Association of NOTCH3 Variant Risk Category With 2-Year Clinical and Radiologic Small Vessel Disease Progression in Patients With CADASIL

Background and Objectives Pathogenic variants in NOTCH3 are the main cause of hereditary cerebral small vessel disease (SVD). SVD-associated NOTCH3 variants have recently been categorized into high risk (HR), moderate risk (MR), or low risk (LR) for developing early-onset severe SVD. The most severe NOTCH3-associated SVD phenotype is also known as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). We aimed to investigate whether NOTCH3 variant risk category is associated with 2-year progression rate of SVD clinical and neuroimaging outcomes in CADASIL. Methods A single-center prospective 2-year follow-up study was performed of patients with CADASIL. Clinical outcomes were incident stroke, disability (modified Rankin Scale), and executive function (Trail Making Test B given A t-scores). Neuroimaging outcomes were mean skeletonized mean diffusivity (MSMD), normalized white matter hyperintensity volume (nWMHv), normalized lacune volume (nLV), and brain parenchymal fraction (BPF). Cox regression and mixed-effect models, adjusted for age, sex, and cardiovascular risk factors, were used to study 2-year changes in outcomes and differences in disease progression between patients with HR-NOTCH3 and MR-NOTCH3 variants. Results One hundred sixty-two patients with HR (n = 90), MR (n = 67), and LR (n = 5) NOTCH3 variants were included. For the entire cohort, there was 2-year mean progression for MSMD (β = 0.20, 95% CI 0.17–0.23, p = 7.0 × 10−24), nLV (β = 0.13, 95% CI 0.080–0.19, p = 2.1 × 10−6), nWMHv (β = 0.092, 95% CI 0.075–0.11, p = 8.8 × 10−20), and BPF (β = −0.22, 95% CI −0.26 to −0.19, p = 3.2 × 10−22), as well as an increase in disability (p = 0.002) and decline of executive function (β = −0.15, 95% CI −0.30 to −3.4 × 10−5, p = 0.05). The HR-NOTCH3 group had a higher probability of 2-year incident stroke (hazard ratio 4.3, 95% CI 1.4–13.5, p = 0.011), and a higher increase in MSMD (β = 0.074, 95% CI 0.013–0.14, p = 0.017) and nLV (β = 0.14, 95% CI 0.034–0.24, p = 0.0089) than the MR-NOTCH3 group. Subgroup analyses showed significant 2-year progression of MSMD in young (n = 17, β = 0.014, 95% CI 0.0093–0.019, p = 1.4 × 10−5) and premanifest (n = 24, β = 0.012, 95% CI 0.0082–0.016, p = 1.1 × 10−6) individuals. Discussion In a trial-sensitive time span of 2 years, we found that patients with HR-NOTCH3 variants have a significantly faster progression of major clinical and neuroimaging outcomes, compared with patients with MR-NOTCH3 variants. This has important implications for clinical trial design and disease prediction and monitoring in the clinic. Moreover, we show that MSMD is a promising outcome measure for trials enrolling premanifest individuals.


