The course of pain hypersensitivity according to painDETECT in patients with rheumatoid arthritis initiating treatment: results from the prospective FRAME-cohort study

The course of pain hypersensitivity according to painDETECT in patients with rheumatoid arthritis... Background: Evidence is emerging that pain in rheumatoid arthritis (RA) exists without underlying inflammation. Our objective was to evaluate the prognostic value of pain classification at treatment initiation using the painDETECT questionnaire (PDQ). Outcomes were change in DAS28-CRP and RAMRIS synovitis score. Methods: RA patients initiating a disease-modifying anti-rheumatic drug (DMARD) or initiating/ switching a biological agent were included. Follow-up time was 4 months. Clinical examination, imaging (MRI, dynamic contrast-enhanced MRI (DCE-MRI)), and patient-reported outcomes were undertaken. The PDQ was used to differentiate pain mechanisms. Mean change (95% CI) was calculated using ANCOVA. Multivariable regression models were used to determine a prognostic value. Results: A total of 102 patients were included; 75 were enrolled for MRI. Mean changes in baseline variables were greatest in the high PDQ classification group (> 18), while limited in the intermediate group (13–18). The 12 patients with high baseline PDQ score all changed pain classification group. No prognostic value of PDQ pain classification was found in relation to change of DAS28-CRP, RAMRIS score, or VAS pain. In the unadjusted model, RAMRIS score at baseline was associated with change in DAS28-CRP. The exploratory variables of DCE-MRI did not differ from other inflammatory variables. Conclusions: In RA patients a high PDQ score (non-nociceptive pain) at baseline was not associated with worse outcomes, in fact these patients had numerically greater improvement in DAS28-CRP. However, pain classification by PDQ was not independently associated with change in DAS28-CRP, RAMRIS score, or VAS pain in the prognostic models. Furthermore, patients classified with a high baseline PDQ score changed pain classification group. Patients with unclear pain mechanism had reduced numerically treatment response. Trial registration: The study was approved by the Regional Ethics Committee of the Capital of Denmark April 18 2013; identification number H-3-2013-049. Keywords: Rheumatoid arthritis, Central sensitization, painDETECT questionnaire, Prognostics, Dynamic contrast- enhanced magnetic resonance imaging (DCE-MRI) * Correspondence: signe.rifbjerg-madsen.02@regionh.dk The Parker Institute, Copenhagen University Hospital, Bispebjerg and Frederiksberg, 2000 Frederiksberg, Copenhagen, Denmark Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 2 of 11 Background In this study, we hypothesized that a high PDQ score Pain in rheumatoid arthritis (RA) has typically been would serve as an indicator of central sensitization and regarded as nociceptive, that is, related to ongoing periph- thus a prognostic factor for a poorer treatment outcome eral inflammation [1]. However, during the last decade, (DAS28-CRP change) in patients with RA initiating or where focus has been on early diagnosis and aggressive intensifying anti-inflammatory treatment. A possible stat- treatment strategies in the treat-to-target regime [2], it has istical interaction between central sensitization (high PDQ become clearer that in a subgroup of RA patients, pain score) and inflammatory load (baseline synovitis defined can become an entity in its own right, probably elicited by hand MRI RAMRIS score) was considered as part of by, but not directly related to, ongoing inflammation [3]. the hypotheses. In the exploratory part of the study, we A substantial proportion of RA patients in stable clinical hypothesized that DCE-MRI would capture change in remission continue to report moderate to severe pain inflammation and thus a possible relation to inflammatory levels [4] and studies have indicated that RA leads to pain mechanisms (low PDQ score) better than conven- widespread pain in 10–20% of patients [5]. Such observa- tional MRI. tions have led to the contention that changes in the peripheral and central nervous system through processes Methods of neural plasticity and central sensitization may play an Design important role [6]. As a rule, sensitization phenomena The Frederiksberg Hospital’s Rheumatoid Arthritis, pain would be expected to extinguish as the tissue heals and assessment and Medical Evaluation (FRAME)-cohort inflammation subsides. However, a state of induced hyper- study was conducted according to a published protocol, sensitivity of the pain system may persist in subsets of which contained a detailed description of the methods patients and lead to chronic pain states in which pain is and prespecified analysis [19, 20]. It was approved by the no longer coupled to the presence of ongoing peripheral Regional Ethics Committee of the Capital of Denmark; inflammation [7]. In such patients, persistent pain hyper- identification number H-3-2013-049. sensitivity may lead to continuous high reports of tender RA patients were recruited from departments and pri- joints and poor global health; subcomponents of the com- vate clinics of rheumatology in the Copenhagen area and monly used composite disease activity score of 28 joints prospectively enrolled from March 2013 to September (DAS28-CRP) and thus overestimation of inflammatory 2014. MRI was included in the examination program from activity. If the anti-inflammatory treatment is intensified May 2013. The examination program was conducted at on this background, little change in DAS28-CRP can be Frederiksberg Hospital. Patients were assessed at treat- expected. Conversely, if inflammatory RA is left un- or ment initiation (baseline) and after 4 months of treatment. not sufficiently treated it will lead to joint destruction and Patients received routine care at the discretion of their loss of function [8]. Identification of underlying pain rheumatologist during the trial period. Add-on of pain- mechanisms therefore has potential importance when killers was allowed. prognosticating the effect of medical treatment on inflam- mation and pain. Patients The painDETECT questionnaire (PDQ) is a self- To be eligible, patients had to fulfil either the 1987 [21]or administered pain classification instrument originally 2010 ACR RA criteria [8]and be ≥18 years. Further, developed to differentiate neuropathic (non-nociceptive) patients had to be scheduled for either (a) treatment initi- from non-neuropathic (nociceptive) pain [9]. It has been ation with any conventional synthetic disease-modifying increasingly used in patients with osteoarthritis and antirheumatic drug (csDMARD) (patients who had not fibromyalgia to assess clinical pain features indicative of received treatment with csDMARD for at least 6 months central sensitization [10–12] and has recently been including newly diagnosed/treatment-naïve patients) or introduced in studies assessing pain mechanisms in (b) treatment initiation or change of any biologic DMARD patients with RA and spondyloarthritis [13–15]. (bDMARD). Magnetic resonance imaging (MRI) is an objective and Major exclusion criteria were intra-articular or intra- sensitive method to assess joint inflammation. The most muscular glucocorticoids administered less than 3 weeks common scoring system in the wrist and metacarpophalan- prior to baseline; treatment with oral corticosteroids at geal (MCP) joints is the OMERACT (outcome measures in doses equivalent to more than 10 mg prednisolone/day rheumatoid arthritis clinical trials) RA MRI scoring (RAM- within the 3 weeks prior to baseline; inability to pause RIS) system [16]. Dynamic contrast-enhanced (DCE) MRI antidepressants, anticonvulsants or other centrally acting is a technique where the sequences are acquired rapidly analgesics; initiation of csDMARD therapy more than and sequentially before and during contrast infusion. DCE- 3 weeks prior to the baseline visit (only patients initiating MRI has been shown to correlate better than conventional csDMARDs); treatment with bDMARD initiated more MRI with the histologic findings of synovitis [17, 18]. than 1 week prior to the baseline visit (only patients Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 3 of 11 initiating a bDMARD). Patients who had contraindica- As standard research procedure a target hand was tions for MRI were excluded from the MRI arm of the chosen for MRI to reflect the general level of joint inflam- study. Furthermore, patients with increased risk of neuro- mation. The most painful hand as reported by the patient pathic pain conditions (e.g. diabetes) were excluded due to was chosen, or, in case of no difference in pain level, the the potential to confound the pain assessment. dominant hand. The examination was carried out in a 3 T Siemens Verio MR scanner according to a published Variables and outcome measures scanning procedure [20]. Conventional coronal and axial The patients underwent an examination program at STIR and 3D coronal T1w GRE VIBE pre- and post- baseline and follow-up, collecting information on contrast images were used for RAMRIS scoring. The wrist demographics and medication and from patient-reported and MCP joints 2–5 were assessed according to the outcomes (PROs). Clinical examination including joint OMERACT RAMRIS [16, 28] and were scored for count and tender point examination conducted by the synovitis (0–3; total score 0–21) and bone marrow edema same assessor at both time points, imaging (MRI and (BMO) (0–3; total score 0–58). Based on previous reports DCE- MRI), and standard blood samples (CRP, immuno- on smallest detectable difference, it was decided that the globulin M-rheumatoid factor [IgM-RF], anti-cyclic RAMRIS synovitis score had to alter by more than 1 unit citrullinated peptide [anti-CCP]) were also performed. to be considered a significant change [29]. All images The following PROs were collected from each patient; were assessed blinded and paired by the same senior the PDQ, the Stanford health assessment questionnaire radiologist (MB). disability index (HAQ-DI), the 36-item short form health For the explorative DCE-MRI analyses the software survey (SF-36), generalized anxiety disorder assessment DYNAMIKA enterprise version 3.2.6 (http://www.ia-grp. (GAD-10) and major depression inventory (MDI). com) was used according to a published procedure [30]. The PDQ is a patient-administered pain classification Only joints with MRI signs of inflammation (“focus tool that was developed in a population of patients with joints”) were included. All images were analyzed paired various pain conditions. It has been further validated for by the same physician (SRM). Regions of interest (ROIs) describing pain phenotypical features in patients with were drawn on all slices where sign of inflammation was inflammatory arthritis by our group [22] and is validated present and collapsed into one volume of interest (VOI) for use on touch screen [9, 23]. It comprises items on pain for each focus joint; wrist and 2nd-5th MCP. It was intensity (three numeric rating scales not included in the decided to include tenosynovitis and capsulitis as ‘signs total score), pain course patterns, pain radiation (from a of inflammation’ in the analyses. Joints with no signal pain drawing) and seven somatosensory signs and symp- were assigned a score of 0. Nvoxel, IRExNvoxel, MExN- toms (rated on a six-category Likert scale). According to a voxel and IRExME were chosen as outcome measures validated algorithm, patients were assigned to one of three [17, 31–35]. The number of enhancing voxels (Nvoxel) pain classification-groups based on a score between − 1 was multiplied by the volume of each voxel in milliliter and 38: > 18 likely neuropathic pain, 13–18 unclear pain (ml) to adjust for different image sizes. The initial rate of mechanism or < 13 likely non-neuropathic pain [9]. Sev- enhancement (IRE) and maximum enhancement (ME) eral studies have used it as indicator of non-nociceptive or represent the degree of perfusion; the IRE reflects the central pain mechanisms [10–12]. initial rate of enhancement of the time intensity curve. HAQ-DI is a measure of limitation of activities of daily ME represents the equilibrium state of the curve and living used for patients with RA. It assesses the patient’s reflects the amount of contrast passing into the ROI. ability to carry out everyday tasks. It includes a visual The composite outcome measures IRExNvoxel and MExN- analogue scale (VAS) evaluation of pain, fatigue and glo- voxel reflect both the volume and degree of perfusion, bal health (GH) [24].The SF-36 assesses eight domains whereas IRExME characterizes the perfusion profile of the concerning general health, which can be summarized voxels derived from the time-intensity curves. into a physical (PCS) and mental (MCS) component summary score. In this study the Danish version of SF- Statistical analysis 36, which uses a 4-week recall period was applied [25]. SAS software (version 9.4, SAS Enterprise Guide 7.1, SAS The GAD-10 is a ten-item instrument developed from Institute Inc., Cary, NC, USA) was used for all statistical the Hamilton six-item anxiety scale. It measures general- analyses. PROC UNIVARIATE statement was used to ized anxiety by scoring the total sum of the items [26]. summarize thedataand for visualinspectionofnormality. The MDI is a questionnaire based on self-reported