic cascade that includes amyloid/tau deposition, structural atrophy, and clinical cognitive impairment, contributing the development of AD.

The glymphatic system is a highly organized fluid transport system that clears cerebral waste products and is currently best understood in animal models [1, 2]. While the glymphatic system can be visualized directly on the magnetic resonance imaging (MRI) using contrast agents [5 –7], there are limitations in human research due to the restrictions on the intrathecal injection of contrast agents into the cerebrospinal fluid (CSF) and the challenges in demonstrating the minute amount of contrast agents injected intravenously [8]. Recently, diffusion tensor image analysis along the perivascular space (DTI-ALPS) has been proposed as a non-invasive method for evaluating the glymphatic system in human brain without necessity of contrast agents [9]. DTI-ALPS can measure water diffusivities along the perivascular spaces, which can reflect glymphatic flow [9, 10]. Its measurement has been shown to be correlated with the direct clearance of intrathecally injected contrast agents, providing evidence that DTI-ALPS can be a reliable method for measuring glymphatic clearance function [10]. Many researchers including Taoka et al. have shown decreased glymphatic flow in patients with AD and/or mild cognitive impairment (MCI) using DTI-ALPS and its correlation with cognitive impairment [9 , 12]. However, the association of glymphatic flow with gray matter volume or cerebral amyloid deposition in AD needs to be further elucidated.

In this study, we aimed to validate impaired glymphatic flow using DTI-ALPS in AD, and explore its association with gray matter volume, cognitive decline, and cerebral amyloid deposition. The purpose of this study was to evaluate DTI-ALPS as an imaging biomarker in reflecting the pathological cascade of AD.

2 MATERIALS AND METHODS

2.1 Study population

This retrospective study was approved by the institutional review board (IRB No. B-2002-738-103) of Seoul National University Bundang Hospital, a tertiary hospital. The requirement to obtain informed consent was waived.

Between March and August 2021, consecutive participants who visited the Dementia Clinic of our institution with symptoms of subjective memory decline were identified. Among them, 98 participants underwent 3T brain MRI including DTI and structural imaging and/or amyloid PET using a tracer of ¹⁸F. Among them, participants who were diagnosed as AD (n = 72) and those whose cognition (n = 15) was normal were included in our study. Participants were excluded if they were diagnosed as probable AD but showed negative amyloid PET results (suspected non-Alzheimer’s disease pathology, n = 5), and if their DTI were not adequate for the processing of DTI-ALPS (n = 2). As a result, a total of 80 participants including 65 diagnosed with AD and 15 with normal cognition were included in this study (Fig. 1).

Fig. 1

Participant flow diagram.

2.2 Clinical assessment

Geriatric neuropsychiatrists with expertise in dementia research administered a face-to-face standardized diagnostic interview as well as physical and neurological examinations using the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet Clinical Assessment Battery (CERAD-K-C) [13] to diagnose cognitive disorders. Laboratory tests, including complete blood counts, chemistry profiles, serological tests for syphilis and apolipoprotein E (APOE) genotyping, were performed for each participant. Research neuropsychologists and trained research nurses administered the CERAD-K-N, which consists of the following neuropsychological tests: Verbal Fluency Test, 15-item Boston Naming Test, Mini-Mental State Examination (MMSE), Word List Memory Test, Constructional Praxis Test, Word List Recall Test, Word List Recognition Test, Constructional Recall Test, Trail Making Test A/B, Digit Span Test, and Frontal Assessment Battery [13, 14].

AD was diagnosed according to the clinical criteria for probable Alzheimer’s disease (AD) dementia from the National Institute on Aging-Alzheimer’s Association [15]. Cognitively normal (CN) was defined by a normal score on cognitive tests.

2.3 MRI acquisition and analysis

All MRI examinations were obtained using a 3T system (Ingenia and Ingenia CX, Philips, Best, Netherlands) with a 32-channel SENSE head coil (Philips Healthcare). DTI was obtained in the axial plane using single-shot echo-planar imaging with b-values of 0 and 1,000 s/mm³ and 32 directions of motion-probing gradients. Color-coded fractional anisotropy (FA) map was automatically generated at the MRI console. The following imaging parameters for DTI were used: repetition time, 9900 ms; echo time, 77 msec; slice thickness, 2 mm; flip angle, 90°, field-of-view, 224×224 mm²; acquisition matrix, 112×112. Along with DTI, three-dimensional (3D) T1-weighted image (T1WI), 3D fluid-attenuated inversion recovery (FLAIR) image, axial T2-weighted image (T2WI), and susceptibility weighted image (SWI) were obtained. The detailed imaging parameters are provided in the Supplementary Material. DTI and SWI were used for DTI-ALPS, 3D T1WI for volumetric analysis, and FLAIR for determining the burden of white matter hyperintensities. This was evaluated by one neuroradiologist (M. K. with 8 years of experience in neuroradiology) who graded white matter hyperintensity in the periventricular and deep white of FLAIR image depending on the size and confluence of lesions according to Fazekas scale [10]. Grading of hyperintensity in the periventricular (scale 0–3) and deep white matter (scale 0–3) was added, and burden of small vessel disease was scaled from 0 to 6.

2.4 DTI-ALPS processing and measurement

We adopted the method for DTI-ALPS processing from the previous publications [9, 16]. For DTI-ALPS calculations, we used an inhouse program named DTI-ALPS Analyzer (MATLAB 9.6, version 2019a, The Mathworks, Inc.). Diffusion coefficients along three orthogonal directions (Dxx, Dyy and Dzz) were extracted from the calculated diffusion tensor, pixel by pixel, for the generation of the diffusion coefficient maps along x, y and z directions (in other words, right-left, anterior-posterior, and cranio-caudal direction along head orientation). These three orthogonal diffusion coefficient maps were used for the DTI-ALPS. Calculated tensors were diagonalized based on singular value decomposition method (SVD), for eigen values and eigen vectors calculation. Using this information, a color-coded FA map was post-processed.

