Abstract
We compared phenotypic characteristics in 35 Greek patients with Parkinson’s disease (PD), carriers of GBA1 mutations (GBA-PD), with 35 Genetically Unidentified PD patients (GU-PD). We found a previously reported higher prevalence of cognitive impairment and a little appreciated more frequent bilateral onset of the disease in GBA-PD vs GU-PD. As far as the exposure to environmental factors, linked to PD, is concerned, our study hints to the possibility that pesticide exposure may be more common in GBA-PD patients, and possibly act synergistically with the mutation carrier status to trigger the disease.
INTRODUCTION
The most common genetic risk factor for Parkinson’s disease (PD) are heterozygous mutations in the gene GBA1 encoding for the lysosomal enzyme glucocerebrosidase, which in the homozygous state result in Gaucher’s disease (GD) [1]. Studies in the Greek population have confirmed this association [2, 3]. Additionally, GBA-associated PD patients have been reported to manifest with a more severe phenotype, in particular earlier age at onset and higher prevalence of cognitive impairment, compared to non-mutation carriers [1]. Herein we compare clinical characteristics of GBA-related PD (GBA-PD) with genetically unidentified PD (GU-PD) patients (i.e., with no known PD genetic determinant) in the Greek population to evaluate whether there are distinct motor and non-motor symptom characteristics in GBA-PD.
MATERIALS AND METHODS
We have been longitudinally collecting DNA samples from PD patients and controls from the outpatient Movement Disorders clinic of Attikon University Hospital of Athens and the General Hospital of Syros. We have screened a large cohort of 114 patients from Syros and 423 patients from Attikon Hospital for the presence of specific GBA1 mutations, known to be associated with GD in the Greek population, and for the p.A53T SNCA mutation [2, 3], focusing on PD patients with a high probability of genetic contribution, based on the presence of family history of PD, and/or early age of onset (≤50 years). Thirty-five PD patients were found to carry in heterozygosity GBA1 mutations, previously reported to be pathogenic (12 L444P, 9 N370S, 11 D409H:H255Q, 1 R120W, 2 IVS10-1G⟶A). All patients were unrelated except for two, who had a father-daughter relationship. One additional patient was compound heterozygous (N370S/IVS6-2A>G), already diagnosed with GD, and excluded from the study. Within GBA-PD, we also categorized patients in 3 subgroups with the most frequent GBA1 mutations (11 D409H, 12 L444P, 9 N370S).
We have compared the 35 GBA-PD patients to 35 sex-, age at the time of the study-, disease duration-, and education-matched PD patients negative for the known GBA1 and SNCA mutations, applying stratified sampling. We selected to perform matching instead of post-hoc adjustment of our two groups for the variables mentioned above, which are known confounders in disease manifestations, because the two groups were very heterogeneous between them, especially regarding age at the time of study and disease onset. We have termed this group GU-PD, to delineate the fact that these most common in our Greek population mutations have been excluded, but that other genetic factors have not been ruled out. In our population, other genetic contributions known to be associated with PD are extremely rare [3]. Severe GBA1 mutations (e.g., L444P) are reportedly associated with an increased PD risk, earlier age at onset and greater cognitive dysfunction compared to mild mutations (e.g., N370S) [4, 5], and therefore we have also compared the phenotypic characteristics between the 3 GBA-PD subgroups. We have also comparatively assessed exposure to environmental factors linked to PD in the two groups, in an effort to begin to understand how other factors in combination with the low penetrance GBA1 mutations may lead to the disease.
A pre-defined questionnaire on demographic data (including smoking, coffee drinking and exposure to pesticides), family history, bilateral or unilateral affected side at disease onset, and non-motor symptoms (dementia, autonomic dysfunction, psychosis and depression) was applied. The presence or absence of such symptoms was based on the clinical impression of the investigator. Autonomic dysfunction was defined by the presence of dysautonomia in at least one the following systems; urogenital, gastrointestinal, cardiovascular, thermoregulatory and pupillomotor. In terms of clinical scales, the modified Hoehn & Yahr (H&Y) scale, the Schwab and England Activities of Daily Living scale, the Unified Parkinson’s Disease Rating scale part III (UPDRSIII) and the standardized Mini-Mental State Examination (sMMSE) were performed. All scales and questionnaires were applied at the time of entry of the subjects in the study. Blood drawing, DNA isolation and examination for specific GBA1 mutations were performed as previously described [2, 3]. The study was approved by the Ethical Committee of Attikon University Hospital. The Mann Whitney U test was performed for the comparison between GBA and GU-PD and the Kruskal Wallis H-test for the GBA subgroups. Dichotomous data were analyzed using the Fisher’s exact test. For possible correlations-associations between categorical data we used the Phi correlation coefficient and between categorical and numerical data the Mann Whitney U test. We additionally present our results after applying Bonferroni correction for multiple comparisons.
