Review
The role of hypoperfusion as biomarker for early detection of dementia
Ellen W.S. Carbo, 10416773 - Literature thesis, 09-08-2015Dementia is a growing disease worldwide and it is currently without a cure. The solitary beneficial medical interventions currently available can only delay the progression of the disease. The earlier the patient starts this intervention in the disease, the more beneficial it will be. Therefore, it is becoming essential to detect dementia in its preclinical phase, which is before the onset of clinical symptoms. Numerous studies have shown that patients suffering from dementia show a pattern of regional-specific hypoperfusion of the cerebral blood flow (CBF), which has been related to – and proceeds - cognitive decline. Therefore, imaging the change in CBF is proposed as a potential biomarker to detect oncoming dementia. However, how to use this biomarker to predict dementia precisely has been a source of controversy in the last years, as multiple factors such as ageing, atrophy and heterogeneity of dementia can strongly confound the results. Therefore, this review gives an overview of current knowledge on CBF hypoperfusion in Alzheimer’s disease and vascular dementia, which are the two most occurring dementia types. Additionally, it describes the (dis)advantages of - and differences in - the three most used imaging modalities: positron emission tomography (PET), single-photon enhanced computed tomography (SPECT) and arterial spin labeling magnetic resonance imaging (ASL). By summarizing comparative cross-sectional studies and longitudinal studies, this review discusses the pattern of hypoperfusion (SPECT/ASL) and hypometabolism (PET) found in dementia. The findings support that the hypoperfusion pattern measured specifically by ASL, present before the onset of clinical symptoms, is a potent biomarker to enhance future research and treatment through the early detection of dementia.
VU assessors: Ananya Chakraborty, Nienke de Wit, Prof. H.E. (Elga) de Vries UvA co-assessor: Dr. J.A. (Jan) Gorter
MSc in Brain and Cognitive Sciences, University of Amsterdam, Neuroscience track
Introduction
At least 5% of the current 60+ population suffers from some form of dementia and this percentage is expected to double worldwide every 20 years (World Alzheimer Report, 2009). The two most common types of dementia are Alzheimer’s disease (AD; 50-70% of dementia cases; World Alzheimer Report, 2009) and vascular dementia (VaD, recently termed ‘vascular cognitive impairment’; 20% of dementia cases; World Alzheimer Report, 2009). One of the biggest risk factors of dementia is age, and due to the increasing long-term survival of the population in developing countries, dementia is becoming a major public health concern worldwide. Nonetheless, knowledge of how to prevent or treat dementia is limited. So far, only medication such as cholinesterase inhibitors can be prescribed to delay the progression of AD (Winblad et al., 2006; Wallin et al., 2007). The timing of the start of medication is important, as delay of treatment has a detrimental effect on treatment efficacy. Likewise, for patients suffering from VaD it is possible to start medication early to prevent subsequent vascular problems, e.g. hypertension (Aggarwal & DeCarli, 2007). Therefore, it is important to identify the early stages of dementia
to explore treatment options to prevent the disorder or at least slow its progression. One potential biomarker garnering significant attention is the change in cerebral blood flow (CBF), which can be measured by different imaging modalities. An altered CBF pattern characterized by local hypoperfusion has been found in both AD and VaD patients. Furthermore, these alterations seem to be present in the prodromal stage, which is before cognitive problems arise, indicating its use as a predictive tool. However, currently there is little consensus on how to use the pattern of hypoperfusion as a biomarker.
This study reviews the current literature on CBF hypoperfusion in dementia to increase the understanding of the arrangement of the hypoperfusion pattern, and the time frame in which it surfaces, by focusing on the following two questions:
Question 1: What is the specific CBF hypoperfusion pattern when comparing dementia patients to healthy controls?
The first half of the review provides a detailed description of the pattern of hypoperfusion present in dementia patients by focusing on three different imaging modalities: positron emission tomography (PET), single-photon enhanced computed tomography (SPECT) and arterial spin labeling magnetic resonance imaging (ASL-MRI, further referred to as ASL). It focuses on comparative cross-sectional studies that compare CBF patterns of AD and VaD patient populations to one another and/or to healthy controls, to localize the pattern present in dementia.
Abbreviations
Aβ, amyloid-beta; CBF, cerebral blood flow; AD, Alzheimer’s disease; MCI, mild cognitive impairment; VaD, vascular dementia; MRI, magnetic resonance imaging; ASL, arterial spin
Question 2: Is this pattern present before the onset of clinical symptoms, and does this have predictive value for the course of the disease? To understand the development and the possible predictive value of the pattern, this review will focus on longitudinal studies that followed one group of patients and comparative cross sectional studies that compared patient-groups in different stages of the disease. Additionally, it will discuss hypoperfusion in populations with risk factors for dementia.
Our final aim is to increase the understanding of the localization of hypoperfusion in the development of dementia, and to determine if imaging of hypoperfusion can be used as a potential biomarker for dementia.
Dementia and the cerebral blood flow
Dementia is diagnosed if cognitive abilities decrease beyond what would be expected during normal ageing and are severe enough to impair normal daily functioning (DSM IV). Dementia is more recently termed as ‘minor or major neurocognitive disorder’ (DSM V). It is a collective name for a highly heterogeneous group of brain disorders, of which VaD and AD are the most common. Vascular factors have increasingly been implicated in the onset and progression of dementia (Austin et al., 2011; Zlokovic, 2011; Iadecola, 2013), and the CBF seems to have a large role in its presence.To understand how CBF changes in dementia, this review first describes the general role of the CBF, and what happens when it becomes insufficient. In a healthy individual at rest, the brain uses 15-20% of all oxygen circulating in the vascular system and it consumes the same percentage of glucose daily - a relatively high amount considering that on average our brain only accounts for 2% of our body weight (Clarke & Sokoloff, 1989). Since the brain has very little capacity to store oxygen and nutrients, these necessities are continuously delivered by the CBF (Brown & Ransom, 2007). This is of vital importance, since interruption of the CBF rapidly results in neuronal and glial death (Moskowitz et al., 2010). For example, once blood flow is hampered due to cardiac arrest and insufficient levels of oxygen are present irreversible brain damage occurs within 5 minutes (Cole & Corday, 1956).
The amount of blood distributed through a unit of tissue within a specific timeframe is termed ‘perfusion’. Reduction in perfusion is called ‘hypoperfusion’. People suffering from dementia seem to have a specific pattern of regional hypoperfusion. Additionally, AD is characterized by a global reduced CBF of up to 20% (Postiglione et al., 1993; Roher et al., 2012). Reductions in CBF by a threshold of around 45% will suppress brain activity and consequently also cognitive functioning. Around this threshold, the damage is reversible once normal CBF levels are re-established (Tatemichi et al., 1995; Marshall et al.,
1999; Marshall, 2012). However, a stroke occurs if the damage is permanent. It has been known that suffering from a stroke will double the risk of suffering from dementia later in life (Leys et al., 2005; Pendlebury & Rothwell, 2009; Allan et al., 2011). Additionally, recurrent stroke is a strong predictor of the onset of dementia (Pendlebury & Rothwell, 2009). Also, medical conditions like hypertension and diabetes have also been included as vascular risk factors for dementia (Viswanathan et al., 2009). Therefore, multiple vascular risk factors and vascular dysfunction are known to contribute to the presence and progression of dementia.
