Effects of Direct Transcranial Current Stimulation on Central Neural Pain Processing in Fibromyalgia

1. BACKGROUND AND SIGNIFICANCE:

1. Fibromyalgia (FM):

*Fibromyalgia is the second most common rheumatologic disorder, behind osteoarthritis, afflicting 2-4% of the population of industrialized countries.(Jacobsen and Bredkjaer, 1992; Wolfe et al., 1990) To fulfill the criteria for FM established by the American College of Rheumatology in 1990, an individual must have both chronic widespread pain involving all four quadrants of the body (and the axial skeleton), and the presence of 11 of 18 pre-defined "tender points" on examination. A positive tender point is identified when an individual complains of pain when approximately four kilograms of pressure is applied to one of these points by an examiner. FM is the prototypical "central" or "non-nociceptive" pain syndrome. Research performed within the past decade has clarified a number of important issues regarding this condition. Multiple studies suggest neurological dysfunction as a hallmark of this disease (Clauw and Crofford, 2003), and this is supported by a number of objective functional neuroimaging abnormalities. (Gracely et al., 2002; Harris et al., 2007; Mountz et al., 1995) Overall the data suggest that the primary abnormality in FM is a generalized disturbance in central nervous system pain processing, leading individuals to sense pain throughout the body in the absence of inflammatory or patho-anatomic damage. (Clauw and Chrousos, 1997; Yunus, 1992) Most FM neuroimaging studies to date have examined brain responses to a painful stimulus, as the imaging of endogenous chronic pain is notoriously difficult. (Baliki et al., 2007). However few studies have examined the modulation of specific brain regions and how this impacts neurotransmitter levels, network connectivity, and structural changes such as cortical thickness within the same subjects.

2. Transcranial Direct Current Stimulation (tDCS):

*Therapies that directly modulate brain activity in specific neural networks might be particularly suited to relieve chronic pain in individuals with FM. Ultimately, this underlies the interest in neurostimulation approaches, which are being explored at multiple levels of the neuroaxis, including the peripheral nerves, spinal cord, deep brain structures, and cortex.(Lefaucheur, 2004) Among the methods of central neurostimulation, two of them, repetitive transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS), are particularly appealing as they can change brain activity in a non-invasive, painless and safe way. TMS is a method of brain stimulation that was developed in 1985 (Barker et al., 1985). It is based on a time-varying magnetic field that generates an electric current inside the skull where it can be focused and restricted to small brain areas by appropriate stimulation coil geometry and size.(Pascual-Leone et al., 1999). This current, if applied repetitively, repetitive TMS (rTMS), induces a cortical modulation that lasts beyond the time of stimulation.(Pascual-Leone et al., 1999) Although tDCS has different mechanisms of action, it induces similar modulatory effects. Several animal studies in the 1960s showed that this technique changes brain activity reliably (Nitsche et al., 2003a, 2003b). tDCS is based on the application of a weak direct current to the scalp that flows between two relatively large electrodes-anode and cathode. Some studies have shown that the efficacy of tDCS depends critically on parameters such as electrode position and current strength.(Nitsche et al., 2003a, 2003b) In fact, application of tDCS for 13 min to the motor cortex can modulate cortical excitability for several hours.(Nitsche and Paulus, 2000; Nitsche and Paulus, 2001) In addition, this technique can be used to obtain clinical gains in neuropsychiatric disorders such as stroke and epilepsy.(Fregni and Pascual-Leone, 2007) In this study we will investigate the modulatory effect of 5 daily tDCS sessions on biochemical, functional, and structural systems and its association with the clinical output in FM.