Introduction
Small vessel disease (SVD) is an important cause of stroke and vascular dementia. 18][9] The most severe end of the NOTCH3-associated SVD (NOTCH3-SVD) spectrum is also known as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), which has a minimal prevalence of 2-5:100,000 individuals worldwide. 102][13][14][15][16][17][18] Variants in NOTCH3 EGFr domains can be stratified into 3 risk categories, namely low risk (LR), moderate risk (MR), and high risk (HR) for developing severe NOTCH3-SVD. 13igh-risk NOTCH3 variants (HR-NOTCH3) are predominantly found in patients with CADASIL, whereas low-risk NOTCH3 variants (LR-NOTCH3) are most frequent in healthy-volunteer population databases, such as UK Biobank.Moderate-risk NOTCH3 variants (MR-NOTCH3) are prevalent in both CADASIL and population cohorts. 13oaggregatory NOTCH3-SVD variants are almost exclusively cysteine-altering missense variants predicted to disrupt intra-EGFr disulfide bridge formation. 19This ultimately leads to NOTCH3 aggregation in the extracellular matrix of the (cerebral) microvasculature. 20Although the molecular mechanisms underlying the association between NOTCH3 variant risk category and disease severity have not yet been elucidated, it has been shown that patients with CADASIL with HR-NOTCH3 variants have higher levels of NOTCH3 aggregation in skin and brain vessels than patients with MR and LR-NOTCH3 variants. 13,21This strongly supports the theory that NOTCH3 aggregation is the pathomechanistic driver of NOTCH3-SVD, 22 setting into motion a cascade of events leading to destruction of cerebral small vessel wall integrity, and thereby reduced cerebrovascular reactivity 23 and cerebral perfusion. 24inical manifestations of severe NOTCH3-SVD, or CADASIL, are mid-adult onset of ischemic stroke and transient ischemic attacks, migraine with aura, mood disorders, and cognitive decline, ultimately leading to vascular dementia. 25Neuroimaging features are progressive symmetrical white matter hyperintensities (WMHs), lacunes, cerebral microbleeds, perivascular spaces, and atrophy.White matter hyperintensities can precede clinical manifestations by decades. 26Diffusion tensor imaging (DTI) has been shown to sensitively capture white matter tract alterations, even in areas with normal-appearing white matter on T2-weighted imaging, 27 suggesting that DTI may provide readily quantifiable biomarkers in early CADASIL/NOTCH3-SVD stages.

Study Participants
Participants of the Disease Variability in NOTCH3-Associated Small Vessel Disease (DiViNAS) study were recruited from the Dutch CADASIL registry, which consists of patients and presymptomatic or paucisymptomatic family members with a genetically confirmed cysteinealtering NOTCH3 variant.All patients were 20 years or older.Participants visited the LUMC at baseline 11 and at 2-year follow-up between May 2019 and December 2023.Details concerning inclusion are presented in Figure 1.At both time points, participants were characterized on a single day, with cerebral MRI, neuropsychological testing, medical history, skin biopsy, and blood withdrawal.If patients were unable to participate in the on-site follow-up study, they were asked whether they were willing to provide 2-year follow-up information through a telephone interview and/or their medical records.This article follows the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines. 30andard Protocol Approvals, Registrations, and Patient Consents The DiViNAS study was approved by the Medical Ethics Committee Leiden-The Hague-Delft (P18.164 and P21.013).All participants gave written informed consent.All procedures were performed in accordance with ethical rules and the principles of the Declaration of Helsinki.

Mortality, Clinical, and Neuropsychological Measures
To determine whether patients were deceased during the follow-up period, the Dutch Personal Records Database was queried.For all patients, clinical information concerning cause of death and the disease course before death was available.Causes of death and premorbid disease course were further detailed by requesting medical records and by interviews with relatives.Participants were assessed for a history of hypertension, hypercholesterolemia, diabetes type 1 or 2, and smoking status (see Table 1 and eMethods for details).Clinical outcomes were defined as previously described. 11Briefly, incident stroke during the follow-up period was assessed using medical history and was defined as either neurologic deficits that lasted longer than 24 hours in the absence of other probable causes or as a diagnosis of ischemic stroke in the medical records; disability was assessed using the modified Rankin Scale (mRS) questionnaire.A neuropsychological test battery was performed at both time points and included the Trail Making Test (TMT) Parts A and B. Scores of the TMT B were corrected for age, sex, educational level, and TMT A-score (TMT Part B given A t-scores, TMT B/ A ) using normative data reported in the literature. 31Patients who were unable to finish the TMT Part A or B in time (<300 seconds) because of too severe cognitive deficits (n = 7) were scored as the lowest t-score in the cohort, which was equal to 15.One patient was unable to complete the TMT because of impaired vision.mRS was available for 146 individuals and TMT B/A for 132 at both time points (Figure 1).

