FDG PET Data is Associated with Cognitive Performance in Patients from a Memory Clinic

Abstract

Background:

Various reasons may lead to cognitive symptoms in elderly, including the development of cognitive decline and dementia. Often, mixed pathologies such as neurodegeneration and cerebrovascular disease co-exist in these patients. Diagnostic work-up commonly includes imaging modalities such as FDG PET, MRI, and CT, each delivering specific information.

Objective:

To study the informative value of neuroimaging-based data supposed to reflect neurodegeneration (FDG PET), cerebral small vessel disease (MRI), and cerebral large vessel atherosclerosis (CT) with regard to cognitive performance in patients presenting to our memory clinic.

Methods:

Non-parametric partial correlations and an ordinal logistic regression model were run to determine relationships between scores for cortical hypometabolism, white matter hyperintensities, calcified plaque burden, and results from Mini-Mental State Examination (MMSE). The final study group consisted of 162 patients (female: 94; MMSE: 6–30).

Results:

Only FDG PET data was linked to and predicted cognitive performance (r(157) = –0.388, p < 0.001). Overall, parameters linked to cerebral small and large vessel disease showed no significant association with cognition. Further findings demonstrated a relationship between white matter hyperintensities and FDG PET data (r(157) = 0.230, p = 0.004).

Conclusion:

Only FDG PET imaging mirrors cognitive performance, presumably due to the examination’s ability to reflect neurodegeneration and vascular dysfunction, thus capturing a broader spectrum of pathologies. This makes the examination a useful imaging-based diagnostic tool in the work-up of patients presenting to a memory clinic. Parameters of vascular dysfunction alone as depicted by conventional MRI and CT are less adequate in such a situation, most likely because they reflect one pathology complex only.

Keywords

Atherosclerosis cognitive performance computed tomography FDG PET magnetic resonance imaging neurodegeneration small vessel disease

INTRODUCTION

Cognitive symptoms in the elderly frequently lead to referrals to a memory clinic for further diagnostic work-up. The spectrum of potential diagnoses in this situation is wide, ranging from the absence of a cognitive disorder or the presence of functional abnormalities to mild cognitive impairment (MCI) and, in the majority of cases, dementia; frequently, mixed diagnoses are reported [1].

With regard to patients with dementia, autopsy studies suggest that mixed pathologies are more common than realized in clinical routine and indeed the prevailing finding, including most commonly Alzheimer’s disease (AD) specific pathology with vascular or other neurodegenerative pathologies such as Lewy bodies [2, 3]. The vascular component is of special interest as based on autopsy studies it accounts for as much as one-third of all dementia cases [4], co-exists in the majority of all major dementias, and lowers the threshold for the clinical syndrome of dementia in neurodegenerative pathologies [3, 5]. It also synergistically interacts with AD pathology [6], increasing the risk of prospective cognitive decline even in asymptomatic older adults [7]. The neuropathology underlying vascular cognitive impairment and dementia is complex; associated factors include macro- and microinfarcts, atherosclerosis, arteriolosclerosis, and hemorrhagic angiopathy [3 , 9].

The diagnostic work-up of patients with cognitive symptoms in general involves, among others, an assessment of cognitive functions and often additionally imaging of the brain. In clinical routine, the most widely used screening test to assess global cognitive function is the Mini-Mental State Examination (MMSE) [10, 11]. Neuroimaging can be divided into modalities providing structural information of the brain such as conventional magnetic resonance imaging (MRI) or computed tomography (CT) and those providing metabolic information, e.g., positron emission tomography (PET) with the radioactive glucose analogue 2-deoxy-2-(¹⁸F)fluoro-D-glucose (¹⁸F-FDG).

MRI is used to exclude potentially treatable causes for cognitive symptoms, to evaluate brain atrophy patterns, and to assess potential vascular lesions within the brain parenchyma [12]. The most frequently found vascular lesions on conventional MRI examinations include white matter hyperintensities (WMHs), recent small subcortical infarcts, or lacunes, among others [13, 14]. These parenchymal alterations are thought to arise from cerebral small vessel pathology and thus are regarded as indicators of vascular damage since direct evaluation of small vessels, in contrast to brain parenchyma, is difficult to impossible with conventional MRI examinations using strength fields of 1.5–3.0 Tesla. Small vessel disease is the most common vascular cause of dementia and a major contributor to mixed dementia [13, 14]. Accordingly, several imaging studies have previously demonstrated that increasing extent of WMHs as seen on MRI is related to worsening of cognitive performance [15, 16].

Although CT is not often used in the elective evaluation of patients with cognitive problems due to its sub-optimal contrast within the brain parenchyma, this technique may provide valuable information about calcifications of larger vessels as a proxy for atherosclerotic plaques/atherosclerosis. Previous research has shown that the amount of CT-quantified calcification volume within extra- and intracranial artery walls is related inversely to cognition [17] as well as positively to a higher future risk for the development of cognitive decline and dementia [18].

PET imaging with FDG displays the regional cerebral glucose consumption. Hypometabolism on FDG PET is supposed to reflect the cumulative loss of neuropil and impaired synaptic/neuronal function [19, 20]. The modality has been established as a useful tool for confirming dementia (including the exclusion of severe neurodegenerative disease in case of a normal FDG PET scan) [21 –23], differentiating between various dementia types based on specific topographic patterns of hypometabolism [21 , 25], and for predicting the risk of future transition from MCI to overt dementia [26]. More recently, FDG PET imaging has been incorporated into the A/T/N classification scheme for AD research as a biomarker of neurodegeneration [20, 27].