3. Proton Magnetic Resonance Spectroscopy (H-MRS) in FM:

*H-MRS neuroimaging obtains chemical spectra from multiple volume-image elements, or voxels, within the human brain using radiofrequencies that excite protons. (Ross and Sachdev, 2004) Specific molecules are identified by their characteristic resonance frequency in the spectrum. Once acquired, spectra are analyzed to determine the relative concentrations of different molecules or central nervous system metabolites within the voxel or region of interest. Typical metabolites identified are: glutamate (Glu), N-acetyl-aspartate (NAA), creatine (Cr), choline (Cho), lactate, lipid, myoinositol, gamma-aminobutyric acid (GABA), and glutamine (Gln). Glu and GABA are of particular importance to brain neurophysiology as they are components of excitatory and inhibitory neurotransmission, respectively. Glu binds to both ionotropic and metabotropic receptors located on postsynaptic neurons and causes excitability (i.e. depolarization). Moreover changes in the strength of Glu neurotransmission are typically indicative of synaptic plasticity, a process proposed to be involved in chronic pain.(Zhuo, 2008) H-MRS methods display multiple features which are amenable to longitudinal studies. High-resolution anatomical scans can be used to isolate identical brain regions on successive sessions that are even weeks apart. Measurement of metabolites within the central nervous system has been largely understudied in the field of pain. Grachev et al. has reported that the level of NAA, a marker for neuronal viability and also function (Nakano et al., 1998; Sager et al., 2001), is lower within the dorsolateral prefrontal cortex of individuals with chronic low back pain as compared to healthy controls.(Grachev et al., 2000) In addition, a recent investigation has begun to implement H-MRS technology to assess functional changes in the concentrations of Glu in response to evoked pain stimuli.(Mullins et al., 2005) Mullins et al. have observed that Glu levels increase by as much as 10% in the anterior cingulate in response to cold pain applied to the foot. Glu in the central nervous system may play a role in FM pathophysiology. A study by Peres et al. found that cerebrospinal fluid levels of Glu were elevated in FM patients possibly having consequences for glutamatergic neurotransmission.(Peres et al., 2004) Administration of ketamine, a glutamate channel blocker, has been found to reduce experimental pain (Graven-Nielsen et al., 2000) and clinical pain (Cohen et al., 2006) in FM. Moreover our group recently demonstrated that long-term treatment of FM patients with acupuncture can lead to changes in Glu levels within the posterior insula and that these changes are highly correlated with changes in pain: greater reductions in Glu are associated with greater reductions in both experimental and clinical pain (Harris et al., 2008). In addition, we have recently compared posterior insula Glu and combined Glu + Gln (Glx) between FM patients and matched controls and have demonstrated that the patients have elevated Glx (and Glu) levels. (Harris et al., 2009).

4. Resting state networks (RSNs) in FM:

- Previous studies have found that in a task-free state (i.e. rest scan), multiple distributed brain areas demonstrate temporal correlation of the fMRI signal or "functional connectivity" in low frequency ranges.(Biswal et al., 1995; Fransson, 2005) In one of the first such studies, Biswal et al. found a significant correlation in resting fMRI signal from sensorimotor cortices of opposite hemispheres.(Biswal et al., 1995) This resting state network (RSN) has been referred to as the sensorimotor network, or SMN. (Beckmann et al., 2005) FM pain is somatic in localization (usually soft tissue), hence resting connectivity in the SMN may demonstrate increased connectivity to the pain processing regions. Other RSNs have also been described, including one anatomically consistent with the Default Mode Network (DMN) (Greicius et al., 2003) [for review see (Buckner and Vincent, 2007; Vincent et al., 2007)]. This network involves brain regions putatively engaged in self-referential cognition that is "deactivated" (more active at rest than during a task state) during a variety of externally focused task conditions. Typically, the DMN (Figure 1) includes the inferior parietal lobule (IPL) (~BA 40, 39), the posterior cingulate cortex (~BA 40, 39), the posterior cingulate cortex (~BA 30, 23, 31) and precuneus (~BA 7), areas of the inferior, medial and superior frontal gyri (~BA 8, 9, 10, 47), the hippocampal formation, and the lateral temporal cortex (~BA 21)(Buckner and Vincent, 2007). Resting fluctuations in the DMN have demonstrated decreased connectivity in Alzheimer's disease (Greicius et al., 2003) and increased connectivity in depression (Greicius et al., 2004), compared to healthy controls. Interestingly, resting state connectivity in the DMN has also been shown to change in response to an intervention or task.(Waites et al., 2005) Waites et al. found increased connectivity between the middle frontal gyrus and posterior cingulate (a component of the DMN) in resting fMRI data following an active (cognitive) task. While the functional significance of spontaneous fluctuations in the DMN remains controversial, Fox and Raichle suggest that resting connectivity in the DMN is fundamental to balancing excitatory and inhibitory inputs to multiple brain networks, thereby setting the "gain" for future task-related response. (Fox and Raichle, 2007) Positive correlations in fMRI signal refer to putatively excitatory connections whereas negative correlations imply putative inhibitory connectivity. We propose that application of tDCS with decrease connectivity in the pain matrix regions that may result in a change within the gain set by the DMN for brain processing within the pain matrix.