Neuroimaging Outcomes
Brain MRI scans at both time points were performed on the same 3T MR scanner (Philips Achieva TX, Philips Medical Systems, Best, the Netherlands) as described previously. 11riefly, the following sequences were included in the protocol: 3D-T1-weighted (T1w) images, 3D-T2-weighted images, T2 fluid-attenuated inversion recovery (FLAIR), susceptibility-weighted images (SWIs), and DTI (eMethods).SVD neuroimaging markers were quantified at both time points (eMethods).Briefly, for longitudinal image registration, a template was created per patient using T1w images of both time points as input.T1w images were rigidly registered to this template, and FLAIR, SWI, and TRACE (i.e., mean of diffusionweighted images) were affine registered to the T1w images of the same time point in the template space.Longitudinally registered images were subtracted for each modality to calculate difference maps across time points for visual rating.Incident lacunes were scored according to the STRIVE criteria 32 and were segmented to calculate changes in total lacune volume (LV) over time.Lacunes and LV at baseline were assessed as previously described. 11Hs were segmented using a fully automated segmentation approach using FLAIR and registered 3D-Tw images as input.Intracranial and brain parenchymal volumes were determined from longitudinally registered T1w and FLAIR images.LV, WMH volume (WMHv), and brain volume were normalized to intracranial volume ([WMHv, LV and brain volume/intracranial volume] × 100) to calculate normalized WMHv (nWMHv), normalized LV (nLV), and brain parenchymal fraction (BPF).
Mean skeletonized mean diffusivity (MSMD) was chosen a priori as a marker for microstructural white matter damage (eMethods), as peak width of the skeletonized mean diffusivity 27 (PSMD) is prone to software updates and thus not as robust in longitudinal analysis as MSMD on Philips MRI scanners. 33MSMD was calculated using a publicly available script and software container (eMethods).An additional sensitivity analysis was performed for PSMD.Three participants were excluded from MRI scanning at one or both time points because of contraindications.For the longitudinal analysis of MRI markers using both time points, nLV was available for 130 participants, nWMHv and BPF for 126 participants, and MSMD for 118 participants (Figure 1).Incident lacunes, nWMHv, and BPF were analyzed at 2 independent sites (Medical Image Analysis Center and Leiden University Medical Center), and intersite pipeline replicability was determined (eMethods).The MRI investigators (M.D., B.G., R.H. and M.C.) were blinded to the NOTCH3 variant risk category.

