Cerebral small vessel disease (CSVD) is a common and complex syndrome with variable clinical presentations, imaging findings, and underlying pathology [1]. Despite its high prevalence and potential for treatment, effective therapies remain limited due to incomplete understanding of its pathophysiology. Studying monogenic forms of CSVD, such as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), offers a valuable approach to overcome the inherent heterogeneity of the broader CSVD population. CADASIL, the most frequent hereditary CSVD, arises from mutations in the highly conserved Notch3 gene [2,3]. Epidemiological data suggest that CADASIL may be underdiagnosed and contributes significantly to recurrent stroke and vascular cognitive impairment [4]. Diagnosis of CADASIL relies primarily on magnetic resonance imaging (MRI), which reveals characteristic features such as white matter hyperintensity, lacunar infarcts, and cerebral microbleeds, all indicative of small vessel damage [[5], [6], [7]]. However, recent research suggests that cerebrovascular dysfunction, including impaired cerebrovascular reactivity (CVR) and cerebral blood flow, precedes the development of parenchymal lesions and clinical symptoms in CSVD patients [8,9]. Therefore, investigating CVR alterations in CADASIL patients may provide crucial insights into the early pathophysiological mechanisms of CSVD.

CVR measures the changes in cerebral blood flow in response to vasoactive challenges. Techniques such as hypercapnic stimulation, acetazolamide injection, and even breath-holding can elevate CO2 levels in the bloodstream, thereby modulating vascular reactivity [10,11]. This makes it a vital tool for studying the pathogenesis of CSVD, neurodegenerative diseases, and traumatic brain injuries [1,[10], [11], [12]]. The Notch3 gene (chromosome 19) encodes the Notch3 receptor, predominantly found in vascular smooth muscle cells (VSMCs) and pericytes, with notable presence in leptomeninges and perforating arteries. In CADASIL, characteristic pathology, visualized using transmission electron microscopy, includes granular osmiophilic material (GOM) accumulation in VSMCs near the basement membrane, along with VSMCs degeneration and loss. These changes severely impair the contractile and relaxation functions of brain blood vessels, particularly affecting cerebral blood flow regulation in the microcirculation and thus neurovascular unit function [2,3,7]. Consequently, we presume that CVR abnormalities are anticipated in CADASIL patients.

Several techniques, including transcranial doppler ultrasound (TCD), arterial spin labeling (ASL)-MRI, and blood oxygen level-dependent (BOLD)-MRI, are used to measure CVR following hypercapnia. Previous studies on CADASIL patients have yielded heterogeneous CVR results. While one TCD study found no CVR impairment in non-demented CADASIL patients [13,14], the other reported reduced CO2 reactivity even in nondisabled CADASIL patients compared to controls [13,14]. An ASL-MRI study, lacking a control group, showed a trend towards impaired CVR in patients with depressive symptoms, disability, or delayed processing speed [15]. A BOLD-7.0T MRI study revealed preserved hypercapnia reactivity in the cortex, subcortical gray matter, and normal-appearing white matter compared to controls; however, CVR was decreased in white matter hyperintensities compared to normal-appearing white matter within the patient group [16]. This inconsistency highlights the need for further investigation into CVR alterations in CADASIL. This study will contribute to a better understanding of microvascular role in CADASIL and these studies will be beneficial for early diagnosis, disease subtyping and mechanistic understanding. However, research evaluating CVR changes in CADASIL patients using resting-state functional magnetic resonance imaging (fMRI) without gas challenges is lacking.

A novel method for mapping CVR utilizes resting-state BOLD fMRI data without gas inhalation. This approach leverages natural respiratory variations to identify a surrogate for arterial CO2 fluctuations within the global BOLD signal. This task-free and stimulus-free technique avoids the risks and complexities associated with traditional hypercapnic gas inhalation methods [11,17]. Validation studies by Liu et al. demonstrated consistent CVR mapping results between this resting-state BOLD method and conventional hypercapnia-based methods in both healthy volunteers and patients with Moyamoya disease [18,19]. Consequently, resting-state BOLD-derived CVR measurements offer a potentially valuable alternative for detecting cerebrovascular dysfunction in clinical settings.

Voxel-based morphometry (VBM) is a powerful technique for quantitatively assessing gray and white matter lesions in MRI scans. By analyzing volumetric, thickness, and density changes at the voxel level, VBM effectively reflects alterations in underlying anatomical structures [20]. Prior research has demonstrated reduced regional cortical volume and thickness in CADASIL patients compared to healthy controls, and has also identified correlations between cortical morphology and clinical characteristics, as well as white matter hyperintensity volume [21,22]. However, studies investigating gray matter density (GMD) in CADASIL patients are limited, and the diagnostic utility of GMD and CVR for CADASIL remains unclear.

This study investigated regional changes in CVR and GMD in CADASIL patients using resting-state fMRI and VBM, respectively. We also examined the association between these alterations and clinical scores, and assessed the diagnostic potential of CVR and GMD for CADASIL.