Clinical and quantitative pathologic correlates of dementia with Lewy bodies

OBJECTIVES: To examine the distribution of cortical Lewy bodies (LB) and their contribution to the clinical syndrome in dementia with LB (DLB) and to address their relationship to the pathologic markers of AD and PD. METHODS: We studied 25 cases meeting neuropathologic criteria for DLB: 13 cases without AD (Braak stage I or II) and 12 cases with concomitant AD changes (Braak stages III to V). Age at onset, disease duration, and clinical symptoms were reviewed for each case. We quantified the regional distribution of LB in substantia nigra, paralimbic areas (cingulate gyrus, insula, entorhinal cortex, and hippocampus), and neocortex (frontal and occipital association areas) using anti-alpha-synuclein immunostaining. We compared the LB pathology between groups of patients with different symptoms at onset or with specific clinical phenotypes. RESULTS: There were no significant differences in clinical symptoms or LB density between cases with or without concomitant AD. LB density showed a consistent gradient as follows: substantia nigra > entorhinal cortex > cingulate gyrus > insula > frontal cortex > hippocampus > occipital cortex. LB density in substantia nigra and neocortex was not significantly different in cases that started with parkinsonism compared with those that started with dementia. There were no significant differences in LB density in any region among patients with or without cognitive fluctuations, visual hallucinations, delusions, recurrent falls, or parkinsonism. Duration of the disease correlated with a global LB burden for each case (p = 0.02) but did not correlate with LB density in any individual area. Paralimbic and neocortical LB density were highly correlated with each other (p<0.0001), but neither of these correlated well with the number of LB in substantia nigra. LB density did not correlate with Braak stage or frequency of neuritic plaques. CONCLUSIONS: There is a consistent pattern of vulnerability to LB formation across subcortical, paralimbic, and neocortical structures that is similar for DLB cases with or without concomitant AD. Paralimbic and neocortical LB do not correlate with LB in substantia nigra, suggesting that DLB should not be considered just a severe form of PD. LB density correlates weakly with clinical symptoms and disease duration.