The wide range of potential diagnoses, the frequent co-existence of mixed pathologies, and the complex role of cerebrovascular disease make the diagnostic process in patients with cognitive symptoms challenging. In the current study, we aim to explore whether and how the specific information derived from commonly used neuroimaging modalities can be linked to cognition in this population. In particular, we investigate the association of 1) hypometabolism as measured by FDG PET, 2) WMHs derived from conventional MRI, and 3) large vessel calcification as depicted by CT (supposed to reflect neurodegeneration, small vessel disease, and atherosclerosis, respectively) with cognitive performance in a heterogeneous group of patients presenting to our memory clinic. We also study associations between these neuroimaging parameters to gain a deeper insight into the information that these measures in fact reflect.

MATERIALS AND METHODS

All procedures performed in the study involving human participants were in accordance with the ethical standards of the Institutional Review Board and with the 1964 Helsinki declaration and its later amendments.

430 consecutive subjects were recruited at the memory clinic of the Department of Psychiatry, University of Munich. Clinical examinations included cognitive testing, cerebrospinal fluid (CSF) sampling, and brain imaging. Imaging was performed at the Department of Radiology and the Department of Nuclear Medicine and included MRI & ¹⁸F-FDG PET. In cases with remaining uncertainty in the final diagnosis additional ¹⁸F-florbetaben PET imaging was performed after recommendation by the interdisciplinary dementia board.

Retrospective analysis of imaging data was approved by the local ethics committee of the LMU Munich (399-09).

Cognitive testing and CSF testing

All patients received a clinical neurological examination and neuropsychological testing consisting of the MMSE and optionally a complete CERAD plus battery that includes Trail Making Test A and B as well as verbal fluency tests. Years of education were recorded, and laboratory parameters for metabolic causes of dementia (vitamin B12, thiamine and folate levels, thyroid and liver function) were assessed. Lumbar CSF was collected for assessment of phosphorylated tau (threshold: p-tau, <61 pg/ml), total tau (threshold: <450 pg/ml) and Aβ₄₂ (threshold: >450 pg/ml).

¹⁸F-FDG PET imaging

¹⁸F-FDG imaging was performed on a PET/CT system (GE Discovery 690, General Electric, Fairfield, CT, USA) combining a LYSO block detector designed PET and a 64-slice CT scanner.

Emission scans were acquired in a 3D mode. CT scans were acquired using a low dose protocol with 120 kV and 11 mAs, the scan range included the complete skull. PET data were corrected for decay and scatter, and reconstructed iteratively by using the CT images for attenuation correction.

For PET imaging, patients fasted for at least 6 h and had serum glucose levels ≤150 mg/dl. A single intravenous dose of 140±7 MBq ¹⁸F-FDG was administered while patients rested in a room with dimmed light and low noise level. Static emission frames of 5 min each were acquired from 30 to 45 min post injection.

MR imaging

MRI examinations were performed on either a 1.5 T system (Avanto, Siemens Healthineers, Erlangen, Germany) or a 3.0 T system (Ingenia, Philips Medical Systems, Best, the Netherlands). WMH load was determined using conventional fluid attenuated inversion recovery (FLAIR) sequences. For the 1.5 T scanner, 2D axial FLAIR images were obtained using the following parameters: repetition time (TR) = 9060 ms, echo time (TE) = 96 ms, inversion time (TI) = 2500 ms, flip angle (FA) = 180°, field of view (FOV) = 256×173 mm, slice thickness = 5 mm, interslice gap = 2 mm. For the 3.0 T scanner, 3D FLAIR images were obtained with the following parameters: TR = 4500 ms, TE = 322 ms, TI = 1650 ms, FA = 90°, FOV = 224×209 mm, slice thickness = 1.5 mm, interslice gap = 1.7 mm.

2.4 Calculation of PET scores

¹⁸F-FDG PET data was processed with the software Neurostat (Department of Radiology and Imaging Sciences, University of Utah, Salt Lake City, UT, USA) to generate three-dimensional stereotactic surface projections (3D-SSP) and calculate z-score maps (with global mean scaling) against reference images from a group of age-matched healthy controls [28]. Visual based quantification of alterations in the PET data was performed by one experienced reader using a simplified approach of the t-sum method published by Herholz and coworkers [29]. To appreciate the specific topographic patterns in hypometabolism in different neurodegenerative processes, visual ratings were performed for 16 cortical sub-regions separately (for each hemisphere: frontal lateral, frontal medial, parietal, temporal, occipital, central region, posterior cingulate, and cerebellar cortex). The degree of hypometabolism in each sub-region was rated as 0 (absent) to 3 (severe) [30] twice at a 2 weeks interval. Scores were averaged across sub-regions and readings resulting in an FDG_global measure for each patient (where low FDG_global values represent the absence or a low degree of hypometabolism and high FDG_global values reflect severe hypometabolism).

White matter hyperintensities

Only WMHs of presumed vascular origin as defined by the standards for reporting vascular changes on neuroimaging (STRIVE) [13] were included into the study. A visual 4-point grading scale was used to classify the individual global WMH load into four categories: “0” - no lesions (including symmetrical, well-defined caps or bands); “1” - focal lesions; “2” - beginning confluence of lesions; “3” - large confluent lesions, with or without involvement of U-fibers [31].

WMH readings were performed based on axial FLAIR images twice by one experienced reader at a 2 weeks interval. The final individual WMH load (WMH_global) was obtained by averaging the scores across the two readings.

