In this study, we compared CSF, plasma, and saliva levels of AD core biomarkers, i.e. Aβ42, Aβ42/40 ratio, and pTau, measured by Lumipulse G platform, in patients with cognitive decline. The levels of all biomarkers significantly differed in the three biofluids, with saliva showing the lowest and CSF the highest levels of Aβ42, Aβ40, and pTau. In CSF, the median concentration of Aβ40 was 8620 pg/mL, which is approximately 30 times greater than its levels in plasma (270.9 pg/mL). When compared to saliva, where Aβ40 was only 0.70 pg/mL, the concentration in CSF was about 12,000 times higher. The median concentration of Aβ42 in CSF was 444.0 pg/mL, much higher than in plasma (26.05 pg/mL) and saliva (0.61 pg/mL). The CSF level was roughly 20 times that in plasma and about 700 times greater than in saliva. The Aβ42/40 ratio showed significant variation across the biofluids. In CSF, it was relatively low (0.049), indicating a much higher concentration of Aβ40 compared to Aβ42. In plasma, the ratio doubled (0.098), suggesting a slightly different balance between Aβ42 and Aβ40. Saliva, however, showed a drastically higher ratio (0.775), reflecting a much different balance between the two biomarkers in this fluid, albeit at much lower concentrations overall. The median concentration of pTau in CSF was 76.00 pg/mL, significantly higher than in plasma (1.79 pg/mL) and saliva (5.58 pg/mL). The concentration in CSF was about 40 times that in plasma and roughly 15 times that in saliva. Differently from Aβ40 and Aβ42, the pTau appeared to be more concentrated in saliva than in plasma. Interestingly, a positive correlation of pTau and Aβ42/40 ratio levels in CSF and plasma was detected, while no correlation between any biomarker in CSF and saliva was found. Thus, our findings suggest that plasma but not saliva could represent a surrogate biofluid for measuring AD core biomarkers. Specifically, plasma Aβ42/40 ratio and pTau could serve as surrogates of the corresponding CSF biomarkers. This is in accordance with Arranz et al. and Martinez-Dubarbie et al., who also explored the correlation between plasma and CSF Aβ42, Aβ42/40 ratio, and pTau levels measured by Lumipulse G platform, finding a moderate correlation for pTau and Aβ42/40 ratio [19, 20].
To date, only Marksteiner et al. measured AD core biomarkers in saliva using Lumipulse [21]. Noteworthy, the Authors collected saliva by salivettes and they did not detect Aβ42 and Aβ40 due to their binding to cotton. Additionally, they do not state the kit used, i.e. Lumipulse G β-Amyloid 1–40 Plasma or CSF, which is important information to understand the sensitivity of the method. Since Fujirebio did not develop a kit for measuring tTau in alternative biological matrices, it is plausible that Authors used kits for CSF. Thus, we cannot compare our findings with those of Marksteiner et al..
Despite the initial enthusiasm for AD biomarkers in saliva, to date evidence in literature are inconsistent to support the use of saliva as a reliable alternative biological matrix to measure AD core biomarkers. First, several technical issues related to sample processing limit saliva for diagnostic purposes. Additionally, the secretion and composition of saliva may be affected by several factors, including medications, lifestyle, age, sex, and diseases [22, 23]. Noteworthy, a bidirectional oral-brain axis connected through almost six routes has been documented [22]. Finally, salivary Aβ42 levels may originate from different sources, including cranial nerves innervating salivary glands, acinar epithelial cells of salivary glands, the transportation from blood to saliva, and the presence of Aβ1–42 protein deposits in peripheral regions, such as the nasal mucosa, lacrimal and lingual glands, which could be released directly or indirectly in saliva [24, 25]. Two recent meta-analyses showed that the salivary pattern of AD is characterized by elevated Aβ42 levels and unchanged or decreased pTau and tTau levels as compared to controls [25, 26].
Overall, our findings suggest that plasma but not saliva could represent a surrogate biofluid for measuring AD core biomarkers. Blood-based AD diagnosis offers several advantages in terms of accessibility and repeatability and the possibility to measure biomarkers by a fully automated platform promotes their widespread diffusion in clinical practice, paving the way to a new revolution in the field of neurodegenerative diseases. Indeed, blood-based biomarkers could aid in supporting Clinicians across the whole path of care of AD patients, from screening, early diagnosis, and monitoring of both disease and therapy. In order to introduce and appropriately use blood-based AD biomarkers in clinical practice, further studies are mandatory to establish reference intervals and decisional cut-offs as well as to evaluate the biological determinants, such as age or sex, and the potential influence of comorbidities. So far, preanalytical variables and their effects on plasma AD core biomarkers measured by the Lumipulse platform have been investigated. Musso et al. reported that hemolysis may alter biomarkers levels [27]. They also described the effect of temperature storage (4°C, -20°C, and − 80°C) on biomarkers concentrations, suggesting that different cut-offs should be used for fresh and thawed samples [27]. However, such an effect was not reported by Mansilla et al. [28]. Thus, further studies are mandatory to clarify the possible effect of temperature storage on plasma biomarkers stability. Two independent studies explored the influence of blood-brain barrier (BBB) integrity, assessed by the CSF/serum albumin quotient, and kidney function on blood-based biomarkers measured by the Lumipulse platform [29,30]. Both found that plasma biomarkers levels are influenced by kidney function, with individuals suffering from renal dysfunction having increased levels. However, the Aβ42/40 ratio spared this effect. This finding is in accordance with Martinez-Dubarbie et al. [20]. On the other hand, they achieved opposite conclusions on the influence of BBB permeability on the biomarkers’ levels [29,30].
In conclusion, the measurement of AD core biomarkers in plasma by fully automated platform hold great promise for routine clinical use [31]. However, some issues must be resolved before their introduction in clinical laboratories.