Evidence Based Chronic Pain Management

Chapter 8 the frequency of the condition. Nevertheless, it is Risk factors clear that nonspecific neck pain including whiplash places a heavy burden on individuals, employers, and Natural history of nonspecific neck pain healthcare services. and whiplash with factors associated with chronic disability Nontraumatic causes About two-thirds of the population will experience Nontraumatic causes neck pain at some time in their lives [7, 8], with the Nonspecific neck pain usually resolves within days or condition being most common in middle age, and weeks, but can recur or become chronic. Once pain in women [9]. The reported prevalence of neck pain becomes persistent, outcome is more unpredictable, varies widely between studies, but has a mean point and there is little consistency in the literature regard- prevalence of 7.6% (range 5.9–38.7%) and a mean ing the duration of symptoms and factors that influ- lifetime prevalence of 48.5% (range 14.2–71%) ence outcome. A systematic review of the clinical [9]. A UK survey found that 18% of 7669 adults course and prognostic factors in nonspecific neck pain had neck pain at the time of the survey, but when found little consensus as to outcome or relevant prog- symptomatic people were re-questioned 1 year later nostic factors, although this was based on poor-qual- (58% responded), half were still symptomatic [10]. ity studies [19]. The systematic review found evidence A Norwegian survey of 10,000 adults also reported that in patients with chronic pain treated in second- that 34% of responders had experienced neck pain ary care or an occupational setting, 20–78% (median in the previous year [11]. Neck pain is second only to 54%) of patients remained symptomatic, irrespective back pain in frequency of musculoskeletal consulta- of the therapy given. Six of the included studies docu- tion in primary care. mented prognostic factors, and the severity of pain at presentation was the best predictor of a poor outcome, Whiplash although previous episodes of neck pain were also Although whiplash injury is very common through- important. Three subsequent studies also considered out the world, the incidence of reported symptoms the factors at presentation which might influence out- and patients who go on to chronic disability or to come at 1 year, and found older age, and concomitant seek compensation varies greatly between coun- low back pain [10, 20] and severity and duration of the tries and even between different regions or social pain [21] to be significant. Patients with chronic spinal groups within the same country. There is no con- conditions were also found to have other chronic pain sistency in the literature about the epidemiology syndromes (69%), chronic physical conditions (55%) and natural history of whiplash, partially because or psychologic problems (35%)[22, 23]. of the poor quality of studies [12, 13]but more specifically because of the complex interactions Neck pain with neurological complications between the individual, legal, economic, and soci- Many patients with neurological abnormality as a etal factors, which may influence presentation and result of nonspecific neck pain will require MRI scan- outcome [12, 13]. In some countries, like Lithuania ning of the cervical spine at an early stage, particularly and Greece where there is no litigation or compen- if there is progressive myelopathy or intractable pain. sation culture, chronic whiplash is very rare and Radiculopathy generally has a favorable outcome, outcome is universally favorable [14, 15]. In other although recovery can be slow. The result of decom- countries like the USA, Canada, The Netherlands, pressive surgery for myelopathy complicating nonspe- Australia and the UK, much higher proportions of cific neck pain is often disappointing. While the rate patients develop chronic whiplash and suffer pro- of progression of the neurological loss may be slowed longed disability [16]. Internationally, whiplash by the surgery, the lost function may not recover or lasting for over 6 months varies between 2% and symptoms may progress at a later date. The poor out- 58% [13, 17] but typically lies between 20% and come following surgery may reflect the irreversible 40% [18]. damage to the cervical cord or compromise to the vas- cular supply to the cord immediately or subsequently. 84

Chronic neck pain and whiplash Whiplash Box 8.1 Therapeutic options for The prognosis for whiplash is also generally favouable nonspecific neck pain [12] but as already mentioned, shows great variability as to the frequency, severity and duration of disability. Likely to be effective This variability in outcome is at least in part related to the culture of litigation and compensation [1] but this Exercise cannot explain all the differences, particularly within Manual therapy (mobilization or manipulation) the same population. Two comprehensive systematic Exercise plus manual therapy reviews [12, 13] found little consistency as to factors influencing outcome, although societal factors like lit- Likely to be ineffective igation and compensation culture were most impor- tant [1, 24, 25]. How other societal factors influence Patient education outcome is complex and even more poorly under- Heat stood [26]. Systematic reviews of prognostic factors following whiplash injury [12, 13, 17] found conflict- Unknown effectiveness ing evidence for pre-existing physical or psychologic factors, or crash-related factors. The most consistent Multimodal therapy predictors of an unfavorable outcome from the sys- Traction tematic reviews and subsequent studies were severity Acupuncture of pain, headache and disability at presentation. Pulsed electromagnetic field therapy Transcutaneous electrical nerve stimulation Drug treatments Soft collar and special pillows Biofeedback Spray and stretch Therapeutic interventions for Box 8.2 Therapeutic options neck pain for neck pain complicated by radiculopathy Methods Literature searches on therapeutic options for treat- Unknown effectiveness ing mechanical neck pain were carried out using the following databases: Chirolars (now called Mantis), Epidural injection Bioethicsline, CINAHL, Current Contents, and Physical therapies Medline, with data being used to prepare and update Immobilization in a collar an article in Clinical Evidence [2]. I will summarize the evidence on treatment modalities currently in Box 8.3 Therapeutic options for use, with an indication of questions which still need acute and chronic whiplash to be answered. Studies relating to specific condi- tions like fibromyalgia and disk prolapse will not be Acute whiplash discussed. Likely to be effective Data on therapy will be considered for patients Early exercise or mobilization with (uncomplicated) nonspecific neck pain, neck pain plus radiculopathy, and whiplash. Therapies will Unknown effectiveness then be categorized as “likely to be effective,” where Early return to normal activities there is at least one high-quality RCT suggesting Early home exercise benefit and reasonable consensus from other studies; Pulsed electromagnetic field therapy “likely to be ineffective,” where there is at least one Multimodal therapy high-quality study suggesting a lack of benefit from Drug treatments the treatment; and “unknown effectiveness,” where there is insufficient or conflicting evidence without Chronic whiplash a consensus (Boxes 8.1–8.3). Relevant systematic Unknown effectiveness Percutaneous radiofrequency neurotomy Multimodal therapy Physical therapies 85

Chapter 8 reviews with be mentioned, but with data mainly management: an RCT (103 women with chronic from the higher quality RCTs. work-related neck pain) [33] compared three exercise regimens (strength training, endurance Most neck pain appears to respond to conservative training, co-ordination exercises) compared to measures, although the effect size is often quite small, stress management over 10 weeks. It found that and the optimal therapeutic approach for uncom- any type of exercise significantly reduced pain plicated neck pain has yet to be established. Even compared with stress management after 10–12 where an initial benefit is shown, this advantage is weeks (P Ͻ 0.05), but there was no significant not sustained. Few modalities of treatment have been difference in outcomes between the different exer- assessed in high-quality randomized studies, but I will cise programs. There was also no significant dif- try to present the best available evidence for the most ference in neck pain among the four groups after commonly used modalities. The evidence is often 3 years [34]. Another RCT (180 women with contradictory because of the poor quality of many chronic neck pain) assessed the rate of change in of the studies, use of interventions in combination, neck strength, pain and disability of a 1-year train- and diverse patient groups. The lack of consistency in ing program comparing high-intensity strength study design makes it difficult to identify which inter- training or lower intensity endurance training vention may be of use in which type of patients. to a control group. The greatest improvement in strength was achieved within the first 2 months for Nonspecific neck pain (see Box 8.1) both treated groups, but with improvement contin- uing to a year. The decrease in pain at 12 months Therapies likely to be effective was 69% for strength training, 61% for endurance 1. Exercise training and 28% for controls, compared to base- Systematic reviews [2,27-30] identified RCTs using line, all being significant (P Ͻ 0.001). The number different exercise strategies, but none of the reviews of patients who were pain free or nearly pain free at could perform a meta-analysis because of heteroge- 12 months was significantly greater for the treated neity among the trials in types of exercise and study groups (53% and 49% respectively) compared to designs. the controls (20%) [35]. However, this study does not describe how randomization and blinding were Positive studies achieved. • Proprioceptive and strengthening exercise versus • Exercise plus infrared versus transcutaneous electrical nerve stimulation (TENS) plus infra- usual care: one RCT (60 people with chronic neck red versus infrared alone: one RCT (218 patients pain) [31] found a proprioceptive and strengthen- with chronic neck pain) [36] compared the effects ing exercise program to be significantly more effec- of twice-weekly therapy for 6 weeks using inten- tive (P Ͻ 0.004) at reducing pain at 10 weeks when sive exercise plus infrared, TENS plus infrared, and compared with usual care (analgesics, nonsteroidal infrared alone. The RCT found that the addition of anti-inflammatory drugs or muscle relaxants). exercise or TENS significantly improved pain com- • Endurance exercise versus strengthening exercise pared to infrared alone at 6 weeks (P ϭ 0.02) and 6 versus no specific exercise program: an RCT (180 months (P ϭ 0.019), but with no difference between female office workers with chronic neck pain) [32] the two combination treatment groups. compared a program of “endurance” (dynamic) or • Exercise plus special pillow versus either alone “strength” (isometric) exercises to a control group versus control (hot/cold pack plus massage): an (no specific exercise), with exercise being carried RCT (151 patients with chronic neck pain) [37] out three times a week for 1 year. Both endurance compared the effects of exercise plus a special and strength exercises significantly improved neck pillow to exercise or a pillow alone or a control pain and disability after 12 months compared to group and found a significant advantage for the the control therapy (P Ͻ 0.001 for exercise groups combination group at 6 weeks, although the dif- versus control). ference was small. • Strength training versus endurance training versus co-ordination exercises versus stress 86

Chronic neck pain and whiplash Negative study [47] compared a single manipulation treatment • Dynamic muscle training versus relaxation versus a single mobilization treatment and found no significant difference between them in imme- training versus advice to continue with ordi- diate improvement in pain (85% with manipula- nary activity: one RCT (393 women office work- tion, 69% with mobilization). In this study, there ers with chronic neck pain) [38] compared three was a transient increase in pain in 5% of people interventions for 12 weeks: dynamic muscle train- receiving manipulation and 6% receiving mobili- ing, relaxation training, and advice to continue zation. The second RCT (336 people with acute or with ordinary activities, and found no significant chronic neck pain) [48] found no significant dif- difference in outcome in pain or disability in the ference between manipulation and mobilization three groups at 3, 6, and 12 months of follow- in “average” pain, “severe” pain and neck disability up. However, the average number of 30-minute scores between a variable number of chiropractic training sessions completed by participants over mobilizations and a variable number of manipu- 12 weeks for both treatment groups was only 40% lations after 6 months. In this RCT, only 35% of of the maximum available, and this might have eligible people agreed to participate, and this may influenced the result. reduce the external validity of the study. A follow- up questionnaire of adverse effects at 2 weeks [49] 2. Manual therapy (manipulation and mobilization found that 30% of the 280 people who responded physiotherapy) reported at least one minor adverse effect such as Systematic reviews [2, 27, 28, 39–44] identified a increased pain or headache associated with manip- number of RCTs comparing manipulation and/or ulation and less commonly with mobilization. mobilization with each other or with other treatments. In the third RCT (70 patients with chronic neck pain) [50], patients received one manipulation or Positive studies one mobilization with assessment before the ther- • Manipulation or mobilization versus other physi- apy and 5 minutes after treatment. Both groups showed significant improvement in pain and range cal treatments, versus usual care versus placebo: of movement, but with a greater benefit for the one RCT (256 people with chronic neck and back manipulation group. pain; 64 with neck and 48 with neck and back • Mobilization versus exercise versus usual care: involvement) [45] compared four treatment groups: one RCT (183 people with neck pain for Ͼ2 weeks) manual treatment (mobilization, manipulation [51] compared three 6-week courses of treatment or both); physical treatments at the discretion of with mobilization, exercise or usual care, with treat- the physiotherapist; usual GP care; and placebo. It ment “success” being defined as “much improved” found that manual treatment significantly improved or “completely recovered.” The RCT found “success” outcomes after 12 months compared with all the to be significantly more common at 7 weeks with other treatments (statistical analysis for people with mobilization compared to exercise or usual care, but neck pain alone was not reported). It was not pos- there was no difference between exercise and usual sible to directly compare the effects of mobilization care. Long-term follow-up of this RCT [52] found versus manipulation. that mobilization was still superior at 26 weeks, but • Manipulation versus mobilization versus exer- not at 1 year. cise: one RCT (119 people with chronic neck pain) [46] compared three treatments: mobilization, Negative studies manipulation, and intensive exercise training. It • Manipulation versus diazepam, anti-inflammatory found no significant difference in pain among the three groups by the end of treatment or after drugs or usual care: one review [40] performed a 12 months, although pain score improved signifi- meta-analysis of three RCTs (155 people with chronic cantly from baseline in all groups. neck and back pain) comparing manipulation to • Manipulation versus mobilization: three RCTs diazepam, anti-inflammatory drugs or usual care. compared manipulation to mobilization. The first It found no significant difference in improvement RCT (100 people with acute or chronic neck pain) 87

Chapter 8 in pain at 3 weeks between manipulation and other found superior pain reduction for the combined treat- treatments, although all treatments improved pain. ment or exercise groups compared to manipulation However, the meta-analysis may have been under- (P ϭ 0.04). powered to detect a clinically important difference. • McKenzie mobilization versus general exercise ver- Negative study sus placebo: one small low-quality RCT (77 people) • Manipulation, mobilization or shortwave dia- [53] compared McKenzie mobilization, exercise and placebo, and found no significant difference in pain thermy plus exercise and advice: one pragmatic between the groups at 6 months and 12 months. multicenter RCT (350 patients with chronic neck • Adverse events associated with manipulation: pain) [57] found no benefit from the addition of manipulation has been associated with occasional manual therapy (63% had mobilization physiother- serious neurologic complications and the estimated apy) or pulsed shortwave diathermy to advice plus risk from case reports of cerebrovascular accident is exercise at 6 weeks or 6 months. 1–3/million manipulations [54], while the estimated risk of all serious adverse effects (such as death Therapies likely to be ineffective or disk herniation) is 5–10/10 million manipula- 1. Patient education alone tions [40]. Two RCTs in people with chronic neck, back or shoul- der pain found no significant benefit from patient 3. Manual therapy plus exercise education (individual advice, pamphlets or group Two systematic reviews by the same Cochrane group instruction) with or without analgesics, stress man- reviewed exercise therapy [30] and manual therapy agement or cognitive behavioral therapy. [41] in patients with nonspecific neck pain, neck pain • Educational pamphlet versus more extensive infor- with radiculopathy and whiplash, and found the best evidence of efficacy was for the combination of manual mation versus cognitive behavioral therapy (CBT): therapy (mobilization or manipulation) with exercise the first RCT (243 people with neck and back pain) when compared with any other treatments. However, [58] compared three interventions: an educational the review did not provide a subgroup analysis in peo- pamphlet, a more extensive information program, ple with uncomplicated nonspecific neck pain. and six sessions of CBT, and found no significant difference among treatments. However, post hoc Positive study analysis suggested that CBT significantly reduced • Manipulation plus strengthening exercises versus time off work compared with an educational pam- phlet (P ϭ 0.05). either treatment alone: one RCT (191 people with • Individualized education plus exercise program chronic neck pain) [55] compared three treatments: versus stress management versus no intervention: low-technology strengthening exercises plus manipu- the second RCT (282 nurses with neck, shoulder or lation (combined treatment), high-technology MedX back pain in the preceding 12 months) [59] com- strengthening exercises (exercise), and manipulation pared three interventions: an individualized edu- alone (manipulation). The RCT found that the com- cation and exercise program, stress management, bined treatment significantly improved patient sat- and no intervention. The RCT found no significant isfaction, objective strength, and range of movement difference in pain among the groups immediately (P Ͻ 0.05) compared with manipulation after 11 after treatment, or at 12 and 18 months. weeks. The RCT also found that both the combined treatment and exercise significantly improved pain 2. Heat and patient satisfaction compared to manipulation Systematic reviews [2, 27, 28] identified two RCTs after 1 year, although it found no significant difference suggesting that heat was less effective than other among treatments in health status, neck disability or therapies in people with uncomplicated neck pain. medication use. The 2-year follow-up to this RCT One RCT of people with chronic neck and back pain (data available for 76% of the original patients) [56] [45] found that heat combined with other physical 88

Chronic neck pain and whiplash treatment was less effective in improving outcomes different control procedures (sham acupuncture, than manipulation or mobilization (see “Manual sham TENS, diazepam, traction, short-wave therapy”). The second RCT [36] found infrared less diathermy, and mobilization) in people with acute effective when used alone than when combined with or chronic neck pain. None of the reviews was able exercise or TENS (see “Exercise”). to perform a meta-analysis, and the RCTs found no consistent difference in pain between acupuncture Therapies of unknown effectiveness and any of the other interventions. The quality of the 1. Multimodal treatment studies was considered “disappointing.” Systematic reviews [2, 60] identified two RCTs which provided insufficient evidence to assess the benefit or 4. Pulsed electromagnetic field treatment (PEMF) cost-effectiveness of multimodal treatment in people Systematic reviews [2, 27, 28, 68] identified one RCT of with uncomplicated neck pain. moderate quality comparing PEMF versus sham PEMF • Multimodal treatment versus CBT versus mini- in people with chronic neck pain. The RCT (81 peo- ple with neck pain and 86 people with osteo-arthritis mal treatment: one RCT (185 patients with non- of the knee) [69] compared true to sham PEMF. specific neck [87%] and/or back [91%] pain) [61] Subgroup analysis in people with chronic neck pain compared minimal therapy (advice to keep active), found that PEMF significantly reduced pain (P Ͻ 0.04) CBT (six sessions of structured CBT), and CBT and pain on passive motion (P ϭ 0.03), compared with plus physical therapy (same regimen as CBT group sham PEMF, but there was no difference in a range of plus physical training). The RCT found that mini- other parameters. Although randomization was con- mal treatment significantly increased the risk of ducted appropriately, baseline characteristics of treated being off work for 15 or more days compared with and placebo groups were, by chance, different and it is CBT plus physical training. It found no significant not clear how much of the observed effect was caused difference in sick leave between CBT and CBT plus by bias introduced by the baseline differences. physical therapy. However, the RCT included peo- ple with neck pain, back pain or both, and did not 5. Transcutaneous electrical nerve stimulation separately report results for those with neck pain (TENS) alone. Systematic reviews [2, 27, 28, 68] identified one RCT • Different forms of multimodal therapy: the sec- [36] that found that TENS plus infrared was equally ond RCT (66 people with chronic neck and shoul- effective to infrared plus exercise, and superior to der pain) [62] compared exercise plus behavioral infrared alone at 6 weeks (see “Exercise”). modification (patient education and advice, with a psychologist acting as an advisor to other staff) 6. Drug treatments (analgesics, NSAIDs or muscle to exercise plus CBT (with CBT administered by relaxants, tricyclic antidepressants) a psychologist). It found no significant difference No systematic review or RCTs examining the effects between the interventions in pain or time off work of drug treatments in people with nonspecific neck after 6 months. pain were found. 2. Traction 7. Facet joint injection Systematic reviews [2, 27, 28, 63, 64] identified two Two systematic reviews [70, 71] found moder- RCTs comparing traction versus sham traction, pla- ate evidence of efficacy for facet joint block using cebo tablets, exercise, acupuncture, heat, collar, and medial branch blocks, but no evidence for cervical analgesics. The RCTs found no consistent differ- intra-articular facet joint blocks. One RCT of medial ence in pain between traction and any of the other branch blocks of the facet joints (60 patients with interventions. facet joint disease confirmed by diagnostic medial branch block) were randomized to receive one of four 3. Acupuncture preparations with the medial branch blocks: bupi- Systematic reviews [2, 27, 28, 65–67] identified 14 vacaine alone, bupivacaine plus sarapin, bupivacaine RCTs comparing needle or laser acupuncture with 89

Chapter 8 plus betamethasone, or all three agents together, with (52 people with chronic cervical brachialgia) [77] the same injections being repeated as required over compared cervical epidural steroid plus lidocaine the next year. Significant pain relief (Ͼ50%) and injection to similar injection into the posterior neck improved disability were observed in 80–87% of all muscles, and found that more people receiving the patients at 3 months, 80–93% at 6 months and 87– epidural injections had reduced pain at 1 year (68% 93% at 12 months, but with no difference between with epidural versus 12% with control), although the treatment groups or patients who did or did not the significance was not reported. receive the steroid. The average number of treatments • Epidural steroid plus lidocaine versus epidural ster- per patient was 3.8 ϩ/Ϫ0.7 in the nonsteroid groups, oid plus lidocaine plus oral morphine: the second and 3.4 ϩ/Ϫ1.0 in the steroid groups, with no signifi- RCT (24 people with cervical radiculopathy for Ͼ1 cant difference among the groups [72]. year) [78] found similar success rates over 1 year for epidural steroid (triamcinolone) alone or when used 8. Soft collars, special pillows, biofeedback, spray with oral morphine (78.5% with epidural alone ver- and stretch sus 80% with epidural plus morphine). There are no data on these measures in patients with nonspecific neck pain. 2. Conservative versus surgical treatment Systematic reviews [2, 79] identified one RCT (81 people Neck pain with radiculopathy (see Box 8.2) with severe radicular symptoms for at least 3 months) [80] comparing three interventions: surgery, physical Therapies of unknown effectiveness treatments, and immobilization. It found no significant Neck pain with radiculopathy usually has a favorable difference among treatments in symptoms after 1 year. prognosis without the need for surgical intervention, but there are very few studies looking at conserva- Treatment of acute whiplash to prevent tive approaches to therapy, like epidural injection chronic disability (see Box 8.3) or a comparison between conservative and surgical The studies of acute whiplash are included to see if treatments. early treatment can influence the development of pro- longed disability. Although there is some evidence that 1. Cervical epidural injection (interlaminar early active interventions can reduce disability, the or transforaminal) studies are insufficiently robust to confirm this [18]. Systematic reviews [2, 73–75] identified two small poor-quality RCTs which provided insufficient evi- Therapies likely to be effective dence on the effects of cervical epidural interlami- 1. Early physiotherapy (mobilization and/or exercises) nar steroid injections for radiculopathy complicating versus immobilization or less active treatment nonspecific neck pain. One RCT (40 patients with Systematic reviews [2, 12, 18, 81] identified RCTs com- radiculopathy confirmed by MRI and diagnostic paring early physiotherapy to less active treatments. transforaminal block) had one transforaminal injec- tion of either steroid or saline plus local anesthetic Positive studies with weekly assessments for 3 weeks. Six of 20 patients • Early mobilization versus immobilization in a col- in each group reported improvement at 3 weeks, with no difference between the groups for any parameter lar: three studies found early mobilization to be at any time [76]. Epidural injections are more inva- more effective than immobilization in a collar. The sive in the cervical than lumbar region, and need to first RCT (61 people with acute whiplash) [82] com- be used with caution. Complications, such as infec- pared early mobilization versus immobilization in tion or abscess formation, have been documented fol- a collar plus rest for 14 days followed by gradual lowing the procedure [73]. mobilization. It found that early mobilization signifi- • Cervical epidural versus posterior neck muscle cantly improved pain and range of movement after 4 and 8 weeks compared with immobilization plus less injection of steroid plus lidocaine: the first RCT active treatment (P Ͻ 0.01). In the second RCT (97 90

