There have been substantial advances in the pharmacotherapy of fibromyalgia (FM), which have occurred in parallel with advances in our understanding of the pathophysiology of FM in the past several years. Consortia of researchers have established a core set of symptom domains, which constitute the condition of FM, including pain, fatigue, sleep and mood disturbance and cognitive dysfunction, which significantly impact a patient’s overall well-being and ability to function. Outcome measures, which assess these domains, both singly and in composite format, are showing increasing reliability to discriminate between the treatment and placebo arms in clinical trials of emerging therapies, which are targeting the pathophysiologic mechanisms of FM. Several different medications, including the serotonin and norepinephrine reuptake inhibitors, duloxetine and milnacipran, and the α 2 δ modulator, pregabalin, have been approved by the Food and Drug Administration (FDA) for the management of FM, based on their clinically meaningful and durable effect on pain in monotherapy trials. They also have been shown to beneficially effect patient global impression of change, function and variably other key symptom domains, such as fatigue, sleep disturbance and cognition. Other medicines, although they have not gone through the formal approval process, have also shown efficacy in multiple domains of FM. Although combination trials have generally not yet been performed, the combined use of medicines with complementary mechanisms of action is rational, and, when done with appropriate caution, will likely be shown to be safe and well tolerated. Adjunctive therapy with medicines targeted at specific symptom domains, such as sleep, as well as treatments aimed at common co-morbid conditions, such as irritable bowel syndrome, or disease states, such as rheumatoid arthritis, should be considered for the purpose of reducing the patient’s overall symptom burden. Current therapies neither completely treat FM symptoms nor benefit all patients; thus, further research on new therapies with different mechanisms and side-effect profiles is needed.
Over the past decade, large clinical trials have demonstrated the ability of several medications, which target specific neurobiologic pathways known to be abnormal in patients with fibromyalgia (FM), to improve FM symptoms and functional impairment, leading to their approval in the United States (US). Historically, clinicians have employed combinations of medicines targeting the key symptoms described by patients, such as non-steroidal anti-inflammatory drugs (NSAIDs), muscle relaxants and narcotic analgesics for pain described in the musculoskeletal system, sedative hypnotics for sleep disturbance and antidepressants for mood disorders. The most common medications used by 2596 patients surveyed by the National Fibromyalgia Association (NFA) in 2005 were acetaminophen, NSAIDs, tricyclic antidepressants (TCAs) and cyclobenzaprine . Yet none of these medications have been formally approved for the treatment of FM, and only a few have been assessed, with mixed results, in controlled trials. Just as our understanding of the pathophysiology of FM is gradually evolving, so is our approach to treatment.
The first step towards establishing the efficacy of treatment in FM is to identify the key clinical domains of FM so that one may assess the impact of treatment on those domains, either singly or in a composite outcome measure. The FM working group of the association, Outcome Measures in Rheumatology Clinical Trials (OMERACT), has identified three levels of symptom domains of FM that should be assessed in clinical trials ( Fig. 1 ). The working group incorporated information from patient focus groups and physician and patient Delphi exercises, data mining of controlled clinical trials and discussions and voting at OMERACT meetings . The ‘core’ set, which should be measured in all clinical trials includes pain, tenderness, fatigue, sleep disturbance, patient ‘global’ sense of well-being as impacted by FM and multidimensional function, which includes not only physical but also social and psycho-emotional aspects of function. A second set of symptoms, which should be assessed in some trials of a medication, include depression and ‘dyscognition’ or cognitive dysfunction. The latter domain has not been fully characterised in FM nor is it best known how to reliably and practically assess it in clinical trials. A third or ‘research’ set of domains, which might be investigated, includes “stiffness,” which is considered important by patients but not well characterised in FM, anxiety and biomarkers such as cerebrospinal fluid neuropeptides and neuroimaging . It is not likely the case that a single medication will optimally benefit all of these domains. Indeed, the Food and Drug Administration (FDA) primarily focusses on efficacy in pain as it considers approval. However, as clinicians, we would like to know that a medicine we recommend can have a clinically meaningful effect in multiple FM symptom domains, including patient global well-being as well as function, and not necessarily just pain alone. The key outcomes reported in more recent drug trials have been efficacy in pain, patient ‘global’ and function, typically as individual variables, although in some trials, they have been evaluated in composite measures in which a patient, to be a successful “responder,” has to achieve a “clinically meaningful” improvement in at least two if not three of these domains simultaneously, and at more enduring time points of at least 3 months, if not 6 months or longer. Farrar, for example, has analysed a number of different trials of pregabalin in different pain conditions and noted that a “very much improved” and “much improved” response on the Patient Global Impression of Change (PGIC) scale correlated with a 30% improvement in pain Visual Analogue Scale (VAS), thus establishing this as a benchmark of clinically meaningful improvement . In that same study, “very much improved” correlated with 50% improvement in pain. Demonstration of achievement of these higher hurdles of response, statistically separated from placebo, reinforces the sense that these agents are truly benefitting specific pathophysiologic mechanisms and separating from the placebo response that can occur from simply being shown attention in a clinical trial. In most studies, other important domains are being assessed as ‘secondary’ measures, including measures of sleep, fatigue and cognition.
