Amitriptyline

For many, amitriptyline has been a mainstay and often first-line agent used to treat FM. The RCT evidence for its popularity arises from a handful of supportive studies between 1986 and 1994 (Carette et al 1986, Goldenberg et al 1986, Jaeschke et al 1991). While doses ranging from 10–50 mg at bedtime for 2–9 weeks demonstrated significant pain reduction, Carrette reported no benefit with 30 mg at bedtime over a 24-week period (Carette et al 1994). Further supportive evidence, however, was found in a fluoxetine-amitriptyline, four-arm, 6-week, cross-over study in 1996 by Goldenberg (Goldenberg et al 1996). Either agent alone decreased pain and fatigue better than placebo, but their combination significantly reduced the Fibromyalgia Impact Questionnaire (FIQ) by 34% (Box 7.1.1).

Arguably, the FIQ has become the preeminent FM assessment tool evaluating pain, function, fatigue, depression, anxiety and work functions (Burckhardt et al 1991). While in need of some updating (Bennett 2005), it remains ostensibly a mandatory inclusion in any FM study or application for regulatory approval for a new FM treatment option. Meta-analyses of amitriptyline demonstrated even stronger evidence for its inclusion into the FM armamentarium, yet the overall benefits were modest (Arnold & Keck 2000, O’Malley et al 2000, Nishishinya et al 2008). Clinical responses ranged from 25–37% benefit, but while sleep, fatigue and pain improved, tender point score did not. Comparing amitriptyline to other types of antidepressant has been attempted, but evidence is weak. Only one study has been conducted with nortriptyline; it was found to be ineffective (Heymann et al 2001).

Cyclobenzaprine

Although often considered a muscle relaxant, cyclobenzaprine is structurally a tricyclic moiety, and has demonstrated similar reduction of FM pain (30%) in several trials (Bennett et al 1988, Quimby et al 1989, Hamaty et al 1989). All RCTs were dosed at 30 mg per day over 4–12 weeks, but again, Carrette found inefficacy at 24 weeks (Carette et al 1994). This study comparing amitriptyline, cyclobenzaprine and placebo yielded a benefit at 6 and 12 weeks, but not at 24 weeks. However, a meta-analysis endorsed the efficacy of cyclobenzaprine for treatment of FM (Tofferi et al 2004). Even very low doses (1–4 mg) at bedtime have demonstrated some analgesic efficacy and have improved sleep stage architecture (Modolfsky 2002).

Selective serotonin reuptake inhibitors

Contrary to the Goldenberg crossover study with amitriptyline, the selective serotonin reuptake inhibitor (SSRI) fluoxetine at a dose of 20 mg each morning was not superior to placebo after 6 weeks (Wolfe et al 1994). Yet, in another study allowing flexible dosing ≤ 80 mg/d, statistically significant efficacy was evident (Arnold et al 2002). Not unexpectedly, depression sub-score of the FIQ contributed to overall FIQ score improvement, but pain and fatigue were also reduced compared to placebo, even after correcting for depression score. Again, tender-point (TP) score did not improve, and this observation has become a theme in FM research. To date, RCTs have only rarely recorded a statistically significant improvement in TP score, which is intriguing given that TPs are the hallmark clinical diagnostic sign of FM.

Other SSRIs have been assessed in RCTs, including a report of sertraline (50 mg) attaining equivalent efficacy with amitriptyline (25 mg) (Celiker & Cagavi 2000). Üçeyler et al (2008) conducted an exhaustive review of antidepressant studies in FM concluding multi-outcome efficacy for amitriptyline (25–50 mg/d), fluoxetine (20–40 mg/d), sertraline (50 mg/d) and paroxetine (20 mg/d) (Patkar et al 2007) but not for citalopram (20–40 mg/d) (Nørregaard et al 1995). Unfortunately for SSRIs as a group, mean pain reduction and mean improved quality of life was only 18% and 16%, respectively, after correcting for placebo effect. Therefore, such options are unlikely to achieve substantial clinical benefits or meaningful remission for most FM patients.

Muscle relaxants and benzodiazepines

The muscle relaxant, carisoprodol, has been found to reduce FM pain in one trial dosed at 1200 mg/d for 8 weeks (Vaerøy et al 1989). Unlike other muscle relaxants, carisoprodol is metabolized to an anxiolytic, meprobamate, a potential drug of abuse (Bailey & Briggs 2002) and is associated with teratogenicity (Timmermann et al 2008) and increased road traffic accidents (Logan et al 2000). This reduces clinical confidence given concern for addictive risk and tolerance, yet its muscle-relaxant properties are considerable. Comorbid positional cervical spinal cord compression (PC3) has been documented in 65% of patients with FM in a recent pilot study using flexion–extension cervical spinal magnetic resonance imaging. Theoretically, this combination of FM and PC3 may be more amenable to treatment with this agent given the benefit of some anxiolytics (benzodiazepines) for spinal cord irritability among patients with cervical spinal cord injury (Merritt 1981, Dahm et al 1989, Winkler et al 1997).