Two neuroradiologists (B.Y.J. with 13 years and M.K. with 8 years of experience in neuroradiology) independently measured DTI-ALPS values. With the reference to the SWI, the axial slice at the level of the lateral ventricular body was chosen where the trans-medullary vessels and perivascular water flow traverse perpendicular to the ventricle [17, 18]. The three bundles of neural fibers identified on the color-coded FA map were considered that contribute to the diffusivity across the perivascular flow: (a) projection neural fibers along the z-axis, (b) association neural fibers along the y-axis, and (c) subcortical neural fibers along the x-axis. On the color-coded FA map, regions-of-interest (ROIs) with a diameter of 5 mm for the areas where projection neural fibers, association neural fibers, and subcortical neural fibers pass were created by the two readers independently, and Dx, Dy, and Dz were automatically generated by the unit of apparent diffusion coefficient (ADC, ×10^–3 mm²/s) in the area of each neural fibers from inhouse software. As a result, a total of nine separate DTI-ALPS measurements from nine ROIs were obtained: the diffusivities in the projection neural fiber (Dxproj, Dyproj, Dzproj), association neural fiber (Dxassoc, Dyassoc, Dzassoc), and subcortical neural fiber (Dxsubc, Dysubc, Dzsubc). As previously described, Dxproj and/or Dxassoc were considered to purely reflect perivascular water flow by excluding diffusivities from the projection and association neural fibers that run perpendicular to the perivascular water flow. Dyproj and Dyassoc were used to account of periventricular white matter hyperintensity. Hence, the following equation was used to calculate ALPS-index to eliminate white matter effect by the division calculation and allow individual glymphatic assessment [9]: $ALPS - index = mean (D xproj, D xassoc) /$ $mean (D yproj, D zassoc)$

ALPS-index close to 1 indicates the impaired perivascular glymphatic water flow, but larger ALPS-index represents higher water diffusivity along the perivascular space, meaning better glymphatic function [9].

2.5 Gray matter volumetric analysis

The high-resolution 3D T1WI from brain MRI was used for automated volumetric analyses using NeuroQuant software package (CorTechs Labs, La Jolla, CA, USA). DICOM files were uploaded to the servers for processing, and the details of this process was previously described elsewhere [19, 20]. The total volume and derived parameters of 61 brain regions were included in the analysis (Supplementary Table 1).

2.6 [¹⁸F]Florbetaben PET protocol and analysis

All participants who underwent PET scans received a bolus intravenous injection of [¹⁸F]Florbetaben 300 MBq (Neuraceq, Piramal, Mumbai, India). Images were acquired with a dedicated PET/CT scanner (Biograph mCTFlow, Siemens Healthcare, Germany). CT scans were acquired prior to the PET emission scan for attenuation correction purposes. Ninety minutes after injection, PET emission scans were acquired for 20 min, comprised of four 5-minute dynamic frames. Images were reconstructed with ordered subset expectation maximization (OSEM) with time-of-flight application, with 21 subsets and 4 iteration numbers, on 400×400 image size.

A nuclear medicine physician (Y.S.S. with 16 years of experience in nuclear medicine) performed visual reading and quantification of the PET images. Visual reading was determined by assessing the presence of amyloid deposition based on brain amyloid plaque load (BAPL) scores [21]. BAPL 1 indicates negative amyloid deposition, while BAPL 2 and BAPL 3 indicates positive amyloid deposition [21]. Amyloid quantification was performed based on a deep neural network model [22]. The amyloid uptake was quantified using standard uptake value ratio (raw SUVR), and was subsequently scaled to the centiloid units for standardization [23]. Both raw and centiloid SUVR were used for the analysis, and bilateral global and 59 regional SUVRs were included in the analysis (Supplementary Table 2).

2.7 Statistical analysis

The demographic findings of the participants were compared with Chi-square test for categorical variables and Mann-Whitney test for continuous variables. Inter-observer agreement on the DTI-ALPS measurements including diffusivities and the ALPS-index between two readers were assessed by interclass correlation coefficient. The ALPS-index amongst AD and CN groups was compared using liner mixed model. A random effect for each participant was added in the linear mixed model to account for repeated measures. Multiple regression analysis was conducted to examine the relationship between the ALPS-index and age, sex, APOE status, and burden of white matter hyperintensities. The correlations between the mean ALPS-index and the gray matter volume, MMSE, and the SUVRs from the amyloid PET were evaluated by Spearman’s correlation coefficient. The least absolute shrinkage and selection operator (LASSO) logistic model was used to select associated volumes with non-zero coefficients. All statistical analyses were adjusted for age, sex, APOE status, and burden of white matter hyperintensities based on the FLAIR images, taking into account multiple comparison correction. Results were considered significant if p < 0.05. Statistical analyses were performed using R software version 4.2.1 (https://www.r-project.org).

3 RESULTS

3.1 Study participants

The clinical characteristics of the two groups of study participants are provided in Table 1 : 65 participants (81%) in AD group and 15 in CN group (17%). Among the 65 AD participants, amyloid PET was performed on 37 individuals, and all of them were confirmed to be amyloid positive. There were significant differences in the age (p = 0.001), burden of white matter hyperintensities (p < 0.001), MMSE scores (p < 0.001), and proportion of amyloid PET positivity (p < 0.001) amongst the study groups.