Calcified plaque burden

Using the 3-dimensional CT data from the PET examination vessel wall calcification was assessed within the intracranial segments of the right and left internal carotid artery (ICA) and vertebral artery (VA), as well as within the middle cerebral artery (MCA) and basilar artery (BA). An experienced reader determined the calcified plaque (CP) score semi-quantitatively for each vessel using a 5-point grading scale: “0” - CP absent; “1” - single, small CP, encompassing less than 10 % of the vessel circumference; “2” - CP involving 10–90 % of the vessel circumference; “3” - CP involving 90–270 % of the vessel circumference; “4” - 270–360 % of the vessel circumference involved [32].

Given that no calcifications were identifiable in the MCA and BA in the vast majority of subjects (CP score > 0 in 1 subject, each), both vessels were removed from further analyses.

As with FDG PET and WMH data, CP score readings for each vessel were performed twice at a 2-week interval. To determine the individual total CP burden, scores were averaged across readings and vessels, yielding the final metric CP_global.

Statistical analyses

Intra-rater reliability of imaging parameters was evaluated with a weighted version of Cohen’s kappa coefficient (κ_w) [33].

Non-parametric partial correlations were run to determine the relations of FDG_global, WMH_global and CP_global with MMSE scores whilst controlling for age, gender, and education level. In addition, we tested the relations between the three neuroimaging parameters using the same procedure. A non-parametric approach was chosen since MMSE scores and neuroimaging parameters are ordinal in nature.

In case of significant partial correlations with MMSE, the predictive value of imaging parameters on cognition was evaluated by means of an ordinal logistic regression using a proportional odds model; age, gender, and education level were included as covariates. For this purpose, MMSE scores were classified according to Tombaugh & McIntyre into 3 categories: MMSE 24–30 (no cognitive impairment), MMSE 18–23 (mild cognitive impairment), and MMSE≤17 (severe cognitive impairment) [34]. Similarly, imaging parameters were classified into categories; for this purpose, the given parameter range was divided into 3 equal parts; e.g., in case of FDG_global (range: 0–1.34), values were categorized as follows: ≤0.45, 0.46–0.9, and ≥0.91.

Analyses were performed for all subjects and additionally for those with MMSE scores <24 only.

All statistical data was computed with SPSS, version 25.0 (IBM, Chicago, IL, USA).

Given the explorative nature of the study, results were considered statistically significant in case of p < 0.05.

RESULTS

Sufficient data for enrollment into the final analysis was given for 162 patients (female: 94). Descriptive statistics of the study sample are shown in Table 1. The leading final diagnoses based on all available information included AD (17.3%), depression without dementia (12.3%), a combination of AD and vascular dementia (11.1%), and prodromal AD (8.0%). Mixed diagnoses were reported for 53 patients (32.7%). Vascular dementia was diagnosed in one third of patients, mostly as part of a mixed diagnosis. Final diagnoses are presented in Supplementary Table 1).

Table 1

Descriptive data for MMSE, age, and education level

Variable	Range	Mean±STD
MMSE	6–30 (6–23)	23.46±4.82 (18.92±4.12)
age in years	44–92 (58–92)	74.25±9.02 (76.46±8.85)
years of education	0–20 (0–20)	12.38±5.46 (11.29±5.31)

MMSE, Mini-Mental State Examination. In parentheses: data for sub-group of patients with MMSE scores <24.

Imaging parameters

Agreement between imaging parameters at reading session 1 and 2 was substantial to near-perfect with κ_w-values ranging from 0.75 to 0.95 (in detail, FDG PET: 0.90–0.95; CP: 0.75–0.88; WMH: 0.86). Values for FDG_global, WMH_global, and CP_global are shown in Supplementary Table 2).

Relationship between FDG_global, WMH_global, CP_global, and MMSE scores

All patients

There was a negative partial correlation between FDG_global and MMSE; no significant associations with MMSE were found for WMH_global and CP_global (Table 2A).

Table 2A

Partial correlations between imaging parameters and MMSE

Imaging parameter	MMSE
FDG_global	r(157) = –0.388, p < 0.001
	[r(61) = –0.455, p < 0.001]
WMH_global	r(157) = –0.122 p = 0.127
	[r(61) = –0.263, p = 0.038]
CP_global	r(157) = –0.086, p = 0.280
	[r(61) = –0.133, p = 0.298]

Data are adjusted for age, gender, and education level. In square brackets: data for sub-group of patients with MMSE scores <24.

WMH_global was positively related to CP_global and FDG_global; no association was present between CP_global and FDG_global (Table 2B).

Table 2B

Partial correlations between FDG_global, WMH_global, & CP_global

Imaging parameter	WMH_global	CP_global
FDG_global	r(157) = 0.230, p = 0.004	r(157) = 0.072, p = 0.370
	[r(61) = 0.376, p = 0.002]	[r(61) = –0.022, p = 0.863]
WMH_global	–	r(157) = 0.194, p = 0.014
		[r(61) = 0.218, p = 0.085]

Data are adjusted for age, gender, and education level. In square brackets: data for sub-group of patients with MMSE scores <24.

Results of the ordinal logistic regression model are shown in Table 3. A strong association existed between FDG_global and cognitive performance. Passing from the group with the lowest level of alterations in FDG PET imaging to those groups with moderate and high alterations resulted in an increasing risk of cognitive deficits, with odds ratios of 2.945 and 28.732, respectively.

Table 3

Odds of low MMSE scores by imaging parameters

Independent Variable	p, final model	Wald	df	p	OR	95% CI of OR
All patients
FDG_global
≤0.45 (reference)	<0.001	–	–	–	–	–
0.46–0.9		8.404	1	0.004	2.945	1.419–6.117
≥0.91		26.946	1	<0.001	28.732	8.085–102.003
Sub-group of patients with MMSE < 24
FDG_global
≤0.45 (reference)	0.008	–	–	–	–	–
0.46–0.9		0.941	1	0.332	2.173	0.453–10.412
≥0.91		10.028	1	0.002	21.285	3.209–141.316
WMH_global
≤1.0 (reference)	0.518	–	–	–	–	–
1.1–2.0		0.104	1	0.747	0.246	0.288–5.675
≥2.1		0.843	1	0.358	1.872	0.491–7.149

MMSE, Mini-Mental State Examination; OR, odds ratio; CI, confidence interval. Data are adjusted for age, gender, and education level. Ordinal logistic regressions were calculated in case of significant partial correlations.