5. White (WM) and Gray Matter (GM) Plasticity in Fibromyalgia:

*The cortical mantle is a highly specialized, folded structure composed of a thin layer of GM. Abnormal variations in the thickness of the cortical mantle might reflect pathophysiological changes of intrinsic structure and integrity of the cortical laminae. Recently, some studies have shown this correlation in chronic pain diseases such as back pain (Apkarian et al., 2004), migraine (DaSilva et al., 2007b; Granziera et al., 2006) and trigeminal neuropathic pain (see preliminary data). The implications of an alteration in these diseases are either degenerative processes or neuroplasticassociated mechanisms. Apkarian and colleagues (Apkarian et al., 2004) found reduction in the gray matter of DLPFC of chronic back pain patients when compared to healthy controls using a volumetric based approach. More recently, such GM volume reduction was also found in the parahippocampus, and cingulate cortex of patients with fibromyalgia when compared to healthy controls. However, it seems that similar changes observed in the GM of fibromyalgia patients may be more related to comorbid affective disorders than the pain endurance (Peres et al., 2004; Wood et al., 2009). Using more sensitive and reliable neuroimaging tools in trigeminal neuropathic pain patients our group found cortical thickness changes that were spatially co-localized with functional allodynic (brush induced pain) activation. In addition, this pattern of concurrent structural and functional changes in chronic pain patients is influenced by somatotopic localization (sensorimotor cortex), known functionality of the specific region (sensory-discriminative and affective-motivational), underline activation/deactivation following allodynic stimulation and the duration of the disorder (see preliminary data). In another study of migraine patients, we found increased cortical thickness of caudal sensorimotor cortex in migraineurs compared to controls (DaSilva et al., 2007a). In the cortical mantle, the thickness changes in the sensory cortex could be due to the chronic sensory stimulation provoked by chronic pain. This is in line with a recent study that showed cortical thickening after sustained stimulation of the motor system (Draganski et al., 2004). In this study, volunteers who have learned to juggle showed transient and selective thickening of the motor cortex, as well as the motion-visual areas (MT/V5), as compared to the pre-learned phase. This suggests that overstimulation of the sensory-discriminative and affective-motivational neuronal systems in chronic pain may induce structural alterations in the cortex that is co-localized with inefficient pain modulation by the opioidergic system at a molecular level.

6. Evaluation of Diffuse Noxious Inhibitory Controls (DNIC):

- There is a body of evidence that suggests that the spontaneous pain and hyperalgesia associated with CMI is due to a dysregulation of intrinsic analgesic systems. The most well known intrinsic analgesic system is the endogenous opioid system, which appears to function normally in CMI. Another system, termed DNIC (Diffuse Noxious Inhibitory Controls), is characterized by widespread analgesia evoked by a noxious stimulus applied to anywhere on the body, such as tourniquet ischemia, or immersion of in painfully hot or cold water. The widespread nature of the DNIC effect, involving convergent second order neurons and a spinal-brain loop, is consistent with the diffuse widespread pain of CMI disorders such as FM. The results of several studies suggest that DNIC may be altered in CMI. Lautenbacher and Rollman observed that DNIC evoked by hot water immersion decreased the sensitivity to painful electrical stimuli in healthy control subjects but had no effect in patients with FM. Marchand [unpublished observations] has reported a similar effect using immersion of an arm in painfully hot water as both the conditioning and pain stimulus. This method shows an effect of DNIC on pain ratings in healthy controls but no effect in FM. Kosek and Hansson found that the DNIC manipulation of tourniquet ischemia decreased the sensitivity to painful pressure in control subjects but not in patients with FM.

