Elderly CADASIL patients with intact neurological status

Article information

J Stroke. 2022;24(3):352-362

Publication date (electronic) : 2022 September 30

doi : https://doi.org/10.5853/jos.2022.01578

Ruiting Zhang^a,b, Elisa Ouin^c, Lina Grosset^a,d, Karine Ighilkrim^e, Jessica Lebenberg^a,d, Stéphanie Guey^a,d, Véronique François^e, Elisabeth Tournier-Lasserve^a,f, Eric Jouvent^a,d, Hugues Chabriat^,a,d

^aParis-Cité University, Inserm U1141 NeuroDiderot, Paris, France

^bDepartment of Radiology, the Second Affiliated Hospital of Zhejiang University, School of Medicine, Hangzhou, China

^cDepartment of Neurology, Amiens University Hospital, Laboratory of Functional Neurosciences1,6 (UR UPJV 4559), Jules Verne Picardy University, Amiens, France

^dLariboisière University Hospital, APHP, Translational Neurovascular Centre and Department of Neurology, Reference Center for Rare Vascular Diseases of the Central Nervous System and the Retina (CERVCO), FHU NeuroVasc, Paris, France

^eDepartment of Geriatrics, Lariboisière University Hospital, APHP, Paris, France

^fDepartment of Neurovascular Molecular Genetics, Saint-Louis Hospital, APHP, Paris, France

Correspondence: Hugues Chabriat Centre NeuroVasculaire Translationnel, Hôpital Lariboisière, 2 rue Ambroise Paré, 75010 Paris, France Tel: +33-1-49952593 Fax: +33-1-49952596 E-mail: hugues.chabriat@aphp.fr

Received 2022 May 16; Revised 2022 August 1; Accepted 2022 August 10.

Abstract

Introduction

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is the most common inherited cerebral small vessel disease (cSVD) [1]. Although the spectrum of clinical manifestations appears very broad and their severity highly varies over decades of disease progression, motor disability and cognitive decline develop inexorably in most older patients with a confirmed diagnosis of CADASIL. In one of the largest samples reported so far, more than 90% of patients aged more than 65 years already experienced one or more strokes and half were severely demented and bedridden [2]. The median life expectancy was estimated around 65 years in men and 71 years in women [2]. Similar findings have been reported in multiple large cohorts of symptomatic patients [3,4]. Thus, although its rate of clinical progression appears highly variable over decades, CADASIL is mainly considered to be one of the most devastating cSVDs that, ultimately, always converges to severe disability with increasing age [1].

In addition to age, different factors have been identified as potentially modulating the disease progression along aging. The ε2 allele of apolipoprotein E was found associated with a larger volume of white matter hyperintensities (WMHs) [5], male sex with a higher risk of cognitive decline and more lacunes and cerebral atrophy [6], smoking with earlier and more incident stroke events [7], higher blood pressure with an increased risk of cognitive decline [8]. More recently, after typical NOTCH3 mutations in domains 7-34 were discovered at an unexpectedly high frequency in the general population, the location of cysteine mutations in epidermal growth factor-like repeat (EGFr) domains 1-6 versus 7-34 was also shown to have a strong detrimental effect on this clinical variability [9]. Some elderly individuals having mutations in EGFr domains 7-34 were even found to present with very limited cerebral lesions [10]. This does not rule out that patients with variants located in EGFr domain 1-6 cannot remain neurologically intact (N_int) or stable [11].

In parallel, during the recent decade and with the spread of genetic testing, diagnosis of CADASIL has been broadened to include patients with more atypical symptoms or much less severe clinical picture. Moreover, the sensitivity of genetic testing in daily practice has been improved by allowing the search of cysteine mutations in all exons encoding for EGFr domains whereas the diagnostic test was initially limited to only few exons. Thus, the number of elderly patients diagnosed with the disease has increased, not only because of the aging of patients diagnosed many years ago but also because of the increased diagnostic rate in the elderly even in presence of transient or benign neurological manifestations or after incidental discovery of white matter signal changes.

Specific analysis of cases who remain neurologically normal despite the significant advancement in age may provide insight into the different sources of clinical variability in the presence of NOTCH3 gene mutations. The demographics, clinical, imaging, and genotype characteristics of this population remain however poorly understood. In the present study, we aimed to better delineate these features by comparing elderly CADASIL patients with intact neurological status to the other patients of identical age but with some permanent symptoms.

Methods

Results

One hundred and twenty-nine patients whose age was higher than 65 years at their last follow-up visit were selected from the entire cohort. Among them, 60.5% were women. The mean age at last visit was 72±5 years (median age, 72; interquartile range, 68 to 74). Eighty-nine patients (69.0%) were identified as the first family case at inclusion.