Table 1
Clinical characteristics of the study participants

AD CN p

n 65 (81%) 15 (19%)

Age (y) 78.3±8.3 70.7±6.2 0.001

Sex (M:F) 17 : 53 8 : 7 0.028

Education (y) 10.1±5.3 12.2±5.4 0.165

APOE ɛ4 (0 : 1:2, NA)^* 26 : 33 : 2, 4 11 : 3:0, 1 0.019

Fazekas score

Periventricular white matter 1.16±0.69 0.40±0.83 <0.001

Deep white matter 1.14±0.59 0.47±0.83 <0.001

MMSE 18.1±5.8 25.5±3.6 <0.001

Amyloid PET positivity^** 37/37 (100%) 0/10 (0%) <0.001

	AD	CN	p
n	65 (81%)	15 (19%)
Age (y)	78.3±8.3	70.7±6.2	0.001
Sex (M:F)	17 : 53	8 : 7	0.028
Education (y)	10.1±5.3	12.2±5.4	0.165
APOE ɛ4 (0 : 1:2, NA)^*	26 : 33 : 2, 4	11 : 3:0, 1	0.019
Fazekas score
Periventricular white matter	1.16±0.69	0.40±0.83	<0.001
Deep white matter	1.14±0.59	0.47±0.83	<0.001
MMSE	18.1±5.8	25.5±3.6	<0.001
Amyloid PET positivity^**	37/37 (100%)	0/10 (0%)	<0.001

Data are expressed as the mean±standard deviation unless otherwise specified. AD, Alzheimer’s disease; APOE, apolipoprotein E; CN, cognitively normal; MMSE, Mini-Mental State Examination. ^*0/1/2 indicates non-carriers/heterozygotes/homozygotes for APOE ɛ4 respectively. ^**The numbers indicate the number of participants with amyloid PET positivity over the number of entire participants who underwent amyloid PET in each group.

3.2 ALPS-index in participants on the AD spectrum

The interobserver agreement of the diffusivities and ALPS-index were excellent except for Dysubc which showed good reliability (Supplementary Table 3). Diffusivities and ALPS-index in the AD and CN groups are presented in Table 2. Our result from multiple regression analysis revealed that age, sex, APOE status, and burden of white matter hyperintensities did not significantly predict the ALPS-index (all, p > 0.05) (Table 3). After adjusting for age, sex, APOE status and burden of white matter hyperintensities, there were significant differences between the AD and CN groups in the ALPS-index (p = 0.017), Dxassoc (p = 0.038), Dyproj (p = 0.011), and Dyassoc (p = 0.011).

Table 2
Diffusivities and ALPS-index in the study participants

AD CN p ^*

Diffusivity

Dxproj 0.673 (0.627–0.718) 0.640 (0.546–0.734) 0.558

Dxassoc 0.781 (0.735–0.828) 0.866 (0.769–0.964) 0.038

Dxsubc 1.011 (0.956–1.066) 0.964 (0.849–1.078) 0.332

Dyproj 0.560 (0.527–0.592) 0.419 (0.351–0.487) 0.011

Dyassoc 1.066 (1.007–1.126) 0.839 (0.716–0.962) 0.011

Dysubc 0.804 (0.760–0.848) 0.752 (0.661–0.843) 0.430

Dzproj 1.185 (1.128–1.243) 0.982 (0.862–1.101) 0.196

Dzassoc 0.453 (0.412–0.494) 0.440 (0.355–0.526) 0.396

Dzsubc 0.653 (0.611–0.695) 0.568 (0.481–0.656) 0.522

ALPS-index 1.476 (1.395–1.556) 1.784 (1.615–1.952) 0.017

	AD	CN	p ^*
Diffusivity
Dxproj	0.673 (0.627–0.718)	0.640 (0.546–0.734)	0.558
Dxassoc	0.781 (0.735–0.828)	0.866 (0.769–0.964)	0.038
Dxsubc	1.011 (0.956–1.066)	0.964 (0.849–1.078)	0.332
Dyproj	0.560 (0.527–0.592)	0.419 (0.351–0.487)	0.011
Dyassoc	1.066 (1.007–1.126)	0.839 (0.716–0.962)	0.011
Dysubc	0.804 (0.760–0.848)	0.752 (0.661–0.843)	0.430
Dzproj	1.185 (1.128–1.243)	0.982 (0.862–1.101)	0.196
Dzassoc	0.453 (0.412–0.494)	0.440 (0.355–0.526)	0.396
Dzsubc	0.653 (0.611–0.695)	0.568 (0.481–0.656)	0.522
ALPS-index	1.476 (1.395–1.556)	1.784 (1.615–1.952)	0.017

Data are expressed as the mean with 95% confidence intervals in the parentheses unless otherwise specified. AD, Alzheimer’s disease; CN, cognitively normal; Dxproj, diffusivity along the x-axis in projection fiber area; Dxassoc, diffusivity along the x-axis in association fiber area; Dxsubc, diffusivity along the x-axis in subcortical fiber area; Dyproj, diffusivity along the y-axis in projection fiber area; Dyassoc, diffusivity along the y-axis in association fiber area; Dysubc, diffusivity along the y-axis in subcortical fiber area; Dzproj, diffusivity along the z-axis in projection fiber area; Dzassoc, diffusivity along the z-axis in association fiber area; Dzsubc, diffusivity along the z-axis in subcortical fiber area. ^*p values were adjusted for age, sex, APOE status, and burden of white matter hyperintensities.

Table 3

Predictive value of clinical factors for the ALPS-index

Variables	B	SE	β	t	p	VIF
Intercept	2.104	0.376		5.592
Age	–0.009	0.005	–0.228	–1.710	0.092	1.360
Sex	0.131	0.087	0.181	1.505	0.137	1.110
APOE status	0.070	0.074	0.108	0.946	0.347	1.004
Burden of white matter hyperintensities	0.021	0.030	0.096	0.694	0.490	1.456

SE, standard error; VIF, variance inflation factor; APOE, apolipoprotein E.

3.3 Association of ALPS-index with gray matter volume

Amongst the volumes of all anatomical locations and volumes as percent of intracranial volume and normative percentiles, there were 34 volumetric measurements with significant correlation with ALPS-index (Table 4).