RESULTS
Results are presented in Tables 1 and 2. GBA-PD patients reported more often family history of the disease compared to GU-PD, however this difference was marginally non-significant, consistent with reduced penetrance of GBA1 mutations. Motor indices, neuropsychiatric features and autonomic function were not different between the two groups. Cognitive impairment was more severe in GBA-PD according to the questionnaire for the presence of dementia (p = 0.001), and the sMMSE score, in which GBA-PD had overall lower scores (p = 0.008), while more GBA-PD compared to GU-PD scored below the cut-off of 25 (8/30 vs 2/34). These results confirm, in a Greek population, the previously reported higher prevalence of cognitive impairment in GBA-PD [1, 4]. This is important because it confirms this finding in another ethnic group with a different panel of GBA1 mutations. We have also found that GBA-PD patients present more frequently (p = 0.027) with a symmetric onset compared to GU-PD; this has only been reported in one previous study [1], and is not widely acknowledged as a feature of GBA-PD. This is potentially important, as symmetric onset has been linked to a more aggressive course of PD [6]. The question arose whether these two features of GBA-PD were correlated, as worse cognitive function and symmetric onset may both indicate a more aggressive disease. However, there was no association between the presence of dementia or the sMMSE score and the symmetric onset of the disease in GBA-PD (Phi = –0.083, p = 0.643 and p = 0.946 respectively). This may be due to the small sample, but may also indicate that these features are independent. Additionally, when we compared all demented patients to non-demented, regardless of the presence of GBA mutation, there was no statistically significant difference as far as symmetric onset is concerned (p = 0.158). As for GBA1 mutation subgroup comparisons, no differences were found, possibly because of the small sample sizes (data not shown).
Demographic characteristics in GBA-PD vs. GU-PD
aMann Whitney U test; bFisher’s Exact test.
Motor and non-motor characteristics in GBA-PD vs. GU-PD
aMann Whitney U test; bFisher’s Exact test; sMMSE, standardized Mini-Mental State Examination.
There is ample evidence for protective effects of smoking and coffee in PD, while pesticide exposure increases PD risk [7–10]; the biologic mechanisms for these possibly causal relations are poorly understood. In preliminary results from our Greek database, comparing a large number of GU-PD to controls, we have confirmed a strong negative association for smoking in GU-PD (Angelopoulou et al., in preparation), a finding also present in two other studies in the Greek population [11, 12]. In our current cohort, as far as smoking or drinking coffee are concerned, there was no difference between GBA-PD and GU-PD in any of the qualitative or quantitative variables assessed. However GBA-PD patients reported more frequently exposure to pesticides compared to GU-PD (p = 0.021). We could interpret these data as showing that pesticides may trigger the expression of PD in GBA1 mutation carriers, more so than in GU-PD; however, 3/9 GBA-PD patients with pesticide exposure were recruited from the Center in Syros, a rural setting, without there being a corresponding recruitment from this Center in the GU-PD cohort.
Applying Bonferroni correction for the 17 tests performed (padjusted = 0.0029), only the higher presence of dementia in GBA-PD compared to GU-PD remained statistically significant. However, as our genetic cohort is relatively small, the application of such a correction would be very conservative. As far as bilateral onset is concerned, this is a phenotypic characteristic that has been reported in a previous study [1], suggesting that it represents a true phenomenon. Furthermore, for some of the tests (e.g., years of smoking and pack years of smoking) there is lack of independence, and thus the Bonferroni method may lead to even further loss of power to detect real differences between the two groups.
DISCUSSION
In conclusion, we confirm distinctive features of GBA-PD that suggest a more severe disease compared to GU-PD, especially regarding the cognitive domain. Importantly, and unlike other studies, GBA-PD were matched to GU-PD for multiple parameters, including age, disease duration and education, rendering further validity to the findings. The fact that the two groups were by design matched for age and disease duration precluded the possibility of assessing in this study whether GBA-PD is associated with an earlier age of onset; in our previous study in sporadic PD however, early age of onset was associated with a much higher chance of carrying a GBA1 mutation [2]. The findings of preferential bilateral onset and increased pesticide exposure in GBA-PD compared to GU-PD should be interpreted with caution, and need replication in further independent studies. Overall, studies of this genetic subtype of PD may lead to insights into disease pathogenesis, as well as aid in patient stratification, prognosis, and eventually monitoring of specific therapeutic interventions.
CONFLICTS OF INTEREST
The authors have no conflict of interest to report.
Footnotes
ACKNOWLEDGMENTS
This work was supported in part by grants THALIS (Hellenic Ministry of Education and Religion THALIS grant number 377206) and SYNERGASIA (GSRT Grant 09-RYN-12876) grants (to L.S.).