A Randomized Pilot rTMS Trial for Knee Arthritis Pain and Depression

Osteoarthritis (OA) is a major public health problem, and involvement of the knee is especially disabling. Symptomatic knee OA has an incidence rate between 40 to 1,020 per 100,000 person years1 and is among the most common causes of disability worldwide. Knee arthritis pain and disability are highly comorbid wiht depression (30-50%). Currently available treatments offer only limited relief. The Pilot project aims to establish feasibility of the rTMS neuromodulation of response to Tai Chi and improvement in pain and comorbid depression in patients with knee OA. There are several ways in which the pilot project will improve scientific knowledge, and clinical practice: 1) The sequential stimulation of two targets (M1 and l-DLPFC) has not been systematically examined for the treatment of comorbid MDD and knee OA. We hypothesize that using a multi-target rTMS strategy combining M1 and l-DLPFC- active targets will be well tolerated and more effective to treat comorbid symptoms than single site rTMS to M1+l-DLPFCsham. This hypothesis will be tested in Aim 1 of this proposal by comparing two experimental conditions: A) M1active and l-DLPFCactive; and B) M1active and l-DLPFCsham. 2) Identifying the relationship between improvement in pain and depression to improvement in pro-inflammatory cytokines would be novel. Adding an rTMS as a neuromodulation technique with novel stimulation sites to assist in the reduction of symptoms of pain and depression is another scalable to clinical use opportunity that will provide pilot data for future clinical trials. We will perform a pilot feasibility trial of rTMS for those presenting with knee osteoarthritis related pain and moderate to severe depression in 30 volunteers who are undergoing Tai Chi intervention. Tolerability and safety of rTMS added to Tai Chi will be assessed along with changes in symptoms of pain and depression, in preparation to future R-01 applications.

Major depressive disorder (MDD) and chronic pain are highly comorbid, especially in women. Pain symptoms can either precede or follow the onset of MDD, suggesting a bidirectional relationship between MDD and chronic pain. Knee OA causes structural and neural pain sensitization changes with segmental and central sensitivity, reduction of pain threshold such as hyperalgesia inside and outside the already sensitized areas, which can lead to generalized pain. Studies also have shown that rTMS can alleviate pain symptoms of different etiologies including FM, headache, neuropathic pain, and complex regional pain syndrome (CRPS). Several studies have highlighted the importance of the interaction between pain and depression, suggesting that the therapeutic effects of rTMS on pain and mood may be mediated by the joint modulation of pain pathways and social and affective regions of the brain. The most effective fibromyalgia pain relief was observed when targeting the primary motor cortex (M1), or the l-DLPFC. Direct comparison of the stimulation of M1, l-DLPFC, and sham on outcome in FM and depressive symptoms and found that while the pain intensity was decreased after M1 stimulation, better overall functional improvement was achieved with rTMS to the l-DLPFC. Our preliminary data show that rTMS treatment for MDD to left DLPFC reduced general pain symptoms in MDD patients, possibly through top-down regulation of pain processing regions, however, rTMS to l-DLPFC failed to improve severe chronic pain symptoms in those with MDD. We hypothesize that targeting two nodes related to depression and pain processing will enhance rTMS treatment outcome for comorbid knee OA and MDD.

We will perform a pilot feasibility trial of the Repetitive Transcranial Magnetic Stimulation (rTMS) protocol for neuromodulation of treatment response to Tai Chi that will compare 2 neuromodulation strategies in 30 subjects with OA knee arthritis pain and comorbid depression. rTMS is a non-invasive treatment that can modulate activity of brain circuits. rTMS most commonly has shown efficacy for relieving pain symptoms with an (M1) site in the left primary motor cortex being the primary stimulation site. rTMS is routinely used for treatment of depression with the stimulation to the left dorsolateral prefrontal cortex (DLPFC), however stimulation at either sits can improve both symptoms due to neuroplasticity of neurocircuits involved, as shown in our preliminary studies.

Pilot Study Research design 30 participants diagnosed with knee OA per parent study and major depressive disorder (MDD, DSM-5R; BDI >20) will be randomized 1:1 into two treatment conditions, using randomization scheme. All stimulation procedures will be double-blinded. The two stimulation protocols consist of 24 sessions, with each session including administration of: Group A) rTMS to M1-active/ l-DLPFC-active; Group B) rTMS to M1-active/l-DLPFC-sham. We anticipate up to 6 (20%; 3 per group) dropouts.

Additional Inclusion/Exclusion criteria for the pilot project are: (1) serious medical or psychiatric conditions other than MDD (DSM5); (2) the BDI score of >20 (moderate-severe depression); (3) stable treatment regimen for pain and depression for at least 4 months; (4) contraindications to TMS[69].

Randomization: We will assign participants 1:1 to group (l-DLPFC sham versus active) randomly by a computer-generated, block-permutation scheme managed by a study team member not involved in study assessments or treatment. For the total sample size of 30 (i.e.,~ 15 per group) participants will be randomly assigned to the treatment condition (sham versus active) using block randomization stratified on sex. A block size of 2 will be used to ensure approximate equal balancing between the anticipated smaller group sizes.

