In a cross-sectional study performed on the ADNI cohort, we investigated predictive role of plasma levels of NFL for regional white matter changes in patients with MCI. Results of voxel-wise partial correlation analysis in our patients revealed a significant correlation between plasma NFL levels and a widespread pattern change in different white matter pathways (Tab2). First, we analyze the relation of plasma NFL and age to other patient's characteristics, including MMSE score, hippocampus, fusiform, ventricles, Entorhinal, mid temporal, and whole brain volume, average amyloid deposition in frontal, anterior cingulate, precuneus, and parietal cortex, average glucose uptake of angular, temporal, and posterior cingulate, CSF level of Aβ, total tau, p tau, and APOE ε4 genotype. In the next, we implemented a voxel-wise analysis to assess the correlation of plasma and CSF biomarkers with each DTI value in ROI regions by using partial correlation controlled for effect of age, sex, and APOE ε4 genotype.
Our findings revealed a significant correlation between plasma and CSF biomarkers with altered white matter microstructural changes in widespread brain regions. Plasma NFL has a negative correlation with FA and positive correlation with RD, AD, and MD values in bilateral Cingulum, bilateral hippocampal cingulum, bilateral Anterior corona radiate, Left Anterior limb of internal capsule, Right Posterior thalamic radiation, Right Retrolenticular part of internal capsule, bilateral Superior corona radiate, bilateral Superior fronto-occipital fasciculus, Right Inferior fronto-occipital fasciculus, Left Superior longitudinal fasciculus, Right Sagittal stratum, Bilateral full corpus callosum, and Bilateral fornix. Our findings are in line with Spotorno et al. indicates higher plasma NFL is associated with lower FA value in superior longitudinal fasciculus, the fronto-occipital fasciculus, the anterior thalamic radiation, and the dorsal cingulum bundle in frontotemporal dementia (12). White matter differences between patients with MCI and healthy controls are mostly observed in the corpus callosum, along with limbic pathways, including the fornix, cingulum, and uncinate fasciculus (13). Crucially, these signature regions are involved in the pathological course of Alzheimer's disease, and alterations in white matter integrity in these regions are known to be associated with disease progression from the early stages of AD (14). Also, serum level of NFL correlates with lower FA and Higher RD, AD, and MD in autosomal dominant Alzheimer's disease (15, 16). By investigating diffusion metrics, information about the different specifications of the White matter can be obtained from DTI values. On the other hand the exact relationship between this four values and the physiopathology mechanism is complex (17). Several researchers have previously been working to clarify the relationship between CSF NFL and white matter damage in AD and MCI (18-21). Also, our findings showed that plasma level of NFL has a strong correlation with patient's age, MMSE score, hippocampus, entorhinal, and ventricles volume, and these results are in line with recent research investigated the predictive role of plasma NFL in neurodegeneration and clinical progression of Alzheimer disease (22, 23). At baseline, high plasma NFL was associated with worse general cognition (.i.e. MMSE score) (24), which reflected damage to larger myelinated axonal processes of neurons (6). As well as we found that higher plasma NFL has a significant correlation with higher levels of tau and p tau in CSF, but there is no correlation with CSF Aβ, similarly with Mattsson et al., but also they found a relation with Aβ (6).
Moreover, based on our results, plasma NFL is elevated in patients with APOE ε4 gene who have a higher risk for developing dementia (25). These findings supported our initial hypothesis that higher plasma NfL levels associated with baseline AD specific changes and altered white matter connections. Coupled with recent evidence, we found a notable correlation between plasma NfL and hypometabolism in angular, temporal, and posterior cingulate (3, 26). With this in mind, plasma NFL level can be utilized as a novel biomarker for neurodegeneration and be considered as candidate instrument to screen neurodegeneration effects in patients with AD (3, 27, 28).
As our expect, and in line with previous studies, our findings revealed a significant correlation between CSF levels of tau, p tau, and Aβ with DTI values in wide regions affected in the earliest stages of AD (19, 29, 30). Meanwhile, our results showed CSF biomarkers correlated with a higher FA value in some regions. Arguably, the correlations between plasma NfL and CSF biomarkers are often not statistically significant within diagnostic groups and may present confounding errors in diagnosis (6). Another important finding in contrast to Buckley et al. is lack of correlation between CSF tau and p tau with CSF Aβ levels, whereas there is a strong correlation between average amyloid deposition in frontal, anterior cingulate, precuneus, and parietal cortex relative to the cerebellum with this two biomarkers (31). According to our results, Aβ is associated with white matter changes in fewer regions comparing to p tau or tau in CSF and plasma NFL. CSF p-tau/Aβ ratio and p-tau levels are predictor for MCI conversion to AD, but cortical amyloid or tau deposition by using PET is the gold standard to detect high risk MCI patients (32).
The apolipoprotein E (APOE) is a most important genetic risk factor for developing AD (33); likewise, recent studies revealed there is a strong correlation between CSF biomarkers and APOE genotype (34, 35); similarly, we found CSF p tau, tau, and amyloid burden are significantly associated with APOE ε4 genotype after implementing a partial correlation adjusted for age and sex.
Generally, plasma NFL track neurodegeneration both in the grey matter or the white matter (36). With these results in mind, our study provides a better understanding of the link between plasma NFL and connectometry changes in Alzheimer's progression. Our findings support the application of the NFL's blood-based measures as a noninvasive and costless biomarker for monitoring people with a high risk of AD in the early stages.