A variety of outcome measures for all of these domains have been used in clinical trials. Pain measures have included pain VAS, measured on paper or electronically, and multi-question assessments such as the Brief Pain Inventory (BPI). Multidimensional function can be measured with instruments such as the Fibromyalgia Impact Questionnaire (FIQ), an instrument validated in FM, and a more general questionnaire such as the Short Form-36 (SF-36). Patient ‘global’ is usually measured as a VAS or Likert scale. A variety of sleep measures, such as the Medical Outcomes Survey (MOS), have been used. Fatigue is typically measured by a multidimensional questionnaire, such as the Multidimensional Fatigue Inventory (MFI) or Multidimensional Assessment of Fatigue (MAF) instruments. Cognition has been measured, in an exploratory fashion, by the a self-report instrument, the Multiple Ability Self-report Questionnaire (MASQ), but more objective administered measures of cognitive function have not yet been used in large clinical trials. These and other measures used in clinical trials have been reviewed , and their psychometric quality of performance evaluated by the OMERACT group .
As discussed in several articles in this issue, the pathophysiology of FM is being more clearly elucidated and a complex interplay between genetics, developmental influences, triggering factors and the neurophysiologic and psycho-emotional substrate of the individual is being understood to yield the mix of symptom domains experienced by the patient.
In parallel with this methodologic and basic science work, a number of large controlled trials with drugs targeting the known neuropathophysiologic pathways of the disease have been completed, generally successfully. As a result of this cumulative work, the FDA and other regulatory agencies have been able to develop a regulatory pathway for approval of FM therapies, resulting in three currently approved therapies (pregabalin, duloxetine and milnacipran), with more agents with diverse mechanisms likely on the way. A similar process is now occurring in Europe. This work has also raised the level of understanding about the illness, its pathophysiology and rational treatment approaches amongst both clinicians and the public, leading to greater acceptance of the FM paradigm and improved likelihood that patients will be steered towards appropriate multimodal (pharmacologic and non-pharmacologic) therapy. The remainder of this article will focus on the evidence for efficacy, tolerability and safety of a variety of medicines used to treat FM, indicating efficacy in several individual domains of the condition and where available, efficacy in multiple domains measured in a composite fashion. It is worth noting that a methodologic research effort is currently underway, amongst the OMERACT FM group, to develop both measures of disease state severity and a responder index, which takes into account the multiple symptom domains of a patient, to provide a more coherent and clinically meaningful approach to the measurement of treatment effect in FM in the future.
Serotonin and norepinephrine modulation in FM
A central hypothesis in the pathophysiology of FM pain is that there is dysregulation of neuropeptide function in the descending and ascending arms of pain signalling pathways in the central nervous system. There is evidence of deficiency of the descending ‘inhibitory’ or ‘modulatory’ pathway between the brain and dorsal horn of the spinal cord and excessive nociceptive neurotransmitter function in the ascending pain pathways .
The demonstration of abnormal sleep physiology in FM patients in the pioneering studies of Moldofsky and Smythe suggested that FM pathology was central and potentially could be improved by the use of central neurotransmitter modulators.
