Animals

Animal experiments suggest that expansion of receptive fields is brought about by the opening of ‘silent’ (ineffective) synapses in the spinal cord by nociceptive input from muscle . Neurokinin-1 (NK-1) and N -methyl- d -aspartate (NMDA) receptors are involved in such an expansion . Within the first 2 h after induction of an experimental myositis, a local anaesthetic block of the muscle afferents can prevent the central sensitisation from developing. If the block is later (2–4 h after induction of the myositis), there is no influence on the development of the central sensitisation . This may underline the importance of an early and effective analgesic therapy for the prevention of development of central hyperexcitability.

As with other tissues, when the muscle is in the state of hyperalgesia or allodynia, the sensitised muscle nociceptors are then more easily activated and may respond to normally innocuous and weak stimuli such as light pressure and muscle movement. Animal studies have found a development of new receptive fields via noxious muscle stimuli . In the context of referred pain, revealing new receptive fields could be the mechanism behind referred pain because of central sensitisation or hyperexcitability . Forming of new receptive fields has been suggested to be the phenomenon of secondary hyperalgesia in deep tissue. Furthermore, an expansion of the receptive fields proximal to the normal receptive field was found in a study where experimental myositis was induced, and, afterwards, application of antagonists to three different NK receptors was effective in preventing the induced hyperexcitability . An important factor, not yet substantiated in patients, is the contralateral spread of sensitisation following a localised muscle trauma .

Human experimental studies

Referred pain is a pain sensation that is perceived at an area other than the site of nociceptive stimulation and may arise from central spinal and other sensitisation events. Referred pain patterns have been described using experimental stimuli applied to different structures, mostly muscles, joints and viscera. It is still controversial whether or not hyperalgesia is present in referred muscle pain areas. Kellgren and Feinstein et al. found hyperalgesia, Steinbrocker et al. found no changes and Graven-Nielsen et al. found hypoalgesia to pressure. In the area of referred pain from OA of the knee, trophic changes in the skin have been found .

Referred muscle pain is probably a combination of central processing and peripheral input as it is possible to induce referred pain in limbs with complete sensory loss resulting from an anaesthetic block . The area of the referred pain is correlated with the intensity of the muscle pain, and the appearance of referred pain is delayed by 20–40 s compared with local muscle pain , indicating that a time-dependent process, perhaps the unmasking of new synaptic connections, is involved in the neural mediation of referred pain. Saline-induced referred pain occurs less frequently in healthy subjects treated with ketamine compared with a placebo treatment , indicating the involvement of central sensitisation.

Clinical studies

The manifestations of expanded receptive fields (central sensitisation) in chronic musculoskeletal pain conditions have been less studied. Nonetheless, Laursen et al. showed generalised sensitisation to pressure stimulation (reduced pressure pain thresholds) in chronic musculoskeletal and visceral pain conditions.

Besides standard clinical examinations, a simple clinical tool to assess patients can be pain drawings to illustrate the localisation and extent of, for example, myofascial pain areas, pain intensity during rest and movement assessed on a visual analogue pain scale, pain quality assessed by the McGill pain questionnaire and the quality of life assessment (e.g., Short Form (36) Health Survey (SF-36)). Information on concomitant sites of pain has been largely neglected, even though their presence draws attention towards the involvement of widespread pathophysiological mechanisms. More advanced signs could include increased hyperalgesic reaction to cutaneous capsaicin in fibromyalgia or rheumatoid arthritis , increased numbers of trigger points in osteoarthritis or spread of pain over time in lateral epicondylitis . Central sensitisation in fibromyalgia and whiplash patients is also suggested, based on increased nociceptive withdrawal reflexes in those patients as compared with controls .

Patients with fibromyalgia, compared with healthy subjects, experience stronger pain intensity and larger referred areas after intramuscular (i.m.) injection of hypertonic saline in the leg ; no spontaneous leg pain was present in those patients. Similar results were obtained in studies on other chronic musculoskeletal pain syndromes, such as whiplash , knee osteoarthritis and low back pain . Expansion of the referred areas in fibromyalgia patients is prevented by ketamine . This may be indicative of generalised sensitisation in those patient populations .