Statistics
Variables with a normal distribution were described using mean ± SD; variables with a skewed distribution were described using median ± interquartile range (IQR).In accordance with the STROBE guidelines, 30 inferential measures including p-values were not reported in the baseline characteristics table.To reduce the number of predictors, instead of defining each cardiovascular risk factor separately, a variable number of CVRF (CVRFn) was created, which was defined as the total number of CVRF at baseline for each individual.CVRF included were hypertension, hypercholesterolemia, diabetes type 1 or 2, and current smoking status (for definitions, see eMethods).To analyze 2-year differences in cumulative incident stroke probability between the HR-NOTCH3 b The effect of NOTCH3 risk category on stroke probability and progression of MSMD and nLV remained statistically significant after correction for multiple testing using the Benjamini-Hochberg procedure (q-value = 0.039 for all 3 outcomes).c Overall p-value derived from univariable mixed-effect models with follow-up time as covariate; if there was a significant difference in progression between the HR and MR-NOTCH3 risk categories ( d ), a p-value for time per risk category was also reported (p HR-NOTCH3 and p MR-NOTCH3 , derived from the linear mixed model below d ).For statistical analyses, MSMD was log-transformed, nWMHv was square root-transformed, and nLV was cube root-transformed.Statistical testing of continuous and ordinal outcomes was performed using mixed-effects models (MMs) with a random intercept.For continuous outcomes, these were linear mixed models (LMMs), and for the ordinal outcome mRS, this was a cumulative link mixed model (CLMM).Participants who had only one available observation for a particular outcome (either at baseline or follow-up) were also included in the MMs.Missing data were assumed to be missing at random.To achieve normal distribution or homoscedasticity of residuals, the following transformations were performed: natural logarithmic transformation of MSMD, cube root transformation of nLV, and square root transformation of nWMHv.For the LMMs, continuous dependent variables were standardized by subtracting the mean and dividing the value by the standard deviation.
For hypothesis testing, likelihood ratio testing was performed comparing MMs with and without the relevant interaction terms with follow-up time as a continuous covariate.To test whether the observed 2-year changes were statistically significant, a first MM was created per outcome with follow-up time, and compared with a null model.To assess the effect of NOTCH3 risk category on 2-year changes in study outcomes, an MM was then created with NOTCH3 risk category and its interaction with follow-up time and the following covariates: sex, baseline age, and baseline CVRFn.To test whether CVRFn and sex also influenced progression, an MM was fitted with additional interaction terms between CVRFn and followup time as well as sex and follow-up time.In case of differential progression between the HR and MR-NOTCH3 groups (i.e., a significant interaction term between NOTCH3 risk category and follow-up time), post hoc t-tests were performed within the MMs to test whether progression was also significant within each NOTCH3 risk category group.
As previous studies have shown that DTI may detect microstructural alterations in normal-appearing white matter 27 and DTI outcomes would require the smallest sample size for demonstrating a treatment effect in CADASIL, 34 additional subgroup analysis of progression of MSMD was performed using a paired t-test of log-transformed MSMD in the following groups: (1) patients younger than 40 years and (2) in premanifest individuals, defined as only minimal deep white matter hyperintensities (Fazekas score <2) and no lacunes at baseline.Baseline descriptives of the analyzed subgroups are provided in eTable 2.
Standardized coefficients (β), odds ratios (ORs) (for LMM: β time and β time × NOTCH3 , and for CLMM: OR time × NOTCH3 ), and hazard ratios were reported with 95% CIs and p-values.In the LMMs, a sensitivity analysis that included pedigree as an additional random effect was performed.If β time × NOTCH3 was significant, sensitivity analyses were performed with models that included the following additional interaction terms: (1) between follow-up time and each cardiovascular risk factor as an independent predictor and (2) between follow-up time and age to examine whether statistical inferences of NOTCH3 risk category would remain significant.Additional correction for multiple testing of the effect of NOTCH3 risk category on all outcomes (n = 7) was performed using a Benjamini-Hochberg procedure, and q-values were reported in Table 2. Two-sided p-values and q-values below 0.05 were considered significant.Statistical analysis was performed using R version 4.2.1.

Data Availability
Data supporting the findings presented in this article are available on reasonable request from the corresponding authors.

Results
Baseline Descriptives and Loss to Follow-Up Follow-up data were obtained of 162 patients of the 186 baseline DiViNAS participants.This included either an onsite visit for the full 2-year follow-up research protocol (n = 133) or clinical data acquisition through clinical records and/or telephone interviews (n = 29).Twenty-four patients were lost to follow-up (Figure 1).Of the 162 follow-up participants, 90 had an HR-NOTCH3 variant, 67 an MR-NOTCH3 variant, and 5 an LR-NOTCH3 variant (eTable 1).
The median follow-up duration was 24.7 months (IQR 2.6 months, Table 1).At baseline, patients with MR-NOTCH3 variants were older, had lower educational levels, and had a higher burden of CVFR compared to patients with HR-NOTCH3 variants (Table 1).HR and MR-NOTCH3 participants did not differ in any of the other baseline descriptives.

