Predictors and Clinical Impact of Incident Lacunes in Cerebral Autosomal Dominant Arteriopathy With Subcortical Infarcts and Leukoencephalopathy
Background and Purpose—Previous studies in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy showed that accumulation of lacunes strongly relate to clinical severity. However, the potential predictors of incident lacunes and their clinical consequences over a short time frame have not been investigated. This study aimed to determine the predictors and clinical impact of such lesions in a large cohort of patients.
Methods—Two hundred and six NOTCH3 mutation carriers (mean age, 49.5±10.6 years) were followed up over 3 years. Incident lacunes were identified using difference imaging from 3-dimensional T1 images. Clinical events and change in different clinical scores such as the Mattis Dementia Rating Scale, Modified Rankin Scale, Barthel index, and time to complete part A and part B of Trail Making Test were recorded. Associations were analyzed with multivariable logistic regression analysis and ANCOVA.
Results—Over a mean period of 3.4±0.7 years, incident lacunes occurred in 51 of 206 patients. Both the number of lacunes (P<0.0001) and systolic blood pressure at baseline (P<0.01) were independent predictors of incident lacunes during follow-up. The results were still significant after excluding patients with systolic blood pressure >140 mm Hg. Incident lacunes were also associated with incident stroke and with change in time to complete Trail Making Test part B, initiation/perseveration subscale of the Mattis Dementia Rating Scale and Barthel Index over the study period.
Conclusions—Systolic blood pressure and the number of prevalent lacunes are independent predictors of incident lacunes in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. These lesions mainly impact executive performances and functional independence over 3 years.
Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is the most frequent hereditary small-vessel disease of the brain. Clinical manifestations in CADASIL develop over decades and include attacks of migraine with aura, mood disturbances, ischemic strokes, and progressive cognitive decline associated with gait and balance disturbances.1 Magnetic resonance imaging (MRI) shows white matter hyperintensities2 from age 20 to 30 years, in all symptomatic individuals. Lacunes,3 microbleeds,3,4 dilated perivascular spaces, and cerebral atrophy5,6 are observed later with the progression of the disease.
Both the number of lacunes and degree of cerebral atrophy have previously been shown to be independently associated with clinical severity, particularly with motor disability6,7 and cognitive decline.7,8 These8 2 measures are also strongly related. Not only the number of lacunes is strongly associated with global cerebral atrophy at cross-sectional level but incident lacunes also promote remote cortical thinning in CADASIL individuals.9 Recently, in a large longitudinal study, the number of lacunes at baseline was found to predict incident stroke and also severe disability at 3 years, even in the absence of stroke events.10 Taken together, these observations suggest that accumulation of lacunes within the cerebral tissue plays a key role in clinical worsening during the course of CADASIL.
The relationship between incident lacunes and clinical features were previously investigated only in a limited sample of 25 CADASIL patients followed up over 7 years.3 The results suggested that these lesions might be associated with the reduction of executive and global cognitive performances at the end of the study period. In a larger cohort, the spatial distribution of incident lacunes has been also recently analyzed.11 The potential predictors and clinical consequences of such lesions over a short time frame as might be expected in a future clinical trial still need to be determined.
Herein, we analyzed the predictors of incident lacunes and their impact on clinical worsening in a large cohort of CADASIL patients followed up over 3 years.
Patients were recruited through a prospective CADASIL study obtained in France and Germany.10 In all subjects, the diagnosis of CADASIL was confirmed by genetic testing. All patients underwent both MRI and clinical examination at study inclusion and at 3-year follow-up. Details of the study protocol have been reported elsewhere.4
Clinical and demographic data were collected at study entry and included age, sex, years of education, and vascular risk factors. History of hypertension was defined as previous diagnosis of hypertension (blood pressure >140/90 mm Hg) or use of antihypertensive treatment for control of blood pressure. Blood pressure was also measured at baseline and at follow-up. A history of transient ischemic attacks or stroke, gait disturbance, and dementia defined according to Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, criteria was obtained at baseline and at follow-up. Subjects were interviewed and asked for cognitive complaints and any permanent difficulties in gait and balance. They also underwent a detailed neurological and neuropsychological examination at baseline and at 3 years. Global cognitive function was assessed using the Mattis Dementia Rating Scale (MDRS). The initiation/perseveration subscale of the MDRS (MDRS-I/P) was analyzed separately. Executive functions were assessed by the time to complete part B of Trail Making Test (TMTB). Part A of the same test was used in parallel to evaluate simple speed processing. Disability was assessed using the modified Rankin scale (mRS). Moderate or severe disability was defined as mRS score ≥3. Functional independence was determined using the Barthel index.
