Validation of abnormal glucose metabolism associated with Parkinson’s disease in Chinese participants based on 18F-fluorodeoxyglucose positron emission tomography imaging

Figures & data

Table 1 Demographic and clinical characteristics of PD patients and healthy controls

Groups	N	Age (years)	Gender (M/F)	Disease duration (years)
Chongqing cohort
Healthy control	48	46.8±7.2	25/23	–
PD	48	59.6±10.6	31/11	2.94±2.06
Shanghai cohort
Healthy control	33	58.1±10.3	15/18	–
PD	33	57.4±10.5	15/18	3.98±4.13

Note: Data are given as mean±SD.

Abbreviations: N, number; PD, Parkinson’s disease.

Figure 1 Brain regions with significant metabolic abnormalities in PD patients relative to normal subjects detected by two-sample t-test analysis of 18F-FDG PET scans with SPM in Chongqing and Shanghai cohorts, respectively. (A) Metabolism in PD patients from the Chongqing cohort was characterized by metabolic increases in the putamen, GP, thalamus, pons, sensorimotor cortex and cerebellum (red), along with metabolic decreases in parieto-occipital areas (blue). (B) Metabolism in PD patients from the Shanghai cohort was similarly characterized by metabolic increases in the putamen, GP, thalamus, pons, sensorimotor cortex and cerebellum (red), covarying with metabolic decreases in parieto-occipital areas (blue). The overlays are depicted in neurological orientation. Thresholds of the color bars represent Z values.

Abbreviations: FDG, fluorodeoxyglucose; GP, globus pallidus; PD, Parkinson’s disease; PET, positron emission tomography; SPM, Statistical Parametric Mapping.

Figure 2 Brain regions with overlapping abnormal metabolism in PD patients relative to normal subjects detected by conjunction analysis of 18F-FDG PET scans with SPM across Chongqing and Shanghai cohorts. (A) Metabolism in PD patients was increased (red) in the putamen, GP, thalamus, pons, sensorimotor cortex and cerebellum, but decreased (blue) in parieto-occipital areas. The thresholds of the color bars represent T values. (B) Group differences in relative metabolic values in select cortical and subcortical regions, obtained post hoc in spherical volumes of interest (4 mm radius) centered at the peak of clusters deemed significant from prior SPM analysis on a voxel-by-voxel basis. Error bars represent mean±SD.

Abbreviations: FDG, fluorodeoxyglucose; GP, globus pallidus; HC, healthy controls; PD, Parkinson’s disease; PET, positron emission tomography; SPM, Statistical Parametric Mapping.

Table 2 Overlapping brain regions with significant metabolic differences between Shanghai and Chongqing patient cohorts relative to healthy controls

Structure	BA	Hemisphere	MNI coordinatesa			Z max	Cluster size (mm^3)
			X	Y	Z
Increased metabolismb
Lentiform nucleus	Lateral GP	L	−22	−16	−6	5.87	15,856
Lentiform nucleus	Lateral GP	R	24	−14	0	7.05	15,992
Paracentral lobule (sensorimotor cortex)	BA 5	L	−18	−34	50	4.92	2,408
Cingulate gyrus	BA 24	L	−14	−6	44	6.48
Cingulate gyrus	BA 24	R	22	6	42	5.68	208
Cerebellum	Anterior lobe	L	−12	−42	−40	7.24	33,592
Cerebellum	Anterior lobe	R	16	−44	−36	6.76
Decreased metabolismb
Parietal lobe	BA 39	L	−44	−70	30	5.26	664
Parietal lobe	BA 7	R	30	−76	46	5.36	3,184
Occipital lobe	BA 17	L	−18	−94	−16	6.22	36,408
Occipital lobe	BA 17	R	16	−100	−2	7.75

Notes:

a MNI standard space.

b Survived at FWE p<0.05, extent threshold =25 voxels (200 mm³).

Abbreviations: BA, Brodmann area; FWE, family-wise error rate; GP, globus pallidus; MNI, Montreal Neurological Institute.

Figure 3 Brain regions with different abnormal metabolism in PD patients relative to normal subjects detected by interaction analysis of 18F-FDG PET scans with SPM across Chongqing and Shanghai cohorts. Metabolism in PD patients from the Shanghai cohort exhibited a greater increase (red) in the right temporal-striatal area, but a more significant decrease (blue) in the bilateral occipital region compared with imaging from the Chongqing cohort. The thresholds of the color bars represent T values.

Abbreviations: FDG, fluorodeoxyglucose; PD, Parkinson’s disease; PET, positron emission tomography; SPM, Statistical Parametric Mapping.

Table 3 Nonoverlapping brain regions with significant metabolic differences between Shanghai and Chongqing patient cohorts

Structure	BA	Hemisphere	MNI coordinatesa			Z max	Cluster size (mm^3)
			X	Y	Z
Decreased metabolismb
Temporal lobe	BA 21	R	44	−8	−12	5.01	2,512
(extending to striatum)
Increased metabolismb
Cuneus	BA 18	R	12	−98	16	4.98	3,352
Inferior occipital gyrus	BA 17	L	−20	−94	−14	5.13	1,304

Notes:

a MNI standard space.

b Survived at FWE p<0.05, extent threshold =25 voxels (200 mm³).

Abbreviations: BA, Brodmann area; FWE, family-wise error rate; MNI, Montreal Neurological Institute.

Figure 4 Diagnostic power for individual expression of PD-related cerebral metabolic activities derived by Shanghai cohort. (A) Individual scores were significantly increased in PD patients relative to normal subjects in both the derivation cohort from Shanghai (left, p<0.001) and the prospective validation cohort from Chongqing (right, p<0.001). The expression was slightly lower in PD patients from Chongqing compared to data from the Shanghai cohort, but no significant difference was observed between the two control groups. (B) The ROC curve for individual scores prospectively calculated in subjects from the Chongqing cohort could significantly discriminate patients from healthy controls, with a sensitivity of 81.3% and a specificity of 85.4% at an optimum cut-off value of 0.6467. The inflection point corresponding to the optimal cut-off value was represented by an asterisk on the ROC curve.

Abbreviations: HC, healthy control; PD, Parkinson disease; ROC, receiver operating characteristic.