P2-207: CSF beta-amyloid 1-42, but not CSF tau, phosphorylated tau or hippocampal atrophy on MRI relate to Alzheimer's disease pathology load in the brain

For more than a decade, CSF biomarker levels of amyloid beta 1–42 (Abeta42), total tau (tau) and phosphorylated tau at threonine 181 (ptau) have been investigated for their use in diagnosing Alzheimer's disease (AD) in addition to hippocampal atrophy on MRI. Little is known however, about the relation between the amount of neuropathological changes and CSF biomarker levels or hippocampal atrophy. Objective: To relate in vivo measured CSF abeta, tau and ptau as well as medial temporal lobe atrophy (MTA-score) on MRI to the amount of amyloid plaques (AP) and neurofibrillary tangles (NFT) at autopsy. 30 autopsied patients were included: 8 AD 3 female (F), 11 FTD (6F), 2 DLB (0F), 4 CJD (4F) and 5 (2F) other neurodegenerative diseases (OD). Average AP load and NFT load were determined semi-quantitatively with a score from none to very many (0–4) in 6 different brain regions (frontal, temporal, parietal, occipital, hippocampal, amygdala) added up and divided by 6. We used linear regression analysis to investigate relationships between CSF biomarker levels and average MTA-score and AP and NFT load. The descriptives are presented in table 1. Across patients, only CSF Abeta42 level was related to both AP load (beta -0,70 p<0,001) and NFT load (beta -0,64 p<0,001). Within the non-AD patients this relation remained significant (beta -0,53 p<0,05), but within AD patients we found no significant relation. In this correlative autopsy study CSF Abeta42 was more specific to AD pathology at autopsy than CSF tau or ptau level or hippocampal atrophy on MRI across patients with varying neurodegenerative diseases. This may support the assumption that Abeta42 is a marker for presence of the disease in any stage, while tau, ptau and MTA, reflect the intensity or progress of the disease process during life. Within AD patients we did not find a correlation between biomarkers and AP or NFT load which may be explained by a ceiling effect of AP and NFT load at autopsy.