Introduction

In the general population there are dramatic individual differences in the severity of cerebrovascular disease with increasing age that promote cognitive decline and clinical symptoms of Alzheimer’s disease (AD) [1–7]. Historically, there has been an underappreciation of the presence, contribution, and significance of cerebrovascular pathology in people with Down syndrome (DS) due to lower frequencies of some of the systemic vascular risk factors that are found in the general population. Indeed, there are several features of aging in people with DS that might suggest protection against cerebrovascular pathology, such as very low rates of hypertension and systemic atherosclerosis [8–10]. On the other hand, people with DS are more likely to have congenital heart disease, obesity, high cholesterol, and sleep apnea, which may increase risk for cerebrovascular disease (see Chapter 12). In this chapter, we first describe cerebrovascular pathology at the microscopic level from postmortem tissue studies of individuals with DS. Next, we review emerging literature that highlights the ability to detect cerebrovascular disease with noninvasive neuroimaging techniques and its potential role in AD in the context of DS.

Cerebrovascular pathology

Triplication of chromosome 21 in adults with DS results in overexpression of amyloid precursor protein, which is thought to mediate increased levels of both fibrillar and soluble forms of beta-amyloid (Aβ) [11–13]. Adults with DS have notable degrees of cerebral amyloid angiopathy (CAA) at autopsy [14, 15] (Fig. 1), referring to the deposition of amyloid within the walls of leptomeningeal and cortical arteries, arterioles, and capillaries [16]. Of historical note, the first isolation and characterization of the beta-amyloid protein in the context of AD was from CAA in an autopsy case from an adult with DS [17]. CAA can lead to micro- and macrohemorrhages as seen in AD in the general population [18] and is more prominent in posterior brain regions. In DS, levels of the shorter, more soluble form of Aβ, Aβ40, are particularly associated with CAA and rise exponentially with age in the cortex [19] and in plasma [20]. Increased age is strongly associated with CAA severity among adults with DS, but not among participants with AD in the general population or in healthy controls [15]. Possible consequences of CAA in DS include vascular dysfunction (impaired constriction and dilation), disruptions in the blood-brain barrier, and microhemorrhages, all of which may contribute to an earlier age of onset of dementia in DS [21].

Although individuals with DS are at increased risk of hemorrhagic lesions due to CAA, interestingly, they have a lower risk for developing intracerebral hemorrhage compared with individuals with a specific form of familial AD who also have three copies of the amyloid precursor protein (APP) gene [22]. This observation raises the possibility that other genes on chromosome 21 and related factors, such as Aβ clearance, inflammation, blood pressure, and vessel disease, could moderate the risk of intracerebral hemorrhage [22].

While hemorrhagic pathology due to CAA is quite common in the adult brain of individuals with DS [15], other forms of vessel disease are quite rare. For example, moderate or severe atherosclerosis is rarely observed in the circle of Willis of adults with DS who come to autopsy, perhaps due to their trisomy or reflecting their younger ages at death [10, 15].

Neuroimaging-derived markers of cerebrovascular disease in adults with Down syndrome

High resolution structural magnetic resonance imaging (MRI) protocols can be used to derive common markers of cerebrovascular disease that parallel observations at autopsy. Clinical MRI protocols that comprise standard T1-weighted anatomical scans, gradient echo or susceptibility weighted imaging, and T2-weighted and T2-weighted fluid attenuated inversion recovery (FLAIR) scans are used to appreciate infarcts, hemorrhagic lesions (including microbleeds), white matter hyperintensities (WMH), and enlarged perivascular spaces. More experimental approaches, such as diffusion tensor imaging (DTI), can estimate the orientation and integrity of white matter tracts in the brain and are often used to characterize microstructural abnormalities due to cerebrovascular disease [23–25].

The frequency and severity of these cerebrovascular abnormalities increase with age in the general population [2, 26] and are strongly associated with worsening cognitive function in older adults (see review [27]). In recent years, there has been interest in the question of whether cerebrovascular disease, such as the abnormalities appreciated with structural MRI, plays a role in the clinical expression and pathogenesis of AD. This question is particularly interesting and relevant in the context of DS for several reasons, as mentioned earlier. There is a low prevalence of classical vascular risk factors that are most strongly associated with cerebrovascular disease, especially hypertension and atherosclerosis, among individuals with DS [10, 28, 29] (see Chapter 2). Given this low prevalence of modifiable vascular risk factors, an observation of increased cerebrovascular markers that increases with a clinical diagnosis of AD in adults with DS would suggest a more fundamental role of cerebrovascular disease in DS and AD. Further, the extent to which these cerebrovascular markers increase with diagnosis or predict future diagnosis of AD in DS may support a hypothesis of a primary pathogenic role of cerebrovascular disease in AD. Understanding the frequency of cerebrovascular disease with MRI and its role in neurodegenerative processes will help identify therapeutic or preventive strategies for AD in DS but would also have relevance for the general population. This approach has been applied to other fully penetrant genetic forms of AD [30, 31].