mood Means (with standard deviations [SDs]) or medians (with symptoms. It holds the ability to generate a DSM-IV and interquartile ranges [IQRs]) were reported and compared International Classification of Diseases (ICD)-10 diagno- by t test or Kruskal-Wallis (Wilcoxon) test, respectively. ses of major (moderate to severe) depression and to rate Delta changes were adjusted for baseline value and com- the severity of symptoms [27]. pared using analysis of covariance (ANCOVA). All analyses Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 4 of 11 were carried out according to the intention-to-treat csDMARD and 54 initiating bDMARD. Of these, 75 principle, i.e. missing data at follow-up was imputed from patients completed the PDQ and had an MRI scan baseline (baseline observation carried forward). A two- performed at baseline and of these, 71 patients had an sided p value less than 0.05 was regarded as being statisti- MRI scan performed at follow-up. This discrepancy was cally significant. primarily caused by administrative delay resulting in no Prior to executing the FRAME-cohort study, a power patients receiving MRI-scan during the first 2 months of calculation was performed based on the assumption that the study period. Exclusion and reasons for dropout are it was feasible to include 100 RA patients during a study further described in the figure text. Due to 3 patients not period of 1½ years [20], as no data for sample size calcula- wishing to receive contrast, 72 patients with complete tion was available. Anticipating a common SD of 1.5 and PDQ were included in the exploratory DCE-MRI study at the correlation between pre- and post-scores being r =0.3 baseline (not shown in Fig. 1). Of these72 patients, in all for a paired t test with a significance level of 0.05, a sample 65 completed the follow-up scan with contrast. of a 100 pairs has a power of 80% (0.797) to detect a mean Baseline characteristics are described in Table 1 (some change of 0.5 DAS28-CRP units. A patient population, data has previously been published [36]). The distribu- who can expect a change of this magnitude in their tion of patients across the PDQ classification groups disease activity, is a reasonable cohort in which to study were n = 66, n = 23, n = 12 for PDQ score < 13, 13–18, > prognostic factors of treatment response. However, this 18, respectively. Statistically significant differences across number was not reached for the MRI subsample. the three PDQ classification groups were observed for The prognostic value of the PDQ score, RAMRIS score, tender joint count (TJC), tender point (TP) count, and their interaction at baseline in relation to change of DAS28-CRP, physical function (HAQ-DI), VAS-fatigue, DAS28-CRP was examined by multivariable regression VAS-pain, VAS-GH, anxiety (GAD-10), depression models using the SAS PROC GLM. As the interpretation (MDI), SF-36 PCS, and SF-36 MCS. A TP count ≥11 of PDQ by nature is trichotomous, the results were was found for higher proportions of patients with higher expressed as least squares means per category. According PDQ classification groups; however, this was not statisti- to the protocol, the model was adjusted for the following cally significant. No differences were found across the prespecified confounders: age (years), sex (male/female), groups for the biochemical and imaging variables. disease duration (month), disease activity (DAS28-CRP at Changes from baseline stratified by PDQ group are baseline), group (csDMARD/bDMARD), antiCCP-positive presented in Table 2. Change in PDQ classification group is (yes/no) and concomitant prednisolone (yes/no). Subse- reported as classification consistency, i.e. number of patient quently, in the fully adjusted model covariates (i.e. possible that did not change classification group. Statistically signifi- confounders) that did not contribute to the model were cant differences in classification consistency were found for removed; age, antiCCP-positive (yes/no) and concomitant the low, intermediate, and high PDQ group; 97%, 83%, and prednisolone (yes/no). 0%, respectively. For all clinical variables and PROs, a U- Secondary outcomes were change in RAMRIS synovitis shaped curve change pattern was observed with greater score and VAS pain. Post hoc, to ensure robustness of change in the high and low PDQ classification group than results, a sensitivity analysis including baseline VAS pain in the intermediate. Remission at follow-up according to as a confounder in the adjusted analysis of change in VAS DAS28-CRP (< 2.6) was found in 41%, 22%, and 42%, pain was performed. respectively. The change in imaging variables did not On an exploratory basis the DCE-MRI variables IRExN- display a distinct pattern. Statistically significant differences voxel (in ml) or MExNvoxel (in ml) for wrist were applied between the classification groups were found for delta in the primary model examining DAS28-CRP and VAS change of CRP, DAS28-CRP, VAS GH, HAQ-DI, and SF36- pain change, replacing the RAMRIS synovitis score. For MCS. However for CRP, no differences between PDQ SRM, inter- and intra-reader agreements, intraclass correl- score < 13 and > 18, and 13–18 and > 18 were found. For ation coefficients (ICC) (absolute agreement) for the four DA28-CRP and VAS GH no differences were found predefined DCE-MRI variables were tested beforehand on between PDQ score < 13 and PDQ score > 18. Regarding data from ten patients using SPSS software. Wrist and HAQ-DI and SF36-MCS there were no difference between MCP joints were tested separately. PDQ score < 13 and 13–18. DCE-MRI variables are presented separately in Table 3. There were neither signifi- Results cant differences nor trends in the DCE-MRI variables Figure 1 illustrates the flow of patients. In all, 151 patients across the groups; however, the variables IRExNvoxel (in fulfilled the inclusion criteria. Of these, 48 patients were ml) and MExNvoxels (in ml) for the wrist also displayed excluded. In total, 103 patients received a baseline assess- the U-shaped change pattern in line with the RAMRIS ment; however two were excluded post hoc. In total, 101 score (BME). Baseline characteristics and delta change by patients completed the PDQ at baseline; 47 initiating initiation group have previously been reported [36, 37]. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 5 of 11 Fig. 1 Flowchart of participants. 1: Refrained from participation (n = 18), comorbidity with risk of neuropathic pain (n = 10), unable to pause antidepressants, anticonvulsants, or other centrally acting analgesics for 1 week (n = 7), initiated DMARD treatment > 3 weeks ago (n = 5), received more than 10 mg prednisolone < 3 weeks ago (n = 4), other reasons (n = 4). 2: Im. Corticosteroid between screening and baseline assessment. 3: MRI not feasible or patients declined. 4: Four patients withdrew from the study; two from each treatment group, three from the MRI subgroup. A further one patient did not participate in the follow-up MRI In the multivariable regression models (Table 4) Discussion change in the predefined outcome variables was To ourknowledge,thisisthe first study to evaluate expressed as least square means. No interaction was the prognostic value of pain classification by the PDQ found between PDQ classification group and RAMRIS score in relation to change in DAS28-CRP, VAS pain, synovitis neither in the protocolized analysis, nor and RAMRIS score in RA patients initiating or escal- between PDQ classification group and the DCE-MRI ating anti-inflammatory treatment. It was hypothe- variables (IRExNvoxel (ml), MExNvoxel (ml)) in the sized that patients in the high PDQ classification exploratory analysis. All interaction statements were group (score > 18) would display constant high TJC therefore left out of the models. No prognostic value and VAS GH as features of persistent pain hypersen- of PDQ classification was found for any of the sitivity (central sensitization) and thus gain little or outcomes in any of the models, sensitivity analysis no change of DAS28-CRP following treatment initi- included (p = 0.44) (data for DCE-MRI is presented as ation. In contrast however, this was not confirmed in additional material, Additional file 1). the multiple regression analysis, we found that these Baseline DAS28 was found to be statistically patients (n = 12) experienced the greatest numerical significantly associated with change in all outcomes change in DAS28-CRP, self-reported disease severity in both adjusted models; DAS28, RAMRIS, and VAS measures, and objective inflammatory parameters, pain (p < 0.01), sensitivity analysis excluded. In this including MRI (Table 2). This result was seen despite analysis, baseline VAS pain was found to be signifi- the fact that baseline inflammatory parameters were cantly associated with change in VAS pain (p <0. thesameacrossthe threePDQ classification groups, 001). Baseline RAMRIS synovitis score was found to and that other baseline characteristics of the high be statistically significantly positively associated with PDQ classification group clinically could indicate DAS28-CRP change in the unadjusted model (p =0. presence of central sensitization (high frequency of 01) and of RAMRIS change in both the unadjusted female sex, a high number of tender joints and tender and the adjusted model (p ≤ 0.02). points, poor mental well-being, a high disability index, In the exploratory DCE-MRI study inter and intra- and high VAS scores). reader reliability showed good to excellent agreement In contrast to the patients in the low and intermedi- (data is shown in Additional file 2). ate PDQ classification -group, all patients in the high Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 6 of 11 Table 1 Baseline characteristics stratified by PDQ group PDQ score 7 (5-9) PDQ score 16 (13-18) PDQ score ≥ 19 p value (n = 66) (n = 23) (n = 12) PDQ score 7 (5–9) 16 (13–18) 21 (20–23) <.0001 Female, n (%) 45 (68.2) 20 (87.0) 31 (91.7) 0.09 Age, years (SD) 56.1 (14.7) 54.6 (18.9) 47.4 (15.6) 0.22 Initiated csDMARD, n (%) 37 (78.7) 6 (12.8) 4 (8.5) Disease duration, months 15.5 (1–104) 53 (9–47) 34.5 (24–149.5) 0.12 Current smoker, n (%) 11 (16.9) 7 (30.4) 2 (16.7) 0.39 Corticosteroid usage, n (%) 11 (16.7) 2 (8.7) 3 (25.0) 0.42 28 Swollen joint count 2 (1–7) 5 (2–8) 2 (1–4) 0.09 28 Tender joint count 5 (3–10) 13 (8–16) 9.5 (6.5–15) <.0001 Tender point count, 0–18 6 (4–14) 10 (7–16) 12 (6–14) 0.02 Tender point count ≥11, n (%) 20 (30.3) 11 (47.8) 7 (58.3) 0.09 DAS28 (SD) 4.2 (1.1) 5.0 (1.2) 4.8 (0.7) 0.007† HAQ-DI, 0–3 0.75 (0.38–1.25) 1.13 (0.88–1.75) 1.63 (1.19–1.88) 0.0004 VAS-fatigue, mm 53.5 (27–71) 75 (52–89) 72.5 (60.5–87) 0.002 VAS-pain, mm 42 (24–60) 69 (50–82) 63 (45.5–80.5) 0.0003 VAS-global health, mm 56 (32–77) 73 (46–85) 78 (61.5–90) 0.03‡ GAD-10 score, 0–50 6 (3–10) 10 (6–17) 9 (9–12) 0.008 MDI score, 0–50 8 (4–12) 11 (6–25) 13.5 (9–19.5) 0.01 SF-36 PCS, 0–100 35 (29–42) 33 (26–37) 28 (24–32) 0.02‡ SF-36 MCS, 0–100 51 (40–57) 43 (31–51) 37 (31–49) 0.005 CRP, mg/mL 8 (3–15) 4 (0.5–19) 3 (0.8–8.5) 0.13 IgM-RF positive, n (%) 41 (62.1) 15 (65.2) 8 (66.7) 1.0 Anti-CCP positive, n (%) 46 (70) 15 (65) 7 (58) 0.72 RAMRIS hand synovitis 7 (5–10) 9.5 (7–11) 7 (6–9) 0.18 RAMRIS hand edema 5 (1–10) 9.5 (3–22) 5 (2–13) 0.12 Values are median (25th, 75th percentiles) unless specified otherwise Unless specified otherwise significant p values reflect difference between PDQ score < 13 and 13–18 and ≥ 19, while there is no difference between PDQ score 13–18 and ≥ 19 Only difference between PDQ score < 13 and 13–18 Only difference between PDQ score < 13 and ≥ 19 RAMRIS hand; wrist+MCP scores One patient receiving MRI had no corresponding PDQ score. 26 patients did not receive MRI scan. Number of patients (n) for RAMRIS parameters: PDQ score < 13, n = 51; PDQ score 13–18, n = 18; PDQ score > 18, n = 6. (a) One missing observation. (b) Three missing observations PDQ classification group changed pain classification pattern across the PDQ classification groups, indicating following treatment initiation or escalation. These find- that the uncoupling of pain mechanism from present ings could indicate that in these particular patients, a inflammation could be found in this group. high PDQ score might not have marked central pain In the multivariable regression analyses, the PDQ did mechanisms uncoupled to ongoing inflammation, but not have prognostic value in relation to change of DAS28- could have reflected reversible inflammatory driven CRP, RAMRIS score, or VAS pain. No interaction between pain hypersensitivity, i.e. normal neuroplasticity. However, PDQ score and baseline RAMRIS synovitis score was regression toward the mean needs to be taken into found in any of the regression analyses including the account. Interestingly, the observed numerically changes exploratory DCE-MRI analyses indicating no relation of the variables were smallest in the intermediate PDQ between ongoing inflammation and pain phenotype (pain group (score 13–18), total hand RAMRIS synovitis score classification group). However, in the light of the relatively excluded. Thus, patients with an unclear pain mechanistic small sub-sample of patients with a high PDQ score at background had the poorest response to medical treat- baseline, it does not seem reasonable to reject any prog- ment also reflected in the U-shaped DAS28 remission nostic value of the PDQ at this point. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 7 of 11 Table 2 Mean change stratified by baseline PDQ group adjusted for baseline value PDQ score < 13 PDQ score 13–18 PDQ score ≥ 19 p value (n = 66) (n = 23) (n = 12) PDQ consistency, N (%) 64 (97) 19 (83) 0 (0) < 0.001 Δ PDQ score −4.10 (−5.74;-2.45) −2.20 (− 4.64;0.25) −5.49(−9.97;-1.02) 0.08 Δ 28 SJC −2.71 (− 3.26;-2.17) − 2.27 (− 3.21;-1.34) − 2.48(− 3.77;-1.19) 0.71 Δ 28 TJC − 4.63(− 5.67;-3.60) −4.36(− 5.14; − 1.58) −5.57(−7.96;-3.18) 0.28 Δ CRP, mg/mL −6.93 (− 8.54;-5.32) −2.81 (− 5.54;-0.08) −6.28 (− 10.10;- 2.47) 0.04a Δ DAS28-CRP − 1.47 (− 1.74; − 1.20) −0.8(− 1.27;-0.34) −1.62 (− 2.25;-1.00) 0.03b Δ Tender point count −2.70 (− 3.87;-1.57) −1.56 (− 3.48; 0.36) − 3.83 (− 6.46;-1.20) 0.35 Δ VAS-fatigue −18.31(− 25.55;-13.06) −8.19 (− 17.07; 0.68) −22.20 (− 34.48;-9.91) 0.09 Δ VAS-pain −23.27(− 28.21;-18.33) −12.99(− 21.40;-4.57) −26.86 (− 38.34;-15.38) 0.07 Δ VAS-global health −26.20(− 31.15;-21.25) −13.49 (− 21.83;-5.15) −31.72 (− 43.44;-20.01) 0.01b Δ HAQ-DI − 0.34 (− 0.44;-0.24) − 0.15 (− 0.32; 0.02) −0.74 (− 0.98;-0.50) < 0.001c Δ MDI-total − 3.04(− 4.39;-1.68) −2.54 (− 4.83;-0.25) −3.68(− 6.80;-0.55) 0.84 Δ GAD10-total −2.24 (− 3.18;-1.30) −1.56 (− 3.15; 0.02) −2.10 (− 4.27; 0.07) 0.76 Δ SF36-MCS 5.5 (3.34; 7.60) 3.70 (0.09; 7.31) 11.38 (6.40;16.36) 0.04c Δ SF36-PCS 6.47 (4.39; 8.56) 4.94 (1.43; 8.44) 8.96 (3.97; 13.94) 0.42 Δ RAMRIS H synovitis −1.69 (− 2.32;-1.07) − 1.81(− 2.90,-0.71) −2.47 (− 4.28;-0.66) 0.72 Δ RAMRIS H edema −2.17 (− 4.00;-0.34) − 0.57(− 3.80;2.66) − 3.46 (− 8.72;1.81) 0.58 Values are means (95% CI). Negative numbers indicate improvement. ANCOVA (BOCF) was used for the analyses unless otherwise indicated. †Chi-square test RAMRIS H hand (W + M). W wrist. M MCP joints (n) for RAMRIS parameters: PDQ score < 13; n = 51, PDQ score 13–18; n = 18, PDQ score > 18; n = 6. (a) No difference between PDQ score < 13 and > 18, and 13–18 and > 18. (b) No difference between PDQ score < 13 and PDQ score > 18. (c) No difference between PDQ score < 13 and 13–18 Table 3 Baseline values and mean changes (adjusted for baseline value) for the exploratory DCE-MRI variables PDQ score < 13 PDQ score 13–18 PDQ score ≥ 19 p value Baseline values Nvoxel in ml wrist 13.17 (5.53–29.40) 14.78(7.22–25.84) 14.21(8.28–22.70) 0.93 Nvoxel in ml MCP 1.65 (0.47–6.57) 3.12(0.80–9.22) 1.69(0.00–3.91) 0.56 IRE x Nvoxel in ml wrist 0.15 (0.05–0.67) 0.25(0.04–0.67) 0.19(0.06–0.56) 0.91 IRE x Nvoxel in ml MCP 0.02 (0.002–0.15) 0.06(0.003–0.24) 0.006(0.00–0.07) 0.31 ME x Nvoxel in ml wrist 21.43(9.76–62.29) 26.65(10.09–56.40) 26.73(14.72–42.67) 0.92 ME x Nvoxel in ml MCP 2.30 (0.67–14.23) 6.70(0.91–21.34) 2.46(0.00–7.42) 0.50 IRE x ME wrist 0.02 (0.01–0.05) 0.03(0.01–0.06) 0.02(0.01–0.05) 0.87 IRE x ME MCP 0.03 (0.01–0.09) 0.04(0.01–0.12) 0.01(0.00–0.05) 0.21 Mean changes Δ Nvoxel in ml wrist − 2.22 (− 4.84;0.41) − 2.47(− 6.93;1.99) − 2.87(− 10.37;4.63) 0.98 Δ Nvoxel in ml MCP − 1.99 (− 3.02;-0.96) − 1.72(− 3.47;0.03) −0.79 (− 3.74;2.17) 0.74 Δ IRE x Nvoxel in ml wrist −0.18(− 0.28;-0.07) −0.11(− 0.29;0.07) −0.17(− 0.46;0.13) 0.80 Δ IRE x Nvoxel in ml MCP −0.07 (− 0.11;-0.03) −0.06(− 0.13;0.003) −0.06(− 0.17;0.06) 0.99 Δ ME x Nvoxel in ml wrist −7.72 (− 14.69;-0.76) −5.55(− 17.37;6.27) −7.90(− 27.80;12.00) 0.95 Δ ME x Nvoxel in ml MCP −5.17(− 7.42;-2.92) − 4.82(− 8.64;-1.00) −3.12(− 9.58;3.33) 0.84 Δ IRE x ME wrist −0.01 (− 0.02;− 0.01) −0.01(− 0.02;0.003) -0.01(− 0.03;0.01) 0.81 Δ IRE x ME MCP − 0.03(− 0.04;-0.02) −0.03(− 0.06;-0.01) −0.03(− 0.07;0.01) 0.96 Baseline values are medians (25th, 75th percentiles) PDQ score < 13, n = 49; PDQ score 13–18, n = 17; PDQ score > 18: n =6 Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 8 of 11 Table 4 Multivariable regression models examining change expressed as least squares means (95% CI) Protocolized models PDQ score < 13 PDQ score 13–18 PDQ score > 18 p value Unadjusted model ΔDAS28 −1.30(− 1.64;-0.97) −1.17(− 1.74;-0.60) − 2.08(− 3.06;-1.10) 0.27 ΔRAMRIS Synovitis −1.68(− 2.30;-1.06) −1.75(− 2.80;-0.70) −2.46(− 4.25;-0.66) 0.72 ΔVAS pain −17.57(− 24.13;-11.02) − 25.45(− 36.56;-14.33) −38.13(− 57.15;-19.11) 0.09 Adjusted ΔDAS28 −1.49 (− 1.81;-1.18) −1.16(− 1.70;-0.61) −2.06(− 2.94;-1.17) 0.17 ΔRAMRIS Synovitis −1.78(− 2.44;-1.26) −1.72(− 2.86;-0.57) −2.48(− 4.34;-0.63) 0.74 ΔVAS pain − 19.36(− 25.89;-12.83) −24.71(− 36.07;-13.35) −38.04(− 56.50;-19.56) 0.15 No interaction was found in any of the models. The protocolized unadjusted models included the baseline PDQ classification groups as a trichotomous variable and RAMRIS synovitis score for the hand. The adjusted models further included female sex, disease duration, initiation group, and DAS28-CRP as covariates Protocolized models, n: PDQ score < 13; n = 51, PDQ score 13–18; n = 18, PDQ score > 18; n =6 In contrast, DAS28-CRP at baseline was a significant with notably the subgroup of patients with a high baseline predictor (p < 0.01) of change in the dependent variable PDQ score was small (n = 12). (DAS28-CRP, RAMRIS synovitis score, VAS pain) in all Inflammation was not only assessed clinically but also adjusted models. Furthermore, RAMRIS synovitis by MRI of one hand to ensure objectivity. Assessing one predicted DAS28-CRP change in the unadjusted model (p hand as a target joint area to reflect general inflammation = 0.01), and RAMRIS synovitis change both in the is a common procedure but can be a limitation to the re- unadjusted and the adjusted model (p ≤ 0.02). The impact gression analysis as it involves a risk of misestimating the of RAMRIS synovitis score on clinical change is not well inflammatory load especially in patients with a non-typical documented [38, 39]. Although our group previously has RA presentation. However, it has not yet been clarified found correlations between RAMRIS synovitis score and how many joints that should be included in a prognostic DAS28-CRP in another cohort of RA patients (in review), image analysis to reflect overall inflammatory load. this may be the first study to indicate a possible prognostic The study was limited by the framework of the PDQ; value of RAMRIS synovitis score in relation to DAS28- the original development for classification of neuropathic CRP change within 4 months. The clinical consistency pain among patients with various chronic pain conditions, and relevance of this finding needs further investigation. the algorithm for cutoff points also being validated in this Overall the DCE-MRI analyses did not add further sample and the lack of a clinical ‘gold standard’ assessing information to conventional MRI. augmented central pain mechanisms, against which the The unexpected behavior of the PDQ may lead to PDQ preferably should be tested. However, patients with speculations about whether the presents finding are due augmented central pain mechanisms express the same to the PDQ not being valid in this population. However, pain features as patients with neuropathic pain, though in the PDQ has, in at least three cross-sectional studies [13, a generalized pattern [12, 44–46]. In our opinion this 14, 36], indicated involvement of non-nociceptive pain sensitive but not specific feature of the questionnaire, mechanisms in subgroups of RA patients. Although we association to tender point count and sign of central failed to demonstrate a prognostic value of the PDQ, the sensitization in quantitative sensory testing and functional implications may still be that in patients with clear indica- MRI findings within other chronic pain conditions vouch tion of non-nociceptive pain mechanisms treatment for the use of the PDQ as indicator of augmented central strategies should involve management of chronic pain, pain processing [10–12, 47–49]. including treatment with classes of medication that target The relatively short follow-up time of 4 months is a central pain mechanisms [6, 7, 40–43], and not only focus limitation that may have reduced the treatment response on medical treatment of the underlying disease. for some patients. Further, the baseline visits including The main strength of this study was the prospective MRI could be up to 3 weeks after starting therapy for pa- design with a rigorous protocol and a prespecified analysis tients initiating a csDMARD and up to 1 week for patients plan. Furthermore, the sample of RA patients was heteroge- initiating a bDMARD, which also may have reduced the neous including patients from outpatient clinics and private observed treatment change from baseline to follow-up, rheumatologists in the Copenhagen area with a broad particularly on DCE-MRI that is known to be very sensi- spectrum of disease severity. Thus, on one hand this het- tive to fast inflammatory changes [31, 35, 50, 51]. Finally, erogeneity gave the results potential for generalizable inter- a direct pain modulating effect of anti-tumor necrosis pretation, on the other hand, it was the main limitation in factor (TNF)-alpha treatment is indicated by the literature relation to pain profiling, i.e. the constitution of the sample [52], and it could therefore be speculated that in some Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 9 of 11 patients treated with anti-TNF blocker the allowed delay for Videnskab og Kunst, Axel Muusfeldts Fond, Dagmar Marshalls Fond, Region Hovedstadens Forskningsfond, and Bjarne Jensens Fond. in baseline assessment may have influenced nociceptive signaling and thereby the study findings. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Conclusions In this study, higher numerically changes of inflamma- Authors’ contributions tory and patient-reported outcomes, and change of PDQ All authors participated in the design of the study. SRM, AWC and MB also classification-group at follow-up were observed in the participated in the data acquisition. All authors participated in drafting the manuscript and have approved it in its final form. group of patients with a high PDQ score at baseline, which may indicate that inflammatory pain can lead to Ethics approval and consent to participate pain hypersensitivity of a reversible character. Interest- This manuscript reports data from clinical research. The study was designed ingly, overall reduced change in variables was found in as a clinical prospective cohort study. It was approved by the Regional Ethics Committee of the Capital of Denmark April 18 2013; identification number patients in the intermediate PDQ classification group H-3-2013-049. with unclear pain mechanism indicating irreversible pain All participants had to sign a consent form. mechanisms. In contrast to our hypothesis, a high PDQ score had no prognostic value in relation to treatment Competing interests MB reports personal fees from Image Analysis, LTD, London, UK, outside the outcome specified as change in DAS28-CRP, RAMRIS submitted work. The remaining authors have no competing interests to score, DCE-MRI, or VAS pain 4 months after initiation declare. or escalation of medical therapy, however, it was found that the RAMRIS synovitis score may have prognostic Publisher’sNote value in relation to DAS28-CRP response. Further large- Springer Nature remains neutral with regard to jurisdictional claims in scale studies are needed to clarify the prognostic value published maps and institutional affiliations. of the PDQ. Author details The Parker Institute, Copenhagen University Hospital, Bispebjerg and Frederiksberg, 2000 Frederiksberg, Copenhagen, Denmark. Department of Additional files Rheumatology, Copenhagen University Hospital, Bispebjerg and Frederiksberg, Copenhagen, Denmark. Department of Rheumatology, Additional file 1: Table showing inter- and intra-reader reliability as Copenhagen University Hospital, Gentofte and Herlev, Hellerup, Denmark. intraclass correlation coefficient (ICC) for the DCE-MRI variables. (DOCX 16 kb) 4 Department of Radiology, Copenhagen University Hospital, Bispebjerg and Additional file 2: Table of multivariable regression models including Frederiksberg, Copenhagen, Denmark. DCE-MRI variables examining change across PDQ categories expressed as least squares means (95% CI). (DOCX 16 kb) Received: 1 November 2017 Accepted: 27 March 2018 Abbreviations References ANCOVA: Analysis of covariance; Anti-CCP: Anti-cyclic citrullinated peptide; 1. Schaible HG, von Banchet GS, Boettger MK, Brauer R, Gajda M, Richter F, et CRP: C-reactive protein; DAS28-CRP: Disease activity score 28 joints – C- al. The role of proinflammatory cytokines in the generation and reactive protein; DCE-MRI: Dynamic contrast- enhanced magnetic resonance maintenance of joint pain. Ann N Y Acad Sci. 2010;1193:60–9. imaging; DMARD: Disease-modifying antirheumatic drug; FM: Fibromyalgia; 2. Smolen JS, Landewe R, Breedveld FC, Buch M, Burmester G, Dougados M, et GAD-10: Generalized anxiety disorder assessment; GH: Global health; al. EULAR recommendations for the management of rheumatoid arthritis HAQ: Health-assessment questionnaire; IRE: Initial rate of enhancement; with synthetic and biological disease-modifying antirheumatic drugs: 2013 IQR: Interquartile range; MCP: Metacarpophalangeal; MCS: Mental component update. Ann Rheum Dis. 2014;73(3):492–509. score; MDI: Major depression inventory; ME: Maximum enhancement; 3. Meeus M, Vervisch S, De Clerck LS, Moorkens G, Hans G, Nijs J. Central MRI: Magnetic resonance imaging; MTX: Methotrexate; Nvoxel: Number of sensitization in patients with rheumatoid arthritis: a systematic literature enhancing voxel; PCS: Physical component score; PDQ: painDETECT review. Semin Arthritis Rheum. 