Two-Year Mortality
Three patients with an HR-NOTCH3 variant and 3 with an MR-NOTCH3 variant died in the interim.Four of these patients died because of CADASIL-related causes, namely stroke or end-stage CADASIL (HR-NOTCH3 ages 55 and 73 years, MR-NOTCH3 ages 71 and 73 years).Another patient died at 60 years due to myelofibrosis and one at 59 years due to (likely) acute myocardial infarction.
Changes in TMT B/A were more pronounced in older patients (eFigure 2).There were no differences in progression rate of TMT B/A between the 2 NOTCH3 risk category groups (β time × NOTCH3 = −0.024,95% CI −0.33 to 0.28, p = 0.87).Sex and CVRFn did not significantly affect the 2-year change in TMT B/A or mRS.Other CADASIL-related clinical signs and symptoms are summarized in eTable 3.
In a sensitivity analysis, the inference of the effect of NOTCH3 risk category on 2-year progression of MSMD and nLV was not changed when including interactions between follow-up time and age or follow-up time and each CVRF separately or when including pedigree of origin.The effect of NOTCH3 risk category on incident stroke probability and progression of MSMD and nLV remained significant after correction for multiple testing (Table 2).Although the HR-NOTCH3 group progressed significantly faster on several neuroimaging measures than the MR-NOTCH3 group, there was still considerable variability within each NOTCH3 risk category.Illustrative examples of variability in 2-year neuroimaging disease progression for both NOTCH3 risk categories are shown in Figure 4.