Even though multiple associations between dementia and CBF have been proposed, the precise relationship is hard to define. Predominantly, due to the large heterogeneity of vascular lesions and of dementia present. Besides AD, this review will discuss the pathologies of VaD and ‘mixed dementia’, because not all studies on AD patients correct for vascular lesions. One study on over 15,000 patients found that about 17% of VaD patients are initially misdiagnosed as probable AD patients (Kirson et al., 2013). Mixed dementia is diagnosed if both AD and VaD pathologies are present in patients. Occasionally, in other literature ‘mixed’ dementia will be used for cases where pathology or symptoms of other types of dementia are present, e.g. dementia with Lewy bodies. However, within this review ‘mixed’ dementia will consistently always refer to a mix of AD and VaD pathology. It has been suggested that in mixed dementia the presence of cerebrovascular impairments works cumulative or even in synergy with AD pathology to cause cognitive decline (Sinka et al., 2010; Savva et al., 2009). Therefore, understanding the pathology of all three types will benefit the understanding of vascular contributions to dementia, as there is so far no gold standard differentiation. Consequently, some of the AD studies described could be investigating a mixed population. This is not necessary negative, as it does reflect the general dementia population better than just focusing on AD.
Vascular dementia
The specific pathology of VaD is difficult to define on a group level, as it can be caused by many different vascular brain injuries (VBI). Examples of VBI are microbleeds, infarcts, lacunes (cerebrospinal fluid-filled brain cavities) and emboli (small gatherings of circulating solid material). Additionally, symptoms of VaD depend heavily on where the vascular lesion(s) occur and the size of the damage (Gold et al., 2007). A stroke in the frontal lobe for instance, will often be noticed rapidly, as proper functioning of the frontal lobe is critical for normal cognition. Compared to AD, memory disturbances and other cognitive impairments are often initially less prominent in VaD. Instead, vascular lesions may appear in the motor area, resulting in movement disabilities such as gait disturbance (Iadecola, 2013). Therefore, diagnosis and possible treatment of VaD depends
on knowledge of what - and where – the VBI is. Adding to difficulty of diagnosis is the possibility that multiple types of VBI can be present, diversely spatially localized throughout the brain.
Alzheimer’s disease
The precise relationship between the vascular system and AD is harder to define than its connection to VaD. Contributing to this difficulty is that vascular factors haven’t been the main focus in AD research for many years. During the 19th century, it was mostly believed that dementia-like symptoms were caused by ‘hardening’ of the arteries (Binswanger, 1894; Libon et al., 1990; Hardy & Allsop, 1991). Alois Alzheimer was the first to describe the presence of protein accumulations in the brain of a dementia patient, after using a silver impregnation method (Alzheimer, 1907). However, his contribution to the understanding of this disorder was not widely accepted until his mentor, Kraepelin (1910), used the term ‘Alzheimer’s disease’ in his book. Alzheimer detected miliary foci and ‘dense bundles of fibrils’. In the years following, these two pathologies would become known respectively as amyloid-beta (Aβ) plaques and neurofibrially tangles (NFT). Together the presence of brain atrophy, Aβ plaques and NFT are the hallmarks of AD diagnosis (Huang & Mucke, 2012).
Besides the Aβ plaques in the parenchyma, in many patients Aβ depositions are also found in the blood vessel wall (Pantelakis, 1954). This pathology is known as cerebral amyloid angiopathy (CAA) (Vinters, 1987).
Since the discovery of these main pathologies, further research has been accelerated by new scientific methods, such as electron microscopy and imaging. Current research mainly focuses on understanding the contribution of individual pathology to development of the disorder, including the vascular contributions (Tanzi & Bertram, 2005; Jellinger, 2006; Gorelick et al., 2011). At this moment, affirmative diagnosis of dementia types can only be set by biopsy or autopsy.
Mixed dementia
An increasing amount of literature extensively describes the presence of mixed dementia, which occurs when AD pathology and cerebrovascular abnormalities are both present (Launer et al., 2008; Schneider et al., 2009; Iadecola, 2013). More specifically, ‘mixed dementia’ can be seen as a wide spectrum with pure classic AD pathology at one end transitioning into pure VaD at the other end (Viswanathan et al., 2009). Nearly 40% of people suffering from dementia are thought to be positioned in the middle of this spectrum (Pitner & Bachman, 2004; Schneider et al., 2009; Jellinger & Attems, 2010; Jellinger, 2013).
Controversies of potential biomarkers of
dementia
Multiple factors that co-exist and influence one-another play an important role in the progression of dementia, e.g. plaques, tangles, hypoperfusion and atrophy. Therefore, it is difficult to separate cause from consequence in dementia (Mazza et al., 2011; Austin et al., 2011). For example, both AD and vascular pathology appear increasingly during normal ageing, and their presence is not always reflected by cognitive decline (Wen & Sachdev, 2004; Pitner & Bachman, 2004; Gold et al., 2007). Aβ plaques and/or NFT are found in approximately 30% of elderly, who show no signs of cognitive impairment (Knopman et al., 2003). Additionally, atrophy is a symptom of ageing in general, and present in all-ageing individuals, also in those who are cognitively stable (Fox & Schott, 2004). However, the rate of atrophy is different for people suffering from dementia: 0.2-0.5 % annually for healthy controls and 2-3 % for AD patients (Fox et al., 1999a/b; Silbert et al., 2003). Additionally, the rate of atrophy increases before the onset and during the progress of clinical symptoms (Fox et al., 1999a/b; Schott et al., 2003; Silbert et al., 2003). Therefore, even if the presence of atrophy is not necessarily dementia-related, the rate of atrophy does have implications for the presence and progression of the disease. As Aβ plaques, NFT and atrophy can be all present during normal ageing, the coexistence of such pathologies could simply be coincidental (Hachinski, 2011). Furthermore, the presence of cognitive decline seems to be a consequence of multiple factors. Thus, it is becoming increasingly important to find a biomarker to predict dementia better than any of the pathologies individually are capable of. More specifically, it is essential to separate changes related to normal ageing from pathologies that predict impending dementia.
Role of hypoperfusion in dementia
One potential biomarker that has been proposed is the pattern of changes in the CBF. Especially hypoperfusion is suggested to have an ‘umbrella role’, indicating hypoperfusion can be a symptom for multiple pathologies (Gao et al., 2013; Mazza et al., 2011). Regional hypoperfusion is common during ageing (Austin et al., 2011). Therefore, it is important to compare to AD patients to normally cognitive elderly people, to find out what is age-related and what is pathology-age-related. The cause for hypoperfusion can be multifactorial, but can generally be approached from two directions: vascular dysfunction or reduced metabolic activity (Gao et al., 2013; Mazza et al., 2011; Austin et al., 2011). In vascular dysfunction, less blood is able to flow towards the hypoperfused regions due to a change in the vasculature, for instance as a result of stroke or traumatic brain injury (TBI) (Austin et al., 2011). Conversely, reduced metabolism can influence the demand of nutrients and oxygen by cells, for instance due to cell death (Gao et al., 2013; Austin et al., 2011).The presence of hypoperfusion often leads to cognitive impairment. General cerebral perfusion reductions can impair cognition, even temporary for instance during a heart attack (Marshall, 2012). Local hypoperfusion due to vascular dysfunction can for instance been attributed to TBI or atherosclerosis, the narrowing of arteries by accumulation of substance on the artery wall. Both TBI and atherosclerosis have been related to cognitive impairment (Vidal et al., 2010; McKhann et al., 2005). Hypoperfusion has been regionally related to the accumulation of Aβ and to neurofibrillary pathology (Bradley et al., 2002; Driscoll et al., 2011; Sojkova et al., 2008).
