Intracranial arteriosclerosis: determinants and clinical consequences

Abstract

Part 1 and Chapter 1 of this dissertation review the available knowledge on intracranial arteriosclerosis. Chapter 1 describes that most studies on the prevalence and risk factors of intracranial arteriosclerosis focus on the intracranial carotid artery. Thus, there is a striking paucity of knowledge on the prevalence and determinants of arteriosclerosis in other intracranial arteries.

For this reason, in Chapter 2 we examined the prevalence of and risk factors for calcification of the vertebrobasilar arteries (VBAC) in a sample of middle-aged cardiovascular patients. Here we found that VBAC occurred in approximately 25% of patients, similar to the prevalence previously observed in elderly people from the general population. With respect to risk factors, we found that, in addition to age, obesity was strongly associated with VBAC. Previous research conducted in people from the general population showed that, in addition to age and obesity, hypertension and smoking were also important risk factors.

Subtypes of intracranial arteriosclerosis

As part of our investigation into the association between intracranial arteriosclerosis and stroke, in Chapter 5 we assessed the prevalence of each subtype of intracranial carotid artery calcification (ICAC) across 5-year age ranges in middle-aged and elderly persons from the general population. Here we saw that the intima subtype was most prevalent until the age group of 65 to 70 years, after which IEL became and remained the most common subtype. An interesting finding was that the prevalence of intimal subtype decreased in the oldest age groups. This could be explained by differential misclassification, in particular with large-volume intima calculi being misclassified as IEL calculi (as is done with large-volume calculi in the leg vessels).

Arteriosclerosis in pseudoxanthoma elasticum (PXE)

PXE as a model disease for intracranial IEL calcification

PXE is often cited in the medical literature as a model disease to study the effects of isolated IEL calcification. However, there is a hypothesis that assumes that PXE patients are at greater risk of developing atherosclerosis. If this is true, it is also conceivable that intima calcification is increased in PXE, which would significantly limit the suitability of PXE as a model disease for isolated IEL calcification. It is therefore important to know whether the burden of intracranial atherosclerosis is indeed increased in PXE patients. In addition, knowledge on the burden of intracranial atherosclerosis in PXE is warranted as this is hypothesized to underlie the increased risk of stroke that is observed in PXE patients. Against this background, we conducted a case-control study in Chapter 3 comparing the prevalence of intracranial atherosclerotic lesions in the arteries of the Circle of Willis between PXE patients and controls (matched for age and sex). We found that the prevalence and distribution of intracranial atherosclerotic lesions were similar between PXE patients and controls. This suggests that the burden of intracranial intimal calcifications is similar between PXE and non-PXE patients (although more certainty about this requires histological examination). Given the increased propensity for calcification of the IEL in PXE patients shown by other research, it seems reasonable to assume that the increased amount of ICAC in PXE patients is indeed mainly composed

of IEL calcification. Also, this finding implies that factors other than intracranial atherosclerosis contribute to the increased risk of stroke observed in PXE.

The natural course of arteriosclerosis in PXE

Due to the increased burden of arteriosclerosis that is observed in PXE, several clinical trials are currently analyzing drugs that aim to inhibit arteriosclerosis in PXE. It is therefore important to better understand the natural course of arteriosclerosis in PXE. For this reason, in Chapter 4, we conducted a study in which we measured the changes of arteriosclerosis throughout the body in PXE patients. We found that calcification progressed in a linear fashion over time for all arteries except for the arteries. The annual rate of change depended highly upon the artery under scrutiny. Also, the impact of cardiovascular risk factors on the progression of arterial calcification was limited.

The clinical consequences of intracranial arteriosclerosis

Stroke

Previous cross-sectional studies found evidence of an association between ICAC and stroke. Also, two longitudinal studies were conducted. An earlier study in the Rotterdam Study (14055 person-years follow-up) showed that both the presence and volume of ICAC predicted first stroke in the general population. In addition, a study conducted in Northern Europe (median follow-up 2.2 years) showed that the amount of intracranial arteriosclerosis predicted recurrent stroke. Against this background, we assessed the association of ICAC subtypes with first stroke in more than two thousand middle-aged and elderly persons from the general population in Chapter 5. Based on our understanding of the pathophysiology of atherosclerosis and the available literature on intracranial atherosclerosis and stroke risk, we expected intima ICAC to underlie the increased risk of stroke. However, we found that although both subtypes increased the risk of stroke, the IEL subtype in particular increased the risk of stroke: the hazard ratio (HR) for stroke for intimal ICAC was 2.11 (95% confidence interval [BI] 1.07-4.13) and for IEL ICAC 2.66 (95% BI 1.39-5.11).

Dementia

Previous cross-sectional studies have reported an inverse association between ICAC burden and cognitive ability. Following this, two longitudinal studies were conducted, the first being an earlier study in the Rotterdam Study with 13397 person-years of follow-up. This study revealed that ICAC predicted cognitive decline and a trend was also observed for increased risk of dementia. In addition, in the Knight ADRC cohort (825 person-years of follow-up), no association was found between ICAC and cognition, nor with a composite outcome of mild cognitive impairment and dementia. Against this background, in Chapter 6 we examined the association of ICAC, ICAC subtypes and VBAC with dementia. We found that both presence (HR 1.53 [95%BI 1.00-2.32]) and volume (HR per 1 standard deviation increase 1.19 [95%BI 1.01-1.40]) of ICAC were associated with increased risk of dementia. In addition, we found that severe VBAC also increased the risk of dementia (HR for the third vs. first volume subtile 1.89 [95%BI 1.00-3.59]). Since vertebrobasilar arteries supply blood to various brain structures important in the etiology of dementia (e.g. subcortical structures, periventricular white matter, the cerebellum), this finding raised the question of whether the effects of VBAC are site-specific, or whether they reflect a general effect of intracranial arteriosclerosis on the brain. In an

attempt to elucidate this further, we examined whether the effects of ICAC and VBAC on dementia were mediated by subcortical brain structural volumes. The amygdala is vascularized by downstream branches of the intracranial carotid artery, the hippocampus by downstream branches of the vertebrobasilar arteries, and the thalamus by downstream branches of both arteries. If it is assume that the effects of arteriosclerosis on dementia are mediated (in part) by atrophy of these subcortical structures, we would expect the volume of the amygdala to mediate the association between ICAC and dementia, the hippocampal volume to mediate the association between VBAC and dementia, and the thalamic volume to mediate both associations. However, none of these effects occurred; we found no mediation via (reduced) subcortical brain structural volumes. Another risk factor for dementia is cerebral small vessel disease (cSVD), in which white matter hyperintensity (WMH) volume, microbleeds and lacunar infarcts are particularly important. Therefore, we also investigated whether cSVD mediated the association between arteriosclerosis and dementia. Both the effects of ICAC (for 13%) and VBAC (for 24%) on dementia were partially mediated by increased WMH volume. No mediation effects occurred due to microbleeds or lacunar infarcts. An interesting finding was that we found that, when we looked at ICAC subtypes, mediation occurred only in the IEL subtype and not in the intima subtype, again underscoring how differences in morphology of calcium (namely those between the two subtypes) determine the effect on the brain.