Evolution of Prodromal Clinical Markers of Parkinson Disease in a GBA Mutation–Positive Cohort

Importance  Numerically, the most important genetic risk factor for the development of Parkinson disease (PD) is the presence of a glucocerebrosidase gene (GBA) mutation.

Objective  To evaluate longitudinally and clinically a GBA mutation–positive cohort and the evolution of the prodromal features of PD.

Design, Setting, and Participants  Participants in a study of the etiology and prodrome of PD were reevaluated in this clinic-based 2-year follow-up report. Patients with type 1 Gaucher disease (GD) and heterozygous GBA mutation carriers were recruited in 2010 from the Lysosomal Storage Disorder Unit at the Royal Free Hospital, London, England. Thirty patients who previously received a diagnosis of type 1 GD, 28 heterozygous GBA mutation carriers, and 26 genetically unrelated controls were included. Exclusion criteria included a diagnosis of PD or dementia for both the patients with GD and the GBA mutation carriers and any existing neurological disease for the controls.

Main Outcomes and Measures  Assessment was performed for clinical markers using standardized scales for hyposmia, rapid eye movement sleep behavior disorder, depression, autonomic dysfunction, cognitive function, and parkinsonian motor signs (using the Unified Parkinson’s Disease Rating Scale motor subscale [UPDRS part III]).

Results  Over 2 years, depression scores were significantly worse for heterozygous carriers (mean baseline, 0.65; mean follow-up, 2.88; P = .01), rapid eye movement sleep behavior disorder scores were significantly worse for patients with GD (mean baseline, 0.93; mean follow-up, 2.93; P < .001) and heterozygotes (mean baseline, 0.10; mean follow-up, 2.30; P < .001), and UPDRS part III scores were significantly worse for patients with GD (mean baseline, 4.29; mean follow-up, 7.82; P < .001) and heterozygotes (mean baseline, 1.97; mean follow-up, 4.50; P < .001). For controls, there was a small but significant deterioration in the UPDRS part II (activities of daily living) score (mean baseline, 0.00; mean follow-up, 0.58; P = .006). At 2 years, olfactory and cognitive assessment scores were lower in patients with GD and heterozygotes compared with controls, but they did not differ significantly from baseline. When the results from the patients with GD and the heterozygotes were combined, a significant deterioration from baseline was observed, as reflected in the Rapid Eye Movement Sleep Behaviour Disorder Questionnaire (mean baseline, 0.51; mean follow-up, 2.63; P < .001), Beck Depression Inventory (mean baseline, 1.72; mean follow-up, 4.44; P = .002), and UPDRS part II (mean baseline, 0.88; mean follow-up, 2.01; P < .001) and part III scores (mean baseline, 3.09; mean follow-up, 6.10; P < .001) (all P < .01), and at 2 years, significant differences in University of Pennsylvania Smell Identification Test, Unified Multiple System Atrophy Rating Scale, Mini-Mental State Examination, Montreal Cognitive Assessment, and UPDRS part II and part III scores were observed between patients with GD/heterozygotes and controls (all P < .05).

Conclusions and Relevance  This study indicates that, as a group, GBA mutation–positive individuals show a deterioration in clinical markers consistent with the prodrome of PD. Within this group of individual, 10% appear to be evolving at a more rapid rate.

Homozygous GBA mutations cause Gaucher disease (GD), a lysosomal storage disease. It is presently estimated that homozygous or heterozygous GBA mutations confer a 20- to 30-fold increased risk for Parkinson disease (PD),^1,2 and at least 7% of patients with PD have GBA mutations.^2,3 The percentage is higher among the Ashkenazi Jewish population.⁴ The percentage of GBA mutation carriers who develop PD has been estimated to be 13.7% at 60 years of age and 29.7% at 80 years of age,⁵ and so a method to determine individual risk for PD expression in this population would be very valuable. In addition, those with dementia with Lewy bodies are 8 times more likely to carry a mutation in GBA than healthy controls, which suggests a role for GBA mutations in other Lewy body disorders.⁶

For any neuroprotective treatment or disease-modifying therapy to be most effective, PD should be detected at as early a stage as possible. The deposition of α-synuclein is not restricted to the brain; for example, deposits have been found in the olfactory bulb, the peripheral nervous system, the enteric nervous system, the heart, and the pelvic plexuses.⁷ This pathology probably underlies the early nonmotor manifestations of PD, which may precede the onset of more typical PD motor symptoms by several years.⁸

Candidate biomarkers have been proposed and may be useful objective measures for the early detection of PD.^9,10 In our study, we have used early clinical markers to quantify nonmotor symptoms such as hyposmia, rapid eye movement sleep behavior disorder, depression, cognition, and autonomic dysfunction.

