I. Introduction

VCID, or Vascular Contributions to Cognitive Impairment and Dementia, is a condition within the broader range of neurodegenerative diseases (NDs) and strongly correlates with circulatory health. Due to a reduction of blood flow––and therefore oxygen and nutrients––to various regions of the brain, VCID causes a decline in thinking skills, confusion, disorientation, and trouble with speaking and walking among other symptoms (Alzheimer’s Association, n.d.). Symptoms may present immediately following a stroke, which blocks major blood vessels in the brain, or by a gradual change seen as a result of multiple mini-strokes or hypoxic conditions (Alzheimer’s Association, n.d.). Although it often coexists with Alzheimer’s disease or Lewy body dementia (commonly known as mixed dementia), between 5–10% of dementia patients exclusively have VCID, and it is believed to be underdiagnosed (Alzheimer’s Association, n.d.).

The challenge presented is in effectively differentiating between these three forms of ND, as each has different biomarkers yet presents similar symptoms of cognitive decline. Alzheimer’s disease (AD), for example, is mainly characterized by plaques of beta-amyloid protein fragments that build up in between nerve cells, as well as tangles of twisted tau protein fibers (Alzheimer’s Association, n.d.). Lewy body dementia (DLB), on the other hand, often presents with mutations to alpha-synuclein, causing protein misfolding that is observable in cerebrospinal fluid (CSF), the assay mechanism known as RT-QuIC (Bousiges & Blanc, 2022). Finally, the indices for matrix metalloproteinase-2, placental growth factor (PIGF), and the albumin brain-plasma ratio are common characterizations for VCID (Alber et al., 2019). The main distinguishing factor for VCID, however, is the emphasis on vascular abnormalities and how they cause the impairment presented by patients. The American Heart Association, American Stroke Association, Alzheimer’s Association, and American Academy of Neurology published a joint statement in 2011 to outline the criteria for mild cognitive impairment (MCI) or dementia caused by vascular changes. The criteria suggest that neurocognitive testing in conjunction with brain imaging with magnetic resonance imaging (MRI) confirms either a recent stroke or other vascular brain changes (Alzheimer’s Association, n.d.). Finally, there must be no evidence that nonvascular factors could contribute to cognitive impairment (Alzheimer’s Association, n.d.).

There are various diagnostic techniques currently in use for neurodegenerative diseases, however, each has limitations of its own. Biomedical imaging, such as MRI, computerized tomography (CT), and positron emission tomography (PET) are known to provide accuracy but are slow, subjective, and cannot be used to predict the disease onset (Carota et al., 2022). Neuropsychological evaluations of symptoms are also commonly used with imaging tests, however, a post- mortem analysis of brain tissue is the only way to confirm the pathology in this case (Carota et al., 2022). Beyond the traditional methods above, researchers additionally look to biomarker detection to better address the long-term proliferation of ND. Biomarkers allow for earlier diagnosis, monitoring of disease progression, differentiation between ND types, and measuring therapeutic responses (Carota et al., 2022). The two main diagnostic techniques currently used for biomarker identification are mass spectrometry and enzyme-linked immunosorbent assay (ELISA), however, both are expensive and slow (Carota et al., 2022).

When referring to VCID specifically, white matter hyperintensities (WMHs) of vascular origin and small vessel disease (SVD) are most prominently expressed on MRI scans (Alber et al., 2019). WMHs show an altered water content in white matter fibers and tracts, and a greater volume of WMHs correlates to a fall in cognitive performance and an increase in strokes (Alber et al., 2019). A study by de Leeuw et al. indicates the role of age as it relates to symptom progression. In a group of participants between 60–70 years old, 87% had subcortical WMHs and 68% had periventricular WMHs (de Leeuw et al., 2002). Additionally, for patients aged 80 to 90 years, 100% had subcortical and 95% had periventricular WMHs (de Leeuw et al., 2002). These results indicate the need for MRI in VCID diagnosis, however, the criteria used to make radiological assessments show varying agreement between interobservers (van Staaten et al., 2003). For VCID, the NINDS-AIREN criteria, a joint effort by the National Institute of Neurological Disorders and Stroke and the Association Internationale pour la Recherche et l’Enseignement en Neurosciences, are most commonly used, and a work by van Staaten et al. (2003) evaluated how inexperienced and experienced interobservers assessed the criterion differently using Cohen’s kappa coefficient. Observers rated 40 MRI studies of patients with clinically-suspected VCID twice, and after the first rating, operational definitions were made and used for the second rating. The results show a kappa of 0.29 after the first rating, indicating a poor level of agreement among observers, and that it only improved to a kappa of 0.38 after the definitions were used in the second rating. Because of these results, it is clear that radiological assessments of MRIs for VCID diagnoses are highly variable, and a more consistent technique is needed.