Chronic neck pain and whiplash people with acute whiplash) [83], there were early compared active physiotherapy (exercise or mobiliza- benefits in pain relief and improved movement with tion) to GP care, with both groups receiving advice on early mobilization compared with immobilization in graded activity. There was substantial improvement in a collar, but similar pain relief after 12 weeks (propor- both groups over time, but no statistically significant tion with neck pain: 2% with mobilization versus 16% differences between the two groups for pain, headache with collar). The RCT did not assess the significance or work disability at 8, 12, 26 or 52 weeks. However, of the difference between the groups. In the third RCT treatment was delayed for up to 4 weeks after the (200 people treated within 48 hours of injury) [84], injury and this might have influenced the result. instruction on mobilization (2–5 physiotherapy vis- its in the first week) was compared to immobilization Therapies of unknown effectiveness with a soft collar. Mobilization significantly reduced 1. Early return to normal activity the proportion of people with neck pain or disability Systematic reviews [2, 18, 81] identified one RCT at 6 weeks compared to those treated with a soft col- (201 people with acute whiplash) [89] which com- lar. However, by 6 weeks, 36% of the collar group and pared advice to “act as usual” versus immobilization 15% of the exercise group had dropped out, and an plus 14 days’ sick leave. The RCT found that advice to intention-to-treat analysis was not carried out. “act as usual” significantly improved some symptoms • Physical treatments versus advice on early mobili- (pain, neck stiffness, headache, poor memory and zation versus rest for 7–14 days followed by mobi- concentration) after 6 months compared with immo- lization: this RCT (247 people with acute whiplash) bilization. However, there was no difference in neck [85] compared three interventions: advice on early range or length of sick leave between the treatment mobilization, physical treatments, or rest for 7–14 groups. Furthermore, a similar proportion of people days followed by gentle mobilization. All participants had severe neck pain after 6 months (11% with “act were given a soft collar and analgesics. Follow-up at as usual” versus 15% with immobilization). 2 years of 167 people responding to a questionnaire found that advice on early mobilization significantly 2. Early home exercise reduced the proportion of people who still had symp- Systematic reviews [2, 18, 29, 81] identified one RCT toms compared with physical treatments or rest (P ϭ (59 people with acute whiplash) [90] which com- 0.02 for early mobilization versus other treatments). pared two home exercise regimens: regular exercise • Immediate mobilization versus mobilization after versus the same exercise regimen plus isometric exer- 96 hours delay versus rest plus a collar: one RCT cises at least three times daily. The RCT found no sig- (97 people with acute whiplash) [86] found that nificant difference between treatments in disability or mobilization (home mobilization and exercise) sig- pain after 3 or 6 months. nificantly improved pain compared with rest plus a collar (P Ͻ 0.001), but only if mobilization was 3. Pulsed electromagnetic field treatment (PEMF) started immediately after injury. If mobilization was Systematic reviews [2, 12, 18, 81] identified one small delayed by more than 96 hours, there was no signifi- RCT (40 people with acute whiplash) [91] which com- cant difference between treatments after 6 months. pared PEMF with sham PEMF, and found the active However, a 3-year follow-up of this RCT [87] found therapy significantly more effective after 4 weeks that mobilization significantly reduced pain and (P Ͻ 0.05), but not after 3 months. sick leave (P Ͻ 0.05) compared with rest plus a neck collar, even if it was delayed for 2 weeks, although 4. Multimodal treatment only people who had received active intervention Systematic reviews [2, 18, 81] identified one RCT (60 within 96 hours had a total cervical range similar to people with whiplash in the previous 2 months) [92] matched controls. which compared multimodal treatment (postural training, psychologic support, eye fixation exercises, Negative study and manual treatment) with physical treatments. The • Active physiotherapy versus GP care: this RCT (80 RCT found that multimodal treatment significantly reduced pain by the end of treatment (P Ͻ 0.05) and patients with whiplash persisting to 4 weeks) [88] 91

Chapter 8 after 1 and 6 months (P Ͻ 0.001) compared with phys- multimodal treatment were satisfied with pain ical treatments. There was a similar benefit for time to control at the end of treatment and their ability to return to work. Further study is necessary to determine perform activities at 3 months (P Ͻ 0.05). if multimodal treatment is a cost-effective approach. • Exercise plus advice versus advice alone: one RCT (134 patients with whiplash of 3–12 months’ dura- 5. Drug treatments (analgesics, NSAIDs, tion) compared three advice sessions to exercise (12 antidepressant drugs, muscle relaxants) sessions over 6 weeks) plus advice. Exercise resulted Systematic reviews [2, 12, 18] identified no studies in a significant benefit at 6 weeks, but not at 12 of efficacy of drug therapies in acute whiplash. One months. Exercise was more effective for patients with RCT (40 patients with acute whiplash of less than 8 higher pain scores. However, 56% of advice patients hours) treated with IV methylprednisolone or pla- and 29% of exercise patients sought further treat- cebo led to a significant reduction in pain at 1 week ment between 6 weeks and 12 months, and could and less sick leave for the steroid-treated group com- have biased the results [96]. pared to placebo, but this benefit was not sustained to 6 months [93]. Future research Treatment of chronic whiplash (see Box 8.3) The lack of consistency in study design, patient pop- Few RCTs have considered treatment for chronic ulation, outcome measures and durations of follow- whiplash, but many of these people are included in up and the use of multiple interventions in the same RCTs of chronic nonspecific neck pain. study complicate comparison between studies. Large well-designed randomized prospective studies using Therapies of unknown effectiveness standardized protocols should clarify efficacy and 1. Percutaneous radiofrequency neurotomy cost-effectiveness of individual treatment modalities. • Percutaneous radiofrequency neurotomy versus The benefit of most therapeutic interventions for sham treatment: systematic reviews [2, 73, 81, 94] nonspecific neck pain is small, and many patients identified one small RCT (24 people with chronic improve with limited or even no treatment [97]. It is whiplash) [5], which found that radiofrequency therefore important that future studies set predeter- neurotomy significantly increased the proportion mined minimum differences, which are considered to of people who were free from pain compared with be clinically relevant. This approach should avoid the sham treatment after 27 weeks (58% with active current difficulty that many studies, while showing treatment versus 8% with sham treatment). The statistically significant differences in outcome, may neurotomy also significantly increased the pain- have little clinical importance. free period (median time for more than half of the pain to return – 263 days with neurotomy versus If patient groups are more homogeneous, meta- 8 days with sham treatment; P ϭ 0.04). Although analyses may be possible to provide more robust this high-quality RCT suggested benefits from this guidance on cost-effective approaches to therapy for therapy, the trial was very small and there is no specific patient groups. supporting evidence of efficacy. Discussion 2. Physical and multimodal treatment Nonspecific neck pain including whiplash is a very • Multimodal treatment versus physical treatment: common cause of disability, and places a heavy burden on individuals, employers and society. systematic reviews [2, 81] identified one RCT (33 However, there are many aspects regarding etio- people with chronic whiplash) [95], which com- pathogenesis and treatment which remain poorly pared physical treatments alone with multimodal understood. Furthermore, if the factors that influ- treatment and found no significant difference ence the progression from acute to chronic pain were between treatments in pain, disability or range of better understood, it might be possible to reduce the movement at the end of treatment or at 3 months. frequency and severity of chronic disability for both However, significantly more people treated with 92

Chronic neck pain and whiplash whiplash and other causes of nonspecific neck pain. tested. While exercise, mobilization, and manipu- Most studies of acute neck pain are in patients follow- lation are mainstays of therapy, any regime must ing whiplash injury, and it is not clear if the findings address postural factors identified in individual from those studies can be generalized to neck pain patients. Reduction in the number of pillows at night from nontraumatic causes. is often important, but there is no evidence to sug- gest that “special” pillows justify the additional cost. There have been some higher quality randomized Stress management, yoga, Pilates and the Alexander controlled trials of therapy in patients with nonspe- Technique all improve neck posture and are useful, cific neck pain, which suggest that exercise [32, 33] but require further study. and manual therapy (mobilization physiotherapy or manipulation) [45, 46, 48] are the treatments of Low-dose tricyclic antidepressants seem to be more choice, and are more effective than less active therapies, effective than simple analgesics or anti-inflammatory but the relative cost-effectiveness of these modalities drugs in reducing pain, particularly at night, but have has not been studied. One high-quality study suggests not been subjected to controlled study. additional benefits if exercise is used in combination with mobilization or manipulation [55, 56], and this Comment approach has been advocated by a Cochrane systematic review group [30, 41] but needs further study. I have outlined the evidence for commonly used ther- apeutic modalities used to treat nonspecific neck pain Cervical radiculopathy usually has a favorable out- with or without radiculopathy, and following whip- come, and there is increasing interest in conservative lash injury, highlighted the shortcomings of the stud- approaches. However, there have only been a few poor- ies presented and the current state of our knowledge. quality studies of epidural injection [77, 78] or con- As the studies are so disparate, meta-analyses were not servative approaches [80] and further study is required. possible, and I have presented the individual studies. There is some evidence that early mobilization can Acknowledgment influence outcome following whiplash injury [82–86] Data on therapy included in contribution by the but a recent systematic review that included 23 stud- author to the publication Clinical Evidence. ies (2344 participants) questioned the quality of the data and robustness of the evidence for early active References interventions [18]. Whether there is a window of opportunity for preventing chronic pain in patients 1. Ferrari R, Russell AJ. Epidemiology of whiplash: an interna- with traumatic and nontraumatic neck pain remains tional dilemma. Ann Rheum Dis 1999; 58: 1–5. uncertain. 2. Binder AI. Neck pain syndromes. Clinical Evidence, Issue 18, Many commonly used first-line strategies such as Dec 2007. BMJ Publishing Group, London. analgesic, anti-inflammatory agents, tricyclic antide- pressants, reduction in the number of pillows, stress 3. Gore DR, Sepic SB, Gardner GM. Roentgenographic find- management, and postural advice have not been stud- ings of the cervical spine in asymptomatic people. Spine ied but remain mainstays of treatment. Other modali- 1986; 11: 521–524. ties such as acupuncture, traction, electrotherapy, and psychotherapy are of uncertain value and also need 4. Boden SD, McCowin PR, Davis DO, Dina TS, Mark AS, further study [2]. Wiesel S Abnormal magnetic–resonance scans of the cervi- cal spine in asymptomatic subjects. A prospective investiga- Author recommendation: tion. J Bone Jt Surg 1990; 72A: 1178–1184. a pragmatic approach to the treatment of nonspecific neck 5. Lord SM, Barnsley L, Wallis BJ, McDonald GJ, Bogduk N. pain, including untested measures Percutaneous radio-frequency neurotomy for chronic cervical zygapophyseal-joint pain. N Engl J Med 1996; 335: 1721–1726. It is not currently possible to treat neck pain solely on the basis of proven evidence-based measures, as 6. Sterling M, Treleaven J, Jull G. Responces to a clinical test of many aspects of therapy have not been adequately mechanical provocation of nerve tissue in whiplash associ- ated disorder. Manual Ther 2002; 7: 89–94. 7. Måkelå M, Heliövaara M, Sievers K. Prevalence, determi- nants, and consequences of chronic neck pain in Finland. Am J Epidemiol 1991; 134: 1356–1367. 93

Chapter 8 8. Cote P, Cassidy D, Carroll L. The Saskatchewan health and 25. McDermott FT. Reduction in cervical “whiplash” after new back pain survey: the prevalence of neck pain and related dis- motor vehicle accident legislation in Victoria. Med J Aust ability in Saskatchewan adults. Spine 1998; 23: 1689–1698. 1993; 158: 720–721. 9. Fejer R, Kyvik KO, Hartvigsen J The prevalence of neck pain 26. Virani SN, Ferrari R. Russell AS. Physician resistance to the in the world population: a systematic clinical review of the late whiplash syndrome. J Rheumatol 2001; 28: 2096–2099. literature. Eur Spine J 2006; 15: 834–848. 27. Aker PD, Gross AR, Goldsmith CH, et al. Conservative 10. Hill J, Lewis M, Papageorgiou AC, Dziedzic K, Croft P. management of mechanical neck pain: systematic overview Predicting persistent neck pain: a 1-year follow-up of a and meta-analysis. BMJ 1996; 313: 1291–1296. population cohort. Spine 2004; 29: 1648–1654. 28. Kjellman GV, Skargren EI, Oberg BE. A critical analysis of 11. Bovim G, Schrader H, Sand T. Neck pain in the general randomised clinical trials on neck pain and treatment effi- population. Spine 1994; 19: 1307–1309. cacy. A review of the literature. Scand J Rehabil Med 1999; 31: 139–152. 12. Spitzer WO, Skovron ML, Salmi LR, Cassidy JD, Duranceau J, Suissa S, Zeiss E. Scientific monograph of the Quebec 29. Sarig-Bahat, H. Evidence for exercise therapy in mechanical Task Force on whiplash-associated disorders: redefining neck disorders. Manual Ther 2003; 8: 10–20. “whiplash” and its management. Spine 1995; 20(suppl 8): 1–73. (Erratum in Spine 1995; 20: 2372.) 30. Kay TM, Gross A, Goldsmith CH, Hoving JL, Brønfort G. Exercises for mechanical neck disorders. Cochrane Database 13. Cote P, Cassidy JF, Carroll L, Frank JW, Bombardier C. A of Systematic Reviews 2005, Issue 3. Art. No.: CD004250. systematic review of the prognosis of acute whiplash and DOI: 10.1002/14651858.CD004250.pub3. a new conceptual framework to synthesize the literature. Spine 2001; 26(19): E445–E458. 31. Revel M, Minguet M, Gregory P, et al. Changes in cervico- cephalic kinesthesia after a proprioceptive rehabilitation 14. Obelieniene D, Schrader H, Bovim, G, Miseviciene I, Sand T. program in patients with neck pain: a randomised control- Pain after whiplash: a prospective controlled inception cohort led study. Arch Phys Med Rehabil 1994; 75: 895–899. study. J Neurol Neurosurg Psych 1999; 66: 279–283. 32. Ylinen J, Takala E, Nykanen M, et al. Active neck 15. Partheni M, Constantoyannis C, Ferrari R, Nikiforidis G, muscle training in the treatment of chronic neck pain in Voulgaris S, Papadakis N. A prospective cohort study of women: a randomized controlled trial. JAMA 2003; 289: the outcome of acute whiplash injury in Greece. Clin Exper 2509–2516. Rheum 2000; 18: 67–70. 33. Waling K, Sundelin G, Ahlgren C, et al. Perceived pain 16. Suissa S, Harder S Veilleux. The relation between initial before and after three exercise programs – a controlled clin- symptoms and signs and the prognosis of whiplash. Eur ical trial of women with work–related trapezius myalgia. Spine J 2001; 10: 44–49. Pain 2000; 85: 201–207. 17. Scholten-Peeters GGM, Verhagen AP, Bekkering GE, et al. 34. Waling K, Jaörvholm B, Sundelin G. Effects of training on Prognostic factors of whiplash-associated disorders: a sys- female trapezius myalgia: an intervention study with a 3– tematic review of prospective cohort studies. Pain 2003; year follow-up period. Spine 2002; 27: 789–796. 104: 303–322. 35. Ylinen JJ, Hakkinen AH, Takala EP, et al. Effects of neck 18. Verhagen AP, Scholten-Peeters GGGM, van Wijngaarden S, muscle training in women with chronic neck pain: one year de Bie R, Bierma-Zeinstra SMA. Conservative treatments follow-up study. J Strength Cond Res 2006; 20: 6–13. for whiplash. Cochrane Database of Systematic Reviews 2007, Issue 2. Art. No.: CD003338. DOI: 10.1002/14651858. 36. Chiu TT, Hui-Chan CW, Chein G. A randomized clinical CD003338.pub3. trial of TENS and exercise for patients with chronic neck pain. Clin Rehabil 2005; 19: 850–860. 19. Borghouts JA, Koes BW, Bouter LM. The clinical course and prognostic factors of nonspecific neck pain: a systematic 37. Helewa A, Goldsmith CH, Smythe HA, Lee P, Obright K, review. Pain 1998; 77: 1–13. Stitt L. Effect of therapeutic exercise and sleeping neck sup- port on patients with chronic neck pain: a randomised clin- 20. Hoving JL, de Vet HCW, Twisk JWR, et al. Prognostic factors ical trial. J Rheumatol 2007; 34: 151–158. for neck pain in general practice. Pain 2004; 110: 639–645. 38. Viljanen M, Malmivaara A, Uitti J, et al. Effectiveness of 21. Kjellman G, Skargren E, Oberg B. Prognostic factors for dynamic muscle training, relaxation training, or ordinary perceived pain and function at one year follow-up in pri- activity for chronic neck pain: randomised controlled trial. mary care patients with neck pain. Disabil Rehabil 2002; 24: BMJ 2003; 327: 475–477. 364–370. 39. Koes BW, Assendelft WJ, van der Heijden GJ, et al. Spinal 22. Cote P, Cassidy JD, Carroll L. The treatment of neck and manipulation and mobilization for back and neck pain: a low back pain: who seeks care? Who goes where? Med Care blinded review. BMJ 1991; 303: 1298–1303. 2001; 39: 956–967. 40. Hurwitz EL, Aker PD, Adams AH, et al. Manipulation and 23. Carroll LJ, Cassidy JD, Cote P. Depression as a risk factor mobilization of the cervical spine: a systematic review of for onset of an episode of troublesome neck and low back the literature. Spine 1996; 21: 1746–1760. pain. Pain 2004; 107: 134–139. 41. Gross A, Hoving JL, Haines T, et al., Cervical Overview 24. Cassidy JD, Carroll LJ, Cote P, Lemstra M, Berglund A, Group. Manipulation and mobilization for mechanical Nygren A. Effect of eliminating compensation for pain and neck disorders. Cochrane Database of Systematic Reviews suffering on the outcome of insurance claims for whiplash 2004, Issue 1. Art. No.: CD004249. DOI: 10.1002/14651858. injury. N Engl J Med 2000; 342: 1179–1186. CD004249.pub2. 94

Chronic neck pain and whiplash 42. Bronfort G, Haas M, Evans RL, et al. Efficacy of spinal 55. Bronfort G, Evans R, Nelson B, et al. A randomized clinical manipulation and mobilization for low back pain and trial of exercise and spinal manipulation for patients with neck pain: a systematic review and best evidence synthesis. chronic neck pain. Spine 2001; 26: 788–797. Spine J 2004; 4: 335–356. 56. Evans R, Bronfort G, Nelson B, et al. Two-year follow-up of 43. Canadian Chiropractic Association, Canadian a randomized clinical trial of spinal manipulation and two Federation of Chiropractic Regulatory Boards, Clinical types of exercise for patients with chronic neck pain. Spine Practice Guidelines Development Initiative, Guidelines 2002; 27: 2383–2389. Development Committee (GDC). Chiropractic clinical practice guideline: evidence-based treatment of adult neck 57. Dziedzic K, Hill J, Lewis M, et al. Effectiveness of manual pain not due to whiplash. J Can Chiropr Assoc 2005; 49(3): therapy or pulsed shortwave diathermy in addition to 158–209. advice and exercise for neck disorders: a pragmatic rand- omized controlled trial in physical therapy clinics. Arthritis 44. Vernon H, Humphreys K, Hagino C. Chronic mechani- Care Res 2005; 53: 214–222. cal neck pain in adults treated by manual therapy: a sys- temic review of change scores in randomised clinical trials. 58. Linton SJ, Andersson T. Can chronic disability be pre- J Manip Physiol Therapeut 2007; 30: 215–227. vented? A randomized trial of a cognitive-behaviour inter- vention and two forms of information for patients with 45. Koes BW, Bouter LM, van Mameren H, et al. Randomised spinal pain. Spine 2000; 25: 2825–2831. clinical trial of manipulative therapy and physiotherapy for persistent back and neck complaints: results of one year fol- 59. Horneij E, Hemborg B, Jensen I, et al. No significant low up. BMJ 1992; 304: 601–605. differences between intervention programmes on neck, shoulder and low back pain: a prospective randomized 46. Jordan A, Bendix T, Nielsen H, et al. Intensive training, study among home-care personnel. J Rehabil Med 2001; 33: physiotherapy, or manipulation for patients with chronic 170–176. neck pain. A prospective, single-blinded, randomized clini- cal trial. Spine 1998; 23: 311–319. 60. Karjalainen KA, Malmivaara A, van Tulder MW, et al. Multidisciplinary biopsychosocial rehabilitation for neck 47. Cassidy JD, Lopes AA, Yong-Hing K. The immediate effect and shoulder pain among working age adults. Cochrane of manipulation versus mobilization on pain and range of Database of Systematic Reviews 2003, Issue 2. Art. No.: motion in the cervical spine: a randomised controlled trial. CD002194. DOI: 10.1002/14651858. CD002194. J Manip Physiol Therapeut 1992; 15: 570–575. 61. Linton SJ, Boersma K, Jansson M, et al. The effects of 48. Hurwitz EL, Morgenstern H, Harber P, et al. A rand- cognitive-behavioral and physical therapy preventive inter- omized trial of chiropractic manipulation and mobilization ventions on pain-related sick leave: a randomized control- for patients with neck pain: clinical outcomes from the led trial. Clin J Pain 2005; 21: 109–119. UCLA neck-pain study. Am J Public Health 2002; 92: 1634–1641. 62. Jensen I, Nygren A, Gamberale F, et al. The role of the psy- chologist in multidisciplinary treatments for chronic neck 49. Hurwitz EL, Morgenstern H, Vassilaki M, et al. Adverse and shoulder pain: a controlled cost-effectiveness study. reactions to chiropractic treatment and their effects on Scand J Rehabil Med 1995; 27: 19–26. satisfaction and clinical outcomes among patients enrolled in the UCLA Neck Pain Study. J Manip Physiol Therapeut 63. van der Heijden GJ, Beurskens AJ, Koes BW, et al. The 2004; 27: 16–25. efficacy of traction for back and neck pain: a systematic, blinded review of randomized clinical trial methods. Phys 50. Martinez-Segura R, Fernandez-de-las-Penas C, Ruiz-Saez M, Ther 1995; 75: 93–104. Lopez-Jimenez C, Rodriguez-Blanco C. Immediate effects on neck pain and active range of motion after a single cervical 64. Graham N, Gross AR, Goldsmith C, and the Cervical high velocity low amplitude manipulation in subjects pre- Overview Group. Mechanical traction for mechanical neck senting with mechanical neck pain: a randomised controlled disorders: a systemic review. J Rehabil Med 2006; 38: 145–152. trial. J Manip Physiol Therapeut 2006; 29: 511–517. 65. White AR, Ernst E. A systematic review of randomized con- 51. Hoving J, Koes B, de Vet H, et al. Manual therapy, physi- trolled trials of acupuncture for neck pain. Rheumatology cal therapy, or continued care by a general practitioner for 1999; 38: 143–147. patients with neck pain. A randomized, controlled trial. Ann Intern Med 2002; 136: 713–722. 66. Smith LA, Oldman AD, McQuay HJ, et al. Teasing apart quality and validity in systematic reviews: an example from 52. Korthals-de Bos I, Hoving J, van Tulder, M, et al. Cost effec- acupuncture trials in chronic neck and back pain. Pain tiveness of physiotherapy, manual therapy, and general 2000; 86: 119–132. practitioner care for neck pain: economic evaluation along- side a randomised controlled trial. BMJ 2003; 326: 911–914. 67. Trinh K, Graham N, Gross A, et al., Cervical Overview Group. Acupuncture for neck disorders. Cochrane Database 53. Kjellman G, Oberg B. A randomised clinical trial compar- of Systematic Reviews 2006, Issue 3. Art. No.: CD004870. ing general exercise, McKenzie treatment and a control DOI: 10.1002/14651858.CD004870.pub3. group in patients with neck pain. J Rehabil Med 2002; 34: 183–190. 68. Kroeling P, Gross A, Goldsmith CH, Houghton PE, Cervical Overview Group. Electrotherapy for neck disorders. 54. Dabbs V, Lauretti WJ. A risk assessment of cervical manipu- Cochrane Database of Systematic Reviews 2005, Issue 2. Art. lation vs NSAIDS for the treatment of neck pain. J Manip No.: CD004251. DOI: 10.1002/14651858.CD004251.pub3. Physiol Therapeut 1995; 18: 530–536. 69. Trock DH, Bollet AJ, Markoll R. The effect of pulsed elec- tromagnetic fields in the treatment of osteoarthritis of 95