The aim of our study was to provide longitudinal data on a GBA-positive cohort at high risk for the development of PD and to identify biomarkers or symptoms indicating progression to early PD. The first clinical evaluation of this cohort has been published,¹¹ and the results presented here represent the 2-year follow-up.

Patients with type 1 GD were recruited from the Lysosomal Storage Disorder Unit at the Royal Free Hospital in London, England, in 2010. Potential heterozygous GBA mutation–positive relatives (78.6% of parents, 10.7% of siblings, and 10.7% of children >21 years of age) and genetically unrelated controls (spouses or partners) were identified by obtaining a detailed family history from each patient with GD, and these heterozygotes and controls were then recruited to participate in our study. Participants were also recruited from the UK Gauchers Association. In all, this unique cohort included 135 participants. Among them, 90 participants have been followed up longitudinally, with target follow-up assessments at 2-year intervals beginning in 2012. Exclusion criteria included a diagnosis of PD or dementia for both the patients with GD and the GBA mutation carriers and any existing neurological disease for the controls. The diagnosis of PD was made according to the UK Parkinson’s Disease Society Brain Bank Criteria.¹² Dementia was diagnosed according to Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) criteria for patients with a Mini-Mental State Examination (MMSE) score of 24 or less. The GBA mutation status in all participants was confirmed by Sanger sequencing of the GBA gene, as previously described.¹¹ The senior researcher was blinded to genotype. Our study was approved by the Hampstead Research Ethics Committee (reference 10/H0720/21). All participants provided written informed consent.

Of the 90 participants who were evaluated at baseline (2010-2011),¹¹ 4 (4.4%) were lost to follow-up because they either declined to participate (n = 2) or could not be contacted (n = 2). In addition, 2 participants (2.2%) died (one died of pneumonia, and the other died of breast cancer). Therefore, 84 participants (93.3%) (30 participants who previously received a diagnosis of type 1 GD, 28 heterozygous GBA mutation carriers, and 26 controls) completed the follow-up evaluation that comprised a structured clinical workup, a standardized clinical history, and a complete neurological assessment, which included the use of the Unified Parkinson’s Disease Rating Scale for activities of daily living and the motor subscale (UPDRS parts II and III, respectively), the University of Pennsylvania Smell Identification Test (UPSIT) for olfactory function, the MMSE and the Montreal Cognitive Assessment (MoCA) for cognitive function, the Rapid Eye Movement Sleep Behaviour Disorder Questionnaire (RBDQ), the Beck Depression Inventory (BDI) for depression, and a subscale of the Unified Multiple System Atrophy Rating Scale (UMSARS) for autonomic dysfunction. Anosmia was interpreted using age- and sex-adjusted normative scores (http://sensonics.com/). All participants were examined independently by a physician trained in movement disorders (M.B.). All procedures were performed and scored identically at follow-up to those performed at baseline. A senior neurologist who is an expert on movement disorders (A.H.V.S.) evaluated participants whenever there was a significant difference between follow-up and baseline UPDRS scores.

The data were analyzed using IBM SPSS Statistics version 21. To assess the differences between the group mean values across the 2 different time points, we performed a 2-way analysis of covariance with the factors “group” (eg, patients with GD vs carriers vs controls) and “time” (time 1 vs time 2). The covariates age, sex, education, and family relationship were added to the design matrix in order to account for differences in these mean values between the groups. Post hoc tests were used to compare the groups at follow-up. Paired t tests were used to compare the scores within each group before and after follow-up. Differences in age, sex, and ethnicity between groups were checked using the 1-way analysis of variance and the χ² test. We also accounted for performing multiple statistical tests across our dependent variables (the UPSIT, UMSARS, RBDQ, MMSE, MoCA, UPDRS part II, UPDRS part III, and BDI scores) by defining a significance threshold for statistical tests of P < .05 and by correcting this for multiple comparisons using the Benjamini-Hochberg false discovery rate. In brief, this procedure involves ordering all P values in ascending order and applying a sequential threshold.

The 84 participants (40 men [47.6%]) had a mean (SD) follow-up duration of 1.9 (0.2) years (range, 1.5-2.3 years). The demographic, clinical, and genetic characteristics, along with statistical comparisons, of the cohort are shown in Table 1. The participants with type 1 GD did not differ significantly from the heterozygous GBA mutation carriers or the controls in terms of age, sex, and ethnicity (all P > .05, determined by use of 1-way analysis of variance and the χ² test). Both the patients with type 1 GD and the heterozygous GBA mutation carriers were significantly more likely than the controls to have a family history of PD (P = .03). As described previously,¹¹ the most common genotype in patients with type 1 GD was N370S/L444P (11 of 30 patients [36.7%]). None of the patients with type 1 GD had features of type III GD (such as generalized seizures or progressive myoclonic epilepsy). For the carriers, the most common genotype was N370S (14 of 28 carriers [50.0%]).

The scores with regard to the prodromal clinical features of PD at baseline and follow-up are reported in Table 2 (see also Figure). Please refer to Table 2 for the exact P values. There was a significant deterioration for GD patients over the mean 2 years of follow-up as reflected in the RBDQ, UPDRS part II, and UPDRS part III scores. Over the same period, the GBA mutation carriers showed a significant deterioration as reflected in the RBDQ, UPDRS part II, UPDRS part III, and BDI scores. There was a marginal but significant deterioration only in the matched controls with regard to the UPDRS part II score. There was no difference between baseline and follow-up scores for all groups for assessments of olfaction, cognition, and autonomic dysfunction.

At the 2-year follow-up, the patients with GD showed a significant difference in mean UPSIT, MMSE, MoCA, UPDRS part II, and UPDRS part III scores when compared with controls. Similarly, at 2 years, GBA mutation carriers showed a significant difference in mean follow-up UPSIT, MMSE, and MoCA scores when compared with controls. When the patients with GD and the GBA mutation carriers were compared at baseline, there was a significant difference in the mean BDI score. At the 2-year follow-up, the patients with GD demonstrated significantly worse mean BDI, UPDRS part II, and UPDRS part III scores compared with the carriers. There was no significant difference in mean UPSIT, UMSARS, MMSE, MoCA, or RBDQ scores between the patients with GD and the carriers at follow-up.

When the results from individuals with homozygous or heterozygous mutations in GBA were combined in a secondary, pooled analysis (Table 3; eFigures 1 and 2 in the Supplement), a significant deterioration in GBA mutation–positive individuals over the 2 years of follow-up was found with regard to mean RBDQ, BDI, UPDRS part II, and UPDRS part III scores. At baseline, GBA mutation–positive individuals showed significant differences in mean UPSIT and MoCA scores when compared with controls.¹¹ At the 2-year follow-up, GBA mutation–positive individuals showed significant differences in mean UPSIT, UMSARS, MMSE, MoCA, UPDRS part II, and UPDRS part III scores when compared with controls.

At baseline, 3 patients with GD had parkinsonian motor signs, but this was insufficient for a diagnosis of PD. As described previously,¹¹ patient GD05 (male, 78 years of age, and Ashkenazi Jewish) had bilateral rigidity with activation maneuver, asymmetric bradykinesia of all limbs, and gait impairment. Patient GD18 (male, 83 years of age, and Ashkenazi Jewish) had a left-arm rest tremor and bilateral arm rigidity with activation maneuver. Patient GD27 (male, 69 years of age, and white British) had flexed posture, bilateral rigidity, and postural and kinetic tremor of the upper limbs. At follow-up, the parkinsonian signs present at baseline in these patients had worsened but did not meet the diagnostic criteria for PD.¹³ Patient GD05 had developed a tremor in both hands (intermittent, present at rest, and worse on intention). Patient GD18 now had bilateral rigidity without activation maneuver or gait impairment. Patient GD27 had developed a head tremor, and the postural and kinetic tremor of his upper limbs had worsened (now present at rest). In addition, 1 patient who did not have parkinsonian signs at baseline had developed them at follow-up. Patient GD11 (male, 73 years of age, and Ashkenazi Jewish) had developed a very slight tremor of his right thumb (non–pill-rolling) present at rest but with no other features of parkinsonism.

Similarly, 2 GBA carriers had parkinsonian motor signs at baseline. As described previously,¹¹ patient C17 (female, 78 years of age, and white British) had bilateral rigidity, masklike facies, and bradykinesia, whereas patient C31 (male, 78 years of age, and white British) had masked facies, bilateral rigidity with activation maneuver, left-arm kinetic tremor, and flexed posture. At follow-up, the parkinsonian signs present in these patients remained unchanged from baseline.

When specific patients with GD and specific GBA heterozygotes, both with features of parkinsonism (6 of 58 participants [10.3%]), were excluded, the follow-up data remained significant. The remaining patients with GD and the remaining GBA mutation carriers (52 of 58 of participants [89.7%]) still showed a significant deterioration as reflected in the RBDQ, UPDRS part II, UPDRS part III, and BDI scores after 2 years (eTable 1 in the Supplement).

When the clinical markers of specific patients with GD (patients GD05, GD11, GD18, and GD27) and specific GBA heterozygotes (patients C17 and C31), both with parkinsonian motor signs, were compared with those of GBA mutation–positive individuals without features of parkinsonism, there were significant differences in age (P = .002) and cognition (P = .009) at baseline (eTable 2 and 3 in the Supplement). Baseline UPSIT scores were also noted to be lower for those individuals with features of parkinsonism, but this difference did not reach statistical significance.

Our study was designed to investigate the progression of clinical biomarkers in a cohort of individuals at high risk for PD. Our results demonstrate that clinical features associated with premotor PD and motor features of PD have both evolved since the initial testing, and they support the hypothesis that some GBA mutation–positive individuals within this cohort are exhibiting clinical features of early neurodegeneration.

Olfactory abnormalities are found in those with PD with mutations in GBA,¹⁴ but they are not a reported feature of GD or its treatment. It has been proposed that the earliest α-synuclein changes occur in the dorsal motor nucleus of the vagus and olfactory bulb,⁸ and evidence suggests that an impaired sense of smell is not simply a consequence of aging but rather is a prodromal phenomenon that may predict PD.¹⁵ In the cohort studied, both patients with type 1 GD and heterozygous mutation carriers were, as a group, hyposmic at baseline.¹¹ At 2 years, follow-up olfactory scores for patients with GD and for heterozygous carriers remained significantly lower than those reported for controls, but they remained unchanged from baseline. This could reflect the short length of the follow-up period, considering that olfactory impairment may progress slowly.

An impaired sense of smell does appear to correlate with other modalities in the prodromal phase of PD (eg, RBD).¹⁶ Owing to its high specificity and long latency to clinical disease, RBD is one of the strongest clinical predictors of neurodegenerative disease and a potential prodromal marker for preventative therapy.¹⁷ The RBDQ carries a high sensitivity, and for those individuals without existing neurological or sleep disorders, it carries a high specificity and, therefore, represents a good tool to detect individuals with RBD.¹⁸ We did identify a significantly increased frequency of symptoms of RBD at the follow-up assessment in GBA mutation–positive individuals compared with controls. It is arguable whether a score of 5 or 6 should be the cutoff point for a scale that is structured to determine whether there is RBD or not. It should be noted, however, that the proportion of RBDQ scores greater than 5 was higher among GBA mutation–positive individuals than among controls at follow-up, albeit this difference did not reach significance (P = .39, determined by use of the χ² test).

Depression can precede the onset of the motor symptoms of PD and is a presenting complaint in 12% to 22% of patients.¹⁹ There was an increase in the number of reports of depressive symptoms in GBA mutation–positive individuals at follow-up. Patients with GD can exhibit moderate to severe psychological complications, similar to patients with other chronic illnesses.²⁰ In addition, BDI scores of 1 to 10 are consistent with minimal depression, and the specificity of depression alone as a clinical marker of prodromal PD is low but may be usefully combined with other features.²¹

Mild cognitive impairment can occur as a prodrome to parkinsonism²² or dementia with Lewy bodies.²³ There are several lines of evidence now for greater cognitive impairment in those with established GBA-related PD vs sporadic PD,^24-26 and this may reflect a higher burden of Lewy body disease in GBA-related parkinsonism.^27,28 Interestingly, in a subgroup of 6 GBA mutation–positive individuals with parkinsonian motor signs, mild cognitive impairment (MoCA score of ≤24 for 5 of 6 individuals [83.3%]) was the main premotor sign present at baseline that could have predicted their motor deterioration. Compared with controls, the remaining GBA mutation–positive individuals demonstrated significantly lower MMSE and MoCA scores at follow-up, albeit these were unchanged from baseline and still within the normal range for cognitive function.

Controls showed a small but significant change in the UPDRS part II score from baseline. Particular aspects of the UPDRS part II that had worsened for controls included a score of 2.11 for getting out of a bed, a car, or a deep chair; a score of 2.12 for walking (eg, use of a walking aid); and a score of 2.5 for dressing (eg, help with buttons). Subjective complaints of stiffness, tremors, and imbalance are associated with an increased risk for the development of PD.²⁹ However, we note that the UPDRS part II was not designed or validated as a tool for activities of daily living for aging controls. We believe that what drove the changes in the controls was a small group of individuals (n = 6) who were older (mean age, 70.5 years [range, 62.6-77.8 years]). A significantly higher follow-up UPDRS part II score for patients with GD distinguished these individuals from age-matched controls.

There were some GBA mutation–positive individuals (10.0%) with significant motor findings identified using the UPDRS part III that did not overlap with normal physiology (eg, bilateral postural tremor) or existing bone/joint abnormalities. These individuals did not meet the diagnostic criteria for PD but could represent a subgroup of GBA mutation–positive individuals who are progressing toward clinical PD.

We considered the effect of concurrent medications. The majority of patients with type 1 GD (83.0%) were receiving enzyme replacement therapy. This type of therapy does not cross the blood-brain barrier and has no reported neurological adverse effects. Furthermore, enzyme replacement therapy has no known effect on dysautonomia. Substrate reduction therapy can induce memory problems.³⁰ However, only 2 patients with GD were receiving substrate reduction therapy when evaluated at baseline and at follow-up, and neither had cognitive impairment.

To our knowledge, this study is the first to undertake the longitudinal follow-up of a large cohort of GBA mutation–positive individuals prior to the development of PD. Much of the work published thus far in the literature has focused on patients with established PD. This has been essential to make important comparisons between sporadic PD and GBA-related parkinsonism and to observe subtle differences. The opportunity to observe patients prospectively within a unique at-risk cohort such as this is essential for defining the optimal time to intervene with neuroprotective therapy.

One limitation of our study was that not all investigators were blinded to the genetic status of individuals. To minimize any observer bias, standardized scores were used, and all follow-up data were reexamined. Other potential criticisms are the use of prodromal markers and their sensitivity, specificity, and positive and negative predictive values. The presence of clinical markers alone may be insufficient to accurately predict a neurodegenerative disorder in the majority of cases. However, clinical markers may be used in combination with other biochemical or imaging markers for prodromal PD to develop a more reliable method for predicting PD.

The data from this cohort suggest that hyposmia is the earliest and most sensitive prodromal marker. Cognitive impairment is also an early feature, and this may relate to the increased cognitive impairment observed in GBA-related PD. Symptoms of RBD, the most specific clinical marker, are now present in GBA mutation–positive individuals. Depressive symptoms have also surfaced but must be interpreted with some caution, considering their low specificity as a marker for PD. There has also, and perhaps most importantly, been a significant decline on the UPDRS, which, together with impaired RBD and depression, suggests that clinical markers in some individuals of this GBA mutation–positive cohort have evolved, in a pattern consistent with the clinical prodrome of PD.

Accepted for Publication: August 22, 2014.

Corresponding Author: Anthony H. V. Schapira_, MD, DSc, FRCP, FMedSci, Department of Clinical Neurosciences, Institute of Neurology, University College London, Rowland Hill St, London NW3 2PF, England (a.schapira@ucl.ac.uk).

Published Online: December 15, 2014. doi:10.1001/jamaneurol.2014.2950.

Author Contributions: Dr Beavan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Beavan, McNeill, Proukakis, Schapira.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Beavan, McNeill.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Beavan.

Obtained funding: McNeill, Mehta, Schapira.

Administrative, technical, or material support: Beavan, McNeill, Hughes, Mehta.

Study supervision: Proukakis, Hughes, Mehta, Schapira.

Conflict of Interest Disclosures: Dr Schapira was a consultant to and received honoraria from Zambon and Lundbeck, and he receives royalties from Elsevier and Oxford University Press. No other disclosures are reported.

Funding/Support: The research study was funded by the Wellcome Trust/MRC Joint Call in Neurodegeneration Award (WT089698) and was supported by the National Institute for Health Research, University College London Hospitals, Biomedical Research Centre.

Role of the Funder/Sponsor: The funding agency had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Information: Dr Schapira is a National Institute for Health Research senior investigator.