However, McCarron et al. [26] did not comment as to whether those haemorrhagic cases with see more APOE ε2 allele also displayed capillary CAA, although this might be worth further investigation. In the present study, the severity of cortical SP was found to be independent of APOE ε4 allele frequency. Previous studies have reported conflicting results. Rebeck et al. [27] reported a greater frequency of SP in APOE ε4 allele homozygotes compared with non-ε4 carriers. However, Greenberg et al. [19] found no difference in SP density in APOE ε4 allele heterozygotes compared with homozygotes, but

did find fewer SP in non-ε4 allele bearers. Attems et al. [16] noted only a weak correlation between SP density and possession of APOE ε4 allele. However, others [11, 15, 21] noted that Aβ plaque count was not associated with possession of APOE ε4 allele. These apparent discrepancies may be explained by a consideration of the actual Aβ peptide species within SP. We have noted that plaque levels of Aβx-42 in AD did not vary according to APOE ε4 allele, but those of Aβx-40 increased in line with APOE ε4 allele copy number [28]. As all morphological forms of SP (that is, both cored

and diffuse) contain Aβx-42, whereas ICG-001 Aβx-40 is present largely, or only, in cored plaques, antibodies, such as 4G8, which are not end-specific to Aβx-40 or Aβx-42, will therefore detect all morphological forms of SP and thus overall ‘counts’ will essentially reflect the numbers/density of Aβx-42-containing SP. The relationship

between Aβ plaque density and CAA appears less clear. Although the present study did not specifically address any possible correlation between the two pathological entities, it was noted that the severity of Aβ plaque deposition did not significantly differ across the four separate phenotypes. Despite this, both Tian et al. [29] and Chalmers et al. [15] reported a negative association between Aβ plaques and CAA severity, whereas others many have suggested that Braak stage (NFT density) is a better correlate with CAA than is SP density [16]. Because of potentially increased risks of associated cerebral haemorrhage or infarction, it is important to be able identify ways of diagnosing CAA during life, particularly the more extensively and severely affected cases. Knowing the APOE genotype may contribute to being able to more accurately predict the type of CAA present, and therefore associated risk. Nonetheless, as shown here, the heterogeneity in pathology with regards to CAA fails to be explained by APOE genotype alone. As findings from Genome Wide Association Studies (GWAS) increase [30, 31], it is possible that risky variants with certain AD susceptibility loci might be identified which selectively promote one type of pathological phenotype over another.