Activated microglia in dementia with Lewy bodies

Methods.

Results.

Control cases demonstrated few activated microglial cells expressing MHC class II antigen in the transentorhinal cortex (11 ± 3 cells/mm²). Patients with pure AD and those with DLB and AD had similar numbers of SP and both had increased microglial activation compared with control subjects (p < 0.003), but did not differ significantly from each another (90 ± 26 cells/mm² and 99 ± 34 cells/mm²) by analysis of variance (ANOVA). Pure DLB cases also showed a greater number of activated microglia than control subjects (48 ± 22 cells/mm²; p < 0.005). Although the limited number of cases studied precludes an absolute conclusion, the degree of inflammation in pure DLB patients appeared to be intermediate between control subjects and AD patients (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Microglial cells immunoreactive for major histocompatibility complex class II antigen in transentorhinal cortex. Each of the groups of demented patients demonstrated significantly more chronic inflammation than the nondemented control subjects (*p < 0.005 and **p < 0.003 by analysis of variance; bars indicate ± SEM). Although the cases of pure dementia with Lewy bodies (DLB) generally had fewer activated microglia than either of the groups with AD pathology, this difference did not reach significance.

There was a positive correlation between the number of activated microglial cells and the number of LB in different brain regions. In four of the pure DLB cases, the pattern of LB and inflammation was transentorhinal > cingulate > frontal. The one case with more LB in the frontal cortex also demonstrated more activated microglia in that region. Double-immunostained sections showed no obvious tendency for activated microglia to aggregate around individual LB or LB-bearing neurons (figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Sections of postmortem tissue from a patient with pure dementia with Lewy bodies double-immunostained for major histocompatibility complex class II and ubiquitin demonstrated no spatial relationship between activated microglial cells (black) and cortical Lewy bodies (arrows). Original magnification ×200 before reduction.

Discussion.

Demonstrating the presence of a significant inflammatory process in the cerebral cortex of patients with DLB, even in the absence of AD pathology, has several important implications. At a very simple level, it indicates that the degree of cortical injury is much more severe than might be suspected by other measures such as the number of LB or the degree of neuronal loss, both of which tend to be modest. The fact that we found activated microglia throughout the affected cortical regions and not just clustered around LB suggests that LB are a marker of a disease process that is, in fact, much more widespread.

Second, the pathologic substrate for dementia in DLB is not well understood. Conflicting results have emerged from studies examining the correlation between the severity of dementia and the number of LB, neuronal loss, synaptic density, and levels of neurotransmitters in the cerebral cortex.6,7 In many cases, the presence of coexisting AD pathology may contribute to the cognitive deficit,7 although it is clear that dementia may exist in cases of pure DLB that lack significant numbers of SP or neurofibrillary tangles. Recognizing that the cortex of a patient with DLB is in a state of chronic inflammation suggests an additional mechanism by which cerebral neurons may become dysfunctional. In addition to proinflammatory mediators, activated microglial cells secrete a variety of potentially neurotoxic substances that could contribute to neuronal degeneration or dysfunction. The situation in DLB may be similar to AD, in which it has been suggested that the accumulation of SP and neurofibrillary tangles alone is insufficient to cause dementia unless accompanied by an inflammatory response.8 In DLB, the presence of cortical LB may be the most specific pathologic marker of disease, although chronic inflammation may be an important final pathway leading to neuronal dysfunction. An interesting way to further examine this possibility would be to assess the correlation between the degree of cerebral inflammation and the severity of dementia. Such a correlation would require a cohort of DLB patients who had been evaluated using a common dementia rating scale; unfortunately this was not the case with the patients used in this study.

Finally, if cortical inflammation does contribute to the development of dementia in DLB, then there are obvious therapeutic implications. Although anti-inflammatory drugs have previously been proposed for the treatment of DLB, this recommendation was based largely on the knowledge that many patients may have coexisting AD pathology.9,10 The finding in this study of a significant inflammatory process, even in the absence of AD pathology, suggests that anti-inflammatory therapy may be of benefit in all cases of DLB.