Clinical and pathologic features of two groups of patients with dementia with Lewy bodies: effect of coexisting Alzheimer-type lesion load

Abstract

The objectives of this study were to examine the clinical and pathologic features of two subgroups of patients with dementia with Lewy bodies (DLB) differing in Alzheimer disease (AD)-type pathology load and to identify clinical variables useful in the differential diagnosis from AD. The records of 64 consecutive demented patients were reviewed. Pathologic diagnoses were independently established [35 AD cases, 11 cases of pure dementia with Lewy bodies (pDLB), and 18 cases of combined AD plus Lewy bodies (AD+LB)], and several neurodegenerative lesions were quantified. Clinical and pathologic data were compared between groups with univariate and multivariate analyses. Compared with the other groups, pDLB cases had more frequent acute-subacute onset of dementia [45% vs. AD (3%) and AD+LB (16%)], early parkinsonism [45% vs. AD (0%) and AD+LB (0%)], early [27% vs. AD (0%) and AD+LB (0%)] and late [73% vs. AD (11%) and AD+LB (16%)] hallucinations, fluctuating course [46% vs. AD (9%) and AD+LB (22%)], delusions [45% vs. AD (11%) and AD+LB (6%)], spontaneous parkinsonism [63% vs. AD (8%) and AD+LB (16%)], less frequent ideomotor apraxia and loss of insight, earlier urinary incontinence [3.2 +/- 1.4 years after onset vs. AD (6.3 years) and AD+LB (5.8 years)], shorter duration of dementia [7.7 +/- 2.4 years vs. AD (9.6 years) and AD+LB (11 years)], milder atrophy in computed tomography scans, greater brain weight, more transcortical spongiosis, wider cortex and subcortex, and less amyloid angiopathy. All pDLB cases but no AD cases had abnormal CA2 neurites. The clinical features of AD+LB patients were similar to those of AD patients other than more frequent acute-subacute onset and fluctuating evolution. Discriminant analyses selected four clinical variables differentiating pDLB from the other two groups as a whole: acute-subacute onset, early parkinsonism, early hallucinations, and early onset of urinary incontinence. Two or more of these features identified pDLB with a sensitivity of 81.8% and a specificity of 95.9%. Differentiation between the three groups (pDLB, AD+LB, and AD) or between both groups with LB (DLB) from AD could be only attained in 70% of cases. We conclude that early symptomatology is the main clue for the diagnosis of pDLB. We identified by discriminant analysis a set of clinical diagnostic criteria similar to those proposed by the Consortium on Dementia With Lewy Bodies. Accuracy was excellent for the diagnosis of pDLB but only mediocre for separating AD+LB as well as the entire DLB group from AD.