The patterns of atrophy and hypoperfusion seen in dementia have been shown to be independent of each other (Chen et al., 2011). For instance, regions with significant tissue atrophy can have normal perfusion. This seems counter-intuitive, and this discrepancy has not been explained so far. It is possible that this dissociation reflects a situation where neuronal death has occurred, but the CBF consistently delivers blood to the region. In a normal setting, perfusion is partially controlled by neurons and astrocytes, as activity consequently increases the demand for nutrients and oxygen. This process is called neurovascular coupling or the hemodynamic response. In this case, the neurovascular connection is unclear, as atrophy decreases the need for CBF. The finding of this dissociation underscores the importance of comparing multiple pathologies or symptoms, as associations might not be intuitive (Chen et al., 2011).
At the moment, neuroimaging is only a supplementary tool to investigate the cause of dementia. However, neuroimaging could be used as a diagnostic and predictive tool. To understand the role of neuroimaging, this review will elaborate next on the different methods to image CBF changes.
Methods for measurement of
reduced CBF
Multiple imaging methods can be used to measure regional CBF (rCBF) in different (in)direct ways. This review will focus on the three currently most often used modalities to localize reduced CBF: PET, SPECT and ASL. In the following sections, it will first elaborate on the three different modalities and their (dis)advantages, before detailing the hypoperfusion findings for each modality.
PET
A PET scan can be used to image selective uptake of molecules by attaching a radioactive tracer to a molecule of choice (Kuhl, 1984). The amount of tracer can be measured, which reflects the presence of the molecule it was attached to. By attaching a tracer to a glucose molecule for instance, one could image metabolic activity. Furthermore, a CT scan is often made to compare
the metabolic image from the PET-scan to the anatomical map. The most commonly used tracer-molecule combination in PET-research on dementia is 18F-fludeoxyglucose (FDG). FDG is a
sugar, analogue to glucose, to which the radioactive tracer 18F is attached. The uptake of 18F-FDG reflects the uptake of glucose by
astrocytes and neurons, and therefore the metabolic activity. During the process of scanning, the 18F-tracer emits positrons. After such a
positron collides with an electron, they annihilate and two gamma photons are emitted in complete opposite directions. The PET scanner measures the gamma photons to determine what levels of
18F are present. As the delivery of glucose is
arranged by the CBF, through neurovascular coupling, the uptake of 18F-FDG also indirectly
reflects the amount of CBF present. However, one always needs to take into account that coupling could be dysfunctional during a disease or in the presence of pathology (Alsop et al., 2000). The
18F-FDG PET scanner is capable of measuring
hypometabolism, and the pattern found can often be matched to the usual hypoperfusion pattern found in SPECT and ASL, even though possible un-coupling due to the pathological disorder should be considered.
SPECT
An imaging modality to measure perfusion directly is SPECT (Leonard et al., 1986), which also requires the addition of an exogenous tracer. The tracer most often used in SPECT-research on
dementia is
technetium-hexamethylpropyleneamine oxime (99m
Tc-HMPAO). The HMPAO molecule, also known under its radiopharmaceutical name exametazime, easily crosses the blood-brain barrier (BBB) and accumulates proportionally to the amount of rCBF (Suess et al., 1992). Attached to HMPAO is the radioactive tracer 99mTc, which emits gamma rays
that can be measured by a SPECT scanner similar to a PET scanner. Consequently, both SPECT and PET measure gamma rays. However, the spatial resolution of PET is higher, because in PET each event has two points of measurement (the positrons) instead of the one point of measurement (gamma ray) in SPECT. Furthermore, considering its ability to differentiate between healthy controls, AD and VaD, 18F-FDG
PET is comparable or even superior to SPECT, as demonstrated in studies comparing both modalities in the same cohorts (Messa et al., 1994; Herholz et al., 2002; Bohnen et al., 2012). Although PET seems to be the optimum choice, SPECT is used more often since it is more broadly available and far less expensive. Besides, solitary clinical diagnosis becomes more accurate by consulting SPECT-images when differentiating between dementias (Dougall et al., 2004). Therefore, even though PET has many advantages, SPECT is often used complementary to clinical examination for an affordable and specific diagnosis.
ASL
The last modality discussed in this review, ASL, does not apply the use of gamma rays, but relies on magnetism using MRI (Detre et al., 1992; Koretsky, 2012 for review history ASL). During an ASL protocol, water below a region of interest is magnetically labeled by applying a 180-degree radiofrequency pulse. Consequently, this water becomes an endogenous contrast tracer, which can be recorded during a MRI scan. The exact same scan will be made afterwards again, but without the magnetization step. This control image will be subtracted from the labeled image, to produce a perfusion image.
There are two main ways to label water, generally classified as continuous ASL (CASL) or pulsed ASL (PASL) (Ferré et al., 2013). In CASL, the arterial water is continuously labeled close to the region of interest in the brain. This leads to a steady supply of magnetized water entering the regions of interest. Conversely, during PASL the labeling takes place during a short pulse, and the water is allowed to flow into the tissue of interest prior to measurement (Ferré et al., 2013). Compared to CASL, PASL has been shown to have lower signal-to-noise ratio, making CASL the optimum choice for the highest amount of clear data (Asllani et al., 2008). Additionally, PASL can only be used to image certain slices, dependent on where the magnetized blood is present at the time. Whereas CASL can be used to cover the whole brain, as the water is continuously labeled, and spreads out more. Especially for research on dementia, the CASL method is beneficial, as some of the main regions of interest – temporal and inferior frontal lobes – are not covered completely by PASL (Asllani et al., 2008). However, currently for research on CBF in dementia both ASL modalities have been used, and will therefore be discussed here.
Compared to PET and SPECT, ASL has some clear benefits: 1) it is completely non-invasive, as it does not require any additional tracer; 2) ASL can easily be added to a standard MRI protocol,
resulting in little extra work and costs for both patient and researcher (Mattson et al., 2014); 3) As ASL is done during the MRI protocol, measuring atrophy and ischemic injury can be done in the same session. This enables more congruent results, as neither the time of measurement or the position of the patient has to change to measure all three aspects; 4) the spatial resolution of ASL is higher than either PET or SPECT (Barkhof et al., 2011). Summarized, ASL is an affordable, less time-consuming, yet more efficient way to image CBF when compared to 18
F-FDG PET and 99mTc-HMPAO SPECT.