2012;41(4):556–67. questionnaire; PRO: Patient-reported outcome; RA: Rheumatoid arthritis; 4. Lee YC, Cui J, Lu B, Frits ML, Iannaccone CK, Shadick NA, et al. Pain persists RAMRIS: Rheumatoid arthritis magnetic resonance imaging score; in DAS28 rheumatoid arthritis remission but not in ACR/EULAR remission: a RF: Rheumatoid factor; ROI: Region of interest; SF-36: 36-item short form longitudinal observational study. Arthritis Res Ther. 2011;13(3):R83. health survey; SJC: Swollen joint count; TJC: Tender joint count; TNF: Tumor 5. Wolfe F, Michaud K. Severe rheumatoid arthritis (RA), worse outcomes, necrosis factor; VAS: Visual analogue scale comorbid illness, and sociodemographic disadvantage characterize ra patients with fibromyalgia. J Rheumatol. 2004;31(4):695–700. Acknowledgements 6. Lee YC, Nassikas NJ, Clauw DJ. The role of the central nervous system in the The authors wish to thank Jack Olsen, Jette Nielsen, and Salomea Hirschorn for generation and maintenance of chronic pain in rheumatoid arthritis, assisting with the data acquisition, Christian Cato Holm for data management osteoarthritis and fibromyalgia. Arthritis Res Ther. 2011;13(2):211. and Robert C. Riis for sharing his DCE-MRI expertise. We also wish to thank the 7. Woolf CJ. Central sensitization: implications for the diagnosis and treatment Departments of Rheumatology at Bispebjerg and Frederiksberg Hospital, Herlev of pain. Pain. 2011;152(3 Suppl):S2–15. and Gentofte Hospital, and Køge Hospital for assisting with the recruitment of 8. Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO III, et al. participants. Finally, we wish to thank the Departments of Radiology at Bispebjerg 2010 Rheumatoid arthritis classification criteria: an American College of and Frederiksberg Hospital for carrying out the MRI scanning. Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62(9):2569–81. Funding 9. Freynhagen R, Baron R, Gockel U, Tolle TR. painDETECT: a new screening The study was supported by grants from The Oak Foundation, Selsbjerg questionnaire to identify neuropathic components in patients with back Holding, The Danish Rheumatism Association, Minister Erna Hamiltons Legat pain. Curr Med Res Opin. 2006;22(10):1911–20. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 10 of 11 10. Amris K, Jespersen A, Bliddal H. Self-reported somatosensory symptoms of 29. Haavardsholm EA, Ostergaard M, Ejbjerg BJ, Kvan NP, Uhlig TA, Lilleas FG, et neuropathic pain in fibromyalgia and chronic widespread pain correlate with al. Reliability and sensitivity to change of the OMERACT rheumatoid arthritis tender point count and pressure-pain thresholds. Pain. 2010;151(3):664–9. magnetic resonance imaging score in a multireader, longitudinal setting. 11. Gwilym SE, Keltner JR, Warnaby CE, Carr AJ, Chizh B, Chessell I, et al. Arthritis Rheum. 2005;52(12):3860–7. Psychophysical and functional imaging evidence supporting the 30. Boesen M, Kubassova O, Parodi M, Bliddal H, Innocenti S, Garlaschi G, et al. presence of central sensitization in a cohort of osteoarthritis patients. Comparison of the manual and computer-aided techniques for evaluation Arthritis Rheum. 2009;61(9):1226–34. of wrist synovitis using dynamic contrast-enhanced MRI on a dedicated scanner. Eur J Radiol. 2011;77(2):202–6. 12. Koroschetz J, Rehm SE, Gockel U, Brosz M, Freynhagen R, Tolle TR, et al. 31. Axelsen MB, Poggenborg RP, Stoltenberg M, Kubassova O, Boesen M, Fibromyalgia and neuropathic pain–differences and similarities. A comparison of 3057 patients with diabetic painful neuropathy and Horslev-Petersen K, et al. Reliability and responsiveness of dynamic contrast- enhanced magnetic resonance imaging in rheumatoid arthritis. Scand J fibromyalgia. BMC Neurol. 2011;11:55. Rheumatol. 2013;42(2):115–22. 13. Ahmed S, Magan T, Vargas M, Harrison A, Sofat N. Use of the painDETECT tool in rheumatoid arthritis suggests neuropathic and sensization 32. Ballegaard C, Riis RG, Bliddal H, Christensen R, Henriksen M, Bartels EM, et al. components in pain reporting. J Pain Res. 2014;7:579–88. Knee pain and inflammation in the infrapatellar fat pad estimated by 14. Koop SM, Ten Klooster PM, Vonkeman HE, Steunebrink LM, van de Laar MA. conventional and dynamic contrast-enhanced magnetic resonance imaging Neuropathic-like pain features and cross-sectional associations in in obese patients with osteoarthritis: a cross-sectional study. Osteoarthr rheumatoid arthritis. Arthritis Res Ther. 2015;17(1):237. Cartil. 2014;22(7):933–40. 15. Wu Q, Inman RD, Davis KD. Neuropathic pain in ankylosing spondylitis - a 33. Bandak E, Boesen M, Bliddal H, Riis RG, Gudbergsen H, Henriksen M. psychophysics and brain imaging study. Arthritis Rheum. 2013;65(6):1494–503. Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study. Osteoarthr Cartil. 2015;23(10):1721–7. 16. Ostergaard M, Peterfy C, Conaghan P, McQueen F, Bird P, Ejbjerg B, et al. 34. Boesen M, Kubassova O, Bouert R, Axelsen MB, Ostergaard M, Cimmino MA, OMERACT Rheumatoid Arthritis Magnetic Resonance Imaging Studies. Core et al. Correlation between computer-aided dynamic gadolinium-enhanced set of MRI acquisitions, joint pathology definitions, and the OMERACT MRI assessment of inflammation and semi-quantitative synovitis and bone RA-MRI scoring system. J Rheumatol. 2003;30(6):1385–6. marrow oedema scores of the wrist in patients with rheumatoid arthritis–a 17. Axelsen MB, Stoltenberg M, Poggenborg RP, Kubassova O, Boesen M, cohort study. Rheumatology (Oxford). 2012;51(1):134–43. Bliddal H, et al. Dynamic gadolinium-enhanced magnetic resonance imaging allows accurate assessment of the synovial inflammatory activity 35. Cimmino MA, Parodi M, Zampogna G, Boesen M, Kubassova O, Barbieri F, et in rheumatoid arthritis knee joints: a comparison with synovial histology. al. Dynamic contrast-enhanced, extremity-dedicated MRI identifies synovitis Scand J Rheumatol. 2012;41(2):89–94. changes in the follow-up of rheumatoid arthritis patients treated with 18. Ostergaard M, Stoltenberg M, Lovgreen-Nielsen P, Volck B, Jensen CH, rituximab. Clin Exp Rheumatol. 2014;32(5):647–52. Lorenzen I. Magnetic resonance imaging-determined synovial membrane 36. Christensen AW, Rifbjerg-Madsen S, Christensen R, Dreyer L, Tillingsoe H, and joint effusion volumes in rheumatoid arthritis and osteoarthritis: Seven S, et al. Non-nociceptive pain in rheumatoid arthritis is frequent and comparison with the macroscopic and microscopic appearance of the affects disease activity estimation: cross-sectional data from the FRAME synovium. Arthritis Rheum. 1997;40(10):1856–67. study. Scand J Rheumatol. 2016;18:1–9. 19. Christensen AW, Rifbjerg-Madsen S, Christensen R, Amris K, Taylor PC, Locht H, 37. Christensen AW, Rifbjerg-Madsen S, Christensen R, Dreyer L, Boesen M, et al. Temporal summation of pain and ultrasound Doppler activity as Ellegaard K, et al. Ultrasound Doppler but not temporal summation of pain predictors of treatment response in patients with rheumatoid arthritis: protocol predicts DAS28 response in rheumatoid arthritis: a prospective cohort study. for the Frederiksberg hospitals Rheumatoid Arthritis, pain assessment and Rheumatology (Oxford). 2016;55(6):1091–8. Medical Evaluation (FRAME-cohort) study. BMJ Open. 2014;4(1):e004313. 38. Ostergaard M, Moller-Bisgaard S. Optimal use of MRI in clinical trials, clinical care and clinical registries of patients with rheumatoid arthritis. Clin Exp 20. Rifbjerg-Madsen S, Christensen AW, Boesen M, Christensen R, Danneskiold- Rheumatol. 2014;32(5 Suppl 85):S-22. Samsoe B, Bliddal H, et al. Can the painDETECT Questionnaire score and MRI help predict treatment outcome in rheumatoid arthritis: protocol for the 39. Suter LG, Fraenkel L, Braithwaite RS. Role of magnetic resonance imaging in Frederiksberg hospital's Rheumatoid Arthritis, pain assessment and Medical the diagnosis and prognosis of rheumatoid arthritis. Arthritis Care Res Evaluation (FRAME-cohort) study. BMJ Open. 2014;4(11):e006058. (Hoboken ). 2011;63(5):675–88. 21. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, 40. Brendbekken R, Eriksen HR, Grasdal A, Harris A, Hagen EM, Tangen T. et al. The American Rheumatism Association 1987 revised criteria for the Return to work in patients with chronic musculoskeletal pain: classification of rheumatoid arthritis. Arthritis Rheum. 1988;31(3):315–24. multidisciplinary intervention versus brief intervention: a randomized 22. Rifbjerg-Madsen S, Waehrens EE, Danneskiold-Samsoe B, Amris K. clinical trial. J Occup Rehabil. 2017;27(1):82-91. Psychometric properties of the painDETECT questionnaire in rheumatoid 41. Tse MM, Yeung SS, Lee PH, Ng SS. Effects of a peer-led pain arthritis, psoriatic arthritis and spondyloarthritis: Rasch analysis and management program for nursing home residents with chronic pain: a test-retest reliability. Health Qual Life Outcomes. 2017;15(1):110. pilot study. Pain Med. 2016;18. 42. Amris K, Luta G, Christensen R, Danneskiold-Samsoe B, Bliddal H, Waehrens EE. 23. Gudbergsen H, Bartels EM, Krusager P, Waehrens EE, Christensen R, Predictors of improvement in observed functional ability in patients with Danneskiold-Samsoe B, et al. Test-retest of computerized health status fibromyalgia as an outcome of rehabilitation. J Rehabil Med. 2016;48(1):65-71. questionnaires frequently used in the monitoring of knee osteoarthritis: a randomized crossover trial. BMC Musculoskelet Disord. 2011;12:190. 43. Amris K, Waehrens EE, Christensen R, Bliddal H, Danneskiold-Samsoe B. Interdisciplinary rehabilitation of patients with chronic widespread pain: 24. Thorsen H, Hansen TM, McKenna SP, Sorensen SF, Whalley D. primary endpoint of the randomized, nonblinded, parallel-group IMPROvE Adaptation into Danish of the Stanford Health Assessment trial. Pain. 2014;155(7):1356–64. Questionnaire (HAQ) and the Rheumatoid Arthritis Quality of Life Scale 44. Baron R, Binder A, Wasner G. Neuropathic pain: diagnosis, pathophysiological (RAQoL). Scand J Rheumatol. 2001;30(2):103–9. mechanisms, and treatment. Lancet Neurol. 2010;9(8):807–19. 25. Bjorner JB, Thunedborg K, Kristensen TS, Modvig J, Bech P. The Danish 45. Bennett MI, Attal N, Backonja MM, Baron R, Bouhassira D, Freynhagen R, SF-36 Health Survey: translation and preliminary validity studies. J Clin et al. Using screening tools to identify neuropathic pain. Pain. 2007; Epidemiol. 1998;51(11):991–9. 127(3):199–203. 26. Bech P, Kastrup M, Rafaelsen OJ. Mini-compendium of rating scales for 46. Rehm SE, Koroschetz J, Gockel U, Brosz M, Freynhagen R, Tolle TR, et al. A states of anxiety depression mania schizophrenia with corresponding cross-sectional survey of 3035 patients with fibromyalgia: subgroups of DSM-III syndromes. Acta Psychiatr Scand Suppl. 1986;326:1–37. patients with typical comorbidities and sensory symptom profiles. 27. Bech P, Rasmussen NA, Olsen LR, Noerholm V, Abildgaard W. The Rheumatology (Oxford). 2010;49(6):1146–52. sensitivity and specificity of the Major Depression Inventory, using the 47. Hochman JR, Gagliese L, Davis AM, Hawker GA. Neuropathic pain symptoms Present State Examination as the index of diagnostic validity. J Affect in a community knee OA cohort. Osteoarthr Cartil. 2011;19(6):647–54. Disord. 2001;66(2–3):159–64. 28. Ejbjerg B, McQueen F, Lassere M, Haavardsholm E, Conaghan P, O'Connor P, 48. Hochman JR, Davis AM, Elkayam J, Gagliese L, Hawker GA. Neuropathic pain et al. The EULAR-OMERACT rheumatoid arthritis MRI reference image atlas: symptoms on the modified painDETECT correlate with signs of central the wrist joint. Ann Rheum Dis. 2005;64(Suppl 1):i23–47. sensitization in knee osteoarthritis. Osteoarthr Cartil. 2013;21(9):1236–42. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 11 of 11 49. MoretonBJ, TewV,das NR WM,Walsh DA,Lincoln NB.Pain phenotype in patients with knee osteoarthritis: classification and measurement properties of painDETECT and self-report Leeds assessment of neuropathic symptoms and signs scale in a cross-sectional study. Arthritis Care Res (Hoboken ). 2015;67(4):519–28. 50. Boesen M, Kubassova O, Cimmino MA, Ostergaard M, Taylor P, Danneskiold-Samsoe B, et al. Dynamic contrast-enhanced MRI can monitor the very early inflammatory treatment response upon intra- articular steroid injection in the knee joint: a case report with review of the literature. Arthritis. 2011;2011:578252. 51. Conaghan PG, Ostergaard M, Bowes MA, Wu C, Fuerst T, van der Heijde D, et al. Comparing the effects of tofacitinib, methotrexate and the combination, on bone marrow oedema, synovitis and bone erosion in methotrexate-naive, early active rheumatoid arthritis: results of an exploratory randomised MRI study incorporating semiquantitative and quantitative techniques. Ann Rheum Dis. 2016;75(6):1024–33. 52. Hess A, Axmann R, Rech J, Finzel S, Heindl C, Kreitz S, et al. Blockade of TNF-alpha rapidly inhibits pain responses in the central nervous system. Proc Natl Acad Sci U S A. 2011;108(9):3731–6. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries � Our selector tool helps you to find the most relevant journal � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Arthritis Research & Therapy Springer Journals

The course of pain hypersensitivity according to painDETECT in patients with rheumatoid arthritis initiating treatment: results from the prospective FRAME-cohort study

Free
11 pages

Loading next page...
