Introduction
Intracranial arterial dolichoectasia (IADE) is a common arteriopathy affecting large intracranial vessels [
1]. It is characterised by abnormal fusiform dilatation (ectasia) and elongation or tortuosity (dolichosis) of the intracranial arteries, mainly involving basilar arteries (BA) in approximately 80% of all cases [
2,
3]. Because of arterial elongation and enlargement, patients with IADE can present with various clinical syndromes, including cerebral ischaemic symptoms due to thromboembolism or local compression (e.g., affecting the cranial nerves or brainstem); obstructive hydrocephalus; or subarachnoid haemorrhage (SAH) [
4‐
6]. Its prevalence in the general population is relatively low, ranging from 0.06% to 5.8%, but its prevalence has been reported to range from 3 to 18% in patients with ischaemic stroke [
1,
7] [
8‐
11].
Recent studies have demonstrated a relationship between cerebral small vessel disease (CSVD) and IADE, which might be important to better understand disease pathogenesis and clinical relevance for diagnosis, prognosis, and treatment in people with cerebrovascular disease [
3,
12]. Advancing age, hypertension, and male sex are associated with IADE, but it is not certain whether these fully explain the observed associations [
13,
14]. Studies on patients with lacunar stroke revealed that IADE was significantly more common in ischaemic stroke attributed to small vessel occlusion than to athero-thromboembolism (36% vs. 19%) and patients with (compared to those without) severe white matter disease (34% vs. 19%) [
7,
15]. However, previous studies are limited in providing a definitive estimate of the strength and consistency of any association with IADE with CSVD. We, therefore, did an updated systematic review and meta-analysis of observational studies to investigate the relationships of IADE among magnetic resonance imaging (MRI)-defined CSVD markers in patients with ischaemic stroke.
Discussion
In our updated systematic review and meta-analysis, we found that 11.4% (95% CI 8.9–13.9) of ischaemic stroke patients had IADE, comparable to previously published data in ischaemic stroke populations [
1‐
3,
6]. Patients with IADE were more likely than patients without IADE to have neuroimaging markers of CSVD, including lacunes, severe leukoaraiosis, CMBs, and WMHs. The included studies already adjusted for potential confounding factors, namely, age, sex, hypertension, ischaemic heart disease or previous myocardial infarction, leukoaraiosis, and stroke subtype, and the finding remained significant [
7,
13,
15,
28‐
30]. Therefore, the associations of IADE with CSVD markers might be independent of these confounders or vascular risk factors.
Our results confirm an association between IADE and CSVD, consistent with the previous cohorts and case–control studies using clinical, imaging, and neuropathological evidence [
7,
13‐
15,
28,
29,
32,
33]. Building on previous work, we show a consistent association of IADE with all of the CSVD markers we investigated and have been able to provide more precise estimates of the increased risk of CSVD associated with IADE. A key question is whether this association is simply due to shared vascular risk factors (e.g., hypertension, diabetes), or whether there are shared aspects of pathophysiology independent of these; our findings indicate that the latter explanation should be considered. Several underlying mechanisms have been proposed to explain the relationship between IADE and CVSD [
34‐
36]. One hypothesis is that pathways associated with blood vessel structure and remodeling are relevant. For example, matrix metalloproteinase (MMP) and tissue inhibitors of metalloproteinase (TIMP) are associated with inflammation and remodeling of extracellular matrix and have been studied in relation to the linkage between CSVD and IADE [
35,
37]. In 510 ischaemic stroke patients, MMP-3 was associated with IADE, suggesting a significance for MMPs in the formation of intracerebral dilative arteriopathy [
37]. A study by DP Zhang et al. on 212 patients confirmed an apparent association and found elevated serum MMP-9 and the ratio of MMP-9 to TIMP-1 (MMP-9/TIMP-1) in vertiginous patients with VBD. Moreover, VBD patients with high-grade WMHs also had significantly high levels of TIMP-1, which correlated with the length and TI of the basilar artery [
35]. MMP might also influence the risk of CSVD; one previous study suggests that mice models exposed to MMP-2 inhibitor exhibited lessened white matter lesions and microglia and astroglia activation following chronic cerebral hypoperfusion [
38], while clinical studies in ischaemic stroke patients with a high burden of CSVD found an independent association with elevated MMP-9 and TIMP-4 [
39,
40]. These results imply that MMP and TIMP might have a role in the pathogenesis of IADE (including VBD) and WMHs or other CVSD markers, with potential relevance for understanding, preventing, and treating cerebrovascular diseases.
IADE may be relevant clinically due to the potential risk of ischaemic stroke. A case–control study of patients with basilar dolichoectasia with and without ischaemic stroke found associations with reduced blood flow velocity, a multi-infarct pattern, and haemodynamic changes related to city and atheromatous alterations in the affected vessels have been proposed as mechanisms of ischaemic stroke (mainly occurring in the pons, but also the thalamus, midbrain and occipital lobes) in patients with BA dolichoectasia [
34]. It can be implied that haemodynamic alterations may play an essential part in dolichoectasia [
3].