Chapter 8 the knee and cervical spine. Report of randomized dou- with collar therapy for whiplash injury. Emerg Med J 2004; ble-blind placebo controlled trials. J Rheumatol 1994; 21: 21: 306–310. 1903–1911. 85. McKinney LA. Early mobilization and outcome in acute 70. Boswell MV, Colson JD, Spillane WF. Therapeutic facet joint sprains of the neck. BMJ 1989; 299: 1006–1008. intervention in chronic spinal pain: a systematic review of 86. Rosenfeld M, Gunnarsson R, Borenstein P. Early interven- effectiveness and complications. Pain Physician 2005; 8: tion in whiplash-associated disorders: a comparison of two 101–114. treatment protocols. Spine 2000; 25: 1782–1787. 71. Boswell MV, Colson JD, Sehgal N, et al. A systematic review 87. Rosenfeld M, Seferiadis A, Carlsson J, et al. Active inter- of therapeutic facet joint interventions in chronic spinal vention in patients with whiplash-associated disorders pain. Pain Physician 2007; 10: 229–253. improves long-term prognosis: a randomized controlled 72. Manchikanti L, Damron K, Cash K, Manchukonda R, clinical trial. Spine 2003; 28: 2491–2498. Pampati V. Therapeutic cervical medial branch blocks in 88. Scholten-Peeters GGM, Neeleman-van der Steen CW, van managing chronic neck pain: a preliminary report of ran- der Windt DA, et al. Education by general practitioners or domised, double-blind, controlled trial NCT0033272, Pain education and exercises by physiotherapists for patients Physician 2006; 9: 333–346. with whiplash-associated disorders? A randomized clinical 73. Boswell MV, Hansen HC, Trescot AM, et al. Epidural ster- trial. Spine 2006; 31: 723–731. oids in the management of chronic spinal pain and radicu- 89. Borchgrevink GE, Kaasa A, McDonagh D, et al. Acute treat- lopathy. Pain Physician 2003; 6: 319–334. ment of whiplash neck sprain injuries: a randomised trial 74. Abdi S, Datta S, Lucas LF. Role of epidural steroids in the of treatment during the first 14 days after a car accident. management of chronic spinal pain: a systematic review Spine 1998; 23: 25–31. of effectiveness and complications. Pain Physician 2005; 8: 90. Söderlund A, Olerud C, Lindberg P. Acute whiplash-asso- 127–143. ciated disorders (WAD): the effects of early mobilization 75. Abdi S, Datta S, Trescot AM, et al. Epidural steroids in the and prognostic factors in long-term symptomatology. Clin management of chronic spinal pain: a systematic review. Rehab 2000; 14: 457–467. Pain Physician 2007; 10: 185–212. 91. Foley-Nolan D, Moore K, Codd M, et al. Low energy high 76. Anderberg L, Annertz M, Persson L, Brandt L, Saveland H. frequency pulsed electromagnetic therapy for acute whip- Transforaminal steroid injections for the treatment of cer- lash injuries. A double blind randomised controlled study. vical radiculopathy: a prospective and randomised study. Scand J Rehabil Med 1992; 24: 51–59. Eur Spine J 2007; 16: 321–328. 92. Provinciali L, Baroni M, Illuminati L, et al. Multimodal 77. Stav A, Ovadia L, Sternberg A, et al. Cervical epidural ster- treatment to prevent the late whiplash syndrome. Scand J oid injection for cervicobrachialgia. Acta Anaesthesiol Scand Rehabil Med 1996; 28: 105–111. 1993; 37: 562–566. 93. Pettersson K, Toolanen G. High-dose methylprednisolone 78. Castagnera L, Maurette P, Pointillart V, et al. Long-term prevents extensive sick leave after whiplash injury. A pro- results of cervical epidural steroid injection with and with- spective randomised double-blind study. Spine 1998; 23: out morphine in chronic cervical radicular pain. Pain 1994; 984–989. 58: 239–243. 94. Niemisto L, Kalso EA, Malmivaara A, Seitsalo S, Hurri 79. Fouyas IP, Statham PF, Sandercock PA. Cochrane review on H. Radiofrequency denervation for neck and back pain. the role of surgery in cervical spondylotic radiculomyelopa- Cochrane Database of Systematic Reviews 2003, Issue 1. thy. Spine 2002; 27: 736–747. Art. No.: CD004058. DOI: 10.1002/14651858.CD004058. 80. Persson LC, Carlsson CA, Carlsson JY. Long-lasting cervical 95. Söderlund A, Lindberg P. Cognitive behavioural compo- radicular pain managed with surgery, physiotherapy, or a nents in physiotherapy management of chronic whiplash cervical collar: a prospective randomised study. Spine 1997; associated disorders (WAD) – a randomised group study. 22: 751–758. Physiother Theory Pract 2001; 17: 229–238. 81. Seferiadis A, Rosenfeld M, Gunnarsson R. A review of treat- 96. Stewart MJ, Maher CG, Refshauge KM, Herbert RD, ment interventions in whiplash-associated disorders. Eur Bogduk N, Nicholas M. Randomized controlled trial of Spine J 2004; 13: 387–397. exercise for chronic whiplash-associated disorders. Pain 82. Mealy K, Brennan H, Fenelon GC. Early mobilization of 2007; 128: 59–68. acute whiplash injuries. BMJ 1986; 292: 656–657. 97. Vernon H, Humphreys BK, Hagino C. The outcome of con- 83. Bonk AD, Ferrari R, Giebel GD, Edelmann M, Huser R. trol groups in clinical trials of conservative treatments for Prospective, randomized, controlled study of activity ver- chronic mechanical neck pain: a systematic review. BMC sus collar, and the natural history for whiplash injury, in Musculoskel Disord 2006; 7: 58. Germany. J Musculoskel Pain 2000; 8: 123–132. 84. Schnabel M, Ferrari R, Vassiliou T, et al. Randomised, con- trolled outcome study of active mobilization compared 96

CHAPTER 9 Pain associated with osteo-arthritis David L. Scott Department of Rheumatology and Weston Education Centre, Kings College London School of Medicine, London, UK Background faster than they are synthesized. The decreased prote- oglycan content and altered collagen matrix result in Pathophysiology deterioration in the physiologic features of cartilage. Osteo-arthritis involves joint cartilage, subchondral bone and synovium. It is characterized by loss of The early stages of cartilage damage are character- articular cartilage, new bone formation both subchon- ized by microfractures and fibrillations. With time drally and at the joint margins, and variable amounts there is gross damage to articular cartilage. The ini- of synovial inflammation. Its clinical features com- tially smooth surface of the cartilage becomes rough prise pain, stiffness, bony joint swelling and reduced and eroded with cracks. It often shows ulceration. joint movement. With time, these impair quality of Many of these early cartilage changes are mediated by life. The eventual outcome of osteo-arthritis is joint proteolytic enzymes, in particular metalloproteases. failure [1, 2]. Key roles are played by collagenase, which is respon- sible for collagen degradation, and stromelysin, which Osteo-arthritis is mechanically driven and chemi- is responsible for proteoglycan degradation. cally mediated. It involves attempted but aberrant repair of the joint cartilage. In the early stages there Changes in the cartilage matrix are accompanied is an imbalance between the destructive and repara- by synovial inflammation, with the involvement of a tive processes in articular cartilage. Its development is range of inflammatory mediators. Proinflammatory linked with a range of genetic, environmental, meta- cytokines, such as tumor necrosis factor α, are impor- bolic and biochemical factors. So far, the search for tant, though there is a balance between cytokine- a single identifiable cause of osteo-arthritis has not driven anabolic and catabolic processes. A range been fruitful, and many believe it is a family of disor- of cytokines have been implicated, including inter- ders and not a single disease entity. leukin-1 and interleukin-6. In early osteo-arthritis a number of “triggering Many other small molecular mediators are involved factors” activate the division and multiplication of in the development of osteo-arthritis. Examples chondrocytes. The most important trigger is exces- include nitric oxide, which is involved in many sive force. As a consequence, the chondrocytes mul- inflammatory conditions in which there are vascular tiply and become metabolically active. Initially the changes, and a range of prostaglandins. These small chondrocytes produce increased amounts of prote- mediators not only regulate cytokines but also result oglycans and collagen. However, these overproduced in pain and inflammation and changes in blood proteoglycans are immature. Over time the collagen vessels. fibers are altered and the proteoglycans break down Classification Evidence-Based Chronic Pain Management. Edited by Osteo-arthritis can be classified based on its symp- C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell toms, findings on examination and radiographic Publishing. assessment. In most patients there is not a discrete onset, there are no specific laboratory abnormalities 97

Chapter 9 and no pathognomonic features. Osteo-arthritis cov- studies in the USA and Europe show about 1% of ers a broad spectrum of clinical features. Consequently the population have knee osteo-arthritis under 35 both its classification and epidemiology are highly years of age and this rises to over 30% above the age varied. of 75 years. Osteo-arthritis can be classified by the joints The natural history of osteo-arthritis of the knee involved (such as knee or hip), and by whether it is is not well characterized. A substantial number of primary or secondary related to metabolic, anatomic, patients show radiological progression and up to 25% traumatic or inflammatory conditions. Primary gen- of osteo-arthritic knees with initially normal joint eralized osteo-arthritis is a recognised subtype in space show major damage after 10 years. However, which there is involvement of the distal and proximal there is no close relationship between X-ray progres- interphalangeal joints of the hand, the first carpomet- sion and clinical features. People with osteo-arthritis acarpal joint, knees, hips and the metatarsophalangeal requiring specialist referral usually have bad out- joints. Most patients have a less well-defined pattern. comes with high levels of physical disability. The extent to which this reflects the disease itself or asso- The American College of Rheumatology has devel- ciated co-morbidities is debatable. oped classification criteria for knee osteo-arthritis [3]. These criteria, which depend on expert physical Risk factors examination and X-rays, are used in reporting clinical Age is the dominant risk factor for osteo-arthritis, trials but are rarely applied in epidemiological studies and in many ways the disorder can be viewed as an or used in everyday clinical practice. inevitability of aging. Sex is also important; about twice as many women get osteo-arthritis, mainly after Epidemiology the menopause. Whether this increased incidence In epidemiological studies, osteo-arthritis is most is due to specific hormonal factors remains uncer- commonly defined by radiological criteria. However, tain. Overall, there is limited evidence to suggest many patients with radiographically defined changes that female sex hormones have a definite effect on in the knee have no symptoms. Radiographic crite- osteo-arthritis. ria proposed by Kellgren and Lawrence over 40 years ago remain the principal method for defining osteo- Osteo-arthritis runs in families, particularly gener- arthritis. An alternative approach is to assess the fre- alized osteo-arthritis. There is less genetic involvement quency of knee pain in later life, usually considered in knee than hand osteo-arthritis. Candidate genes for as over 55 or 60 years of age. Osteo-arthritis is found osteo-arthritis include the vitamin D receptor gene, at some sites in most people older than 65 years, and insulin-like growth factor 1 genes, cartilage oligomeric over 80% of those over the age of 75 years. protein genes, and the HLA region. It seems likely that genes affecting osteo-arthritis will influence its occur- Autopsy studies show pathological features of rence in many joints. Along with these genetic risks osteo-arthritis are very common. Almost everyone are racial differences in the development of the dis- over the age of 65 years will have autopsy evidence ease. For example, in North America the frequency of of cartilage damage. Cartilage erosions, subchondral osteo-arthritis is higher in black women. bone changes and osteophytes in the knees are seen in over 60% of people who die in their seventh and Weight is a very important risk factor for knee eighth decades. osteo-arthritis. It is thought that overloading knee joints leads to cartilage breakdown and failure of A North American study – the National Health ligamentous and other structural support. The effect and Nutrition Examination Survey – reported that of weight on osteo-arthritis is particularly impor- knee osteo-arthritis is rare (0.1%) in people aged tant because obesity is a serious and growing public 25–34. It rises to over 30% in those aged 75 years or health problem in the Western world. In persons who older [4]. Another large North American study – the are overweight, weight loss can reduce the risk for Framingham Study – found the prevalence of knee osteo-arthritis. osteo-arthritis was 30% among those aged 65–74 years. The prevalence of radiographic osteo-arthritis Some occupations increase the risk of osteo-arthri- rises steeply with age [5]. Overall population-based tis. For example, heavy physical labor, especially lifting, 98

Pain associated with osteo-arthritis may increase the risk of developing osteo-arthritis of The medical management of pain in osteo- the knee, as can kneeling and squatting. Some sports, arthritis is by definition nonsurgical. However, it is especially high-impact sports, also increase the risk important to appreciate that surgical intervention of osteo-arthritis. Participation in some competitive in osteo-arthritis can be particularly important in sports is associated with the subsequent development overcoming pain as well as having positive effects of knee osteo-arthritis. However, moderate regular in reducing disability and specifically improving running has low, if any, risk. Sports that increase risks mobility. Joint replacement surgery is particularly are those that demand high-intensity, acute, direct effective, but it lies outside the themes explored in joint impact as a result of contact with other par- this review. ticipants, playing surfaces or equipment, for example injuries to the knees of soccer players. Most trials use simple assessments of pain and dis- ability. Pain is recorded using visual analog scales or Finally, there is incomplete evidence that continuous five-point Likert scales. Pain can be recorded globally exposure to oxidants contributes to the development or in specific situations such as at rest or during exer- of osteo-arthritis as well as other common age-related cise. There are also specific scales that record pain, diseases. As micronutrient antioxidants provide defense stiffness and function in osteo-arthritis; the most against tissue injury, high dietary intake of such micro- widely used is the Western Ontario and McMaster nutrients might protect against osteo-arthritis, though Osteo-Arthritis (WOMAC) Scale [6]. the evidence for this is incomplete. As virtually all trials assess global pain, changes in Interventions this measure are the focus of the treatment reviews that follow. The benefit of this assessment has been Assessing benefit in osteo-arthritis counterbalanced by an assessment of adverse events. The primary goals of treatment are reducing pain, In all cases the most recent systematic review has been stiffness and disability. Associated aims are to improve the basis for judging efficacy. In some situations effi- the quality of life and ensure adverse effects are mini- cacy and adverse events have been assessed separately, mal. All treatments achieve these goals to a greater or for example with oral nonsteroidal anti-inflammatory lesser extent. A long-term goal is to reduce progressive drugs (NSAIDs). This is because no single review joint damage, though no treatment yet achieves this. examines both. The overall benefits of effective treat- ments are summarized in Table 9.1. Table 9.1 Strength of evidence for treatments in knee osteo-arthritis: largest systematic reviews compared for effective treatments Treatment Placebo-controlled Patients Effectiveness Adverse effects Overall value trials Paracetamol 7 1966 SMD –0.13 Minimal Low efficacy compensated by few Oral NSAID/COXIBs 23 10,845 Pooled effect size Multiple serious adverse effects 976 0.32 adverse effects Efficacy limited by high Local NSAIDs 7 Pooled effect size Minimal rates of adverse effects 4856 0.41 Efficacy only in short term Opioids 18 Pooled effect size Multiple 1973 0.79 unpleasant Efficacy limited by high Local steroid 28 2542 adverse effects rates of adverse effects injections 40 Mean weighted Uncommon Hyaluronic acid difference 21.9 Efficacy only in short term injections Mean weighted Uncommon difference 9.0 Sustained efficacy but limited by need for repeated injections 99

Chapter 9 Interventions supported by evidence Opioid analgesics Opioid analgesics, particularly mild opioids, have Simple analgesia with paracetamol been widely suggested to be both effective and rela- Paracetamol is the most commonly used analgesic. tively safe for treating moderate to severe osteo- A single 1000 mg dose of paracetamol provides Ͼ50% arthritis that does not respond to first-line treatment. pain relief over 4–6 hours in moderate or severe pain The effects are mainly mediated centrally by changing compared with placebo. There are virtually no con- pain perceptions in the brain. There is some evidence traindications, significant drug–drug interactions or for peripheral effects of opioids in arthritis. The dis- side effects at the recommended dosage. It is well advantage of opioids is their significant adverse effects tolerated by patients with peptic ulcers. Despite many including both gastrointestinal problems like nausea years of use, the mechanism of action of paraceta- and central problems like drowsiness. mol is not well understood. It may be centrally active, producing analgesia by elevating the pain threshold A systematic review (18 RCTs, 4856 people with through prostaglandin synthetase inhibition in the osteo-arthritis) [8] compared opioid analgesics with hypothalamus. At therapeutic dosages it does not placebo. The opioids included oxycodone, fentanyl and inhibit prostaglandin synthetase in peripheral tissues, morphine, tramadol, tramadol with paracetamol, and so has no anti-inflammatory activity. codeine. The pooled effect sizes of all opioids com- pared to placebo on pain intensity was −0.79 (95% CI A systematic review (seven randomized control- −0.98 to −0.59). The heterogeneity of these trials was led trials (RCTs), 1966 people with osteo-arthritis) substantial; however, there was no evidence to suggest [7] compared paracetamol with placebo. Five of that the conclusions differed for type of opioid, the the RCTs reviewed reported that paracetamol was way in which pain was measured and the methodo- superior to placebo. Two showed no benefit of para- logic quality of the study. The most frequent adverse cetamol over placebo. Overall, paracetamol was sig- events with opioids were nausea, constipation, dizzi- nificantly better than placebo in several assessments ness, somnolence and vomiting. The average treatment of pain including pain response, pain on motion, and discontinuation rate for toxicity was 25% with opioids overall pain, and also physician and patient global and 7% with placebos. The number needed to harm assessment. Dichotomous outcomes were recorded in (NNH) for all opioids compared to placebo was 5; for a single study and the number needed to treat (NNT) strong and weak opioids it was 4 and 9, respectively. for a pain response was 4 (95% confidence interval (CI) 2–24) and for an improvement in patient glo- Another systematic review specifically evaluated bal assessment was 2 (95% CI 2–13). Overall pain, tramadol and tramadol and paracetamol combina- recorded as a continuous variable in five studies, tions (11 RCTs, 1939 people with osteo-arthritis) showed a standard mean difference (SMD) of Ϫ0.13 [9]. Three trials included placebo controls and they (95% CI Ϫ0.22 to Ϫ0.04) with a NNT of 16. showed a weighted mean difference in favor of trama- dol of Ϫ8.5 (95% CI Ϫ12.1 to Ϫ4.9). Two placebo- In terms of toxicity, the total number of patients controlled studies reported data on pain relief from reporting any adverse event, the relative risk of para- 216 participants who received active treatment and cetamol to placebo was 1.02 (95% CI 0.89–1.17). For 218 participants who received placebo, and showed the toxicity outcome of total number of withdrawals due that tramadol increased by 53% the likelihood of a to toxicity, the relative risk was 1.24 (95% CI 0.87–1.77). moderate improvement compared to placebo. This Consequently, there were no significant differences in improvement was equivalent to a number needed to toxicity between paracetamol and placebo in these trials. benefit of 8 (95% CI 5–25). Common adverse events with tramadol included vomiting, dizziness, constipa- The overall assessment of paracetamol is that it is tion, somnolence, tiredness and headache. In placebo- beneficial, with a favorable ratio of benefits to risks. controlled trials minor adverse events occurred 2.2 However, its effect is minor and therefore only a small times more often with tramadol, and the NNH was number of patients will wish to take it. It has one spe- 5 (95% CI 4–7). For adverse events that were severe cific disadvantage: it needs to be taken in substantial enough for treatment to be withdrawn, the NNH was amounts on a regular basis and most patients are 8 (95% CI 7–12). reluctant to take 1 g paracetamol four times daily. 100

Pain associated with osteo-arthritis Oral NSAIDs such as myocardial infarction and renal problems. Nonsteroidal anti-inflammatory drugs are a diverse Older conventional NSAIDs had a substantial risk of group. Their name distinguishes them from anti- upper gastrointestinal adverse events, many of which inflammatory steroids (glucocorticoids) and non- were serious, and the newer COXIBs such as celecoxib narcotic analgesics. NSAIDs are one of the most reduced these significantly. However, COXIBs have frequently used groups of drugs overall, although resulted in other problems, particularly the onset of their benefits must be set against significant, some- myocardial infarctions when used for prolonged peri- times fatal, gastrointestinal and renal toxicity and also ods of time at high dosages. The older clinical tri- the recently described cardiovascular risks with cyclo- als were short term and did not investigate cardiac oxygenase-2 (COX-2) inhibitors. risks in detail. However, recent trials of COXIBs have almost entirely used conventional NSAIDs as com- COX was originally purified in the 1970s. By 1990 parators. Consequently it is difficult to estimate the it was realized that the enzyme had two isoforms. overall risks of conventional NSAIDs and COXIBs. The COX-1 isoform is responsible for the produc- tion of “housekeeping” prostaglandins critical for In a long-term randomized trial lasting up to normal renal function, gastric mucosal integrity 4 years, 61–63% of patients receiving oral NSAIDs and vascular hemostasis. By contrast, COX-2 is an and 50% of those receiving placebo reported adverse inducible enzyme. It is upregulated in macrophages, events [11]. One systematic review evaluated the abil- monocytes and other inflammatory cells by various ity of COXIBs and other gastroprotective strategies stimuli including IL-1 and other cytokines. NSAIDs with conventional NSAIDs to reduce upper gastroin- can be classified according to their relative effect on testinal risks [12]. There was strong evidence that COX-1 and COX-2. Generally, the risk of gastrointes- COXIBs, in comparison to conventional NSAIDs, tinal adverse effects is reduced with increasing COX-2 reduce events by about 50% in meta-analyses of ran- selectivity. However, other factors are involved in the domized trials (52, 474 patients) and large obser- causation of gastrointestinal toxicity because, para- vational studies in clinical practice (3093 bleeding doxically, certain NSAIDs that are relatively COX-2 events). Evidence on the efficacy of NSAIDs plus selective have been associated with a higher incidence gastroprotection with acid suppressants (proton of gastrointestinal adverse events. NSAIDs have many pump inhibitors and histamine antagonists), based actions other than their effect on COX. These include on the surrogate measure of endoscopic ulcers, sug- uncoupling oxidative phosphorylation; inhibiting gested that NSAIDs with added protection was more lysosomal enzyme release; inhibiting complement damaging than COXIBs. Another systematic review activation; antagonizing the generation of activity in evaluated thrombotic cardiovascular adverse events kinins; inhibiting free radicals. [13] and reported that in placebo studies, COXIBs gave a 42% relative increase in the incidence of seri- The most recent large systematic review of efficacy ous vascular events, which was chiefly attributable to of oral NSAIDs compared them against placebo (23 an increased risk of myocardial infarction. Overall, trials, 10,845 people with osteo-arthritis) [10]. This the incidence of serious vascular events was similar review found that NSAIDs significantly reduced pain between a selective COX-2 inhibitor and any tradi- compared with placebo. The pooled effect size for a tional NSAIDs (1.0%/year versus 0.9%/year). reduction in pain was 0.32 (95% CI 0.24–0.39). Those trials that used short durations of treatment of less Topical NSAIDs than 6 weeks showed no difference in effect sizes for Two systematic reviews have compared topical NSAIDs reduced pain. with placebo. The first systematic review (seven RCTs, 976 people with osteo-arthritis) [14] found that topi- The disadvantage of conventional oral NSAIDs cal NSAIDs significantly reduced pain compared with (such as ibuprofen, naproxen and diclofenac) and placebo in weeks 1 and 2 (effect size 0.41, 95% CI newer COX-2 specific drugs, which have been termed 0.18–0.63) but not in weeks 3 and 4 (effect size ϩ0.08, COXIBs (such as celecoxib, lumiracoxib and etori- 95% CI Ϫ0.04 to ϩ0.2. The second systematic review coxib) is their propensity to cause adverse effects. (three RCTs, 790 people with osteo-arthritis) [15] These include upper gastrointestinal ulcers and bleeding, small bowel problems, cardiac problems 101