 
/lp/springer_journal/the-course-of-pain-hypersensitivity-according-to-paindetect-in-0nbPgoSpfg
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Rheumatology; Orthopedics
eISSN
1478-6362
D.O.I.
10.1186/s13075-018-1581-4
Publisher site
See Article on Publisher Site

Abstract

Background: Evidence is emerging that pain in rheumatoid arthritis (RA) exists without underlying inflammation. Our objective was to evaluate the prognostic value of pain classification at treatment initiation using the painDETECT questionnaire (PDQ). Outcomes were change in DAS28-CRP and RAMRIS synovitis score. Methods: RA patients initiating a disease-modifying anti-rheumatic drug (DMARD) or initiating/ switching a biological agent were included. Follow-up time was 4 months. Clinical examination, imaging (MRI, dynamic contrast-enhanced MRI (DCE-MRI)), and patient-reported outcomes were undertaken. The PDQ was used to differentiate pain mechanisms. Mean change (95% CI) was calculated using ANCOVA. Multivariable regression models were used to determine a prognostic value. Results: A total of 102 patients were included; 75 were enrolled for MRI. Mean changes in baseline variables were greatest in the high PDQ classification group (> 18), while limited in the intermediate group (13–18). The 12 patients with high baseline PDQ score all changed pain classification group. No prognostic value of PDQ pain classification was found in relation to change of DAS28-CRP, RAMRIS score, or VAS pain. In the unadjusted model, RAMRIS score at baseline was associated with change in DAS28-CRP. The exploratory variables of DCE-MRI did not differ from other inflammatory variables. Conclusions: In RA patients a high PDQ score (non-nociceptive pain) at baseline was not associated with worse outcomes, in fact these patients had numerically greater improvement in DAS28-CRP. However, pain classification by PDQ was not independently associated with change in DAS28-CRP, RAMRIS score, or VAS pain in the prognostic models. Furthermore, patients classified with a high baseline PDQ score changed pain classification group. Patients with unclear pain mechanism had reduced numerically treatment response. Trial registration: The study was approved by the Regional Ethics Committee of the Capital of Denmark April 18 2013; identification number H-3-2013-049. Keywords: Rheumatoid arthritis, Central sensitization, painDETECT questionnaire, Prognostics, Dynamic contrast- enhanced magnetic resonance imaging (DCE-MRI) * Correspondence: signe.rifbjerg-madsen.02@regionh.dk The Parker Institute, Copenhagen University Hospital, Bispebjerg and Frederiksberg, 2000 Frederiksberg, Copenhagen, Denmark Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 2 of 11 Background In this study, we hypothesized that a high PDQ score Pain in rheumatoid arthritis (RA) has typically been would serve as an indicator of central sensitization and regarded as nociceptive, that is, related to ongoing periph- thus a prognostic factor for a poorer treatment outcome eral inflammation [1]. However, during the last decade, (DAS28-CRP change) in patients with RA initiating or where focus has been on early diagnosis and aggressive intensifying anti-inflammatory treatment. A possible stat- treatment strategies in the treat-to-target regime [2], it has istical interaction between central sensitization (high PDQ become clearer that in a subgroup of RA patients, pain score) and inflammatory load (baseline synovitis defined can become an entity in its own right, probably elicited by hand MRI RAMRIS score) was considered as part of by, but not directly related to, ongoing inflammation [3]. the hypotheses. In the exploratory part of the study, we A substantial proportion of RA patients in stable clinical hypothesized that DCE-MRI would capture change in remission continue to report moderate to severe pain inflammation and thus a possible relation to inflammatory levels [4] and studies have indicated that RA leads to pain mechanisms (low PDQ score) better than conven- widespread pain in 10–20% of patients [5]. Such observa- tional MRI. tions have led to the contention that changes in the peripheral and central nervous system through processes Methods of neural plasticity and central sensitization may play an Design important role [6]. As a rule, sensitization phenomena The Frederiksberg Hospital’s Rheumatoid Arthritis, pain would be expected to extinguish as the tissue heals and assessment and Medical Evaluation (FRAME)-cohort inflammation subsides. However, a state of induced hyper- study was conducted according to a published protocol, sensitivity of the pain system may persist in subsets of which contained a detailed description of the methods patients and lead to chronic pain states in which pain is and prespecified analysis [19, 20]. It was approved by the no longer coupled to the presence of ongoing peripheral Regional Ethics Committee of the Capital of Denmark; inflammation [7]. In such patients, persistent pain hyper- identification number H-3-2013-049. sensitivity may lead to continuous high reports of tender RA patients were recruited from departments and pri- joints and poor global health; subcomponents of the com- vate clinics of rheumatology in the Copenhagen area and monly used composite disease activity score of 28 joints prospectively enrolled from March 2013 to September (DAS28-CRP) and thus overestimation of inflammatory 2014. MRI was included in the examination program from activity. If the anti-inflammatory treatment is intensified May 2013. The examination program was conducted at on this background, little change in DAS28-CRP can be Frederiksberg Hospital. Patients were assessed at treat- expected. Conversely, if inflammatory RA is left un- or ment initiation (baseline) and after 4 months of treatment. not sufficiently treated it will lead to joint destruction and Patients received routine care at the discretion of their loss of function [8]. Identification of underlying pain rheumatologist during the trial period. Add-on of pain- mechanisms therefore has potential importance when killers was allowed. prognosticating the effect of medical treatment on inflam- mation and pain. Patients The painDETECT questionnaire (PDQ) is a self- To be eligible, patients had to fulfil either the 1987 [21]or administered pain classification instrument originally 2010 ACR RA criteria [8]and be ≥18 years. Further, developed to differentiate neuropathic (non-nociceptive) patients had to be scheduled for either (a) treatment initi- from non-neuropathic (nociceptive) pain [9]. It has been ation with any conventional synthetic disease-modifying increasingly used in patients with osteoarthritis and antirheumatic drug (csDMARD) (patients who had not fibromyalgia to assess clinical pain features indicative of received treatment with csDMARD for at least 6 months central sensitization [10–12] and has recently been including newly diagnosed/treatment-naïve patients) or introduced in studies assessing pain mechanisms in (b) treatment initiation or change of any biologic DMARD patients with RA and spondyloarthritis [13–15]. (bDMARD). Magnetic resonance imaging (MRI) is an objective and Major exclusion criteria were intra-articular or intra- sensitive method to assess joint inflammation. The most muscular glucocorticoids administered less than 3 weeks common scoring system in the wrist and metacarpophalan- prior to baseline; treatment with oral corticosteroids at geal (MCP) joints is the OMERACT (outcome measures in doses equivalent to more than 10 mg prednisolone/day rheumatoid arthritis clinical trials) RA MRI scoring (RAM- within the 3 weeks prior to baseline; inability to pause RIS) system [16]. Dynamic contrast-enhanced (DCE) MRI antidepressants, anticonvulsants or other centrally acting is a technique where the sequences are acquired rapidly analgesics; initiation of csDMARD therapy more than and sequentially before and during contrast infusion. DCE- 3 weeks prior to the baseline visit (only patients initiating MRI has been shown to correlate better than conventional csDMARDs); treatment with bDMARD initiated more MRI with the histologic findings of synovitis [17, 18]. than 1 week prior to the baseline visit (only patients Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 3 of 11 initiating a bDMARD). Patients who had contraindica- As standard research procedure a target hand was tions for MRI were excluded from the MRI arm of the chosen for MRI to reflect the general level of joint inflam- study. Furthermore, patients with increased risk of neuro- mation. The most painful hand as reported by the patient pathic pain conditions (e.g. diabetes) were excluded due to was chosen, or, in case of no difference in pain level, the the potential to confound the pain assessment. dominant hand. The examination was carried out in a 3 T Siemens Verio MR scanner according to a published Variables and outcome measures scanning procedure [20]. Conventional coronal and axial The patients underwent an examination program at STIR and 3D coronal T1w GRE VIBE pre- and post- baseline and follow-up, collecting information on contrast images were used for RAMRIS scoring. The wrist demographics and medication and from patient-reported and MCP joints 2–5 were assessed according to the outcomes (PROs). Clinical examination including joint OMERACT RAMRIS [16, 28] and were scored for count and tender point examination conducted by the synovitis (0–3; total score 0–21) and bone marrow edema same assessor at both time points, imaging (MRI and (BMO) (0–3; total score 0–58). Based on previous reports DCE- MRI), and standard blood samples (CRP, immuno- on smallest detectable difference, it was decided that the globulin M-rheumatoid factor [IgM-RF], anti-cyclic RAMRIS synovitis score had to alter by more than 1 unit citrullinated peptide [anti-CCP]) were also performed. to be considered a significant change [29]. All images The following PROs were collected from each patient; were assessed blinded and paired by the same senior the PDQ, the Stanford health assessment questionnaire radiologist (MB). disability index (HAQ-DI), the 36-item short form health For the explorative DCE-MRI analyses the software survey (SF-36), generalized anxiety disorder assessment DYNAMIKA enterprise version 3.2.6 (http://www.ia-grp. (GAD-10) and major depression inventory (MDI). com) was used according to a published procedure [30]. The PDQ is a patient-administered pain classification Only joints with MRI signs of inflammation (“focus tool that was developed in a population of patients with joints”) were included. All images were analyzed paired various pain conditions. It has been further validated for by the same physician (SRM). Regions of interest (ROIs) describing pain phenotypical features in patients with were drawn on all slices where sign of inflammation was inflammatory arthritis by our group [22] and is validated present and collapsed into one volume of interest (VOI) for use on touch screen [9, 23]. It comprises items on pain for each focus joint; wrist and 2nd-5th MCP. It was intensity (three numeric rating scales not included in the decided to include tenosynovitis and capsulitis as ‘signs total score), pain course patterns, pain radiation (from a of inflammation’ in the analyses. Joints with no signal pain drawing) and seven somatosensory signs and symp- were assigned a score of 0. Nvoxel, IRExNvoxel, MExN- toms (rated on a six-category Likert scale). According to a voxel and IRExME were chosen as outcome measures validated algorithm, patients were assigned to one of three [17, 31–35]. The number of enhancing voxels (Nvoxel) pain classification-groups based on a score between − 1 was multiplied by the volume of each voxel in milliliter and 38: > 18 likely neuropathic pain, 13–18 unclear pain (ml) to adjust for different image sizes. The initial rate of mechanism or < 13 likely non-neuropathic pain [9]. Sev- enhancement (IRE) and maximum enhancement (ME) eral studies have used it as indicator of non-nociceptive or represent the degree of perfusion; the IRE reflects the central pain mechanisms [10–12]. initial rate of enhancement of the time intensity curve. HAQ-DI is a measure of limitation of activities of daily ME represents the equilibrium state of the curve and living used for patients with RA. It assesses the patient’s reflects the amount of contrast passing into the ROI. ability to carry out everyday tasks. It includes a visual The composite outcome measures IRExNvoxel and MExN- analogue scale (VAS) evaluation of pain, fatigue and glo- voxel reflect both the volume and degree of perfusion, bal health (GH) [24].The SF-36 assesses eight domains whereas IRExME characterizes the perfusion profile of the concerning general health, which can be summarized voxels derived from the time-intensity curves. into a physical (PCS) and mental (MCS) component summary score. In this study the Danish version of SF- Statistical analysis 36, which uses a 4-week recall period was applied [25]. SAS software (version 9.4, SAS Enterprise Guide 7.1, SAS The GAD-10 is a ten-item instrument developed from Institute Inc., Cary, NC, USA) was used for all statistical the Hamilton six-item anxiety scale. It measures general- analyses. PROC UNIVARIATE statement was used to ized anxiety by scoring the total sum of the items [26]. summarize thedataand for visualinspectionofnormality. The MDI is a