Experimental blinding: The utilization of a specialized A/P treatment coil (MagVenture Cool-B65 A/P coil) will allow for double blinding of the experimental condition for all involved personnel including TMS technicians, TMS physicians, raters, and patients. The symmetrical coil has no indication of active vs. placebo sides but, internally, one side has a magnetic shield that renders treatment ineffective. Sham stimulation mimics the auditory TMS experience through a clicking noise. Additionally, scalp electrodes placed on the left side of the forehand below the magnet deliver electrical stimulation to the skin with each TMS train to mimic the sensory experience (i.e., tingling sensations associated with receiving active TMS). Prior research suggests that treatment-naïve participants are unable to distinguish sham from active TMS using this arrangement. The study team member responsible for randomization will assign a participant to group A or B using the randomization scheme above. They will then inform the TMS technicians of a unique participant code that, when entered into the stimulator using its research interface, will direct the technician to use one side or another of the A/P coil - thereby determining whether the patient receives sham or active treatment. Neither the technicians nor the physicians delivering TMS, nor the raters will know which patient corresponds to which treatment arm. The study team members will also write down whether they believe study participants received sham or active treatment.

Neuronavigation: Neuroanatomic coil placement will be performed using the (Brainsight). The stimulation coordinates for the anatomical locations of l-DLPFC and M1 will be defined using the reliable targeting approach described in [78], based on a 3-dimensional brain reconstruction provided by the ANT Visor2 system. Left DLPFC defined by this algorithm corresponds to the junction between Brodmann Area (BA) 9 and BA 46, while the M1 target corresponds to the posterior border of hand representation. Coil position and angle will be maintained throughout and consistent across sessions.

rTMS treatments [70] will be performed with the MagVenture MagPro X100 stimulator equipped with the Cool-B70 A/P for sham-controlled stimulation. The A/P coil supports double-blinding of treatment. The coil has two sides, one of which is shielded so that no energy is imparted to the brain when the shielded side is applied to the subject's head. Small skin electrodes are applied to the scalp under the coil that impart a small microcurrent to the scalp simulating the sensation of active treatment. The operator receives a code from the device instructing them which side of the coil to use for each subject, ensuring complete treatment blinding. Active treatment to l-DLPFC will consist of 1800 pulses of intermittent theta burst stimulation (iTBS) at 120% MT. Active treatment to M1 will consist of 1500 pulses of 10 Hz stimulation with 20 trains of 10-s on and 30-s intertrain intervals at 100% MT. Intensity of active stimulation at both targets will be ramped up to the target intensity over the first three treatment sessions to increase tolerability.

Electric field modeling of TMS. Modeling will be performed to confirm that a therapeutic level of stimulation is delivered to the correct neuroanatomical areas. We will utilize simNIBS 3.0 (www.simnibs.org) to model current distributions with a finite element method based upon the magnetic dipole model for the MagVenture A/P treatment coil positioned to create a tangent with regard to brain structure. Electrical conductivities will be assigned to different tissue types with assumed isotropic conductivity. The intensity was set to 50% of maximum system output, a typical level needed to achieve 120% of MT. These results indicate that the therapeutic level of stimulation is largely confined to the target structures, and that the ratio of target to non-target activity is high. This targeting technique will ensure accurate targeting the DLPFC and M1 sites.

Randomization allocation concealment and blinding. Participants will be randomly allocated to either: (1) M1active/ l-DLPFCactive or (2) M1active/ l-DLPFCsham based on a 1:1 allocation ratio with the allocation list generated by the statistician. Participants, study physicians, and the PI will be blinded to the type of rTMS they will receive due to the coil. All participants will be given the same instructions and information about the rTMS intervention. Researchers conducting laboratory-based outcome assessment and will be blinded to group allocation. Unblinding will be allowed when an adverse or unexpected event occurs.

Measures of feasibility and safety. Feasibility and safety of the rTMS and Tai Chi intervention will be assessed using the following measures: (1) the number of sessions attended by each participant (attendance rate >80% is considered feasible)[71]; (2) the number of dropouts in each group (dropout rate <20% is considered feasible)[71]; (3) the proportion of participants recruited from the total number screened; (4) willingness of each participant to undergo interventions at baseline on an 11-point NRS with "not at all willing" at 0 and "very willing" at 10 (80% of participants score 7 or more are considered feasible); (5) the number of adverse events and the details of each event. Each adverse event will be considered separately. One or more serious adverse events will be considered unsafe.

Inflammatory markers: Inflammation is a characteristic feature of OA and depression, and a panel of cytokine/chemokines will be obtained to characterize a pattern of inflammation and to assess feasibility of using the panel to track changes in inflammation with TCC intervention[72]. Whole blood will be centrifuged at 2000 rpm for 10 min and plasma will be immediately stored at -80 0 C. We will use the Human 38-plex magnetic cytokine/chemokine kits (EMD Millipore, HCYTMAG60K-PX38). T