Pattern of reduced CBF in dementia
All the three modalities discussed in the previous section are used in multiple experiments to image the pattern of hypometabolism (PET) and hypoperfusion (SPECT/ASL). In this section, the findings of these studies will be described using comparative cross-sectional studies to differentiate between healthy subjects and dementia-patients based on the hypoperfusion pattern.PET
The general consensus in 18F-FDG PET studies is
that AD patients have general global decline, but specifically reductions in the temporoparietal regions, including the precuneus (for elaborate reviews: Minoshima et al., 1997; Patwardhan et al., 2004; Mosconi et al., 2008; Bohnen et al., 2012). Additionally, the posterior cingulate cortex (PCC) is metabolically less active. Both the precuneus and the PCC seem to be affected very early on in AD, whereas the primary visual cortex, cerebellum, thalamus and basal ganglia are relatively spared throughout the whole disease. Research on other dementia subtypes is not as extensive as on AD. Especially studies on VaD have been limited, mostly due to the large diversity of the possible causes. Obviously, for each individual VaD patient hypometabolism will be present in regions damaged by the patient-specific vascular lesions. One study on a subtype of VaD
Figure 1. Direct comparison between 18F-FDG PET and 99mTc-HMPAO SPECT through scans of one AD
patient. The scans show good correspondence. Note the congruency in the temporoparietal cortex (white arrows), which is one of the main regions influenced in the AD specific pattern (adapted from Herzholz et al., 2002).
with mainly subcortical lesions, called subcortical ischemic vascular dementia (SIVaD), found that compared to AD, SIVaD patients have reduced metabolism mainly in the deep gray nuclei, cerebellum, primary cortices, middle temporal gyrus, and anterior cingulate gyrus (Kerrouche et al., 2006). Regions such as the cerebellum were found relatively intact in AD patients. Comparatively, reduced metabolism in AD versus SIVaD occurred mainly in the hippocampal region and orbitofrontal, PCC, and posterior parietal cortices (Kerrouche et al., 2006). Both AD and VaD patients showed congruent hypometabolism compared to healthy controls in posterior parietal, prefrontal and anterior hippocampal regions, as well as in the PCC and the precuneus. Additionally, hypometabolism in these regions was for both groups linearly correlated with cognitive ability (Kerrouche et al., 2006).
SPECT
99mTc-HMPAO SPECT shows decreased CBF for
AD patients in bilateral temporal and parietal lobes, including the precuneus and PCC (for complete review: Henderson, 2012). The patterns detected by PET and SPECT are very similar, which indicates that perfusion and metabolism are robustly coupled, despite the dementia pathology present (Messa et al., 1994; Herzholz et al., 2002; figure 1). Especially the detected changes in the temporoparietal cortex and PCC are very comparable. However, findings in regions such as the frontal, temporobasal, and temporomesial cortices and in the cerebellum can be incongruent, mainly due to imaging specific variables such as attenuation and scatter correction (Herholz et al., 2002). For VaD patients the patterns are highly variable, due to its heterogeneous origin. However, there is often more anterior hypoperfusion in VaD than in AD, which could confound the results for AD populations that haven’t been checked for vascular lesions (Schuff et al., 2009; Henderson, 2012). Additionally, perfusion of subcortical areas in VaD shows more prominent defects when compared to AD (Henderson, 2012).
ASL
The two different approaches for ASL, CASL (Asllani et al., 2008) and PASL (Du et al., 2006), show great potential to separate AD patients and healthy controls in comparative cross-sectional studies based on the CBF pattern (supplementary table 1, p. 14; for complete review: Alsop et al., 2010). A study with CASL that found these results excluded patients with a history of clinical stroke or any type of imaging evidence for infarcts, suggesting that they investigated a ‘pure’ AD population, versus a mixed dementia population (Asllani et al., 2008). Using CASL, hypoperfusion was found in the temporal, parietal, frontal, and PCC in AD group when compared to healthy controls (Alsop et al., 2000).
The findings of reduced rCBF in temporal-parietal cortices and the PCC are mostly (Johnson et al., 2005), but not always (Wolk & Detre, 2012)
congruent with earlier studies using PET or SPECT (supplementary table 1, p. 14). Even though some results are conflicting, overall most studies with ASL show good agreement with FDG PET and HMPAO SPECT studies, even after accounting for underlying cortical gray matter atrophy (figure 2; Johnson et al., 2005; Asllani et al., 2008). One example of such an explorative study, using 18 controls and 13 AD patients, compared both 18F-FDG PET and CASL-MRI
modalities on different aspects (Musiek et al., 2012). The pattern of CBF was very similar upon visual inspection by two separate experts (Musiek et al., 2012). As ASL and PET show overlap in regions of respectively hypoperfusion and hypometabolism, this support the idea that direct coupling between these two effects is intact in AD (Herholz et al., 2002; Asllani et al., 2008).
Conversely, findings in the frontal lobe seem to be conflicting even within each individual modality (Schuff et al., 2009; Musiek et al., 2012; Raji et al., 2010). The most forward explanation for this discrepancy is that the rCBF in the frontal lobe depends on the presence of common vascular lesions in VaD. Therefore, the heterogeneity that can be found within a patient group would influence these results greatly. Additionally, many of these studies base the subdivision of patients on clinical diagnosis, and have no post-mortem confirmation. Thus, some patients of the proposed AD group could actually have mixed dementia confounding the results.
Summary on imaging modalities on CBF in dementia
Considering the three modalities ASL seems to show more promise than either PET or SPECT, due to the main benefits: better spatial resolution, less time-consuming and more affordable. All three of the above-mentioned modalities show a characteristic AD rCBF pattern: reduced rCBF in the temporal and parietal lobes, specifically in the precuneus and PCC. Additionally, this reduction seems to correlate to deteriorating cognition. Even though the pattern for VaD is very heterogeneous, the average hypoperfusion in VaD seems to be focused towards the frontal lobes more than found in AD (Schuff et al., 2009).
The predictive role of reduced CBF in
dementia
All studies described so far, focused on the difference in rCBF to differentiate between healthy, cognitively normal controls and people suffering from different dementias. Even though this is important as a baseline, for clinical purposes it is essential to know how the CBF changes during the years as early diagnosis of which dementia-subtype the patient is suffering from is important for all parties involved. Early diagnosis is especially essential as each subtype follows its own progression course. Furthermore, there is a growing body of evidence that the onset
of pathophysiological processes and metabolic changes precedes clinical symptoms by a long time (Cerami et al., 2015). Therefore, early imaging could be increasingly successful in recognizing oncoming dementia in a prodromal stage.
Initially, the diagnosis of dementia just relies on cognitive testing by a clinical neuropsychologist or observations in a clinical setting. However, clinical diagnosis alone is often unable to give a correct diagnosis (Henderson, 2012). To be able to give a definite diagnosis of any subtype of dementia, up until today pathological confirmation by biopsy or autopsy is always needed. To improve diagnosis ante-mortem, researchers have been trying different imaging methods since as early as 1948. The most recent and/or influential studies will be described below.
Longitudinal studies
Multiple longitudinal studies followed patients prior to clinical diagnosis of dementia to decipher if their CBF pattern was predictive of later dementia (Yoshikawa et al., 2003). Studies in PET and SPECT support the hypothesis that the characteristic pattern of hypoperfusion in the bilateral parietal and temporal cortices is present already in preclinical stages of AD (Jagust, 2000; Foster et al., 1983; Haxby et al., 1990; Jagust et al., 1988; Holman et al., 1992; Beason-Held et al., 2013), as do studies using ASL (Alsop et al., 2000, 2010; Johnson et al., 2005; Xu et al., 2007; Yoshiura et al., 2009; Dai et al., 2009; Chao et al., 2009, 2010; Hu et al., 2010; Chen et al., 2011; Alexopoulos et al., 2012; Wolk & Detre, 2012; Wang et al., 2013). A study using PET showed that specifically the PCC has decreased glucose metabolism at very early stage AD, before any other region (Maddock et al., 2001). This is of special interest, as the PCC has been associated
with memory and learning. Furthermore, the PCC has even been implicated in autobiographical memories, such as remembering friends and family (Maddock et al., 2001), which is often impaired in people suffering from AD. Thus, the pattern found all three modalities shows promise to investigate prodromal AD, with the PCC as an extra region of interest.