Table 4
Correlation of the ALPS-index with gray matter volumetric measurements

Partial r p

^Total volume of the primary motor cortex as percent of ICV 0.339 0.002

Volume of the left temporal lobe as percent of ICV 0.293 0.008

Total volume of the primary motor cortex 0.360 0.001

Volume of the left parahippocampal gyrus as normative percentile 0.222 0.048

Volume of the left parahippocampal gyrus as percent of ICV 0.263 0.018

Volume of the right primary motor cortex 0.336 0.002

Total volume of the amygdala 0.331 0.003

Volume of the right primary motor cortex as percent of ICV 0.303 0.006

Volume of the right lingual gyrus as percent of ICV 0.245 0.029

Volume of the right pars opercularis as percent of ICV 0.274 0.014

Total volume of the temporal lobe as percent of ICV 0.341 0.002

Total volume of the lingual gyrus as percent of ICV 0.255 0.022

Volume of the right amygdala 0.324 0.003

Volume of the left primary motor cortex 0.306 0.006

Volume of the right entorhinal cortex 0.330 0.003

Volume of the right pars opercularis 0.309 0.005

Volume of the left cortical gray matter as percent of ICV 0.275 0.014

Volume of the left amygdala 0.303 0.006

Volume of the right hippocampus as percent of ICV 0.373 0.001

Volume of the left temporal lobe 0.284 0.011

^Volume of the right hippocampus 0.393 0.0003

Volume of the right temporal pole as normative percentile 0.265 0.017

^Volume of the right temporal pole 0.372 0.001

Total volume of the temporal lobe 0.320 0.004

Volume of the left fusiform gyrus as percent of ICV 0.229 0.041

Volume of the left superior and transverse temporal gyrus as percent of ICV 0.299 0.007

Volume of the right entorhinal cortex as percent of ICV 0.303 0.006

^Total volume of the entorhinal cortex 0.343 0.002

Volume of the right temporal pole as percent of ICV 0.357 0.001

Volume of the right temporal lobe as percent of ICV 0.349 0.002

Total volume of the cortical gray matter as percent of ICV 0.320 0.004

Volume of the right cortical gray matter as percent of ICV 0.340 0.002

^*Total volume of the entorhinal cortex as percent of ICV 0.340 0.002

Total volume of the pars opercularis 0.285 0.010

	Partial r	p
^*Total volume of the primary motor cortex as percent of ICV	0.339	0.002
Volume of the left temporal lobe as percent of ICV	0.293	0.008
Total volume of the primary motor cortex	0.360	0.001
Volume of the left parahippocampal gyrus as normative percentile	0.222	0.048
Volume of the left parahippocampal gyrus as percent of ICV	0.263	0.018
Volume of the right primary motor cortex	0.336	0.002
Total volume of the amygdala	0.331	0.003
Volume of the right primary motor cortex as percent of ICV	0.303	0.006
Volume of the right lingual gyrus as percent of ICV	0.245	0.029
Volume of the right pars opercularis as percent of ICV	0.274	0.014
Total volume of the temporal lobe as percent of ICV	0.341	0.002
Total volume of the lingual gyrus as percent of ICV	0.255	0.022
Volume of the right amygdala	0.324	0.003
Volume of the left primary motor cortex	0.306	0.006
Volume of the right entorhinal cortex	0.330	0.003
Volume of the right pars opercularis	0.309	0.005
Volume of the left cortical gray matter as percent of ICV	0.275	0.014
Volume of the left amygdala	0.303	0.006
Volume of the right hippocampus as percent of ICV	0.373	0.001
Volume of the left temporal lobe	0.284	0.011
^*Volume of the right hippocampus	0.393	0.0003
Volume of the right temporal pole as normative percentile	0.265	0.017
^*Volume of the right temporal pole	0.372	0.001
Total volume of the temporal lobe	0.320	0.004
Volume of the left fusiform gyrus as percent of ICV	0.229	0.041
Volume of the left superior and transverse temporal gyrus as percent of ICV	0.299	0.007
Volume of the right entorhinal cortex as percent of ICV	0.303	0.006
^*Total volume of the entorhinal cortex	0.343	0.002
Volume of the right temporal pole as percent of ICV	0.357	0.001
Volume of the right temporal lobe as percent of ICV	0.349	0.002
Total volume of the cortical gray matter as percent of ICV	0.320	0.004
Volume of the right cortical gray matter as percent of ICV	0.340	0.002
^*Total volume of the entorhinal cortex as percent of ICV	0.340	0.002
Total volume of the pars opercularis	0.285	0.010

ICV, intracranial volume. ^*Identified as significant volumetric measurements by LASSO.

When LASSO was used to identify significant volumes, volumetric measurements of four brain regions (total volume of the entorhinal cortex itself and as a percentage of intracranial volume, total volume of the primary motor cortex as a percentage of intracranial volume, volume of the right temporal pole, and volume of the right hippocampus) were identified (Fig. 2).

Fig. 2

Illustration of regions at which their volumetric measurements were identified to be associated with ALPS-index by the least absolute shrinkage and selection operator (LASSO) logistic model including (a) total volume of the entorhinal cortex, (b) total volume of the primary motor cortex, (c) volume of the right temporal pole, and (d) volume of the right hippocampus.

3.4 Correlation of ALPS-index with MMSE

Figure 3 shows the correlation between the ALPS-index and MMSE in all participants and participants with AD. When adjusted for age, sex, APOE status, and burden of white matter hyperintensities, there was a positive correlation between ALPS-index and MMSE (partial r = 0.435, p < 0.001 in all participants; partial r = 0.396, p = 0.003 in the participants with AD).

Fig. 3

Scatterplot of diffusion tensor image analysis along the perivascular space index (ALPS-index) and MMSE. All data are plotted with orange circles, and data from AD participants are plotted with blue circles. There was a significant positive correlation between the ALPS-index and MMSE in all participants (orange dashed line; partial r = 0.435, p < 0.001) and in AD (blue dashed line; partial r = 0.396, p = 0.003).

3.5 Association of ALPS-index with cerebral amyloid deposition

There were 47 participants (58.8%, 47/80) who underwent amyloid PET (37 participants with AD and 10 participants with CN), and 37 participants (78.7%, 37/47) showed amyloid PET positivity including all participants with AD (100%, 37/37) (Table 1). There was no significant correlation between the ALPS-index and cerebral amyloid deposition quantified as raw global (p = 0.659), centiloid global (p = 0.590), and bilateral regional SUVRs (lowest p = 0.178).

