Aims and objectives
The aim of the present study protocol is to evaluate if MORE is able to restore DA function in FM patients, in particular with regard to the DA responses to reward, and to reduce pain and negative mood in FM.
The objectives are as follows:
To compare the 18F-DOPA influx in striatal regions and neural responses to reward with fMRI before the MORE intervention between FM patients and a group of healthy controls
To compare the 18F-DOPA influx and neural responses to reward measured with fMRI after the MORE intervention between FM patients that participated in MORE versus those assigned to a wait-list control condition.
Explore changes in FM related pain and mood symptoms after the MORE intervention
Explore changes in self-report measures of behavioral and biological stress in FM patients after the MORE intervention
Our hypothesis are:
FM patients show a lower 18F-DOPA influx in striatal regions and decreased neural response to reward measured with fMRI than a healthy control group before the MORE intervention
FM patients participating in the MORE intervention show an increased 18F-DOPA influx and increased neural responses to reward measures with fMRI after the MORE intervention compared to a wait-list control group.
We expect FM patients participating in the MORE intervention to show significant changes in pain, mood and FM related outcomes compared to a wait-list control group.
We expect changes in everyday self-report measures of stress and pain levels including biological stress measures in FM patients versus healthy controls before the MORE intervention.
This is a multi-center RCT with 3 time points: before the intervention (pre-test measures, T1), after completion of the intervention (post-test measures, T2) and 3-months after completion of the intervention (follow-up measures, T3) (see Fig. 1). Participants in the wait-list control condition will perform the T1 and T2 measures adjuvant eight weeks apart. Study evaluation will be done by comparing within and between the groups. Healthy controls will perform T1 measures only. The potential effect will be assessed on a series of outcome measures (Table 1). Measures will take place at baseline (T0), T1 T2 and T3. Before T1, we will perform the screening of the participants with regard to the inclusion and exclusion criteria (T0). Pre-test and post-test measures include each a 18F-DOPA PET scan at rest and a fMRI measure with use of the reward task, the ambulatory assessment (AA) measures, the clinical and pain measures as well as the fMRI measures. The follow-up consists in questionnaires assessments only, including the self-reported pain and clinical measures associated with PET and fMRI meausres. The questionnaires will be presented through an online survey that is suitable for RCT’s (Research Electronic Data Capture, redcap) and available at the Universities Fribourg,and that can be filled at home. The measures that are not self-reported will be performed by trained assessors, who will be blind to the treatment condition to minimize expectation bias of the assessors.
The wait-list control will undergo the same measures at pre-test and post-test, with a post-test planned on average after 8 weeks. They will not participate in any specific intervention, but will be given the opportunity to participate in MORE after the follow-up assessment.
The study is a multi-center study and includes the University Fribourg, Department of Psychology; the University Hospital Lausanne (CHUV), Center for integrative and complementary medicine & Pain Center, and University Hospital Zurich, Department of Consultation-Liaison Psychiatry and Psychosomatic Medicine. The intervention will be given in each of these centers in order to access a larger pool of participants. The PET-scans will be performed at the PET-Center of the Department of Nuclear Medicine at the University Hospital Zurich. The fMRI measures will be performed at the Department for Neuroradiology at the University Hospital Zurich.
The following groups of participants will be included: a) 80 female subjects fulfilling the classification criteria of the American College of Rheumatology for FM (ACR 2011 criteria) (Wolfe, 2010) and without any psychiatric condition according to the ICD-10 and b) 20 healthy women without a history of chronic pain or any mental disorder for the PET-measures; and a total of 15 healthy women for the daily assessments. The groups will be matched on age. All participants will be right-handed and older than 18 years old. Subjects will be excluded if they are pregnant, if they have a history of neurological disorders, current substance or tobacco abuse, current and past substance dependence, schizophrenia spectrum disorder, any other form of chronic pain (except FM for the FM group), and if they have been treated with medication affecting the central DA system in the three months preceding the scanning session (including opioids, neuroleptics, antileptics, antidepressants and lithium). Participating in another RCT or some form of individual or group psychotherapy focusing on pain management is prohibited.