In the realm of medical diagnostic technology, Raman spectroscopy (RSS) holds a role in biomarker identification and is known for efficiency, cost sensitivity, and effectiveness. By utilizing the principles of the photoelectric effect and quantized energy levels, RSS technology currently aids researchers in diagnosing certain cancers such as gliomas, and has recently been adapted for AD. This research paper will review the effectiveness of RSS as a diagnostic technique––especially for neurodegenerative disorders such as AD––and determine how RSS could be used for the separate but closely related disorder, VCID (“Vascular Dementia”), to improve patient outcomes by monitoring disease progression and implementing early identification.

II. Raman Spectroscopy

Discovered in 1928 by C. V. Raman & K.S. Krishnan, Raman spectroscopy measures the outcome of light interacting with chemical bonds in a sample and provides information regarding the crystallinity and molecular interactions, chemical structure, and phase and polymorphy (HORIBA, n.d.). As light interacts with the molecule, the photon’s electromagnetic field polarizes the molecule’s electron cloud, transferring energy to the molecule in the process. This forms a short-lived complex of photon/molecule called the “virtual state” of a molecule, which is very unstable and almost immediately reemits the photon as scattered light. The scattered light is the result of two processes occurring: Rayleigh scattering and Raman scattering (Edinburgh Instruments Ltd., n.d.). The dominant of the two––Rayleigh scattering––expresses the elastic process of the incident photon’s energy and wavelength from the laser beam, meaning that the energy of the molecule is unchanged, or rather conserved, after the interaction (Edinburgh Instruments Ltd., n.d.). The inelastic process used for observational study is called Raman scattering and occurs once every ten million photons (Edinburgh Instruments Ltd., n.d.). When the molecule gains energy, as it is excited to a higher vibrational level, while the photon loses energy and the wavelength increases, the process is known as Stokes Raman scattering (Edinburgh Instruments Ltd., n.d.). Comparatively, Anti-Stokes Raman scattering is the loss of energy in the molecule by relaxing to a lower vibrational level, as the photon gains energy and its wavelength decreases (Edinburgh Instruments Ltd., n.d.). Due to the Boltzmann distribution in which the majority of molecules will be at a ground energy level, Stokes Raman scattering is statistically more probable, provides a greater intensity of a scatter, and is, therefore, more commonly measured in RSS (Edinburgh Instruments Ltd., n.d.). Analysis of the absolute and relative Raman scattering peaks (HORIBA, n.d.) requires accounting for the different incident lasers used––as each produces different wavelengths––and as such, the scattering must be mathematically converted to a Raman shift away from the incident wavelength (Edinburgh Instruments Ltd., n.d.).

As a diagnostic technique, Raman microscopy couples a Raman spectrometer with an optical microscope to provide highly magnified visual representations of a sample, as well as the Raman analysis via a microscopic laser spot (HORIBA, n.d.). The Raman microscope was developed in the 1960s by Professor Michel Delhaye and Edouard DaSilva in Lille, France (HORIBA, n.d.). Motorized mapping stages, which use a spatial resolution of 0.5–1μm, produce images including thousands of Raman spectra coming from various sample positions (HORIBA, n.d.). The images shown can express false colors that differentiate the distribution of chemical components within the sample (HORIBA, n.d.). Sample types may include solids, powders, liquids, gases, inorganic/organic/biological materials, and more, however, metals (and their alloys) cannot be RSS samples (HORIBA, n.d.). More recently, various developments in Raman microscopy have specialized the technique: SRS (stimulated Raman scattering), SERS (surface-enhanced Raman scattering), TERS (tip-enhanced Raman scattering), Transmission Raman scattering, as well as others (HORIBA, n.d.). Each has specific functionalities: Transmission Raman scattering is utilized for bulk material analysis, SRS and SERS enhance Raman signals (SERS through the binding of nanoparticle metals to molecules), and TERS uses a metal tip to focus light rather than a metal surface (HORIBA, n.d.).