Chapter 9 evaluated trials involving topical diclofenac. Compared significantly improved pain compared to placebo for with placebo, topical diclofenac significantly reduced up to 6 months post injection. The weighted mean dif- pain scores with a standardized mean difference –0.33 ference in the first 3 months after injection was –7.7 (95% CI –0.48 to –0.18). The main adverse effect (95% CI –11.3 to –4.1). In the second 3 months after of topical treatment is local skin irritation; systemic injection the weighted mean difference in pain was –9.0 adverse effects were no more common than with (95% CI –14.8 to –3.2). The second systematic review placebo. (11 RCTs, 1443 people with osteo-arthritis) compared hyaluronan injection with placebo [18]. It showed that Corticosteroid injections hyaluronic acid injections were moderately effective Local steroids have been widely used for almost in relieving knee pain in patients with osteo-arthritis 50 years in arthritis. A number of formulations are up to 10 weeks post injection, but not thereafter. Local available including hydrocortisone and methylpred- skin reactions are reported with hyaluronic acid prepa- nisolone; most clinicians favor longer-acting steroids rations but there is no evidence of a significant increase such as methylprednisolone. compared to placebo treatment. One systematic review (28 RCTs, 1973 people with Interventions refuted by evidence osteo-arthritis) compared intra-articular corticoster- oids with placebo and other treatments [16]. It found Antioxidant vitamin and selenium that intra-articular corticosteroids significantly supplements reduced pain at 1 week compared with placebo; the One systematic review evaluated antioxidants (nine weighted mean difference was –21.9 (95% CI –29.9 trials, 567 patients) [19]. Seven trials examined vita- to –13.9). There was some evidence of benefit at 2 min E and other trials examined selenium ACE, and 3 weeks but no evidence for a sustained effect vitamin A and vitamin C. The authors found no con- beyond 4 weeks. Adverse effects are uncommon; vincing evidence that selenium, vitamin A, vitamin C symptom flare, tissue atrophy, fat necrosis, calcifica- or the combination product selenium ACE was effec- tion and vascular necrosis have all been reported and tive in the treatment of osteo-arthritis. there is a theoretical risk of infection but virtually no evidence that it is a relevant issue. Other alternative treatments A detailed recent systematic review of alternative Hyaluronic acid injections treatments in osteo-arthritis [20] found no evidence Hyaluronic acids are large glucosaminoglycans in to support the use of homeopathy, magnet therapy, synovial fluid. They have high but variable molecu- tai chi, leech therapy, music therapy, yoga, imagery lar weight and viscosity. The reason why they may and therapeutic touch. For these various treatments improve knee osteo-arthritis is uncertain and a the evidence was weak and contradictory. number of mechanisms have been suggested, includ- ing providing lubrication and shock absorption. The Commonly used interventions currently usual source is avian and there may be some protein unproven contamination that can cause allergic reactions. Some products are produced from bacterial sources and so Glucosamine avoid this potential problem. The degree of cross- Glucosamine is a sugar, a sulfated amino-monosac- linking of the hyaluronic acid is variable. Usually charide. It is one constituent of the disaccharide units 3–5 weekly injections comprise a course of treatment. in cartilage proteoglycans. Experimentally it can alter Recent research is aiming to reduce this to a single chondrocyte metabolism, and this was part of the injection. rationale underlying its clinical use, though whether or not oral glucosamine reaches chondrocytes in the Two systematic reviews have compared various joint is uncertain. Its classification as a drug, food formulations of hyaluronic acid to placebo. The first supplement, nutriceutical or complementary therapy systematic review (40 trials, 2542 people with osteo- is debatable. arthritis) compared intra-articular hyaluronan and hyaluronan derivatives with placebo [17]. Hyaluronan 102

Pain associated with osteo-arthritis One systematic review of glucosamine (15 trials, 2613 concluded there was some evidence to support the use people with osteo-arthritis) found that glucosamine of avocado/soybean unsaponifiables, topical capsai- hydrochloride was not effective, but there was some evi- cin, and devil’s claw. For example, an analysis of three dence supporting the use of glucosamine sulfate [21]. trials involving capsaicin showed that the odds ratio The pooled effect size of glucosamine hydrochloride on favoring capsaicin affecting pain was 4.36. However, pain was 0.06 (95% CI –0.08 to +0.20) and the pooled the numbers of trials and their size were insufficient effect size of glucosamine sulfate on pain was 0.44 (95% to determine a definite benefit from any of these spe- CI 0.18–0.70). However, there was considerable hetero- cific herbal treatments. geneity amongst trials and therefore the overall impact of glucosamine was uncertain. Comparison of treatments The short-term effects of different treatments for Adverse reactions to glucosamine sulfate were eval- osteo-arthritis have been compared in a single sys- uated in an earlier systematic review of 12 trials; only tematic review of placebo-controlled trials (63 trials, seven of 1486 patients receiving glucosamine sulfate 14,060 people with osteo-arthritis) [27]. Opioids and were withdrawn for glucosamine-related toxicity and oral NSAID therapy in patients with moderate to only 48 reported any glucosamine-related adverse severe pain had maximum efficacies compared to pla- reactions [22]. cebo at 2–4 weeks with improvements in 100 mm vis- ual analog pain scores of 10.5 mm (95% CI 7.4–13.7) Chondroitin and 10.2 mm (95% CI 8.8–11.2) respectively. There Chondroitin is a glycosaminoglycan that is found in was some evidence that the efficacy of opioids was the proteoglycans of articular cartilage. It is an ingre- inflated by high withdrawal rates and “‘best-case”’ dient found commonly in dietary supplements used scenarios reported in intention-to treat analyses. By to treat osteo-arthritis. It is commonly combined comparison, intra-articular steroid injections and with glucosamine. Most chondroitin is made from topical NSAIDs had maximum efficacies at 1–3 weeks extracts of cartilaginous cow and pig tissues; other of 14.5 mm (95% CI 9.7–19.2) and 11.6 mm (95% sources such as shark cartilage are also used. Since CI 7.4–15.7), respectively. Paracetamol, glucosamine chondroitin is not a uniform substance, and is natu- sulfate and chondroitin sulfate had maximum mean rally present in a wide variety of forms, the composi- efficacies of 4.7 mm or less. These effects are summa- tion of each supplement varies. Only a minority of rized in Figure 9.1. The benefits of treatments other chondroitin taken by mouth is absorbed. than NSAIDs, opioids and local steroids appear highly questionable in the short term. One systematic review of chondroitin (20 trials, 3846 people with osteo-arthritis) found the symp- Future research to improve tomatic benefit of chondroitin is minimal or nonex- management of pain in istent [23]. The initial meta-analysis showed a large osteo-arthritis effect size on pain of –0.75 (95% CI –0.99 to –0.50). However, there was not only a substantial degree Ideally we need new drugs that are more effective in of heterogeneity between trials but the results also controlling pain in osteo-arthritis. The problem with showed that the benefit of chondroitin was high this goal is that very few such treatments have been in early trials and fell to nonexistent in larger, more introduced over the last 25 years, with the excep- recent trials. Overall, the authors dismissed its impact tion of new NSAIDs, including the COXIBs. There as being not clinically relevant. The overall risk of have been some improvements in how conventional adverse events was small; the pooled relative risk was drugs are delivered, particularly focusing on transder- 0.99 (95% CI 0.76–1.31). mal routes of administration which reduce adverse events. However, given the size of the problem, these Herbal treatments advances have been disappointingly small. Hundreds of herbal remedies are used for osteo- arthritis, but very few have any evidence supporting One innovation that in the medium term may their use. A review of systematic reviews in this area improve pain control is the imaging of pain using [24], using data from two systematic reviews [25, 26], 103

Chapter 9 20 Difference in pain VAS over placebo Slight improvement 10 Minimal perceptible improvement 0 Topical Opioids Paracetamol Glucosamine Chondroitin Figure 9.1 Changes in pain over 2–4 Oral NSAID weeks in placebo-controlled trials in osteo-arthritis. VAS, visual analog NSAID scale. Reproduced from & coxibs Bjordal et al. [27] with permission from Elsevier. functional magnetic resonance imaging (fMRI). The I prefer to use simple analgesics and topical NSAIDs approach has been applied to chronic pain because, in mild disease, NSAIDs in patients in whom there unlike in acute pain, the central nervous system (CNS) is an inflammatory component, and injections of is involved through pain centralization. The resulting hyaluronic acid in patients in whom there has been neural activity causes changes in capillary blood flow little improvement with these simpler medical treat- which can be visualized by fMRI. Positron emission ments. In terms of balancing benefits with risks of tomography and fMRI studies using experimental pain adverse events, my own perspective is that the advan- models in arthritis and other conditions have dem- tages of therapy invariably outweigh its risks. onstrated the existence of the CNS pain matrix – the changes in CNS activity which accompany chronic Many patients like to take self-prescribed glu- pain. fMRI has helped to define the neurobiology of cosamine and chondroitin. My views are invariably acupuncture and has been used to assess fibromyal- neutral on whether this is sensible, though if asked gic pain, chronic back pain and response to NSAIDs I personally think they are very safe but regrettably in arthritis. Borsook & Becerra [28] believe fMRI may ineffective. Not every expert has reached the same con- revolutionize the understanding of pain, help evalu- clusion on the published evidence and I try to respect ate different pain states and develop new treatments. the views of other clinicians on this complex problem. Although no research has yet been undertaken in osteo-arthritis, it is inevitable that fMRI will be a pow- References erful instrument in assessing pain in these patients and for identifying and evaluating new treatment strategies. 1. Dieppe PA, Lohmander LS. Pathogenesis and management of pain in osteoarthritis. Lancet 2005; 365: 965–973. The other key need is to integrate physical and drug treatments. Pain control in osteo-arthritis is not 2. Felson DT. An update on the pathogenesis and epidemiology just a matter of better drugs. Improved fitness and of osteoarthritis. Radiol Clin North Am 2004; 42: 1–9. regular exercise are equally if not more important. As well as prescribing drug treatment, clinicians ought 3. Altman R, Asch E, Bloch D, et al. Development of criteria for to prescribe exercise regimens for patients as these the classification and reporting of osteoarthritis. Classification appear equally effective. of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Author recommendations Association. Arthritis Rheum 1986; 29: 1039–1049. There is no single simple treatment for osteo-arthritis, 4. Dillon CF, Rasch EK, Gu Q, Hirsch R. Prevalence of knee and therapy needs to be individually tailored. Equally, osteoarthritis in the United States: arthritis data from the the disease is variable and over time patients will need Third National Health and Nutrition Examination Survey a range of different treatments. 1991–94. J Rheumatol 2006; 33: 2271–2279. 5. Felson DT, Zhang Y, Hannan MT, et al. The incidence and natural history of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis Rheum 1995; 38: 1500–1505. 6. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status 104

Pain associated with osteo-arthritis instrument for measuring clinically important patient rel- of osteoarthritis of the knee. Cochrane Database of evant outcomes to antirheumatic drug therapy in patients Systematic Reviews 2006, Issue 2. Art. No.: CD005328. DOI: with osteoarthritis of the hip or knee. J Rheumatol 1988; 15: 10.1002/14651858.CD005328.pub2. 1833–1840. 17. Bellamy N, Campbell J, Welch V, Gee TL, Bourne R, Wells 7. Towheed T, Maxwell L, Judd M, Catton M, Hochberg MC, GA. Viscosupplementation for the treatment of osteoarthri- Wells GA. Acetaminophen for osteoarthritis. Cochrane tis of the knee. Cochrane Database of Systematic Reviews Database of Systematic Reviews 2006, Issue 1. Art. No.: 2006, Issue 2. Art. No.: CD005321. DOI: 10.1002/14651858. CD004257. DOI: 10.1002/14651858.CD004257.pub2. CD005321.pub2. 8. Avouac J, Gossec L, Dougados M. Efficacy and safety of opi- 18. Modawal A, Ferrer M, Choi HK, Castle JA. Hyaluronic acid oids for osteoarthritis: a meta-analysis of randomized con- injections relieve knee pain. J Fam Pract 2005; 54: 758–767. trolled trials. Osteoarthritis Cartilage 2007; 15(8): 957–965. 19. Canter PH, Wider B, Ernst E. The antioxidant vitamins 9. Cepeda MS, Camargo F, Zea C, Valencia L. Tramadol A, C, E and selenium in the treatment of arthritis: a sys- for osteoarthritis: a systematic review and metaanalysis. tematic review of randomized clinical trials. Rheumatology J Rheumatol 2007; 34: 543–555. (Oxford) 2007; 46(8): 1223–1233. 10. Bjordal JM, Ljunggren AE, Klovning A, Slordal L. Non- 20. Ernst E. Complementary or alternative therapies for oste- steroidal anti-inflammatory drugs, including cyclo-oxygen- oarthritis. Nat Clin Pract Rheumatol 2006; 2: 74–80. ase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of 21. Vlad SC, Lavalley MP, McAlindon TE, Felson DT. randomised placebo controlled trials. BMJ 2004; 329: 1317. Glucosamine for pain in osteoarthritis: why do trial results 11. Scott DL, Berry H, Capell H, et al. The long-term effects of differ? Arthritis Rheum 2007; 56: 2267–2277 non-steroidal anti-inflammatory drugs in osteoarthritis of the 22. Towheed TE, Anastassiades TP. Glucosamine therapy for knee: a randomized placebo-controlled trial. Rheumatology osteoarthritis. J Rheumatol 1999; 26: 2294–2297. (Oxford) 2000; 39: 1095–1101. 23. Reichenbach S, Sterchi R, Scherer M, et al. Meta-analysis: 12. Moore RA, Derry S, Phillips CJ, McQuay HJ. Nonsteroidal chondroitin for osteoarthritis of the knee or hip. Ann Intern anti-inflammatory drugs (NSAIDs), cyxlooxygenase-2 selec- Med 2007; 146: 580–590. tive inhibitors (coxibs) and gastrointestinal harm: review of 24. Soeken KL. Selected CAM therapies for arthritis-related clinical trials and clinical practice. BMC Musculoskelet Disord pain: the evidence from systematic reviews. Clin J Pain 2006; 7: 79. 2004; 20: 13–18. 13. Kearney PM, Baigent C, Godwin J, Halls H, Emberson JR, 25. Little CV, Parsons T, Logan S. Herbal therapy for treating Patrono C. Do selective cyclo-oxygenase-2 inhibitors and osteoarthritis. Cochrane Database of Systematic Reviews traditional non-steroidal anti-inflammatory drugs increase 2000, Issue 4. Art. No.: CD002947. DOI: 10.1002/14651858. the risk of atherothrombosis? Meta-analysis of randomised CD002947. trials. BMJ 2006; 332: 1302–1308. 26. Long L, Soeken KL, Ernst E. Herbal medicines for the treat- 14. Lin J, Zhang W, Jones A, Doherty M. Efficacy of topical ment of osteoarthritis: a systematic review. Rheumatology non-steroidal anti-inflammatory drugs in the treatment 2001; 40: 779–793. of osteoarthritis: meta-analysis of randomised controlled 27. Bjordal JM, Klovning A, Ljunggren AE, Slordal L.Short- trials. BMJ 2004; 329: 324–326. term efficacy of pharmacotherapeutic interventions in 15. Towheed TE. Pennsaid therapy for osteoarthritis of the osteoarthritic knee pain: a meta-analysis of randomised knee: a systematic review and metaanalysis of randomized placebo-controlled trials. Eur J Pain 2007; 11: 125–138. controlled trials. J Rheumatol 2006; 33: 567–573. 28. Borsook D, Becerra LR. Breaking down the barriers: fMRI 16. Bellamy N, Campbell J, Welch V, Gee TL, Bourne R, applications in pain, analgesia and analgesics. Mol Pain Wells GA. Intraarticular corticosteroid for treatment 2006; 2: 30. 105

C HAP TE R 10 Pain associated with rheumatoid arthritis Paul Creamer1 and Sarah Love-Jones2 1Southmead Hospital, Bristol, UK 2 Frenchay Hospital, Bristol, UK Background Rheumatoid arthritis (RA) is a multisystem inflam- Figure 10.1 Rheumatoid hands, showing typical matory disorder principally affecting the synovial lin- deformities associated with advanced disease. ings of joints and tendons (Fig. 10.1). rheumatica, can be a presenting symptom, especially Pain is the most common symptom of RA and in the elderly. is primarily located in and around affected joints. Early disease commonly affects metacarpophalan- Other symptoms of RA include stiffness, especially geal (MCP), metatarsophalangeal (MTP), wrist and morning stiffness. The duration of early morning stiff- proximal interphalangeal (PIP) joints, with knees, ness is a good measure of the degree of inflammation shoulders and hips also involved in many patients. or “disease activity.” Systemic symptoms such as fatigue, Affected joints are painful, swollen, warm and tender. fever, anemia and weight loss also occur, especially dur- The other cardinal sign of inflammation, redness ing inflammatory episodes. Extra-articular involvement (erythema), is unusual in RA and should alert the may affect eyes, nervous system, heart, lungs and skin. doctor to the possibility of super-added infection. Other synovial joints can be affected, including, Rheumatoid arthritis is a phasic disease. Periods for example, temporomandibular (TMJ), giving of inflammation (often described by the patient as jaw pain, or cricoarytenoid, giving throat pain and “flares”) are interspersed with quiescent phases in hoarse voice. Neck pain is common in RA and usu- which the inflammatory response is reduced. Joints ally reflects facet joint inflammation: however, an may still be painful at these times, due to damage that important complication is atlantoaxial disease, which has already been done. Although RA rarely becomes can result in subluxation and potentially cervical cord completely “burnt out” or noninflammatory, as disease compression. In addition to joints, other structures duration increases the cause of pain becomes predom- lined by synovium, such as tendons and bursae, can inantly mechanical or due to secondary osteo-arthritis. also become inflamed and painful. It is important Assessment of pain in RA must attempt to determine to consider the anatomic source of pain, since that directs treatment. “Wrist pain” may therefore be due to synovitis of the wrist joint itself or tenosynovitis of the extensor tendons; “elbow” pain may arise from the elbow joint or the olecranon bursa, and so on. Muscle pain and stiffness, similar to that seen in polymyalgia Evidence-Based Chronic Pain Management. Edited by C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell Publishing. 106

Pain associated with rheumatoid arthritis Table 10.1 Differentiation between inflammatory and noninflammatory pain in RA Inflammatory Noninflammatory Pain ϩϩ ϩϩ ϩϩ (soft tissue) ϩ/Ϫ (bony) Swelling ϩϩ (early morning) ϩ/Ϫ (inactivity/gelling) ϩϩϩ Ϫ Stiffness ϩϩϩ Ϫ Extra-articular features Fatigue, malaise, weight Raised Normal loss ϩϩϩ ϩϩϩ ESR, CRP Response to analgesia ϩϩϩ Ϫ Response to DMARDs CRP, C-reactive protein; DMARDs, disease-modifying antirheumatic drugs; ESR, erythrocyte sedimentation rate. if the pain is inflammatory or noninflammatory, as the treatment for these will be different. Some clues to aid differentiation are listed in Table 10.1. Pathophysiology The primary site of change in RA is the synovium, Figure 10.2 X-ray changes of RA. which becomes swollen, hypertrophic and inflamed. Lymphocytes and plasma cells invade the tissue synovial tissue outstripping the capacity of the nerve and excess synovial fluid production leads to joint supply. This results in a drop-out in laminae I and II swelling and effusion. Erosion of bone, due to of the dorsal horn. Aβ (proprioceptive) fibers replace local pressure from inflamed tissue or the effects of the pain fibers in these laminae so normal joint move- destructive cytokine release, occurs at the periarticu- ment can be perceived as pain. lar regions of the joint. These changes are first seen on X-ray at the MCP, MTP and PIP joints and around Inflammation, such as that seen in RA, has little effect the ulna styloid at the wrist (Fig. 10.2). It is now clear on Aβ fibers but Aδ and C become sensitized with an that X-ray is a relatively insensitive way of detecting increased response to movement or pressure. Previously erosions and MRI or ultrasound can detect lesions at “silent” fibers develop mechanically sensitive fields so a preradiographic stage (potentially allowing earlier, that simply moving the joint results in stimulation of more aggressive treatment). Untreated, continued pain afferents. The mechanisms responsible for this bone and cartilage destruction leads to irreversible sensitization are complex and involve many mediators deformity and secondary osteo-arthritis. such as bradykinin and prostaglandins; hence the effect is blocked by aspirin and other NSAIDs. Peripheral sen- Mechanisms of pain in RA sitization also results in spinal cord changes, leading Primary articular nerves consist of unmyelinated to expanded receptive fields so that a spinal neurone (IV, C fibers 80%) and myelinated (III, Aδ and II Aβ, responds to stimuli from a wider area. Opioid receptors 20%) fibers. Aβ fibers end in corpuscles. Aδ and C are found in the synovium and their expression may be fibers are free nerve endings; in the normal joint these increased with inflammation of the joint. are widely distributed throughout the synovium, cap- sule, ligaments and bone. The synovial membrane is not only richly innervated with C fibers but also with sympathetic nerves, whereas cartilage has no inner- vation at all. In RA there is a loss of C fibers from the synovium, possibly due to the rapid growth of 107

Chapter 10 Epidemiology secondary to the development of permanent damage or erosions in the joints. This change is irreversible The epidemiology of RA pain closely reflects that of so treatment is directed not only at alleviating the RA itself. RA is the most common inflammatory joint patient’s symptoms but also at reducing the chance disease, with an annual incidence in males of between of erosive progression. No single drug is capable of 0.15 and 0.46 per 1000 and in females between 0.24 halting progression. and 0.88 per 1000 [1], different incidence rates reflecting the different methods used to detect disease. Of the 600,000 people in the UK with RA, 10,000– Although peak age presenting to secondary care is 12,500 will have severe disease, which has failed 30–50 years, community studies suggest that inci- to respond to current therapy. RA is not a benign dence increases with age until the seventh decade. In disease, but carries significant morbidity and a stand- all except the oldest age groups, women are affected ardized mortality ratio of over 1.5 [3]. Survival rates more than men. The overall prevalence of RA in the for patients with RA are comparable with Hodgkin’s UK is about 1% [2], with similar levels found across disease, diabetes mellitus, stroke and triple-vessel cor- Europe and North America. Both the incidence and onary artery disease [4]. Functional disability (reflect- severity of RA may be declining over time, particularly ing disease activity) and low socio-economic status in females. are major contributors to early mortality. The cause of RA is unknown but, as for most dis- Even in the early stages of the disease, patients expe- eases, represents an interaction between genetic and rience significant morbidity. One study of patients environmental causes. There is increased familial with a mean of 3 years’ disease duration found that aggregation and greater concordance in monozy- over half reported significant impact on three or more gotic (12–15%) than dizygotic (4%) twins. The most areas (work, income, daytime rest, leisure, mobility, clearly associated genetic link is with the major histo- housing and social support) while 42% were regis- compatibility complex antigen HLA DR4, presence of tered work disabled [5]. Radiographic damage (which which conveys a sixfold increase in risk of incidence predicts disability) occurs mostly in the first years of of RA (and also increased disease severity). However, the disease [6]. Early effective treatment is therefore HLA-DR4 accounts for only about 30–50% of genetic vital to control symptoms, prevent joint damage and susceptibility and other, as yet undetermined genes improve function. must be involved. The mechanism by which this gene acts as a susceptibility factor for autoimmune events A number of risk factors have been identified remains unclear. which are associated with an unfavorable prognosis in terms of pain as well as function and joint destruc- Environmental triggers for RA have not been clari- tion. These include: fied, despite extensive searches for infectious organ- • evidence of joint damage (erosions on plain X-ray) isms. Many patients blame their RA on a stressful event; though this is unlikely to be causal, there at baseline is increasing evidence that stress may modify the • persistently raised inflammatory markers (CRP, immune system and be associated with exacerbations of RA. Hormones clearly play a role in RA. Before the ESR) menopause, there is an excess of females affected but • functional disability at baseline (as measured, for this gender difference disappears after the menopause. Pregnancy usually results in amelioration of symptoms example by Health Assessment Questionnaire) though a postpartum flare is often seen. Finally, the • presence of extra-articular features such as nodules contraceptive pill may be weakly protective against RA. • psychosocial problems • rheumatoid factor seropositivity Risk factors • presence of susceptibility genes (HLA-DR4 0401, The long-term joint consequences of RA (pain, 0404 or 0405). disability and need for total joint replacement) are Rheumatoid arthritis is an expensive disease to treat [1, 4]. Using various models [7], it is estimated that the annual cost of treating RA in the UK is £0.8–£1.3 billion [1]; 50% of this is accounted for by hospitalization and long-term care, affecting only those with most severe disease. Five percent of patients (mostly those requiring 108

