By Harini Sarva, MD

Assistant Professor of Clinical Neurology, Weill Cornell Medical College; Assistant Attending Neurologist, NewYork Presbyterian Hospital

Dr. Sarva reports no financial relationships relevant to this field of study.

SYNOPSIS: This observational study of 807 autopsy-confirmed cases of Lewy body disease (LBD) used Braak neurofibrillary tangle staging, frequency of neuritic plaques, and Lewy body staging to demonstrate that those individuals who had diffuse LBD had a shorter disease duration than those with transitional LBD localized to the limbic region.

SOURCE: Graff-Radford J, Aakre J, Savica R, et al. Duration and pathologic correlates of Lewy body disease. JAMA Neurol 2017; doi:10.1001/jamaneurol.2016.4926 [Epub ahead of print].

This retrospective, observational study of autopsy findings and clinical data from 807 subjects with Lewy body disease (LBD) from the National Alzheimer’s Coordinating Center included individuals who had transitional LBD and diffuse LBD. Clinical data included age at onset of cognitive symptoms, sex, Unified Parkinson’s Disease Rating Scale scores, Mini Mental State Exams (MMSE), neuropsychological evaluations, age at death, and presence of core dementia with Lewy bodies (DLB) features (hallucinations, fluctuations, parkinsonism, and probable REM sleep behavior disorder [RBD]). The mean age of onset of cognitive decline was 70.9 years, and the mean age at death was 79.2 years. Sixty-three percent were men. Vascular pathology also was assessed — 126/806 had microinfarcts; 94/656 had lacunes; and 36/654 had large vessel infarcts. Of the 255 with transitional LBD, 73.7% met criteria for high or intermediate likelihood of dementia caused by Alzheimer’s disease (AD) pathology; 72.6% of 401 subjects with diffuse LBD met criteria for AD. Those with transitional LBD were older at time of death and had less parkinsonism and neuritic plaques compared with those with diffuse LBD.

Of the original 807 subjects, 766 had information on disease duration. Among all 766 subjects, univariate regression models suggested that older age and male sex were associated with shorter disease duration from time of cognitive symptom development. The presence of APOEε4 did not correlate with disease duration. After adjusting for age at onset, gender, and APOEε4 status, higher Braak staging (neurofibrillary tangles in the cortex) and increased frequency of neuritic plaques did not correlate with disease duration. However, between-group comparison suggested that those with diffuse LBD had shorter disease duration compared with those with transitional LBD. After correction for age, gender, APOEε4 status, Braak staging, and frequency of neuritic plaques, the presence of lacunar infarcts was associated with a shorter disease duration.

When specifically analyzing those with clinical diagnosis of DLB during life (n = 238), older age, positive APOEε4 status, and diffuse LBD pathology were associated with shorter disease duration, whereas male gender, Braak staging, neuritic plaques, and transitional LBD were not. When comparing the clinical features of diffuse LBD and transitional LBD, the former had more of the core features of DLB and a faster cognitive decline, not only in the MMSEs but also in the Category Fluency Test and the Wechsler Adult Intelligence Scale—Revised Digit Symbol subtest. Of the four core features of DLB, the two that had the best correlation with shorter disease duration were parkinsonism and probable RBD reported at the final visit.


DLB is the second most common form of dementia after AD and has a shorter disease duration than AD. Despite known diagnostic criteria, DLB is still difficult to accurately recognize clinically, because of the clinical variability of these patients and the pathological overlap with AD and vascular dementia. Predicting prognosis of these patients is even more challenging. This complexity in diagnosis not only hampers our ability to symptomatically treat early but also to prognosticate and prepare the patient and families for the future course of the disease.

This study by Graff-Radford et al further adds to the understanding of the different subgroups of LB pathology and the potential clinical trajectories. However, the retrospective, autopsy-driven nature of this study reiterates the need for better diagnostic assessments during life that would enable long-term observational studies of living cohorts. The development of reliable LB pathology-specific functional imaging may aid in better earlier diagnosis, particularly with the presence of AD and vascular pathology in this cohort. With such a high percentage of those with AD-like pathology, the role of amyloid in disease duration was not assessed and also may contribute to shorter disease duration. From a treatment standpoint, we know that atypical parkinsonian patients, such as those with DLB, do not have a consistent robust response to levodopa, yet those with parkinsonism at the final clinical evaluation were found to have a shorter disease duration. Whether parkinsonism correlates to the diffuse pathological changes or motoric issues (falls, contractures from severe rigidity) from poorly treated parkinsonism is unclear. Lastly, the importance of treating vascular risk factors in conjunction with parkinsonism and RBD again is raised, with the correlation between shorter disease duration and lacunar strokes, which not only add to cognitive burden but can lead to significant morbidity. Thus, while adding to our knowledge of pathology, this study highlights the need for the development of reliable biomarkers to study these patients during their lifetimes to accurately correlate clinical findings and prognosis.