Abstract
Objective: To investigate the location and severity of MRI signal abnormalities in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL).
Background: One hallmark of this arteriopathy due to mutations of Notch 3 gene is the presence of MRI signal abnormalities in both symptomatic and asymptomatic patients.
Methods: MRIs of 75 patients (43 with symptoms) were reviewed by a neuroradiologist masked to their clinical status. After assessing the presence of MRI lesions on T1- and T2-weighted images (T1-WI, T2-WI) in different subcortical regions, the severity of hyperintensities on T2-WI was scored using a global rating scale and a regional semiquantitative scale in the periventricular white matter (PV), deep white matter (WM), basal ganglia (BG), and infratentorial areas (IT).
Results: Sixty-eight patients (90%) had hyperintensities on T2-WI located in the white matter, more frequent in PV (96%) and WM (85%) than in the superficial white matter (25%). Hyperintensities also occurred in BG (60%) and brainstem(45%). Forty-seven patients (62%) presented with hypointensities on T1-WI. In one-third of the affected inviduals, white matter hyperintensities occurred in the absence of small deep infarcts on T1-WI. The frequency and severity scores calculated for PV, WM, BG, or IT hyperintensities increase dramatically with age. These scores were higher in symptomatic compared with asymptomatic gene carriers. Dementia, Rankin score >1, or both occurred only in the presence of diffuse white matter signal abnormalities.
Conclusion: Our results suggest that different subcortical areas have different vulnerabilities to ischemia in CADASIL. The age effect we observed may show an accumulation of lesions with aging during the course of the disease. A prospective study is needed to investigate if the rating of MRI lesions is of prognostic value in CADASIL.
Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is an inherited arterial disease caused by mutations of Notch 3 gene on chromosome 19.1 The disease can start with attacks of migraine with aura at a mean age of 30 years.2 The most frequent clinical manifestations are subcortical transient ischemic attacks or strokes, usually occurring between 40 and 50 years of age, sometimes associated with severe mood disturbances. The disease leads to death in about 20 years after a period of subcortical dementia associated with pseudobulbar palsy and urinary incontinence.2 One of the hallmarks of CADASIL is the presence of striking white matter MRI signal abnormalities in all symptomatic patients, but also in asymptomatic individuals with the mutated gene.3 These neuroimaging findings within families have been essential for the genetic analysis of affected pedigrees before gene identification.3 MRI shows a more or less diffuse leukoencephalopathy predominating in the periventricular areas, often associated with typical small infarcts in the white matter, basal ganglia(BG), and pons in affected patients.4 The severity and clinical significance of various lesions on MRI have not been systematically assessed. This study aims to investigate whether the frequency of MRI signal abnormalities in CADASIL varies among several subcortical regions, and if the signal abnormalities' severity is related to age, clinical presentation, or both. For this purpose, we reviewed the MRIs performed in 75 CADASIL subjects.
Patients and methods. Before the recent gene identification, MRI investigations were performed in pedigrees affected by CADASIL to confirm their linkage to the corresponding locus on chromosome 19.3 We selected 10 unrelated affected pedigrees originating from France and linked to chromosome 19.5 At least one member from each of these families was later shown to carry a deleterious mutation in Notch 3 gene. All members of these families were genotyped with microsatellite markers spanning the CADASIL interval; of these, 75 members carried the affected haplotype and were selected for this study. The error risk of indirect genotypic diagnostic test as compared with direct screening for Notch 3 mutations is less than 1/10000 in the absence of recombination events at both flanking markers.1,5
Clinical evaluation. The following clinical data were obtained at the time of the MRI study: vascular risk factors; presence of the four main symptoms of the disease (attacks of migraine with aura [IHS criteria],6 ischemic stroke, mood disturbances, and dementia [DSM-III criteria]7); and handicap index with Rankin Scale, mainly evaluating mobility and disability in daily and instrumental activities.8
MRI investigations and evaluation. Neuroimaging was performed at 1 to 1.5 tesla in different radiologic centers in France between 1991 and 1996. T1-weighted images (T1-WI) (time to echo [TE] = 5 to 20 ms, time to repeat [TR] = 400 to 600 ms) and T2-weighted images (T2-WI) (TE = 120 ms, TR= 2000 to 7500 ms) were obtained with contiguous slices (5-, 7-, or 9-mm thickness). T2-WI were always obtained from axial planes using the neuro-ocular plane for reference. T1-WI were from axial, sagittal, or both planes. MRI signal abnormalities were systematically assessed by the same board-certified neuroradiologist (C.L.) masked to the clinical status of the patients.