- Together, these results are consistent with the hypothesis that the pain and tenderness in FM may be due to tonically inactive DNIC analgesic systems. These results, however, do not specify causality and could also represent a mechanism in which DNIC is tonically activated in CMI in response to the widespread ongoing pain of the disease. These alternative mechanisms cannot be separated using conventional psychophysical tests. Performing the tests in the fMRI scanner will differentiate these mechanisms because in one case the DNIC system remains "OFF" in the patient populations, in the other case the DNIC system is constantly "ON." fMRI analysis of activity in brainstem regions (e.g., caudal medulla) implicated in intrinsic DNIC analgesia will provide evidence for tonic ON or OFF activity in these regions, and in addition further specify the neuroanatomical locus of this abnormal pain processing. The fMRI analysis will provide crucial evidence of whether FM results from a DNIC defect or whether the DNIC abnormality is simply one of the signs of the disease.

2. RATIONALE (proposed research, and potential benefits to patients and/or society):

1. There is not much information about FM disease and treatment options available. This study seeks to gain a better, more wholesome understanding about Fibromyalgia. People who suffer from this disease experience constant, chronic pain; which ultimately results in absence from school, work, etc. If a feasible treatment is available for Fibromyalgia sufferers, they will have an increase in life satisfaction and the bio-power (people able to work and perform more tasks) will increase.

3. SPECIFIC AIMS (Research Objectives):

a.The main goal of this Collaborative Proposal is to investigate biochemical, functional, and structural neuroimaging changes following non-invasive brain stimulation in patients with chronic widespread pain: fibromyalgia (FM). Additionally, we aim to:

*Determine the effects of tDCS on the excitatory neurotransmitter glutamate (Glu) within the insula (posterior and anterior) and thalamus in individuals with FM. Glu levels within the insula and thalamus will be reduced following tDCS, reflecting a down regulation of excitatory neurotransmission in these pain regions.

- Investigate whether long-term therapy with tDCS normalizes gray matter thickness in target and cortical areas associated with pain perception and modulation in FM. Cortical thickness in FM patients, will return to comparable age- and sex-matched pain-free control participant levels following tDCS. These effects will be specifically detected in pain modulatory regions (e.g. dorsal lateral prefrontal cortex) of FM patients.

- Explore the effects of long-term tDCS on intrinsic connectivity between pain processing and modulatory regions and other brain networks (e.g. default mode network, sensory motor network) in FM. Our preliminary data suggest that FM patients display enhanced connectivity between various pain processing regions and the default mode network, a specific brain network that is active during periods of inactivity. We propose that tDCS will decrease connectivity between pain modulatory regions and other networks such as the default mode network thus resulting in a reduction in pain symptoms.

4. RECRUITMENT METHODS:

a.Potential subjects will be recruited by public advertisement in the School of Dentistry clinics, including MCOHR, and the Chronic Pain and Fatigue Research Center in addition to other University of Michigan clinics. They will also be recruited via UMClinicalStudies.org, the DaSilva lab webpage (with a flyer for the study listed under current research), ClinicalTrials.gov, and the U.S. National Institutes of Health. In addition, subjects may be recruited by the PI or study staff in a private setting. The potential subject's healthcare providers will be able to suggest the availability of the study and inform them of a place where they will be able to find more information about participating in the study.

5. STUDY PROCEDURES:

1. This study requires a total of 15 visits, broken down by the following: 1 baseline visit, 3 MRI visits, 10 tDCS testing visits, and 1 final follow-up/debriefing visit. Patient participant will last for a total of 5 consecutive weeks. During this time, we will be collecting clinical and psycophysical evaluation: the MRI photo imaging, DNIC/MAST pain tolerance data (computerized), and pain questionnaires (verbal), Quantitative Sensory Test (QST).

2. No drugs will be administered during this study

3. Devices used will include: MRI, tDCS, MAST/DNIC

6. RISKS/DISCOMFORT:

1. While these therapies are non-invasive, study participants may experience unpleasantness from the constant stimulus from the MAST/DNIC procedures; however, the pressure is not enough to cause any damage to the nail bed. The study participant is encouraged to inform the researchers of any discomfort/side effects they experience during any point of the study as the priorities of the research team is to keep the study participant safe. With regards to the tDCS testing, the participant may experience a temporary tingling sensation and minor skin irritation/redness as a result of the brain stimulation pads