CSF proteomic profiles related to cognitive decline in MCI A+ depend on tau levels

Individuals with mild cognitive impairment (MCI) and an abnormal amyloid biomarker (A+) are at considerable increased risk of developing dementia. Still, these individuals vary greatly in rates of cognitive decline, and the mechanisms underlying this heterogeneity remain largely unclear. One factor related to an increased risk of progression to dementia is having an abnormal tau status (T+), but this still explains only part of the variance. Furthermore, previous work has indicated that MCI A+ individuals with T- or T+ are characterized by distinct molecular processes as reflected by distinct CSF proteomic profiles. As such, it could be hypothesized that differences in rates of cognitive decline in A+ MCI with abnormal or normal tau status may be explained by distinct underlying mechanisms. We studied this question using an untargeted CSF proteomic approach in individuals with MCI and abnormal amyloid. We measured untargeted Tandem Mass Tag (TMT) mass spectrometry proteomics in CSF of 80 A+ MCI individuals from the Amsterdam Dementia Cohort [age 66 ± 7.9 years, 52 (65%) T+]. For each protein, we tested if CSF levels were related to time to progression to dementia using Cox survival models; and with decline on the Mini-Mental State Examination (MMSE) with linear mixed models, correcting for age, sex and education. We validated our results in the independent Alzheimer's Disease Neuroimaging Initiative (ADNI) that employed the orthogonal CSF Soma logic protein measures in 245 CSF A+ MCI individuals [age 73 ± 7.2 years, 135 (55%) T+]. In total, we found 664 (29%) proteins to be related to cognitive decline in A+T+ and 718 (31%) proteins in A+T-. In A+T+, higher levels of 393 proteins that were associated with synaptic plasticity processes, and lower levels of 271 proteins associated with the immune function processes predicted a steeper decline on the MMSE and faster progression to dementia. In A+T-, higher levels of 306 proteins that were related to blood-brain barrier impairment and lower levels of 412 proteins associated with synaptic plasticity processes predicted a steeper decline; 67% of pathways associated with a decline in A+T+ and 58% in A+T- were replicated in ADNI. In conclusion, cognitive decline in A+ MCI individuals with and without tau may involve distinct underlying pathophysiology. These findings suggest that treatments aiming to delay cognitive decline may need tailoring according to the underlying mechanism of these patient groups, and that amyloid and tau levels could aid in stratification of selecting patients.