Discussion
In this prospective single-site 2-year follow-up study stratifying for NOTCH3 variant risk category, we show that NOTCH3 risk category is a major predictor of the rate of clinical and neuroimaging disease progression in patients with CADASIL.5][36][37][38][39][40][41][42][43][44][45][46] For the whole cohort, clinical and neuroimaging outcome measures significantly progressed in a trial-sensitive time frame of 2 years.Patients with HR-NOTCH3 variants, however, were shown to progress significantly faster than patients with MR-NOTCH3 variants with respect to 2-year incident stroke, MSMD, and nLV, after controlling for sex, age, and cardiovascular risk factors at baseline.
Of all outcome measures included in the study, the microstructural marker MSMD showed the most significant 2-year change for the whole cohort, which is in line with a previous study that showed that diffusion MRI had the smallest sample size estimates among clinical and neuroimaging markers in CADASIL. 34Diffusion tensor imaging measures have previously been shown to correlate strongly with cognitive outcomes in CADASIL and in SVD in general. 27The 2-year increase in MSMD was even significant in young (<40 years) and premanifest individuals.Moreover, there was a significant difference in 2-year MSMD progression between the HR and MR-NOTCH3 groups, with MSMD increasing 46% more in the HR-NOTCH3 group.Differences in progression of neuroimaging markers between the HR and MR-NOTCH3 groups were most pronounced for normalized lacune volume (previously shown to predict clinical worsening in CADASIL 42,44 ), which only showed a significant 2-year increase in the HR-NOTCH3 group, with half of the patients having at least 1 incident lacune.We found no significant differences between patients with HR and MR-NOTCH3 variants for 2-year progression of nWMHv or BPF.There is evidence that WMH may have a heterogeneous etiology in CADASIL, with some hyperintensities being edematous in To ensure participant anonymity, age ranges are given and sex is not specified.
nature while others are probably ischemic. 47,48BPF measures in patients with CADASIL have shown some counterintuitive outcomes, with patients with more severe phenotypes having higher BPF values (suggesting less brain atrophy), which may be explained by the presence of brain swelling masking brain atrophy. 48tients with HR-NOTCH3 variants had a 4-fold higher incident stroke probability than patients with MR-NOTCH3 variants, and the lifetime stroke rate was higher in patients with HR-NOTCH3 variants independent of the number of previous strokes.The 2-year incidence of stroke was 13.4%, which is lower than what has been previously reported in studies with similar follow-up durations (19.8%-22%). 36,43his may be attributable to differences in standards of medical care, for example, indication for referral to a neurologist or performing an MRI and, therefore, a stroke diagnosis.A high proportion of patients did have incident lacunes (32.8%), which suggests that, similar to previous studies, 41,44 covert infarction was frequent in our cohort.
In a multivariable model, only NOTCH3 risk category and age were predictors of 2-year cumulative stroke probability.In line with previously published studies, 36,43,44 there was a significant 2-year decline of disability and a borderline significant decline in executive function.There was no significant difference in these outcomes between the NOTCH3 risk categories.Likely, measures for disability and cognition are not sensitive enough to capture differences between the groups in such a short time frame.A previous study using a computational approach found a stronger predicted progression on the Matthias Dementia Rating Scale for patients with variants in EGFr domains 7-34 (i.e., predominantly MR and LR-NOTCH3 variants), compared with patients with EGFr domains 1-6 (i.e., HR-NOTCH3 variants). 29This is in contrast to our results, as we found no single neuroimaging, clinical, or cognitive outcome which progressed faster in the MR-NOTCH3 group than in the HR-NOTCH3 group.This is likely attributable to differences in study design (i.e., a computational vs longitudinal study design) and in NOTCH3 risk category classification (i.e., using EGFr 1-6 vs EGFr 7-34 12 instead of the updated 3-tiered NOTCH3 risk category classification 13 ).
This study has confirmed that, even in a short follow-up period, clinical and neuroimaging disease progression is significantly faster in HR-NOTCH3 patients than in MR-NOTCH3 patients.This has important implications for the development of disease guidelines, supporting a differential approach to HR vs MR-NOTCH3 patients concerning frequency of disease monitoring and disease management, as well as improved individualized disease prediction.Patients with HR-NOTCH3 variants may benefit from increased clinical surveillance compared with patients with MR-NOTCH3 variants.Although a patient-centered approach, which, for example, includes age, assessment of cardiovascular risk factors, and disease stage at diagnosis should guide clinical decision making, it is clear that NOTCH3 risk category is an important factor to take into account in individuals genetically diagnosed with a cysteine-altering NOTCH3 variant.
HR-NOTCH3 variants have previously been shown to be associated with higher vascular NOTCH3 protein aggregation load. 13,21Given the faster rate of short-term disease progression in patients with HR-NOTCH3 variants, it is tempting to speculate that disease progression is mediated by vascular NOTCH3 aggregation.This potential causal relation between disease progression and increase in vascular NOTCH3 aggregation load merits further study, especially considering the fact that anti-NOTCH3 aggregation approaches for patients with CADASIL are in preclinical development. 49Such disease-modifying therapies will likely have the strongest beneficial effect in premanifest individuals, and as vascular NOTCH3 aggregation has been shown to precede clinical symptoms, 50 NOTCH3 aggregation load could be a promising target-engagement biomarker.Moreover, we show that young and premanifest individuals have a significant 2-year progression in MSMD, which could, therefore, potentially be used as a biomarker in early disease stage intervention clinical trials.
The strength of this study is its prospective, longitudinal nature, with all data uniformly gathered at 1 site, including state-of-theart neuroimaging acquisition and quantification.A limitation is that we had to make a selection of the most relevant clinical and neuroimaging outcome measures to decrease multiple testing effects.We selected neuroimaging markers which have been shown to be consistently present in patients with CADASIL or are strongly correlated with disease outcome measures. 26,27,42,45he effect of NOTCH3 risk category on progression of other neuroimaging markers, for example, CMB or PVS, merits further research.Likewise, to limit the number of independent variables in the prediction models, we did not study the effect of individual CVRF on disease progression, but instead used the total burden of CVFR as a predictor.Finally, given the limited number of patients with LR-NOTCH3 variants in our cohort, which is because low-risk variants are rare in CADASIL pedigrees, we could not determine disease progression in this subgroup of patients.
In conclusion, we show that patients with CADASIL show a significant progression of neuroimaging markers, executive function, and disability in a clinical trial-sensitive time frame of 2 years.We show the importance of taking NOTCH3 variant risk category into account because patients with HR-NOTCH3 variants have a significantly faster disease progression compared with patients with MR-NOTCH3 variants.This has important implications for disease prediction and monitoring in the clinic, biomarker development, and selection of patients and outcome measures in clinical trials.
Pamelen, and Ruth Berghuis for performing neuropsychological testing and assisting in MRI scanning during the research days and Marissa Wolswijk and Gijs Vermeij for technical support during MRI scanning.

