Annual Meeting Research: Dementia Biomarkers Draw Attention

Neurodegeneration as Marker for Dementia

Markers of neurodegeneration are the strongest positive and negative predictors of short-term incident dementia in MCI, irrespective of amyloidosis status, researchers have shown. FDGPET is the strongest individual positive predictive biomarker, they report.

Findings come from an analysis of 73 MCI patients who were clinically followed for at least one year (mean of 28±17 months) to ascertain progression to AD. Patients were assessed for markers of amyloidosis (abnormal CSF Abeta42) and neurodegeneration (hippocampal atrophy, TP hypometabolism on FDG-PET, and abnormal CSF tau).

Of the 73 subjects enrolled, 29 MCI patients progressed to AD (pMCI) and 44 remained stable (sMCI). Among IWG criteria, positivity to any biomarker had lowest negative likelihood ratios (0.00), while positivity to FDG-PET had highest positive likelihood ratios (5.82). Among NIA-AA criteria, positivity to neurodegeneration (FDG-PET, MRI or CSF tau markers, irrespective of amyloidosis status) had lowest negative likelihood ratios (0.06), while positivity to AB42 and FDG-PET or AB42 and hippocampal volume atrophy had highest positive likelihood ratios (6.45 and 5.56).
— IN3-2.005, Frisoni, et al.

Predictive Value of CSF Biomarkers for AD

Assessment of CSF Aβ-42, T-tau, and P-tau evaluation is useful in patients with MCI, in order to assess the probability of progression to Alzheimer’s disease, according to a new study. These CSF biomarkers taken together provide a sensitivity of 89 percent, a specificity of 77 percent, a PPV of 73 percent, and a VPN of 91 percent in predicting AD.

Eighty-eight patients (49 females and 39 males; mean age 68.3 ± 8.2 yrs) diagnosed with MCI were recruited for the study. Of these, 42 percent had amnestic MCI and 58 percent had non-amnestic MCI. All patients underwent lumbar puncture, and CSF Aβ-42, T-tau and P-tau concentrations were measured. Patients were re-evaluated at three years follow up. Aβ-42 had an AUC of 0.88, a sensitivity of 86%, and a specificity of 79% to predict AD (cut-off: 500 mg /mL), while AD profile had an AUC of 0.97, a sensitivity of 97%, and a specificity of 96% (cut-off: 1).
—IN3-2.003, Rubino, et al.

Amyloid Deposition and Hippocampal Volume Predict Incident Dementia in Non- Demented Elderly

Amyloid load does not appear to accurately predict dementia, new MRI findings suggest. Specifically, while the amount of amyloid deposited in specific brain regions was a strong predictor of dementia, a binary measure of PiB “positivity” was not. Findings suggest that atrophy and amyloid deposition must be present before some individuals develop symptoms of dementia (MCI to dementia). Alternatively, atrophy may precede the detection of amyloid deposition, requiring less amyloid to express clinical symptoms.

A total of 191 non-demented elderly subjects had a brain MRI and positron emission tomography scan with Pittsburgh compound B (PiB) in 2009 and were re-examined clinically in 2011. At the initial scan, 151 of the subjects were cognitively normal and 40 were classified as mild cognitive impairment (MCI); 105 were PiB+ (55 percent). In 2011, 23 (12 percent) of subjects progressed to dementia; of these, 17 (74 percent) of these were PiB+ in 2009. Incident dementia was associated with average cortical PiB retention in five brain regions, and with hippocampal volumes. The full assessment of the initial cohort is: 114 remained normal in 2011, 24 progressed to MCI, 10 progressed from normal to dementia, 27 remained as MCI, 13 progressed from MCI to dementia, and three MCI subjects converted to normal.

Higher cortical PiB retention was observed in subjects who progressed from MCI to dementia, but not among those who progressed from normal to dementia. Hippocampal volumes were smaller both in subjects who progressed from normal to dementia and from MCI to dementia compared to the other groups.
—IN3-2.004, Lopez, et al.