TCAs were initially investigated after identification of alpha–delta non-rapid eye movement (NREM) sleep disruptions in FM patients caused by abnormalities in central serotonergic neurotransmission . Tertiary amine tricyclics (TCAs), such as amitriptyline, were studied because of their effect as serotonin–norepinephrine (NE) reuptake inhibitors, leading to the increased synaptic concentration of these neuropeptides . Three meta-analyses of TCA trials in FM have demonstrated beneficial effect on sleep quality and more variable and moderate effect on pain. FM patients demonstrated benefit for pain, sleep and overall sense of well-being with amitriptyline doses of 25–50 mg at bedtime, cyclobenzaprine (officially classed as a muscle relaxant, although a TCA by structure), 10–40 mg in short-term studies and, to a lesser extent, other medications classed as TCAs . Not all controlled trials were able to demonstrate benefit and, when amitriptyline was studied over a 24-week period, statistically significant separation from placebo was lost, raising questions about the ability of these agents to provide durable benefit . A further drawback to TCA use has been tolerability problems arising from anticholingeric, antiadrenergic, antihistaminergic and quindine-like effects, including sedation, constipation and cardiovascular issues, which diminish compliance and limit their long-term utility .
Another category of antidepressants evaluated for efficacy in FM have been the selective serotonin reuptake inhibitors (SSRIs), such as fluoxetine, citalopram and paroxetine . Several controlled trials with these agents in FM have shown mixed results regarding efficacy in pain, even though these agents are generally better tolerated than TCAs . Wolfe et al. found that fluoxetine in a fixed dose of 20 mg per day was not superior to placebo . On the other hand, Arnold et al., when allowing adjustable dosing, found that higher doses resulted in modest symptom benefit in FM patients . It is generally agreed that SSRIs have less impact on the pain of FM than the TCAs or the newer serotonin–norepinephrine reuptake inhibitors (SNRIs), even though they may be helpful for clinical domains, such as mood disorder or fatigue .
Building on the observation of benefit from agents, which have dual serotonergic and noradrenergic properties, there have been several large trials of newer SNRIs. These agents have greater neuroreceptor selectivity, are potentially more potent and are better tolerated than the older TCAs. The first of this newer class of SNRIs, venlafaxine, has been suggested to aid in the management of neuropathic pain and the prevention of migraine and tension headaches . However, venlafaxine shows conflicting results for efficacy in FM, with small open trials suggesting benefit but a larger controlled trial, albeit with low-dose (75 mg), not showing clear benefit over placebo .
Two newer SNRIs, duloxetine and milnacipran, have demonstrated more positive benefits for FM patients. Duloxetine is approved in the US for the treatment of depression, the pain of diabetic peripheral neuropathy and FM, and most recently, chronic musculoskeletal pain. Its efficacy in FM was demonstrated in two pivotal registry trials . The first study assessed both a 60-mg daily and 60-mg bid dose versus placebo in 354 women with FM for 12 weeks. Greater than or equal to 30% response of the BPI average pain severity score was seen in 55%/54%/33% of 60 qd/60 bid/placebo groups, respectively and ≥50% in 41%/41%/23%, respectively (each dose group statistically significant relative to placebo). The Patient Global Impression of Improvement (PGII) and total FIQ scores (used as a measure of function and quality of life), also were significantly improved. About a quarter of the patients had a concomitant diagnosis of major depressive disorder (MDD); they were allowed to participate and the response rates in the MDD subgroup were similar to the response rates of the group without MDD. A logistic regression methodology, known as Path analysis, demonstrated that the majority of the drug’s impact on pain was a direct analgesic effect and not indirectly mediated through its effect on depression. The idea that the effect of antidepressants on FM symptomatology occurs only because of their effects on psychiatric elements is a common misconception. The most common side effect was nausea, which was generally mild and typically improves with continued use of duloxetine. In general, patients did not gain weight.
A second pivotal study assessed 520 FM patients of both genders for 6 months’ placebo-controlled versus duloxetine in a similar dosing regimen . Approximately a third of the duloxetine-treated patients had a ≥50% improvement in BPI average pain score compared with approximately a fifth of the placebo group, which was statistically significant. Other secondary measures that showed improvement at both 3 and 6 month’s assessments in one or both duloxetine dose arms included PGII, total FIQ and several components of the fatigue measure, MFI. Path analysis in this study similarly showed that the majority of the pain-relieving effect of duloxetine was direct, as opposed to indirect effect through improvement in mood. Measures of sleep in duloxetine trials have shown that it does not specifically help with sleep, nor does it interfere with sleep. Based on the similarity of efficacy of the two different dosages of duloxetine that have been tested in FM, the FDA has approved a dose of 60 mg once a day.