Patients will be recruited from the interdisciplinary outpatient clinic for pain at the University Hospital Zurich, from the Department of Consultation-Liaison Psychiatry and Psychosomatic Medicine at the University Hospital Zurich, from the Zugerklinik, from the Neurology Department and interdisciplinary outpatient clinic for pain at the Canton Hospital Fribourg, from the outpatient pain clinic at the Divisiont of Anesthesiology as well as the service of rheumatology at the Lausanne University Hospital, from private practices, and from adds on websites of the Rheumaliga, as well as patient groups such as the Fibromyalgieforum Schweiz in the German-speaking and French-speaking parts of Switzerland. The healthy controls will be recruited through ads, word of mouth and from our previous and current studies as well. When interested participants call into one of the study telephone lines, they will be told about the study and screened for eligibility using pre-determined scripts. If potentially eligible, an in-person visit (90 min duration) at the study center will be scheduled. This first visit can be the start of the study, or this start can be reported, depending on the clinical situation. At study start, eligibility will be confirmed, consent provided and the initial assessment completed.
The 80 FM participants will be assigned to the MORE treatment or to a wait-list condition in a randomized way (see flow-chart, Fig. 1). The randomization sequence is generated by the website Research Randomizer (http:// www.randomizer.org). The list is concealed by instructed office personnel in the study center of the University Fribourg, Switzerland, who is unaware of any patient information. The randomized participants will join the MORE group or wait-list control. Participants will not be blinded regarding assignment to intervention. Group numbers should not exceed a maximum of 12 per group. As the study progresses, participants coming off the waitlists will also be added to the ongoing groups. We will begin with one MORE group running in Zurich and in Fribourg, but will add additional groups as numbers require.
Description of the MORE Intervention
On the basis of the MORE treatment manual (Garland, 2013), sessions will offer instruction in applying mindfulness and related skills to the following topics: discriminating between nociception, pain, and suffering; gaining awareness of automaticity and coping habits in chronic pain; disrupting the link between negative emotions, catastrophizing, and pain experience through reappraisal; refocusing attention from pain and life stressors to savour pleasant experiences; cultivating self-transcendence and meaning in life; and developing a mindful recovery plan. For this Trial the original MORE session 5 targeting opioid craving will be modified to address mindful self-care strategies. Mindfulness training will involve mindful breathing and body scan techniques, with an emphasis on developing metacognitive awareness and shifting attention from affective to sensory processing of pain. During this process, patients are taught to decompose pain experience into its constituent sensations (e.g., heat, tightness, tingling), as well as to increase awareness of the center, edges, and permeability (versus solidity) of these sensations, and then to notice and savor any adjacent or distal pleasant sensations (Garland, 2020). Sessions will be video-taped for control of therapists’ adherence with the manual. MORE participants will be asked to engage in daily 15-minute mindfulness, reappraisal, and savoring practice sessions at home guided by a MP3 file.
A detailed overview of questionnaires and measures is given in Table 1. Schedule of enrolment, interventions, and assessments is provided in Table 2. Data management including data entry, coding, security and storage will be provided by the web-based application Research Electronic Data Capture (REDCap) http://dx.doi.org/10.5195/jmla.2018.327 that is available at the University of Fribourg. It is Health Insurance Portability and Accountability Act (HIPAA)–compliant and highly secure.
Data to be collected at baseline include age, sex, marital status, education level, professional status.
Measures of functioning includes a measure of quality of life with the World Health Organization quality of life questionnaire WHOQOL Brief questionnaire (Skevington, 2004) and the Fibromyalgia Impact Questionnaire-Revised (FIQ-R; Bennett, 2009) that also includes measure of fatigue. Measures of sleep quality include the medical outcomes study sleep scale (MOS; (Stewart, 1988 Hays, 1992) that has been widely used in rheumatology research and is recommended by OMERACT-10.
