In past years, a small subgroup of patients with GD was found to have a propensity to develop parkinsonism during their adult life. Most of the patients have type I GD. Parkinson disease (PD) is a neurodegenerative disease that is common among adults throughout the world, affecting 1% of people over the age of 65. PD is extraordinarily complex in cause and capacity to be treated. The presenting manifestations include bradykinesia, tremor, rigidity, asymmetric manifestations and postural instability. The neuropathology includes a loss of dopamine neurons and the eventual depigmentation of the substantia nigra (SN).
The first indication of a relationship between GD and parkinsonism came in the form of scattered case reports in the medical literature.4-6
Each described a patient with Gaucher disease, or a family member who developed symptoms of parkinsonism. In our clinic, concern first arose about a possible association between the two diseases in the late 1990’s when we evaluated and treated a woman with relatively mild Gaucher disease who developed a tremor at age 42, followed by rapid deterioration of her gait.6
A pallidotomy performed at age 47 was not helpful, and motor and cognitive decline ensued despite enzyme replacement therapy for Gaucher disease. While the presence of these two disorders could have been unrelated, the previous case reports as well as scattered clinical descriptions from colleagues prompted a solicitation and evaluation of similar cases.
Within a few years, we collected seventeen male and female individuals from several ethnicities, described in a 2003 report.7
These subjects suffered from mild Gaucher manifestations with a mean age at diagnosis of 35 years. However, their parkinsonism presented relatively early, at a mean age of 48 years. Classic parkinsonian symptoms were described. Although four patients were treated with enzyme replacement therapy, Parkinson disease symptoms persisted. Additionally, several subjects described a family history of parkinsonism, even among family members without GD. Of note, while a variety of different genotypes were encountered among the 17 probands, mutation N370S, previously considered to be associated exclusively with non-neuronopathic GD, was the most common allele found.
More recently, subjects carrying GBA
mutations were evaluated prospectively at the NIH Clinical Center with detailed neurological examinations.8
The findings of ten subjects (7 males, 3 females) with parkinsonism and GBA
mutations were recently summarized. Mutations N370S, L444P, c.84dupG and recombinant alleles were all identified in patient DNA samples. The subject group had a mean age of disease onset of 49, disease duration was 7.8 years (1.2−16), and the mean Uniform Parkinson Disease Rating Scale (UPDRS III) motor score off therapy was 26.3. Cognitive changes were reported by half of the subjects, substantiated by formal neuropsychiatric testing. Six patients were diagnosed with Parkinson disease, three patients had Lewy body dementia, and one further patient in the series was diagnosed with a “Parkinson plus” syndrome. Atypical manifestations included electroencephalographic abnormalities, myoclonus and seizures. Olfactory dysfunction was the most common non-motor finding. Thus, it appears that glucocerebrosidase mutations are associated with a broad spectrum of synucleinopathies, including classic PD, and less common disorders such as LBD. This study is being expanded, and is utilizing different functional and imaging techniques to identify potential biomarkers for PD in this population, as well as a better phenotypic assessment of these subjects.
The next major finding was the discovery of an increased incidence of GBA
mutations in cohorts with parkinsonism. One study screened DNA extracted from brain tissue of patients with Parkinson disease for mutations in the GBA
Mutations were found more frequently than expected. Of the 57 different DNA samples collected from five separate American brain banks, eight (14%) had confirmed mutant alleles; two of these mutant samples were GBA
homozygotes, and six were heterozygotes. Five of the identified mutant alleles had the common N370S mutation, which is associated with non-neuronopathic GD. This study also included 44 age-matched controls without pathological evidence of PD, none of which carried mutations. The GBA
mutations were found in DNA samples from patients with an earlier age of onset. DNA samples were also collected from brain bank samples in Britain, revealing that two (8%) of the 26 cases with PD carried GBA
Other studies have screened cohorts of Parkinson patients for one or more specific GBA
mutations. One such clinic-based study in Northern Israel screened 99 Ashkenazi patients with classic PD symptoms.11
DNA samples were tested for several common glucocerebrosidase mutations. Strikingly, 31 (31.3%) patients screened carried a N370S or c.84insG mutation, representing a five-fold increase above the two control groups screened, 74 patients suffering from Alzheimer's disease (AD), and 1543 young Ashkenazi controls. The controls collected were less than ideal however, because of the lack of both selectivity and neurological evaluations in the control groups.
