Association of Dual LRRK2 G2019S and GBA Variations With Parkinson Disease Progression

IMPORTANCE: Despite a hypothesis that harboring a leucine-rich repeat kinase 2(LRRK2) G2019S variation and a glucocerebrosidase (GBA) variant would have a combined deleterious association with disease pathogenesis, milder clinical phenotypes have been reported in dual LRRK2 and GBA variations Parkinson disease (PD) than in GBA variation PD alone. OBJECTIVE: To evaluate the association of LRRK2 G2019S and GBA variants with longitudinal cognitive and motor decline in PD. DESIGN, SETTING, AND PARTICIPANTS: This longitudinal cohort study of continuous measures in LRRK2 PD, GBA PD, LRRK2/GBA PD, and wild-type idiopathic PD used pooled annual visit data ranging from 2004 to 2019 from the Mount Sinai Beth Israel, Parkinson Disease Biomarker Program, Harvard Biomarkers Study, Ashkenazi Jewish-LRRK2-Consortium, Parkinson Progression Marker Initiative, and SPOT-PD studies. Patients who were screened for GBA and LRRK2 variations and completed either a motor or cognitive assessment were included. Data were analyzed from May to July 2020. MAIN OUTCOMES AND MEASURES: The associations of LRRK2 G2019S and GBA genotypes on the rate of decline in Montreal Cognitive Assessment (MoCA) and Movement Disorders Society-Unified Parkinson Disease Rating Scale–Part III scores were examined using linear mixed effects models with PD duration as the time scale. RESULTS: Among 1193 individuals with PD (mean [SD] age, 66.6 [9.9] years; 490 [41.2%] women), 128 (10.7%) had GBA PD, 155 (13.0%) had LRRK2 PD, 21 (1.8%) had LRRK2/GBA PD, and 889 (74.5%) had idiopathic PD. Patients with GBA PD had faster decline in MoCA than those with LRRK2/GBA PD (B [SE], −0.31 [0.09] points/y; P < .001), LRRK2 PD (B [SE], −0.33 [0.09] points/y; P < .001), or idiopathic PD (B [SE], −0.23 [0.08] points/y; P = .005). There was a LRRK2 G2019S × GBA interaction in MoCA decline (B [SE], 0.22 [0.11] points/y; P = .04), but not after excluding severe GBA variations (B [SE], 0.12 [0.11] points/y; P = .28). Patients with GBA PD had significantly worse motor progression compared with those with idiopathic PD (B [SE], 0.49 [0.22] points/y; P = .03) or LRRK2 PD (B [SE], 0.77 [0.26] points/y; P = .004). CONCLUSIONS AND RELEVANCE: These findings suggest that longitudinal cognitive decline in patients with GBA PD was more severe than in those with LRRK2/GBA PD, which more closely resembled LRRK2 PD. This further supports the notion of a dominant association of LRRK2 on GBA in individuals who carry both and raises the possibility of an LRRK2 × GBA interaction. However, the biological basis of a dominant association or interaction is not clear and is apparently contrary to basic investigations. Study of a larger cohort of individuals with severe GBA variation is warranted.