III. Current Diagnostic Applications of RSS Technology

There are many current diagnostic uses of RSS technology, especially in the realm of cancer research. Although Raman spectroscopy has been used for the past few decades to detect breast, prostate, colorectal, skin, gastric, and nasopharyngeal tumors, it is increasingly becoming an intraoperative auxiliary method to remove said tumors (Zhang et al., 2023). Neurosurgeons, for example, utilize RSS technology for glioma removal procedures, as they can examine the tissue properties mid-surgery to optimize the margins of removal and therefore improve patient outcomes (Zhang et al., 2023). The rapidity, high accuracy, and non-invasive nature of RSS technology allow for these advancements and could even change how surgical-oncology procedures are done (Zhang et al., 2023). Currently, however, the only method of Raman spectroscopy approved by the FDA is stimulated Raman histology (SRH) which uses the SRS technique (Zhang et al., 2023).

Outside of the operating room, RSS holds an important role in analyzing molecular components to enhance cancer research. Gliomas in particular, which affect astrocytes, oligodendrocytes, and ependymal cells [14], can be characterized by the effects seen on the target p53 protein [18]. p53 is an intrinsically disordered protein that contains structured & disordered domains, depending on the conditions and coexisting conformations (Signorelli et al., 2019). The protein contains a C and an N terminus which bind to the central DNA-binding domain (DBD) (Signorelli et al., 2019). A study by Signorelli et al. (2019) utilized Raman spectroscopy to examine the effects on the secondary structures of p53 and its DBD when bound to Azurin (AZ). Azurin is a copper-containing, electron transport anticancer protein secreted by Pseudomonas aeruginosa. In the study, the Fermi doublet ratios relative to the Raman peaks of Tyrosine and Tryptophan were analyzed given that the amino acids accurately express the protein side chain environments. Additionally, the Fermi ratios for Amide I were analyzed, as they show hydrophobic and hydrophilic changes to the α-helices, β-sheets, and random coil motifs resulting from the p53-AZ interaction (Figure 1). The results indicate that the binding of AZ to p53 stabilizes the protein and causes an in vivo and in vitro intracellular increase, thereby assisting its tumor suppressor function. The use of RSS technology specifically in this study allowed researchers to go beyond the previously discovered possible binding sites of AZ and determine the specific secondary structure changes from the interaction.

In addition to its role in cancer, Alzheimer’s disease diagnoses recently began utilizing Raman spectroscopy as a preliminary method. A study by Carota et al. (2022) investigated this potential use of RSS by analyzing the spectra of blood serum samples and their multiple vibrational modes. From hemoglobin to carotenoids to phenylalanine, the software assigned various molecules from the samples based on the modes measured (Figure 1). The results most notably found a correlation between the level of carotenoids present in the blood serum and the severity and progression of AD. NDs are often characterized by the presence of neuroinflammation and autophagy failures, which is a cellular process that, when functioning, removes misfolded toxic proteins including twisted tau fibers (Manochkumar et al., 2021). Carotenoids are known to be anti-neuroinflammatory compounds as well as autophagy modulators (Manochkumar et al., 2021). Therefore, the experimental changes in carotenoid levels measured by RSS and its association with the NINDS criteria and Clinical Dementia Rating (CDR) of patients confirm the beneficial use of RSS for NDs (Carota et al., 2022). Additionally, this study offers a window into RSS’s non-invasive potential, given that AD biomarkers for tau proteins are obtained via cerebrospinal fluid, which requires more invasiveness than blood serum draws (Carota et al., 2022). Given the relationship between carotenoids and NDs overall, it is clear that the same methodology and RSS techniques could be translated to diagnosing VCID.

