MISMATCH BETWEEN CLINICAL AND NEUROPATHOLOGICAL DIAGNOSIS OF DEMENTIA: HOW CAN BIOMARKERS HELP?

Abstract

With upcoming clinical trials for dementia, an accurate clinical diagnosis reflecting the underlying neuropathological cause, becomes ever more important. Literature reports a mismatch between clinical and neuropathological diagnosis (7%- 32%). Biomarkers reflecting neuropathology may reduce this mismatch. This study focuses on the influence of CSF, amyloid-PET, and FDG-PET biomarkers in the (dis)concordance between clinical and neuropathological diagnosis. We included 87 patients from the Amsterdam Dementia Cohort of whom autopsy was performed between 2000–2014. Initial and final clinical diagnoses were compared to neuropathological diagnoses. Initial diagnosis was based on structured history, neurological examination, neuropsychological assessment, and MRI. Final diagnosis was defined at clinical follow-up (mean, range: 3, 0–46 months) with use of biomarkers. Biomarker tests were performed in 56 out of 87 patients; 45 CSF; 2 FDG-PET; 5 CSF+FDG-PET; 2 CSF+amyloid-PET; 2 CSF+FDG-PET+amyloid-PET. Patient characteristics per pathological diagnosis did not differ between groups and are summarized in table 1. Table 2 shows the clinical-pathological mismatch for initial and final diagnosis per diagnostic group. Based on initial diagnosis, 24 patients (28%) had a clinical-pathological mismatch, of which 5 in the AD group, 8 in the group with other dementias, and 4 in the frontotemporal dementia (FTD) group. Based on final diagnosis, 15 patients (17%) had a clinical pathological mismatch, of which 2 in the AD group, 6 in the group with other dementias, and 4 in the FTD group. The decrease in clinical-pathological mismatch from initial to final diagnosis was highest in the AD group (8%), followed by the group with other forms of dementia (4%) and the FTD group (2%). Six of 15 final mismatches had biomarkers (5 patients CSF, 1 CSF+ FDG-PET). Two patients had an AD biomarker profile in CSF, reflecting the pathological diagnosis, however, this had not changed the clinical diagnosis in AD. One patient had an FTD distribution on FDG-PET, not congruent with the pathological diagnosis. In the other 4 patients, CSF was helpful by excluding AD. The use of CSF and PET biomarkers may decrease the clinical-pathological diagnostic mismatch for AD. The higher mismatch in other forms indicates the need for more neuropathologically based biomarkers.