Dissecting the Subtypes of Parkinson's Disease

Abstract & Commentary

By Claire Henchcliffe, MD, Assistant Professor, Department of Neurology and Neuroscience, Weill Cornell Medical College. Dr. Henchcliffe serves on the speaker's bureau of GlaxoSmithKline, Teva Neuroscience, Boehringer Ingelheim, Novartis, and Allergan.

Synopsis: Dementia in Parkinson's disease is associated with a non-tremor-predominant phenotype, and the MAPT H1/H1 genotype. A rapidly progressing subtype without dementia has been identified, associated with another non-motor feature, depression.

Sources: Selikhova M, Williams DR, Kempster PA, et al. A clinico-pathological study of subtypes in Parkinson's disease. Brain 2009;132:2947-2957. Williams-Gray CH, Evans JR, Goris A, et al. The distinct cognitive syndromes of Parkinson's disease: 5 year follow-up of the CamPaIGN cohort. Brain 2009;132:2958-2969.

Selikhova and colleagues examined clinical records and cerebral a-synuclein and amyloid A pathology in 242 cases with autopsy-proven Parkinson's disease (PD). They divided cases into four clinically defined groups: earlier disease onset (onset < 55 years old: EDO, 25%); onset > 55 years old that was either tremor-dominant (TD, 36%) or non-tremor-dominant (NTD, 38%); and rapid disease progression without dementia (death within 10 years of onset, 8%). Mean disease duration was not significantly different between the TD and NTD groups (13.5 years vs. 12.9 years, p > 0.05). However, the NTD phenotype was significantly more associated with dementia (61% vs. EDO 16% vs. TD 32%), present in 43% total, and associated with shorter duration from onset to death (13.8 ± 5.7 years vs. 17.0 ± 7.6 years). Significantly higher Lewy body scores were measured in NTD vs. EDO and TD, particularly in frontal and transentorhinal cortex. Neurofibrillary pathology was more highly represented in NTD than EDO or TD cases. Those with rapid disease progression were more likely to have the TD phenotype, but were also more likely to have been diagnosed with depression at PD onset (40%) vs. EDO (20%), TD (15%), or NTD (24%).

In the accompanying study, Williams-Gray and colleagues report five-year follow-up from a longitudinal cohort of 126 patients, together with a cross-sectional cohort of 386 cases. The importance of dementia was emphasized by its 17% incidence within the first five years from diagnosis. Baseline features of age > 72 years, difficulty copying intersecting pentagons, and impaired semantic fluency (20 words in 90 seconds) led to an odds ratio (OR) of 88 for developing dementia within five years of PD onset. Microtubule-associated protein tau (MAPT) gene alleles H1 and H2 were tested. Presence of the H1/H1 genotype resulted in an OR of 21.1 for development of dementia. The authors suggest this is likely due to an effect of genotype upon levels of tau, since in autopsy brain tissue from patients with PD or dementia with Lewy bodies (n = 61), presence of the H1 allele increased 4-repeat tau transcription by approximately 20%.


Parkinson's disease (PD) is both clinically and genetically heterogeneous. These two studies use different approaches to suggest how better understanding the various subtypes might be translated into the clinic to determine a more accurate prognosis. First, it has long been thought that tremor predicts a more benign course. This is challenged by Selikhova's finding of a rapidly progressing subgroup of cases, most of whom fall into the tremor-dominant (TD) category. Their data suggest that in the TD subtype, "red flags" include not only faster motor evolution and development of axial symptoms, but also older age of onset and early depression. Older-onset patients with depression may therefore warrant more careful and frequent monitoring, despite a TD phenotype. Second, dementia in PD is distressing and common, but not universal. Selikhova found that a bradykinesia (rather than tremor) -dominant phenotype was strongly associated with clinical dementia, as well as heavier cortical Lewy body load and Alzheimer's-disease-related pathology. The authors suggest this association has a biological basis, but their study does not address what that might be. However, in the accompanying study, presence of the MAPT H1/H1 genotype was associated with more than a 20-fold increased risk of dementia. This strongly suggests a role for tau protein in the pathology underlying dementia in PD. Despite the need for caution in interpreting this finding (post-mortem confirmation of diagnosis is needed), it does suggest the future possibility for adding genotyping to careful phenotypic subtyping to stratify risk for dementia in PD.