Neuropathology of aminergic nuclei in Alzheimer's disease

Abstract

In order to comprehensively evaluate pathological involvement of the locus coeruleus (LC) and cortically projecting raphe nuclei in Alzheimer's disease (AD), we have recently completed a study (Zweig, et al., 1988) in which numbers of neurons containing neuromelanin within the LC and large nucleolus-containing neurons within the dorsal raphe nucleus (DR) and the central superior (raphe) nucleus (CSN) were determined in 25 cases of AD and 12 age-matched controls. Numbers of neurofibrillary tangles (NFTs) within these regions were also counted. Pathological results were compared with clinical data, including psychiatric evaluations, available for 21 of the AD cases. Neuronal loss in AD cases was most severe within LC at mid level (p less than .01) and within DR, caudally (p less than .05). Counts of NFTs within LC were also highest at mid level (p less than .05) in comparison with caudal level). Neuronal loss was not demonstrated within CSN, although NFTs were abundant within this nucleus. At individual levels, neuronal and NFT counts did not correlate. Relative severity of neuronal loss or NFTs was usually consistent from level to level within nuclei; internuclear correlations were weaker. Cases of AD complicated by depression had significantly fewer neurons at mid LC and rostral CSN levels than nondepressed cases (p less than .05). There was also a trend (nonsignificant) suggesting increased neuronal loss at all levels of LC and DR in depressed cases. Neuronal loss correlated negatively with age, particularly within LC. NFT counts correlated negatively with duration of illness, particularly within DR. As all but 3 individuals had severe dementia, NFT counts may reflect rate of progression of disease. Neuronal loss and NFTs frequently occur in the LC and cortically projecting raphe nuclei in AD (Curcio and Kemper, 1984, Hirano and Zimmermann, 1962, Ishii, 1966, Marcynuik et al., 1986, Yamamoto and Hirano, 1985, Bondareff and Mountjoy, 1986, Iverson et al., 1983, Mann et al., 1985, Tabaton et al., 1986). We have recently completed a comprehensive evaluation of pathological involvement of the LC, the DR, and the CSN in a series of aged controls and AD patients for whom detailed clinical information, including psychiatric evaluations, were available (Zweig et al., 1988). This report summarizes the findings of that study.