Mild cognitively impaired
The next studies discussed are not longitudinal in the traditional sense, namely following subjects over multiple years, but they do compare the potential pre-dementia stage of mild cognitively impaired patients (MCI) to AD patients and healthy controls. Therefore, these studies do attribute to the understanding of the development of dementia. MCI patients show decreased cognitive ability, but not bad enough to be categorized as dementia as it does not affect their daily functioning sufficiently. Previously, a meta-analysis of 41 cohorts showed that about 40% of people with MCI develop dementia later in life (Mitchell & Shiri-Feshki, 2009).
Multiple studies with ASL found that the CBF pattern for MCI patients seems to be intermediate between healthy controls and AD patients; this suggests that there is a slope both cognitively and pathologically (Dai et al., 2009; Binnenwijzend et al., 2013, figure 3). One PASL study investigated two separate MCI groups, divided by which cognitive domain was impaired (Chao et al., 2009). MCI patients with memory deficits had hypoperfusion in the bilateral PCC matching with the general MCI pattern in SPECT, PET and ASL. Conversely, MCI patients with deficits in the domain of executive functioning showed the same pattern, but additional hypoperfusion in the left middle frontal cortex as well as the left precuneus. These results support the idea that MCI hypoperfusion is an intermediate form of AD hypoperfusion. However, it also indicates that researchers need to be very cautious when including subjects, as such heterogeneous populations can easily confound results.
The risk factor ApoE4 mutation
Another way to predict impending dementia is to focus on patients who are genetically at risk to develop this disease. A well-known genetic risk factor for AD is a mutation of the apolipoprotein E gene (ApoE) on chromosome 19, resulting in the presence of the e4 allele (ApoE4). People with one copy of ApoE4 are 4 times more likely to develop dementia, while people carrying two copies are 9 or even 12 times more likely to develop it (Carlsson et al., 2009; Kim et al., 2009). Additionally, having two ApoE4 copies has been associated with an earlier onset of dementia (Kim et al., 2009). However, not all ApoE4 carriers develop dementia, nor do all dementia patients have an ApoE4 copy. Therefore, even though the presence of ApoE4 is a major risk factor, it is neither necessary nor sufficient for dementia. Figure 2. Regional hypoperfusion found with PASL
in an AD group (n = 20) compared to a group of cognitively normal control subjects (n = 23). All colored voxels were significantly different (p < 0.05), and yellow regions showed the greatest group difference (adapted from Johnson et al., 2005).
An FDG PET experiment in MCI patients with and without ApoE4 mutation showed hypometabolism in AD-typical regions, mainly the temporoparietal cortex and PCC (Mosconi et al., 2004). However, the ApoE4-MCI group also showed hypometabolism in frontal areas such as the anterior cingulate cortex (ACC) and inferior frontal cortex (IFC). The amount of hypoperfusion in the pattern could predict future AD with an 84% accuracy (p < 0.003). Moreover, prediction for the ApoE4-MCI group showed even more improved diagnostic value: 100% sensitivity, 90% specificity, and 94% accuracy (p < 0.0005) (Mosconi et al., 2004). This result indicates that the hypometabolism pattern has an important role in the prediction of dementia, especially when taking risk factors into account. Additionally, CBF reduction may occur prior to atrophy and AD pathology in ApoE4 carriers, indicating that CBF is in this group a promising biomarker (Winkler et al., 2014).
Discussion
The main finding of this literature review is that hypoperfusion is a potential biomarker for impending dementia.
Patients with AD show a characteristic pattern of hypoperfusion in the bilateral temporoparietal lobes, including the PCC and precuneus, and this pattern is already present before the onset of cognitive problems. Additionally, the pattern can be found in patients with high risk of developing dementia, such as people with the ApoE4 mutation and MCI patients. To image this pattern both PET and SPECT can be used, respectively reflecting hypometabolism and hypoperfusion. However, the most promising modality seems to be ASL, which is another approach to measuring hypoperfusion. ASL is proposed as the optimal technique as it finds the same pattern as PET and SPECT, but it is cheaper, non-invasive and less time consuming.
Early detection of oncoming dementia is important, as treatment needs to be started as early as possible to have the largest beneficial effect. Additionally, to improve future research on dementia, it is beneficial to be able to predict who will develop dementia. Other possible biomarkers than hypoperfusion for oncoming dementia have so far been insufficient. Aβ plaques, NFT and atrophy – the main pathologies of AD – can all be present in various amounts during normal ageing, and are often not reflected by cognitive impairment (Wen & Sachdev, 2004; Pitner & Bachman, 2004; Gold et al., 2007, Fox & Schott, 2004). Therefore, the use of these pathologies as biomarkers can be questioned. Even though the presence of atrophy has shown little potential as a biomarker, the rate of atrophy occurrence does have predictive value for future dementia (Fox et al., 1999a/b; Schott et al., 2003; Silbert et al., 2003, Dickerson et al., 2011). However, multiple annual MRI scans are needed to calculate if atrophy rate is higher than expected from normal ageing, and therefore possible indication of future dementia. Conversely,
imaging the hypoperfusion pattern can be done in a singular session. Furthermore, hypoperfusion can be imaged before discernable atrophy is present (Mendez et al., 2007). Therefore, the hypoperfusion pattern has more potential as a biomarker than the increased rate of atrophy.
Limitations
Research on the use of hypoperfusion as a biomarker is limited by a number of factors. First of all, since affirmative AD diagnosis can only be confirmed post-mortem, it is possible a mixture of dementia types is present in the patient cohorts used in the ‘AD’-studies on the hypoperfusion pattern. To establish if the diagnosis of the subtype of dementia set during the acquisition of the data was correct, patients in cohorts need to be diagnosed post-mortem. Misdiagnosis of dementia type during life happens regularly (Kirson et al. 2013). The most occurring faulty diagnosis during life is when people who suffer from VaD or mixed dementia are instead diagnosed with AD (Pitner & Bachman, 2004; Schneider et al., 2009; Jellinger & Attems, 2010; Kirson et al., 2013). The vascular impairments that these patients suffer from influence the pattern when scanning for hypoperfusion. Therefore, the misdiagnosis of probable ‘pure’ AD can confound the results found in different cohorts. By using post-mortem diagnosis, this limitation can be controlled for.
Secondly, it is important to realize that global hypoperfusion is also present during normal ageing (Chen et al. 2011). On average, the whole-cortex CBF is reduced about 0.38% each year as seen with PASL (Chen et al., 2011). Additionally, a CASL study found that solitary gray-matter decrease by 0.45% each year (Parkes et al., 2004). The parietal, especially the precuneus, and temporal lobes – important regions within the specific AD pattern – are also affected by normal ageing (Chen et al., 2011). However, studies on the hypoperfusion pattern have taken this into account, and their healthy control groups do not significantly differ on age (Alsop et al., 2000, 2010; Johnson et al., 2005; Xu et al., 2007; Yoshiura et al., 2009; Dai et al., 2009; Chao et al., 2009, 2010; Hu et al., 2010; Chen et al., 2011; Alexopoulos et al., 2012; Wolk & Detre, 2012; Wang et al., 2013). Therefore, these studies show that the amount of hypoperfusion in the specific AD pattern is predictive of AD, despite ageing contributions.