Sub-group of patients with MMSE < 24

In the sub-group of patients with mild to severe cognitive impairment (n = 66), negative partial correlations were present between FDG_global and MMSE as well as between WMH_global and MMSE; no significant association was found between CP_global and MMSE (Table 2A).

Inter-parameter analyses demonstrated an association between WMH_global and FDG_global only (Table 2B).

The ordinal logistic regression model still demonstrated a strong association between FDG_global and cognitive performance. Passing from the group with the lowest level of alterations in FDG PET imaging to that with a high degree of alterations resulted in an increasing risk of cognitive deficits, with an odds ratio of 21.285. No significant effect of WMH_global on cognitive performance was present in this analysis (Table 3).

DISCUSSION

In the current study, we evaluated the relation of commonly used imaging-derived proxies for neurodegeneration, cerebral small vessel disease, and cerebral large vessel atherosclerosis with cognitive performance in a heterogeneous group of patients presenting to our university’s memory clinic (MMSE scores 6–30). The results demonstrate that in this population only information derived from FDG PET imaging, generally widely accepted as a biomarker of neurodegeneration, was linked to and predicted cognitive performance. In contrast, parameters linked to cerebral small and large vessel disease, specifically WMHs and atherosclerotic calcified plaques in large intracranial arteries, showed no statistically significant association with cognition when accounting for the patients’ age, gender, and level of education; only in a sub-group of patients with mild to severe cognitive impairment (MMSE < 24), a relatively weak association between WMHs and cognitive performance was detected. Further findings of this study demonstrate a relationship between proxies for small and large vessel disease as well as between small vessel disease and the information captured by FDG PET imaging.

Relationship between FDG_global and MMSE scores

In addition to the established roles of FDG PET brain imaging for the diagnostic work-up of cognitive decline and dementia, our results show that the degree of the brain’s metabolism using a global quantification approach (FDG_global) is highly linked to and predictive of cognitive performance. In detail, compared to patients who lacked or only had minor abnormalities in brain FDG metabolism, those with moderate reductions in metabolism (FDG_global 0.46-0.9) had a 3-fold higher risk of a low MMSE score; in those with highly reduced metabolism (FDG_global ≥0.91), this risk increased by the factor of approx. 29. Similar results were obtained in a sub-group of patients with mild to severe cognitive impairment (MMSE < 24).

The association between globally quantified metabolism in FDG PET brain imaging and cognitive performance has been scarcely investigated so far. But one of the very few available studies also emphasizes the role of global quantification demonstrating that this approach better indicates cognitive performance specifically in AD and MCI patients when compared to amyloid PET [35]. Our results extend this finding in the sense that the information depicted by FDG PET imaging using a global quantification approach is apparently in general a good indicator of cognitive performance, i.e., also when assessing a heterogeneous group of patients where cognitive symptoms may have different underlying pathologies.

Relationship between parameters of vascular pathology and MMSE scores

In contrast to the results of prior studies, no statistically significant relations of imaging-derived proxies for cerebrovascular disease— in our case WMH_global and CP_global— with cognitive performance were identifiable in the current work when assessing all patients. Only in patients with mild to severe cognitive impairment (MMSE < 24) a weak association between WMH_global and cognitive performance was present (r(61) = –0.263, p = 0.038) suggesting that overall the presence of WMHs is unspecific. This is in line with some older reports demonstrating a poor specificity of this parameter given generally the significant numbers of normal elderly with moderate to severe WMHs of uncertain etiology [36, 37]. However, we admit that the poor link between imaging parameters of vascular pathology and cognitive performance may also be related to the small number of subjects studied (thus implying an effect size of vessel pathology on cognition too small to be detected).

It is unlikely that the visual approaches used to grade global WMH load and calcified plaque burden in vessel walls were not sufficiently qualified to detect associations with cognitive performance for two reasons. First, intra-rater agreement between measurements was substantial to near-perfect with κw-values between 0.75 and 0.88, suggesting a high degree of ratings reliability. Second, comparative studies have shown that visual scores as used here generally highly correlate with more sophisticated, quantitative measures [38, 39]. Nonetheless, quantitative approaches could be more sensitive for the detection of correlations given their continuous measures.

Relationship between cerebral small and large vessel disease

Cerebral small and large vessel disease is often investigated separately although the vessels are structurally and functionally connected. This connection also manifested in the positive correlation between WMH_global and CP_global (r(157) = 0.194, p = 0.014) in our study. The causal relationship for this finding is speculative, but it seems obvious that factors assumed to increase the risk for vascular disease, e.g., hypertension, generally act likewise ubiquitously in the vasculature. There also seems to exist a direct effect of pathology in large upstream arteries on small downstream ones, especially in the brain, where small arteries are particularly vulnerable to high-pressure fluctuations from larger afferent vessels [40]; stiffening of these larger vessels, e.g., by atherosclerotic changes, has been suggested to increase the pulsatile stress on the brain’s microvasculature [40] and thus to contribute to small vessel disease associated alterations in brain parenchyma including WMHs [41 –43]. On the other hand, small vessel disease may also directly contribute to large artery stiffness and atherosclerosis by increasing peripheral resistance and blood pressure, eventually leading to a vicious circle of aggravation between macro- and microcirculation pathology [44].