Among the 129 elderly selected CADASIL patients, 23 (17.8%) individuals were considered as N_int. Their main characteristics are summarized in Table 1. Thirteen of them were diagnosed after MRI evaluation of attacks of migraine with aura, one after the occurrence of repeated isolated auras, three had stroke and one had TIA, three had an MRI after a depressive episode associated in one case with chronic headache, and one patient had syncope-like episodes. Only in one patient, MRI and genetic testing were performed after the diagnosis of CADASIL in her brother who had a recent ischemic stroke. Vascular risk factors were present in 65% of N_int subjects, hypertension was diagnosed in 26% of them.

Table 1.

Characteristics of the N_int patients

The comparison between N_int and N_ps patients is detailed in Table 2. The results showed that age, the sex ratio, and frequency of vascular risk factors did not differ between the two groups. Migraine with aura were significantly more frequent in the N_int than in the N_ps group. Notably, N_int patients had significantly less stroke than the N_ps patients. Transient episode of mood disturbance, such as episodes of anxiety or depression that needed a medical treatment, in the N_int group were significantly less frequent than in the N_ps group. Also, the different clinical scores (mRS, Barthel index, NIHSS, MMSE, and MDRS) obtained in each group confirmed the large clinical integrity of N_int patients compared to the N_ps group. The same differences were observed from inclusion in the study when only baseline data were considered in the analysis (Supplementary Table 1).

Table 2.

Comparison of the main demographic, clinical and genetic features between N_int and N_ps patients

The analysis of MRI parameters detailed in Table 3 showed that the volume of WMHs did not differ between the N_int and N_ps group, nor did the Fazekas scores obtained separately for the periventricular and deep WMHs. Conversely, the load of lacunes and microbleeds largely differ between the two groups. N_int patient also present with a higher normalized brain volume than N_ps patients. These results did not differ when only patients from distinct families were considered in the analysis (comparison restricted to only the first family members participating to the cohort study and considered as “index” patients, see Supplementary Tables 2 and 3).

Table 3.

Comparison of MRI data between N_int and N_ps patients

The main imaging findings in the 23 N_int patients are summarized using FLAIR images on Figures 1 and 2. These data clearly confirmed the large variability in the load of WMHs in N_int patients whose normalized WMHs volume was ranging from 1.0% to 13.3%. Eleven patients harbored a variant located in EGFr domains 1-6, while 12 patients harbored a variant located in EGFr domains 7-34. The figures showed that patients with variants in EGFr domain 1-6 had a higher load of WMHs than patients with variants in EGFr domain 7-34 (normalized WMHs volume, 8.9% vs. 5.2%, P=0.027), especially in the temporal poles (WMHs score, 3 vs. 1, P=0.023). In the entire group (including both N_int and N_ps individuals), patients with variants in EGFr domain 1-6 also had a higher load of WMHs in the temporal poles than the others with variants in EGFr domain 7-34 (WMHs score, 3 vs. 2, P<0.001), but the total lesion load did not differ between the two groups (normalized WMHs volume, 8.0% vs. 6.3%, P=0.234). The results did not change when only the subgroup of N_int patients who never had a stroke (n=12) was compared to all other patients older than 65 years (n=117, see Supplementary Tables 4 and 5).

Figure 1.

Magnetic resonance fluid-attenuated inversion recovery images of the 23 elderly neurologically intact (N_int) patients at the level of basal ganglia. (A) Eleven patients had their cysteine mutation in epidermal growth factor-like repeat (EGFr) domain 1-6 (Patient #1 to #11), (B) 12 others in EGFr domain 7-34 (Patient #12 to #23). Only three patients (Patient #1, #2, and #5) were from the same family. The patients were sorted according to their white matter hyperintensities load (from the highest to the lowest) in the two EGFr groups, respectively.

Figure 2.

Magnetic resonance fluid-attenuated inversion recovery images of the neurologically intact (N_int) patients already presented in Figure 1 (same order) at the level of temporal pole. (A) Eleven patients had their cysteine mutation in epidermal growth factor-like repeat (EGFr) domain 1-6 (Patient #1 to #11), (B) 12 others in EGFr domain 7-34 (Patient #12 to #23).

Discussion

In this large prospective cohort of patients diagnosed with CADASIL, we observed that 17.8% of individuals can remain N_int despite their age extending beyond 65 and up to 76 years. Only anecdotal elderly cases up to 86 years have been previously described in the literature as “minimally symptomatic,” “pauci-symptomatic,” or “with no history of stroke or vascular cognitive decline.” [9,22-30] In the present study, the drastic selection of N_int cases from a large sample of consecutive elderly patients allows, for the first time, to delineate the spectrum of imaging and genetic characteristics associated with neurological preservation despite aging in CADASIL.