questionnaire based on self-reported mood Means (with standard deviations [SDs]) or medians (with symptoms. It holds the ability to generate a DSM-IV and interquartile ranges [IQRs]) were reported and compared International Classification of Diseases (ICD)-10 diagno- by t test or Kruskal-Wallis (Wilcoxon) test, respectively. ses of major (moderate to severe) depression and to rate Delta changes were adjusted for baseline value and com- the severity of symptoms [27]. pared using analysis of covariance (ANCOVA). All analyses Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 4 of 11 were carried out according to the intention-to-treat csDMARD and 54 initiating bDMARD. Of these, 75 principle, i.e. missing data at follow-up was imputed from patients completed the PDQ and had an MRI scan baseline (baseline observation carried forward). A two- performed at baseline and of these, 71 patients had an sided p value less than 0.05 was regarded as being statisti- MRI scan performed at follow-up. This discrepancy was cally significant. primarily caused by administrative delay resulting in no Prior to executing the FRAME-cohort study, a power patients receiving MRI-scan during the first 2 months of calculation was performed based on the assumption that the study period. Exclusion and reasons for dropout are it was feasible to include 100 RA patients during a study further described in the figure text. Due to 3 patients not period of 1½ years [20], as no data for sample size calcula- wishing to receive contrast, 72 patients with complete tion was available. Anticipating a common SD of 1.5 and PDQ were included in the exploratory DCE-MRI study at the correlation between pre- and post-scores being r =0.3 baseline (not shown in Fig. 1). Of these72 patients, in all for a paired t test with a significance level of 0.05, a sample 65 completed the follow-up scan with contrast. of a 100 pairs has a power of 80% (0.797) to detect a mean Baseline characteristics are described in Table 1 (some change of 0.5 DAS28-CRP units. A patient population, data has previously been published [36]). The distribu- who can expect a change of this magnitude in their tion of patients across the PDQ classification groups disease activity, is a reasonable cohort in which to study were n = 66, n = 23, n = 12 for PDQ score < 13, 13–18, > prognostic factors of treatment response. However, this 18, respectively. Statistically significant differences across number was not reached for the MRI subsample. the three PDQ classification groups were observed for The prognostic value of the PDQ score, RAMRIS score, tender joint count (TJC), tender point (TP) count, and their interaction at baseline in relation to change of DAS28-CRP, physical function (HAQ-DI), VAS-fatigue, DAS28-CRP was examined by multivariable regression VAS-pain, VAS-GH, anxiety (GAD-10), depression models using the SAS PROC GLM. As the interpretation (MDI), SF-36 PCS, and SF-36 MCS. A TP count ≥11 of PDQ by nature is trichotomous, the results were was found for higher proportions of patients with higher expressed as least squares means per category. According PDQ classification groups; however, this was not statisti- to the protocol, the model was adjusted for the following cally significant. No differences were found across the prespecified confounders: age (years), sex (male/female), groups for the biochemical and imaging variables. disease duration (month), disease activity (DAS28-CRP at Changes from baseline stratified by PDQ group are baseline), group (csDMARD/bDMARD), antiCCP-positive presented in Table 2. Change in PDQ classification group is (yes/no) and concomitant prednisolone (yes/no). Subse- reported as classification consistency, i.e. number of patient quently, in the fully adjusted model covariates (i.e. possible that did not change classification group. Statistically signifi- confounders) that did not contribute to the model were cant differences in classification consistency were found for removed; age, antiCCP-positive (yes/no) and concomitant the low, intermediate, and high PDQ group; 97%, 83%, and prednisolone (yes/no). 0%, respectively. For all clinical variables and PROs, a U- Secondary outcomes were change in RAMRIS synovitis shaped curve change pattern was observed with greater score and VAS pain. Post hoc, to ensure robustness of change in the high and low PDQ classification group than results, a sensitivity analysis including baseline VAS pain in the intermediate. Remission at follow-up according to as a confounder in the adjusted analysis of change in VAS DAS28-CRP (< 2.6) was found in 41%, 22%, and 42%, pain was performed. respectively. The change in imaging variables did not On an exploratory basis the DCE-MRI variables IRExN- display a distinct pattern. Statistically significant differences voxel (in ml) or MExNvoxel (in ml) for wrist were applied between the classification groups were found for delta in the primary model examining DAS28-CRP and VAS change of CRP, DAS28-CRP, VAS GH, HAQ-DI, and SF36- pain change, replacing the RAMRIS synovitis score. For MCS. However for CRP, no differences between PDQ SRM, inter- and intra-reader agreements, intraclass correl- score < 13 and > 18, and 13–18 and > 18 were found. For ation coefficients (ICC) (absolute agreement) for the four DA28-CRP and VAS GH no differences were found predefined DCE-MRI variables were tested beforehand on between PDQ score < 13 and PDQ score > 18. Regarding data from ten patients using SPSS software. Wrist and HAQ-DI and SF36-MCS there were no difference between MCP joints were tested separately. PDQ score < 13 and 13–18. DCE-MRI variables are presented separately in Table 3. There were neither signifi- Results cant differences nor trends in the DCE-MRI variables Figure 1 illustrates the flow of patients. In all, 151 patients across the groups; however, the variables IRExNvoxel (in fulfilled the inclusion criteria. Of these, 48 patients were ml) and MExNvoxels (in ml) for the wrist also displayed excluded. In total, 103 patients received a baseline assess- the U-shaped change pattern in line with the RAMRIS ment; however two were excluded post hoc. In total, 101 score (BME). Baseline characteristics and delta change by patients completed the PDQ at baseline; 47 initiating initiation group have previously been reported [36, 37]. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 5 of 11 Fig. 1 Flowchart of participants. 1: Refrained from participation (n = 18), comorbidity with risk of neuropathic pain (n = 10), unable to pause antidepressants, anticonvulsants, or other centrally acting analgesics for 1 week (n = 7), initiated DMARD treatment > 3 weeks ago (n = 5), received more than 10 mg prednisolone < 3 weeks ago (n = 4), other reasons (n = 4). 2: Im. Corticosteroid between screening and baseline assessment. 3: MRI not feasible or patients declined. 4: Four patients withdrew from the study; two from each treatment group, three from the MRI subgroup. A further one patient did not participate in the follow-up MRI In the multivariable regression models (Table 4) Discussion change in the predefined outcome variables was To ourknowledge,thisisthe first study to evaluate expressed as least square means. No interaction was the prognostic value of pain classification by the PDQ found between PDQ classification group and RAMRIS score in relation to change in DAS28-CRP, VAS pain, synovitis neither in the protocolized analysis, nor and RAMRIS score in RA patients initiating or escal- between PDQ classification group and the DCE-MRI ating anti-inflammatory treatment. It was hypothe- variables (IRExNvoxel (ml), MExNvoxel (ml)) in the sized that patients in the high PDQ classification exploratory analysis. All interaction statements were group (score > 18) would display constant high TJC therefore left out of the models. No prognostic value and VAS GH as features of persistent pain hypersen- of PDQ classification was found for any of the sitivity (central sensitization) and thus gain little or outcomes in any of the models, sensitivity analysis no change of DAS28-CRP following treatment initi- included (p = 0.44) (data for DCE-MRI is presented as ation. In contrast however, this was not confirmed in additional material, Additional file 1). the multiple regression analysis, we found that these Baseline DAS28 was found to be statistically patients (n = 12) experienced the greatest numerical significantly associated with change in all outcomes change in DAS28-CRP, self-reported disease severity in both adjusted models; DAS28, RAMRIS, and VAS measures, and objective inflammatory parameters, pain (p < 0.01), sensitivity analysis excluded. In this including MRI (Table 2). This result was seen despite analysis, baseline VAS pain was found to be signifi- the fact that baseline inflammatory parameters were cantly associated with change in VAS pain (p <0. thesameacrossthe threePDQ classification groups, 001). Baseline RAMRIS synovitis score was found to and that other baseline characteristics of the high be statistically significantly positively associated with PDQ classification group clinically could indicate DAS28-CRP change in the unadjusted model (p =0. presence of central sensitization (high frequency of 01) and of RAMRIS change in both the unadjusted female sex, a high number of tender joints and tender and the adjusted model (p ≤ 0.02). points, poor mental well-being, a high disability index, In the exploratory DCE-MRI study inter and intra- and high VAS scores). reader reliability showed good to excellent agreement In contrast to the patients in the low and intermedi- (data is shown in Additional file 2). ate PDQ classification -group, all patients in the high Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 6 of 11 Table 1 Baseline characteristics stratified by PDQ group PDQ score 7 (5-9) PDQ score 16 (13-18) PDQ score ≥ 19 p value (n = 66) (n = 23) (n = 12) PDQ score 7 (5–9) 16 (13–18) 21 (20–23) <.0001 Female, n (%) 45 (68.2) 20 (87.0) 31 (91.7) 0.09 Age, years (SD) 56.1 (14.7) 54.6 (18.9) 47.4 (15.6) 0.22 Initiated csDMARD, n (%) 37 (78.7) 6 (12.8) 4 (8.5) Disease duration, months 15.5 (1–104) 53 (9–47) 34.5 (24–149.5) 0.12 Current smoker, n (%) 11 (16.9) 7 (30.4) 2 (16.7) 0.39 Corticosteroid usage, n (%) 11 (16.7) 2 (8.7) 3 (25.0) 0.42 28 Swollen joint count 2 (1–7) 5 (2–8) 2 (1–4) 0.09 28 Tender joint count 5 (3–10) 13 (8–16) 9.5 (6.5–15) <.0001 Tender point count, 0–18 6 (4–14) 10 (7–16) 12 (6–14) 0.02 Tender point count ≥11, n (%) 20 (30.3) 11 (47.8) 7 (58.3) 0.09 DAS28 (SD) 4.2 (1.1) 5.0 (1.2) 4.8 (0.7) 0.007† HAQ-DI, 0–3 0.75 (0.38–1.25) 1.13 (0.88–1.75) 1.63 (1.19–1.88) 0.0004 VAS-fatigue, mm 53.5 (27–71) 75 (52–89) 72.5 (60.5–87) 0.002 VAS-pain, mm 42 (24–60) 69 (50–82) 63 (45.5–80.5) 0.0003 VAS-global health, mm 56 (32–77) 73 (46–85) 78 (61.5–90) 0.03‡ GAD-10 score, 0–50 6 (3–10) 10 (6–17) 9 (9–12) 0.008 MDI score, 0–50 8 (4–12) 11 (6–25) 13.5 (9–19.5) 0.01 SF-36 PCS, 0–100 35 (29–42) 33 (26–37) 28 (24–32) 0.02‡ SF-36 MCS, 0–100 51 (40–57) 43 (31–51) 37 (31–49) 0.005 CRP, mg/mL 8 (3–15) 4 (0.5–19) 3 (0.8–8.5) 0.13 IgM-RF positive, n (%) 41 (62.1) 15 (65.2) 8 (66.7) 1.0 Anti-CCP positive, n (%) 46 (70) 15 (65) 7 (58) 0.72 RAMRIS hand synovitis 7 (5–10) 9.5 (7–11) 7 (6–9) 0.18 RAMRIS hand edema 5 (1–10) 9.5 (3–22) 5 (2–13) 0.12 Values are median (25th, 75th percentiles) unless specified otherwise Unless specified otherwise significant p values reflect difference between PDQ score < 13 and 13–18 and ≥ 19, while there is no difference between PDQ score 13–18 and ≥ 19 Only difference between PDQ score < 13 and 13–18 Only difference between PDQ score < 13 and ≥ 19 RAMRIS hand; wrist+MCP scores One patient receiving MRI had no corresponding PDQ score. 26 patients did not receive MRI scan. Number of patients (n) for RAMRIS parameters: PDQ score < 13, n = 51; PDQ score 13–18, n = 18; PDQ score > 18, n = 6. (a) One missing observation. (b) Three missing observations PDQ classification group changed pain classification pattern across the PDQ classification groups, indicating following treatment initiation or escalation. These find- that the uncoupling of pain mechanism from present ings could indicate that in these particular patients, a inflammation could be found in this group. high PDQ score might not have marked central pain In the multivariable regression analyses, the PDQ did mechanisms uncoupled to ongoing inflammation, but not have prognostic value in relation to change of DAS28- could have reflected reversible inflammatory driven CRP, RAMRIS score, or VAS pain. No interaction between pain hypersensitivity, i.e. normal neuroplasticity. However, PDQ score and baseline RAMRIS synovitis score was regression toward the mean needs to be taken into found in any of the regression analyses including the account. Interestingly, the observed numerically changes exploratory DCE-MRI analyses indicating no relation of the variables were smallest in the intermediate PDQ between ongoing inflammation and pain phenotype (pain group (score 13–18), total hand RAMRIS synovitis score classification group). However, in the light of the relatively excluded. Thus, patients with an unclear pain mechanistic small sub-sample of patients with a high PDQ score at background had the poorest response to medical treat- baseline, it does not seem reasonable to reject any prog- ment also reflected in the U-shaped DAS28 remission nostic value of the PDQ at this point. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 7 of 11 Table 2 Mean change stratified by baseline PDQ group adjusted for baseline value PDQ score < 13 PDQ score 13–18 PDQ score ≥ 19 p value (n = 66) (n = 23) (n = 12) PDQ consistency, N (%) 64 (97) 19 (83) 0 (0) < 0.001 Δ PDQ score −4.10 (−5.74;-2.45) −2.20 (− 4.64;0.25) −5.49(−9.97;-1.02) 0.08 Δ 28 SJC −2.71 (− 3.26;-2.17) − 2.27 (− 3.21;-1.34) − 2.48(− 3.77;-1.19) 0.71 Δ 28 TJC − 4.63(− 5.67;-3.60) −4.36(− 5.14; − 1.58) −5.57(−7.96;-3.18) 0.28 Δ CRP, mg/mL −6.93 (− 8.54;-5.32) −2.81 (− 5.54;-0.08) −6.28 (− 10.10;- 2.47) 0.04a Δ DAS28-CRP − 1.47 (− 1.74; − 1.20) −0.8(− 1.27;-0.34) −1.62 (− 2.25;-1.00) 0.03b Δ Tender point count −2.70 (− 3.87;-1.57) −1.56 (− 3.48; 0.36) − 3.83 (− 6.46;-1.20) 0.35 Δ VAS-fatigue −18.31(− 25.55;-13.06) −8.19 (− 17.07; 0.68) −22.20 (− 34.48;-9.91) 0.09 Δ VAS-pain −23.27(− 28.21;-18.33) −12.99(− 21.40;-4.57) −26.86 (− 38.34;-15.38) 0.07 Δ VAS-global health −26.20(− 31.15;-21.25) −13.49 (− 21.83;-5.15) −31.72 (− 43.44;-20.01) 0.01b Δ HAQ-DI − 0.34 (− 0.44;-0.24) − 0.15 (− 0.32; 0.02) −0.74 (− 0.98;-0.50) < 0.001c Δ MDI-total − 3.04(− 4.39;-1.68) −2.54 (− 4.83;-0.25) −3.68(− 6.80;-0.55) 0.84 Δ GAD10-total −2.24 (− 3.18;-1.30) −1.56 (− 3.15; 0.02) −2.10 (− 4.27; 0.07) 0.76 Δ SF36-MCS 5.5 (3.34; 7.60) 3.70 (0.09; 7.31) 11.38 (6.40;16.36) 0.04c Δ SF36-PCS 6.47 (4.39; 8.56) 4.94 (1.43; 8.44) 8.96 (3.97; 13.94) 0.42 Δ RAMRIS H synovitis −1.69 (− 2.32;-1.07) − 1.81(− 2.90,-0.71) −2.47 (− 4.28;-0.66) 0.72 Δ RAMRIS H edema −2.17 (− 4.00;-0.34) − 0.57(− 3.80;2.66) − 3.46 (− 8.72;1.81) 0.58 Values are means (95% CI). Negative numbers indicate improvement. ANCOVA (BOCF) was used for the analyses unless otherwise indicated. †Chi-square test RAMRIS H hand (W + M). W wrist. M MCP joints (n) for RAMRIS parameters: PDQ score < 13; n = 51, PDQ score 13–18; n = 18, PDQ score > 18; n = 6. (a) No difference between PDQ score < 13 and > 18, and 13–18 and > 18. (b) No difference between PDQ score < 13 and PDQ score > 18. (c) No difference between PDQ score < 13 and 13–18 Table 3 Baseline values and mean changes (adjusted for baseline value) for the exploratory DCE-MRI variables PDQ score < 13 PDQ score 13–18 PDQ score ≥ 19 p value Baseline values Nvoxel in ml wrist 13.17 (5.53–29.40) 14.78(7.22–25.84) 14.21(8.28–22.70) 0.93 Nvoxel in ml MCP 1.65 (0.47–6.57) 3.12(0.80–9.22) 1.69(0.00–3.91) 0.56 IRE x Nvoxel in ml wrist 0.15 (0.05–0.67) 0.25(0.04–0.67) 0.19(0.06–0.56) 0.91 IRE x Nvoxel in ml MCP 0.02 (0.002–0.15) 0.06(0.003–0.24) 0.006(0.00–0.07) 0.31 ME x Nvoxel in ml wrist 21.43(9.76–62.29) 26.65(10.09–56.40) 26.73(14.72–42.67) 0.92 ME x Nvoxel in ml MCP 2.30 (0.67–14.23) 6.70(0.91–21.34) 2.46(0.00–7.42) 0.50 IRE x ME wrist 0.02 (0.01–0.05) 0.03(0.01–0.06) 0.02(0.01–0.05) 0.87 IRE x ME MCP 0.03 (0.01–0.09) 0.04(0.01–0.12) 0.01(0.00–0.05) 0.21 Mean changes Δ Nvoxel in ml wrist − 2.22 (− 4.84;0.41) − 2.47(− 6.93;1.99) − 2.87(− 10.37;4.63) 0.98 Δ Nvoxel in ml MCP − 1.99 (− 3.02;-0.96) − 1.72(− 3.47;0.03) −0.79 (− 3.74;2.17) 0.74 Δ IRE x Nvoxel in ml wrist −0.18(− 0.28;-0.07) −0.11(− 0.29;0.07) −0.17(− 0.46;0.13) 0.80 Δ IRE x Nvoxel in ml MCP −0.07 (− 0.11;-0.03) −0.06(− 0.13;0.003) −0.06(− 0.17;0.06) 0.99 Δ ME x Nvoxel in ml wrist −7.72 (− 14.69;-0.76) −5.55(− 17.37;6.27) −7.90(− 27.80;12.00) 0.95 Δ ME x Nvoxel in ml MCP −5.17(− 7.42;-2.92) − 4.82(− 8.64;-1.00) −3.12(− 9.58;3.33) 0.84 Δ IRE x ME wrist −0.01 (− 0.02;− 0.01) −0.01(− 0.02;0.003) -0.01(− 0.03;0.01) 0.81 Δ IRE x ME MCP − 0.03(− 0.04;-0.02) −0.03(− 0.06;-0.01) −0.03(− 0.07;0.01) 0.96 Baseline values are medians (25th, 75th percentiles) PDQ score < 13, n = 49; PDQ score 13–18, n = 17; PDQ score > 18: n =6 Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 8 of 11 Table 4 Multivariable regression models examining change expressed as least squares means (95% CI) Protocolized models PDQ score < 13 PDQ score 13–18 PDQ score > 18 p value Unadjusted model ΔDAS28 −1.30(− 1.64;-0.97) −1.17(− 1.74;-0.60) − 2.08(− 3.06;-1.10) 0.27 ΔRAMRIS Synovitis −1.68(− 2.30;-1.06) −1.75(− 2.80;-0.70) −2.46(− 4.25;-0.66) 0.72 ΔVAS pain −17.57(− 24.13;-11.02) − 25.45(− 36.56;-14.33) −38.13(− 57.15;-19.11) 0.09 Adjusted ΔDAS28 −1.49 (− 1.81;-1.18) −1.16(− 1.70;-0.61) −2.06(− 2.94;-1.17) 0.17 ΔRAMRIS Synovitis −1.78(− 2.44;-1.26) −1.72(− 2.86;-0.57) −2.48(− 4.34;-0.63) 0.74 ΔVAS pain − 19.36(− 25.89;-12.83) −24.71(− 36.07;-13.35) −38.04(− 56.50;-19.56) 0.15 No interaction was found in any of the models. The protocolized unadjusted models included the baseline PDQ classification groups as a trichotomous variable and RAMRIS synovitis score for the hand. The adjusted models further included female sex, disease duration, initiation group, and DAS28-CRP as covariates Protocolized models, n: PDQ score < 13; n = 51, PDQ score 13–18; n = 18, PDQ score > 18; n =6 In contrast, DAS28-CRP at baseline was a significant with notably the subgroup of patients with a high baseline predictor (p < 0.01) of change in the dependent variable PDQ score was small (n = 12). (DAS28-CRP, RAMRIS synovitis score, VAS pain) in all Inflammation was not only assessed clinically but also adjusted models. Furthermore, RAMRIS synovitis by MRI of one hand to ensure objectivity. Assessing one predicted DAS28-CRP change in the unadjusted model (p hand as a target joint area to reflect general inflammation = 0.01), and RAMRIS synovitis change both in the is a common procedure but can be a limitation to the re- unadjusted and the adjusted model (p ≤ 0.02). The impact gression analysis as it involves a risk of misestimating the of RAMRIS synovitis score on clinical change is not well inflammatory load especially in patients with a non-typical documented [38, 39]. Although our group previously has RA presentation. However, it has not yet been clarified found correlations between RAMRIS synovitis score and how many joints that should be included in a prognostic DAS28-CRP in another cohort of RA patients (in review), image analysis to reflect overall inflammatory load. this may be the first study to indicate a possible prognostic The study was limited by the framework of the PDQ; value of RAMRIS synovitis score in relation to DAS28- the original development for classification of neuropathic CRP change within 4 months. The clinical consistency pain among patients with various chronic pain conditions, and relevance of this finding needs further investigation. the algorithm for cutoff points also being validated in this Overall the DCE-MRI analyses did not add further sample and the lack of a clinical ‘gold standard’ assessing information to conventional MRI. augmented central pain mechanisms, against which the The unexpected behavior of the PDQ may lead to PDQ preferably should be tested. However, patients with speculations about whether the presents finding are due augmented central pain mechanisms express the same to the PDQ not being valid in this population. However, pain features as patients with neuropathic pain, though in the PDQ has, in at least three cross-sectional studies [13, a generalized pattern [12, 44–46]. In our opinion this 14, 36], indicated involvement of non-nociceptive pain sensitive but not specific feature of the questionnaire, mechanisms in subgroups of RA patients. Although we association to tender point count and sign of central failed to demonstrate a prognostic value of the PDQ, the sensitization in quantitative sensory testing and functional implications may still be that in patients with clear indica- MRI findings within other chronic pain conditions vouch tion of non-nociceptive pain mechanisms treatment for the use of the PDQ as indicator of augmented central strategies should involve management of chronic pain, pain processing [10–12, 47–49]. including treatment with classes of medication that target The relatively short follow-up time of 4 months is a central pain mechanisms [6, 7, 40–43], and not only focus limitation that may have reduced the treatment response on medical treatment of the underlying disease. for some patients. Further, the baseline visits including The main strength of this study was the prospective MRI could be up to 3 weeks after starting therapy for pa- design with a rigorous protocol and a prespecified analysis tients initiating a csDMARD and up to 1 week for patients plan. Furthermore, the sample of RA patients was heteroge- initiating a bDMARD, which also may have reduced the neous including patients from outpatient clinics and private observed treatment change from baseline to follow-up, rheumatologists in the Copenhagen area with a broad particularly on DCE-MRI that is known to be very sensi- spectrum of disease severity. Thus, on one hand this het- tive to fast inflammatory changes [31, 35, 50, 51]. Finally, erogeneity gave the results potential for generalizable inter- a direct pain modulating effect of anti-tumor necrosis pretation, on the other hand, it was the main limitation in factor (TNF)-alpha treatment is indicated by the literature relation to pain profiling, i.e. the constitution of the sample [52], and it could therefore be speculated that in some Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 9 of 11 patients treated with anti-TNF blocker the allowed delay for Videnskab og Kunst, Axel Muusfeldts Fond, Dagmar Marshalls Fond, Region Hovedstadens Forskningsfond, and Bjarne Jensens Fond. in baseline assessment may have influenced nociceptive signaling and thereby the study findings. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Conclusions In this study, higher numerically changes of inflamma- Authors’ contributions tory and patient-reported outcomes, and change of PDQ All authors participated in the design of the study. SRM, AWC and MB also classification-group at follow-up were observed in the participated in the data acquisition. All authors participated in drafting the manuscript and have approved it in its final form. group of patients with a high PDQ score at baseline, which may indicate that inflammatory pain can lead to Ethics approval and consent to participate pain hypersensitivity of a reversible character. Interest- This manuscript reports data from clinical research. The study was designed ingly, overall reduced change in variables was found in as a clinical prospective cohort study. It was approved by the Regional Ethics Committee of the Capital of Denmark April 18 2013; identification number patients in the intermediate PDQ classification group H-3-2013-049. with unclear pain mechanism indicating irreversible pain All participants had to sign a consent form. mechanisms. In contrast to our hypothesis, a high PDQ score had no prognostic value in relation to treatment Competing interests MB reports personal fees from Image Analysis, LTD, London, UK, outside the outcome specified as change in DAS28-CRP, RAMRIS submitted work. The remaining authors have no competing interests to score, DCE-MRI, or VAS pain 4 months after initiation declare. or escalation of medical therapy, however, it was found that the RAMRIS synovitis score may have prognostic Publisher’sNote value in relation to DAS28-CRP response. Further large- Springer Nature remains neutral with regard to jurisdictional claims in scale studies are needed to clarify the prognostic value published maps and institutional affiliations. of the PDQ. Author details The Parker Institute, Copenhagen University Hospital, Bispebjerg and Frederiksberg, 2000 Frederiksberg, Copenhagen, Denmark. Department of Additional files Rheumatology, Copenhagen University Hospital, Bispebjerg and Frederiksberg, Copenhagen, Denmark. Department of Rheumatology, Additional file 1: Table showing inter- and intra-reader reliability as Copenhagen University Hospital, Gentofte and Herlev, Hellerup, Denmark. intraclass correlation coefficient (ICC) for the DCE-MRI variables. (DOCX 16 kb) 4 Department of Radiology, Copenhagen University Hospital, Bispebjerg and Additional file 2: Table of multivariable regression models including Frederiksberg, Copenhagen, Denmark. DCE-MRI variables examining change across PDQ categories expressed as least squares means (95% CI). (DOCX 16 kb) Received: 1 November 2017 Accepted: 27 March 2018 Abbreviations References ANCOVA: Analysis of covariance; Anti-CCP: Anti-cyclic citrullinated peptide; 1. Schaible HG, von Banchet GS, Boettger MK, Brauer R, Gajda M, Richter F, et CRP: C-reactive protein; DAS28-CRP: Disease activity score 28 joints – C- al. The role of proinflammatory cytokines in the generation and reactive protein; DCE-MRI: Dynamic contrast- enhanced magnetic resonance maintenance of joint pain. Ann N Y Acad Sci. 2010;1193:60–9. imaging; DMARD: Disease-modifying antirheumatic drug; FM: Fibromyalgia; 2. Smolen JS, Landewe R, Breedveld FC, Buch M, Burmester G, Dougados M, et GAD-10: Generalized anxiety disorder assessment; GH: Global health; al. EULAR recommendations for the management of rheumatoid arthritis HAQ: Health-assessment questionnaire; IRE: Initial rate of enhancement; with synthetic and biological disease-modifying antirheumatic drugs: 2013 IQR: Interquartile range; MCP: Metacarpophalangeal; MCS: Mental component update. Ann Rheum Dis. 2014;73(3):492–509. score; MDI: Major depression inventory; ME: Maximum enhancement; 3. Meeus M, Vervisch S, De Clerck LS, Moorkens G, Hans G, Nijs J. Central MRI: Magnetic resonance imaging; MTX: Methotrexate; Nvoxel: Number of sensitization in patients with rheumatoid arthritis: a systematic literature enhancing voxel; PCS: Physical component score; PDQ: painDETECT review. Semin Arthritis Rheum. 