Pain related outcomes:
Pain severity and functional interference The Brief pain inventory (BPI) is used to obtain information on self-reported measures of pain severity and functional interference using the Brief pain inventory (BPI, Cleeland, 1994). Pain magnitude is queried by four items that ask about pain now, worst pain, least pain and average pain. Items use numeric rating scales anchored by 0 (no pain) to 10 (most severe pain). Pain interference consists of seven items that ask about how pain interferes with aspects of daily living using numeric rating scales anchored by 0 (no interference) to 10 (completely interferes).
Current pain intensity The one item Verbal Rating Scale (VRS) from the SF-36 (Ware, 1992) is used to measure current pain intensity. The VRS has proven itself over decades as a valid, reliable and change sensitive measure of subjective pain (Ware, 1992).
Pain interference with functional impairment To measure the degree to which pain interferes with function in major life areas, we will use the 7-item Pain Disability Index (PDI) (Dillmann, U. et al.,1994).
Other clinical measures:
Severity of depressive symptoms To assess the severity of depressive symptoms, the Beck Depression Inventory (BDI)-II (Hautzinger, 1991, Beck, 1961), a self-report questionnaire will be used.
State and Trait Anxiety The State-Trait Anxiety Inventory (STAI) will be used as a measure of state and trait anxiety (Laux, 1981).
Mood states The Profile of Mood States (POMS) (McNair et al., 1971) will be used to assess transient, distinct mood states. The POMS measures six different dimensions of mood swings over a period of time with high sensitivity to change.
Quality of Life To measure quality of life, we will use the quality of live scale from the World Health Organization WHOQOL Brief questionnaire consisting of26-items, (Skevington et al. 2004). The WHOQOL Brief has good to excellent psychometric properties of reliability and performs well in preliminary tests of validity (WHOQOL group, 1998).
Function, Impact and Overall symptoms of Fibromyalgia The revised Fibromyalgia Impact Questionnaire (FIQ-R) (Bennet, 2009) is a commonly used and validated 9-item instrument in the evaluation of fibromyalgia (FM) patients.
Sleep quality will be assessed using the Sleep quality, medical outcomes study sleep scale MOS (Stewart, 1988, Hays 1992) that includes 12 items assessing sleep disturbance, sleep adequacy, somnolence, quantity of sleep, snoring, and awakening short of breath or with a headache.
Mechanisms underlying the MORE intervention Mechanisms related to the MORE intervention we will assess with the following questionnaires: Nonreactivity: Five Facet Mindfulness Questionnaire (Baer et al., 2006) the reinterpretation of pain sensations (Subscale of Coping Strategies Questionnaire, (Brown & Nicassio, 1987), positive reappraisal (Subscale of Cognitive Emotion Regulation Questionnaire CERQ) (Garnefski et al., 2007), and savouring (Savouring Beliefs Inventory) (Bryant et al., 2003), according to (Garland et al., 2014). They aim to understand the mechanisms related to the MORE intervention.
PET / fMRI measures
Non-invasive 18 F-DOPA PET and fMRI imaging of the brain (high resolution, anatomical, T1 weighted imaging, resting state will be used to study presynaptic and functional changes in the brain.
