Treating Refractory Schizophrenia With rTMS

Background and Significance

There is increasing evidence from neuropsychological and imaging studies that cerebellar function is relevant not only to motor coordination, but equally to cognition and behavior (Rapoport et al., 2000). Selective modulation of cerebello-thalamocortical pathways, in turn, is believed to provide an additional means of modulating cortical function. Repetitive transcranial magnetic stimulation (rTMS) can modulate cortical excitability focally in conscious subjects. Trains with slow frequency (i.e.,1 Hz) are known to suppress cortical excitability (Chen et al., 1997), whereas facilitation occurs if frequencies higher than 5 Hz are used (Berardelli et al., 1998). With respect to rTMS of the cerebellum, a major impact on cognitive function (Oliveri et al., 2007) has been described.

The cerebellum is a very good candidate to be the generator for intracortical inhibition; its stimulation can modulate the cortical inhibition. Invasive studies by Robert Heath at Tulane University revealed that the cerebellum is strongly connected to 2 structures at the core of the proposed abnormal circuitry in schizophrenia, the septal nuclei and the hippocampus (HC). According to his theory and findings, the septal nuclei are involved in positive mood regulation, pleasure. The firing of the HC was correlated with negative affect and sadness (Heath et al, 1980). By stimulating the fastigial nucleus and vermis of the cerebellum, the septal nuclei were facilitated to fire and the HC was inhibited. The other component of what Heath referred to as the "aversive system", the amygdala was also inhibited. This central role of the cerebellum in this circuit is analogous to its role in "smoothing" the flow of movements. When considering emotion and cognition, the cerebellum has a smoothing function. Aside from direct monosynaptic connections between these sites, there is evidence that the deep cerebellar nuclei are connected to the parietal cortex, temporal cortex, as well as the cingulate gyrus. These are all areas that have limbic function. The cerebellum is also directly connected to the midbrain reticular activating system (RAS). This region is responsible for levels of consciousness and arousal. By potentiating the activation of the RAS, wthe investigators e can increase the decreased level of arousal, which in many schizophrenia patients is akin to psychomotor retardation and mutism (catatonic). Midline deep cerebellar nuclei efferents have been traced to the hypothalamus, central nuclei of the thalamus, which are also associative (cognitive) and limbic in function. The Locus Ceruleus and the substantia nigra, in the brain stem are also monosynaptically connected to the cerebellum.

The cerebellum is connected to thalamus and motor cortex (frontal cortex) through cerebello-thalamo-cortical pathway. And as stated above it is also connected to a vast array of limbic structures, making it a good choice to use to modulate abnormal activity in these structures.

Purkinje cells, the output neurons of cerebellar cortex reduce the excitatory drive from the deep cerebellar nuclei via the ventrolateral thalamus to inhibitory neurons in the motor cortex. Activation of Purkinje cells will inhibit the thalamic drive to intracortical inhibitory neurons, therefore, decrease the intracortical inhibitory interneuron activity and decrease in SICI and CSP. On the other hand, inhibition of Purkinje cerebellar cells is expected to have opposite effect and release the thalamus from inhibitory control, increase the thalamic drive to stimulate the inhibitory interneurons which can be demonstrated by increase in SICI and CSP Indeed, applying inhibitory rTMS at frequency of 1 Hz resulted in increase in SICI (Langguth et al., 2008).

The midline deep cerebellar nuclei, those that are anatomically and phylogenetically related to the vermis, also send collaterals to the reticular activating system (RAS) of the brain stem. By increasing the excitatory (Glu) drive on the RAS, the subject will experience increased awareness and connection to their environment.