4 DISCUSSION

In this study, we validated glymphatic dysfunction by DTI-ALPS in the participants on the AD spectrum and demonstrated significant association with gray matter volume loss and cognitive decline, while no significant association was found with cerebral amyloid deposition. The ALPS-index was lower in AD than CN even after adjusting for age, sex, APOE status, and burden of white matter hyperintensities. Volume loss in the entorhinal cortex, hippocampus, temporal pole and primary motor cortex showed significant association with the ALPS-index, and there was a positive correlation with MMSE. As a non-invasive imaging biomarker, this result may imply that glymphatic dysfunction demonstrated by DTI-ALPS may reflect structural degeneration on MRI and cognitive decline in the pathological cascade of AD, independent from the early process of cerebral amyloid accumulation [24].

In accordance with previous publications [9 , 12], we successfully demonstrated impaired glymphatic function in AD compared to CN by reduced ALPS-index, which was correlated with the degree of cognitive impairment measured by MMSE, even after adjusting for multiple covariates attributing to AD. Noteworthy was that we included the burden of white matter hyperintensities as a covariate, which was not considered in many previous studies [11, 12]. White matter hyperintensities as small vessel disease have been shown to be closely related to glymphatic function [10, 25], and AD is known to have similar risk factors to cerebral small vessel disease, accounting for higher incidence of small vessel disease in patients with AD [26, 27]. Moreover, white matter hyperintensities are particularly relevant to the DTI-ALPS, since the ROIs used for measuring perivascular diffusivities are located in the periventricular and deep white matter where white matter hyperintensities are commonly present. Indeed, several studies have suggested the possibility of increasing Dy component in white matter hyperintensities due to white matter degeneration [9 , 28]. Therefore, measuring diffusivities and calculating ALPS-index without adjusting for white matter hyperintensities questions the ability of DTI-ALPS to purely reflect glymphatic dysfunction independently of disrupted white matter integrity seen in small vessel disease. To address this issue, we employed the Fazekas scale to quantify the burden of white matter hyperintensities, encompassing both periventricular and deep white matter, which we subsequently incorporated as a covariate in our statistical analysis. We believe that this adjustment can further strengthen the evidence for glymphatic dysfunction in AD and its association with clinically present cognitive decline, independent of the presence of white matter lesions.

Our results of ALPS-indices may seem higher than the previously reported values [9 –12], especially in the CN group. However, the ALPS-index is not an absolute value; rather, it is a relative value that can vary based on patient factors, such as age or disease state, and also on imaging parameters, such as head position, the number of motion-probing gradients, or echo time [29]. Accordingly, various studies have reported wide range of ALPS-indices, and many studies have reported an ALPS-index close to 2 or higher in their study cohorts as in ours [30 –32]. Similarly, ALPS-index values smaller than 1 have been previously reported [33 –35]. Therefore, we believe that it is feasible to compare ALPS-index values measured within a specific study cohort under identical imaging conditions.

In addition, the Dxproj values, originally known to represent perivascular glymphatic flow in the x-axis direction, showed no significant differences between our study groups. We can speculate that, despite the exclusion of the diffusivity from the major neural fiber tracts, other factors such as minor neural bundles in the white matter or individual variation in the white matter integrity can still influence the Dx metrics. This speculation could be supported by the significant differences in Dy values. Therefore, we believe that relying on the ALPS-index for the assessment of perivascular glymphatic water flow is crucial. It allows for the cancellation of the underlying white matter diffusivity, isolating the contribution of perivascular diffusivity from the water flow, and mitigating the effects the white matter disintegrity.

According to the amyloid cascade hypothesis, structural change can precede the clinically evident cognitive decline [24]. Therefore, we could postulate that ALPS-index could be correlated with the degree of brain atrophy measured by high-resolution MRI, if the glymphatic impairment contributes the structural damage in AD. As expected, we were able to demonstrate the positive correlation between the ALPS-index and the volumetric measurements of the entorhinal cortex, hippocampus, temporal pole, and primary motor cortex, which could suggest the relationship between global decrease in the glymphatic flow and the regional brain atrophy. As the atrophy of hippocampus, entorhinal cortex, and temporal lobe is well established in AD [36 –38], the correlation of glymphatic dysfunction with atrophy of these locations further supports the role of glymphatic dysfunction in the pathogenesis of AD. In addition, previous studies showed that the hippocampi are susceptible for glymphatic dysfunction [36, 39], and delayed removal of CSF-derived toxic molecules and accumulation of neuroinflammatory triggers are postulated to underlie neuronal loss [40]. Notably, Kamagata et al. [12] recently demonstrated that lower ALPS-index was associated with lower [¹⁸F]fluorodeoxyglucose (FDG) uptake in PET images. [¹⁸F]FDG uptake in the brain, particularly in the temporal and parietal lobes, is known to be reduced in patients with AD due to neuronal loss and decreased glucose metabolism, making it a biomarker for neurodegeneration [41, 42]. Therefore, our result of a significant correlation between impaired glymphatic function and regional brain atrophy specific to the AD is in accordance with the correlation between glymphatic function and the [¹⁸F]FDG uptake, as both can reflect the neurodegeneration resulting from the neuronal loss in the brain.

In our study, the ALPS-index did not show significant association with cerebral amyloid deposition. This may seem to contradict the results of the recent studies which have revealed negative correlation of the ALPS-index with the deposition of amyloid on PET images and with CSF Aβ₄₂ [11, 12]. There are several potential explanations for this finding. First, the higher mean age (age±standard deviation, 78.3±8.3) of our participants with AD compared to previous studies may be a factor, as age is well-recognized to increase amyloid PET SUVR and positivity [43]. Indeed, the percentage of amyloid-positive patients in AD group (100%) was higher compared to the previous research. The higher age can indicate that the amyloid plaque burden in the brain may have already reached a plateau, which may interfere the correlation with the ALPS-index. Second, unlike previous studies, we had removed the impact of white matter hyperintensities in comparing the ALPS-index between the study groups. Without this adjustment, the observed correlation between amyloid burden and the ALPS-index in other studies may reflect the relationship between small vessel disease and the glymphatic system [10, 44].

