Pain Avoidance Behavior and Its Relation to Risk for Opioid Use in Chronic Pain Patients

This study is designed to find behavioral and neurophysiological correlates of pain avoidance behavior among chronic pain patients, controlling for risk for opioid use disorder. Further, traits and risk factors that contribute to pain avoidance behavior will be investigated. The knowledge gained will broaden the current understanding of mechanisms involved in pain avoidance behavior in chronic pain patients, and help devise novel interventions.

Overall Research Design. The overarching goal of this study is to gain mechanistic insights into maladaptive pain avoidance behavior and risk factors that could provide a pathway for OUD in chronic pain patients. We aim to explore the neurophysiological constructs (brain oscillations) of the maladaptive pain avoidance behavior. Advancing this knowledge will open new opportunities for future clinical interventions to prevent opioid addiction in chronic pain patients. Aim-1 is based on a mechanistic framework termed predictive coding, and is designed to elicit cue-pain associative learning impairments that form the basis for maladaptive pain avoidance behavior in chronic pain patients. Aim-2 is designed to examine the influence of OUD related personality traits and risk factors in moderating maladaptive pain avoidance behavior. We strongly believe that aim-1 and 2 will aid the development of future clinical paradigms for early identification of patients vulnerable to OUD, treatment resistance and other psychodynamic factors that increase the risk for poor outcomes. Aim-3 further enhances the scope of the project by incorporating the study of brain oscillations during pain avoidance behavior that could help the development of non-addictive neuromodulation based interventions to promote extinction of learned maladaptive pain avoidance behavior. Subjects. The study will evaluate patients with Complex Regional Pain Syndrome (CRPS) who present with central sensitization i.e. allodynia (pain response to a non-painful stimulus) and/or hyperalgesia (exaggerated pain response to a nociceptive stimulus), and age/gender matched healthy controls. The advantage of the CRPS model of chronic pain is that it is common and a common cause of post-injury disability. Additionally, patients typically present with an affected and non-affected extremity/site facilitating within-subject comparison. Patients will be considered candidates after review of their history, neurological and psychological pain assessment. Healthy controls will be recruited through fliers posted throughout our campus and by referrals from other studies. Sample Size. Sample size is 30 patients and 30 age/gender-matched controls. The sample size was chosen based on the recommended power size for cue-based pain modulation studies. Study Design. This cross-sectional study will present the participants with an instrumental pain avoidance learning task, while we record continuous MEG brain signals, and other autonomic measurements. All participants, upon initial screening and enrollment, will report to our MEG facility on the day of testing. They will undergo: 1) training to understand the task; 2) scalp measurements for head tracking inside the MEG scanner; and 3) actual testing in the MEG. A T1-weighted MRI (for anatomical co-registration of MEG data) will be acquired separately, if a suitable dataset is not already available. Health History and clinical assessments. All patients will complete neuropsychological and pain assessments to rate their current pain, disability, fear, anxiety, and depression using standardized scales. In addition, we will administer Temperament and Character Inventory (TCI) questionnaire to assess the traits for aim-2. OUD risk classification: The Cleveland Clinic Alcohol and Drug Recovery Center (ADRC) staff has standard procedures to determine the risk for OUD based on biological, psychiatric and social risk factors derived from health history questionnaires. Each patient enrolled will get an ADRC assessment that will include one of three determinations: "Low Risk", "At Risk", or "Substance Use Disorder." The risk for OUD will be used as a covariate in our analysis for all aims, as necessary. Effects of Drugs. We will convert all medicine dosages that patients currently take to that of a representative medicine in class dosage (e.g. pain killers, antidepressants) and then use that as a covariate in the analysis, as necessary, to examine drug effects on the results. Instrumental pain avoidance task. The task will comprise 150 trials (3 runs of 50 trials), during which subjects will select a visual cue (out of 2 choices) with the lowest probability of receiving a painful stimulus. They start out by trial and error, but will soon learn to associate a cue with higher probability of receiving a painful outcome and avoid the cue. At least twice during the testing, the pain outcome contingencies of the cues will reverse and the subjects will have to adjust accordingly. The probabilities (for receiving a painful outcome) associated with each cue will be independent of one another and varied from trial to trial probabilistically. Each trial starts with the presentation of the two options (circle or square, randomly displayed to the left or right) for 1.8s, during which participants enter their decision by pressing the left or right button of the response unit. Inactivity will result in the computer randomly selecting a response. After a feedback period of 200ms and an anticipatory period of 2s, the fixation point changes from a cross (+) to an asterisk (*) displayed for 3s to mark the period during which participants could receive a painful or no stimulus. Following a painful stimulus, participants will rate their pain by clicking on a computerized visual analog scale (CoVAS, Medoc Inc.) presented on the screen. The next trial starts after a jittered inter-trial interval of 3 - 5s. To minimize the effects of sensitization/habituation, the painful stimulus will be delivered to a slightly different location after each run. Participants will perform a limited practice session with visual cues and pain outcome probability different from the test session. During practice, they will be carefully instructed about all aspects of the experiment, except the actual probabilities of pain. Patient's non-affected extremity will be tested before the affected extremity. Experimental pain: For this proposal, a pneumatic pressure pain device (PPD) developed at the University of Colorado (Wager lab) will be used. The PPD uses a pneumatic piston to deliver pain stimuli, and can be adapted to deliver pain to any part of the body. We propose to use either the thumb or the index finger. Prior to experimental task, participants will be tested for threshold limits to identify pain stimuli intensity. We will use a stepwise procedure, where pressure in the piston will be increased in steps of 1kg/cm (ramp and hold for 1s) from lowest possible level to a level where participants pain perception is equal to 8 on a VAS (0, no sensation; 10, worst possible pain). Electrophysiology/Neurophysiology. Our group has proven success in collecting pain event-related electrophysiological recordings from patients, including MEG/electroencephalography (MEEG). Here, we will be collecting these data while subjects perform the task described above. Simultaneous measurement of scalp MEEG (2400 samples/sec) will be performed in a magnetically shielded room (Imedco, Switzerland). The 306-channel MEG system (Elekta, Finland) at Cleveland Clinic records both MEG and up to 128 channels of EEG. For scalp EEG, international 10-10 electrode placement will be employed. The MEG system has additional channels with appropriate gains for recording autonomic measures such as electrocardiogram (EKG) and skin conductance responses (SCR) using MEG compatible Ag/AgCl electrodes and standard procedures. SCR electrodes will be attached on the thenar and hypothenar eminences on the same extremity where pain will be delievered. Patients will be evaluated in a seated position. Prior to the experimental session, fiduciary points (nasion, right and left auricular), along with head surface points and head position indicator (HPI) locations will be tracked on subject's scalp using a Fastrack digitizer (Polhemus, USA) to allow for anatomical co-registration with MRI. The HPI tracks the position and orientation of the head to account for variance due to movement. Subjects will be monitored inside the MEG suite using a surveillance camera. An intercom system allows for communication with the subjects. An E-prime stimulus presentation system (Psychology Software Tools, USA) integrated with state-of-the-art projector (Panasonic, Japan) will be used to provide cues. Participants will undergo 3 - 5 mins of resting state MEEG measurements prior to the task.