Relation between parameters of vascular pathology and FDG_global

Our data demonstrates an association between WMH_global (but not CP_global) and FDG_global both in the whole group analysis (r(157) = 0.230, p = 0.004) as well as in the sub-group of patients with mild to severe cognitive impairment (r(61) = 0.376, p = 0.002). This is not surprising as previous work has already demonstrated that cerebral vascular lesions, e.g., in patients with vascular dementia and small vessel disease, are also detected by FDG PET imaging, usually demarcating as a widespread pattern of scattered areas of focal cortical and subcortical hypometabolism [45]. Our approach of investigating the FDG uptake on a global scale thus most likely must have additionally captured vascular pathology— presuming the vascular origin of WMHs— and not neurodegeneration only.

In this context, recent literature indicates that FDG PET is not a unique biomarker of neuronal hypometabolism, but also tracks vascular dysfunction, in particular failure of the blood brain barrier transport [46]. But the mechanism of action of FDG PET may be even more complicated as others have claimed that FDG PET signal is in part also driven by glucose utilization by astrocytes [47]. Thus, cerebral FDG hypometabolism seems to be rather a summation of multiple biological processes [48].

The question remains why only information from FDG PET imaging was linked to and predicted cognitive performance in our heterogeneous group of patients. Given the aforementioned, we speculate that this is caused by the method’s ability to capture not only one pathological process (i.e., neurodegeneration) but several of them. Such an “all-in-one” feature of FDG PET imaging will inevitably show pathological test results independent of the underlying pathology— neurodegeneration, vascular dysfunction/small vessel disease, or a mixture of them. Given a link between these pathologies and cognition, results from FDG PET imaging will thus also mirror cognitive performance with high sensitivity and specificity, with increasingly pathological imaging results reflecting a higher degree of cognitive decline. This said, our findings also implicate that parameters of presumed cerebral vascular pathology alone, since they reflect one pathology only, must be inevitably less accurate in reflecting a complex construct as cognition, at least in a heterogeneous group with various underlying pathologies as studied here.

In general, our results underpin the role of FDG PET as a very useful imaging-based tool in the work-up of patients with cognitive symptoms, especially in the beginning of the diagnostic process where various underlying pathologies including mixed diagnoses must be considered. In this scenario, the examination may deliver important information based on the presence of abnormalities (with the absence of any abnormalities implying a low probability of both neurodegeneration and cerebrovascular disease) and their topographic distribution.

Strengths and limitations

The strength of our study includes the assessment of cognitive performance and neuroimaging data supposed to reflect different brain pathologies using one single data set of patients with cognitive symptoms. In this way, between-parameter interactions can be depicted more clearly. Moreover, the group of patients studied was heterogeneous with regard to the underlying pathologies as well as the degree of cognitive decline (MMSE scores 6–30); these characteristics allow the identification of imaging parameters, which reflect the cognitive status under various circumstances with high accuracy.

Relative weaknesses of the current work are also acknowledged. First, MMSE does not test specifically for several cognitive domains, e.g., executive function, thus its use as the only test representing cognition is not optimal. We chose this approach since MMSE results were available in the majority of subjects (in comparison to other test results) thus allowing for robust analyses. Future studies should examine the transferability of our results on other neuropsychological tests that more comprehensively capture global cognitive function. Second, as already mentioned above, our study included a relatively small number of subjects. Moreover, the majority of them had normal MMSE scores (≥24). Another limitation is the method of investigating the association between cognition and large vessel atherosclerosis since calcification as visualized by CT is only part of the atherosclerotic process within the arterial wall. However, prior research has demonstrated that calcification volume is highly correlated with the histologically quantified plaque area and thus is a suitable measure for the total atherosclerotic plaque burden [49]. A final limitation is the use of different scanners as intuitively a 3.0 T MRI system is more sensitive in WMH depiction than a 1.5 T system and this potentially could result in differences in WMH load depending on the scanner used for patient examination. Scientific data on this topic, to the best of our knowledge, is scarce, especially with regard to visual rating systems. Interestingly, one of the very few available studies did not find any differences in the extent of WMHs when comparing a 1.5 T with a 7.0 T scanner [50], suggesting that the same may be true for the comparison between 1.5 T and 3.0 T. And even if differences between scanners in WMH detection exist, we think that a visual rating system as used here with an easy-to-perform attribution of WMHs to four categories is a relatively robust approach and thus the impact of different scanners on the presented results should be of small importance.

Conclusion

Our study shows that across the investigated parameters only information derived from FDG PET imaging was linked to and predicted cognitive performance in a heterogeneous group of patients with a variety of possible causes for cognitive symptoms including mixed diagnoses. The reason for this finding is probably the ability of FDG PET imaging to not only depict neurodegeneration— although this is apparently the main target— but also other pathological processes such as vascular dysfunction/small vessel disease; this makes the examination a very sensitive and specific imaging-based mirror of cognitive performance, independent of the underlying pathology, and underlines its usefulness as a diagnostic tool in the general work-up of patients with cognitive symptoms, e.g., in the setting of a memory clinic.

Parameters of vascular dysfunction as depicted by conventional MRI and CT were less adequate to reflect cognitive performance in such a situation, most likely because they reflect one pathology complex only and thus are insensitive to other pathologies that may come along with cognitive worsening. In case of WMHs, also their lower specificity for vascular pathology may play a role.

Footnotes

ACKNOWLEDGMENTS

Parts of this paper originated from the doctoral thesis of Rebecca Henkel.

No financial and material support was obtained for this research.

Authors’ disclosures available online ().