Written informed consent was obtained from all participants or a close relative if the patient was too severely disabled. This study was approved by the ethics committees of both participating institutions (Paris and Munich).
Scanner details and MRI sequences have been reported elsewhere.4 The protocol included millimetric 3-dimensional T1-weighted images, axial slices of 5 mm thickness of fluid-attenuated inversion recovery images, and T2*-weighted gradient-echo planar images. Brain MRI was performed at baseline and repeated after 3 years by using the same protocol at each center. Validated methods were used by board-certified neurologists and neuroradiologists for evaluating white matter hyperintensities, lacunes, microbleeds, and brain volumes as previously described,4,12 who were blinded to the clinical details. Special care was taken to exclude enlarged perivascular spaces.
Identification of Baseline and Incident Lacunes
Baseline lacunes were segmented manually using 3-dimensional T1 images by 2 trained raters. Hypointense lesions on T1-weighted images with a signal identical to cerebrospinal fluid, sharp delineation, and a diameter >2 mm were selected for this segmentation as previously reported.9 In difficult cases, a consensus was obtained after reviewing all MRI data available. The number of lacunes was thereafter counted.
Incident lacunes were identified using difference imaging from 3-dimensional T1 images obtained at baseline and at 3-year follow-up as detailed previously.9 Briefly, after intensity normalization and image registration, the T1 scan at baseline was subtracted from the follow-up T1 scan, thereby highlighting incident hypointensities as additional dark intraparenchymal areas (Figure) on a homogenous background. The location of each incident lacune was also recorded as previously reported.11
The baseline characteristics of the population was compared between patients with and without incident lacunes using χ2 test for dichotomous clinical variables and Student t test for continuous variables. Adjusted multivariable logistic regression model was used to find predictors of incident lacunes; candidate covariates were those associated with incident lacunes in univariate analysis (P<0.15). For continuous variables, the decision as to whether to treat the variable as a continuous or categorical variable was based on the lowest value of the Akaike information criterion for the corresponding univariate logistic regression model. MRI parameters (number of lacunes, normalized volume of white matter hyperintensities, and brain parenchymal fraction) were dichotomized according to their median values. Microbleeds were dichotomized according to their presence or absence.
Incident strokes were assessed irrespective of whether the subject already experienced a stroke. Incident dementia and incident moderate or severe disability were considered only in individuals without dementia or without moderate or severe disability at baseline (mRS 0–2). Two multivariable logistic regression models were used to test the potential impact of incident lacunes on clinical events during follow-up: model 1 was adjusted for age and sex, and model 2 was adjusted for age, sex, and MRI parameters at baseline.
Change of clinical scores (MDRS, MDRS-I/P, Barthel index, and mRS) and of time from completion of part A and part B of Trail Making Test were calculated by subtracting the baseline value from that obtained at 3-year follow-up. The association between incident lacunes and modifications of clinical scores during follow-up was tested using ANCOVA with adjustment for age, sex, or MRI parameters as in models 1 and 2.
Tests were 2 sided. The significance level was fixed at 5%. All analyses were conducted using SAS (release 9.3; SAS Statistical Institute, Cary, NC).
Two hundred and six patients who had a complete set of MRI sequences (both at baseline and 3 years) with high-quality images allowing the different measures were included in this study. The baseline characteristics of this population are detailed in Table 1. Half of individuals were <50 years of age, and one fifth were treated for hypertension or diagnosed hypertensive at baseline. The mean systolic blood pressure (SBP) in the population was 126 mm Hg, and the mean diastolic blood pressure was 75 mm Hg.