Of the CSVD markers we investigated, cerebral microbleeds (CMBs) were the most strongly associated with IADE. CMBs are small (generally 2–10 mm), round or ovoid hypointensity lesions detected on T2*-gradient echo (GRE)/susceptibility-weighted imaging (SWI) and commonly found in patients with CSVD, including hypertensive arteriopathy (arteriolosclerosis) and cerebral amyloid angiopathy (CAA) [
41,
42]. Our results confirm the previously reported associations between CMBs and large vessel dilatative arteriopathy, IADE and its subset (VBD), in various populations including ischaemic stroke [
32,
33,
43]. The degree of basilar artery tortuosity, BA dolichosis, was independently associated with deep CMBs (Odds ratio (OR) 4.14,
P = 0.002) [
32]. In patients with VBD, CMBs are more frequently documented in the posterior circulation brain regions, including the cerebellum, thalamus, and occipital lobe [
43] and a high CMBs burden (> 10 CMBs) was more frequent in vascular territories supplied by vessels arising from dolichoectatic vessels in the posterior region [
31]. This anatomical link between large and small artery pathology is consistent with either shared pathophysiological mechanisms or vulnerability rather than simply shared risk factors [
29,
31,
36,
43]. Importantly, results from meta-regression of gender heterogeneity on the influencing of CMBs in AIS patients with IADE showed that gender might not affect the incidence or burden of CMBs, in agreement with previous studies [
44,
45]. Even though we did not find evidence that the heterogeneity in the relationship between IADE and CMBs is driven by differences in hypertension in the meta-regression analysis, previous studies have emphasised the strong association between IADE and the presence of CMBs after adjusting for age, sex, and hypertension [
13]. Further studies are required to establish the causative relation between CMBs in patients with IADE, especially after correction for the traditional vascular risk factors.
In our systematic review, white matter hyperintensities were observed in 22.8% of ischaemic stroke patients with IADE, over double the risk of WMHs compared to individuals without IADE. A cross-sectional study by Fierini et al. in a cerebrovascular outpatient service confirmed the high prevalence of moderate to severe WMHs in patients with IADE [
33]. Subsequent studies have also suggested a relationship between BA dolichoectasia and CSVD [
30,
32,
33,
46]. BA diameter significantly correlated with the presence of WMHs, and BA dolichoectasia was approximately three times associated with the severity of WMHs [
32]. Although a study of 469 Chinese AIS patients failed to establish the association between IADE and intracranial atherosclerosis (ICAS), they found that IADE had been related to older age, hypertension, multilacunes, and WMHs [
30]. An autopsy-based study of 381 patients with stroke emphasised the association of CSVD and its consequence because IADE-positive patients were more significantly to develop SVD than IADE-negative patients. Notably, plaque formation in the affected arteries of IADE patients was elevated considerably, and IADE-positive patients exhibited no evidence of CAA [
28]. These findings might be hypothesised that CVSD and IADE in AIS individuals could have common underlying pathophysiologic processes.
Brain atrophy, defined as cortical or subcortical brain volume reduction that is not associated with significant traumatic brain injury or infarction, is recognised as one of the neuroimaging features of CSVD on brain MRI [
47‐
49]. We did not find studies reporting any association between IADE and brain atrophy. However, our results did show an association of IADE with lacunes, which have in turn been associated with regional cortical and subcortical grey matter volume loss in individuals with vascular mild cognitive impairment [
50]. Additional studies are required to establish more directly whether IADE is associated with brain atrophy. In addition, it is important to differentiate true lacunar infarcts related to in situ arteriolosclerosis from branch atheromatous disease (BAD), i.e., occlusion of the perforator orifice due to junctional plaque, because the underlying vascular pathology might be different and BAD-related strokes are associated with early neurological deterioration (END) in acute stroke due to mall vessel occlusion [
51,
52].
Most previous studies demonstrated the prevalence of IADE or VBD in acute ischaemic stroke populations, while data in intracerebral haemorrhage (ICH)—an important and clinically severe manifestation of CSVD—are limited. In 2012, a study of 481 acute stroke patients showed the prevalence of VBD in patients with ICH was approximately double that seen in patients with brain infarcts (12.1% vs. 6.4%) [
53], although the prevalence of VBD in that study was lower than in other studies despite using similar criteria to diagnose VBD [
4,
54]. Therefore, additional studies are needed to investigate IADE in patients with ICH. In addition, the utility of VBD as a predictor of stroke outcomes or mortality requires additional investigation. The involvement of the basilar artery was an independent risk factor for transient or fixed posterior circulation dysfunction or neurological morbidity, whereas the mortality in affected patients appears to be anticipated more by traditional vascular risk factors than by VBD characteristics [
55]. Increasing evidence is expanding the range of MRI features of lesions related to CSVD, including recent small subcortical infarcts (RSSI), cortical cerebral microinfarcts, and cortical superficial siderosis (cSS); future research concerning associations between these SVD markers and IADE in both ischaemic stroke and ICH may help to better understand the mechanisms underlying the associations we have reported.
In the present study, we comprehensively examined all published reports of associations of IADE with a range of neuroimaging markers of CSVD. However, we acknowledge some limitations, including the small number of included studies that limit our ability to identify publication bias and sources of heterogeneity in meta-regression [
56]. Second, the overall quality of the included studies is only fair, and the certainty of evidence for the outcomes is relatively broad, ranging from very low to moderate. Five of seven publications have an NOS score of less than seven. Finally, the effect of gender differences might restrict the applicability of the present outcomes. While several studies found the potential risk of male gender for IADE development in AIS patients, our meta-regression analysis on CMBs outcomes showed a negative relationship between the proportions of male participants and CMBs risk.