Network modulation by the subthalamic nucleus in the treatment of Parkinson's disease

Abstract

Deep brain stimulation of the subthalamic nucleus (STN DBS) has become an accepted tool for the treatment of Parkinson's disease (PD). Although the precise mechanism of action of this intervention is unknown, its effectiveness has been attributed to the modulation of pathological network activity. We examined this notion using positron emission tomography (PET) to quantify stimulation-induced changes in the expression of a PD-related covariance pattern (PDRP) of regional metabolism. These metabolic changes were also compared with those observed in a similar cohort of patients undergoing STN lesioning. We found that PDRP activity declined significantly (P < 0.02) with STN stimulation. The degree of network modulation with DBS did not differ from that measured following lesioning (P = 0.58). Statistical parametric mapping (SPM) revealed that metabolic reductions in the internal globus pallidus (GPi) and caudal midbrain were common to both STN interventions (P < 0.01), although declines in GPi were more pronounced with lesion. By contrast, elevations in posterior parietal metabolism were common to the two procedures, albeit more pronounced with stimulation. These findings indicate that suppression of abnormal network activity is a feature of both STN stimulation and lesioning. Nonetheless, these two interventions may differ metabolically at a regional level.

Fig. 1

(A) Parkinson’s disease-related pattern (PDRP) identified by network analysis of [¹⁸F]-fluorodeoxyglucose (FDG) PET scans from 30 PD patients and 30 age-matched normal volunteer subjects. This pattern was characterized by relative metabolic increases in the pallidum and thalamus (left) and in the pons and cerebellum (bottom). These changes covaried with metabolic decreases in the lateral premotor and parieto-occipital association areas (right). [The display represents voxels that contribute significantly to the network at P = 0.001 and that were demonstrated to be reliable (P < 0.001) by bootstrap estimation (Efron and Tibshirani, 1994). Voxels with positive region weights (metabolic increases) are color coded from red to yellow; those with negative region weights (metabolic decreases) are color coded from blue to purple.] (B) Bar histogram of the change in PDRP expression (ΔPDRP; mean±SE) quantified in hemispheres undergoing either STN deep brain stimulation (DBS, filled bar) or lesioning (LESION, shaded bar), and in non-operated control hemispheres (CNTL, open bar). Significant reductions in network activity were observed with both interventions (P < 0.05, see text). However, the degree of network modulation was not different for the two treatment groups (P = 0.58). [Asterisks refer to comparisons with control hemispheres; P < 0.05].

Fig. 2

Top: regions with metabolic reductions occurring in both the STN stimulation and lesioning groups (Table 3; see text). Metabolic decrements with treatment were detected in the midbrain (left), the internal segment of the globus pallidus (GPi) (middle), and in the anterior cingulate region (right). [SPM{t}maps were superimposed on a single-subject MRI brain template; surgically treated hemispheres appear on the right.] Bottom: bar histograms illustrate rates of glucose utilization for each of the significant clusters displayed above. Metabolic values (mean±SE) are presented for each treatment condition. The treatment-induced decrease in GPi metabolism was greater in magnitude in the STN lesion group (P < 0.01). [OFF/PRE (filled bars) and ON/POST (open bars) correspond to the stimulation (DBS) and lesioning (LESION) groups, respectively. Regional changes with intervention appear in parenthesis.]

Fig. 3

Left: regions with metabolic increases occurring in both STN stimulation and lesioning groups (see Table 3 and Fig. 2). Metabolic increases with treatment were detected in the parietal association cortex. [SPM{t}maps were superimposed on a single-subject MRI brain template; surgically treated hemisphere appears on the right.] Right: bar histograms illustrate rates of glucose utilization for the significant cluster in the posterior parietal cortex. Metabolic values (mean±SE) are presented for each treatment condition. The treatment-induced increase in parietal metabolism was greater in magnitude in the DBS group (P < 0.05). [OFF/PRE (filled bars) and ON/POST (open bars) correspond to the stimulation (DBS) and lesioning (LESION) groups, respectively. Regional changes with intervention appear in parenthesis.]