PET Data acquisition
PET images will be acquired at the Department of Nuclear Medicine at the University Hospital Zurich. MRI overlay images will be acquired at 3 Tesla at the Department of Neuroradiology at the University Hospital Zurich, using T1 weighted sequence (MP-RAGE) to provide an anatomical framework for image analysis. 18-F DOPA is a well-validated measure of presynaptic DA function (Heiss and Hilker, 2004); and previous studies showed differences in 18-F DOPA binding between FM and healthy participants at rest (Wood, 2007). PET-measures will be performed before and after the MORE intervention (after 8 weeks for the wait-list control group). One hour before scanning they will receive as routinely an oral dose of Carbidopa 100mg, a peripheral DOPA decarboxylase blocker and Entacopone 400mg, a peripheral catechol O-methyltransferase antagonist, and an additional 50mg dose of Carbidopa 30 minutes before 18 F-DOPA injection in order to provide increased availability of 18F-DOPA for striatal uptake. The purpose of the premedication regimen is to limit the metabolism of the 18 F-DOPA tracer by peripheral enzymes, i.e. DOPA decarboxylase and catechol O-methyltransferase, thereby maximizing central uptake (Koopmans et al.2007 ). Subjects will lay quietly on a gurney for 30 minutes before administration of the tracer to allow them to habituate into the environment and to relax. Subjects will then be injected with approximately 2.5mCi of 18F- DOPA. Scans will be acquired on a PET/CT Discovery 690 scanner. Dynamic scanning will start 90 minutes (Koopmans et al. 2007 ) after injection of tracer and continued for 100 minutes.
fMRI data acquisition
The fMRI data acquisition will take place at the Neuroradiology Department at the University Hospital Zurich. To measure brain structure and function, we use a 3.0 Tesla whole-body scanner. The measures include an anatomical scan, a diffusion MRI and a resting state functional MRI and a task based functional MRI. The total time in the scanner will be about 75 minutes. The subjects will perform the Fribourg reward task, an event related fMRI task adapted from the spatial delayed response task to measure neural responses to monetary reward (Gaillard et al. 2019).
Ambulatory assessment measures and physiological measures
To investigate the effects of the MORE intervention on daily affects, stress and reward experiences we will use ambulatory assessment (AA) measures. Primary outcomes related to the AA measures include 1) self-report measures of pain; 2) self-report measures and physiological measures (CAR, alpha-amylase awakening, and daily profile of cortisol and alpha-amylase) of stress, 3) measures of positive affect and 4) measures of reward experience. They will be compared before and after the MORE intervention between both FM groups. AA self-reports of pain, stress and reward and positive experience in daily life will be performed using an adaptation of the Experience Sampling Method (Wilhelm & Schoebi, 2007) to assess the stress and reward experience in the daily living environment. The items related to pain are defined according to a pain diary that we have developed, tested and validated in 50 chronic pain patients at the University of Fribourg. The self-reported items measuring stress, positive and negative affect as well as reward experiences have additionally been adapted and validated in the framework of another research project (Guillod, 2017). All self-assessments are rated on a 7-point Likert scale. The participants will receive an Ipod and perform self-assessments during one week five times a day, during the week preceding or following the PET-measures before and after the MORE intervention. At the end of each day, participants randomly assigned to MORE will record the number of minutes spent engaged in the mindfulness, reappraisal, and savoring practices taught in the MORE intervention to provide a measure of adherence with the home practice associated with the MORE training. Biological measures of stress in everyday life will be assessed by collecting saliva samples 6 times a day (2 for the cortisol awakening response (CAR) and for the salivary alpha-amylase (AA) awakening, 4 for daily profile) for totally 3 days. Saliva samples will be obtained by using the passive drooling method (Salicap, IBL International, Hamburg, Germany). The CAR resp. AA awakening profile and the daily profiles will be used as biological indexes of stress reactivity in everyday life (Wilhelm & Schoebi, 2007).