Pain associated with rheumatoid arthritis total joint replacement) account for more than 25% of Given that pain is the direct result of either the inflam- the total costs [7]. matory process underlying RA or of joint damage consequent on that inflammation, management of Although the primary aim in treating RA is first RA pain is essentially the management of RA itself. to reduce symptoms, especially pain, there is good The rest of this chapter, therefore, considers the evidence that early aggressive treatment can reduce management of RA with special reference to pain, long-term complications and is highly cost-effective but it should be appreciated that therapies that treat in the long term. the disease will also relieve pain (for evidence base see Box 10.1). Assessing pain in rheumatoid arthritis The need for early referral Early referral to a hospital specialist is desirable in Pain is almost universal in RA and is usually bilateral, all patients presenting with a persistent inflamma- symmetric and involves multiple joint sites. tory polyarthritis. Although RA is the most likely diagnosis, there is a differential including viral, Several techniques have been used to measure pain reactive, seronegative and crystal arthritis. Early in RA, including numerical rating scales, verbal rating accurate diagnosis enables the correct management scales, visual analog scales (VAS), questionnaires and to be instituted without delay. A vital part of early behavioral observation methods. The Arthritis Impact management is referral to other members of the Measurement Scales (AIMS) pain questions are valid multidisciplinary team for advice and education on and sensitive to change [8]. A simple 100 mm hori- joint protection, splinting, exercises and consideration zontal VAS with anchors of “no pain” at 0 mm and of work- and family-related issues. For most patients, “worst possible pain” at 100 mm is probably the sim- RA is a life-long diagnosis and care will be shared by plest and most sensitive tool [9]. Other pain assess- ment scales include the Brief Pain Inventory [10] and Box 10.1 Evidence base for the McGill Pain Questionnaire [11]. pain-reducing interventions in RA The overall pain experience in RA appears simi- Interventions supported by evidence lar to that of osteo-arthritis (OA) [12,13]. The words used to describe the pain are similar in both – Simple analgesia: paracetamol, codeine and other “aching,” “sharp,” “stabbing” – hence pain descrip- opiates, cannabinoids tors are not helpful in distinguishing OA from RA. NSAIDs In both conditions the affective component is more DMARDs intense than the sensory component, indicating the Steroids: oral, intra-articular importance of emotional factors in the pain experi- Biologic therapy ence. Overall pain intensity increases with disease Education duration. RA pain may differ from that of fibromy- Physiotherapy algia in that fibromyalgia patients report a greater Hydrotherapy variety of pain descriptors and the pain is less clearly Orthotics localized to joints, though the overall intensity is Fish oils similar [14]. Surgery Management of rheumatoid Commonly used interventions currently arthritis pain unproven As in other conditions, effective therapy of pain in RA Glucosamine should combine optimum analgesia with enhanced Hot and cold function and minimal side effects. Decisions regarding Ultrasound pain control should take into account type of pain Hydrotherapy (inflammatory versus mechanical, central versus peripheral) and the patient’s psychosocial situation. 109

Chapter 10 the consultant, GP and members of the multidiscipli- that a patient will have a larger improvement in nary team. functional status, for example will cease to be house- bound or be able to return to work. One of the major reasons for early referral is to ensure optimal pharmacologic treatment. The aims The Disease Activity Scale (DAS) is a similar of treatment are to improve symptoms, reduce disease weighted composite score of disease activity in RA, progression and prevent disability. Drug therapy con- developed by a group of European rheumatologists sists of analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs), disease-modifying antirheumatic Modalities of pain relief drugs (DMARDs) and steroids. Early intra-articular aspiration and injection with corticosteroid are often Pharmacologic extremely effective. Management paradigms for RA Analgesics have changed considerably in the last 10 years with the Although there is little controlled evidence regarding knowledge that early use of disease-modifying therapy efficacy of analgesics in RA [16], it is widely believed can reduce joint damage and, hence, long-term dis- that simple analgesics such as paracetamol are very ability. Early diagnosis and prompt intervention with helpful, providing quick pain relief with generally low disease modifying drugs have become standard. risk of toxicity. Simple analgesics are often used in combination with NSAIDs and DMARDs. Economic evaluation of managing early arthritis suggests that costs are high in the first few months Analgesic compounds with central opioid agonist of disease, with approximately half the cost being activity, such as pentazocine and oxycodone, have associated with work absence. Aggressive treatment been available clinically for a number of years for of disease in the early stages is therefore likely to be treatment of RA pain but they have not been utilized cost-effective. extensively due to their dysphoric side effects and concerns about dependence. Mild opioids such as To fully understand the clinical results to be pre- codeine are frequently used to supplement other ther- sented, it may be helpful to describe how severity is apy in controlling pain despite a paucity of data from defined in RA. The standard method of assessing dis- controlled studies, particularly on long-term use. ease activity and response to treatment is the American College of Rheumatology (ACR) response criteria [15]. A retrospective study of 640 patients with chronic The following variables are measured at baseline and at arthritis (including RA) found that 290/640 patients set intervals after commencement of therapy. had used opiates (codeine or oxycodone) in the past • Number of tender joints (of 28 assessed) 3 years [17]. Although this was an uncontrolled study, • Number of swollen joints (of 28 assessed) opioids significantly reduced pain severity scores, • Health Assessment Questionnaire from 8.2 to 3.6 (on a 0–10 scale) (P Ͻ 0.001). Mild • Patient assessment of disease activity (VAS) side effects were reported in 38%; nausea, dyspepsia, • Patient assessment of pain (VAS) constipation, and sedation were the most common. • Physician global assessment (VAS) The mean ϮSD initial dosage was 2.1 Ϯ 1.7 30 mg • Erythrocyte sedimentation rate or C-reactive codeine equivalent/day. Dosage escalations occurred in 32 patients and were attributable to worsening of protein the underlying arthritis in all but four patients, who If a 20% or greater reduction is seen in the number also displayed other abuse behaviors. The authors of tender joints, the number of swollen joints and concluded that prolonged treatment of rheumatic dis- at least three of the remaining five variables, the ease pain with mild opiates reduces pain severity and patient is said to have achieved an “ACR 20” response. is associated with only mild toxicity. Development An ACR 20 response is defined as the minimal of tolerance, requiring dose escalation, is rarely seen. acceptable response to be achieved for a patient This study and others suggest that concerns about to have responded to a drug and may mean that a opioid efficacy, toxicity, tolerance, and abuse or addic- patient becomes independent in an activity of daily tion should no longer be used to justify with-holding living such as bathing, which can have a large impact opioids from patients with well-defined rheumatic on quality of life. An “ACR 50” response is a 50% or disease pain. greater reduction. An ACR 50 response usually means 110

Pain associated with rheumatoid arthritis The use of opioid medication in RA patients study of 58 patients with RA, a cannabis-based enrolled in Tennessee’s Medicaid program was ana- medicine was found to be well tolerated and to sig- lyzed [18] and found to increase from 38% to 55% nificantly reduce pain and disease activity (as meas- (P Ͻ 0.001) in the study period of 1995–2004. This ured by DAS). Further studies are required to define study also showed an increase in the use of DMARD the role of cannabinoids in RA. and a decrease in the use of glucocorticoids in RA patients. NSAIDs There is good evidence that NSAIDs are effective at The consensus statement of an international expert reducing pain in RA and are usually the first drugs pre- panel focused on the six WHO step III opioids – scribed following symptom onset, often prior to refer- buprenorphine, fentanyl, hydromorphone, methadone, ral to a rheumatologist. Most patients will experience morphine and oxycodone – most often used in the about a 50% reduction in pain following introduction management of chronic severe pain in the elderly [19]. of NSAIDs. One placebo-controlled study of etoricoxib This population included those with RA. The panel and naproxen found that the percentage of patients concluded that opioids are efficacious in noncancer who achieved an ACR 20 responder criteria response pain (treatment data mostly level Ib or IIb) but needed was 41% in the placebo group, 59% in the etoricoxib individual dose titration and consideration of their group, and 58% in the naproxen group [26]. respective tolerability profiles. Many NSAIDs are available. Although there is A recent review of the evidence of efficacy of no evidence that one NSAIDs is consistently more opioids for chronic pain is relevant for RA patients effective than another at relieving RA pain, there is and concludes that there is strong evidence to support individual variation and it is worth switching NSAIDs the initial effectiveness of opioids for the treatment of in an attempt to find one that suits a particular chronic pain, with much less clarity about the long- patient. In general, a long-acting preparation is term effectiveness [20]. preferred, often given at night to maximize relief of early morning pain and stiffness. Selective COX-2 Opioid receptors are also present in periper- inhibitors (COXIBS) are more effective than placebo hal tissue [21], where opiates may exert an anti- in RA [27,28] though they are no more effective than inflammatory effect. Compared with standard traditional NSAIDs [29, 30]. therapy of an intra-articular dexamethasone applica- tion, intra-articular morphine resulted in greater pain The efficacy of NSAIDs at relieving pain can reduction and a significant reduction of the leukocyte present a diagnostic difficulty in that symptoms and count in the synovial fluid in patients with OA and signs of inflammation may be masked, leading to a RA [22]. This analgesic effect is due only in part to delay in diagnosis and hence initiating disease-modi- interaction with opioid receptors. The prolonged fying therapy [30]. Sometimes, therefore, it is neces- analgesic effect can also be explained by other mecha- sary to withdraw NSAIDs temporarily to fully assess nisms, such as opioid-mediated inhibition of inflam- the degree of disease-related inflammation. mation. Expression of opiate receptors is increased with inflammation [23] and in experimental arthritis, Nonsteroidal anti-inflammatory drugs are associ- administration of peripherally acting opioids shows ated with significant toxicity, notably on the gastroin- antinociceptive activity. Other animal studies have testinal (GI) system. The addition of cytoprotective confirmed that opiates are anti-inflammatory in a drugs such as misoprostil or proton pump inhibi- dose-dependent, time-dependent, stereoselective and tors may allow patients with RA to continue to take antagonist reversible manner [24]. The potential for NSAIDs, although it should be remembered that the opiates to reduce inflammation in human RA has not correlation between GI symptoms and serious GI tox- been studied. icity (perforation, ulcers, bleeds) is poor and addition of cytoprotection does not fully protect against seri- Two cannabinoid receptors (CB1 and CB2) are ous complications. COXIBS are probably associated expressed on neurones and immune cells, particu- with lower GI risks in RA [18] though this effect larly in inflammation. Receptor agonists such as may relatively small [31]. Other side effects which anandamide are analgesic. In the only double-blind may limit use of NSAIDs include effects on the liver placebo-controlled study of its kind [25], a 5-week 111

Chapter 10 and renal system [32]. About 15% of patients with steroid use is limited by toxicity, though in the short asthma experience worsening on starting an NSAID; term, harmful effects on, for example, bone mineral patients should be warned about this and may have density may be outweighed by an increase in strength to discontinue the drug if they are adversely affected following reduction in inflammatory mediators and but asthma should not be regarded as an absolute an improvement in patient mobility [39]. contraindication to NSAIDs. Disease-modifying antirheumatic drugs Recently it has become clear that while attention Disease-modifying antirheumatic drugs differ from on toxicity has focused on GI problems, there is likely NSAIDs in their ability to modify the natural history of to be a small but detectable risk of increased heart the disease and affect disease-related parameters such attacks and strokes in people taking coxibs. This prob- as systemic markers of inflammation (ESR, CRP) and lem was particularly linked to rofecoxib and this drug radiographic progression of erosions [40]. By reducing has now been withdrawn. Several subsequent studies inflammation, they also reduce pain – hence the “man- have considered whether other COX-2 drugs and also agement of RA pain” must include these drugs. The nonselective NSAIDs might carry a similar increased conventional pyramidal approach in which DMARDs risk. Data are conflicting [33, 34] but it seems likely are introduced gradually and only if NSAIDs fail to that a small increase in cardiac risk is seen with all provide adequate symptomatic relief has been super- NSAIDs [35] though there are theoretical reasons, seded by a more aggressive approach, with the early backed up by some data, to suggest that this is greater introduction of DMARDs, either singly or in combi- with coxibs than with nonselective NSAIDs. Current nation, with or without steroids, in the hope of con- guidelines suggest that all NSAIDs should be used trolling synovitis in the early stages of the disease and with caution in patients with cardiac disease. preventing joint damage [41, 42]. The patient should be continually assessed clinically, serologically and Corticosteroids radiologically to ensure that the chosen therapy is truly There is good evidence that corticosteroids can suppressing the disease; failure to do this should be an achieve rapid control of symptoms, including pain indication to modify DMARDs therapy. and maintenance of function [36], and are par- ticularly useful during flares of disease activity. Oral An ideal DMARDs should reduce synovitis, limit doses from 7.5 mg to 20 mg daily prednisolone are joint damage and improve function with minimal tox- often used, as are intravenous infusions of between icity and continued efficacy [43]. Unfortunately, none 500 mg and 3 g methylprednisolone. Local injec- of the DMARDs currently available fulfill these crite- tion of inflamed joints, using a long-acting steroid ria completely. The most frequently used DMARDs such as triamcinolone, is an excellent and safe way to are shown in Table 10.2. Oral gold, penicillamine and relieve pain rapidly. Steroids can also be given as an cyclophosphamide are used less frequently. intramuscular depot injection, for example methyl- prednisolone 120 mg (3 ml). This delivers the equiv- All DMARDs share certain limitations, listed alent of about 4 mg prednisolone daily for about 6 below. weeks. Although the average daily dose is small, some patients find such injections more effective than oral Delayed action. Most DMARDs take 2–3 months to prednisolone, possibly because the drug is released work and are rarely discontinued for lack of effect continuously rather than in the pulsatile fashion of before 6 months. There is therefore an inevitable oral therapy. window in which inflammation and joint damage may progress unchecked. Further, the delay in Evidence for a long-term disease-modifying effect response can lead to reduced compliance if the of prednisolone is conflicting. Combination therapy expectations of the patient are unrealistic. studies in which high doses of steroids are used ini- tially and then tailed off [37] show greater efficacy Toxicity. Disease-modifying antirheumatic drugs are than monotherapy. Addition of a fixed dose of pred- toxic. The major problems encountered are listed in nisolone 7.5 mg daily to standard DMARD results in Table 10.2. Some side effects resolve but others require a reduction in erosive progression [38]. Long-term cessation of treatment. On average 20–40% of patients 112

Pain associated with rheumatoid arthritis Table 10.2 Disease modifying antirheumatic drugs (DMARD) and their side effects DMARD Dose Common side effects Methotrexate 7.5 mg po or s/c weekly, increasing GI symptoms, teratogenicity, bone marrow suppression, Sulfasalazine up to 25 mg pulmonary hypersensitivity, abnormal liver function tests 1 g bd, increasing to max 1.5 g bd GI symptoms, rash, headache, abnormal liver function tests, bone marrow suppression (commonest in first Gold 10 mg IM test dose, 50 mg IM weekly 6 months), reversible reduced sperm count Hydroxychloroquine for 12–20 weeks then 50 mg monthly Rash, pruritus, mouth ulcers, proteinuria, bone marrow 200–400 mg/day; max 6 mg/kg/day suppression, vasomotor injection reactions Azathioprine 2.5 mg/kg/day Rash, retinopathy (rare, much less common than with chloroquine) Cyclosporine 2.5 mg/kg/day in divided 12 hrly doses, max 5 mg/kg/day Bone marrow suppression, GI disturbances, flu-like symptoms, increased risk of lymphoproliferative disorders Hypertension, raised creatinine, hyperkalemia, hypertrichosis, gum hyperplasia, hepatotoxicity will need to stop treatment due to toxicity within 2 certain drugs but since the mechanisms of action are years. When informed of the risks, many patients poorly understood, it is not possible to target specific elect not to take these drugs, though compliance can parts of the complex inflammatory process responsi- be improved by good communication, education and ble for RA. support from other members of the team. It should be stressed, for example, that although there is an Current practice is largely with sequential mono- undeniable risk of toxicity with DMARDs therapy, it therapy until a response is found. The choice of agent must be remembered that inadequate disease control is based on familiarity, likely compliance, tolerability of also carries significant risks in terms of quality of life, side effects, co-morbidity and concurrent medication as disability and mortality. well as consideration of which DMARDs have already been tried and failed. Sometimes changing the route of The need for monthly blood tests can be a disin- administration (e.g. subcutaneous injections of meth- centive to start or continue treatment and is costly otrexate) can improve efficacy. It is difficult to conclude in time and money. The safety of various DMARDs that a DMARDs is ineffective unless it has been tried at during pregnancy and breastfeeding is an important the maximum tolerable dose and for at least 4–6 months. issue as many patients will be of child-bearing age. Biologic therapy Unpredictable efficacy. Not only are currently available Although early aggressive use of traditional DMARDs, DMARDs relatively toxic but they are also of limited alone or in combination, has proven efficacy in RA, a efficacy. Only 70% of patients will respond to a given number of problems remain, as outlined above. The DMARDs in terms of a noticeable improvement in search for novel therapeutic targets has resulted in pain, swelling and function and even this may be lost major breakthroughs in RA treatment. with time. Data on the ability of DMARDs to limit radiographic damage are also disappointing although Cytokines are protein or glycoprotein molecules this may be improved with earlier introduction of that deliver important intercellular messages regulat- therapy. The results on long-term outcome are also a ing chronic inflammation and tissue damage in RA. cause for concern, as levels of disability remain high Cytokines such as tumor necrosis factor α (TNF-α), despite the use of DMARDs therapy for over 20 years. interleukin-1 (IL-1), interleukin-6 (IL-6), and granu- locyte macrophage-colony stimulating factor (GM- There is no way of predicting which patients will CSF) are abundant in inflamed joints and promote respond to which DMARDs. It is possible that subsets the influx of inflammatory neutrophils and mono- of patients exist who are more likely to respond to cytes to the joints. 113

Chapter 10 Tumor necrosis factor α appears to have a pivotal Abatacept is a selective co-stimulation modulator, role in perpetuating inflammation and joint damage which reduces T cell activation [48]. Rituximab selec- in RA [44]. Following evidence from animal arthritis tively depletes B cells from the circulation and is given models in which antibodies against TNF were shown by infusion every 9–18 months [49]. to prevent or reduce symptoms when administered either before or after the onset of disease, drugs to Cost efficacy of biologic therapies has been exam- specifically block the effects of TNF in humans were ined in detail, for example by the National Institute introduced for the treatment of RA in 1999. Three for Health and Clinical Excellence (NICE) in the UK. such drugs are now available: infliximab, etanercept Accurate estimates are extremely difficult to obtain and adalimumab. It is clear that such biologic thera- since cost savings may not be obtained for many years pies can produce a dramatic reduction in symptoms in the future. Simple assessment of incremental cost- and can halt or even reverse the joint damage caused effectiveness ratio gives about £66,000 per quality- by RA. Anti-TNF-α agents have potent analgesic adjusted life-year (QALY) for etanercept and £99,000 effect as well as their anti-inflammatory action. for infliximab. However, models which attempt to include future cost savings from reduced work loss, About 40% patients treated with anti TNF-α achieve reduced need for hospitalization and joint replace- an ACR 50 at 1 year [45, 46]. Although the ACR 50 ment give figures of £16,300 and £29,000 per QALY measures response to a range of outcomes, it is heavily respectively. It is also arguable that the pain reduction influenced by pain and a reduction in ACR 50 is inevi- achieved by these drugs has an inherent value over tably associated with a reduction in pain. For example, and above that of reduced service utilization. the ARMADA trial of adalimumab [47] reported a 55% ACR 50 response rate, with a 47% reduction in Tricyclic antidepressants self-reported pain levels. Such benefits are sustained The tricyclic antidepressants such as amitriptyline are for several years. Anti-TNF-α therapy has also been frequently used to treat chronic pain associated with shown capable of reducing erosive progression. osteo-arthritis and RA. A systematic review [50] of papers published between 1966 and 2007 on antide- All anti-TNF-α drugs are given by injection: IV infu- pressants in rheumatologic conditions looked at 78 sion every 8 weeks (infliximab), subcutaneous injection clinical studies and 12 meta-analyses from 140 papers. every 2 weeks (adalimumab) or twice weekly (etaner- The strongest evidence of an analgesic effect of anti- cept). Although generally well tolerated, anti-TNF- depressants has been obtained in fibromyalgia. A α drugs care associated with side effects, commonly weak analgesic effect is observed for chronic low back injection site reactions, mild flu-like illness or rash. pain, with an efficacy level close to that of analgesics The rate of infection is probably increased and serious such as NSAIDs. Specific comment was difficult to infections such as septic arthritis or tuberculosis are make in relation to RA and ankylosing spondylitis as well documented. Current practice includes advice to most trials identified in the review included patients screen for tuberculosis before starting treatment. There with osteoarthritic conditions. The authors concluded is a theoretical risk of increased malignancy: currently that TCAs have weak analgesic effects in RA patients it is unclear whether increased rates observed reflect with or without depressive symptoms. For patients drug toxicity or simply the greater disease severity of with ankylosing spondylitis, amitriptyline may mod- patients in whom these drugs are used. These drugs ify a number of symptoms including pain, fatigue and are expensive, costing about £10,000 per annum, but sleep disorders. increasingly these costs will be offset by reduced need for hospital admissions and joint replacement surgery. Complementary therapies Up to 60% patients with arthritis use some form of If a patient fails to respond to one anti-TNF-α complementary therapy such as acupuncture, Alexander drug, they may respond to another and it is worth technique, aromatherapy, massage, homeopathy and switching medication. However, a number of other reflexology. Many of these may provide relaxation and targets have been identified as contributing to temporary symptom relief, including pain reduction. the inflammatory process in RA and drugs have They may also improve the patient’s general well-being beendesigned specifically to neutralize their effect. A number of these are now available for use in RA. 114

Pain associated with rheumatoid arthritis and improve self-efficacy, but a robust evidence base for and improve function). A number of studies have most of these interventions does not exist. examined the effect of exercise on RA though many are limited by potential bias and limited patient Many patients also take nutritional supplements selection. De Jong et al. [55] used a program of (neutriceuticals). Omega-3 fatty acids, found in fish bi-weekly group exercises involving bicycle train- oils and some plant oils, are the first components of a ing at 70–90% predicted maximum heart rate, physiologic cascade that results in inhibition of pain- nine strengthening exercises and a sporting activ- producing mediators such as prostaglandin E2, LTB4 ity to deliver impact to joints and bones. This pro- and TNF-α. Fish oils have a modest but clear effect on gram improved function and disability, was safe for pain and swollen joints [51] and can reduce the need patients and did not increase disease activity though for NSAID. However, at least 3 g daily are required a specific effect on pain was difficult to measure. and many patients find such large doses impractical. Other aerobic and strengthening programs [56] Benefit may not be seen for up to 12 weeks. have confirmed these findings and also demon- strated improved muscle strength. The mechanism γ-linolenic acid (GLA) is an omega-6 essential fatty for any reduction in pain is unclear but may involve acid found primarily in vegetable oils. Several studies enhanced self-efficacy and modulation of central have suggested benefit in RA though trial methodol- pain as well as direct effects on muscle and joint. ogy is variable [52]. There is no evidence that glu- Hydrotherapy is also of value [57]. Many patients cosamine or chondroitin have an effect in RA. use hot and cold treatments – hot water bottles or heated wheat bags for heat; ice packs or a bag of fro- Nonpharmacologic zen peas for cold – and report pain relief, though Education firm evidence for benefit is lacking. Education is important in providing knowledge to patients about their disease. Other components Transcutaneous electrical nerve stimulation may include self-help principles, coping with stress Transcutaneous electrical nerve stimulation (TENS) and problem solving, nutrition, complementary may be of benefit, especially when combined with acu- therapy, physician/patient interaction, pacing skills puncture [58]. TENS is a form of electrotherapy and and so on. Benefits may include: reduction in pain, is thought to produce analgesia according to the gate improved self-efficacy, improved pain behavior, control theory [59]. Conventional TENS (C-TENS) is reduced medication and health resource utilization given at high stimulation frequency with low intensity, [53]. Education is generally considered to be cost whereas acupuncture-like TENS (AL-TENS) is given at effective, though the evidence base supporting this low frequency and high intensity. A systematic review intervention is very limited. Timing of education is [58] included three RCTs, involving 78 people, in important and the content will differ according to which C-TENS and AL-TENS were compared to pla- the stage of disease. cebo and each other. Administration of 15 minutes of AL-TENS a week, for 3 weeks, resulted in a significant Hammond & Freeman [54] found an educational- decrease in rest pain (67% relative benefit, 45 points behavioral training joint protection program to be absolute benefit on 100 mm VAS), but not in grip pain, more effective than a standard arthritis education compared to placebo. AL-TENS did result in a clinical program: although pain improved in both groups, beneficial improvement in muscle power scores with joint protection adherence, early morning stiffness, a relative diference of 55%, and an absolute benefit of function and hand deformities were all better in the 0.98, compared to placebo. No significant difference education-behavioral group. These benefits persisted was found between one 20-minute treatment dura- 4 years after the intervention. tion of C-TENS versus AL-TENS, or C-TENS versus placebo on decrease in mean scores for rest pain or Physiotherapy and exercise grip pain, or on the number of tender joints. Results Three types of exercise may have a role in RA: range showed a statistically significant reduction in joint of movement (to help maintain or increase flex- tenderness, but no clinical benefit from C-TENS over ibility); strengthening exercise (to maintain mus- cle strength) and aerobic/endurance exercise (to improve cardiovascular function, reduce weight 115