Compared with studies in the general population or in late onset AD, there is relatively little research that has used neuroimaging modalities to examine cerebrovascular disease in DS or in AD in the context of DS. In the following sections, we review this emerging literature, focusing on the most common markers of small and large vessel cerebrovascular disease. Because there is evidence that white matter microstructural abnormalities can also result from cerebrovascular disease, we also discuss recent studies that have used diffusion tensor imaging in DS.

Microbleeds. Cerebral microbleeds are small, punctate hypointense lesions detected with susceptibility-weighted and gradient-recalled echo T2* MRI sequences that reflect hemosiderin deposits from small hemorrhages [32, 33] (see Fig. 2). The regional distribution of cerebral microbleeds provides insight into their underlying etiology. Microbleeds that are distributed in deep subcortical nuclei, such as the basal ganglia and thalamus, are typically associated with vessel damage secondary to hypertensive arteriopathy [34], whereas microbleeds distributed in “lobar” regions reflect the presence of cerebral amyloid angiopathy, as evidenced by autopsy studies [35, 36]. Because triplication of chromosome 21 in individuals with DS causes Aβ overproduction [11–13, 37], it is not surprising, as noted before, that autopsy studies in people with DS confirm widespread CAA, deposition of amyloid protein in the brain’s small vessels. This arterial deposition of beta-amyloid can weaken and rupture vessel walls, causing small hemorrhagic lesions, appreciated as microbleeds on MRI [18, 38, 39]. In autopsy samples from individuals with DS, CAA severity corresponded with the number of microbleeds detected, particularly in posterior regions [38]. This distribution pattern is consistent with previous studies of adults with clinical CAA from the general population [32].

Carmona-Iragui and colleagues [30] found that about 16% of adults with DS had detectable lobar microbleeds, but this frequency increased with diagnosis of AD; lobar microbleeds were detected among 38.5% of individuals with DS who were symptomatic for AD compared with 12% among those who were asymptomatic. These findings are similar to a study that showed 15% of older adults with DS had lobar microbleeds, which increased with diagnosis of clinical AD [40]. Examination of cerebral microbleeds on MRI scans in adults with DS likely underestimates the severity of underlying CAA and the number of microbleeds detectable at autopsy because of the lack of sensitivity of MRI protocols to visualize these small lesions [41].

Although MRI-defined lobar microbleeds increase with diagnosis of clinical AD, whether microbleeds themselves, or the underlying CAA they reflect, promote cognitive dysfunction is not entirely clear. Some studies in the general population [33, 42, 43] show a reliable relationship between number of microbleeds and cognition, whereas others do not [44, 45]. We are not aware of any studies that examined the relationship between presence or number of microbleeds and cognitive decline or incident dementia among adults with DS.

White matter hyperintensities. White matter hyperintensities (WMH; Fig. 2) are areas of increased signal visualized on T2-weighted, including FLAIR, MRI scans that are considered radiological markers of small vessel cerebrovascular disease [46, 47]. They are classically linked to vascular risk factors, particularly hypertension, increase with age in the general population [48–50], and have been implicated as a pathogenic and contributing feature to AD [31, 51, 52], vascular cognitive impairment [53, 54], and vascular dementia [55, 56]. The primary pathophysiological underpinnings are hypothesized to be ischemic damage due to local blood flow abnormalities [54, 57]. As reviewed before, individuals with DS tend to have lower levels of vascular risk factors that are classically linked to small vessel cerebrovascular disease, including hypertension, resulting in few studies that have examined WMH in DS. We identified only two published studies that included measurement of WMH in individuals with DS with high-field MRI among other imaging markers of aging [30, 58]. In 30 adults with DS and 30 age- and sex-matched adults without DS, Roth and colleagues [58] found that people with DS had more severe WMH that increased with age. They interpreted the findings as evidence of accelerated brain aging among adults with DS but did not speculate on the underlying etiology of the observed increase in WMH severity. Similarly, Carmona-Iragui reported a strong correlation of WMH severity with increased age, with markers of CAA, and with diagnosis of AD, which was stronger among adults with DS than in control groups [30]. These authors concluded that their observations supported a hypothesis that WMH are partially related to Aβ deposition. In our own work [40], we found that WMH volume increases monotonically across Alzheimer’s-related diagnostic groups, increasing in severity across older adults with DS with no evidence of dementia, mild cognitive impairment, possible dementia, and definite dementia. Further, the AD diagnosis-related increase in WMH favored a more posterior distribution, which is consistent with previous work in sporadic AD [59] and highlights the anatomical colocalization with other radiological correlates of AD, including hypometabolism [60] and microbleeds [61–63] (see later discussion here and Chapter 2).

