Anterior thalamic nucleus stimulation modulates regional cerebral metabolism: An FDG-MicroPET study in rats

Abstract

The mechanism underlying the antiepileptic function of deep brain stimulation (DBS) of the anterior thalamic nucleus (ATN) remains unknown, presumably related to functional lesioning of target. We measured the regional normalized cerebral metabolic rate of glucose (nCMRglc) with ¹⁸F-fluorodeoxyglucose (FDG)-MicroPET in animals receiving either ATN stimulation or lesioning. Bilateral ATN stimulation reversibly increased glucose uptake in the target region, the thalamus and hippocampus, and decreased glucose uptake in the cingulate cortex and frontal cortex. However, bilateral ATN lesioning decreased glucose uptake only in the target region. Animals with bilateral ATN lesions showed no metabolic changes after ATN stimulation. Thus, bilateral DBS of the ATN reversibly induces metabolic activation of the target area and modulates energy metabolism in remote brain regions via efferent or afferent fibers in non-epileptic rats. DBS of the ATN may work by a different mechanism than ATN lesioning.

Introduction

Deep brain stimulation (DBS), usually with high-frequency stimulation, is a new option for intractable epilepsy. The anterior thalamic nucleus (ATN) is a DBS target that can influence seizure propensity based on its connectivity to the cortex and limbic structures (Upton et al., 1985). ATN stimulation is anticonvulsant in several seizure models (Hamani et al., 2004, Hamani et al., 2008, Mirski et al., 1997, Takebayashi et al., 2007), and proved to be feasible and effective clinically (Hodaie et al., 2002, Kerrigan et al., 2004, Lim et al., 2008).

DBS may block seizures through local inhibition of the target area, a functional lesion effect (Grill et al., 2004), as suggested by initial animal studies (Benazzouz et al., 2000, Burbaud et al., 1994). DBS also produces clinical benefits analogous to those achieved by surgical lesioning of the target in movement disorders (McIntyre et al., 2004). Both DBS and lesioning of the ATN have similar anticonvulsant activity on pilocarpine-induced status epilepticus (Hamani et al., 2004) and on kanic acid-induced focal cortical seizures (Takebayashi et al., 2007). Clinically, some investigators even proposed that DBS of the ATN may work via the physical lesion caused by the implantation of the electrodes, a microthalamotomy effect (Hodaie et al., 2002, Lim et al., 2007).

However, the lesion hypothesis may be too simple to explain how DBS controls seizures, because DBS can induce a complex pattern of inhibition and excitation in the brain (Vitek, 2002). Besides the direct target, DBS can also modulate the activity of remote areas anatomically connected with the target by activating axonal elements (Brown et al., 2004, Hashimoto et al., 2003, Hilker et al., 2008). For example, DBS of the ATN decreases motor cortex excitability in patients with intractable epilepsy, as measured electrophysiologically (Molnar et al., 2006). However, the electrophysiologic methods can only measure selected downstream areas, whereas Positron Emission Tomography (PET) imaging can measure DBS-induced changes in the regional cerebral metabolism rate (rCMR) of the entire brain, including some deep subcortical areas.

Therefore, we measured rCMR with ¹⁸F-fluorodeoxyglucose (FDG) small animal PET (MicroPET) in rats receiving either high frequency stimulation or lesioning of the ATN. We used a region of interest (ROI) approach to measure the metabolic effects of DBS. The rCMR in each ROI is presented as a normalized cerebral metabolic rate of glucose (nCMRglc). The nCMRglc provides a qualitative index of neuronal activity and has been used in the previous DBS studies (Fukuda et al., 2001, Hilker et al., 2008). In the present study, we studied the DBS-induced changes in nCMRglc in rats not previously treated with proconvulsant agents to avoid disease-related confounding factors.