Thirdly, the inevitable presence of atrophy during ageing has consequences for the accuracy of CBF imaging through the ‘partial volume effect’ (PVE) (Thomas et al., 2011). The PVE is the loss of apparent activity in structures, due to limited spatial resolution. The degree of PVE is dependent on the size of the structure. Smaller structures such as atrophied regions tend to be more affected. Correcting for PVE reduces the average hypoperfusion reduction from 20-30% (Parkes et al., 2004; Restom et al., 2007) to 15% (Asllani et al., 2008). Even though the intensity of
the reduction is less, the pattern of hypoperfusion doesn’t change significantly (Binnenwijzend et al., 2013; figure 3). Therefore, it is recommended to correct for PVE, but earlier research uncorrected for atrophy on the hypoperfusion pattern can still be taken into account.
Future perspectives
For the individual patient the predictive value of the hypoperfusion biomarker is so far not well known. However, for future development, personalized medicine is starting to play a big role, and it is important to consider the individualization of diagnosis and treatment. Essentially, most past research on hypoperfusion patterns was done on group level, to understand the general development of dementia. Progress for the diagnosis of the individual patient has been made by considering risk factors, such as the presence of ApoE4, stroke or MCI. These risk factors are neither necessary nor sufficient for the presence of dementia. However, their presence does relate to a higher occurrence of this disorder. Therefore, considering research on the predictive value of the hypoperfusion pattern in individuals, the risk factors need to be taken into account. One multi-risk factor study already investigated the added value of testing for ApoE4 in a MCI cohort, to see if controlling for this genetic risk factor increases the sensitivity (Mosconi et al., 2004). Indeed, the accuracy of prediction of future dementia in a MCI cohort was increased when ApoE4 was taken into account (Mosconi et al., 2004). Considering these results, it is recommended to consider risk factors
in future research when analyzing cohorts, as the presence of risk factors could attribute to a more specialized individual diagnosis.
Another future approach to understand the individual pattern is to investigate the regional progression with which this characteristic hypoperfusion AD pattern develops. Possibly, the specific AD-pattern is an end stage for multiple regional pathways of hypoperfusion. One study already showed that two groups of patients suffering from MCI in different cognitive domains have slightly different hypoperfusion patterns. MCI patients with memory deficits showed hypoperfusion in the bilateral PCC, while MCI patients had besides hypoperfusion in the bilateral PCC additional reduction of the CBF in the left frontal cortex and left precuneus. As there is a high chance for all MCI patients to progress into AD patients, this finding implies that there are multiple regional hypoperfusion pathways to get to the AD specific pattern (Mitchell & Shiri-Feshki, 2009). If understanding of the progression of the pattern can be increased, it could be possible to detect dementia even before the complete pattern is present. Potentially, this could even be used for diagnosis in individual cases.
The next step to understanding the hypoperfusion pattern would be to consider why this specific pattern emerges. As described previously in this review, hypoperfusion can result from multiple causes, all of which can be divided into two main categories (Gao et al., 2013; Mazza et al., 2011; Figure 3. Reduced CBF maps measured by ASL for an uncorrected hypoperfusion measurement and PVC-corrected cortical CBF. Persons whose cognitive problems could not be confirmed in a clinical setting or by psychological testing were placed in the subjective complains group (n = 70). AD patients (n=71) showed the greatest decrease in CBF, while the hypoperfusion of the MCI group (n=31) was in between the hypoperfusion of the AD and SC group. Correcting for PVC reduced global hypoperfusion, but did not significantly changed the pattern (Adapted from Binnenwijzend et al., 2013).
Austin et al., 2011). On the one hand, vascular dysfunction can result in less blood being physically able to get to the hypoperfused tissue. On the other hand, local cell death can result into less blood being requested as metabolic need decreases. Since it is still unknown precisely what causes the hypoperfusion in AD, either or even both of these causes could be responsible. Theoretically, understanding what drives the hypoperfusion could lead to comprehension why some regions are affected more than others. Therefore, for future research it is essential to investigate how regional vascular and cellular components contribute to the emergence of the AD specific hypoperfusion pattern. Conversely, this caveat does not influence the use of hypoperfusion as a biomarker for dementia, as its predictive value is independent of understanding its precise cause.
Conclusion
This review supports the view that the AD-specific hypoperfusion pattern measured with ASL is a potential biomarker to detect future dementia. Two main benefits result from early detection. First, for the individual patient medication can be started as soon as possible, which is related to a more beneficial effect. Secondly, research on dementia can be further aided by better diagnosis and prediction of who will develop dementia. Ideally, dementia can be detected before onset of clinical symptoms. This would improve overall understanding of how this disorder develops, and maybe even aid in finding a way to prevent or at least delay progression of dementia.
Acknowledgements
Foremost, I would like to express my gratitude to my main supervisor Ananya Chakraborty for the time and effort she took to help me during the whole process of writing this thesis. Her never-tiring input and ideas were great motivation to deliver a good final product. Besides Ananya, I would like to thank Nienke de Wit for her help and guidance in the first weeks, including her help with setting up the main story lines. I am grateful to both Ananya and Nienke for their advice to help me improve my scientific writing skills in just a short time period. My sincere thanks also goes to Elga de Vries, for the initial approval and help on picking a topic. Lastly, I would like to thank my co-assessor Jan Gorter, who was willing to continue to provide support from the University of Amsterdam despite a drastic subject change.
References
Aggarwal NT, DeCarli C. Vascular dementia: emerging trends. Semin Neurol 2007, 27: 66-77.
Alexopoulos P, Sorg C, Forschler A, et al. Perfusion abnormalities in mild cognitive impairment and mild dementia in Alzheimer's disease measured by pulsed arterial spin labeling MRI. Eur Arch Psychiatry Clin
Neurosci 2012, 262: 69–77.
Allan LM, Rowan EN, Firbank MJ, Thomas AJ, Parry SW, Polvikoski TM, O’Brien JT, and Kalaria RN Long term
incidence of dementia, predictors of mortality and pathological diagnosis in older stroke survivors. Brain 2011, 134, 3716–3727.
Alsop DC, Dai W, Grossman M, Detre J. Arterial spin labeling blood flow MRI: Its Role in the early
characterization of Alzheimer’s disease. J Alzheimers Dis 2010, 20: 871–880.
Alsop DC, Detre J, Grossman M. Assessment of cerebral blood flow in Alzheimer’s disease by spin-labeled magnetic resonance imaging. Ann Neurol 2000, 47: 93–100. Alzheimer A. About a peculiar disease of the cerebral cortex [in German]. Centralblatt fur Nervenheilkunde und
Psychiatrie 1907, 30: 177–179.
Asllani I, Habeck C, et al. Multivariate and Univariate Analysis of Continuous Arterial Spin Labeling Perfusion MRI in Alzheimer’s Disease Iris. Cereb Blood Flow Metab 2008, 28: 725–736.
Austin BP, Nair VA, Meier TB, Xu, G, Rowley HA, Carlsson CM, Johnson SC, Prabhakaran V. Effects of
Hypoperfusion in Alzheimer ’ s Disease. J Alzheimers Dis 2011, 26: 123–133.
Barkhof F, Fox NC, Bastos-Leite AJ, Scheltens P. Neuroimaging in Dementia. Springer Science & Business Media, 2011, p 33.
Beason-Held LL, Goh JO, An Y, et al. Changes in brain function occur years before the onset of cognitive impairment. J Neurosci 2013, 33: 18008–18014.
Binnewijzend M, Kuijer JP, Benedictus MR, et al. Cerebral blood flow measured with 3D pseudocontinuous arterial spin-labeling MR imaging in Alzheimer disease and mild cognitive impairment: a marker for disease severity.