Primary outcome measures
Effects of the MORE intervention on the DA response to monetary reward by comparing 18F DOPA influx before and after the 8-week MORE intervention
Secondary outcome measures:
2. The percent BOLD signal change in striatal activity during the reward task before and after the 8-week MORE intervention
3. The correlation between daily practice of mindful breathing and savouring (minutes) and increased striatal activity and 18F DOPA influx
4. Changes in pain related outcomes after the MORE intervention
5. Changes in other clinical measures after the MORE intervention
6. Effects of the MORE intervention on daily affect, stress and reward experiences measured with AA measures
State of the art image processing techniques will be used to analyse the PET and the fMRI images. PET image pre-processing will be done using PMOD Version 4.1 or higher (PMOD software, PMOD technologies Zurich) and SPM (Wellcome Department of Imaging Neuroscience, London, UK) for the 18-F DOPA PET study. The most often used method to quantify biochemical function from 18-F DOPA PET image is the multiple time graphical approach (MTGA) that provides rate constants (Ki) for the storage of 18F- DOPA within regions of interest (ROIs) placed over the striatum (Patlak and Blaberg, 1995). Images from each dynamic DOPA PET dataset will be aligned and parametric images of 18F-DOPA influx (Ki) will be created for each subject. The Ki images will be transformed into standard stereotactic space. Regional Ki values of striatal regions (putamen, caudate, nucleus accumbens) will be analyzed with statistical parametric mapping for comparison of regional Ki values on a voxel by voxel basis. Group comparisons will be tested using independent samples T-test; treatment group x time comparisons will be analysed with two factorial ANOVA’s including Ki values obtained in the nucleus accumbens and the caudate to test the primary outcomes. The reward task will be performed in the fMRI scanner and we will correlate 18-FDopa influx with the striatal activation obtained with the fMRI task. In order to measure behavioural differences in the fMRI task results, in the ambulatory assessment data and in the intervention program’s efficacy between the groups tested, we plan to use different kind of variance analyses and multilevel analyses. The number of participants for each part of the study has been determined in order to assure a good statistical power according with this specific research domain (e.g. including potential dropout, outliers). fMRI data analysis will be performed with state-of-the art software freely available online. T1 weighted scans will be analysed with regard to cortical thickness and subcortical volumes using surface based morphometry implemented in FreeSurfer software suite 5.0.1.(http://surfer.nmr.mgh.harvard.edu). This includes a fully automated method. Briefly, this processing includes 1) motion correction, 2) removal of non-brain tissue, 3) automated Talairach transformation, 4) segmentation of the subcortical white matter and deep gray matter volumetric structures (amygdala, hippocampus, thalamus, caudate, putamen, pallidum, nucleus accumbens, ventricles), 5) intensity normalization, 6) tessellation of the gray matter/white matter boundary, 7) automated topology correction, and 8) surface deformation following intensity gradients to optimally place the gray/white and gray/cerebrospinal fluid borders at the location where the greatest shift in intensity defines the transition to the other tissue class. Freesurfer morphometric procedures have been demonstrated to show good test-retest reliability across scanner manufacturers and across field strengths. Obtained subcortical volumes will be normalized by individual intra-cranial volumes for further statistical analyses. Exploratory approaches will involve a vertex-based analysis across the whole brain. Diffusion MRI scans will be analysed for structural connectivity with FMRIB software library 4.1.9 (FSL, http://www.fmrib.ox.ac.uk/fsl). Functional MRI scans will be analysed for activity and functional connectivity with the latest version of Statistical Parametric Mapping (SPM, http://www.fil.ion.ucl.ac.uk/spm). Functional MRI data will be pre-processed according to the following steps: 1) slice timing correction, 2) realignment, 3) linear and non-linear normalization on to a standard EPI template, 4) voxel re-sampling to 2x2x2 mm3 5) smoothing with a Gaussian kernel of 6 mm full width at half maximum, 6) detrending, 7) filtering (such that frequencies 0.01 < f < 0.08 Hz passed the filter), and 8) regressing out the variance of nuisance covariates. With REST toolbox 1.6, for each subject and each specified region of interest (atlas-based specification), mean signal time courses will be extracted and cross-correlated. Next, correlations will be r-to-z transformed for group-level statistics. These z-values can be used for comparisons across (cross-sectional) and within (longitudinal) groups or for parametric correlations with psychometric measures. The main analyses of the measures related to the primary outcomes of the clinical effects of MORE, including measures of pain intensity and severity, measures of functioning, measures of sleep, and measures of mood outcomes will be performed with two-factorial (time x treatment group) ANCOVA controlling for pain score intensity at pre-test according to (Garland et al., 2014) as well as one ANCOVA controlling for depressive symptoms at pre-test. Measures related to the primary outcomes of the clinical effects of MORE, including measures of pain intensity and severity, measures of functioning, measures of sleep, and measures of mood will be performed with ANOVA’s. In case of not normally distributed data, we will use a transformation in Z-scores, rather than using non-parametric tests. We will perform per protocol analyses to test our hypotheses. Intent-to-treat analyses will be used for sensitivity analyses (Garland et al., 2014). To account for missing data we will use multiple imputation procedures.