Based on the known effectiveness of SNRIs in helping the symptom of fatigue and subjectively measured cognitive dysfunction (see milnacipran discussion later), a more rigorous study using objective cognition assessment batteries was performed in a duloxetine FM trial. In this trial, significant cognitive impairment was not noted and duloxetine treatment did not have a positive or negative effect on cognition . It is possible that other approaches to cognition assessment, other than those used in this trial, will be necessary to further elucidate cognitive impairment and potential benefit of treatment .
In addition to diabetic peripheral neuropathic pain (DPNP) and FM, duloxetine has demonstrated effectiveness in patients with chronic low back pain (CLBP) and chronic pain due to osteoarthritis (OA) , leading to the recent FDA approval of duloxetine for the management of chronic musculoskeletal pain. This is the first time that the FDA has granted such a broad label to an analgesic medication, and highlights the key role of central pain mechanisms in a variety of chronic pain conditions.
Milnacipran, an SNRI approved for the treatment of depression in parts of Europe and Asia, is approved for the management of FM in the US. The drug displays serotonin and norepinephrine reuptake inhibitor as well as mild N -methyl-D-aspartate (NMDA) inhibition properties. It has a greater ratio of norepinephrine to serotonin effect than other agents discussed here, tends to have very little in the way of drug–drug adverse interactions and is dosed twice daily. Studies of milnacipran, in contrast to duloxetine, screened out subjects with depression. The primary outcome measures used in two pivotal trials of milnacipran in FM were ‘FM responders’, defined as patients who fulfilled composite responder indices of 30% improvement in pain VAS, “very much” or “much improved” PGIC scale and six-point improvement of SF-36 physical component score (PCS) and ‘FM pain responders’ defined as response of pain and patient global, but did not include the SF-36 function component .
The first study included 888 FM patients of both genders, randomised 1:1:2 to placebo, 50-mg bid and 100-mg bid milnacipran, respectively . At the primary endpoint of 15 weeks, 33% of observed cases (OCs) achieved FM responder status in both milnacipran dose arms versus 19% of placebo patients. Using a more strict baseline observation carried forward (BOCF) analysis, approximately 19% of treated versus 12% of placebo patients did so. Both analyses showed statistically significant separation of milnacipran treatment from placebo. FM-pain-responder status was achieved by 45% of milnacipran-treated versus 27% of placebo-treated patients using an OC analysis, and 27% versus 19% using a BOCF analysis. Secondary measures that showed statistically significant improvement included fatigue as measured by the MFI and a self-reported measure of cognition, the MASQ. Efficacy was generally maintained through the second time point of assessment, 27 weeks. The most prominent side effects were mild to moderate nausea and headache, which typically resolved with continued use of the medication. The patients did not gain weight, not unlike the experience with duloxetine. A 12-month treatment group showed that long-term treatment with milnacipran maintains positive effects and is safe .
A second pivotal trial of milnacipran in FM assessed 1196 patients randomised 1:1:1 to placebo, 50-mg bid and 100-mg bid of milnacipran for 15 weeks . The same measures of response were used and the results were similar, showing statistically significant separation of response of both dose arms compared with placebo in both FM-responder and FM-pain-responder rates per BOCF and OC methods of analysis. Key secondary measures that showed statistically significant improvement included PGIC, function (SF-36) and fatigue (MFI). The adverse event profile was similar to that of the first study. An important point to note with this agent as well as duloxetine is that patients tend to accommodate better to the drug when a gradual initial dose titration is employed. Further, similar to duloxetine, milnacipran neither helps nor hinders sleep quality. Based on the data from these trials, the FDA has approved milnacipran for the management of FM in both 50-mg bid and 100-mg bid dosages.
Two other phase 3 trials of milnacipran in FM have been completed and have confirmed the results from the pivotal trials, including a trial conducted in Europe in 884 patients, using the 100-mg bid dose and a trial in 1025 patients using a 50-mg bid dose .
A trial assessing the effect of adding milnacipran to pregabalin in patients experiencing an inadequate effect with pregabalin has been conducted . This demonstrated additive benefit in measures such as PGIC, pain, function, fatigue and cognition and importantly, the combination was well tolerated.