For decades, the cerebellum has been thought to be predominantly involved in motor performance and cognitive operations. Recently, however, a growing body of evidence indicates that the cerebellum is also involved in emotion. The first evidence for cerebellar involvement in emotion came from the work of Robert G. Heath during the early fifties. Although his initial work predominantly involved the electrical stimulation of the septum, he then began research on stimulation of the cerebellum, thinking that it might provide a better entry to the emotional circuitry of the brain. Several cerebellar pacemaker studies by Heath did indeed demonstrate positive effects on mood and personality in patients with psychiatric illness after electrical stimulation of the cerebellum. Moreover, Schmahmann and Sherman provided clinical support for the role of the cerebellum and particularly the vermis in the regulation of emotion and mood. Given its modulatory role on emotion, the midline cerebellar vermis together with the fastigial nucleus and the flocculonodular lobe have been designated the limbic cerebellum (Schutter and van Honk 2005). Furthermore, additional evidence for the involvement of the cerebellum in schizophrenia was supported by genetic, structural and functional imaging data (Sandyk et al., 1991; Nopoulos et al., 1999; Ichimiya et al., 2001; Varnas et al., 2007) as well as by clinical evidence (Deshmukh et al., 2002; Ho et al., 2004; Varambally et al., 2006). For example, in an animal model for schizophrenia using prenatal infection of mice with human influenza virus, the animal developed behavioral changes similar to those of schizophrenia and was associated with altered expression of cerebellar genes (Fatemi et al., 2008). Some studies reported smaller bilateral cerebellar volumes as compared to controls in first episode schizophrenia patients (Bottmer et al., 2005). One of the first studies to demonstrate the importance of a dysfunctional cerebellar circuitry in schizophrenia was a positron emission tomography (PET) study (Andreasen et al., 1996). The authors examined memory performance in schizophrenia patients and correlate it to blood flow in cerebello-thalamo-cortical pathway. They used two tasks for memory, namely an easy and a relatively difficult one. While patients with schizophrenia showed normal performance on the easy practiced memory task they already demonstrated decreased blood flow in the cerebello-thalamo-cortical pathway. By contrast, in the relatively more difficult memory task, schizophrenia patients performed worse than healthy controls and displayed significantly lowered frontal and cerebellar blood flow (Andreasen et al., 1996).

Consistent with the assumed disruption of the cerebellothalamo- cortical pathway in schizophrenia is evidence from two proton magnetic resonance spectroscopic imaging (HMRS) studies. Lower levels of N-acetylaspartate (NAA), a marker for neuron density and viability, were found in the thalamus and cerebellar vermis (Deicken et al., 2001) in patients with schizophrenia. In keeping with these findings, lower NAA levels in the vermis and cerebellar cortex have also been found (Ende et al., 2005) as well as in mediodorsal region of the thalamus (Ende et al., 2001). In addition, poor executive functioning in patients with schizophrenia was associated with volumetric reductions in the cerebello-thalamo-cortical network (Rusch et al., 2007). Moreover, a diffusion tensor imaging (DTI) study has shown that patients with schizophrenia demonstrate abnormality in the connectivity between cerebellum and thalamus with possible difference between the right and left cerebellum (Magnota et al., 2008). Investigating connectivity between the cerebellum and thalamus in schizophrenia using diffusion tensor tractography: A pilot study). Another DTI study found neuronal disorganization in the superior peduncle with neuronal disorganization being associated with poor cognitive performance (Okugawa et al., 2006). Finally, the activity of right and left cerebellum may not be the same. For example, impaired working memory in schizophrenia is associated with over and under-activation along the cerebellothalamo- cortical pathway with under-activation of the left DLPFC and right cerebellum and over-activation of the left cerebellum (Mendrek et al., 2005).

To date, cerebellar involvement in schizophrenia remains a subject of ongoing study. It was shown that motor impairments in schizophrenia are related to cerebellar malfunction. Several studies report that the cerebellum is indeed involved in cognitive (Eyler et al., 2004; Aasen et al., 2005; Kiehl et al., 2005) and affective (Paradiso et al., 2003; Takahashi et al., 2004; Stip et al., 2005) impairments. This study aims to clarify the role of the cerebellum in development of negative symptoms through its regulation of cortical inhibition, activation of the septal region with reciprocal inactivation of the hippocampus, and RAS activation.

Experimental Design and Methods/Procedures

1. rTMS over the vermis of the cerebellum
2. 5 sessions/week for 1 week
3. Randomization as explained below.

Patients will be randomly assigned to either the high frequency or low frequency Medial Cerebellum Target Treatment protocol. Each group will then be entered into a randomized, double-blind, sham-controlled, parallel-design clinical trial that consists of three main phases: (1) Baseline Psychiatric and Psychometric Testing Exam; (2) 5 rTMS treatments, double-blind, in 5 treatment sessions/week with active or sham rTMS for over a period of 1 week; and (3) a follow-up period of 3 weeks. The patients will then be reassigned to the other Frequency (either high or low) Arm of the study. The protocol will then be repeated. Patients and the investigators, except the investigator who applied rTMS, will be blinded to the treatment arm.