Chapter 10 placebo in relief of joint tenderness. No statistically significant difference was shown between 15 days of treatment with C-TENS or AL-TENS in relief or joint pain, although there was a clinically important benefit of C-TENS over AL-TENS on patient assessment of change in disease (risk difference 21%, number needed to treat (NNT) 5). Acupuncture Figure 10.3 Splinting in RA. Acupuncture is used as an adjunct therapy for the treatment of RA pain. A Cochrane Database review of a walking stick can reduce knee pain when held [60] evaluated the effects of acupuncture and in the contralateral hand by “unloading” the joint. electroacupuncture on the objective and subjective Metatarsalgia (pain over the MTP joint in the fore- measures of disease activity in patients with RA. Two foot) is a common problem in early RA. Provision of comparative controlled studies involving 84 patients an insole with forefoot support can effectively reduce were included. One study used acupuncture while pain [62]. the other used electroacupuncture. In the acupunc- ture study, although not statistically significant, pain Surgery in the treatment group improved by four points on The major indications for surgery in RA are pain and a 0–100 mm VAS versus no improvement in the pla- functional loss despite full medical treatment. In well- cebo group. In the second, electroacupuncture study, selected patients surgery is highly effective at reliev- a significant decrease in knee pain was reported in the ing pain, even at the expense of some mobility. RA experimental group, 24 hours post treatment, when patients are probably at greater risk of complications compared to the placebo group (weighted mean dif- than other arthritis patients, due to co-morbidity and ference (WMD) Ϫ2.0 with 95% CI Ϫ3.6 to Ϫ4.0). A drug effects. In general, surgery is of benefit to patients significant decrease was found also at 4 months post with advanced, established RA in which irreversible treatment (WMD Ϫ0.2, 95% CI Ϫ0.36 to –0.04). destructive change has occurred. Exceptions to this Although the results show both acupuncture and might include synovectomy or arthroscopic lavage for electroacupuncture may be beneficial to reduce symp- persistently inflamed joints but in general inflamma- tomatic knee pain in patients with RA, the reviewers tion is better treated with drugs. concluded that the poor quality of the trial, including the small sample size, precluded its recommendation. Synovectomy is a useful way of reducing a large mass of inflammatory tissue and is often combined Occupational therapy with another procedure – for example, wrist synovec- Occupational therapy interventions include advice on tomy may be combined with excision of the radial joint protection, energy conservation and problem- head. Tenosynovectomy and tendon reconstruction solving skills, instruction about assistive devices and can also be performed. provision of splints. There is good evidence that “instruction on joint protection” is beneficial and Joint fusion (arthrodesis) eliminates joint motion provision of splints (Fig. 10.3) has been shown to and hence pain though it does increase stress on adja- decrease pain [61]. As with education, the timing of cent joints: wrist, IP joint and talonavicular joints are intervention is important, positive changes being all amenable to fusion. more likely if the patient perceives it as relevant to their needs at the time. Joint replacement (e.g. hip, knee) is highly effective at reducing pain in appropriately selected patients: at Orthotics 3 years pain decreased by 29% in RA patients having In addition to splints, other forms of orthoses can be very helpful in reducing pain. Simple provision 116

Pain associated with rheumatoid arthritis total hip replacemement (compared with 56% in Many patients describe flares of RA, with OA). For total knee replacement, a 19% fall in pain in exacerbation of all symptoms including pain, after RA was seen compared with 48% in OA [63]. stressful events such as bereavement or divorce. The relationship between mood, stress and the immune Surgery in RA is often multiple and requires careful response is another potential area of research. Early advance planning. Some practical rules include: work suggests that abnormalities in the hypotha- • stage surgery to avoid immobilization and preserve lamic-pituitary-adrenal axis may provide a link between psychosocial stress and disease activity function in RA. • replace lower limb joint before upper, as rehabili- Personal recommendations tation using crutches or sticks is much harder on for management replaced joints • replace hips before knees as during total hip Management of pain in RA requires a multidiscipli- replacement the knee will be manipulated and nary team approach and early referral to secondary stressed care is essential [64]. Confirmation of the diagnosis is • correct foot problems if possible before large joint by clinical assessment supported by simple investiga- replacement. tions such as a measure of the acute phase response (CRP or plasma viscosity), measurement of rheu- Future research into pain matoid factor (RF) and plain radiography of hands associated with rheumatoid and feet. If there is diagnostic doubt (for example, arthritis if a positive RF is found but the patient is clinically atypical for RA, raising the possibility that the RF Pain is an almost inevitable consequence of RA: is false positive), measurement of anti-CCP anti- research into better ways of treating the disease will bodies and use of more sensitive methods to detect therefore improve outcomes in terms of pain. The inflammation such as ultrasound of the small joints introduction of biologic therapies has revolution- of the hand should be employed. NSAID should be ized our treatment: future research should continue started early, unless contraindicated. Confirmation to identify new components of the immune response of diagnosis should lead to aggressive therapy, with that might be amenable to targeted therapy. This is disease-modifying drugs being introduced if signs of already happening to some extent: B cells, IL-1, IL-6, inflammation persist 3 months after onset. My pre- IL-10, adhesion molecules and co-stimulatory mol- ferred drug is methotrexate, initially 7.5 mg weekly, ecules are all being investigated as potential targets increasing by 2.5 mg every month until disease con- for inhibition. New biologic drugs should maxi- trol is achieved. Alternatives would be sulfasalazine mize efficacy whilst minimizing toxicity, especially 1 g bd. Prednisolone 7.5 mg daily with appropriate infection. Data on the long-term effects of these bone protection should be added for the first 2 years drugs should continue to be collected, for example in all patients at high risk of progression and others using national registries such as the BSR Biologics in whom the inflammatory response is inadequately Database. Cost is currently a limiting factor in use controlled. of biologics and novel (cheaper) ways of making such drugs would potentially allow extension of Failure to respond to a DMARD (as measured by their use. Finally, identifying those patients likely to DAS) should trigger a switch to another DMARD. respond to a particular drug would be of great value Failure of three or more DMARDs, alone or in in reducing the time and money lost by using expen- combination, should result in anti-TNF therapy being sive therapies that are subsequently found to be inef- initiated. Failure of one anti-TNF drug is followed by fective in that individual. Genetic study of response rituximab or a second anti-TNF drug. to drugs (pharmacogenomics) will offer the poten- tial to, for example, identify whether one particular On diagnosis, all patients should be assessed by a individual is more likely to respond to anti-TNF physiotherapist, occupational therapist and specialist whist another may respond to anti-B cell and a third nurse and the pain-controlling measures outlined to anti-IL-10. 117

Chapter 10 above initiated. Surgical referral, ideally to a surgeon 2. Silman AJ. Epidemiology of rheumatoid arthritis. APMIS with a special interest in surgical management of RA, 1994; 102: 721–728. is undertaken as required. 3. Pincus T, Callahan LF. Taking mortality in rheumatoid Potential barriers to implementation of this algo- arthritis seriously – predictive markers, socioeconomic sta- rithm include: tus and comorbidity. J Rheumatol 1986; 13: 841–845. • delayed presentation by patients to primary care • delayed recognition and referral from primary to 4. Markenson JA. Worldwide trends in the socio-economic impact and long-term prognosis of rheumatoid arthritis. secondary care Semin Arthritis Rheum 1991; 110: 4–12. • centrally directed political constraints such as 5. Albers JM, Kuper HH, van Riel PL, et al. Socio-economic a move to place all rheumatology care in the consequences of rheumatoid arthritis in the first years of community the disease. Rheumatology 1999; 38: 423–430. • lack of provision of a multidisciplinary team • lack of resources to monitor patients closely (essen- 6. van der Heijde DMFM, van Leeuwen MA, van Riel PLCM, tial if lack of response is to be picked up quickly et al. Biannual radiographic assessments of hands and feet and changes made) in a three year follow up of patients with early rheumatoid • financial constraints on provision of drugs, espe- arthritis. Arthritis Rheum 1992; 35: 26–34. cially biologic therapies • lack of good patient-centered outcome measures. 7. Gabriel SE, Crowson CS, Luthra HS, et al. Modelling the lifetime costs of rheumatoid arthritis. J Rheumatol 1999; 26: Conclusion 1269–1274. Pain is the most important symptom of RA: about 8. Buchbinder R, Bombardier C, Yeung M, Tugwell P. Which 50% patients describe themselves as being in outcome measures should be used in rheumatoid arthritis “constant pain.” Management of RA pain is best clinical trials? Clinical and quality-of-life measures, respon- achieved by a multidisciplinary approach tailored siveness to treatment in a randomized controlled trial. to the individual and using a variety of drug and Arthritis Rheum 1995; 38: 1568–1580. nondrug treatments. The introduction of TNF-α blockade has revolutionized the treatment of RA, 9. Bellamy N, Campbell J, Syrotuik J. Comparative study of though cost, nonresponse, increased risk of infec- self-rating pain scales in rheumatoid arthritis patients. Curr tion and possible increased risk of malignancy are Med Res Opin 1999; 15: 121–127. likely to limit their use to those patients at risk of severe disease. Induction of remission with TNF- 10. Cleeland CS, Ryan KM. Pain assessment: global use of the α blockade and maintenance with traditional Brief Pain Inventory. Ann Acad Med Singapore 1994; 2: DMARDs is a potentially attractive approach for 129–138. the future as it may reduce the need for long-term biologic therapy. 11. Melzack R. The McGill Pain Questionnaire. Pain 1975: 3: 277–299. Many other biologic therapies are now in develop- ment for use in RA (and other automimmune dis- 12. Charter RA, Alexis M, Nehemkis AM, et al. The nature of eases). With early diagnosis and effective treatment, a arthritis pain. Br J Rheumatol 1985; 24: 53–60. complete cure for RA may be a realistic goal within the next 10 years. 13. Helliwell PS. The semeiology of arthritis: discriminat- ing between patients on the basis of their symptoms. Ann References Rheum Dis 1995; 54: 924–926. 1. Scott DL, Shipley M, Dawson A, Edwards S, Symmons DP, 14. Leavitt F, Katz RS, Golden HE, Glickman PB, Layfer LF. Woolf AD. The clinical management of rheumatoid arthri- Comparison of pain properties in fibromyalgia patients tis and osteoarthritis: strategies for improving clinical effec- and rheumatoid arthritis patients. Arthritis Rheum 1986; tiveness. Br J Rheumatol 1998; 37: 546–554. 29: 775–781. 15. Felson DT, Anderson JJ, Boers M, et al. The American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995; 386: 727–735. 16. Wienecke T, Gøtzsche PC. Paracetamol versus nonster- oidal anti-inflammatory drugs for rheumatoid arthritis. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD003789. DOI: 10.1002/14651858.CD003789.pub2. 17. Ytterberg SR, Mahowald ML, Woods SR. Codeine and oxy- codone use in patients with chronic rheumatic disease pain. Arthritis Rheum 1998; 41: 1603–1612. 18. Grijalva CG, Chung CP, Stein CM, Mitchel Jr EF, Griffin MR. Concise Report: changing patterns of medication use in patients with rheumatoid arthritis in a Medicaid popula- tion. Rheumatology 2008; 47: 1061–1064. 19. Pergolizzi J, Boger RH, Budd K, et al. Opioids and the man- agement of chronic severe pain in the elderly: Consensus 118

Pain associated with rheumatoid arthritis statement of an International expert panel with focus on effects. A randomized, double-blind, placebo-controlled the six clinically most often used WHO step III opioids clinical trial. Ann Intern Med 2002; 136: 1–12. (Buprenorphine, fentanyl, hydromorphone, methadone, 37. Landewé RB, Boers M, Verhoeven A, et al. COBRA morphine, oxycodone). Pain Pract 2008; 8(4): 287–313. combination therapy in patients with early rheumatoid 20. Ballantyne JC, Shin NS. Efficacy of opioids for chronic pain: arthritis: long-term structural benefits of a brief interven- a review of the evidence. Clin J Pain 2008; 6: 469–478. tion. Arthritis Rheum 2002; 46: 347–356. 21. Stein C. Peripheral mechanisms of opioid analgesia. Anesth 38. Kirwan JR. The effect of glucocorticoids on joint destruc- Analg 1993; 76: 182–191. tion in rheumatoid arthritis. The Arthritis and Rheumatism 22. Stein A, Yassoouridis A, Szopko C, et al. Intraarticular mor- Council Low-Dose Glucocorticoid Study Group. N Engl phine versus dexamethasone in chronic arthritis. Pain 1999; J Med 1995; 333: 142–146. 3: 525–532. 39. Haugeberg G, Strand A, Kvien T, Kirwan J. Reduced loss of 23. Kidd BL, Urban LA. Mechanisms of inflammatory pain. Br hand bone density with prednisolone in early rheumatoid J Anaesth 2001; 87: 3–11. arthritis. Arch Intern Med 2005; 165: 1293–1297. 24. Walker JS. Anti-inflammatory effects of opioids. In: 40. van der Heijde DM, van Riel PL, Nuver-Zwart IH, Gribnam Machels H, Stein C (eds) Immune Mechanisms of Pain and FW, van de Putte LB. Effects of hydroxychloroquine and Analgesia. Landes Bioscience, Austin, TX, 2002. sulphasalazine on progression of joint damage in rheuma- 25. Blake DR, Robson P, Ho M, Jubb RW, McCabe CS. toid arthritis. Lancet 1989; i: 1036–1038. Preliminary assessment of the efficacy, tolerability and 41. van der Heide A, Jacobs JW, Bijlsma JW, et al. The effective- safety of a cannabis-based medicine (Sativex) in the treat- ness of early treatment with “second line” anti-rheumatic ment of pain caused by rheumatoid arthritis. Rheumatology drugs. A randomized, controlled trial. Ann Intern Med 1996; 2006; 45: 50–52. 124: 699–707. 26. Collantes E, Curtis SP, Lee KW. A multinational rand- 42. Boers M, Verhoeven AC, Markusse HM, et al. Randomised omized, controlled, clinical trial of etoricoxib in the treat- comparison of combined step down prednisolone, ment of rheumatoid arthritis. BMC Fam Pract 2002; 3: 10. methotrexare and sulphasalazine with sulphasalazine 27. Matsumoto AK, Melian A, Mandel DR, et al. Etoricoxib alone in early rheumatoid arthritis. Lancet 1997; 350: RA Study Group. A randomized, controlled, clinical trial 309–318. of etoricoxib in the treatment of RA. J Rheumatol 2002; 29: 43. Edmonds JP, Scott DL, Furst DE, Brooks PM, Paulus HE. 1623–1630. Anti-rheumatic drugs: a proposed new classification. 28. Bensen W, Weaver A, Espinoza L, et al. Efficacy and safety Arthritis Rheum 1993; 36: 336–339. of valdecoxib in treating the signs and symptoms of RA: 44. Feldmann M, Brennan FM, Chantry D, et al. Cytokine a randomized, controlled comparison with placebo and production in the rheumatoid joint: implications for treat- naproxen. Rheumatology 2002; 41: 1008–1016. ment. Ann Rheum Dis 1990; 49: 480. 29. Emery P, Zeidler H, Kvien KT, et al. Celecoxib versus 45. Lipsky PE, van der Heijde DM, St Clair EW, et al. Infliximab diclofenac in long-term management of rheumatoid arthri- and methotrexate in the treatment of rheumatoid arthritis. tis: randomized double-blind comparison. Lancet 1999; N Engl J Med 2000; 343: 1594–1602. 354: 2106–2111. 46. Weinblatt ME, Kremer JM, Bankhurst AD, et al. A trial 30. Marzo-Ortega H, Green MJ, Karim Z, et al. NSAIDs of etanercept, a recombinant tumour necrosis factor alter the presentation of early inflammatory arthritis. receptor:Fc fusion protein in patients with rheumatoid Rheumatology 2000; 39(suppl 1): 42. arthritis receiving methotrexate. N Engl J Med1999; 340: 31. Juni P, Rutjes AW, Dieppe PA. Are selective COX 2 inhibi- 253–259. tors superior to traditional NSAIDs? BMJ 2002; 324: 47. Weinblatt ME, Keystone EC, Furst DE, et al. Adalimumab, 1287–1288. a fully human anti-tumor necrosis factor monoclonal anti- 32. Brater DC, Harris C, Redfern JS, Gertz BJ. Renal effects of body, for the treatment of rheumatoid arthritis in patients COX-2-selective inhibitors. Am J Nephrol 2001; 21: 1–15. taking concomitant methotrexate: the ARMADA trial. 33. Solomon DH, Schneeweiss S, Glynn RJ et al. Relationship Arthritis Rheum 2003; 48: 35–45. between selective cyclooxygenase-2 inhibitors and acute MI 48. Genovese MC, Becker JC, Schif M, et al. Abatacept for in older adults. Circulation 2004; 109: 2068–2073. rheumatoid arthritis refractory to tumor necrosis factor α 34. Ray WA, Stein CM, Daugherty JR, et al. COX-2 selec- inhibition. N Engl J Med 2005; 353: 1114–1123. tive NSAIDs and risk of serious CHD. Lancet 2002; 360: 49. Smolen JS, Keystone EC, Emery P, et al., the Working Group 1071–1073. on the Rituximab Consensus. Consensus statement on the 35. Graham D, Campen D, Hui R, et al. Risk of acute MI use of rituximab in patients with rheumatoid arthritis. Ann and sudden cardiac death in patients treated with COX 2 Rheum Dis 2007; 66: 143–150. selective and non-selective NSAIDs: nested case-control 50. Perrot S, Javier R-M, Marty M, LeJeunne C, Laroche F. Is study. Lancet 2005; 365: 475–481. there any evidence to support the use of anti-depressants in 36. van Everdingen AA, Jacobs J, Siewertsz D, et al. Low-dose painful rheumatological conditions? Systematic review of prednisone therapy for patients with early active RA: pharmacological and clinical studies. Rheumatology 2008; clinical efficacy, disease-modifying properties, and side 47(8): 1117–1123. 119

Chapter 10 51. Cleland LG, James MJ. Fish oil and rheumatoid arthritis: 58. Brosseau L, Yonge KA, Welch V, et al. Transcutaneous antiinflammatory and collateral health benefits. J Rheumatol electrical nerve stimulation (TENS) for the treatment of 2000; 27: 2305–2307. rheumatoid arthritis in the hand. Cochrane Database of Systematic Reviews 2003, Issue 2. Art. No.: CD004377. DOI: 52. Little CV, Parsons T. Herbal therapy for treating rheuma- 10.1002/14651858.CD004377. toid arthritis. Cochrane Database of Systematic Reviews 2000, Issue 4. Art. No.: CD002948. DOI: 10.1002/14651858. 59. Melzack R, Wall PD. Pain mechanisms: a new theory. CD002948. Science 1965; 150: 971–979. 53. Riemsma RP, Kirwan JR, Taal E, Rasker HJJ. Patient edu- 60. Casimiro L, Barnsley L, Brosseau L, Milne S, Welch cation for adults with rheumatoid arthritis. Cochrane V, Tugwell P, Wells GA. Acupuncture and electroacupunc- Database of Systematic Reviews 2003, Issue 2. Art. No.: ture for the treatment of rheumatoid arthritis. Cochrane CD003688. DOI: 10.1002/14651858.CD003688. Database of Systematic Reviews 2005, Issue 4. Art. No.: CD003788. DOI: 10.1002/14651858.CD003788.pub2. 54. Hammond A, Freeman K The long-term outcomes from a randomized controlled trial of an educational–behavioural 62. Steultjens EM, Dekker J, Bouter LM, et al. Occupational joint protection programme for people with rheumatoid therapy for rheumatoid arthritis: a systematic review. arthritis. Clin Rehabil 2004; 18: 520–528. Arthritis Care Res 2002; 47: 672–685. 55. de Jong Z, Munneke M, Zwinderman AH, et al. Is a long- 63. Woodburn J, Barker S, Helliwell PS. A randomized con- term high-intensity exercise program effective and safe in trolled trial of foot orthoses in rheumatoid arthritis. patients with rheumatoid arthritis? Results of a randomized J Rheumatol 2002; 29: 1377–1383. controlled trial. Arthritis Rheum 2003; 48: 2415–2424. 64. Kirwan, JR, Currey H, Freeman M, Snow S, Young P. Overall 56. van den Ende FC, Breedveld FC, Le Cessie S, et al. Effect of long term impact of total hip and knee joint replacement intensive exercise on patients with active rheumatoid arthritis: surgery on patients with OA and Ra. Br J Rheumatol 1994; a randomised clinical trial. Ann Rheum Dis 2000; 59: 615–621. 33: 357–360. 57. Hall J, Skevington SM, Maddison PJ, Chapman K. A rand- 65. Luqmani R, Hennell S, Estrach C, et al. BSR and BHPR omized and controlled trial of hydrotherapy in rheumatoid guideline for the management of rheumatoid arthritis (the arthritis. Arthritis Care Res 1996; 9: 206–215. first 2 years). Rheumatology 2006; 45(9): 1167–1169. 120

CHAP TE R 11 Fibromyalgia Winfried Häuser1, Kati Thieme2, Frank Petzke3 and Claudia Sommer4 1Klinikum Saarbrücken, Department of Internal Medicine, Saarbrücken, Germany 2Center for Neurosensory Disorders, Thurston Arthritis Research Center, University of North Carolina, Chapel Hill, NC, USA 3Uniklinik Köln, Department of Anesthesiology, and Postoperative Intensive Care Medicine, Köln, Germany 4Universität Würzburg, Department of Neurology, Würzburg, Germany Background Definition and classification Many randomized controlled trials have been con- According to the criteria of the American College ducted on the pharmacologic and nonpharmacologic of Rheumatology (ACR), fibromyalgia syndrome is treatment of fibromyalgia syndrome (FMS). A sub- defined as chronic (Ͼ3 months) widespread pain stantial number of systematic reviews have also been (CWP) (including pain on both sides of the body, published, in which the evidence from these trials above and below the waist, and axial pain) and ten- has been summarized. Recently, three evidence-based derness on manual palpation in at least 11 out of 18 guidelines gave recommendations on the manage- defined tender points [4]. Chronic widespread pain ment of FMS. The recommendations of the American and tender points do not capture the essence of all Pain Society [1] were based on a systematic search of FMS-associated symptoms [5, 6]. Other key symp- the literature until April 2004. The European League toms of FMS are fatigue and nonrestorative sleep. Against Rheumatism EULAR [2] conducted a system- Most patients complain of additional somatic and atic search of the literature up to December 2005. The psychological symptoms [7, 8]. German Association of Pain Therapy, an umbrella organization of the medical and psychological socie- The term “fibromyalgia syndrome” is preferable to ties involved in the treatment of patients with chronic “fibromyalgia” because the definition of FMS according pain, co-ordinated an interdisciplinary guideline with to the ACR-criteria is based on a combination of symp- 10 scientific societies and two patient self-help organ- toms (CWP) and clinical findings (tenderness). Because izations. The systematic search included literature no consistent anatomic or specific pathophysiological published up to December 2006 and the recommen- mechanisms have yet been identified [9], FMS can be dations were built following a structured consensus classified as a functional somatic syndrome [10, 11]. process [3]. This chapter on evidence-based medicine “Fibromyalgia” is listed in the International Classification for FMS provides an overview of the current evidence of Diseases of the World Health Organization within on treatment of FMS and summarizes how this evi- Chapter M, “Diseases of the musculoskeletal system and dence has been translated into the US-American and connective tissue,” with the code M 79.7 [12]. German guideline recommendations. Fibromyalgia syndrome is not a distinct nosologic Evidence-Based Chronic Pain Management. Edited by entity like a myocardial infarction following an occlu- C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell sion within the coronary vasculature. Symptoms of Publishing. FMS are more like other continuous medical vari- ables such as blood pressure or coronary sclerosis for which clinically relevant limits have been defined to differentiate normal from borderline and pathological 121