Main Characteristics of Incident Lacunes
The mean time (±SD) between the baseline and follow-up examination was 3.4±0.7 years. Overall, 51 of the 206 patients (24.8%) included in the study presented one or more incident lacunes during the follow-up period (total number of incident lacunes=80). Among patients with incident lacunes, 35 (68.6%) had 1, 9 (17.6%) had 2, and 7 (13.7%) had 3 to 10 incident lacunes. Incident lacunes were most frequently detected in the centrum semiovale (n=26), basal ganglia (n=15), and within the frontal pole (n=10). They were less frequent in other areas (corpus callosum [n=9], occipital pole [n=7], thalamus [n=5], and others [n=8]).
Baseline Predictors of Incident Lacunes Over 3 Years
The main characteristics of patients with and without incident lacunes are summarized in Table 2. Cardiovascular risk factors including the frequency of hypertension did not differ between the 2 groups. However, in univariate analysis, the mean SBP at baseline was 5 mm Hg higher in patients with incident lacunes than in others (P=0.05). This difference was confirmed when SBP was analyzed by tertiles. The percentage of patients having their baseline SBP >130 mm Hg among individuals with incident lacunes was nearly twice that observed in the other group (49% versus 27%; P=0.018). In addition, the number of lacunes at baseline was also twice higher in patients with incident lacunes (P<0.0001). In the adjusted multivariable logistic regression model considering both demographic and clinical factors, as well as MRI parameters, both the baseline SBP (tertile ≥130 mm Hg; odds ratio=2.7; 95% confidence interval: 1.1–6.9) and number of lacunes (odds ratio=7.4; 95% confidence interval: 2.7–20.3) were independently associated with incident lacunes. To determine whether these results are also observed in patients who had their SBP at entry within the normal range, we excluded all individuals having SBP >140 mm Hg. The 2 previous factors remain highly significant (Table 3).
Impact of Incident Lacunes on Clinical Manifestations Over 3 Years
During follow-up, incident stroke (all were presumably of ischemic type in the absence of recent hemorrhage on computed tomography scan or MRI) occurred in 40 of 202 patients (19.8%). Moderate or severe disability occurred in 9 of 179 subjects (5.0%) and incident dementia in 12 of 171 patients (7.0%).
In univariate analysis, incident strokes occurred twice as often in patients with incident lacunes compared with those without (P=0.013). After adjustment for age and sex, incident strokes were still independently associated with incident lacunes (odds ratio=2.6; 95% confidence interval: 1.2–5.6; model 1). However, when further adjusting for the baseline MRI parameters, this association did not remain significant (P=0.068; model 2; Table 4). Although nonsignificant, worsening of disability or incident dementia occurred more than twice as frequent in patients with incident lacunes compared with other individuals.
Impact of Incident Lacunes on Clinical Scores Over 3 Years
In the group of patients without incident lacunes, a significant decrease of the Barthel index score and increase of the mRS score were detected during follow-up. In the group of patients with incident lacunes, not only a larger variation of these clinical scales was observed but also a significant decrease of the total MDRS score and of the MDRS-I/P subscore at 3 years (Table 5).
In univariate analysis, the decrease of the MDRS-I/P score was found 4-fold larger in the presence of incident lacunes than in their absence (P=0.0024), and the reduction of the Barthel index was also more than 10 times larger in patients with incident lacunes (P=0.011). A significant increase of TMTB time was observed in patients with incident lacunes, whereas a slight decrease was detected in those without (P=0.0002). Adjustment for age, sex, and education level did not change these results (model 1). After additional adjustment for baseline MRI parameters (model 2), only the association between incident lacunes and increase of TMTB time was still significant.
There are 3 main findings in the present study. First, incident lacunes are observed in ≈1 in 4 individuals over 3 years in CADASIL patients. Second, both the level of SBP and number of lacunes on MRI observed at baseline are major and independent predictors of these lesions. Third, incident lacunes obviously impact the clinical course of the disease and contribute to the development of motor and cognitive decline.