An open trial of two benzodiazepines, lorazepam or clonazepam (2 mg at bedtime), among 202 patients with FM demonstrated widespread and considerable reduction in pain over 2 weeks (Holman 1998) sustained for 49% of subjects at 52 weeks without dose escalation or abuse (Holman 1999). Both agents also decrease symptoms of restless leg syndrome (RLS), found to be more common among patients with FM (34% vs. 2% controls) (Yunus & Aldag 1996), and may reduce cervical spinal cord irritability (Fenollosa et al 1993). Studies with other benzodiazepines have yielded mixed findings. Alprazolam (0.5–3.0 mg/d) for 6 weeks was ineffective (Russell et al 1991), while temazepam (15–30 mg at bedtime) for 12 weeks was beneficial (Hench et al 1989). Neither agent reduces RLS or spinal cord irritability, but temazepam may influence sleep deficit in FM.

NSAIDs, steroids and opiates

Non-steroidal anti-inflammatory drugs (NSAIDs), including naproxen (Goldenberg et al 1986) and ibuprofen (Yunus et al 1989), as well as corticosteroids (Clark et al 1985) have been uniformly ineffective in FM RCTs. Tramadol, a centrally acting analgesic with combined mu-opioid activity and inhibition of serotonin/norepinephrine reuptake, has been studied in three RCTs (Biasi et al 1998, Russell et al 2000, Bennett et al 2003). A variety of pain measures improved statistically with good tolerability. In an interesting design based on time to study withdrawal for inefficacy over 91 days with 315 subjects, those in the tramadol arm were less likely to withdraw for any reason (48% [tramadol] vs. 62% [placebo]) and noted better pain relief and improved FIQ scores. Most common adverse effects (AE) were nausea (20%) and constipation (15%). Tizanidine, an alpha-2 receptor antagonist useful for reduction of muscular spasticity, has also been beneficial and reduces cerebrospinal fluid (CSF) neuroamines and substance P levels in patients with FM (4–24 mg/day) (Russell et al 2002).

Curiously, there have been no opioid RCTs in patients with FM. Nevertheless, their use is not uncommon, yet clinicians may assume inefficacy without justification. And, as understanding of cervical spinal cord research expands into a reappraisal of chronic widespread pain (CWP) and FM (Chapter 5), clinicians may eventually discover a less biased and more evidence-based approach to narcotic usage.

Sodium oxybate

To date, only one commercial compound induces stage 4 sleep: sodium oxybate (Lapierre et al 1990). This agent is FDA approved for treatment of narcolepsy with cataplexy and is a naturally occurring CNS metabolite found in the hippocampus and basal ganglia (Snead 3^rd & Liu 1984). Given its potential as a drug of abuse and its ease of non-commercial manufacture, it is highly regulated and available in the US only through a central pharmacy or in the UK on a case-by-case basis. Nevertheless, Scharf et al (2003) initially conducted a double-blind cross-over trial over 1 month with a 2-week wash out period. A total of 18 subjects completed the trial dosed at 6.0 g at bedtime (two divided doses, 4 hours apart) and were monitored by polysomnogram (PSG), tender point index and subjective measures of improvement in daily diaries. A variety of pain and fatigue scores improved by 29–33% in the active arm compared with 6–10% in the placebo arm (P < 0.005). Tender point score decreased by 8.4 points with sodium oxybate compared with an increase of 0.4 points in the placebo arm (P = 0.008). PSG measures of inappropriate CNS arousal, including alpha wave intrusion, sleep latency and rapid eye movement (REM) sleep decreased with treatment, while stage 3 and 4 slow-wave sleep increased compared with placebo (P < 0.005).