Analyses of the secondary outcomes related to the treatment mechanisms will be performed using two-factorial (treatment group x time) ANOVAs to test the pre-post-test differences. The AA measures yield intensive longitudinal data that are clustered, as they represent series of measurements that stem from different individuals. A multilevel approach to analyse these data takes into account clustering and can therefore accommodate these data and provides flexible tools to investigate within-subject phenomena, such as responses to stressors or rewarding experiences (see Bolger & Laurenceau, 2013). We will use a software that allows for the simultaneous modelling of within-subject and between-subject aspects of the data, and the examination of associations among individual difference variables, and individual differences in within-subject parameters (e.g., Mplus 7.3, (Muthen & Muthen, 1998–2012). The sample size of this study is based on: 1) our previous research (Ledermann et al. 2016) that showed large effect sizes (Cohen’s d 1.1; F = 1.3) in the comparison between our healthy controls (N = 17) and FM participants without depression (N = 17) for the differences in DA binding in response to unpredictable reward. 2) Earlier studies investigating neurophysiological changes in MORE with EEG that reported significant changes (F1,25 = 4.9; ηpartial2 = .17) in response to reward before and after the MORE intervention in chronic pain patients with sample sizes of N = 11 (active group) and N = 18 (support group). Using this effect size, a power analysis with G*Power (Faul et al. 2007) showed that 46 participants are necessary for an ANOVA with 2 factors and repeated measures to be significant. Compared to Garland et al. 2015 we will use a waiting-list design in this study and not an active intervention as control, what might increase the effect size of the results. In addition, results obtained with EEG cannot be directly compared to PET measures. For these reasons, we interpolate the sample size between the different estimations obtained that the sample size after the intervention should be between N = 12 based on Backman et al. 2011 and N = 46 based on Garland et al. 2015, taking into account that a strict minimum of participants should be exposed to radioactivity for the PET measures. With regard to the clinical measures, the RCT by Garland et al. 2014 evidenced significant reductions in pain severity with medium effect size (Cohen’s d = 0.5) after the MORE intervention in a starting group of 115 participants and of 69 at post-intervention measures (N = 31 for MORE and N = 38 for support group). Based on these estimates, an ANCOVA to be significant should have a minimal sample of N = 25 participants after treatment in each group. Finally, with regard to the daily measures of positive affect, pain and stress, Garland et al. 2017 found significant effects associated with the MORE intervention compared to a support group therapy in a group of N = 55 (MORE. N = 26, Support Group: N = 29) using multilevel analyses. In order to be able to do group comparisons with the AA measures, we will add 15 healthy participants who will only do the AA part. On the basis of these analyses, we consider that an estimation of 30 participant in each condition (waiting-list versus MORE) after treatment should be sufficient to obtain significant results for our aims. In their randomized clinical trial of the MORE intervention in chronic pain patients with opioid abuse, Garland et al. 2014 found 20% drop-out between beginning and completion of the intervention. We will therefore postulate 20% drop-out in total, giving a starting sample of 38 participants in each group, that we will round up to 40 to account for data loss related to technical problems, giving a total sample of 80.