Chapter 11 conditions. Within this context, FMS can be concep- Prevalence in inpatients tualized as the extreme of a continuum of distress In Israel, the prevalence of FMS in patients hospitalized caused by pain, fatigue, sleeping problems, and cog- on internal medicine wards was reported to be 15% nitive disturbances [13–15]. Chronic pain in different [26]. We found no data on the prevalence of FMS in sites of the body and fatigue are common symptoms pain clinics. The percentage of inpatients with chronic in the general population as well as in several somatic widespread pain in German pain clinics was reported diseases and mental disorders [16, 17]. CWP has been to be 25–30% [27] (Schiltenwolf, personal communi- the focus of independent research and results have cation, 2008). The percentage of inpatients with the been included in this review. Tender points can be primary diagnosis of FMS in a German interdisci- found in painful conditions other than FMS as well plinary secondary care pain clinic was 3.6% (Gockel, as in a small percentage of subjects without pain [18]. personal communication, 2008). The lack of data on FMS, according to the ACR criteria, defines a distinct FMS in German pain clinics can be explained by the clinical syndrome at the end of a continuum of pain skepticism of most pain therapists about the use of this sites, fatigue and tender points[18, 19]. diagnostic category. Most patients with FMS are coded as “somatoform pain disorder” (Schiltenwolf, personal Prevalence communication, 2008). The fact that the diagnosis of a somatoform pain disorder and inpatient multidiscipli- Prevalence in the general population nary treatment leads to higher remuneration than the A review of 10 studies from different Western countries diagnosis of FMS within the German diagnosis-related reported a prevalence of FMS according to the ACR system might also explain the coding preferences of criteria in the general adult population of between German pain therapists. 0.7% and 3.3%. The prevalence in women was between 1.0% and 4.9%, in men between 0.0% and 1.6%. The Course of fibromyalgia syndrome sex ratio of women to men was 2–21:1 [20]. Most patients were between 40 and 60 years old but FMS can A review of longitudinal studies on the natural course also be diagnosed in children and adolescents [21]. of FMS demonstrated that the symptoms of FMS per- sist in the long run. Some patients adapt to the symp- Prevalence in general practice toms and the associated restrictions and report a better Only a few data on the prevalence of FMS in general long-term satisfaction with their health status [11]. practice are available. Within the UK General Practice Research Database, an increase of the incidence of Risk factors FMS from less than 1 per 100,000 in 1990 to 35 per 100,000 in 2001 was found. The rising incidence rates A risk factor can be defined as a characteristic, con- most likely reflect trends in diagnostic labeling rather dition or behavior that increases the possibility of a than true changes in incidence [22]. In a Dutch study disease. The association between the condition and using a postal questionnaire, 83% of the GPs reported the disease should be demonstrated not only in clini- one or more FMS patients in their practice. The esti- cal populations but also in population-based studies. mated prevalence of FMS as recognized by GPs is low, Risk factors can be assumed to be etiologic factors, if with 157 patients per 100,000 [23]. The low preva- they contribute to the cause of a disease. The causal lence in primary care can be explained by the fact that role of a factor should be demonstrated by prospec- GPs are reluctant to use the tender point examination tive studies in the general population. Moreover, and the diagnostic category of “FMS” [24]. pathophysiologic studies should demonstrate an impact of the factor on a pathophysiologic mechanism of the Prevalence in rheumatologic practices disease. Therefore prospective population-based stud- In a cross-sectional study in a population of patients ies on CWP and FMS have the highest significance cared for by rheumatologic practices in public Spanish for the assessment of risk factors of FMS. The current hospitals, the prevalence of fibromyalgia was 12% knowledge on risk factors of FMS is summarized in (2.2% in men and 15.5% in women) [25]. Table 11.1. 122

Fibromyalgia Table 11.1 Risk factors for occurrence and chronicity of chronic widespread pain and fibromyalgia syndrome in adults Occurrence Chronicity Biologic risk factors Family aggregation Physical co-morbidity Sociodemographic factors Inflammatory rheumatoid disorder Psychologic factors Age Mental co-morbidity Female sex Negative life events Occupational factors Lower social class index Psychologic distress Depressed mood Somatization Family aggregation Mental disorders Occupational psychologic stressors Somatization Occupational mechanical burdens Risk factors for occurrence cases, respectively. This high variability may depend on the psychosocial characteristics of patients, since Biologic risk factors most of the studies were performed in tertiary care Genetics settings. Even referring to the lower percentages, There is a high aggregation of FMS in families of the occurrence of mental disorders is significantly FMS patients. The mode of inheritance is unknown higher in FMS subjects compared to the general but is most probably polygenic. There is evidence population [30]. Co-morbid post-traumatic stress of polymorphisms of genes in the serotoniner- disorders have been reported in 30–60% of FMS gic, dopaminergic, and catecholaminergic systems. patients [9]. Moreover, elevated frequencies of men- These polymorphisms are not specific for FMS and tal disorders have been described in relatives of FMS are associated with other functional somatic disor- patients [30, 31]. The FMS/mental disorder aggrega- ders and depression [28]. An analysis of participants tion suggests a common pathophysiology, and alter- of the Swedish Twin Registry examining proband- ations of neurotransmitter systems may constitute wise concordance rates and tetrachoric correla- the shared underlying factors [30]. tions suggested modest genetic influences for both women and men with CWP. Genetic and shared Depressed mood environmental influences explained approximately A prospective Norwegian population-based study half of the total variance, with no indication of sex found that depressed mood predicted the manifesta- differences in either the type or magnitude of these tion of FMS in patients with only local pain on the influences [29]. initial assessment [32]. Inflammatory rheumatoid disorder Functional somatic syndromes An association between FMS and inflammatory rheu- The prevalence of other functional somatic syndromes matologic disorders (rheumatoid arthritis, systemic such as chronic fatigue syndrome or irritable bowel lupus erythematosus, Sjögren’s syndrome) has been symptoms has been reported as 30–80%, depending reported in 1–50% of patients [9]. on the setting and the diagnostic methods used [33]. The frequent aggregation of functional somatic syn- Psychologic risk factors droms suggests a common pathophysiology [17]. Mental disorders Compared with controls, FMS patients show a signi- Somatization ficantly higher prevalence of depressive and anxi- Two British population-based studies found that soma- ety disorders, reported in 20–80% and 13–63.8% of tization predicted the manifestation of CWP [34, 35]. 123

Chapter 11 Somatization was a stronger predictor of the manifes- are more common in women than in men [33]. tation of CWP than work-related mechanical factors Women tend to report more physical and psychologic [34]. A population-based British study demonstrated symptoms and exhibit a more intensive healthcare- that subjects were at substantially increased odds of seeking behavior than men [41]. developing CWP if they displayed features of somatiza- tion, healthcare-seeking behavior and poor sleep [36]. Social class Within the framework of a German population- Childhood adversities based cross-sectional study, subjects with FMS had a Numerous retrospective case–control studies have 3.6-fold risk of having a lower social class level, when shown that traumatic experiences (maltreatment, compared to people without chronic pain [43]. sexual abuse, emotional neglect) during childhood are more frequently reported by FMS patients in clinical Physical and occupational factors populations than by medically ill or healthy controls Physical injury [31]. However, these studies are biased due to recall set- The German FMS guideline group found two pro- ting and response bias [37]. A British population-based spective studies with conflicting results regarding the case–control study found an association between CWP occurrence of FMS after a traffic injury with initial and hospital treatment in childhood, but not with sex- neck pain [9]. ual abuse or chronically ill parents [38]. A systematic review of prospective studies on sexal abuse in child- Occupational factors hood and chronic pain syndromes in adulthood found A British population-based prospective survey found no clear evidence for a causal relationship to FMS [37]. that pushing/pulling heavy weights, repetitive wrist movements, kneeling and local pain complaints at Negative life events in adulthood baseline were associated with new-onset chronic Although retrospective studies in clinical samples sug- widespread pain. However, the strongest predictor gest that the onset of FMS is frequently associated with was a high score on the Illness Behavior Scale [44]. various types of negative life events [31], prospective Harkness et al. [45] demonstrated in a prospective population-based studies failed to demonstrate an British study that those who reported low job satis- increased risk of FMS-like pain complaints following faction, low social support, and monotonous work the World Trade Center terrorist attack [39]. had an increased risk of new-onset widespread pain. Daily hassles Risk factors for chronicity Workplace bullying, high workload and low decision Somatization latitude were associated with an increased incidence In a British prospective, population-based study, per- of diagnosed FMS within a Finnish prospective sistent chronic widespread pain was strongly associ- population-based study [40]. ated with baseline test scores for high psychologic distress and fatigue. In addition, these subjects were Sociodemographic factors more likely to display a pattern of illness behavior Sex characterized by frequent visits to medical practition- The mechanisms of gender differences in FMS are not ers for symptoms disrupting daily activities [46]. fully understood. An interaction between biologic, psy- chological and sociocultural factors has been postulated Co-morbidities [41]. Between puberty and menopause, adult women An American prospective study with FMS patients usually show lower responses of the hypothalamic- from both tertiary care and community settings found pituitary-adrenal axis (HPA) and autonomic responses that poorer health status and more co-morbidity than men of the same age [42]. Female gender is a risk at baseline were predictors of poorer health status 6 factor for psychologic distress and some mental disor- months later. After controlling for these co-variates, ders (affective and anxiety disorder, PTSD) which are psychologic distress still contributed significantly to associated with FMS. Functional somatic syndromes the model [47]. 124

Fibromyalgia Negative life events aspects of the altered function were related to psycho- Norwegian FMS patients were followed for 4.5 years. social distress [52]. Another prospective population- Receiving a permanent disability pension or having based cohort study of the same study group revealed experienced an excess of major negative life events that dysfunction of the HPA axis (high levels of corti- predicted a negative outcome [48]. sol post dexamethasone, low levels in morning saliva and high levels in evening saliva) within a group of Protective factors subjects free of CWP but at future risk based on their psychosocial profile helped to distinguish those who An Australian prospective study with FMS patients would and would not develop new-onset CWP [35]. from community rheumatologic practices found that Thus, dysfunction of the HPA axis can be triggered by regular physical exercise, rather than medication or psychosocial distress. On the other hand, coping with specific physical therapies, correlated highly with low psychosocial distress can be impeded by a genetically FMS activity scores. Analysis of mood and coping based hyporeactivity of the HPA axis. strategies at the 2-year review showed low correla- tions with current FMS activity [49]. Norwegian FMS The biopsychosocial model of FMS postulates that patients were followed for 4.5 years. An adequate there is a heterogeneity in the genetic and psychologic physical activity level and increasing age predicted predispositions as well as in the vegetative, endocrine a positive outcome [48]. Another Norwegian study and central nervous system reactions. Different etio- reassessed women with FMS who had participated in logic factors and pathophysiologic mechanisms lead exercise and patient education programs after 6 and to a common pathway of symptomatology currently 8 years. Adjusting to the new situation and distrac- classified as FMS [9]. tion from symptoms, as well as frequent participation in physical activities, were associated with a benign Diagnosis long-term outcome [50]. History taking Pathophysiology Patients presenting with chronic mono- or oli- golocular pain should be asked for potential other Several potential pathophysiologic mechanisms for pain locations. A pain diagram helps to identify FMS have been described, but their causal relation- patients with CWP. If CWP is diagnosed, other key ship are unclear because of the cross-sectional nature symptoms of FMS (fatigue/nonrestorative sleep and of all these studies. Potential mechanisms include increased morning stiffness/swelling of the fingers or central nervous system (CNS) pain-processing abnor- the hands) should be actively explored. The German malities, hyporeactivity of the HPA axis, increased sys- interdiscplinary guideline recommends the establish- temic proinflammatory and reduced anti-inflammatory ment of a symptom-based diagnosis of FMS if symp- cytokine profiles, and disturbances in the dopaminer- toms in all the following three symptom domains are gic and serotonergic systems [9]. Potential linked or reported by the patient: (1) CWP, (2) fatigue or non- underlying genetic mechanisms have been described restorative sleep or sleep disturbances, and (3) sensa- above. Cross-sectional studies suggested that FMS tions of stiffness or swelling in the hands or the feet may be related to hypofunctional stress systems, par- or the face [11]. ticularly in the autonomic nervous system (ANS) and the HPA axis. Studies have demonstrated that patients If the key symptoms of FMS are reported, patients with FMS exhibit lowered sympathoadrenal reactivity should be screened for symptoms of other functional to stress [51]. somatic syndroms and mental disorders as well as current psychosocial stressors. Moreover, restrictions A prospective British population-based study of daily activities and subjective illness attributions assessed relationships between psychosocial and should be asked for. Finally all types of medication biologic variables relevant for CWP. Abnormalities used by the patient should be assessed since arthral- of HPA function were more marked in people with gia, myalgia and fatigue can also be side effects of established CWP than in those at risk of CWP. Some medication. Misuse of medication should be actively explored [11]. 125

Chapter 11 Physical examination a fatigue score 6 on a 11-point visual analog scale There is an ongoing debate on the utility of the ten- (VAS). The use of standardized somatic symptom der point examination in the clinic. The ACR crite- scales and questionnaires such as the Patient Health ria were primarily developed for the classification of Questionnaire PHQ [59] can be considered in primary FMS [4] to identify a group of patients with similar care. Restrictions of daily life associated with FMS clinical features for future systematic studies. Their symptoms can be assessed by the Brief Pain Inventory practicability and validity for clinical diagnosis have [60] or the Fibromyalgia Impact Questionnaire [61]. never been tested outside a rheumatologic setting. Nevertheless, a history of CWP Ͼ 3 months and the Blood tests and diagnostic imaging finding of tenderness in at least 11 of 18 tender points The following routine blood tests are recommended with manual palpation using approximately 4 kg of by the German interdisciplinary FMS guideline for pressure have become the gold standard of FMS diag- patients with CWP (potential differential diagnosis nosis in clinical studies. Their use was recently recom- are indicated in parentheses): mended by an editorial in the Journal of Rheumatology • blood sedimentation rate, C-reactive protein, red [18] but the diagnostic use of tender point examina- tion in clinical practice has been criticized by various and white cell blood count (polymyalgia rheumat- authors, for the following reasons [53, 54]. ica, rheumatoid arthritis) • Fibromyalgia syndrome is the only functional • creatinine kinase (muscle disease) • calcium (hypercalcemia) somatic syndrome (FSS), which is defined and diag- • thyroid-stimulating hormone (hypothyreosis) nosed by symptoms and a clinical sign. FSS in other • depending on history and examination, further blood medical disciplines such as irritable bowel syn- tests can be necessary if other differential diagnoses drome, chronic tension headache, and nonspecific are suspected. low back pain are “symptoms-only diagnoses.” Without cinical signs, routine testing for antibodies • The tender point count has been shown to be influ- associated with inflammatory rheumatologic diseases enced by the interaction between patient and exam- is not recommended. A Canadian study in outpa- iner [55]. Despite efforts to standardize the procedure, tient secondary care found no predictive value of such as the manual tender point survey [56], it has not the assessment of antibodies associated with inflam- been shown to be reproducible across different clinical matory rheumatologic diseases in patients with a settings. It is a poor marker of change in clinical stud- history of CWP and fatigue in the absence of other ies and controversial in legal situations. features such as joint swelling, typical rashes or organ • There is general agreement that tender points are a involvement [62]. If no other diseases which require marker of psychophysiologic distress [18] and not imaging studies for diagnosis are suspected on clini- an objective measure of tenderness [57]. Therefore cal grounds, X-rays or other diagnostic imaging stud- the limited time of medical consultation should be ies are not recommended. A Norwegian longitudinal used to explore psychosocial distress rather than to study in primary care demonstrated a low diagnostic press on tender points. Most nonrheumatologists value of imaging studies in patients with CWP [63]. are reluctant to use the tender point examination because of the time involved [24]. Referral to specialists • A complete physical examination including ortho- In a case of suspected medical, neurologic or psychi- pedic and neurologic examination is recommended atric disease or the presence of dysfunctional coping to reveal signs of internal or neurologic disorders styles, referral to a specialist is recommended [24]. mimicking the key symptoms of FMS [11]. Treatment Questionnaires Wolfe [58] developed survey criteria for FMS. Here, Problems of evidence-based medicine FMS is simply diagnosed by a questionnaire, the The majority of the numerous therapies used in regional pain scale. FMS is diagnosed if the patient FMS have been systematically reviewed recently, indicates pain in at least 11 of 19 pain sites and reports 126

Fibromyalgia including aerobic exercise [64, 65], complementary Collaboration and alternative therapy [66], multicomponent ther- Various healthcare providers may be involved in the apy [67], patient education [24], physical therapy treatment of FMS. Although there may be some vari- and physiotherapy [65], pharmacologic therapy [68], ations between countries, exercise therapists, general and psychotherapy [69]. The reviews demonstrate practitioners, neurologists, orthopedic surgeons, pain that there is no therapy that works in every patient. therapists, psychiatrists, physiotherapists, psycholo- If defined in the study results, responders and non- gists, rheumatologists and others may all be involved. responders can be found with similar rates in phar- The primary care physician has a central role in the macologci and nonpharmacologic therapies. Overall co-ordination of the management. It is important evidence is limited to middle-aged women with FMS that information and treatment are consistent across without co-morbidity. The German FMS guideline professions and specialties, and that healthcare pro- group found only one RCT in children and adoles- viders closely collaborate with each other, with cents, which demonstrated a superiority of cognitive patients and their self-help organizations [24]. behavioral therapy over self-monitoring at the end of therapy [70]. Most studies excluded patients with rel- Systematic reviews of the effectiveness evant physical and mental co-morbidities [71]. Men of treatments for FMS were either excluded or no separate analysis for men was performed. Interventions strongly supported by evidence Evidence on the most frequently used treatments of The question of how to manage a life-long disor- FMS is summarized in Box 11.1. Of the pharmaco- der remains unanswered by all controlled studies. logic treatments, only duloxetine, milnacipran and There is evidence for the short-term effectiveness of pregabalin had been approved for use in FMS in the some pharmacologic and nonpharmacologic thera- USA, but not by the EMEA for use in FMS in Europe. pies. Most studies lasted between 4 and 12 weeks and assessed the outcomes at the end of the treatment Antidepressants period. There currently is no evidence of a long-term The German FMS guideline group systematically effectiveness (Ͼ6 months) of pharmacologic and reviewed 26 RCT on antidepressants. Amitriptyline, physical therapies. There is only limited evidence for studied in 13 RCT, was efficient in reducing pain with the long-term effectiveness of aerobic exercise [65], a moderate magnitude of benefit (pain reduction by a multicomponent therapy [67] and cognitive behav- mean of 26%, improvement in quality of life by 30%). ioral therapy (CBT) [69], based on follow-up periods Selective serotonin reuptake inhibitors (SSRI) were ranging from 6 to 24 months. There is limited evi- studied in 12 RCT, which also gave positive results, dence of the cost-effectiveness of operant pain ther- except for the two studies on citalopram and one apy and spa therapy [72, 73]. on fluoxetine. Three RCT on the dual serotonin and noradrenaline reuptake inhibitors (SNRI) duloxetine General principles of management of FMS and milnacipran reported positive results at the end of therapy. Results concerning the effectiveness of the Self-management monoamine oxidase inhibitor (MAOI) moclobemide Currently, FMS cannot be cured by any therapy and (300–600 mg/day) were conflicting [71]. Eighteen overall treatment effects are modest at best. According RCT (median duration 8 weeks, range 4–38 weeks) to expert opinion, the aims of therapy are the pres- involving 1427 subjects were suitable for meta-analy- ervation or improvement of daily functioning and sis. There was strong evidence for an association of coping with symptoms and disabilities. Coping with antidepressants with reduction in pain (standard symptoms includes both the acceptance of symptoms mean difference (SMD) –0.43, 95% confidence inter- and of some limitations (e.g. hard physical work) as val (CI) –0.55 to –0.30), fatigue (SMD –0.13, 95% CI well as continous self-management (physical activ- –0.26 to –0.01), depressed mood (SMD –0.26, 95% ity, stress management) to reduce the impact of the CI –0.3 to –0.12) and sleep disturbances (SMD –0.32, symptoms. Therefore the patient has the major task 95% CI –0.46 to –0.18). There was strong evidence in FMS therapy[1, 24]. 127

Chapter 11 Box 11.1 Evidence of treatments for adult fibromyalgia syndrome Interventions strongly supported by evidence (systematic reviews) Aerobic exercise Amitriptyline Balneo- and spa therapy Cognitive behavioral therapy Cyclobenzaprine Duloxetine Fluoxetine Pregabalin Multicomponent treatment programs Interventions supported by evidence (at least two RCT with consistent results) Homeopathy Hypnotherapy/guided imagery Mindfulness meditation Patient-centered communication Tramadol Tropisetron Vegetarian diet Whole-body heat therapy Written emotional disclosure Commonly used interventions currently unproven (only one RCT with low quality, RCTs with conflicting results or no RCT available) Acetaminophen Acupuncture Body awareness therapy Lidocaine infusions Lymph drainage Massage Metamizol Muscle relaxants other than cyclobenzaprine Osteopathy Qi-gong Physiotherapy Psychodynamic therapy Tender point injections Tramadol/acetaminophen Interventions strongly refuted by evidence (systematic reviews) Anxiolytics Biofeedback as single intervention Citalopram Corticosteroids Neuroleptics Nonsteroidal agents Patient education as single intervention Relaxation therapy as single intervention 128

Fibromyalgia for an association of antidepressants with improved (SMD 0.65, 95% CI Ϫ0.09 to 1.39). Strength and flex- health-related quality of life (HRQOL) (SMD –0.31, ibility remain underevaluated; however, strength train- 95% CI –0.42 to –0.20). Effect sizes for pain reduc- ing may have a positive effect on FMS symptoms. tion were large for tricyclic antidepressants (TCA) (SMD –1.64, 95% CI –2.57 to –0.71), medium for Multicomponent treatment (MT) programs MAOI (SMD –0.54, 95% CI –1.02 to –0.07) and small There is no internationally accepted definition of mul- for SSRIs (SMD –0.39, 95% CI –0.77 to –0.01) and ticomponent therapy. The existing systematic reviews SNRIs (SMD –0.36, 95% CI –0.46 to –0.25)[74]. on MT agree that MT should include at least one edu- cational or other psychologic therapy and at least one Balneo- and spa therapy exercise therapy [67, 78]. The German FMS guideline McVeigh et al. [75] systematically reviewed 10 RCT group meta-analyzed 9/14 RCT, with 1119 subjects with with balneo- or spa therapy or pooled-based exer- a median treatment time of 24 hours included in the cises. Improvements were demonstrated in pain, meta-analysis. There is strong evidence that MT reduces health status, anxiety, fatigue, in addition to func- pain (SMD Ϫ0.37, 95% CI Ϫ0.62 to Ϫ0.13), fatigue tion and aerobic capacity. However, for the most part, (WMD Ϫ0.85, 95% CI Ϫ1.50 to Ϫ0.20), depressive improvements tended to be shor-lived. symptoms (SMD Ϫ0.67, 95% CI Ϫ1.08 to Ϫ0.26) and limitations of HRQOL (SMD Ϫ0.59, 95% CI Cyclobenzaprine Ϫ0.90 to –0.27) and improves self-efficacy pain (SMD Tofferi et al. [76] included five RCT with cycloben- 0.54, 95% CI 0.26Ϫ0.82) and physical fitness (SMD zaprine, a muscle relaxant with additional profile of 0.30, 95% CI 0.02Ϫ0.57) post treatment. There is no a TCA, in their meta-analysis. The odds ratio (OR) evidence of the efficacy of MT on pain, fatigue, sleep for a global improvement was 3.0 (95% CI 1.6–5.6). disturbances, depressive symptoms, HRQOL and self- Cyclobenzaprine is not licensed in most European efficacy pain in the long term. There is strong evidence countries. that the positive effects on physical fitness (SMD 0.30, 95% CI 0.09–0.51) can be maintained in the long term Cognitive behavioral therapy (CBT) (median follow-up 7 months) [79]. The German FMS guideline group reviewed 14 RCT on CBT. Most studies lasted between 6 and 15 weeks, Pregabalin and most therapies comprised 6–30 hours of interven- In a meta-analysis of four RCTs with PGB and one tion. Twelve of the 14 studies found a superiority of RCT with gabapentin with a parallel design there was a CBT in most outcomes at the end of the therapy. Nine strong evidence for the reduction of pain (SMD Ϫ0.28, of the 14 studies performed follow-ups and 5/9 stud- 95% CI Ϫ0.36 to Ϫ0.20; P < 0.001), improved sleep ies reported a persistant reduction of FMS symptoms (SMD Ϫ0.39, 95% CI Ϫ0.48 to Ϫ0.39; P < 0.001), after 6–24 months [69]. Twenty-two out of 33 relevant and improved health-related quality of life (HRQOL) studies investigating 1209 subjects could be meta-ana- (SMD –0.30, 95% CI Ϫ0.46 to Ϫ0.15; P < 0.001), but lyzed. CBT had large effects on pain, self-efficacy and not for depressed mood (SMD Ϫ0.12, 95% CI –0.30 disability only at follow-up. Hypnotherapy has a large to 0.06; P = 0.18). There was strong evidence for a not effect on the improvement of self-efficacy pain and a substantial reduction of fatigue (SMD = Ϫ0.16, 95% medium effect on the improvement of pain post treat- CI Ϫ0.23 to Ϫ0.09, P < 0.001) and of anxiety (SMD = ment and at follow-up [77]. Ϫ0.18, 95% CI Ϫ0.27 to Ϫ0.10). The external validity was limited because patients with severe somatic and Exercise mental disorders were excluded [80]. Busch and co-workers [64] systematically reviewed 34 studies. Meta-analysis of six studies provided moder- Interventions supported by evidence ate-quality evidence that aerobic-only exercise training Some further interventions supported by evidence at American College of Sports Medicine-recommended (at least two RCT with consistent results or majority intensity levels has positive effects on global well-being of RCT with consistent results) are listed in Box 11.1. (SMD 0.49, 95% CI 0.23Ϫ0.75) and physical function In a systematic search of the literature up to December (SMD 0.66, 95% CI 0.41Ϫ0.92) and possibly on pain 2006, the German FMS guideline group found six RCT 129

