Could analysis of a patient’s blood plasma via vibrational spectroscopy facilitate early differential diagnosis and treatment of Alzheimer disease (AD) and other dementias? A team of British researchers says yes.1 The team conducted a multicenter study in which attenuated total reflection chemometric analysis of blood plasma via Fourier transform infrared (ATR-FTIR) spectroscopy was used to identify biomarkers in 347 patients with various neurodegenerative diseases and 202 healthy, age-matched controls.
The investigators aimed to identify patients with AD and also to detect biomarkers that could shed further light on a patient’s diagnosis. Thus, in applying and comparing 6 chemometric techniques via ATR-FTIR, they identified variables that differentiate subjects with AD from healthy controls. Because the apolipoprotein E (APOE) e4 allele is highly associated with AD risk, the investigators included APOE genotype and participant age as confounding factors. Analysis was conducted with and without APOE genotype and age information and compared.
The team also looked into the identification of early AD; AD progression; lipid-to-protein, phosphate-to-carbohydrate, and RNA-to-DNA ratios related to age; and how spectral wave numbers could allow for differentiation of AD from other neurodegenerative diseases, such as dementia with Lewy bodies (DLB), frontotemporal dementia (FTD), progressive supranuclear palsy (PSP), and Parkinson disease. Data were primarily reported on DLB because the sample size (n = 34), compared with those of FTD, PSP, and Parkinson disease, was considered sufficient to provide information of a “robust” quality.
The diagnostic accuracy of blood-based chemometric analysis using ATR-FTIR spectroscopy was in the range of that of existing biomarker-based technologies, reported to vary from approximately 70% to about 95%. Use of ATR-FTIR spectroscopy excluding APOE e4 genotype information identified AD, which accounted for 164 cases in the cohort, with 70% sensitivity and specificity. When APOE e4 information was included, the degree of sensitivity and specificity increased to 86% among subjects with 1 to 2 e4 alleles, and sensitivity increased to 72% and specificity to 77% in subjects who did not carry e4 alleles. Early AD (n = 14) was identified with 80% sensitivity and 74% specificity, and differentiation of AD from DLB was achieved with 90% sensitivity and specificity. Significant differentiation also was achieved for AD versus FTD.
The investigators pointed out that there is no single definitive medical test for the diagnosis of AD and that most available biomarker testing procedures has been impractical for general clinical use because of high cost and labor-intensive sample processing. They noted that blood-based vibrational spectroscopy techniques, such as FTIR and also Raman, are increasingly being used for diagnosing, screening, and monitoring diseases. In addition, these techniques have an advantage over conventional molecular testing because they are more cost-effective than enzyme-linked immunosorbent assay and can simultaneously analyze a range of different molecules at once. Therefore, vibrational spectroscopy may be ideal for the clinical diagnosis of complex, multifactorial diseases, the study authors stated.
The research team concluded that blood-based vibrational spectroscopy could provide an alternative to current blood, cerebrospinal fluid, or magnetic resonance testing in the diagnosis and monitoring of AD and other neurodegenerative diseases. The team acknowledged that a larger, prospective study of healthy subjects or subjects with mild cognitive impairment would be necessary to confirm whether vibrational spectroscopy could be a screening tool for AD. The ultimate aim, according to the investigators, would be to identify those persons in whom AD was destined to develop long before symptoms emerge so that the disease could be slowed down or prevented.
1. Paraskevaidi M, Morais CLM, Lima KMG, et al. Differential diagnosis of Alzheimer’s disease using spectrochemical analysis of blood. Proc Natl Acad Sci U S A. 2017 Sep 5. [Epub ahead of print]