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

References

McWhirter

, Ritchie

, Stone

, Carson

(2020) Functional cognitive disorders: A systematic review. Lancet Psychiatry 7, 191–207.

Schneider

, Arvanitakis

, Bang

, Bennett

(2007) Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology 69, 2197–2204.

Kapasi

, DeCarli

, Schneider

(2017) Impact of multiple pathologies on the threshold for clinically overt dementia. Acta Neuropathol 134, 171–186.

Power

, Mormino

, Soldan

, James

, Yu

, Armstrong

, Bangen

, Delano-Wood

, Lamar

, Lim

, Nudelman

, Zahodne

, Gross

, Mungas

, Widaman

, Schneider

(2018) Combined neuropathological pathways account for age-related risk of dementia. Ann Neurol 84, 10–22.

Toledo

, Arnold

, Raible

, Brettschneider

, Xie

, Grossman

, Monsell

, Kukull

, Trojanowski

(2013) Contribution of cerebrovascular disease in autopsy confirmed neurodegenerative disease cases in the National Alzheimer’s Coordinating Centre. Brain 136, 2697–2706.

Hachinski

, Einhaupl

, Ganten

, Alladi

, Brayne

, Stephan

BCM

, Sweeney

, Zlokovic

, Iturria-Medina

, Iadecola

, Nishimura

, Schaffer

, Whitehead

, Black

, Ostergaard

, Wardlaw

, Greenberg

, Friberg

, Norrving

, Rowe

, Joanette

, Hacke

, Kuller

, Dichgans

, Endres

, Khachaturian

(2019) Special topic section: Linkages among cerebrovascular, cardiovascular, and cognitive disorders: Preventing dementia by preventing stroke: The Berlin Manifesto. Int J Stroke 15, 961–984.

Rabin

, Schultz

, Hedden

, Viswanathan

, Marshall

, Kilpatrick

, Klein

, Buckley

, Yang

, Properzi

, Rao

, Kirn

, Papp

, Rentz

, Johnson

, Sperling

, Chhatwal

(2018) Interactive associations of vascular risk and beta-amyloid burden with cognitive decline in clinically normal elderly individuals: Findings from the Harvard Aging Brain Study. JAMA Neurol 75, 1124–1131.

Arvanitakis

, Capuano

, Leurgans

, Bennett

, Schneider

(2016) Relation of cerebral vessel disease to Alzheimer’s disease dementia and cognitive function in elderly people: A cross-sectional study. Lancet Neurol 15, 934–943.

Iadecola

, Duering

, Hachinski

, Joutel

, Pendlebury

, Schneider

, Dichgans

(2019) Vascular cognitive impairment and dementia: JACC Scientific Expert Panel. J Am Coll Cardiol 73, 3326–3344.

10.

Hort

, O’Brien

, Gainotti

, Pirttila

, Popescu

, Rektorova

, Sorbi

, Scheltens

, Dementia

ESPo

(2010) EFNS guidelines for the diagnosis and management of Alzheimer’s disease. Eur J Neurol 17, 1236–1248.

11.

Sorbi

, Hort

, Erkinjuntti

, Fladby

, Gainotti

, Gurvit

, Nacmias

, Pasquier

, Popescu

, Rektorova

, Religa

, Rusina

, Rossor

, Schmidt

, Stefanova

, Warren

, Scheltens

, Dementia

ESPo

, Cognitive

(2012) EFNS-ENS Guidelines on the diagnosis and management of disorders associated with dementia. Eur J Neurol 19, 1159–1179.

12.

Filippi

, Agosta

, Barkhof

, Dubois

, Fox

, Frisoni

, Jack

, Johannsen

, Miller

, Nestor

, Scheltens

, Sorbi

, Teipel

, Thompson

, Wahlund

, European Federation of the Neurologic S (2012) EFNS task force: The use of neuroimaging in the diagnosis of dementia.e131-140. Eur J Neurol 19, 1487–1501.

13.

Wardlaw

, Smith

, Biessels

, Cordonnier

, Fazekas

, Frayne

, Lindley

, O’Brien

, Barkhof

, Benavente

, Black

, Brayne

, Breteler

, Chabriat

, Decarli

, de Leeuw

, Doubal

, Duering

, Fox

, Greenberg

, Hachinski

, Kilimann

, Mok

, Oostenbrugge

, Pantoni

, Speck

, Stephan

, Teipel

, Viswanathan

, Werring

, Chen

, Smith

, van Buchem

, Norrving

, Gorelick

, Dichgans

, STandards for ReportIng Vascular changes on nEuroimaging (STRIVE v1) (2013) Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 12, 822–838.

14.

Pantoni

(2010) Cerebral small vessel disease: From pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol 9, 689–701.

15.

de Groot

, de Leeuw

, Oudkerk

, van Gijn

, Hofman

, Jolles

, Breteler

(2000) Cerebral white matter lesions and cognitive function: The Rotterdam Scan Study. Ann Neurol 47, 145–151.

16.

van der Flier

, van Straaten

, Barkhof

, Verdelho

, Madureira

, Pantoni

, Inzitari

, Erkinjuntti

, Crisby

, Waldemar

, Schmidt

, Fazekas

, Scheltens

(2005) Small vessel disease and general cognitive function in nondisabled elderly: The LADIS study. Stroke 36, 2116–2120.

17.

Bos

, Vernooij

, Elias-Smale

, Verhaaren

, Vrooman

, Hofman

, Niessen

, Witteman

, van der Lugt

, Ikram

(2012) Atherosclerotic calcification relates to cognitive function and to brain changes on magnetic resonance imaging.S. Alzheimers Dement 8, 104–111.

18.