Chapter 11 on balneo- and spa therapy, two studies on home- Treatment should begin with confirming the diag- opathy, five studies on hypnotherapy/guided imagery, nosis and patient education. Self-management and three studies on patient-centered communication, two patient-centered communication are regarded as RCT on vegetarian diet, three RCT on whole-body heat key principles of FMS therapy. Both guidelines rec- therapy, two on written emotional disclosure, three on ommended first, second and third lines of therapy tropisetron, and three on tramadol (one in combina- depending on the course of symptoms, the existence of tion with acetaminophen), fulfilling the criteria defined relevant restrictions of daily activities and the response above [24, 65, 66, 68, 69]. Since December 2006 further to different treatment modalities. The APS recom- RCT on pharmacologic, psychotherapeutic and physi- mended individualized exercise and CBT as first-line cal treatment have been published which support the therapy. The German guideline recommended aerobic efficacy of the interventions mentioned above. exercise, CBT, amitriptyline and treatment of physical and mental co-morbidities as single interventions or Regarding further treatment options, we wish to combined as first-line therapies. highlight two RCT with consistent results on mind- fulness meditation [81, 82]. In cases of persisting symptoms after first-line treatment, the APS recommended tender point Interventions not supported by evidence injections, local manual therapies and acupuncture It is important to note that evidence is lacking for for focal pain, tramadol, SNRI, anticonvulsants for the effectiveness of frequently used pharmacologic generalized pain, and psychiatric treatment in cases treatments of FMS [7, 11], such as acetaminophen, of major mood disorder as second-line therapy. The metamizol, opioids other than tramadol or muscle German guideline recommended multicomponent relaxants other than cyclobenzaprine, as well as for therapy in cases of persisting restrictions of daily frequently used nonpharmacologic therapies such as activities as second-line therapy. massage, osteopathy or lymph drainage. Multidisciplinary pain management, psychopharma- The German FMS guideline group found inconsist- cology, opioids, experimental therapies and combina- ent results in eight RCT on patient education as sin- tions is the third-line therapy recommended by the APS gle intervention and seven studies with massage with in cases of persisting symptoms. In cases lacking adap- conflicting results [24, 65]. Furthermore, inconsistent tation to symptoms or persistent restrictions of daily results of five controlled trials each on biofeedback functioning, the German guideline recommended either and relaxation therapy (including autogenic training no therapy or self-management (aerobic exercise, stress and progressive muscle relaxation) were found [69]. management, pool-based exercise), or booster multi- component therapy, or psychotherapy (hypnotherapy, Interventions strongly refuted by evidence written emotional disclosure), or pharmacologic ther- There is evidence for the noneffectiveness of the apy (duloxetine or fluoxetine or paroxetin or prega- most frequently used drug class in FMS, nonsteroidal balin, or tramadol with or without acetaminophen), agents. Furthermore, a systematic review found RCT or complementary/alternative therapies (homeopathy, with negative results for anxiolytics, corticosteroids vegetarian diet) as third-line therapy. The choice of and neuroleptics [68]. treatment options should be based on informed patient consent, the patient’s preferences and co-morbidities, Mayhew & Ernst [83] systematically reviewed five and the treatment options locally available [24]. RCT on acupuncture (traditional Chinese acupunc- ture and electroacupuncture) and concluded because Guidelines and implementation of conflicting results that acupuncture is not effective in FMS therapy. The development and dissemination of guidelines do not automatically mean that patients and healthcare A stepwise treatment approach to FMS providers will read, understand and use these guidelines. A graded treatment approach to functional somatic The following strategies were designed to dissiminate syndromes is recommended [10]. The American Pain and implement the German guideline. Society (APS) [1] and the German Interdisciplinary Guideline Group [24] both recommend a stepwise The results of the systematic reviews and recom- treatment approach. mendations of the guideline are published in scientific 130

Fibromyalgia journals and presented at the annual scientific meet- active treatment approaches (aerobic exercise, CBT) ings of the societies involved. The scientific version of should be preferred. the guideline (complete and short version) is avail- able on the homepages of the societies involved. The Evidence-based guidelines for the management German Society of General and Family Medicine will of FMS should be developed by an interdisciplinary develop and validate a pocket version of the guideline approach and include patients in all stages of devel- for primary care physicians. A patient version of the opment and implementation. It is important to note guideline is available on the homepages of the medical that many of the recommendations are not based on scientific societies as well as of the self-help organiza- clinical studies, but rather the agreement of interdis- tions involved and published in the respective journals ciplinary experts and patients and their joint vision of the self-help organizations involved. The recom- of the management of FMS in a specific health system mendations were presented and discussed at regional and societal context. patient–doctor meetings. Thus the strategies aim at modifying not only the practice of healthcare provid- References ers but also the knowledge and demands of patients. Patients are accessed by self-help organizations, by 1. Burckhardt CS, Goldenberg D, Crofford L, et al. EULAR regional meetings, the internet and patients’ journals. evidence based recommendations for the management of The evaluation of the implementation strategies out- fibromyalgia syndrome. Ann Rheum Dis 2008; 67: 536–541. lined will be a major challenge for the future. 2. Carville SF, Arendt-Nielsen S, Bliddal H, et al. EULAR evi- Producing evidence of dence based recommendations for the management of effectiveness in the future fibromyalgia syndrome. Ann Rheum Dis 2008; 67: 536–541. Promising strategies include the identification of pre- 3. Bernardy K, Klose P, Üçeyler N, Kopp I, Häuser W. dictors of a positive treatment outcome [84] and of Methodological fundamentals of the development of the relevant subgroups that may benefit from more spe- guideline. Report of the methodology. Schmerz 2008; 22(3): cific intervention tailored to coping strategies and co- 244–254. (German) morbidity of mental disorders [69, 85]. Furthermore, there is a need to develop strategies to motivate 4. Wolfe F, Smythe HA, Yunus MB, et al. The American patients for continuous self-management, such as College of Rheumatology criteria for the classification of maintaining the benefits of aerobic exercise, CBT, and fibromyalgia: report of the Multicenter Criteria Committee. MT after the end of treatment. Arthritis Rheum 1990; 33: 160–172. Authors’ recommendations 5. Turk DC, Flor H. Primary fibromyalgia is greater than tender points: toward a multiaxial taxonomy. J Rheumatol Researchers should take into consideration that causal 1989; 19: 80–86. relationships between FMS, stress/depression, central augmentation of pain processing, and hyporeactiv- 6. Mease PJ, Clauw DJ, Arnold LM, et al. Fibromyalgia syn- ity of the HPA axis might not be linear but recursive. drome. J Rheumatol 2005; 32: 2270–2277. Future research should leave simple linear etiopatho- genetic models and prospectively study the interac- 7. Bennett RM, Jones J, Turk DC, Russell IJ, Matallana L. An tions between biologic and psychosocial variables and internet survey of 2569 people with fibromyalgia. BMC their underlying genetic factors. Musculoskelet Disord 2007; 8: 27. The main therapeutic challenges are the development 8. Häuser W, Zimmer C, Felde E, Köllner V. What are the key of cost-effective treatments tailored to defined FMS sub- symptoms of fibromyalgia? Results of a survey of the German groups and the long-term efficacy of such approaches. Fibromyalgia Association. Schmerz 2008; 22: 176–183 (German). The aim of therapy is to enhance the self-manage- 9. Sommer C, Häuser W, Gerhold K, et al. Etiology and patho- ment of the patient to better cope with the symptom physiology of fibromyalgia syndrome and chronic wide- load, and to improve daily functioning. Therefore spread pain. Schmerz 2008; 22: 267–282 (German). 10. Henningsen P, Zipfel S, Herzog W. Management of func- tional somatic syndromes. Lancet 2007; 369: 946–955. 11. Eich W, Häuser W, Friedel E, et al. Definition, classification and diagnosis of fibromyalgia syndrome. Schmerz 2008; 22: 256–266 (German). 12. World Health Organization. International Classification of Diseases Version 2007. www.who.int/classifications/apps/ icd/icd10online/?gm70.htm+m797. 13. Croft P, Burt J, Schollum J, Thomas E, Macfarlane G, Silman A.More pain, more tender points: is fibromyalgia just one end of a continuous spectrum? Ann Rheum Dis 1996; 55: 482–485. 14. Ruiz Moral R, Muñoz Alamo M, Pérula de Torres L, Aguayo Galeote M. Biopsychosocial features of patients with 131

Chapter 11 widespread chronic musculoskeletal pain in family medi- 34. Macfarlane GJ, Hunt IM, Silman AJ. Role of mechanical and cine clinics. Fam Pract 1997; 14: 242–248. psychosocial factors in the onset of forearm pain: prospec- 15. Wolfe F. The relation between tender points and fibromyalgia tive population based study. BMJ 2000; 321: 676–679. symptom variables: evidence that fibromyalgia is not a dis- crete disorder in the clinic. Ann Rheum Dis 1997; 56: 268–271. 35. McBeth J, Silman AJ, Gupta A, et al. Moderation of psycho- 16. Brähler E, Schumacher J, Brähler C. First all-Germany social risk factors through dysfunction of the hypothalamic- standardization of the brief form of the Gissen Complaints pituitary-adrenal stress axis in the onset of chronic widespread Questionnaire GBB-24. Psychother Psychosom Med Psychol musculoskeletal pain: findings of a population-based prospec- 2000; 50: 14–21 (German). tive cohort study. Arthritis Rheum 2007; 56: 360–371. 17. Henningsen P, Derra C, Türp JC, Häuser W. Functional somatic pain syndromes: summary of hypotheses of their 36. Gupta A, Silman AJ, Ray D, et al. The role of psychoso- overlap and etiology. Schmerz 2004; 18: 136–140 (German). cial factors in predicting the onset of chronic widespread 18. Harth M, Nielson WR. The fibromyalgia tender points: pain: results from a prospective population-based study. use them or lose them? A brief review of the controversy. J Rheumatology (Oxford) 2007; 46: 666–671. Rheumatol 2007; 34: 914–922. 19. Russel IJ. Is fibromyalgia a distinct clinical entity? The 37. Raphael KG, Chandler HK, Ciccone DS. Is childhood abuse clinical investigator’s evidence. Baillière’s Best Pract Res Clin a risk factor for chronic pain in adulthood? Curr Pain Rheumatol 1999; 13: 445–454. Headache Rep 2004; 8: 99–110. 20. Gran JT. The epidemiology of chronic generalized musculoskel- etal pain. Best Pract Res Clin Rheumatol 2003; 17: 547–561. 38. McBeth J, Morris S, Benjamin S, et al. Associations between 21. Michels H, Gerhold K, Häfner R, et al. Juvenile fibromyalgia adverse events in childhood and chronic widespread pain syndrom. Schmerz 2008; 22: 339–345 (German). in adulthood: are they explained by differential recall? J 22. Gallagher AM, Thomas JM, Hamilton WT, White PD. Incidence Rheumatol 2001; 28: 2305–2309. of fatigue symptoms and diagnoses presenting in UK primary care from 1990 to 2001. J R Soc Med 2004; 97: 571–575. 39. Raphael KG, Natelson BH, Janal MN, et al. A community- 23. Bazelmans E, Vercoulen JH, Swanink CM, et al. Chronic based survey of fibromyalgia-like pain complaints follow- fatigue syndrome and primary fibromyalgia syndrome as ing the World Trade Center terrorist attacks. Pain 2002; recognized by GPs. Fam Pract 1999; 16: 602–604. 100: 131–139. 24. Häuser W, Eich W, Herrmann M, Nutzinger DO, Schiltenwolf M, Henningsen P. Fibromyalgia syndrome: classification, diag- 40. Kivimäki M, Leino-Arjas P, Virtanen M, et al. Work stress nosis, and treatment. Dtsch Arztebl Int 2009; 106: 383–391. and incidence of newly diagnosed fibromyalgia: prospective 25. Gamero RF, Gabriel SR, Carbonell AJ, Tornero MJ, Sánchez cohort study. J Psychosom Res 2004; 57: 417–422. MI. Pain in Spanish rheumatology outpatient offices: EPIDOR epidemiological study. Rev Clin Esp 2005; 20: 41. Yunus MB. The role of gender in fibromyalgia syndrome. 157–163 (Spanish). Curr Rheumatol Rep 2001; 3: 128–134. 26. Buskila D, Neumann L, Odes LR, Schleifer E, Depsames R, Abu-Shakra M. The prevalence of musculoskeletal pain and 42. Kajantie E, Phillips DI. The effects of sex and hormonal fibromyalgia in patients hospitalized on internal medicine status on the physiological response to acute psychosocial wards. Semin Arthritis Rheum 2001; 30: 411–417. stress. Psychoneuroendocrinology 2006; 31: 151–178. 27. Gralow I, Hürter A, Schwerdt C, Hannich HJ, Meyer B, Voss S. Concept of a day hospital. First clinical experience. 43. Schochat T, Beckmann C. Sociodemographic characteris- Schmerz 1996; 41: 242–249 (German). tics, risk factors and reproductive history in subjects with 28. Buskila D, Sarzi-Puttini P, Ablin JN. The genetics of fibro- fibromyalgia – results of a population-based case-control myalgia syndrome. Pharmacogenomics 2007; 8: 67–74. study. Z Rheumatol 2003; 62: 46–59 (German). 29. Kato K, Sullivan P, Evengård B, Pedersen N. Importance of genetic influences on chronic widespread pain. Arthritis 44. McBeth J, Harkness EF, Silman AJ, et al. The role of Rheum 2006; 54: 1682–1686. workplace low-level mechanical trauma, posture and 30. Fietta P, Fietta P, Manganelli P. Fibromyalgia and psychiatric environment in the onset of chronic widespread pain. disorders. Acta Biomed 2007; 78: 88–95. Rheumatology (Oxford) 2003; 42: 1486–1494. 31. van Houdenhove B, Luyten P. Stress, depression and fibro- myalgia. Acta Neurol Belg 2006; 106: 149–156. 45. Harkness EF, Macfarlane GJ, Nahit E, et al. Mechanical 32. Forseth KO, Husby G, Gran JT, et al. Prognostic fac- injury and psychosocial factors in the work place predict tors for the development of fibromyalgia in women with the onset of widespread body pain: a two-year prospective self-reported musculoskeletal pain. A prospective study. J study among cohorts of newly employed workers. Arthritis Rheumatol 1999; 26: 2458–2467. Rheum 2004; 50: 1655–1664. 33. Henningsen P, Zimmermann T, Sattel H. Medically unex- plained physical symptoms, anxiety, and depression: a 46. McBeth J, Macfarlane GJ, Hunt IM, et al. Risk factors for meta-analytic review. Psychosom Med 2003; 65: 528–533. persistent chronic widespread pain: a community-based study. Rheumatology (Oxford) 2001; 40: 95–101. 47. Dobkin PL, de Civita M, Abrahamowicz M, Baron M, Bernatsky S. Predictors of health status in women with fibromyalgia: a prospective study. Int J Behav Med 2006; 13: 101–108. 48. Wigers H. Fibromyalgia outcome: the predictive values of symptom duration, physical activity, disability pension, and critical life events – a 4.5 year prospective study. J Psychosom Res 1994; 41: 235–243. 49. Granges G, Zilko P, Littlejohn GO. Fibromyalgia syndrome: assessment of the severity of the condition 2 years after diagnosis. J Rheumatol 1994; 21: 523–529. 132

Fibromyalgia 50. Mengshoel AM, Haugen M. Health status in fibromyalgia– 68. Sommer C, Häuser W, Berliner M, et al. Pharmacological a followup study. J Rheumatol 2001; 28: 2085–2089. therapy of fibromyalgia syndrome. Schmerz 2008; 22: 313–323 (German). 51. Okifuji A, Turk DC. Stress and psychophysiological dys- regulation in patients with fibromyalgia syndrome. Appl 69. Thieme K, Häuser W, Batra A, et al. Psychotherapy in Psychophysiol Biofeedback 2002; 27: 129–141. patients with fibromyalgia syndrome. Schmerz 2008; 22: 295–302 (German). 52. McBeth J, Chiu YH, Silman AJ, et al. Hypothalamic-pitui- tary-adrenal stress axis function and the relationship with 70. Kashikar-Zuck S, Swain NF, Jones BA, Graham TB. Efficacy chronic widespread pain and its antecedents. Arthritis Res of cognitive-behavioral intervention for juvenile primary Ther 2005; 7: R992–R1000. fibromyalgia syndrome. J Rheumatol 2005; 32: 1594–1602. 53. Wolfe F. Stop using the American College of Rheumatology 71. Üçeyler N, Häuser W, Sommer C. A systematic review on criteria in the clinic. J Rheumatol 2003; 30: 1671–1672. the effectiveness of treatment with antidepressants in fibro- myalgia syndrome. Arthritis Rheum 2008; 59: 1279–1298. 54. Biewer W, Conrad I, Häuser W. Fibromyalgia. Schmerz 2004; 18: 118–124. 72. Thieme K, Gromnica-Ihle E, Flor H. Operant behavio- ral treatment of fibromyalgia: a controlled study. Arthritis 55. Ohrbach R, Crow H, Kamer A. Examiner expectancy effects Rheum 2003; 49: 314–320. in the measurement of pressure pain thresholds. Pain 1998; 74: 163–170. 73. Zijlstra TR, Braakman-Jansen LM, Taal E, Rasker JJ, van de Laar MA. Cost-effectiveness of spa treatment for fibro- 56. Okifuji A, Turk DC, Sinclair JD, Starz TW, Marcus DA. A myalgia: general health improvement is not for free. standardized manual tender point survey. I. Development Rheumatology (Oxford) 2007; 46: 1454–1459. and determination of a threshold point for the identifica- tion of positive tender points in fibromyalgia syndrome. J 74. Häuser W, Bernardy K, Üçeyler N. Treatment of fibromyal- Rheumatol 1997; 24: 377–383. gia syndrome with antidepressants – a meta-analysis. JAMA 2009; 301(2): 198–209. 57. Petzke F, Gracely RH, Park KM, Ambrose K, Clauw DJ. What do tender points measure? Influence of distress on 4 75. McVeigh JG, McGaughey H, Hall M, Kane P. The effective- measures of tenderness. J Rheumatol 2003; 30: 567–574. ness of hydrotherapy in the management of fibromyalgia syn- drome: a systematic review. Rheumatol Int 2008; 29: 119–130 58. Wolfe F. Pain extent and diagnosis: development and validation of the regional pain scale in 12 995 patients. J 76. Tofferi JK, Jackson JL, O’Malley PG.Treatment of fibro- Rheumatol 2003; 30: 369–378. myalgia with cyclobenzaprine: a meta-analysis. Arthritis Rheum 2004; 51: 9–13. 59. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary 77. Füber N, Bernardy K, Häuser W, Spinath FM. Efficacy of care study. Primary Care Evaluation of Mental Disorders. psychotherapy in fibromyalgia syndrome – a metaanalysis. Patient Health Questionnaire. JAMA 1999; 282: 1737–1744. Schmerz 2008; 2(suppl): 165 (German). 60. Cleeland CS, Ryan KM. Pain assessment: global use of the 78. Burckhardt CS. Multidisciplinary approaches for manage- Brief Pain Inventory. Ann Acad Med Singapore 1994; 23: ment of fibromyalgia. Curr Pharm Des 2006; 12: 59–66. 129–138. 79. Häuser W, Bernardy K, offenbächer M, Arnold B, Schiltenwolf 61. Burckhardt CS, Clark SR, Bennett RM. The Fibromyalgia M. Efficacy of multicomponent treatment in fibromyalgia Impact Questionnaire: development and validation. J syndrome – a meta-analysis of randomized controlled clini- Rheumatol 1991; 18: 728–733. cal trials. Arthritis Rheum 2009; 61(2): 216–224. 62. Suarez-Almazor ME, Gonzalez-Lopez L, Gamez-Nava 80. Häuser W, Bernardy K, Üceyler N, Sommer C. Treatment of JI, Belseck E, Kendall CJ, Davis P. Utilization and pre- fibromyalgia syndrome with gabapentin and pregabalin – a dictive value of laboratory tests in patients referred to meta-analysis of randomized controlled trails. PAIN 2009, rheumatologists by primary care physicians. J Rheumatol june 16 epub. 1998; 25: 1980–1985. 81. Grossman P, Tiefenthaler-Gilmer U, Raysz A, Kesper U. 63. Lindgren H, Bergman S. The use and diagnostic yield of Mindfulness training as an intervention for fibromyalgia: radiology in subjects with longstanding musculoskeletal evidence of postintervention and 3-year follow-up benefits pain – an eight year follow up. BMC Musculoskelet Disord in well-being. Psychother Psychosom 2007; 76: 226–233. 2005; 6: 53. 82. Sephton SE, Salmon P, Weissbecker I, et al. Mindfulness 64. Busch AJ, Schachter CL, Overend TJ, Peloso PM, Barber KA. meditation alleviates depressive symptoms in women with Exercise for fibromyalgia: a systematic review. J Rheumatol fibromyalgia: results of a randomized clinical trial. Arthritis 2008; 35(6): 1130–1144. Rheum 2007; 57: 77–85. 65. Schiltenwolf M, Häuser W, Felde E, et al. Physiotherapy, 83. Mayhew E, Ernst E. Acupuncture for fibromyalgia – a exercise and physical therapy in fibromyalgia syndrome. systematic review of randomized clinical trials. Schmerz 2008; 22: 303–312 (German). Rheumatology 2007; 46: 801–804. 66. Langhorst J, W. Häuser W, Irnich D, et al. Complementary 84. Thieme K, Turk DC, Flor H. Responder criteria for operant and alternative therapies in fibromyalgia syndrome. and cognitive-behavioral treatment of fibromyalgia syn- Schmerz 2008; 22: 324–333. drome. Arthritis Rheum 2007; 57: 830–836. 67. Arnold B, Häuser W, Bernardy K, et al. Multicomponent 85. Koulil SV. Cognitive-behavioural therapies and exercise treatment of fibromyalgia syndrome. Schmerz 2008; 22: programmes for patients with fibromyalgia: state of the art 334–338 (German). and future directions. Ann Rheum Dis 2007; 66: 571–581. 133