In a second study, 188 subjects were randomized (1:1:1) to placebo and sodium oxybate doses of 4.5 g and 6.0 g in divided nightly doses (Russell et al 2005). Outcomes were more challenging and included a triple outcome measure of ≥20% improved 10 cm visual analogue pain (PVAS), ≥ 20% improved total FIQ, and either ‘much better’ or ‘very much better’ on patient global impression of change (PGIC). This triple primary outcome was achieved by 13% receiving placebo, 35% receiving 4.5 g (P = 0.005) and 27% receiving 6.0 g (P = 0.05) doses. Dose response was lacking possibly due to greater early withdrawal from overly rapid initial dose escalation in the 6.0 g arm and the use of a categorical outcome variable. Based on four PSG evaluations per subject, sleep quality measures also improved, including increased stage 4 sleep (Moldofsky & Alvarez-Horine 2008). The most common adverse events (AEs) were nausea and dizziness.

A large phase 3 trial designed for application to the FDA for approval for use in FM was reported by Jazz Pharmaceuticals (Palo Alto, CA, USA) (2008) as successful with details to follow. Although sodium oxybate does not directly address dysautonomia or have a known cervical spinal cord effect, it directly induces stage 4 sleep, which may begin to explain its clinical benefit in FM. Further research will be required to determine how this agent will be incorporated into the host of new options available to clinicians and combination trials are underway.

Gabapentin

Gabapentin has been evaluated in an FM RCT reported by Arnold et al (2007). In this 12-week RCT, 1800 mg per day of gabapentin was compared to placebo in 150 age/gender-matched subjects. Using the Brief Pain Inventory (BPI) on a 0–10 scale, gabapentin reduced pain by a difference of –0.92 [95%CI = –1.75, –0.71] compared with controls (P = 0.015). However, given the mean baseline pain score of 5.8, the relative benefit reflected only a 16% reduction in pain. Mean FIQ benefit was –8.4 (P = 0.001), or an 18% improvement. The most common AEs include dizziness, somnolence, oedema, light-headedness and weight gain.

Pregabalin

The first medication to achieve FDA approval for treatment of FM was pregabalin, an analogue of gabapentin with greater target affinity at the α₂-δ receptor. Initially, in an 8-week study, three pregabalin doses (150 mg/d, 300 mg/d and 450 mg/d) and placebo were randomized (1:1:1:1) to 529 patients (Crofford et al 2005). Mean baseline pain score was 7 (range: 1[no pain]–10[worst pain]) for a group comprised of 90% women with a mean FM duration of 8 years. All patients discontinued concomitant medications and 77% completed the trial. Improvement was noted in all active arms as early as week 2, and was sustained through week 8. Maximum pain benefit relative to placebo (in the 450 mg arm) was moderate (–0.93 [1–11 scale]), while the degree of statistical significance (P = 0.0009) likely reflected an ample study size.

Additional RCTs assessed up to 600 mg/d, but pain reduction was again modest (–0.66 [P = 0.007]) (Mease et al 2008a). With both studies, response was consistent, but equivalent to a 13% and 9% reduction in pain from baseline, respectively. Lower doses in the initial study (150 mg and 300 mg) were not statistically superior to placebo, but enlarging the second study to 748 subjects enhanced statistical validity of benefit associated with the 300 mg arm in the second study (–0.43 [0–10 scale]; P = 0.0449). In the first study, 29% of subjects achieved a ≥50% reduction in pain compared to 13% receiving placebo (P = 0.003). A non-significant difference in ‘responders’ defined as ≥ 30% reduced pain was reported for the 300 mg (43%), 450 mg (43%), 600 mg (44%) and placebo (35%) in the second study. Still, secondary outcomes favoured pregabalin over placebo, including assessments of sleep, patient assessment of global change, and several other domains of quality of life. Important AEs were dose related and included dizziness, somnolence, weight gain and dry mouth. The rationale for selection of pregabalin as an FM option beyond its FDA approval include its consistent, albeit modest response, improvement in secondary outcomes, and its mechanism of action, which may further implicate a spinal cord concern. Positional cervical cord compression (PC3) imbedded in the FM presentation could be aggravated by enhanced range of motion prevalent among patients with JHS. Although not yet studied, this dysautonomic concern may be responsive to pregabalin, since pregabalin is already specifically approved by the EMEA for treatment of central, cervical spinal pain in Europe (Pfizer Inc 2006). Further studies are required to evaluate whether pain reduction in the pregabalin FM RCTs is due to an effect on comorbid PC3 embedded within FM or due to an effect on FM itself.