Incident lacunes were detected in ≈7.3% patients per year in the present study. This is more than twice the annual incidence observed in the Cardiovascular Health study13 or in the Rotterdam Scan study,14 2 large population-based cohorts of elderly individuals with frequent sporadic small-vessel disease. As previously reported,11 incident lacunes were also detected much more frequently than incident strokes in CADASIL patients. In the present study, only one third of individuals with incident lacunes had a clinical stroke during follow-up. Also, the youngest subject of the cohort whose MRI showed an incident lacune was 35 years old and did never report stroke manifestations. These data indicate that not only incident lacunes occur early and much more frequently than stroke events during the course of CADASIL but also severe damage of cerebral tissue can progress in the absence of acute clinical manifestations. On the contrary, although incident strokes were found more than twice as frequent in patients with incident lacunes, ≈1 in 6 individuals without any incident lacune reported typical stroke manifestations during follow-up. This further illustrates that ischemic stroke manifestations in CADASIL are not necessarily related to complete tissue infarction with cavitation in line with previous studies showing that about one third of acute ischemic lesions of lacunar type on diffusion-weighted images or T2-weighted sequence do not develop a cavity.15,16
We previously showed that the load of lacunes was a strong predictor of ischemic stroke over 3 years in a prospective study of CADASIL patients. Herein, we show that the number of lacunes is also a strong predictor of incident lacunes over the same time frame, as previously suggested in a smaller group of patients followed up for 7 years.3 This effect remains highly significant after adjustment for age, sex, and even after considering the other MRI markers of the disease. These findings emphasize that a higher load of lacunes at baseline predicts a faster progression of cerebral lesions during follow-up as already suggested in a predictive model of clinical worsening.10 They also support that, at a certain stage of the disease, the pathogenic mechanism leading to accumulation of lacunes may either accelerate rapidly or promote itself an exponential progression of focal ischemic lesions.
In this study, we observed that baseline SBP is a strong and independent predictor of incident lacunes in CADASIL patients. Hypertension has been repeatedly shown to promote the extension of white matter hyperintensities and incident small deep infarcts in sporadic small-vessel disease, particularly in elderly populations.17,18 In our cohort, the predictive effect of SBP on incident lacunes was detected while the mean blood pressure in the whole cohort was only 126 mm Hg. In addition, only 26 patients had their SBP >140 mm Hg at entry in the study. This small number of true hypertensive subjects at baseline and the management of all vascular risk factors as recommended during the study period might explain the lack of association between incident lacunes and hypertension in our results. Interestingly, exclusion of all subjects with SBP >140 mm Hg from multivariable analysis did not alter the predictive effect of SBP level on incident lacunes. Patients with SBP between 130 and 140 mm Hg had a 4-fold higher risk to develop incident lacunes than patients with SBP <120 mm Hg. SBP level within the normal range was also previously related to the load of microbleeds in a large cross-sectional study of CADASIL patients.4 SBP was also found to predict altered cerebral microstructure (as reflected in an increase in water diffusivity)19 and brain atrophy5 in smaller samples of CADASIL individuals who were mostly normotensive. Taken together, all these data strongly suggest that SBP level, even in the absence of hypertension, may well influence the natural history of cerebral lesions in CADASIL. The lack of association between baseline SBP and clinical worsening in a previous analysis does not contradict this interpretation. The influence of blood pressure level on the clinical course may be masked by other confounding factors or much stronger predictors as the number of lacunes at baseline. It might be also mediated by incident lacunes and more difficult to establish.