Yet, considering AD is a multifactorial disorder encompassing various pathologic process including aging, tau pathology, and neuroinflammation, in addition to amyloid deposition, all of which can contribute to the glymphatic dysfunction, it is possible that ALPS-index may not be directly correlated with amyloid deposition. In fact, it is recognized that amyloid deposition does not correlate with cognitive decline, since it starts earlier than tau accumulation, structural change, and clinical deterioration according to the AD cascade [24]. According to the AD cascade theory [24], amyloid deposition starts early in the CN phase and saturates by the time of MCI phase. Subsequent brain structure changes lead to detectable MRI-observable atrophy, continuing from late CN phase to MCI and AD stages. Clinical cognitive deterioration, detected by MMSE, occurs last, worsening from MCI to AD. Therefore, our result of ALPS-index correlating with brain volume and MMSE score, but not with amyloid SUVRs, may suggest that glymphatic dysfunction may be linked to later disease stages rather than early amyloid deposition, capturing downstream effects. Another evidence to support this suggestion was that the brain regions that correlated with the ALPS-index in our study were the entorhinal cortex, hippocampus, and temporal lobe, which are more closely associated with tau pathology than amyloid typically starting from the frontal base and posterior cingulate gyrus. Therefore, future prospective studies including larger number of participants with MCI and CN and with longer period of follow-up for disease progression are necessary to fully elucidate the underlying temporal relationship of glymphatic dysfunction, amyloid and tau deposition, neurodegeneration, and cognitive impairment in AD cascade.

Our study has limitations. First, this was a retrospective, single center study with a relatively small number of participants with CN relative to AD. Furthermore, despite statistical adjustments, the unbalanced distribution of clinical characteristics including age, sex, and the burden of white matter hyperintensities among the study groups could still have potentially influenced the results. Second, DTI-ALPS is an indirect way of evaluating glymphatic function based on the measurement of diffusivity along the deep medullary arterioles and veins at the level of the lateral ventricle body while offsetting the effects of projection and association fibers. However, the ROIs placed at the projection and association areas cannot solely include perivascular space, and the surrounding white matter structure may affect its measurement. We attempted to partly account for this by adjusting for burden of white matter hyperintensities in the statistical analyses of this study, but other variables that affect white matter integrity may need to be considered. In addition, the glymphatic function in the periventricular area may not reflect glymphatic alterations directly affecting each region of the brain. Although regional assessment of glymphatic flow is not currently possible with non-contrast MRI technique, future studies that assess glymphatic function in specific anatomical structures of the brain can enhance the understanding of the amyloid and tau pathology in AD affecting relevant regions. Third, incorporating volumetric measurements of white matter hyperintensities for adjustment could have enhanced the accuracy of our results when comparing the ALPS-index. Further studies using volumetric assessment of the white matter hyperintensities are warranted. Fourth, lack of longitudinal data as well as the small number of participants with CN limits the possibility of risk stratification of ALPS-index as a predictive or prognostic biomarker.

In conclusion, we corroborated glymphatic dysfunction measured by DTI-ALPS in the AD spectrum after adjusting from multiple covariates and demonstrated association with gray matter volume loss and cognitive decline independent of cerebral amyloid deposition. This suggests that decreased glymphatic flow measured by DTI-ALPS may serve as a marker of neurodegeneration correlated with structural atrophy and cognitive decline in AD, but further investigation is needed to determine its temporal relationship with cerebral amyloid deposition in the AD cascade.

AUTHOR CONTRIBUTIONS

Minjae Kim (Data curation; Formal analysis; Investigation; Methodology; Writing – original draft; Writing – review & editing); Yoo Sung Song (Data curation; Formal analysis; Investigation; Methodology; Resources; Software; Visualization; Writing – original draft; Writing – review & editing); Kyunghwa Han (Formal analysis; Investigation; Methodology; Validation; Visualization; Writing – original draft; Writing – review & editing); Yun Jung Bae, MD, PhD (Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Resources; Software; Supervision; Validation; Writing – original draft; Writing – review & editing); Ji Won Han (Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Supervision; Validation; Writing – original draft; Writing – review & editing); Ki Woong Kim (Formal analysis; Methodology; Validation; Writing – original draft; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

We thank Dr. Inseong Kim (Siemens Healthineers Ltd., Seoul, Republic of Korea) for the DTI-ALPS Analyzer. We also thank Prof. Kyunghwa Han at the Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Yonsei University College of Medicine, for providing statistical analyses.

FUNDING

Dr. Yun Jung Bae has received research support from the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1F1A1064530) and the Seoul National University Bundang Hospital Research Fund (grant No. 02-2023-0011).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

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

References

Iliff

, Wang

, Liao

, Plogg

, Peng

, Gundersen

, Benveniste

, Vates

, Deane

, Goldman

, Nagelhus

, Nedergaard

(2012) A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid β, Sci Transl Med 4, 147ra111.

Iliff

, Lee

, Yu

, Feng

, Logan

, Nedergaard

, Benveniste

(2013) Brain-wide pathway for waste clearance captured by contrast-enhanced MRI. J Clin Invest 123, 1299–1309.

Wang

, Gu

, Masters

, Wang

(2017) A systemic view of Alzheimer disease –insights from amyloid-β metabolism beyond the brain. Nat Rev Neurol 13, 612–623.

Sevigny

, Chiao

, Bussière

, Weinreb

, Williams

, Maier

, Dunstan

, Salloway

, Chen

, Ling

, O’Gorman

, Qian

, Arastu

, Li

, Chollate

, Brennan

, Quintero-Monzon

, Scannevin

, Arnold

, Engber

, Rhodes

, Ferrero

, Hang

, Mikulskis

, Grimm

, Hock

, Nitsch

, Sandrock

(2016) The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease. Nature 537, 50–56.