Animal studies

Repeated strong C-fibre stimulation of somatic nociceptive fibres causes a frequency-dependent increase in neuronal excitability, which outlasts the stimuli. The resulting response of spinal cord neurones to successive stimuli of this type is a progressive increase in the magnitude of the Aδ- and C-fibre input , and is often followed by the development of afterdischarge. Windup has been used as a model of neural plasticity and central sensitisation in the spinal cord. Postsynaptic actions of the neurotransmitters (e.g., substance P (SP) and glutamate) released by repeated noxious stimuli can contribute to the enhanced excitability.

Windup and central sensitisation are not identical phenomena, as windup does not persist for a long time after stimulation, whereas sensitisation can be long-lasting. Nonetheless, there are apparent similarities in transmitters and pathways underlying windup and sensitisation, and windup initiates and maintains central sensitisation . It has been shown that windup increased the receptive field area of dorsal horn neurones , which is a feature of central sensitisation.

Human experimental studies

In humans, the initial phase of the windup process can be measured by increased pain response to consecutive stimuli. Various modalities have been used to induce temporal summation in muscles (electrical, pressure and algesic substances) . Temporal summation of pain stimuli applied to skin, joint and muscle showed the highest reproducibility for muscle tissue . If i.m. chemical stimulations are repeated, the pain will increase and, at the same time, the referred pain area is increased . Increased temporal summation to pressure is found under experimental conditions with experimentally induced muscle hyperalgesia .

Clinical studies

Muscular temporal stimulation is significantly facilitated in chronic pain patients with fibromyalgia or whiplash compared with controls. Facilitated temporal summation suggests the involvement of central sensitisation in patients with chronic musculoskeletal pain . The increased efficacy of temporal summation in fibromyalgic patients has been reproduced with cutaneous heat stimulation , indicating that there is no tissue-specific sensitisation. In fibromyalgic patients, the exaggerated temporal summation is partly inhibited by ketamine . Recently, facilitated temporal summation of pressure pain was found for OA knee pain patients compared with matched controls . Interestingly, this facilitated temporal summation was manifested on the OA knee and on the lower leg muscle.

Animal studies

The inhibitory control of nociceptive neuronal excitability in animals is manifested via inhibitory circuits operating principally at the segmental level and via pathways that originate at higher central nervous system levels, such as cerebral cortex, thalamus and brainstem (e.g., periaqueductal grey (PAG), raphe nuclei and rostroventral medial medulla (RVM)). One manifestation of such inhibitory influences is that associated with diffuse noxious inhibitory control (DNIC) that has been especially explored in the spinal system, and reported to be selectively expressed on ‘convergent’ (wide dynamic range, WDR) nociceptive neurones . The inhibitory bulbospinal pathways are serotoninergic and inhibit the activity of especially the WDR spinal dorsal horn neurones, but additional non-serotoninergic mechanisms (e.g., opioid) are also involved.

In addition, descending facilitatory influences on nociceptive neuronal excitability are manifested via pathways that originate in the midbrain and brainstem structures (e.g., PAG, raphe nuclei and RVM). These are key structures in the descending modulatory repertoire, which, together with the descending inhibitory influences, allow for the bidirectional control of spinal cord activity through descending facilitatory and inhibitory networks .

Human experimental studies

In human experimental and clinical studies, it is not possible to dissociate between the two competing systems, inhibition and facilitation; but only the net effect can be assessed. Most human studies have addressed DNIC although they may not only assess DNIC. Painful heterotopic conditioning tonic stimuli (thermal, mechanical, electrical or chemical) decrease pain perception induced by phasic noxious stimulation given elsewhere in the body . Interestingly, in humans, it was found that μ-opioid receptors were activated in multiple cortical and subcortical brain regions during saline-induced muscle pain; for example, the μ-opioid receptor availability in the amygdala was reduced by muscle pain and was negatively correlated to the pain intensity . Indeed, the amygdala is involved in the antinociception via descending inhibition . Many of the conditions with impaired DNIC have a female predominance, and it has been found experimentally that females show less efficient inhibition as compared with males , indicating that gender may be an important parameter. Moreover, the μ-opioid receptor availability in the amygdala during saline-induced muscle pain is higher in females compared with males , illustrating the more pronounced hypoalgesic effect in men compared with women . A later study reported that the level of μ-opioid receptor activation during saline-induced muscle pain was related to genetic variations. This might further add to the variability of somatosensory changes in experimental muscle pain studies.