On T1-WI, the presence of hypointensities (identical to CSF signal) was systematically assessed on both sides for cortex, white matter, caudate, putamen, pallidum, thalamus, internal capsule, external capsule, cerebellum, mesencephalon, pons, and medulla.
On T2-WI, the presence or absence of hyperintensities was systematically assessed in cortex, periventricular white matter (PV) including caps and bands, deep white matter (WM) including internal capsule, superficial white matter in contact with the cortex, external capsule, BG, brainstem, and cerebellar hemispheres. The study also included the already validated semiquantitative visual scoring of hyperintensities of Scheltens et al.9,10 and a global rating derived from that of Bots et al.11 Using the Scheltens method, four scores were calculated: the PV hyperintensities score rates caps or bands into 0 (absent), 1 (≤5-mm thick), or 2 (>5-mm, ≤10-mm thick); the WM hyperintensities score is calculated after summing the rates obtained separately in the frontal, parietal, temporal, and occipital lobes based on the size and number (n) of the lesions as 0 (absent), 1 (≤3 mm; n≤ 5), 2 (≤3 mm; n > 5), 3 (4 to 10 mm; n ≤ 5), 4 (4 to 10 mm; n > 5), 5 (>10 mm; n ≥ 1), or 6 (confluent); the BG hyperintensities score was similarly calculated for the regions including the caudate nucleus, putamen, globus pallidus, and thalamus; and the infratentorial regions (IT) score was also calculated after summing the different rates similarly obtained in the cerebellum, mesencephalon, pons, and medulla. Thereafter, a global rating was performed on T2-WI using a scale of four grades: A (no lesion), B (punctiform or slight periventricular hyperintensities, or both), C (nodular or moderate periventricular hyperintensities, or both) and D (confluent lesions or severe periventricular hyperintensities, or both).
Data analysis. Descriptive statistics were obtained for the different clinical variables and for the presence of MRI lesions in the various areas of the brain (percentage of patients with signal abnormalities in each brain area). Comparisons after grouping for age (<40 years, 40 to 55, and >55 years) or for grades of severity (A, B, C, D) were tested using the chi-square test or analysis of variance (ANOVA). The statistical analysis was computed using the Statview 4.5 statistical software (Abacus Concepts, Berkeley, CA).
Results. Clinical data. Of the 75 family members having the affected haplotype, 33 (44%) had at least one vascular risk factor(22 were smokers, 11 had hypercholesterolemia, and 7 women used contraceptive pills), but only 3 were treated for hypertension. Forty-three (57%) had at least one of the four typical manifestations of the disease [mean age± SD = 50.8 ± 11 years], whereas 25 (43%) who were younger[mean age = 38.7 ± 10 years, p < 0.0001] were asymptomatic at the time of the MRI.
Among the 43 symptomatic patients, 13 (30%) had a history of migraine with aura, 11 (26%) had transient ischemic attacks (TIAs) [mean age = 47 ± 13 years], 18 (42%) had one or recurrent completed strokes [mean age = 56± 9 years], and 8 (18%) previously had severe episodes of mood disturbances. At the time of the MRI, eight (18%) family members presented with dementia [mean age = 57 ± 9.5 years]. Eleven (26%) had symptoms of pseudobulbar palsy [55 ± 9.4 years]. None had oculomotor palsy or abnormal movements. The Rankin score was >3 in only seven (16%) symptomatic patients.
MRI results. Seven subjects (aged 20 to 33 years) had normal MRI results. All were asymptomatic except one who had rare attacks of migraine without aura. Sixty-eight family members (90.6%), including all symptomatic individuals, had MRI signal abnormalities on T2-WI. Forty-seven patients (62%) (33 symptomatic and 14 asymptomatic) had hypointensities on T1-WI.
Results on T1-WI. Frequency of hypointensities(signal equal to that of CSF) in the different brain regions in subjects with abnormal MRI. Sixty-nine percent of subjects with MRI signal abnormalities had at least one hyposignal on T1-WI (figure 1). The frequency of hypointensities in the white matter was 44% in the PV, 28% in the WM, and 25% in the superficial white matter. These hypointensities were also found in BG (29%): thalamus (23%), putamen (17%), pallidum (12%), and caudate (3%). They were found in internal capsule in 12% of cases and in external capsule in 24% of cases. In the brainstem (26%), they occurred in 4% of cases in the mesencephalon and in 23% in the pons.