Our analysis showed a larger progression of clinical manifestations and scores in the group of patients with incident lacunes at the univariate level. Not only stroke events were more frequent, but also moderate or severe disability and dementia occurred twice more often in this group compared with individuals without incident lacunes. However, after adjustment for age, sex, and education level, a larger cognitive decline was detected in individuals with incident lacunes only using tests related to executive performances and speed processing such as the MDRS-I/P subscore and TMTB time. Alterations of executive functions and reduced cognitive processing speed are cognitive hallmarks of CADASIL.20 Recent studies showed that prevalent lacunes are associated with executive dysfunction as assessed by the TMTB time and the Stroop interference subtest in this disorder21 and that the load of such lesions can predict cognitive decline as assessed by the MDRS-I/P subscore at follow-up.10 The key role of lesion location on cognition in CADASIL has also been recently documented.22 The frequent location of incident lacunes in frontal lobes, basal ganglia, and thalamus in the present cohort may explain alterations of strategic networks involved in executive control, particularly the dorsolateral prefrontal–subcortical circuit23 that connects the prefrontal cortex to various subcortical and cortical areas.24 In contrast, when we analyzed global clinical scales or the occurrence of dementia after 3-year follow-up, only a significant association between incident lacunes and the Barthel index was detected. This discrepancy may have different sources. Accumulation of lacunes may not be the only contributor to cognitive and motor decline in CADASIL. White matter lesions, increase of microstructural tissue loss, accumulation of microbleeds, or focal atrophy processes may also be involved in the occurrence of clinical worsening. Elsewhere, aging processes may also participate to the acceleration of tissue lesions and clinical aggravation.
The prospective design, large number of patients, use of validated clinical scales, and imaging methods following the best clinical trial standards are the main strengths of this study. There are also various potential limitations. A selection bias is highly probable because of the exclusion of the most severe cases who could not return for follow-up visits or who were excluded because of head movements and insufficient quality images. Other limitations are the lack of detailed analysis of antihypertensive drugs and of variations of blood pressure during follow-up, and the lack of diffusion tensor imaging data that can provide fully automated and robust imaging measures of white matter tract alterations in small-vessel disease.25
Finally, this study shows that incident lacunes are more frequent than stroke events during the course of CADASIL, have a significant impact on clinical worsening, and may be related to SBP level. Additional studies are needed to further understand other potential mechanisms involved in clinical worsening and for determining whether a reduction of SBP can have a beneficial effect during the course of the disorder.
We acknowledge all patients who participated in this research, their families, and the association CADASIL France for their active collaboration.
Sources of Funding
This study was supported by grants from the French Ministry of Health (Regional and National PHRC AOR 02-001), Association de Recherche en NEurologie Vasculaire, the Vascular Dementia Research Foundation, and the Fondation Leducq (Transatlantic Network of Excellence on the Pathogenesis of Small Vessel Disease of the Brain; http://www.fondationleducq.org). Dr Ling is funded by the French Chinese Foundation for Science and Applications and the China Scholarship Council.
- Received October 25, 2016.
- Revision received December 4, 2016.
- Accepted December 7, 2016.
- © 2016 American Heart Association, Inc.
- Viswanathan A,
- Guichard JP,
- Gschwendtner A,
- Buffon F,
- Cumurcuic R,
- Boutron C,
- et al
- Jouvent E,
- Duchesnay E,
- Hadj-Selem F,
- De Guio F,
- Mangin JF,
- Hervé D,
- et al
- Chabriat H,
- Hervé D,
- Duering M,
- Godin O,
- Jouvent E,
- Opherk C,
- et al
- Duering M,
- Csanadi E,
- Gesierich B,
- Jouvent E,
- Hervé D,
- Seiler S,
- et al
- Wardlaw JM,
- Smith EE,
- Biessels GJ,
- Cordonnier C,
- Fazekas F,
- Frayne R,
- et al
- Longstreth WT Jr,
- Dulberg C,
- Manolio TA,
- Lewis MR,
- Beauchamp NJ Jr,
- O’Leary D,
- et al
- Vermeer SE,
- Den Heijer T,
- Koudstaal PJ,
- Oudkerk M,
- Hofman A,
- Breteler MM
- Loos CM,
- Staals J,
- Wardlaw JM,
- van Oostenbrugge RJ
- Gouw AA,
- van der Flier WM,
- Fazekas F,
- van Straaten EC,
- Pantoni L,
- Poggesi A,
- et al
- Holtmannspötter M,
- Peters N,
- Opherk C,
- Martin D,
- Herzog J,
- Brückmann H,
- et al
- Liem MK,
- van der Grond J,
- Haan J,
- van den Boom R,
- Ferrari MD,
- Knaap YM,
- et al
- Baykara E,
- Gesierich B,
- Adam R,
- Tuladhar AM,
- Biesbroek JM,
- Koek HL,
- et al