Radiology 2013, 267: 221–230.
Binswanger O. Defining the general progressive paralysis [in German]. Klin Wochenschr 1894, 49: 1103–1105. Bohnen NI, Djang DSW, Herholz K, Anzai Y, Minoshima S. Effectiveness and Safety of 18F-FDG PET in the
Evaluation of Dementia: A Review of the Recent Literature.
J Nucl Med 2012, 53: 59–71.
Bradley KM, O’Sullivan VT, Soper ND, et al. Cerebral perfusion SPET correlated with Braak pathological stage in Alzheimer’s disease. Brain 2002, 125: 1772–1781. Brown AM, Ransom BR. Astrocyte Glycogen and Brain Energy Metabolism. Glia 2007, 55: 1263–1271.
Carlsson CM, Gleason CE, Puglielli L, Asthana S. Chapter 65. Dementia Including Alzheimer’s Disease. Hazzard’s Geriatric Medicine and Gerontology. McGraw-Hill Medical; New York: 2009, 797-811.
Cerami C, Della Rosa PA, Magnani G, et al. Brain metabolic maps in Mild Cognitive Impairment predict heterogeneity of progression to dementia. NeuroImage
Clin 2015, 7: 187–194.
Chao LL, Buckley ST, Kornak J, et al. ASL perfusion MRI predicts cognitive decline and conversion from MCI to dementia Alzheimer Dis Assoc Disord 2010, 24: 19–27. Chao LL, Pa J, Duarte A, et al. Patterns of cerebral hypoperfusion in amnestic and dysexecutive MCI.
Chen JJ, Rosas HD, Salat DH. Age-Associated
Reductions in Cerebral Blood Flow Are Independent from Regional Atrophy. Neuroimage 2011, 55: 468–478. Clarke DD, Sokoloff L. Circulation and energy metabolism of the brain. In: Basic Neurochemistry. New York: Raven Press, 1989: pp. 565–590.
Cole SL, Corday E. Four-minute limit for cardiac resuscitation. J Am Med Assoc 1956, 161:1454–1458. Dai W, Lopez OL, Carmichael OT, Becker JT, Kuller LH, Gach HM. Mild cognitive impairment and alzheimer disease: patterns of altered cerebral blood flow at MR imaging. Radiology 2009, 250: 856–866.
Detre JA, Leigh JS, Williams DS, Koretsky AP. Perfusion imaging. Magn Reson Med 1992, 23: 37–45.
Dickerson B, Albert M, Blacker D, et al. Alzheimer’s-signature MRI biomarker predicts likelihood of Alzheimer's dementia in cognitively normal adults. Alzheimers Dement 2011, 7: S109.
Dougall NJ, Bruggink S, Ebmeier KP. Systematic review of the diagnostic accuracy of 99mTc-HMPAO-SPECT in dementia. Am J Geriatr Psychiatry 2004, 12: 554–70. Driscoll I, Zhou Y, An Y, Sojkova J, Davatzikos C, Kraut M A, Ferrucci L, Mathis CA, Klunk WE, Wong DF, Resnick SM. Lack of association between 11C-PiB and longitudinal brain atrophy in non-demented older individuals. Neurobiol
Aging 2011, 32: 2123–2130.
DSM IV. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. Washington, DC: American Psychiatric Association, 1994 DSM V. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. Arlington, VA: American Psychiatric Association, 2013 Du AT, Jahng GH, Hayasaka S, Kramer JH, Rosen HJ, Gorno-Tempini ML, Rankin KP, Miller BL, Weiner MW, Schuff N. Hypoperfusion in frontotemporal dementia and Alzheimer disease by arterial spin labeling MRI. Neurology 2006, 67: 1215–1220.
Ferré J-C, Bannier E, Raoult H, Mineur G, Carsin-Nicol B, Gauvrit J-Y. Arterial spin labeling (ASL) perfusion: techniques and clinical use. Diagnostic and Interventional
Imaging 2013, 94: 1211–23.
Foster NL, Chase TN, Fedio P, et al. Alzheimer’s disease: focal cortical changes shown by positron emission tomography. Neurology 1983, 33: 961-965.
Fox NC, Scahill RI, Crum WR, Rossor MN. Correlation between rates of brain atrophy and cognitive decline in AD.
Neurology 1999a, 52: 1687–1689.
Fox NC, Warrington EK, Rossor MN. Serial magnetic resonance imaging of cerebral atrophy in preclinical Alzheimer’s disease. Lancet 1999b, 353: 2125. Fox NC, Schott JM. Imaging cerebral atrophy: Normal ageing to Alzheimer’s disease. Lancet 2004, 363: 392– 394.
Gao Y, Zhang J, Liu H, Wu G, Xiong L, Shu M. Regional Cerebral Blood Flow and Cerebrovascular Reactivity in Alzheimer’s Disease and Vascular Dementia Assessed by Arterial Spin- labeling Magnetic Resonance Imaging. Curr
Neurovasc Res 2013, 10: 49–53.
Gold G, Giannakopoulos P, Herrmann FR, Bouras C, Kövari E. Identification of Alzheimer and vascular lesion thresholds for mixed dementia. Brain 2007, 130: 2830– 2836.
Gorelick PB, Scuteri A, Black SE, et al. Vascular contributions to cognitive impairment and dementia: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011, 42: 2672–2713.
Hachinski V. Stroke and Alzheimer disease: fellow travelers or partners in crime? Arch Neurol 2011, 68: 797– 798.
Hardy J, Allsop D. Amyloid deposition as the central event in the aetiology of Alzheimer's disease.
Trends Pharmacol Sci 1991, 12: 383-388.
Haxby JV, Gray CL, Koss E, et al. Longitudinal study of cerebral metabolic asymmetries and associated neuropsychological pat- terns in early dementia of the Alzheimer type. Arch Neurol 1990, 47: 753–760. Henderson T. The diagnosis and evaluation of dementia and mild cognitive impairment with emphasis on SPECT perfusion neuroimaging. CNS Spectr 2012, 17: 176-206. Herholz K, Schopphoff H, Schmidt M, et al. Direct comparison of spatially normalized PET and SPECT scans in Alzheimer’s disease. J Nucl Med 2002, 43: 21–26. Holman BL, Johnson KA, Gerada B, et al. The scintigraphic appearance of Alzheimer disease: a prospective study using technetium-99m-HMPAO SPECT.
J Nucl Med 1992, 33: 181–185.
Hu WT, Wang Z, Lee VM, Trojanowski JQ, Detre JA, Grossman M, et al. Distinct cerebral perfusion patterns in FTLD and AD. Neurology 2010, 75: 881–888.
Huang Y, Mucke L. Alzheimer mechanisms and therapeutic strategies. Cell 2012, 148: 1204–1222. Iadecola C. The Pathobiology of Vascular Dementia.
Neuron 2013, 80: 844–866.
Jagust WJ. Neuroimaging in dementia. Neurol Clin 2000,
18: 885-902.
Jagust WJ, Friedland RP, Budinger TF, et al. Longitudinal studies of regional cerebral metabolism in Alzheimer disease. Neurology 1988, 38: 909–912.
Jellinger KA. Alzheimer 100 - Highlights in the history of Alzheimer research. J Neural Transm 2006, 113: 1603– 1623.
Jellinger KA. Pathology and pathogenesis of vascular cognitive impairment-a critical update. Front Aging
Neurosci 2013, 5: 1–19.