These findings have since been supported by studies of GBA
mutations in other groups of patients with PD around the world. A study conducted by investigators at Columbia University screened 160 Ashkenazi Jewish probands with PD and 92 clinically evaluated control subjects of Jewish ancestry from a New York City clinic.12
When each subject was screened for the N370S mutation, 17 probands (10.7%) carried the mutation while only 4.3% of the controls had a mutation. Later, the same group revisited the question with a larger sample, sequencing GBA
in a cohort of 278 subjects with PD (178 Ashkenazi Jews) alongside 179 controls (85 Ashkenazi Jews) enrolled in the Genetic Epidemiology of Parkinson Disease (GEPD) study.13
They reported that 13.7% of patients with PD carried mutations in GBA
, while only 4.5% of the controls were positive for mutations. GBA
mutations were most frequent among patients with an age at onset below 50 years of age (22%). In this study, the age at onset was 1.7 years earlier among patients with GBA mutations, when controlling for age, sex, family history, and Jewish ancestry. A larger study in Ashkenazi Jews conducted by Gan-Or, et al
. in Israel examined a cohort of 420 PD patients, 333 elderly controls, and 3805 young controls, all of whom were screened for eight common GBA
The elderly and PD patients were also screened for the LRRK2 G2019S mutation, and while one-third of the patients screened carried this mutation, only four had both G2019S and a GBA
mutation. In this study, GBA
mutations were found in 17.95% of the PD patients, a significantly higher frequency than the 4.2% observed among elderly and 6.35% among young controls. Among the identified mutations were two early frame-shift mutations, which conferred a high odds ratio. Finally, this study again found that the age at onset of PD was lower among those carrying GBA
There have also been studies conducted in non-Jewish cohorts. A study from Toronto screened 88 unrelated Canadian PD subjects for seven different mutations in the GBA
gene, including two very rare alleles.15
The subjects were selected because of either a family history of PD, or because of early onset of symptoms. The control group consisted of 122 clinically screened Canadians. Mutations were observed in 5.6% of patients diagnosed with PD, while only 0.8% of the controls carried a mutation.
Two further studies performed in South America gave similar results. One study using samples from Venezuela sequenced the entire GBA
gene among 33 PD patients and 29 control subjects. Four (12%) unrelated PD patients carried GBA
mutations, each with earlier onset disease.16
In a study from Brazil, 62 probands were investigated for GBA
Two (3.5%) were found to have common mutations while none were found among 267 age and sex-matched controls.
Another study initiated by Bras, et al
. examined a cohort of Portuguese PD patients.18
A total of 230 patients were examined and compared to 430 controls. The PD sample group identified 14 heterozygotes for GBA
mutations (N370S, N396T, D409H, and L444P), while the control group had three N370S. Additionally, two polymorphisms, E326K and T369M, previously described as non-pathogenic, were found among controls and patients. Two novel variants and one mutation of unknown significance were also identified. A 6.1% frequency of GBA
mutations was observed among the PD patients, compared to a frequency of 0.7% among the controls. Since in this study, as well as others using non-Ashkenazi cohorts, the GBA
mutations identified included many rare alleles, it appears to be necessary to sequence all patient samples in studies among non-Ashkenazi PD patients.
Several additional studies have screened for the presence or absence of two common GBA
mutations, N370S and L444P in large Caucasian PD cohorts. Mata, et al.
found the two mutations in 2.9% of 721 PD subjects from Seattle versus 0.4% of 554 controls.19
However, Toft, et al.
, screening 318 probands from Norway found a mutation frequency of 2.3 %, which did not differ significantly from the 1.7% seen in 412 controls.20
A group in Italy reported the presence of one of these two mutations in 2.8% of patients and 0.2% of controls.21
Because it had been argued that ethnicity could have confounded our initial results, we focused on a specifically non-Jewish cohort of 184 Chinese patients from Taiwan with Parkinson disease, as well as 92 clinically screened and ethnically/age-matched controls using DNA plates from the NINDS Human Genetic Resource Center (Coriell).22 GBA
sequencing demonstrated that mutations were present in 5.4% of the PD patients, while only in 1.1% of controls. Several mutant alleles were found including mutations L444P and D409H, as well as rare or novel mutations such as L174P, R131S, R163Q, S271G, and Q497R. These results demonstrated that the link between Gaucher and Parkinson diseases was independent of the ethnicity of the group studied. These results were also confirmed in two other Asian PD cohorts.23, 24
A recent paper explored the frequency of GBA
alterations in familial cases of PD.25
In this study, 12.6% of familial PD patients screened were shown to have GBA
alterations, while they were only seen in 5.3% of controls. However, when subjects carrying E326K and T369M were removed, the frequency of other GBA
mutations was 4.1% in cases versus 1.1% in controls, which is similar to non-familial cohorts.
For the most part these studies have independently reached similar results demonstrating that GBA mutations are seen in patients with PD at a higher frequency than expected. Although it is premature to attempt to determine the risk of PD among GBA mutation carriers, even heterozygotes appear to be at increased risk. A large meta-analysis, pooling genotypic data from 16 different centers from around the world is in progress. This study will more definitively establish the Odds Ratios across different ethnicities and provide information that ultimately may be useful for genetic counseling.