Observations of prevalent WMH in adults with DS that increase in severity among those diagnosed with clinical AD contribute to questions about the etiology of WMH and about their role in the pathogenesis of AD. Findings from individuals with and at risk for sporadic AD suggest that WMH, and the underlying small vessel cerebrovascular disease they represent, may contribute additively to AD clinical presentation [52]. As in DS, individuals with presenilin 1, presenilin 2, and APP mutations, who have early onset AD but relatively low vascular risk, also show increased WMH up to 20 years prior to expected symptom onset, particularly in posterior regions [31]. Thus these “genetic forms” of AD provide neuroimaging evidence of a fundamental involvement of small vessel cerebrovascular changes in AD that cannot be attributable solely to exposure to vascular risk factors. Some have argued that the involvement of WMH in familial AD is a reflection of CAA [64] or Wallerian-like abnormalities secondary to neurodegenerative changes [65]. However, our work demonstrated that markers of CAA do not fully mediate the relationship between AD and WMH [66] and the regional distribution and timing of the emergence of WMH do not support a purely neurodegenerative role. In the case of DS, inflammatory changes are quite prevalent [67] and could play a role in WMH pathophysiology [68]. Furthermore, as noted before, individuals with DS have higher rates of sleep apnea, congenital heart defects, and hypotension. These factors have not yet been examined fully in the context of white matter abnormalities in DS and should be the topic of future research.

Infarcts. Despite the prominent role of cerebrovascular disease and DS, little empirical evidence using neuroimaging techniques has been published about the presence and severity of ischemic infarcts (Fig. 2) in DS. In a single case report, Collacott and colleagues describe a 55-year-old person with DS who has had five cerebrovascular accidents, followed by a stepwise progressive global deterioration [69]. They raise a critical issue that as longevity of individuals with DS increases, the risk and contribution of larger vessel pathology to dementia in DS also increases. Furthermore, the association of Moyamoya disease, a chronic occlusive cerebrovascular disorder characterized by stenosis of arteries, in individuals with DS can lead to larger hemorrhagic strokes in adults [70]. While there are multiple case reports of stroke due to Moyamoya disease within developmental cases and young adults [70–72], there are few such examples in older adults with DS [22].

White matter structural integrity. Diffusion tensor imaging (DTI) assesses the microstructural integrity and orientation of white matter tracts by approximating the diffusion of water molecules. Because myelinated axons are composed of lipid sheaths that are organized in specific orientations, water molecules tend to diffuse along the axonal processes. Fractional anisotropy (FA) is the most common metric applied to DTI and refers to the degree to which water molecules diffuse in the same orientation within a voxel. Disruption or less organized white matter fibers yield lower FA values than fibers that are more intact or better organized. Diffusion tensor imaging characteristics in DS related to developmental processes have been described previously [73], but we are aware of only three reports that used DTI in adults with DS to capture potential neurodegenerative changes [74–76]. We examined white matter FA values in adults with DS and non-DS controls [75] and found lower FA values, indicating less tissue coherence, in predominantly frontal white matter tracts among individuals with DS compared with individuals without DS. These differences were most pronounced for DS participants with clinical AD and suggest white matter microstructural degeneration is related to AD. However, the predominantly periventricular distribution, which mirrors typical WMH distribution patterns in the general population [77, 78], also suggests a cerebrovascular role in the observed white matter changes. These findings were similar to those recently reported among 45 DS individuals without dementia across the adult lifespan [74]. Compared with age-matched controls from the general population, individuals with DS evidenced lower FA in major white matter pathways, including prominent changes in the frontal-subcortical circuits [74]. However, the relationship between FA values and age did not differ in the two participant groups, suggesting developmental white matter microstructural changes in people with DS and not “accelerated aging.” Another study [76] replicated the initial DTI findings, and reported decreased FA in adults without DS compared with control participants, within the anterior thalamic radiation, inferior fronto-occipital fasciculum, the inferior longitudinal fasciculum, and the corticospinal tract, bilaterally. The authors postulated that these white matter differences reflect early degenerative changes due to AD.

Diffusion tensor imaging studies in DS suggest widespread white matter microstructural differences compared with individuals from the general population. Because DTI does not detect specific pathologies, the observed effects could reflect a combination of individual differences in white matter development, neurodegenerative changes due to AD, age-related decline in white matter microstructure, and the effects of cerebrovascular disease.

Summary

This chapter describes vascular pathology in the brain and neuroimaging studies about the involvement of cerebrovascular disease in aging and AD among adults with DS. Compared with the general population, individuals with DS have a unique pattern of risk factors. On the one hand, they appear to be “protected” from some vascular risk factors that are strongly associated with cerebrovascular disease and dementia in the general population, most notably hypertension. On the other hand, other factors, like morphological heart abnormalities, sleep apnea, and overproduction of beta-amyloid protein due to triplication of chromosome 21, are more common and may increase risk for cerebrovascular disease but have not yet been examined in relation to cerebrovascular abnormalities on postmortem tissue or neuroimaging. This profile in many ways is consistent with what has been observed in autopsy studies: there is very little evidence of frank vessel disease (e.g., atherosclerosis) but clearly increased numbers of small hemorrhagic lesions due to CAA compared with the general population. Other pathological markers of cerebrovascular disease have not been studied comprehensively in people with DS or in AD in the context of DS. While there are only a few neuroimaging studies that examined markers of cerebrovascular disease in aging and AD among adults with DS, results are somewhat consistent with autopsy studies in demonstrating the presence of microbleeds. But neuroimaging studies also suggest much more widespread ischemic cerebrovascular pathology in older adults with DS that is greater than what would be expected given the vascular risk profile and that appears to be associated with clinical AD. Future work will need to address the factors that mediate these initial observations.