Figure 1. Typical MR images in a patient with symptomatic cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy show hypointensities on T1-weighted image and hyperintensities on T2-weighted image located in basal ganglia and white matter.
Frequency of hypointensities according to age. The frequency of hypointensities according to age is shown in figure 2. In all areas, this frequency remains <60% regardless of age group. The frequency of hypointensities increases with age in the WM (p = 0.02), BG (p < 0.01), and IT(p = 0.0006).
Figure 2. Frequency of hypointensities on T1-weighted images according to age in the periventricular (PVH) and deep white matter (WMH), in basal ganglia (BGH), and in infratentorial areas(ITH). The frequency of hypointensities significantly increases according to age in the deep white matter, basal ganglia, and infratentorial areas.
Results on T2-WI. Frequency of hyperintensities in the different brain regions in subjects with abnormal MRI. All subjects with MRI signal abnormalities had hyperintensities located in the white matter. The signal abnormalities were found to be symmetrical in 85% of cases. The frequency of hyperintensities was 96% in the PV, 85% in the WM, and 25% in the superficial white matter. Hyperintensities were also frequent in the external capsule (53%), BG (60%), and brainstem (45%). They rarely occurred in the cerebellum (6%) or in the cortex (3%).
Frequency and severity of hyperintensities according to age (figure 3). On T2-WI, the frequency of hyperintensities was calculated in three groups according to age: <40 years (n = 32); 40 to 55 years (n = 22); and >55 years (n = 21). The frequency of hyperintensities in the PV and WM, BG, and brainstem as observed in these three groups is presented in figure 4. Hyperintensities are constant in the PV after 55 years of age. The frequency of hyperintensities increases according to age in BG (p < 0.0002) and in IT (mostly in brainstem, p < 0.001).
Figure 3. Hyperintensities of different severity(upper row: left and middle = grade B, right = grade C; lower row: grade D) at the same brain level in five different patients with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy of ages varying from 23 to 60 years.
Figure 4. Frequency of hyperintensities on T2-weighted images according to age in the periventricular (PVH) and deep white matter (WMH), in basal ganglia (BGH), and in infratentorial areas(ITH). Hyperintensities in basal ganglia and infratentorial regions are found less frequently than in the white matter, but their frequency increases according to age (BGH: p < 0.0002; ITH: p < 0.001).
Grades for global severity of MRI lesions dramatically increase with age(p < 0.0001) (figure 5). Grade A was observed only in individuals younger than 40 years. Above 55 years of age, over 80% of subjects had confluent white matter lesions (grade D). Punctiform lesions occurred only in family members younger than 55 years. Also, hyperintensities scores calculated in the PV and WM, BG, and IT significantly increase according to age (table).
Figure 5. Severity grades of hyperintensities according to age.
Table Mean scores obtained in patients for hypointensities in the periventricular (PV) and deep white matter (WM), basal ganglia (BG), and infratentorial areas (IT) using Scheltens' method*
Severity of signal abnormalities and clinical presentation. PV, BG, and IT hyperintensities scores were significantly higher in symptomatic compared with asymptomatic individuals of different ages: <40 years, 40 to 55 years, and >55 years. The WM hyperintensities score was significantly higher in symptomatic compared with asymptomatic individuals only before 40 years of age (figure 6). ANOVA showed a strong interaction between the presence of symptoms and the severity of hyperintensities score in PV (p = 0.0002), WM (p = 0.01), BG (p = 0.002), and IT (p = 0.004).
Figure 6. Severity of MRI hyperintensities in the periventricular (PVH) and deep white matter (WMH), in basal ganglia (BGH), and in infratentorial areas (ITH) according to age and to the clinical status(symptomatic versus asymptomatic). A significant difference for each score between symptomatic and asymptomatic subjects within each class of age is indicated with a star.
All patients with MRI grade A were asymptomatic. Conversely, 61% with MRI grade B and 64% with MRI grade C were symptomatic. Eighty-two percent of subjects with MRI grade D had symptoms, including all demented patients and all individuals with Rankin score >1 (figure 7).
Figure 7. Clinical status and Rankin score of all subjects according to the grade of MRI lesions (each subject is represented by a black dot).