Bos

, Vernooij

, de Bruijn

, Koudstaal

, Hofman

, Franco

, van der Lugt

, Ikram

(2015) Atherosclerotic calcification is related to a higher risk of dementia and cognitive decline. Alzheimers Dement 11, 639–647 e631.

19.

Jack

Jr , Knopman

, Jagust

, Shaw

, Aisen

, Weiner

, Petersen

, Trojanowski

(2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol 9, 119–128.

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.

Mosconi

, Tsui

, Herholz

, Pupi

, Drzezga

, Lucignani

, Reiman

, Holthoff

, Kalbe

, Sorbi

, Diehl-Schmid

, Perneczky

, Clerici

, Caselli

, Beuthien-Baumann

, Kurz

, Minoshima

, de Leon

(2008) Multicenter standardized 18F-FDG PET diagnosis of mild cognitive impairment, Alzheimer’s disease, and other dementias. J Nucl Med 49, 390–398.

22.

Laforce

Jr , Buteau

, Paquet

, Verret

, Houde

, Bouchard

(2010) The value of PET in mild cognitive impairment, typical and atypical/unclear dementias: A retrospective memory clinic study. Am J Alzheimers Dis Other Demen 25, 324–332.

23.

Nobili

, Arbizu

, Bouwman

, Drzezga

, Agosta

, Nestor

, Walker

, Boccardi

, EANM-EAN Task Force for the Prescription of FDG-PET for Dementing Neurodegenerative Disorders (2018) European Association of Nuclear Medicine and European Academy of Neurology recommendations for the use of brain (18) F-fluorodeoxyglucose positron emission tomography in neurodegenerative cognitive impairment and dementia: Delphi consensus. Eur J Neurol 25, 1201–1217.

24.

Minoshima

, Foster

, Sima

, Frey

, Albin

, Kuhl

(2001) Alzheimer’s disease versus dementia with Lewy bodies: Cerebral metabolic distinction with autopsy confirmation. Ann Neurol 50, 358–365.

25.

Foster

, Heidebrink

, Clark

, Jagust

, Arnold

, Barbas

, DeCarli

, Turner

, Koeppe

, Higdon

, Minoshima

(2007) FDG-PET improves accuracy in distinguishing frontotemporal dementia and Alzheimer’s disease. Brain 130, 2616–2635.

26.

Herholz

, Nordberg

, Salmon

, Perani

, Kessler

, Mielke

, Halber

, Jelic

, Almkvist

, Collette

, Alberoni

, Kennedy

, Hasselbalch

, Fazio

, Heiss

(1999) Impairment of neocortical metabolism predicts progression in Alzheimer’s disease. Dement Geriatr Cogn Disord 10, 494–504.

27.

Jack

Jr , Bennett

, Blennow

, Carrillo

, Feldman

, Frisoni

, Hampel

, Jagust

, Johnson

, Knopman

, Petersen

, Scheltens

, Sperling

, Dubois

(2016) A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 87, 539–547.

28.

Minoshima

, Frey

, Koeppe

, Foster

, Kuhl

(1995) A diagnostic approach in Alzheimer’s disease using three-dimensional stereotactic surface projections of fluorine-18-FDG PET. J Nucl Med 36, 1238–1248.

29.

Herholz

, Salmon

, Perani

, Baron

, Holthoff

, Frolich

, Schonknecht

, Ito

, Mielke

, Kalbe

, Zundorf

, Delbeuck

, Pelati

, Anchisi

, Fazio

, Kerrouche

, Desgranges

, Eustache

, Beuthien-Baumann

, Menzel

, Schroder

, Kato

, Arahata

, Henze

, Heiss

(2002) Discrimination between Alzheimer dementia and controls by automated analysis of multicenter FDG PET. Neuroimage 17, 302–316.

30.

Beyer

, Schnabel

, Kazmierczak

, Ewers

, Schonecker

, Prix

, Meyer-Wilmes

, Unterrainer

, Catak

, Pogarell

, Perneczky

, Albert

, Bartenstein

, Danek

, Buerger

, Levin

, Rominger

, Brendel

(2019) Neuronal injury biomarkers for assessment of the individual cognitive reserve in clinically suspected Alzheimer’s disease. Neuroimage Clin 24, 101949.

31.

Wahlund

, Barkhof

, Fazekas

, Bronge

, Augustin

, Sjogren

, Wallin

, Ader

, Leys

, Pantoni

, Pasquier

, Erkinjuntti

, Scheltens

, European Task Force on Age-Related White Matter C (2001) A new rating scale for age-related white matter changes applicable to MRI and CT. Stroke 32, 1318–1322.

32.

Strobl

, Kuhlin

, Stahl

, Sabel

, Helck

, Schindler

, Witt

, Bamberg

, Reiser

, Saam

(2018) Intracranial arterial calcifications as a prognostic factor for subsequent major adverse cardiovascular events (MACE). Radiol Med 123, 456–462.

33.

Cohen

(1968) Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 70, 213–220.

34.

Tombaugh

, McIntyre

(1992) The Mini-Mental State Examination: A comprehensive review. J Am Geriatr Soc 40, 922–935.

35.

Khosravi

, Peter

, Wintering

, Serruya

, Shamchi

, Werner

, Alavi

, Newberg

(2019) 18F-FDG is a superior indicator of cognitive performance compared to 18F-Florbetapir in Alzheimer’s disease and mild cognitive impairment evaluation: A global quantitative analysis. J Alzheimers Dis 70, 1197–1207.

36.

Hunt

, Orrison

, Yeo

, Haaland

, Rhyne

, Garry

, Rosenberg

(1989) Clinical significance of MRI white matter lesions in the elderly. Neurology 39, 1470–1474.