Antiepileptic drugs
It has long been known that certain drugs originally developed for the treatment of epilepsy could have pain-relieving effects. The clearest evidence for this has been with pregabalin and gabapentin. Pregabalin, approved for the treatment of diabetic peripheral neuropathy, post-herpetic neuralgia, as well as adjunctive therapy in partial-onset seizures, was the first pharmaceutical approved by the FDA for FM treatment. It binds to the α 2 δ auxiliary protein associated with voltage-gated calcium channels, and modulates neuronal calcium influx. The result is a reduction of release of several pain pathway neurotransmitters, such as glutamate and substance P, which play a role in pain processing. Efficacy in FM was first established in an 8-week trial in 529 FM patients, using dosages of 150-mg, 300-mg and 450-mg total daily dose (divided tid) versus placebo. The two higher dose arms showed statistically significant improvements in pain, patient global impression of improvement with therapy, sleep disturbance, fatigue and health-related quality of life in FM patients . Statistically significant responder rates, using pain VAS, showed ≥50% response in 29% of the 450-mg group versus 13% in placebo and ≥30% in 48% and 27% of 450-mg and placebo groups, respectively. The 300-mg and 150-mg groups did not separate from placebo statistically. Phase 3 trials of pregabalin confirmed these observations, including efficacy in the three key domains of pain, patient global improvement and function, and supported longer-term durability of effect out to 6 months . An important secondary domain that demonstrated improvement was sleep. Adverse effects experienced by some patients included dizziness and sedation, which tended to improve and resolve during the initial period of use, as well as weight gain in a small percentage of patients. Based on these results, the FDA granted approval for the use of pregabalin for the management of FM in dosages of 300–450 mg. As with the SNRIs, gradual dose titration improves initial tolerability with this agent.
A single controlled trial with gabapentin showed efficacy in multiple FM symptom domains, including pain, PGIC, function and sleep . The median dose used by 150 patients randomised to gabapentin versus placebo was 1800 mg in divided dose per day (range 1200–2400 mg allowed and titrated to tolerance). The side-effect profile was similar to that of pregabalin. Although mechanistically similar to pregabalin, the pharmacokinetic and pharmacodynamic profile of gabapentin in not as favourable as that of pregabalin.
Antiepileptic drugs
It has long been known that certain drugs originally developed for the treatment of epilepsy could have pain-relieving effects. The clearest evidence for this has been with pregabalin and gabapentin. Pregabalin, approved for the treatment of diabetic peripheral neuropathy, post-herpetic neuralgia, as well as adjunctive therapy in partial-onset seizures, was the first pharmaceutical approved by the FDA for FM treatment. It binds to the α 2 δ auxiliary protein associated with voltage-gated calcium channels, and modulates neuronal calcium influx. The result is a reduction of release of several pain pathway neurotransmitters, such as glutamate and substance P, which play a role in pain processing. Efficacy in FM was first established in an 8-week trial in 529 FM patients, using dosages of 150-mg, 300-mg and 450-mg total daily dose (divided tid) versus placebo. The two higher dose arms showed statistically significant improvements in pain, patient global impression of improvement with therapy, sleep disturbance, fatigue and health-related quality of life in FM patients . Statistically significant responder rates, using pain VAS, showed ≥50% response in 29% of the 450-mg group versus 13% in placebo and ≥30% in 48% and 27% of 450-mg and placebo groups, respectively. The 300-mg and 150-mg groups did not separate from placebo statistically. Phase 3 trials of pregabalin confirmed these observations, including efficacy in the three key domains of pain, patient global improvement and function, and supported longer-term durability of effect out to 6 months . An important secondary domain that demonstrated improvement was sleep. Adverse effects experienced by some patients included dizziness and sedation, which tended to improve and resolve during the initial period of use, as well as weight gain in a small percentage of patients. Based on these results, the FDA granted approval for the use of pregabalin for the management of FM in dosages of 300–450 mg. As with the SNRIs, gradual dose titration improves initial tolerability with this agent.
A single controlled trial with gabapentin showed efficacy in multiple FM symptom domains, including pain, PGIC, function and sleep . The median dose used by 150 patients randomised to gabapentin versus placebo was 1800 mg in divided dose per day (range 1200–2400 mg allowed and titrated to tolerance). The side-effect profile was similar to that of pregabalin. Although mechanistically similar to pregabalin, the pharmacokinetic and pharmacodynamic profile of gabapentin in not as favourable as that of pregabalin.
Related posts:
Chronic widespread pain or fibromyalgia? That is the question
Chronic widespread pain: From peripheral to central evolution
Trauma and work-related pain syndromes: Risk factors, clinical picture, insurance and law interventions
Translational musculoskeletal pain research
Non-pharmacological treatment of chronic widespread musculoskeletal pain
Multidisciplinary approach to fibromyalgia: What is the teaching?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
