Basic Science and Pathogenesis

Evaluation of post-mortem brain tissue provides insight in biological heterogeneity in dementia, often showing multiple coexisting pathologies across diagnoses. Individual or combined contributions of specific (co-)pathologies to ante-mortem diagnostic accuracy remains largely unknown. We hypothesized that discordance between the primary pathological and clinical diagnosis is related to the presence of (multiple) co-pathologies.

We extracted clinical and pathological diagnostic data of 202 dementia patients from the Amsterdam Dementia Cohort (ADC) with available post-mortem autopsy reports from the Netherlands Brain Bank (NBB) between 1993-2022. Pearson chi-square tests were performed to compare the number of co-pathologies between diagnostically concordant and discordant cases, stratified by primary pathological diagnosis (R v.4.2.1).

13 primary pathological diagnoses were observed in our cohort (Table 1;Figure 1), most frequently AD (n = 89, 44%), FTLD-TDP (n = 28, 14%), and LBD (n = 23, 11%). In total, 158/202 (78%) cases had concordant clinical and primary pathological diagnoses. Amongst pathological groups the highest concordance was found in AD (80/89, 90%), followed by FTLD-TDP (25/28, 89%) and LBD (15/23, 65%). Age at death did not differ between concordant and discordant cases. Next, we characterized the occurrence of co-pathologies (Table 1;Figure 2), and observed ≥1 co-pathologies most frequently in LBD cases (21/23, 91%), followed by AD (73/89, 82%), and FTLD-TDP (22/28, 79%). In AD cases, prevalent co-pathologies were CAA (55/89, 62%), LBD (36/89, 40%), and ARTAG (15/89, 17%). In FTLD-TDP cases, prevalent co-pathologies were hippocampal sclerosis (9/28, 32%), and CAA (8/28, 29%). In LBD the most frequent co-pathology was AD (16/23, 70%). On average, the number of co-pathologies was slightly higher in concordant compared to discordant cases, but this difference was non-significant (p = 0.4). In individual pathological groups (AD, FTLD-TDP, LBD) there was also no significant difference in the number of co-pathologies between concordant and discordant cases (p_AD=0.7, p_FTLD-TDP=0.4, p_LBD=0.5).

Our results highlight the pathological heterogeneity within a large memory clinic cohort, and show that neurodegenerative diseases rarely present in isolation. Although diagnostic discordance was not explained by the number of co-pathologies or age at death, additionally investigating severity of co-pathologies might provide insight into interactions between co-pathologies and clinical phenotypes.