Résumé
Background
Plasma markers are now available for all categories in the AT(N) classification. However, their agreement with CSF markers is not fully established. A blood signature to generate the AT(N) classification would facilitate early diagnosis and follow‐up of patients with Alzheimer’s disease (AD) through an easy and minimally invasive approach. We aimed to assess the agreement between plasma and CSF markers for the AT(N) classification.
Method
We measured Aβ, pTau181 and neurofilament light (NfL) in plasma samples of the SPIN cohort (Barcelona, Spain). Plasma Aβ‐peptides were quantified by Aβ immuno‐purification followed by mass spectrometry analysis. Plasma levels of pTau181 and NfL were measured using the ultrasensitive Single Molecule Array (Simoa, Quanterix). We assessed the correlation between plasma and CSF markers and calculated areas under the ROC curve (AUC) for individual plasma biomarkers to detect positivity/negativity in their respective category. We also calculated the accuracy of plasma biomarkers for the detection of the A+T+ profile.
Result
We included a total of 150 participants from the SPIN cohort, comprising patients with AD (n=27), Lewy body dementia (n=52), frontotemporal lobar degeneration‐related syndromes (n=25) and 46 cognitively normal participants.
As shown in Figure 1, we found moderate correlation between plasma Aβ composite and CSF Aβ1‐42/Aβ1‐40 (Rho=‐0.5, p<0.001) and between plasma pTau181 and CSF pTau181 in the entire cohort (Rho=0.51, p<0.001). NfL levels in plasma showed high correlation with those in CSF (Rho=0.78, p<0.001). The plasma Aβ composite, pTau181 and NfL yielded AUC of 0.75, 0.78 and 0.88 to discriminate positive and negative participants in their respective categories. These accuracies were similar when other CSF cutoffs were applied (Figure 2). The combination of all three markers did not outperform pTau181 alone (AUC=0.81) to discriminate A+T+ from A‐T‐ participants (Figure 3).
Conclusion
Plasma biomarkers are useful to detect the AT(N) categories, and their use can differentiate patients with pathophysiological evidence of AD. This signature could be applied in clinical routine as an early step for the diagnosis of AD or to detect AD copathology, optimizing the selection of potential candidates for disease modifying therapies.