2012;41(4):556–67. questionnaire; PRO: Patient-reported outcome; RA: Rheumatoid arthritis; 4. Lee YC, Cui J, Lu B, Frits ML, Iannaccone CK, Shadick NA, et al. Pain persists RAMRIS: Rheumatoid arthritis magnetic resonance imaging score; in DAS28 rheumatoid arthritis remission but not in ACR/EULAR remission: a RF: Rheumatoid factor; ROI: Region of interest; SF-36: 36-item short form longitudinal observational study. Arthritis Res Ther. 2011;13(3):R83. health survey; SJC: Swollen joint count; TJC: Tender joint count; TNF: Tumor 5. Wolfe F, Michaud K. Severe rheumatoid arthritis (RA), worse outcomes, necrosis factor; VAS: Visual analogue scale comorbid illness, and sociodemographic disadvantage characterize ra patients with fibromyalgia. J Rheumatol. 2004;31(4):695–700. Acknowledgements 6. Lee YC, Nassikas NJ, Clauw DJ. The role of the central nervous system in the The authors wish to thank Jack Olsen, Jette Nielsen, and Salomea Hirschorn for generation and maintenance of chronic pain in rheumatoid arthritis, assisting with the data acquisition, Christian Cato Holm for data management osteoarthritis and fibromyalgia. Arthritis Res Ther. 2011;13(2):211. and Robert C. Riis for sharing his DCE-MRI expertise. We also wish to thank the 7. Woolf CJ. Central sensitization: implications for the diagnosis and treatment Departments of Rheumatology at Bispebjerg and Frederiksberg Hospital, Herlev of pain. Pain. 2011;152(3 Suppl):S2–15. and Gentofte Hospital, and Køge Hospital for assisting with the recruitment of 8. Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO III, et al. participants. Finally, we wish to thank the Departments of Radiology at Bispebjerg 2010 Rheumatoid arthritis classification criteria: an American College of and Frederiksberg Hospital for carrying out the MRI scanning. Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62(9):2569–81. Funding 9. Freynhagen R, Baron R, Gockel U, Tolle TR. painDETECT: a new screening The study was supported by grants from The Oak Foundation, Selsbjerg questionnaire to identify neuropathic components in patients with back Holding, The Danish Rheumatism Association, Minister Erna Hamiltons Legat pain. Curr Med Res Opin. 2006;22(10):1911–20. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 10 of 11 10. Amris K, Jespersen A, Bliddal H. Self-reported somatosensory symptoms of 29. Haavardsholm EA, Ostergaard M, Ejbjerg BJ, Kvan NP, Uhlig TA, Lilleas FG, et neuropathic pain in fibromyalgia and chronic widespread pain correlate with al. Reliability and sensitivity to change of the OMERACT rheumatoid arthritis tender point count and pressure-pain thresholds. Pain. 2010;151(3):664–9. magnetic resonance imaging score in a multireader, longitudinal setting. 11. Gwilym SE, Keltner JR, Warnaby CE, Carr AJ, Chizh B, Chessell I, et al. Arthritis Rheum. 2005;52(12):3860–7. Psychophysical and functional imaging evidence supporting the 30. Boesen M, Kubassova O, Parodi M, Bliddal H, Innocenti S, Garlaschi G, et al. presence of central sensitization in a cohort of osteoarthritis patients. Comparison of the manual and computer-aided techniques for evaluation Arthritis Rheum. 2009;61(9):1226–34. of wrist synovitis using dynamic contrast-enhanced MRI on a dedicated scanner. Eur J Radiol. 2011;77(2):202–6. 12. Koroschetz J, Rehm SE, Gockel U, Brosz M, Freynhagen R, Tolle TR, et al. 31. Axelsen MB, Poggenborg RP, Stoltenberg M, Kubassova O, Boesen M, Fibromyalgia and neuropathic pain–differences and similarities. A comparison of 3057 patients with diabetic painful neuropathy and Horslev-Petersen K, et al. Reliability and responsiveness of dynamic contrast- enhanced magnetic resonance imaging in rheumatoid arthritis. Scand J fibromyalgia. BMC Neurol. 2011;11:55. Rheumatol. 2013;42(2):115–22. 13. Ahmed S, Magan T, Vargas M, Harrison A, Sofat N. Use of the painDETECT tool in rheumatoid arthritis suggests neuropathic and sensization 32. Ballegaard C, Riis RG, Bliddal H, Christensen R, Henriksen M, Bartels EM, et al. components in pain reporting. J Pain Res. 2014;7:579–88. Knee pain and inflammation in the infrapatellar fat pad estimated by 14. Koop SM, Ten Klooster PM, Vonkeman HE, Steunebrink LM, van de Laar MA. conventional and dynamic contrast-enhanced magnetic resonance imaging Neuropathic-like pain features and cross-sectional associations in in obese patients with osteoarthritis: a cross-sectional study. Osteoarthr rheumatoid arthritis. Arthritis Res Ther. 2015;17(1):237. Cartil. 2014;22(7):933–40. 15. Wu Q, Inman RD, Davis KD. Neuropathic pain in ankylosing spondylitis - a 33. Bandak E, Boesen M, Bliddal H, Riis RG, Gudbergsen H, Henriksen M. psychophysics and brain imaging study. Arthritis Rheum. 2013;65(6):1494–503. Associations between muscle perfusion and symptoms in knee osteoarthritis: a cross sectional study. Osteoarthr Cartil. 2015;23(10):1721–7. 16. Ostergaard M, Peterfy C, Conaghan P, McQueen F, Bird P, Ejbjerg B, et al. 34. Boesen M, Kubassova O, Bouert R, Axelsen MB, Ostergaard M, Cimmino MA, OMERACT Rheumatoid Arthritis Magnetic Resonance Imaging Studies. Core et al. Correlation between computer-aided dynamic gadolinium-enhanced set of MRI acquisitions, joint pathology definitions, and the OMERACT MRI assessment of inflammation and semi-quantitative synovitis and bone RA-MRI scoring system. J Rheumatol. 2003;30(6):1385–6. marrow oedema scores of the wrist in patients with rheumatoid arthritis–a 17. Axelsen MB, Stoltenberg M, Poggenborg RP, Kubassova O, Boesen M, cohort study. Rheumatology (Oxford). 2012;51(1):134–43. Bliddal H, et al. Dynamic gadolinium-enhanced magnetic resonance imaging allows accurate assessment of the synovial inflammatory activity 35. Cimmino MA, Parodi M, Zampogna G, Boesen M, Kubassova O, Barbieri F, et in rheumatoid arthritis knee joints: a comparison with synovial histology. al. Dynamic contrast-enhanced, extremity-dedicated MRI identifies synovitis Scand J Rheumatol. 2012;41(2):89–94. changes in the follow-up of rheumatoid arthritis patients treated with 18. Ostergaard M, Stoltenberg M, Lovgreen-Nielsen P, Volck B, Jensen CH, rituximab. Clin Exp Rheumatol. 2014;32(5):647–52. Lorenzen I. Magnetic resonance imaging-determined synovial membrane 36. Christensen AW, Rifbjerg-Madsen S, Christensen R, Dreyer L, Tillingsoe H, and joint effusion volumes in rheumatoid arthritis and osteoarthritis: Seven S, et al. Non-nociceptive pain in rheumatoid arthritis is frequent and comparison with the macroscopic and microscopic appearance of the affects disease activity estimation: cross-sectional data from the FRAME synovium. Arthritis Rheum. 1997;40(10):1856–67. study. Scand J Rheumatol. 2016;18:1–9. 19. Christensen AW, Rifbjerg-Madsen S, Christensen R, Amris K, Taylor PC, Locht H, 37. Christensen AW, Rifbjerg-Madsen S, Christensen R, Dreyer L, Boesen M, et al. Temporal summation of pain and ultrasound Doppler activity as Ellegaard K, et al. Ultrasound Doppler but not temporal summation of pain predictors of treatment response in patients with rheumatoid arthritis: protocol predicts DAS28 response in rheumatoid arthritis: a prospective cohort study. for the Frederiksberg hospitals Rheumatoid Arthritis, pain assessment and Rheumatology (Oxford). 2016;55(6):1091–8. Medical Evaluation (FRAME-cohort) study. BMJ Open. 2014;4(1):e004313. 38. Ostergaard M, Moller-Bisgaard S. Optimal use of MRI in clinical trials, clinical care and clinical registries of patients with rheumatoid arthritis. Clin Exp 20. Rifbjerg-Madsen S, Christensen AW, Boesen M, Christensen R, Danneskiold- Rheumatol. 2014;32(5 Suppl 85):S-22. Samsoe B, Bliddal H, et al. Can the painDETECT Questionnaire score and MRI help predict treatment outcome in rheumatoid arthritis: protocol for the 39. Suter LG, Fraenkel L, Braithwaite RS. Role of magnetic resonance imaging in Frederiksberg hospital's Rheumatoid Arthritis, pain assessment and Medical the diagnosis and prognosis of rheumatoid arthritis. Arthritis Care Res Evaluation (FRAME-cohort) study. BMJ Open. 2014;4(11):e006058. (Hoboken ). 2011;63(5):675–88. 21. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, 40. Brendbekken R, Eriksen HR, Grasdal A, Harris A, Hagen EM, Tangen T. et al. The American Rheumatism Association 1987 revised criteria for the Return to work in patients with chronic musculoskeletal pain: classification of rheumatoid arthritis. Arthritis Rheum. 1988;31(3):315–24. multidisciplinary intervention versus brief intervention: a randomized 22. Rifbjerg-Madsen S, Waehrens EE, Danneskiold-Samsoe B, Amris K. clinical trial. J Occup Rehabil. 2017;27(1):82-91. Psychometric properties of the painDETECT questionnaire in rheumatoid 41. Tse MM, Yeung SS, Lee PH, Ng SS. Effects of a peer-led pain arthritis, psoriatic arthritis and spondyloarthritis: Rasch analysis and management program for nursing home residents with chronic pain: a test-retest reliability. Health Qual Life Outcomes. 2017;15(1):110. pilot study. Pain Med. 2016;18. 42. Amris K, Luta G, Christensen R, Danneskiold-Samsoe B, Bliddal H, Waehrens EE. 23. Gudbergsen H, Bartels EM, Krusager P, Waehrens EE, Christensen R, Predictors of improvement in observed functional ability in patients with Danneskiold-Samsoe B, et al. Test-retest of computerized health status fibromyalgia as an outcome of rehabilitation. J Rehabil Med. 2016;48(1):65-71. questionnaires frequently used in the monitoring of knee osteoarthritis: a randomized crossover trial. BMC Musculoskelet Disord. 2011;12:190. 43. Amris K, Waehrens EE, Christensen R, Bliddal H, Danneskiold-Samsoe B. Interdisciplinary rehabilitation of patients with chronic widespread pain: 24. Thorsen H, Hansen TM, McKenna SP, Sorensen SF, Whalley D. primary endpoint of the randomized, nonblinded, parallel-group IMPROvE Adaptation into Danish of the Stanford Health Assessment trial. Pain. 2014;155(7):1356–64. Questionnaire (HAQ) and the Rheumatoid Arthritis Quality of Life Scale 44. Baron R, Binder A, Wasner G. Neuropathic pain: diagnosis, pathophysiological (RAQoL). Scand J Rheumatol. 2001;30(2):103–9. mechanisms, and treatment. Lancet Neurol. 2010;9(8):807–19. 25. Bjorner JB, Thunedborg K, Kristensen TS, Modvig J, Bech P. The Danish 45. Bennett MI, Attal N, Backonja MM, Baron R, Bouhassira D, Freynhagen R, SF-36 Health Survey: translation and preliminary validity studies. J Clin et al. Using screening tools to identify neuropathic pain. Pain. 2007; Epidemiol. 1998;51(11):991–9. 127(3):199–203. 26. Bech P, Kastrup M, Rafaelsen OJ. Mini-compendium of rating scales for 46. Rehm SE, Koroschetz J, Gockel U, Brosz M, Freynhagen R, Tolle TR, et al. A states of anxiety depression mania schizophrenia with corresponding cross-sectional survey of 3035 patients with fibromyalgia: subgroups of DSM-III syndromes. Acta Psychiatr Scand Suppl. 1986;326:1–37. patients with typical comorbidities and sensory symptom profiles. 27. Bech P, Rasmussen NA, Olsen LR, Noerholm V, Abildgaard W. The Rheumatology (Oxford). 2010;49(6):1146–52. sensitivity and specificity of the Major Depression Inventory, using the 47. Hochman JR, Gagliese L, Davis AM, Hawker GA. Neuropathic pain symptoms Present State Examination as the index of diagnostic validity. J Affect in a community knee OA cohort. Osteoarthr Cartil. 2011;19(6):647–54. Disord. 2001;66(2–3):159–64. 28. Ejbjerg B, McQueen F, Lassere M, Haavardsholm E, Conaghan P, O'Connor P, 48. Hochman JR, Davis AM, Elkayam J, Gagliese L, Hawker GA. Neuropathic pain et al. The EULAR-OMERACT rheumatoid arthritis MRI reference image atlas: symptoms on the modified painDETECT correlate with signs of central the wrist joint. Ann Rheum Dis. 2005;64(Suppl 1):i23–47. sensitization in knee osteoarthritis. Osteoarthr Cartil. 2013;21(9):1236–42. Rifbjerg-Madsen et al. Arthritis Research & Therapy (2018) 20:105 Page 11 of 11 49. MoretonBJ, TewV,das NR WM,Walsh DA,Lincoln NB.Pain phenotype in patients with knee osteoarthritis: classification and measurement properties of painDETECT and self-report Leeds assessment of neuropathic symptoms and signs scale in a cross-sectional study. Arthritis Care Res (Hoboken ). 2015;67(4):519–28. 50. Boesen M, Kubassova O, Cimmino MA, Ostergaard M, Taylor P, Danneskiold-Samsoe B, et al. Dynamic contrast-enhanced MRI can monitor the very early inflammatory treatment response upon intra- articular steroid injection in the knee joint: a case report with review of the literature. Arthritis. 2011;2011:578252. 51. Conaghan PG, Ostergaard M, Bowes MA, Wu C, Fuerst T, van der Heijde D, et al. Comparing the effects of tofacitinib, methotrexate and the combination, on bone marrow oedema, synovitis and bone erosion in methotrexate-naive, early active rheumatoid arthritis: results of an exploratory randomised MRI study incorporating semiquantitative and quantitative techniques. Ann Rheum Dis. 2016;75(6):1024–33. 52. Hess A, Axmann R, Rech J, Finzel S, Heindl C, Kreitz S, et al. Blockade of TNF-alpha rapidly inhibits pain responses in the central nervous system. Proc Natl Acad Sci U S A. 2011;108(9):3731–6. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries � Our selector tool helps you to find the most relevant journal � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Journal

Arthritis Research & TherapySpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off