Ringstad

, Valnes

, Dale

, Pripp

, Vatnehol

S-AS

, Emblem

, Mardal

K-A

, Eide

(2018) Brain-wide glymphatic enhancement and clearance in humans assessed with MRI, JCI Insight 3, e121537.

Eide

, Ringstad

(2015) MRI with intrathecal MRI gadolinium contrast medium administration: A possible method to assess glymphatic function in human brain. Acta Radiol Open 4, 2058460115609635.

Öner

, Barutcu

, Aykol

, Tali

(2017) Intrathecal contrast-enhanced magnetic resonance imaging-related brain signal changes: Residual gadolinium deposition. Invest Radiol 52, 195–197.

Edeklev

, Halvorsen

, Løvland

, Vatnehol

SAS

, Gjertsen

, Nedregaard

, Sletteberg

, Ringstad

, Eide

(2019) Intrathecal use of Gadobutrol for glymphatic MR imaging: Prospective safety study of 100 patients. AJNR Am J Neuroradiol 40, 1257–1264.

Taoka

, Masutani

, Kawai

, Nakane

, Matsuoka

, Yasuno

, Kishimoto

, Naganawa

(2017) Evaluation of glymphatic system activity with the diffusion MR technique: Diffusion tensor image analysis along the perivascular space (DTI-ALPS) in Alzheimer’s disease cases. Jpn J Radiol 35, 172–178.

10.

Zhang

, Zhou

, Wang

, Gong

, Chen

, Zhang

, Cai

, Chen

, Fang

, Sun

, Lou

(2021) Glymphatic clearance function in patients with cerebral small vessel disease. Neuroimage 238, 118257.

11.

Hsu

, Wei

, Toh

, Hsiao

, Lin

, Yen

, Liao

, Ro

(2022) Magnetic resonance images implicate that glymphatic alterations mediate cognitive dysfunction in Alzheimer disease, Ann Neurol 93164–174.

12.

Kamagata

, Andica

, Takabayashi

, Saito

, Taoka

, Nozaki

, Kikuta

, Fujita

, Hagiwara

, Kamiya

, Wada

, Akashi

, Sano

, Nishizawa

, Hori

, Naganawa

, Aoki

(2022) Association of MRI indices of glymphatic system with amyloid deposition and cognition in mild cognitive impairment and Alzheimer disease, Neurology 99, e2648–2660.

13.

Lee

, Lee

, Kim

, Jhoo

, Kim

, Lee

, Kim

, Han

, Woo

(2002) Development of the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet (CERAD-K): Clinical and neuropsychological assessment batteries, J Gerontol B Psychol Sci Soc Sci 57, P47–53.

14.

Lee

, Lee

, Kim

, Jhoo

, Kim

, Yoon

, Woo

, Ha

, Woo

(2004) A normative study of the CERAD neuropsychological assessment battery in the Korean elderly. J Int Neuropsychol Soc 10, 72–81.

15.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack Jr CR , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) 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 7, 263–269.

16.

Bae

, Choi

, Kim

, Choi

, Cho

, Kim

(2021) Altered glymphatic system in idiopathic normal pressure hydrocephalus. Parkinsonism Relat Disord 82, 56–60.

17.

Okudera

, Huang

, Fukusumi

, Nakamura

, Hatazawa

, Uemura

(1999) Micro-angiographical studies of the medullary venous system of the cerebral hemisphere. Neuropathology 19, 93–111.

18.

Taoka

, Fukusumi

, Miyasaka

, Kawai

, Nakane

, Kichikawa

, Naganawa

(2017) Structure of the medullary veins of the cerebral hemisphere and related disorders. RadioGraphics 37, 281–297.

19.

Brewer

, Magda

, Airriess

, Smith

(2009) Fully-automated quantification of regional brain volumes for improved detection of focal atrophy in Alzheimer disease. AJNR Am J Neuroradiol 30, 578–580.

20.

England

, Gillis

, Hampstead

(2014) RBANS memory indices are related to medial temporal lobe volumetrics in healthy older adults and those with mild cognitive impairment. Arch Clin Neuropsychol 29, 322–328.

21.

Barthel

, Gertz

, Dresel

, Peters

, Bartenstein

, Buerger

, Hiemeyer

, Wittemer-Rump

, Seibyl

, Reininger

, Sabri

(2011) Cerebral amyloid-β PET with florbetaben (18F) in patients with Alzheimer’s disease and healthy controls: A multicentre phase 2 diagnostic study. Lancet Neurol 10, 424–435.

22.

Kang

, Kim

, Shin

, Kim

, Choi

, Lee

(2023) Fast and accurate amyloid brain PET quantification without MRI using deep neural networks. J Nucl Med 64, 659–666.

23.

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, 1–15.e11-14

24.

Jack

, Knopman

, Jagust

, Petersen

, Weiner

, Aisen

, Shaw

, Vemuri

, Wiste

, Weigand

, Lesnick

, Pankratz

, Donohue

, Trojanowski

(2013) Tracking pathophysiological processes in Alzheimer’s disease: An updated hypothetical model of dynamic biomarkers. Lancet Neurol 12, 207–216.

25.

Tang

, Zhang

, Liu

, Xue

, Ren

, Huang

, Shi

, Fu

(2022) The association between glymphatic system dysfunction and cognitive impairment in cerebral small vessel disease. Front Aging Neurosci 14, 916633.

26.

Kim

, Hong

, Jeon

(2020) Cerebral small vessel disease and Alzheimer’s disease: A review. Front Neurol 11, 927.

27.

Kivipelto

, Helkala

, Laakso

, Hänninen

, Hallikainen

, Alhainen

, Soininen

, Tuomilehto

, Nissinen

(2001) Midlife vascular risk factors and Alzheimer’s disease in later life: Longitudinal, population based study. BMJ 322, 1447–1451.

28.