This cross-sectional study will present the participants with an instrumental pain avoidance learning task, while we record continuous Magnetoencephalography (MEG) brain signals, and other autonomic measurements. All patients will complete behavioral, neuropsychological and pain assessments to rate their current pain, disability, fear, anxiety, and depression using standardized scales.

Inclusion Criteria: 18 years and older Diagnosis of CRPS I, CRPS II Six months or more of medically refractory severe pain Average daily pain for the past 30 days reported as ≥ 5 on a 0 - 10 scale Presence of allodynia Unilateral pain or pain affecting predominantly one extremity. Predominantly will be operationally defined at >=75 of pain on one extremity. Willingness to complete behavioral, clinical and neurophysiological measurement procedures Able to provide informed consent Able to undergo structural brain MRI Exclusion Criteria: 17 years and younger All other pain diagnosis Less than six months or more of medically refractory severe pain Average daily pain for the past 30 days reported as < 5 on a 0 - 10 scale Absence of allodynia Bilateral pain Unwillingness to complete behavioral, clinical and neurophysiological measurement procedures Unable to provide informed consent Claustrophobic or unable to undergo MRI, MEG or unsafe to undergo MRI (i.e. due to MRI unsafe medical implants).

Roy M, Shohamy D, Daw N, Jepma M, Wimmer GE, Wager TD. Representation of aversive prediction errors in the human periaqueductal gray. Nat Neurosci. 2014 Nov;17(11):1607-12. doi: 10.1038/nn.3832. Epub 2014 Oct 5.

Gopalakrishnan R, Burgess RC, Malone DA, Lempka SF, Gale JT, Floden DP, Baker KB, Machado AG. Deep brain stimulation of the ventral striatal area for poststroke pain syndrome: a magnetoencephalography study. J Neurophysiol. 2018 Jun 1;119(6):2118-2128. doi: 10.1152/jn.00830.2017. Epub 2018 Jan 31.