Clinical studies

Less efficient DNIC-like mechanisms have been shown in musculoskeletal pain conditions such as myofascial TMD , chronic low back pain , fibromyalgia , painful OA and chronic tension-type headaches . Interestingly, the patients in the OA study, who presented first with a deficient DNIC-like pain inhibition, showed normal inhibition after surgery in a pain-free state , suggesting that the chronic pain maintained the DNIC dysfunction and that ongoing pain from one site may interact with DNIC evoked by pain from another area. DNIC stimuli applied to patients with neuropathic pain cause a decrease in the area of brush-evoked pain, but result in no changes in the intensity of brush-evoked pain or spontaneous pain intensity , suggesting differential effects of DNIC on different aspects of nociceptive processing. The mechanism of descending control seems intact in short- and long-term rheumatoid arthritis patients compared with controls .

Animal studies

Muscle nociception is normally evoked by overuse, trauma or inflammation. Damaged muscle tissue release, for example, potassium ions, prostaglandin E2, bradykinin and serotonin and the nerve endings release, for example, neuropeptides, such as SP, calcitonin gene-related peptide (CGRP) or somatostatin . Most of the sensitising substances released in muscles are similar to those released in other tissues, but adenosine triphosphate (ATP) is present in muscle cells in high concentration and seems more specific for muscles. During muscle inflammation, the density of nerve endings containing SP and nerve growth factor (NGF) increases .

When the muscle is in the state of hyperalgesia or allodynia, the sensitised muscle nociceptors are more easily activated and may respond to normally innocuous and weak stimuli, such as light pressure and muscle movement.

Human experimental studies

Microneurographic recordings in volunteers showed that the responses of human muscle nociceptors to injections of algesic substances were comparable to those of animal nociceptors . Experimental induction and assessment of pain and peripheral sensitisation in humans can be performed by a variety of chemical, mechanical, thermal and electrical methods . Generally, the induced sensitisation is short lasting but may still mimic some of the signs and symptoms seen in patients. As the assessment possibilities are limited, it can in most cases be difficult to dissociate the peripheral from the central manifestations.

Short-lasting (minutes) muscle hyperalgesia can be studied after i.m. injections of capsaicin and glutamate or combined i.m. injections of serotonin (5-HT) and bradykinin. Ernberg et al. found that i.m. co-injection of 5-HT and the 5-HT(3) receptor antagonist granisetron reduced the spontaneous pain evoked by injection of 5-HT and prevented allodynia/hyperalgesia to mechanical pressure stimuli. Thus, peripheral serotonergic receptors could be involved in the regulation of musculoskeletal pain disorders. It is also known that the muscular glutamate receptor is involved in pain, as the glutamate-evoked pain can be reduced by co-administration of an NMDA-antagonist .

Short-lasting (days) muscle hyperalgesia can be induced if a muscle is eccentrically exercised, and it will become sore after 1–2 days. This model of delayed-onset muscle soreness shows hyperalgesia to pressure, and has been used as a model of muscle hyperalgesia in patients . Interestingly, the increased sensitivity to pressure was detected at different sites among subjects , and these localised spots of hyperalgesia might link to the initial mechanisms for hyperalgesic trigger points in myofascial pain patients.

Long-lasting (week) muscle hyperalgesia can be induced in healthy subjects by i.m. injection of NGF . This hyperalgesia might, however, also be related to central sensitisation .

Clinical studies

Muscle damage, injury and inflammation of deep tissue are typical examples of conditions with a likely contribution of peripheral sensitisation. This phenomenon is potentially also involved in the manifestation of trigger points in myofascial pain patients . Increased sensitivity of trigger points has often been assessed by pressure or electrical i.m. stimulation . The continued presence of hyperalgesia (temporal summation) and the presence of multiple hyperalgesic points (spatial summation) might sensitise the spinal cord and supraspinal structures by continued nociceptive afferent barrage to the central nervous system.

OA pain and management as an example

As an example, the lack of pain management efficacy (maximal 20–25% alleviation of initial pain) of current medications in OA indicates that the mechanisms involved are not responsive to the therapies, and alternative approaches may be needed.