IV. Diagnostic Application of RSS Technology for VCID

Biomarkers associated with SVD and WMHs may be analyzed using RSS technology to better quantify the factors distinguishing VCID from other types of ND. For patients with clinically diagnosed SVD––a prominent characteristic of VCID––there is a reported increase in the metalloproteinase-2 index, albumin brain-plasma ratio, low molecular weight neurofilament marker (NF-L), extracellular metalloproteinase matrix metalloproteinase-9, and the tissue inhibitor of metalloproteinase-1 (Alber et al., 2019). Although there are many biomarkers for VCID that RSS could be used to detect, there is already experimental evidence for the effective use of surface-enhanced Raman scattering (SERS) for matrix metalloproteinase-2 (Gong et al., 2015). Matrix metalloproteinase (MMP) is a family of endopeptidases that hold a role in the proteolytic degradation of extracellular matrix (ECM) components (Gong et al., 2015). When MMP activity is enhanced, it cleaves calcitonin gene-related peptide (CGRP), causing vasoconstriction (Fernandez-Patron et al., 2000). Vasoconstriction results in hypertension, a major cause of stroke, and it is therefore a significant risk factor for VCID (Johansson, 1999).

In a study by Gong et al. (2015), SERS evaluated levels of matrix metalloproteinase-2 and 7 to reveal their function as cancer biomarkers. To detect MMP, the study performed reactions where MMP peptides were cleaved by their corresponding enzyme. Gold nanoparticles (AuNPs) were bound to a specialized SERS-based bimetallic-film-over-nanosphere (BMFON) substrate, and MMP peptide chains inhibited this binding. Once the peptides were cleaved, AuNP binding sites were open; by tagging the AuNPs with either 4-aminothiophenol (4-ATP) or 2-naphthalenethiol (NT) and measuring the intensities of their SERS peaks, researchers then identified the concentrations of the MMP enzymes. Figure 2 below presents the data of vibrational modes measured in the study. SERS mapping generates chemical images based on the Raman peaks to determine if the peptide cleavage was successful. The results of the study indicate that SERS is a highly sensitive technology that when coupled with the cleave-and-bind mechanism, can detect MMP by measuring the concentrations of its enzymes and producing signals. Given the success of this study with regard to MMP and cancer diagnoses, such technology could be implemented into the list of diagnostic techniques used for VCID to efficiently improve diagnostic objectivity across the board.

In the United States, 74.5% of individuals aged 60+ present with hypertension, which increases their risk of stroke and therefore VCID (Ostchega et al., 2020). Moreso, VCID accounts for an annual global social cost of $604 billion, or 1% of the global gross domestic product (GDP) (Zhang et al., 2020). Because of its prevalence and the resulting monetary challenges presented to patients, a more efficient, cost-effective diagnostic technique such as RSS technology could improve patients’ access to proper care. Given that hypertension rates are also higher in historically marginalized populations such as the Black community (Ostchega et al., 2020), restructuring accessibility is a necessity for today’s healthcare system. If used in tandem with the current NINDS-AIREN criteria for MRI evaluations, Raman spectroscopy could more accurately track the severity and progression of VCID while helping researchers differentiate between VCID and other forms of ND.

V. Conclusion

About a century after its development, Raman spectroscopy (RSS technology) has been successfully applied to medical diagnoses for various types of cancer by coupling it with an optical microscope. Recent studies show its effectiveness for diagnosing neurodegenerative diseases such as Alzheimer’s disease, as the intensities of vibrational peaks represent biomarker concentrations. Due to the subjective nature of the current criteria used to evaluate white matter hyperintensities on magnetic resonance imaging (MRI), translating RSS technology to “Vascular Dementia” (VCID) would provide quantified evidence to either support or disprove previous claims of disease progression and severity. There is already evidence for the ability of surface-enhanced Raman scattering (SERS)––a type of RSS––to identify matrix metalloproteinase-2, a biomarker for VCID. Given the likelihood of individuals to present with hypertension at older ages, an efficient, cost-effective technique such as RSS could aid in the diagnosis of VCID, differentiate between types of neurodegenerative diseases, and ultimately support the development of future treatment options.