Duloxetine

Of five important duloxetine RCTs, two pivotal trials secured the FDA indication for treatment of FM. Earlier, an initial proof of concept trial randomized 207 patients (1:1) to 60 mg/d and placebo for 12 weeks (Arnold et al 2004). Interestingly, it did not achieve its dual primary outcomes: FIQ total score (treatment difference –5.53 [P = 0.027]) and FIQ 0–10 pain VAS subscore (treatment difference –0.63 cm [P = 0.13]). But, further sub-analysis revealed that treatment response in the 22 men unexpectedly favoured placebo over duloxetine. Also, a secondary measure of pain, the Brief Pain Inventory (BPI) did achieve statistical significance. Consequently, a second 12-week trial of 354 patients eliminated men and employed the BFI as the primary outcome (Arnold et al 2005). Subjects were randomized (1:1:1) to 60 mg daily, 60 mg twice per day and placebo. Pain decreased significantly using the BPI (P < 0.001) and was independent of mood. Reduction of pain by ≥ 50% was achieved by 41% receiving 60 mg, 41% receiving 120 mg and 23% receiving placebo (P = 0.003).

In a third study by Russell et al (2008), and the second pivotal study submitted for FDA review, 520 subjects (5% men and 25% with major depressive disorder [MDD]) were randomized to duloxetine doses of 25 mg, 60 mg, 120 mg and placebo. Pain reduction was significant overall as well as greater for those with MDD, but did not distinguish 60 mg from the 120 mg dose. Adverse events and premature discontinuation were dose-dependent. In a fourth study by Chappell et al (2007) of non-responders to 60 mg were randomized to either continuation of 60 mg or an increased dose of 120 mg/d. By 60 weeks, treatment at 120 mg was no more likely to achieve >30% reduced pain than continuing the 60 mg/d dosage.

Safety was reported from two prior RTC extensions and was a primary endpoint of a fifth recent study of over 60 weeks (Chappell et al 2008). Important safety concerns with duloxetine in FM trials have been nausea, headache, insomnia, dizziness and constipation as well as concern raised over hepatic failure, hyponatraemia and orthostatic hypotension.

Of note, all antidepressants as a class, including duloxetine, carry a label warning from the FDA highlighting an increased risk of suicide, especially among children and young adults. Neither duloxetine nor milnacipran is approved for use among children. In terms of potential drug interactions, some analgesics commonly used by patients with FM require a cautionary note. Tramadol, previously reviewed, should not be co-administered with duloxetine given the risk of developing serotonin syndrome. And, through cytochrome P450 enzyme system interactions, duloxetine may prolong effects of methadone, oxycodone and propoxyphene, but not fentanyl, codeine and morphine.

Polypharmacy is a commonly employed strategy in FM, since no treatment is sufficiently effective. Nevertheless, cautious appraisal and reappraisal of patient safety, tolerability and efficacy remains fundamental.

Milnacipran

Recently, milnacipran also successfully competed for FDA approval as a treatment of FM. An initial proof of concept trial of milnacipran dosed 100 mg/d, 100 mg twice per day and placebo randomized 125 patients (3:3:2) and defined its primary outcome as either ≥ 30% or ≥ 50% reduction in pain at 12 weeks (Gendreau et al 2005). Only the 100 mg twice per day dose achieved a statistically significant outcome of ≥ 50% decreased pain compared to placebo (37% [100 mg twice a day; P = 0.04], 22% [100 mg daily; P = 0.55], 14% placebo). Other pain assessments, including Gracely pain scores, demonstrated a significant benefit, yet the 10 cm VAS pain scale did not (P = 0.595).

Two subsequent and much larger 3-month studies, documented efficacy for both the 100 mg daily and 100 mg twice per day dosing (P value range = 0.004–-0.025) (Mease et al 2008b, Clauw et al 2008). These studies (N = 888 and N = 1196) used a composite outcome defined as ≥ 30% pain reduction combined with a rating of ‘much improved’ and ‘very much improved’ on the PGIC, somewhat similar to the sodium oxybate trials. Secondary outcomes, including the FIQ total score and Multi-Dimensional Fatigue Inventory (MFI) also improved significantly. A third large study randomized 884 patients to 100 mg twice daily and placebo (1:1) for 12 weeks to demonstrate efficacy with the same composite outcome (P = 0.003) as well as significant improvement in FIQ (P = 0.015) and other secondary measures (Short Form 36 Health Survey, MFI, and FIQ subscores) (Branco et al 2008). Important AEs included nausea, headache, constipation, hyperhidrosis, hot flushes, dizziness, palpitations, dry mouth and hypertension.

Pramipexole

Related posts:

Fibromyalgia and hypermobility The physiology of pain Physiotherapy and occupational therapy in the hypermobile adolescent Principles of rehabilitation and considerations for sport, performance and fitness Regional complications in joint hypermobility syndrome Neuromuscular physiology in joint hypermobility

Stay updated, free articles. Join our Telegram channel

Join