As evident from the previous sections, peripheral and central pain sensitisation in OA seems to be a very prominent feature. A hallmark in OA is progressive degeneration of the articular cartilage and subsequent joint space narrowing. In the majority of patients, the aetiology of OA is not known. Among the known risk factors of OA are age, significant traumata, obesity, altered gait, altered biomechanics (e.g., varus or valgus deformity) and excessive loading.

Experimental and clinical observations suggest that the structural integrity of articular cartilage is dependent on normal subchondral bone turnover, intact chondrocyte function and ordinary biomechanical stresses . An increasing line of evidence suggests that there are strong interrelationships between the subchondral bone and the articular cartilage . As bone and cartilage are closely interrelated, interventions affecting pain related to bone turnover may, in addition, be related to the osteoarthritis relevant pain.

Many and diverse structural features of the joint have been implicated in the pain associated with OA, including but not limited to presence of osteophytes in the patellofemoral compartment, focal or diffuse cartilaginous abnormalities, subchondral cysts, bone marrow oedema, subluxation of the meniscus, meniscal tears and Baker cysts. However, pain perception is often highly individual, and very few features correlate with pain robustly.

Normal adult articular cartilage is thought to be vascular and neural. However, very recently, sympathetic and sensory nerves were both identified present within vascular channels in the articular cartilage in both mild and severe OA. Perivascular and free nerve fibres and nerve trunks were observed within the subchondral bone marrow and within the marrow cavities of osteophytes. Nerve endings localised in damaged human articular cartilage suggest that vascularisation and the associated innervation of articular cartilage may contribute to tibiofemoral pain in OA across a wide range of structural disease severity. These findings imply that the musculoskeletal pain associated with osteoarthritis may result from a combination of the heretofore-accepted central and bone-derived effects.

Current recommendations for symptomatic treatment of OA list paracetamol (4 g per day) as the first-line analgesic. However, in a recent survey of patients with OA, only a third of the patients found paracetamol effective. Over 60% of the patients responded better to non-steroidal anti-inflammatory drugs (NSAIDs) . Current recommendations suggest that long-term use of NSAIDs should be avoided for management of OA pain. It is obvious that current OA management regimes do not at all focus on possible peripheral and central manifestation of this chronic pain condition. One reason for this is the lack of understanding of the underlying causes of pain in OA.

A systematic review and meta-analysis of randomised placebo-controlled trials has recently been published to estimate the analgesic efficacy of NSAIDs, including selective cyclo-oxygenase-2 (COX-2) inhibitors (coxibs), in patients with OA in the knee . The study concluded that NSAIDs reduced short-term pain in OA of the knee slightly better than placebo. However, the analysis did not support long-term use of NSAIDs for this condition, as serious adverse effects can be associated with oral NSAIDs. Another recent meta-analysis study evaluated randomised controlled trials of short duration only (less than 4 weeks) assessing the efficacy of topical NSAIDs in OA . After 2 weeks, there was no evidence of efficacy superior to placebo. No trial data support the long-term use of topical NSAIDs in OA.

Several ongoing clinical trials are currently exploring the role of calcitonin in OA-related pain, and the data will be interesting to follow as calcitonin may interact with some of the central manifestations related to the central sensitisation. Calcitonin is found to significantly affect cartilage-specific collagen type II fragments, as a marker of cartilage degradation . Preliminary evidence of oral calcitonin in patients with OA showed a significant effect on pain and joint function . A series of ongoing OA studies may better characterise the fundamental and intrinsic effect, and the mode of action of calcitonin analgesia in OA.

General management regimes and new compounds in the pipeline

The management regimes for rheumatoid arthritis and OA have changed dramatically in recent years. Early referral for specialist opinion is recommended for any patient with rheumatoid arthritis or suspected synovitis of undetermined cause . Effective communication and education is vital, together with ready access to a multidisciplinary team. The early use of either conventional disease-modifying anti-rheumatic drug (DMARD) therapy or biologic DMARD therapy offers the prospect of very significant disease control, and in the case of biologic therapy, disease remission .

Current opinion favours the use of a combination of conventional or biologic DMARDs, ideally within the first 3 months of diagnosis; this should include methotrexate and at least one other DMARD, plus short-term glucocorticoids . In some patients, the very rapid reduction in pain scores following introduction of anti-tumour necrosis factor agents suggests a direct analgesic effect over and above more general anti-inflammatory effects, although this remains unproven .