Jellinger KA, Attems J. Prevalence of dementia disorders in the oldest-old: an autopsy study. Acta Neuropathol 2010, 119: 421–433.
Johnson N, Jahng G-H, Weiner MW, et al. Pattern of cerebral hypoperfusion in Alzheimer disease and mild cognitive impairment measured with arterial spin-labeling MR imaging: initial experience. Radiology 2005, 234: 851– 859.
Kerrouche N, Herholz K, Mielke R, Holthoff V, Baron J-C. 18FDG PET in vascular dementia: differentiation from Alzheimer’s disease using voxel-based multivariate analysis. J Cereb Blood Flow Metab 2006, 26: 1213–1221.
Kim J, Basak JM, Holtzman DM. The Role of
Apolipoprotein E in Alzheimer’s Disease. Neuron 2009, 63, 287–303.
Kirson NY, Hunter CA, Desai U, Cummings AKG, et al. Excess costs associated with misdiagnosis of Alzheimer’s disease among US medicare beneficiaries with vascular dementia or Parkinson’s disease. Boston, MA: Alzheimer’s Association International Conference 2013, Alzheimer’s &
Dementia 2013, 9: S845-S846.
Knopman DS, Parisi JE, Salviati A, Floriach-Robert M, Boeve BF, Ivnik RJ, Smith GE, Dickson DW, Johnson KA, Petersen LE, McDonald WC, Braak H, Petersen RC. Neuropathology of cognitively normal elderly. J
Neuropathol Exp Neurol 2003, 62: 1087-1095.
Koretsky AP. Early Development of Arterial Spin Labeling to Measure Regional Brain Blood Flow by MRI.
Neuroimage 2012, 62: 602–607.
Kraepelin E. Psychiatrie: ein Lehrbuch fur Studierende und Arzte. II. A Band, Klinische Psychiatrie. Verlag Johann Ambrosius Barth, Leipzig, 1910.
Kuhl DE. Imaging Local Brain Function with Emission Computed Tomography. Radiology 1984, 150: 625–631. Launer LJ, Petrovitch H, Ross GW, Markesbery W, White LR. AD brain pathology: vascular origins? Results from the HAAS autopsy study. Neurobiol Aging 2008, 29: 1587– 1590.
Leonard JP, Nowotnik DP, Neirinckx RD. Technetium-99m-d,1-HM-PAO: a new radiopharmaceutical for imaging regional brain perfusion using SPECT — a comparison with iodine-123 HIPDM. J Nucl Med 1986, 27: 1819–1823. Leys D, Hénon H, Mackowiak-Cordoliani MA, Pasquier F. Poststroke dementia. Lancet Neurol 2005, 4: 752–759. Libon D, Scanlon M, Swenson R, Branch Coslet H. Binswanger's disease: some Neuropsychological Considerations. J Geriatr Psychiatry Neurol 1990, 3: 31-40.
Maddock RJ, Garrett AS, Buonocore MH. Remembering Familiar People: The Posterior Cingulate Cortex and Autobiographical Memory Retrieval. Neuroscience 2001,
104: 667–676.
Marshall RS. Effects of altered cerebral hemodynamics on cognitive function. J Alzheimers Dis 2012, 32: 633–642. Marshall RS, Lazar RM, Mohr JP, Pile-Spellman J, Hacein-Bey L, et al. Higher cerebral function and hemispheric blood flow during awake carotid artery balloon test occlusions. J Neurol Neurosurg Psychiatry 1999, 66: 734– 738.
Mattsson N, Tosun D, Insel PS, et al. Association of brain amyloid-β with cerebral perfusion and structure in Alzheimer’s disease and mild cognitive impairment. Brain 2014, 137: 1550–1561.
Mazza M, Marano G, Traversi G, Bria P, Mazza S. Primary cerebral blood flow deficiency and Alzheimer’s disease: Shadows and lights. J Alzheimers Dis 2011, 23: 375–389. McKhann GM, Grega MA, Borowicz LM Jr, Bailey MM, Barry SJ, Zeger SL, Baumgartner WA, Selnes OA. Is there cognitive decline 1 year after CABG? Comparison with surgical and nonsurgical controls. Neurology 2005, 65: 991–999.
Mendez MF, Shapira JS, McMurtray A, Licht E, Miller BL. Accuracy of the clinical evaluation for frontotemporal dementia. Arch Neurol 2007, 64: 830–835.
Messa C, Perani D, Lucignani G, et al. High-resolution technetium-99m- HMPAO SPECT in patients with probable Alzheimer’s disease: comparison with fluorine-18-FDG PET. J Nucl Med 1994, 35: 210–216.
Minoshima S, Giordani B, Berent S, Frey K a, Foster NL, Kuhl DE. Metabolic reduction in the posterior cingulate cortex in very early Alzheimer’s disease. Ann Neurol 1997,
42: 85–94.
Mitchell AJ, Shiri-Feshki M. Rate of progression of mild cognitive impairment to dementia - Meta-analysis of 41 robust inception cohort studies. Acta Psychiatrica
Scandinavica 2009, 119, 252–265.
Mosconi L, Tsui WH, Herholz K, et al. Multicenter standardized 18F-FDG PET diagnosis of mild cognitive impairment, Alzheimer’s disease, and other dementias. J
Nucl Med 2008, 49: 390–398.
Mosconi L, Perani D, Sorbi S, Herholz K, Nacmias B, Holthoff V, et al. MCI conversion to dementia and the APOE genotype: a prediction study with FDG-PET.
Neurology 2004, 63: 2332–2340.
Moskowitz MA, Lo EH, Iadecola C. The science of stroke: mechanisms in search of treatments. Neuron 2010, 67: 181–19.
Musiek E, Chen Y, et al. Direct Comparison of FDG-PET and ASL-MRI in Alzheimer’s Disease. Alzheimers Dement 2012, 8: 51–59.
Pantelakis S. A particular type of senile angiopathy of the central nervous system: congophilic angiopathy, topography and frequency. Monatsschr Psychiatr Neurol 1954, 128: 219–256.
Parkes LM, Rashid W, Chard DT, Tofts PS. Normal Cerebral Perfusion Measurements Using Arterial Spin Labeling: Reproducibility, Stability, and Age and Gender Effects. Magn Reson Med 2004, 51: 736–743.
Patwardhan MB, McCrory DC, Matchar DB, Samsa GP, Rutschmann OT. Alzheimer disease: operating
characteristics of PET—a meta-analysis. Radiology 2004,
231: 73–80.
Pendlebury ST, Rothwell PM. Prevalence, incidence, and factors associated with pre-stroke and post-stroke dementia: a systematic review and meta-analysis. Lancet
Neuro 2009, 8: 1006–1018.
Pitner JK, Bachman DL. A synopsis of the practice parameters on dementia from the american academy of neurology on the diagnosis of dementia. Consult Pharm 2004, 19: 52-63.
Postiglione A, Lassen NA & Holman BL. Cerebral blood flow in patients with dementia of Alzheimer's type. Aging 1993, 5: 19-26.
Raji CA, Lee C, Lopez OL, Tsay J, Boardman JF, Schwartz ED, et al. Initial experience in using continuous arterial spin-labeled MR imaging for early detection of Alzheimer disease. AJNR 2010, 31: 847–855. Restom K, Bangen KJ, Bondi MW, Perthen JE, Liu TT. Cerebral blood flow and BOLD responses to a memory encoding task: A comparison between healthy young and elderly adults. NeuroImage 2007, 37: 430–439.