Discussion. The clinical presentation of the symptomatic family members in our series does not differ from the usual phenotype of the disease with ischemic manifestations occurring during mid-adulthood, and dementia in the older affected members.2 Forty-three percent of the gene carriers were asymptomatic, but had MRI signal abnormalities as previously reported in numerous families affected by CADASIL.2,12-16
Our results confirm that MRI signal abnormalities are mainly located in the subcortical areas and that cortical or cerebellar lesions are extremely rare in CADASIL. This study also shows that the frequency of the lesions varies widely among different subcortical areas. When MRI results are abnormal, hyperintensities on T2-WI are found consistently in the white matter, and are much more frequent in the PV and WM than in the superficial white matter. In particular, the presence of periventricular hyperintensities(frontal and occipital caps) in 96% of the subjects suggests that the diagnosis of CADASIL should be questioned in the absence of such signal abnormalities. In contrast, hyperintensities are less frequent in BG, and even less frequent in the brainstem. This discrepancy could reflect the varying vulnerability of these regions to ischemia secondary to this arterial disease.17 Pathologic studies in CADASIL have shown that although all cerebral arteries are involved, lesions predominate in the deepest parts of the brain.17,18 Our results also confirm that the regions irrigated by the perforating arteries are the most altered in CADASIL. Furthermore, the length of the perforating arteries may play a role in the regional variations of frequency and severity of lesions as the longest perforating arteries go to the PV, the medium-sized to the BG, and the shortest to the brainstem according to the decreasing frequency of signal abnormalities within these different areas.19 The discrepancy between the pattern of MRI lesions and the systemic distribution of the vessel wall alterations previously observed in several affected patients18,20,21 supports that the ultrastructural arterial changes in CADASIL mostly alter the cerebral microcirculation.
The presence of small deep infarcts identified on T1-WI is usually a sign of small arterial diseases of the brain.22,23 However, in the current study, T2 hyperintensities in the white matter occur in the absence of hypointensities in one third of the affected individuals. Furthermore, these hyperintensities occur in 40% of the patients without typical infarcts in BG and in two thirds without similar lesions in the white matter. Thus, the diagnosis of CADASIL could be possible with only the presence of white matter hyperintensities on T2-WI. These results concur with those of Uhlenbrock et al., whose study showed that low signal intensity lesions on T1-WI were uncommon in subcortical arteriosclerotic encephalopathy and that hyperintensities on T2-WI mainly represent demyelination.23 In several affected patients of the first families identified with CADASIL, the diagnosis of multiple sclerosis or a metabolic disorder were initially suspected.14,16,24 In such cases, the predominant location of the hyperintensities in the WM and PV, its symmetric distribution, and the frequent involvement of the external capsule should be helpful in differentiating CADASIL from other disorders involving the white matter. However, these features are nonspecific in adults, as they are also observed in other conditions such as eclampsia, angiitis, and metachromatic leukodystrophy, or after radiotherapy.22 Therefore, in most cases, the clinical presentation (absence of optic and peripheral nerve involvement or medullar involvement) and the family history are essential to avoid misdiagnosis.2 Skin biopsy18,21 or genotypic analysis1 are decisive for diagnostic confirmation.
The frequency and severity of MRI lesions increase dramatically with age, so that over 55 years of age, all patients have grade C or D (the two most severe grades) white matter lesions. These results suggest that the brain lesions accumulate with age during the course of the disease. Furthermore, the results indicate that age is crucial when discussing the diagnosis of CADASIL in patients with white matter hyperintensities. Thus, the severity of MRI lesions in CADASIL dramatically differs from that reported in normal aging. In the Rotterdam study, Breteler et al. found that only 3% of their subjects aged between 65 to 69 years had confluent lesions, severe periventricular hyperintensities, or both.25
In our study, one-fifth of the subjects with diffuse white matter signal abnormalities were asymptomatic at the time of the study. However, without extensive neuropsychologic assessment, we cannot exclude a subtle cognitive deficit in these patients, as previously reported in nondemented patients with CADASIL.26 Despite the effect of age on the MRI lesions in this series, we found that signal abnormalities on T2-WI were more severe in symptomatic individuals compared with asymptomatic individuals. We also observed that all demented patients and all patients with functional disability had the most severe grade of lesions on MRI. Therefore, although the presence of diffuse white matter lesions is not systematically correlated with severity of the clinical presentation, a prospective follow-up is needed to determine its prognostic value.
Acknowledgment
We are indebted to Dr. C. Tzourio for his kind and helpful comments concerning the data analysis.
Footnotes
-
Supported in part by the Assistance Publique des Hôpitaux de Paris(Réseau EMUL 93704), Association Française contre les Myopathies, and Association pour la Recherche sur la Sclérose en Plaques.
Received October 26, 1997. Accepted in final form March 25, 1998.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan