Clinical course and biomarkers in patients with type 1 Gaucher disease and Parkinson's disease: A retrospective cohort study

Abstract

Introduction

Pathogenic variants in GBA1, the gene encoding beta-glucocerebrosidase, are significant genetic risk factors for Parkinson's disease (PD). Patients with type 1 Gaucher disease (GD1), caused by biallelic GBA1 variants, have approximately six-fold higher lifetime risk for PD compared to the general population. However, few studies have described the clinical presentation of GD-associated PD (GD-PD) in GD1 patients. We describe the variable presentation of PD and evaluate clinical characteristics and biomarkers in a cohort of patients with concurrent diagnoses of GD-PD.

Methods

Retrospective case-control analysis of GD1 patients within a single academic clinic from 2010 to 2020 comparing against age- and sex-matched controls.

Results

Among eleven patients diagnosed with GD-PD, the median age of GD1 diagnosis was 33 years (interquartile range/IQR 28–54 years), and all were on treatment with enzyme replacement therapy (n = 8, 72.7%) or substrate reduction therapy (n = 3, 27.3%) at follow-up. The average age of onset of PD symptoms was 54.6 years old with a median age of onset of 53 years old (IQR 47–59) and four patients developing symptoms before the age of 50. The majority (n = 9, 81.8%) were treated with levodopa/carbidopa with or without deep brain stimulation. Within this limited cohort, neither clinical phenotype nor systemic GD biomarkers distinguished GD-PD cases from controls.

Conclusions

Data from this single-center cohort demonstrate considerable variability in GD-PD among GD1 individuals and suggest that GD1 patients may have increased risk of GD-PD before the age of 50. Routine screening for PD in GD1 patients starting at age 40 may aid early diagnosis and close monitoring for initiation of PD-directed treatment.

Highlights

• •Gaucher disease patients can develop Parkinsonism before age 50.
• •In these patients, disease biomarkers do not correlate with presence of Parkinsonism.
• •Gaucher disease symptom burden does not correlate with the presence of Parkinsonism.

1. Introduction

Gaucher disease (GD) is the most common inherited disorder in the Ashkenazi Jewish population. GD has an estimated carrier frequency of 1:18 and a disease prevalence of 1:855 [1]. GD is caused by biallelic pathogenic variants in the GBA1 gene, encoding the protein beta-glucocerebrosidase; two disease-causing GBA1 variants cause deficient activity of beta-glucocerebrosidase and lysosomal accumulation of glucosylceramide (GL1) with its deacylated metabolite glucosylsphingosine (lyso-GL1) [2]. This multi-systemic condition presents on a spectrum with type 1 GD (GD1) being the most common presentation. GD1 is highly variable in clinical presentation and includes cytopenia, organomegaly and bone disease (osteopenia, lytic lesions, and osteonecrosis). GD1 patients are at an increased risk for hematologic malignancies (most commonly multiple myeloma) [3] and Parkinson's disease (PD) [1]. Two treatments modalities exist for GD – providing exogeneous enzyme as an infusion for enzyme replacement therapy (ERT) and decreasing production of GL1 by means of an oral inhibitor of glucosylceramide synthesis for substrate reduction therapy (SRT). Both modalities improve symptoms, halt disease progression, and prevent complications [[4], [5]].

Despite the rarity of GD, pathogenic variants in GBA1 are the most common genetic risk factor for PD [[1], [6], [7]]. Type 1 GD patients have a 12% lifetime risk of PD compared to 1.5% in the general population [1], often presenting with akinetic subtypes including asymmetric tremor, bradykinesia, postural instability, and gait abnormalities as well as variable cognitive impairment [8]. Variants in GBA1 have also been associated with Lewy body dementia; between 3% and 30% of individuals with DLB harbor at least one pathogenic variant [[9], [10]] with higher estimates correlating with increased ascertainment of Ashkenazi Jewish individuals. The mechanism for this increased risk remains unclear, with some studies suggesting that that glucocerebrosidase deficiency leads to accumulation of alpha-synuclein, the protein that pathologically aggregates in neurons in PD patients [6]. Unfortunately, ERT and SRT have limited roles in GD-associated PD (GD-PD). Current FDA approved SRT and ERT do not penetrate the blood-brain barrier and have limited central nervous system (CNS) activity [[11], [12]]. Additionally, at least one trial using the CNS penetrant SRT venglustat in GBA1-associated PD (GBA1-PD) has not shown a clinical benefit [13].

Despite the observed association between GD1 and PD, few factors identify GD1 individuals at risk of developing PD. GD1 patients develop PD younger on average (mean = 54 years) with more rapid motor progression and cognitive changes compared to idiopathic PD patients [[1], [6], [14], [15]]. However, disease severity and treatment status generally do not correlate with risk [16]. The GBA1 genotype - PD phenotype association is also complicated. Previous work has identified an increased risk for PD among patients with more severe GBA1 variants [[17], [18], [19], [20], [21]]; though this association has been challenged by other work [16] and complicated by the observation that other non-disease causing GBA1 variants can still be associated with increased PD risk [[22], [23], [24], [25]]. Here, we describe an exploratory case-control study of a cohort of GD1 patients with GD-PD alongside age- and sex-matched GD1 controls without apparent PD to compare GD disease activity (e.g. via blood biomarkers and imaging) and PD manifestations.

2. Materials & methods

We conducted an exploratory retrospective chart review utilizing electronic health records (EHRs) of patients with known diagnoses of GD1 that were followed at a single site (Mount Sinai Hospital, New York, New York) and diagnosed with GD-PD between 2010 and 2020 by one or more movement disorders specialists (including RNA). Such individuals satisfied criteria for PD and were evaluated for atypical and secondary PD. This study was approved by the Institutional Review Board at the Icahn School of Medicine at Mount Sinai Hospital.

Clinical information was obtained regarding the patient's GD1 and GD-PD manifestations. The patient's GBA1 genotype, GD treatment, baseline and most recent biomarker levels (ferritin, chitotriosidase, lyso-GL1) were reviewed. In this cohort of individuals with predominantly Ashkenazi Jewish ancestry, GBA1 genotype was primarily ascertained through GBA1 targeted mutation testing with phase inferred in conjunction with leukocyte glucocerebrosidase enzyme activity testing. One case (Case 3) was found to have GD through targeted sequencing for GBA1 p.N409S and LRRK2 p.G2019S variants. Other than this case, individuals were not tested for other monogenic causes of PD (e.g. LRRK2, PRKN, SYNA, PINK1). Chitotriosidase levels were not corrected for chitotriosidase genotype, including the common duplication variant or single nucleotide variants known to affect absolute levels [26]. Lyso-GL1 was measured using tandem mass spectrometry by a single lab (Sema4 Laboratories). Baseline and most recent (”follow-up”) results were obtained for abdominal MRI to evaluate liver and spleen volumes, MRI of the femurs and lumbar spine to evaluate bone marrow infiltration/disease burden, and dual x-ray absorptiometry/DEXA scans to assess for bone fragility.

For GD-PD, all patients at this site were screened yearly using a standardized review of symptoms related to PD (including bradykinesia, gait difficulties, hyposmia/anosmia, REM sleep behavior disorder, constipation) as well as an exam assessing rigidity, tremor, movement speed, and gait. Before 2018, all individuals at the site above the age of 50 (including controls) were screened for PD; starting 2018, all individuals above the age of 40 were screened. Data collected for individuals found to have PD included presenting symptoms, treatment, motor and non-motor symptoms, and family history of PD was obtained, as well as records of standard neurological scoring for stage of PD (Hoehn and Yahr scale), and longitudinal course of PD (Unified Parkinson's disease Rating Scale, part III; UPDRS-III). Age- and sex-matched patients with GD1 but without GD-PD or other neurological abnormalities were selected as controls by manual chart review; most cases (n = 10) were matched 1:2 with controls and a single patient was matched 1:1 with a control due to limited availability of matching control participants in the sample.

Exploratory comparisons between cases and controls were performed using non-parametric Mann-Whitney rank-sum testing for two-group continuous variables, and Fisher exact testing for associations of binary/categorical variables as appropriate. Exploratory hypothesis testing was performed at a significance level of α = 0.05 with Bonferroni correction of multiple hypothesis testing as needed; reported p-values are unadjusted. Analyses were conducted in Python using the Scipy statistics module (v.1.16.2).

3. Results

During the ten-year study period, this site ascertained 11 GD1 cases with PD alongside 21 age- & sex-matched GD1 individuals without PD (Table 1; Supplementary Table 1). Most control individuals were seen annually in-person for most of the ascertainment period with a median of 9 years (IQR 9–10 years; Table 1) with additional in-person visits limited in the tenth year due to COVID−19 precautions. Of the eleven patients diagnosed with GD-PD, most were male (n = 8, 73%), Ashkenazi Jewish (n = 9, 82%), and had a p.Asn409Ser [N409S; formerly known as p.Asn370Ser/N370S] homozygous GBA1 genotype (n = 6, 55%). The median age of cases was 62.0 years (interquartile range 55–70) with a median age of GD onset of 33.0 years (interquartile range 27.5–54). Approximately half (6/11; 54%) of patients were on GD-directed therapy (with ERT) at time of PD symptom onset; the remaining individuals had not been treated for GD. At time of record review, most individuals were on GD-directed treatment, either with ERT (n = 6, 45%) or SRT with eliglustat (n = 3, 27%). The median duration of treatment prior to onset of PD symptoms among those on treatment was 8 years (interquartile range 6–18). Most cases and control individuals were either homozygous or compound heterozygous with the common N409S allele (10/11 and 20/21, respectively); there was no meaningful difference in the proportion of individuals who were homozygous for N409S between the two groups (6/11 vs 15/21, P = 0.44). The most common GD manifestations at diagnosis were cytopenias, followed by splenomegaly, hepatomegaly, and osteoporosis. One patient had a splenectomy at age 10. The control group had similar characteristics, including average age of GD diagnosis and proportion of individuals on treatment at the time of record review.

Table 1.

Clinical and Genetic Characteristics of Study and Control Cohorts^⁎

	Patients with GD and PD	Controls	P value
Age, median (IQR)	62 (55–70)	63 (53–72)	0.95
Sex Assigned at Birth			1.0
Male	8	15
Female	3	6
Self-Identified Ancestry#			1.0
Ashkenazi Jewish	9	18
Italian	0	3
European	1	0
Hispanic	1	0
GBA Genotype†			0.44
p.N409S/p.N409S	6	15
p.N409S/p.L483P	2	1
p.N409S/p.R535H	1	1
Other††	2	4
GBA Genotype Severity			0.84
Mild	7	16
Severe	3	4
Undefined	1	1
Age at GD Diagnosis, median (IQR)	33 (28–54)	38 (30–48) (n = 20)	1.0
Age at PD Symptom Onset, median (IQR)	53 (47–59) (n = 9)	–	n/a
Family History of PD			0.28
Yes	5	4
No	5	15
Unknown	1	2
GD treatment			0.87
ERT	6	13
SRT	3	4
None	2	4
Difference in Case and Control Age (yr), median (IQR)		-1 (−5–2)	n/a
Duration of Control Follow-Up (yr), median (IQR)	–	9 (9–10)	n/a
Total Time with PD Symptoms at Last Follow-Up (yr), median (IQR)	5 (5–9.5)	–	n/a
GD treatment at PD Symptom Onset			n/a
ERT	6	–
SRT	0	–
None	5	–
PD treatment			n/a
Levodopa	8	–
DBS	3	–
None	2	–

^⁎Number of case and controls n = 11 and 21, respectively, unless otherwise noted.

^#p-value calculated for proportion of Ashkenazi Jewish versus non-Ashkenazi Jewish individuals.

^†The noted genotypes reflect updated nomenclature, including: p.N409S (formerly p.N370S), p. L483P (formerly p.L444P), p.535H (formerly p.R496H). P-value reflects proportion of p.N409S/p.N409S individuals compared to proportion of non-p.N409S homozygous genotypes.

^††Additional genotypes include p.N409S/R535H, R535H/c.84GG, and p.N409S/p.I200N (n = 1 case with each genotype); as well as p.L483P/p.P237R, p.N409S/IVS2 + 1, p.N409S/c.84GG, and p.N409S/p.G416S (n = 1 control with each genotype).

All individuals had follow-up imaging and blood-based biomarker measurements at least one year post-GD1 diagnosis (Supplementary Table 1). Timing relative to GD treatment onset was variable due to the inherent variability in GD1 activity and clinical findings that would prompt initiation (Supplementary Table 1). Most measurements in cases were at least one year after onset of PD symptoms (Supplementary Table 1).

3.1. Imaging

Imaging findings for case and control groups are summarized in Table 2. Median baseline spleen volume for cases was 338 cc (interquartile range/IQR 299–625) and baseline liver volume was 2006.5 cc (IQR 1697.3–2542.3). Median control spleen volumes were ∼2 fold higher at baseline compared to cases but with an overlapping range (median 763 cc with IQR 436–1078; p = 0.21). Liver volumes were comparable (median 2037 cc, p = 0.93) to cases. Most GD-PD patients (n = 9, 81.8%) had evidence of mild or diffuse infiltration with or without fibrosis in their femora; one patient did not have a baseline femur MRI available. Ten individuals had bone density testing available at baseline. Five out of 10 individuals (45.5%) had osteopenia at the hip (T-score range − 1.9 to −1.5), and seven out of 10 (63.6%) had osteopenia at the spine (T-score range −2.2 to −1.1). No cases had osteoporosis of the hip or spine at baseline.

Table 2.

GD Clinical Characteristics in Patients with and without PD^⁎.

	Patients with GD and PD				Controls				P value
	n	Initial Median (IQR)	n	Follow-up Median (IQR)	n	Initial Median (IQR)	n	Most Recent Baseline Median (IQR)
Organ size (cc)
Spleen	9	338.0 (299.0–625.0)	6	287.0 (224.5–465.8)	17	763.0 (436.0–1078)	16	302.0 (254.8–409.8)	Baseline: 0.21 Follow up: 0.94
Liver	10	2007 (1697–2542)	7	1568 (1475–2054)	20	2037 (1814–2232)	19	1725 (1514–1948)	Baseline: 0.93 Follow up: 0.93
DEXA (T-score)
Hip	9	−1.0 (−1.50--0.70)	7	−1.3 (−1.5 to −0.90)	19	-0.90 (−1.4 - -0.15)	18	−0.75 (−1.4–0.60)	Baseline: 0.26 Follow up: 0.28
Spine	9	−1.1 (−1.4–0.80)	8	−1.0 (−2.0 - -0.27)	20	−1.05 (−1.6–0.55)	19	−0.4 (−1.5–0.0)	Baseline: 0.96 Follow up: 0.31
Ferritin (ng/ml; ref 22–275)	10	822 (343–1010)	10	237 (166–664)	21	344 (166–685)	21	147 (61–254)	Baseline: 0.07 Follow up: 0.07
Chitotriosidase (nmol/h/ml; ref 〈120)	11	230 (169–3010)	10	72 (43–262)	21	236 (118–1600)	21	78 (44–262)	Baseline: 0.39 Follow up: 0.91
Lyso-GL1 (ng/mL; ref ≤ 1)	6	4.4 (1.8–8.3)	7	2.4 (1.1–3.3)	17	4.5 (3.0–6.5)	13	4.5 (3.1–6.3)	Baseline: 0.78 Follow up: 0.05

⁎Biomarkers listed as median (IQR).

After treatment, organomegaly, bone marrow infiltration, and bone density overall improved or stabilized. Follow up imaging in cases revealed decreased spleen (median 287 cc, IQR 224.5–465.8) and liver volumes (median 1568 cc, IQR 1475–2054) that was comparable to controls (median 302 cc/p = 0.94 and 1725 cc/p = 0.93, respectively; Table 2). Of the seven cases with available follow up MRI studies of the femora, six did not have worsening bone marrow infiltration or evidence of fibrosis. One case developed diffuse infiltration with minimal activity and a medullary infarction of the bilateral proximal tibia while on chronic ERT and prior to onset of PD symptoms. Five cases had osteopenia or osteoporosis of the hip or spine including one case with osteoporosis at both sites; the prevalence of osteopenia and osteoporosis by follow-up was similar between cases and controls (5/11 versus 11/27, respectively; p = 1.0; Table 2).

3.2. Blood-based activity biomarkers

Compared to control patients at baseline, GD-PD cases had higher median but overall clinically comparable ferritin levels (cases: 822 ng/ml with IQR 343–1010; controls: 344 ng/ml with IQR 166–685; p = 0.05; Table 2). Other biomarkers at baseline were not clinically nor statistically different between cases and controls (Table 2). Ferritin, chitotriosidase, and lyso-GL1 levels decreased on average at follow-up within the six to twelve months after initiation of treatment. At follow-up, ferritin, chitotriosidase, and lyso-GL1 were clinically and statistically similar between cases and controls (Table 2).

3.3. PD characteristics

The median age of onset of PD symptoms was 53.0 years old (IQR 47–59; Table 1) and the average age of onset of PD symptoms was 54.6 years old. The earliest age of onset was 41 years old with four out of 11 patients developing PD symptoms before the age of 50 years old (including two p.N409S homozygous individuals; Table 3). Multiple patients were initially suspected of having GD-PD by their GD provider (n = 3, 27%), who prompted a referral to neurology; and conversely, two patients were diagnosed with GD through PD screening programs. Five of the 11 patients (45%) had a first or second degree relative with PD. GD genotype and treatment status was not clearly different between cases and controls (Table 1). Four patients were deceased with a median time of 5.5 years between onset of PD symptoms and death; cause of death was largely unrelated to PD symptomatology (Table 1, Table 3).

Table 3.

PD Clinical Manifestations in GD Cohort.

Case #/Sex	GBA1 Genotype	PD Symptom Onset (yr)	Presenting Symptoms	Follow-up After PD Onset (yr)	Duration of PD Prior to Death (yr)	How Diagnosis was Made	Motor Symptoms	Levodopa Response	Deep Brain Stimulator Placement	UPDRS-III Score	H&Y Stage	Outcome of Motor Disease	Anosmia	Hallucinations/Psychosis	Autonomic Dysfunction	Cognition	MOCA
1/M	p.N409S/p.N409S	unknown (prior to age 57)	Heaviness in legs, stiffness, slowness, resting tremor	13	–	unknown	Dyskinesia and bradykinesia, micrographia, soft and slurred speech. Increased tone in neck, right upper extremity, and left lower extremity; and, positive glabellar release sign. Reduced rapid alternating movements in lower extremities (left greater than right). Stooped posture, unsteady gait with decreased right arm swing	Limited duration of ON-time (2 h) on IR levodopa/carbidopa. Continued freezing of gait. Peak-dose dyskinesias improved with amantadine.	Yes - Still using Sinemet per prior schedule and dosing with significant improvement in gait and tremor.	–	–	Stablized	No	Hallucinations	Orthostatic hypotension	Mild cognitive impairment (after onset of PD symptoms)	7/30
2/M	p.N409S/p.L483P	60.5	Resting and intention tremor	5	5	referred to neurology by LSD clinic	Symmetric slowness and stiffness. Occasional myoclonic jerks of the outstretched hands. Decreased arm swing.	Limited duration of ON-time (1 h) on IR levodopa/carbidopa. Peak-dose dyskinesias.	No	22 → 20	–	Died before management could be optimised (cause of death: pneumonia)	No	Hallucinations (visual and olfactory)	Orthostatic hypotension	Dementia (suspected dementia w/ Lewy bodies)	28/30
3/F	p.N409S/p.N409S	59	Occasional hand tremor, softening of voice, difficulties with activities of daily living	5	–	separate diagnosis by neurologist	Bradykinesia, monotone and slurred speech, hypomimia. Neck and extremity rigidity, severe impairment of bilateral hand pronation.	Limited duration of ON-time IR levodopa/carbidopa. Significant but subjectively tolerable symptom burden in OFF-state (residual tremor, rare fall).	No	33 → 17	II	Stabilized	Yes	No	No	No impairment	–
4/F	p.N409S/p.R535H	55	Right arm action tremor and right foot dragging	21	–	unknown	Dysarthric, soft speech with hypomimia.	Persistent bradykinesia in ON state on IR levodopa/carbidopa. Treatment limited by intolerable peak-dose dyskinesias.	No	25 → 25	–	Unstable.	No	Hallucinations	Orthostatic hypotension	Dementia (suspected dementia with Lewy bodies)	16/30
5/M	p.R535H/c.84GG	41	Right arm stiffness and right leg heaviness	5	–	separate diagnosis by neurologist (initial diagnosis stiff-man syndrome, revised after abnormal DAT scan and response to levodopa/carbidopa)	Excessive movement of R fingers without tremor or dyskinesia. Normal gait with decreased R arm swing	Limited duration of ON-time with minimally impaired gait and motor function. Tolerable peak-dose dyskinesias.	No	15	I	Stable	Yes	No	Constipation	No impairment	–
6/M	p.N409S/I200N	53	Intention tremor, decrease in fine motor control, severe fatigue	2	2	referred to neurology by LSD clinic	Dyskinesia. Limb rigidity. Resting tremor. Reduced arm swing.	Persistent bradykinesia and unsteady gait in ON-state.	No	–	–	Died before management could be optimised (cause of death: cardiac arrest)	No	No	Constipation	No impairment	–
7/M	p.N409S/p.N409S	unknown (prior to age 65)	Tremor	6	6	referred to neurology by LSD clinic	Limb rigidity. Intention tremor. Recurrent falls.	Did not initiate levodopa prior to death.	No	–	–	Died before management could be optimised (cause of death: renal cell carcinoma)	No	–	No	Cognitive impairment (attributed to frailty/multifactorial decline)	24/30
8/M	p.N409S/p.N409S	47	Right hand tremor, balance difficulty	13	13	separate diagnosis by neurologist	Choreiform dyskinesia. Soft speech and hypomimia. Decreased arm swing	Persistent bradykinesia, rigidity, and gait instability in ON state. Treatment limited by dyskinesias.	Yes - Residual rigidity and bradykinesia after DBS placement incompletely controlled on extended-release levodopa-carbidopa/Rytary and still limited by dyskinesias.	59 → 16	III → II	Stable. Died subsequently (cause of death: complications of BMT)	No	No	Constipation	No impairment	–
9/F	p.N409S/p.L483P	49	Balance and gait difficulty	5	–	separate diagnosis by neurologist	Rigidity. No tremor noted.	Limited duration of ON-time on IR levodopa/carbidopa QID. Treatment limited by intermittent dyskinesias.	No	–	–	Stable	No	No	Orthostatic hypotension, constipation	No impairment
10/M	p.N409S/p.N409S	83	Micrographia, shuffling gait, balance difficulty, hyposmia	2	–	separate diagnosis by neurologist	Soft speech, hypomimia, micrographia. Increased tone in neck and extremities, and positive glabellar release sign. Decreased rapid alternating movements without tremor. Decreased arm swing.	Patient deferred treatment with levodopa/carbidopa.	No	–	–	Stable	Yes	No	Constipation	Dementia (suspected dementia w/ Lewy bodies)	15/30
11/M	p.N409S/p.N409S	44	Decreased right hand dexterity	5	–	separate diagnosis by neurologist	Asymmetric bradykinesia with loss of right arm dexterity as well as facial movement abnormalities. Dystonic posturing of right arm. Right endpoint tremor. Decreased right arm swing.	Persistent unilateral arm rigidity and discomfort on IR levodopa/carbidopa; other symptoms 40% improved from pre-treatment.	No	–	–	Unstable.	Yes	No	Constipation	No impairment	–

The most common PD presentation was stiffness and postural gait instability (n = 6/11, 55%; Table 3). Nine out of 11 individuals were treated with levodopa/carbidopa therapy; one individual died of unrelated causes before treatment could be initiated (Case 7), and another deferred for personal reasons (Case 10). Eight of these nine individuals on treatment had limited ON-times with titration complicated by peak-dose dyskinesias. Two individuals underwent deep brain stimulator placement. Patients who had DBS placement improved but still had residual symptoms that required maintaining their prior levodopa treatment.

Five individuals had movement disorder assessment data (UPDRS-III with or without Hoehn & Yahr staging), and four of these individuals had at least two measurements (Table 3). Three individuals with multiple measurements and treated with levodopa only had stable or mild improvement with levodopa therapy (UPDRS-III difference range − 16 to 0 between initial evaluation and follow-up), and the one individual with DBS (case 8) improved from UPDRS-III score of 59 to 16 (with an improvement of Hoehn and Yahr staging from III to II).

Seven patients had non-motor symptoms including cognitive deficits, hallucinations, hyposmia/anosmia, and orthostatic hypotension (Table 3). One patient was diagnosed with Lewy body dementia (Case 2) with two others possibly with DLB after evaluation with a movement disorders specialist (Case 4 and Case 10).

4. Discussion

In this retrospective exploratory study, we reviewed the clinical characteristics and presentation of PD in eleven patients with concurrent GD1 and compared characteristics with age- and sex-matched GD1 controls. Of note, GD1 individuals with GD-PD had a considerably earlier age of onset compared to sporadic PD. In this limited cohort, neither demographic characteristics nor systemic GD biomarker levels were distinguishable between patients who developed versus did not develop GD-PD.

There are several limitations to this study. Small sample size limits quantitative comparison with controls, which warrants interpretation with caution. Study cases and controls were reviewed from a single academic clinic which increases the risk of ascertainment bias related to referral. Many patients were of Ashkenazi Jewish ancestry, which may limit generalizability to other populations. Select patients had limited information on neurological outcomes (including UPDRS-III). Additionally, given that GD1 controls are also at an increased lifetime risk of GD-PD, a subset of this group could have converted to PD in the future.

Of note, individuals were not uniformly tested for other monogenic risk factors/causes of Parkinsonism (e.g. variants in LRRK2, PINK1). Specifically, LRRK2 variants including p.G2019S have carrier frequency of up to 28% individuals among Ashkenazi Jewish individuals[27]. Individuals with a p.G2019S variant and a GBA1 variant also appear to have different trajectories of disease compared to individuals with one or more GBA1 variants alone[[28], [29]]. In this predominantly Ashkenazi Jewish sample with family history of PD, we cannot exclude a role of other variants that would contribute to PD burden or disease course. This is a recurrent confound in ascertained cohorts of GD-PD patients, and deserves broader evaluation in future work.

Within these limitations, the findings here provide evidence that PD symptoms develop earlier in individuals with GD-PD compared to individuals who develop sporadic PD. In our study cohort, the average age of onset of GD-PD symptoms was 54.6 years old; this is at least 3–8 years younger compared to the typical average age of onset seen for sporadic PD (estimated between 58 and 63 years of age; [[30], [31]]). Most common presenting symptoms included stiffness and gait imbalance. Approximately half of patients had non-motor symptoms including cognitive decline, hallucinations, decreased sense of smell, and orthostatic hypotension. This early age of onset and prevalence of non-motor symptoms are consistent with epidemiological studies of PD in patients with Gaucher disease [[8], [24]]. This range of symptoms and prevalence is comparable to that seen in individuals with sporadic PD [[32], [33]]. However, despite optimal GD1 management, GD1 patients are at risk for PD earlier than their non-GD counterparts.