Study Funding
Glossary BPF = brain parenchymal fraction; CADASIL = cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; CLMM = cumulative link mixed model; CVRF = cardiovascular risk factors; CVRFn = number of cardiovascular risk factors at baseline; DiViNAS = Disease Variability in NOTCH3-Associated Small Vessel Disease; DTI = diffusion tensor imaging; EGFr = epidermal growth factor like repeat; FLAIR = fluid-attenuated inversion recovery; HR/MR/ LR-NOTCH3 = high risk/moderate risk/low risk NOTCH3 variants; IQR = interquartile range; LMM = linear mixed model; mRS = modified Rankin Scale; MSMD = mean skeletonized mean diffusivity; OR = odds ratio; nLV = normalized lacune volume; nWMHv = normalized white matter hyperintensity volume; PSMD = peak width of the skeletonized mean diffusivity; SVD = small vessel disease; STROBE = Strengthening the Reporting of Observational studies in Epidemiology; SWI = susceptibility-weighted image; TMT = Trail Making Test; TMT B/A = Trail Making Test B given A t-scores; WMH = white matter hyperintensity.individualizeddisease prediction, to tailor disease monitoring and management, and to enable patient stratification in future clinical trials.We performed a single-center, prospective 2-year follow-up study to investigate whether NOTCH3 risk category is associated with clinical and neuroimaging disease progression in patients with CADASIL.

Figure 2
Figure 2 The HR-NOTCH3 Group Had a Higher 2-Year Cumulative Probability of Experiencing a Stroke Compared with the MR-NOTCH3 Group

Figure 3
Figure 3 The HR-NOTCH3 Group Had Higher Mean 2-Year Progression of MSMD and nLV Compared With the MR-NOTCH3 Group

Figure 4
Figure 4 MRI FLAIR Sequences at Baseline and Follow-Up: Examples of Variable Progression of Neuroimaging Markers

Table 1
Baseline Descriptives of DiViNAS 2-Year Follow-Up Study Participants

Table 2
Results at Baseline and 2-Year Follow-Up Abbreviations: BPF = brain parenchymal fraction; CVRFn = number of cardiovascular risk factors at baseline; IQR = interquartile range; mRS = modified Rankin Scale; MSMD = mean skeletonized mean diffusivity; nLV = normalized lacune volume; nWMHv = normalized white matter hyperintensity volume; TMT B/A = Trail Making Test B given A t-scores.a p-Value derived from multivariable Cox regression with NOTCH3 risk category, sex, baseline age, and baseline CVRFn as covariates.
MR-NOTCH3 groups, multivariable Cox regression was performed adjusted for sex, baseline age, and baseline CVRFn.Analysis of recurrence rate of lifetime stroke was performed using an Andersen-Gill Cox repeated-measures model, adjusted for the number of previous strokes, CVRF, and sex.
d p-Value derived from multivariable mixed-effect models after correction for sex, baseline age, and baseline CVRFn.e nWMHv, nLV, and BPF are normalized and are expressed as a percentage of the intracranial volume (in milliliters).The median (IQR) non-normalized WMH volume in the HR-NOTCH3 group was 55.1 mL (81.3) at baseline and 62.5 mL (79.4) at follow-up; for the MR-NOTCH3 group, this was 18.1 mL (35.5) at baseline and 21.8 mL (40.4) at follow-up.and