37.

Awad

, Spetzler

, Hodak

, Awad

, Carey

(1986) Incidental subcortical lesions identified on magnetic resonance imaging in the elderly. I. Correlation with age and cerebrovascular risk factors. Stroke 17, 1084–1089.

38.

Subedi

, Zishan

, Chappell

, Gregoriades

, Sudlow

, Sellar

, Wardlaw

(2015) Intracranial carotid calcification on cranial computed tomography: Visual scoring methods, semiautomated scores, and volume measurements in patients with stroke. Stroke 46, 2504–2509.

39.

Valdes Hernandez Mdel

, Morris

, Dickie

, Royle

, Munoz Maniega

, Aribisala

, Bastin

, Deary

, Wardlaw

(2013) Close correlation between quantitative and qualitative assessments of white matter lesions. Neuroepidemiology 40, 13–22.

40.

O’Rourke

, Safar

(2005) Relationship between aortic stiffening and microvascular disease in brain and kidney: Cause and logic of therapy. Hypertension 46, 200–204.

41.

Poels

, Zaccai

, Verwoert

, Vernooij

, Hofman

, van der Lugt

, Witteman

, Breteler

, Mattace-Raso

, Ikram

(2012) Arterial stiffness and cerebral small vessel disease: The Rotterdam Scan Study. Stroke 43, 2637–2642.

42.

Webb

, Simoni

, Mazzucco

, Kuker

, Schulz

, Rothwell

(2012) Increased cerebral arterial pulsatility in patients with leukoaraiosis: Arterial stiffness enhances transmission of aortic pulsatility. Stroke 43, 2631–2636.

43.

Aribisala

, Morris

, Eadie

, Thomas

, Gow

, Valdes Hernandez

, Royle

, Bastin

, Starr

, Deary

, Wardlaw

(2014) Blood pressure, internal carotid artery flow parameters, and age-related white matter hyperintensities. Hypertension 63, 1011–1018.

44.

Laurent

, Briet

, Boutouyrie

(2009) Large and small artery cross-talk and recent morbidity-mortality trials in hypertension. Hypertension 54, 388–392.

45.

Heiss

(2018) The additional value of PET in the assessment of cerebral small vessel disease. J Nucl Med 59, 1660–1664.

46.

Sweeney

, Montagne

, Sagare

, Nation

, Schneider

, Chui

, Harrington

, Pa

, Law

, Wang

DJJ

, Jacobs

, Doubal

, Ramirez

, Black

, Nedergaard

, Benveniste

, Dichgans

, Iadecola

, Love

, Bath

, Markus

, Salman

, Allan

, Quinn

, Kalaria

, Werring

, Carare

, Touyz

, Williams

SCR

, Moskowitz

, Katusic

, Lutz

, Lazarov

, Minshall

, Rehman

, Davis

, Wellington

, Gonzalez

, Yuan

, Lockhart

, Hughes

, Chen

CLH

, Sachdev

, O’Brien

, Skoog

, Pantoni

, Gustafson

, Biessels

, Wallin

, Smith

, Mok

, Wong

, Passmore

, Barkof

, Muller

, Breteler

MMB

, Roman

, Hamel

, Seshadri

, Gottesman

, van Buchem

, Arvanitakis

, Schneider

, Drewes

, Hachinski

, Finch

, Toga

, Wardlaw

, Zlokovic

(2019) Vascular dysfunction-The disregarded partner of Alzheimer’s disease. Alzheimers Dement 15, 158–167.

47.

Zimmer

, Parent

, Souza

, Leuzy

, Lecrux

, Kim

, Gauthier

, Pellerin

, Hamel

, Rosa-Neto

(2017) [(18)F]FDG PET signal is driven by astroglial glutamate transport. Nat Neurosci 20, 393–395.

48.

, Xu

, Li

, Guo

, Cui

, Chen

, Huang

, Dong

, Tan

, Yu

, Alzheimer’s Disease Neuroimaging Initiative (2019) FDG-PET as an independent biomarker for Alzheimer’s biological diagnosis: A longitudinal study. Alzheimers Res Ther 11, 57.

49.

Rumberger

, Simons

, Fitzpatrick

, Sheedy

, Schwartz

(1995) Coronary artery calcium area by electron-beam computed tomography and coronary atherosclerotic plaque area. A histopathologic correlative study. Circulation 92, 2157–2162.

50.

Theysohn

, Kraff

, Maderwald

, Barth

, Ladd

, Forsting

, Ladd

, Gizewski

(2011) 7 tesla MRI of microbleeds and white matter lesions as seen in vascular dementia. J Magn Reson Imaging 33, 782–791.

FDG PET Data is Associated with Cognitive Performance in Patients from a Memory Clinic

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

INTRODUCTION

MATERIALS AND METHODS

Cognitive testing and CSF testing

18F-FDG PET imaging

MR imaging

2.4 Calculation of PET scores

White matter hyperintensities

Calcified plaque burden

Statistical analyses

RESULTS

Imaging parameters

Relationship between FDGglobal, WMHglobal, CPglobal, and MMSE scores

All patients

Sub-group of patients with MMSE < 24

DISCUSSION

Relationship between FDGglobal and MMSE scores

Relationship between parameters of vascular pathology and MMSE scores

Relationship between cerebral small and large vessel disease

Relation between parameters of vascular pathology and FDGglobal

Strengths and limitations

Conclusion

Footnotes

ACKNOWLEDGMENTS

References

¹⁸F-FDG PET imaging

Relationship between FDG_global, WMH_global, CP_global, and MMSE scores

Relationship between FDG_global and MMSE scores

Relation between parameters of vascular pathology and FDG_global