Consistent with prior reports, clinical status and visceral disease activity did not clearly differentiate GD1 individuals with and without PD in our cohort. None of the disease biomarkers measured (ferritin, chitotriosidase, lyso-GL1) were clinically distinguishable between GD-PD cases and controls at baseline nor at follow-up. In contrast to other work [[20], [24], [34]], we found that the proportion of individuals with the homozygous p.N409S genotype and the number of individuals with more severe variants (e.g. c.84GG and p.L483P) were comparable to matched controls. Given limitations in study power, these findings cannot definitively compare clinical status, disease activity, and GD1 biomarkers with and without GD-PD. However, these findings underscore the need for further work to determine the role of these biomarkers.

There are two explanations for why clinical status or visceral disease biomarkers may have been similar between individuals in this study with and without GD-PD. From a therapeutics/biochemical perspective, ERT and most SRTs do not penetrate the blood-brain barrier and have limited ability to treat neurodegenerative disease in GD-PD and GBA1-PD [[11], [12]]. Hence, treating with these modalities may trivially uncouple disease in the CNS from systemic disease. Alternatively, GBA1 or sphingomyelin pathophysiology in the CNS is suspected to have another possible mechanism that includes toxic gain of function due to protein misfolding and proteotoxic stress [35] and autophagosomal/lysosomal trafficking defects [36].

Both scenarios necessitate further study using other types of biomarker assessment to elucidate the cause of increased risk for PD in GD1 patients. A number of biomarkers are being broadly explored to monitor neurodegenerative injury in PD including CSF neurofilament light chain [37]; this biomarker have not been adapted specifically for GD1 and GBA1-PD, but may prove useful in future work for understanding GBA1-PD. Additionally, alpha-synuclein seed aggregation assays (SAAs) have emerged as strong indicators of overall synucleinopathic burden in the peripheral and central nervous systems. SAAs use primary tissue/fluid specimens as a source for alpha-synuclein amyloid seeds to perform quantitative measurement of amyloid quantity. These assays use olfactory mucosa brushings or skin biopsy specimens to monitor PNS involvement, and CSF specimens to monitor CNS involvement. These assays have been implemented broadly for PD to establish diagnosis [[38], [39]] and to provide cohort stratification for development of PD symptoms and comorbidities [40]; and, emerging evidence has shown promise in using SAAs to quantify and stratify individuals with alpha-synuclein burden GBA1-PD [[40], [41]].

Alongside prior work, findings from this cohort suggest that early screening for PD among individuals with GD1 may identify individuals who would benefit from early management. Early detection of GD-PD is a management goal among recent consensus guidelines, though there are no specific recommendations on age at which to begin screening or screening tools [[42], [43]]. Individuals with GD-PD have variable responses to levodopa therapy; however, their responses tend to be less than individuals with sporadic PD [[8], [24]]. Often, individuals with GD-PD require deep brain stimulator therapy (DBS) placement approximately 4 years earlier in their disease course than sporadic PD [44]. Since patients with both GBA1-associated PD and GD-PD have had positive motor outcomes compared to sporadic PD [[45], [46], [47]], these patients may benefit from early trial of therapy for expedited consideration of DBS placement. Given the findings of this study, we recommend screening for motor and pre-motor signs starting at the age of 40 years old with a low threshold for referral to a movement disorders specialist.

CRediT authorship contribution statement

Ayuko Iverson: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Drew B. Sinha: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Luca Fierro: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Data curation. Chanan Stauffer: Writing – review & editing, Resources, Investigation. Hetanshi Naik: Writing – review & editing, Visualization, Formal analysis, Data curation. Roy N. Alcalay: Writing – review & editing, Conceptualization. Manisha Balwani: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Conceptualization.

Ethics declaration

The Icahn School of Medicine Institutional Review Boards (IRB) approved the study as described herein (IRB #19–00697). All clinical data were de-identified. The study adhered to the principles set out in the Declaration of Helsinki.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The following authors have nothing to declare.

Appendix A. Supplementary data

Supplementary material

Data availability

The authors are unable or have chosen not to specify which data has been used.