In the N_int group, 14 individuals already suffered attacks of migraine with aura or isolated auras. This frequency is twice that observed in the N_ps group and much higher than the average frequency of 40% usually reported in cohorts of symptomatic CADASIL patients [31,32]. This is possibly related to our selection of N_int patients among cases recruited in a clinical center, who were mostly symptomatic and thus diagnosed after mild or transient clinical manifestations. Also, this high prevalence of migraine with aura in N_int elderly individuals further supports the absence of link between migraine with aura and clinical disability or cognitive decline in CADASIL [31,33]. Other transient clinical manifestations were depression episodes in two N_int cases, possibly favoured by the presence of confluent white matter abnormalities [34]. Interesting, only a single N_int woman was totally asymptomatic and diagnosed after her brother had a stroke at young age, which led to genetic testing. The results did not change when only “index” patients were considered in the analysis (supplementary data). Finally, we did not observe a significant difference in the frequency of cardiovascular risk factors between the N_ps and N_int groups, suggesting that these factors are not involved in this clinical contrast. Conversely, N_int patients were more educated than N_ps patients, thus whether additional protective factors related to differences in the level of regular exercise, diet, behavior, access to care, health awareness or to the environment cannot be ruled out [35].

As expected, N_int elderly patients had much less lacunes and microbleeds as well as less cerebral atrophy than N_ps patients. These imaging markers are strongly related to the development of disability and dementia in CADASIL patients [13,36]. In contrast, the global amount of WMHs did not differ between the N_int and N_ps groups. This discrepancy suggests that although the disease has visible effects in the white matter, the development of lacunes and loss of cerebral tissue along aging does not occur or remains limited in most N_int patients. In line, accumulating data support that some WMHs in CADASIL may result more from water accumulation than from myelin or axonal loss as observed in ischemic cerebral tissue [37,38].

The location of NOTCH3 mutation in different EGFr domains might be involved in this relative preservation despite aging. About half of N_int CADASIL patients harbored cysteine mutations within EGFr domain 7-34. This is higher than the average frequency of such mutations in European cohorts of symptomatic diagnosed cases [3]. Since a selection bias related to the inclusion of only survivors aged more than 65 years cannot be excluded in our cohort, this proportion might be even underestimated. Recent pathology study of skin biopsies suggested that patients with an EGFr 7-34 variant might have less NOTCH3 extracellular domains accumulation in the wall of their brain vessels than patients with an EGFr 1-6 variant [39]. This would explain the higher prevalence of these mutations in the general population with a much larger number of silent or mild forms of the disease [3,40]. In the present study, the location of NOTCH3 variants within EGFr domains did not differ between N_int and N_ps patients. Therefore, if the impact of mutation location on the level of NOTCH3 accumulation is actually proven at cerebral vascular level, other additional factors are obviously involved. Various protection mechanisms may contribute to this relative preservation at multiple level, such as, genetic or hormonal modulation of brain tissue susceptibility to chronic hypoperfusion, anatomical or physiological characteristics of the cerebral microvascular network or processes involved in the triggering and adaptation of cerebral blood flow autoregulation at tissue level.

Obviously, among elderly N_int patients, a large amount of WMHs involving both temporal lobes was observed in most cases having a mutation in EGFr domain 1-6 while the extent of WMHs appeared much less and the temporal poles largely spared in presence of a mutation in EGFr domain 7-34. At the opposite extreme, very beginning and limited WMHs were detected in a 66-year-old patient having a cysteine mutation in EGFr domain 30 (case 22) but such beginning WMHs were also detected in a 73-year-old patient who had a mutation in EGFr domain 4 (case 11). Altogether, these results suggest that the location of NOTCH3 mutation might modulate the mechanisms involved in the development of WMHs, especially in the white matter adjacent to the cortex as within the temporal poles. The factors needed to delay by several decades the appearance of WMHs in some individuals remain also unknown.

The strengths of this study are multiple. The selected sample of elderly CADASIL patients is large. Each case was extensively investigated. Both clinical results, MRI data and the location of their mutation in the NOTCH3 gene were analyzed in detail. The collection of data was obtained by experienced clinicians over 15 years and always using the same imaging and clinical parameters. The study has also some limitations. It is based on data driven from the hospital, consequently, in the N_int group, totally asymptomatic individuals are probably largely underestimated. On the opposite side, in the N_ps group, elderly cases who were severely disabled or dependent are also most likely underrepresented. They have more difficulties to participate in research activities and to travel to the referral center. The most severe CADASIL cases who died before the age of 65 years are also not accounted in such a comparison study. Besides, due to the very long duration of the disease, we didn’t analyze all treatments that may have been used over decades in all individuals. In addition, subtle psychiatric disturbances might have not been taken into account when defining “intact” neurological status. Finally, the interpretation of our data based on imaging data, particularly in N_int patients should remain cautious in the absence of pathological examination.

Conclusions

We found that nearly one in five CADASIL patients can remain N_int after the age of 65 years. Their different clinical and imaging profile cannot be fully explained by the location of NOTCH3 mutation inside or outside EGFr domains 1-6. In some N_int individuals, severe ischemic lesions do not develop conversely to WMHs that can become extensive. In others, all types of cerebral lesions, including WMHs appear extremely limited and might be delayed by several decades for unknown reasons. The factors involved in this relative cerebral tissue preservation along aging remain to be determined.

Supplementary materials

Article information Continued