Yokota

, Vijayasarathi

, Cekic

, Hirata

, Linetsky

, Ho

, Kim

, Salamon

(2019) Diagnostic performance of glymphatic system evaluation using diffusion tensor imaging in idiopathic normal pressure hydrocephalus and mimickers. Curr Gerontol Geriatr Res 2019, 5675014.

29.

Taoka

, Ito

, Nakamichi

, Kamagata

, Sakai

, Kawai

, Nakane

, Abe

, Ichikawa

, Kikuta

, Aoki

, Naganawa

(2022) Reproducibility of diffusion tensor image analysis along the perivascular space (DTI-ALPS) for evaluating interstitial fluid diffusivity and glymphatic function: CHanges in Alps index on Multiple conditiON acquIsition eXperiment (CHAMONIX) study. Jpn J Radiol 40, 147–158.

30.

Bae

, Kim

J-M

, Choi

, Ryoo

, Song

, Nam

, Yoon

I-Y

, Cho

, Kim

(2023) Altered brain glymphatic flow at diffusion-tensor MRI in rapid eye movement sleep behavior disorder, Radiology 307, e221848.

31.

Saito

, Kamagata

, Andica

, Uchida

, Takabayashi

, Yoshida

, Nakaya

, Tanaka

, Kamio

, Sato

, Nishizawa

, Akashi

, Shimoji

, Wada

, Aoki

(2023) Glymphatic system impairment in corticobasal syndrome: Diffusion tensor image analysis along the perivascular space (DTI-ALPS). Jpn J Radiol 41, 1226–1235.

32.

Morita

, Kamagata

, Andica

, Takabayashi

, Kikuta

, Fujita

, Samoyeau

, Uchida

, Saito

, Tabata

, Naito

, Someya

, Kaga

, Tamura

, Miyata

, Akashi

, Wada

, Taoka

, Naganawa

, Watada

, Kawamori

, Abe

, Aoki

(2023) Glymphatic system impairment in nonathlete older male adults who played contact sports in their youth associated with cognitive decline: A diffusion tensor image analysis along the perivascular space study. Front Neurol 14, 1100736.

33.

, Chang

, Wang

, Lirng

, Wu

, Pan

, Wang

, Chen

(2024) Impaired glymphatic and meningeal lymphatic functions in patients with chronic migraine. Ann Neurol 95, 583–595.

34.

Jiang

, Liu

, KongY , Chen

, Rosa-Neto

, Chen

, Ren

, Chu

, Wu

, Frontotemporal Lobar Degeneration Neuroimaging Initiative (2023) Regional glymphatic abnormality in behavioral variant frontotemporal dementia. Ann Neurol 94, 442–456.

35.

Georgiopoulos

, Tisell

, Holmgren

, Eleftheriou

, Rydja

, Lundin

, Tobieson

(2023) Noninvasive assessment of glymphatic dysfunction in idiopathic normal pressure hydrocephalus with diffusion tensor imaging. J Neurosurg 8, 1–9.

36.

Pearson

, Esiri

, Hiorns

, Wilcock

, Powell

(1985) Anatomical correlates of the distribution of the pathological changes in the neocortex in Alzheimer disease. Proc Natl Acad Sci U S A 82, 4531–4534.

37.

Jack

, Jr. , Petersen

, Xu

, Waring

, O’Brien

, Tangalos

, Smith

, Ivnik

, Kokmen

(1997) Medial temporal atrophy on MRI in normal aging and very mild Alzheimer’s disease. Neurology 49, 786–794.

38.

Desikan

, Cabral

, Fischl

, Guttmann

, Blacker

, Hyman

, Albert

, Killiany

(2009) Temporoparietal MR imaging measures of atrophy in subjects with mild cognitive impairment that predict subsequent diagnosis of Alzheimer disease. AJNR Am J Neuroradiol 30, 532–538.

39.

Zhang

, Li

(2022) Glymphatic system dysfunction in central nervous system diseases and mood disorders. Front Aging Neurosci 14, 873697.

40.

Carotenuto

, Cacciaguerra

, Pagani

, Preziosa

, Filippi

, Rocca

(2021) Glymphatic system impairment in multiple sclerosis: Relation with brain damage and disability. Brain 145, 2785–2795.

41.

Marcus

, Mena

, Subramaniam

(2014) Brain PET in the diagnosis of Alzheimer’s disease, Clin Nucl Med 39, e413–e422.

42.

Kato

, Inui

, Nakamura

, Ito

(2016) Brain fluorodeoxyglucose (FDG) PET in dementia. Ageing Res Rev 30, 73–84.

43.

Fleisher

, Chen

, Liu

, Ayutyanont

, Roontiva

, Thiyyagura

, Protas

, Joshi

, Sabbagh

, Sadowsky

, Sperling

, Clark

, Mintun

, Pontecorvo

, Coleman

, Doraiswamy

, Johnson

, Carpenter

, Skovronsky

, Reiman

(2013) Apolipoprotein E ɛ4 and age effects on florbetapir positron emission tomography in healthy aging and Alzheimer disease. Neurobiol Aging 34, 1–12.

44.

, Su

, Fu

, Wang

, Nguchu

, Qiu

, Dong

, Cheng

(2022) Glymphatic dysfunction correlates with severity of small vessel disease and cognitive impairment in cerebral amyloid angiopathy. Eur J Neurol 29, 2895–2904.

Impaired Glymphatic Flow on Diffusion Tensor MRI as a Marker of Neurodegeneration in Alzheimer’s Disease: Correlation with Gray Matter Volume Loss and Cognitive Decline Independent of Cerebral Amyloid Deposition

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

1 INTRODUCTION

2 MATERIALS AND METHODS

2.1 Study population

2.3 MRI acquisition and analysis

2.4 DTI-ALPS processing and measurement

2.5 Gray matter volumetric analysis

2.6 [18F]Florbetaben PET protocol and analysis

2.7 Statistical analysis

3 RESULTS

3.1 Study participants

4 DISCUSSION

AUTHOR CONTRIBUTIONS

Footnotes

ACKNOWLEDGMENTS

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References

2.6 [¹⁸F]Florbetaben PET protocol and analysis