The general principles for the use of analgesic agents in rheumatoid arthritis and other inflammatory arthropathies are the same as for other musculoskeletal disorders. Acetaminophen (paracetamol) plus codeine combinations are widely used, although objective evidence for efficacy is limited by the paucity of clinical trials (Number To Treat = 2.2). Adverse events limit widespread applicability, although titration of the dose to effect is useful in overcoming these problems . The long-term use of stronger opioids in chronic musculoskeletal conditions remains controversial , but recent developments in oral and transdermal sustained release formulations have increased both the safety and utility of these agents.

NSAIDs continue to be used widely for symptomatic therapy of rheumatoid arthritis , although their long-term use has been reduced in consequence of increased awareness of significant toxicity issues. Gastrointestinal events, including perforation, ulceration and bleeding, are well documented . Other well-recognised problems include oedema and renal insufficiency, with the development of coxibs (COX-2 inhibitors) serving to highlight the additional cardiovascular risks associated with these agents. Other clinically useful strategies include combining NSAIDs with tramadol or NSAIDs with weak opioids, although there are surprisingly few adequately designed randomised controlled trials to provide objective support for these approaches.

The role of antidepressants in relieving pain and depression in rheumatoid arthritis is not clear, although some studies suggest an effect on both pain and morning stiffness . For the most part, these agents remain useful as adjuvant therapy and are not considered front-line analgesic agents in most musculoskeletal disorders.

As is the case for rheumatoid arthritis, clinical guidelines for the management of OA stress the importance of both pharmacological and non-pharmacological approaches . Patient education and self-management programmes reduce overall pain scores and improve general well-being . Weight loss and adherence to an exercise programme also play a vital role in symptomatic knee OA . In general, we say that 5-kg weight loss is equivalent to the effect of one NSAID dose. Other interventions shown to be of benefit include knee taping in patients with patellofemoral OA , as well as wedged insole and shoe orthoses . Transcutaneous nerve stimulation may play a role in some individuals .

For many years, oral analgesics, such as acetaminophen and NSAIDs, have been the mainstay treatment for symptomatic OA. To an even greater extent than in rheumatoid arthritis, the increased awareness of NSAID toxicity has seen a move away from the long-term use of these agents in symptomatic OA. Combination therapies as discussed above are used widely. Stronger opioids, such as transdermal fentanyl, have been shown to be effective in reducing pain scores and improving function in patients with knee and hip OA . Overall, however, the relative lack of efficacy and the many safety issues surrounding both NSAIDs and opioids have led the pharmaceutical industry to invest massively in finding new compounds for managing, for example, OA pain, and promising results were published using tanezumab, which is a humanised anti-NGF monoclonal antibody, in OA pain . The OA trial on tanezumab was stopped in 2010 the Food and Drug Administration (FDA) stopped the OA trial on tanezumab due to side effects such as paraesthesia and expected subchondral changes.

Recently, an oral, highly selective p38α inhibitor (ARRY-797), which decreases production of prostaglandin E2, as well as cytokines, such as tumour necrosis factor and interleukin-1, were shown to be effective in ankylosing spondylitis .

The elevation of fatty acid amides, such as anandamide, has been associated with analgesic efficacy in rodent models. A recent experimental compound (PF-04457845) has been shown to be an irreversible inhibitor of fatty acid amide hydrolase (FAAH) in rodents, healthy human volunteers and patients, but it did not show efficacy in OA .

TRPV1 antagonists have recently been suggested as a new possible treatment modality for OA, and experimental TRPV1 antagonists have shown analgesic effect but unfortunately also increased body temperature .

Review articles on the most recent clinical trials have been published by Berenbaum and Graverand-Gastineau .

Other pharmacological approaches include the use of topical therapies such as NSAID creams and gels, which have a much better safety record (adverse events <1.5%) compared with oral administration. Systemic reviews also support the use of topically applied capsaicin, with a limited number of trials reporting benefit in OA . Approximately one-third of patients report local adverse events with topical capsaicin, usually burning discomfort at the site of application.

Intra-articular steroid injections are widely used to control symptoms in OA, although the duration of symptom relief may be relatively short with effects lasting only a few weeks . Intra-articular hyaluronic acid (hyluronan), which is a high-molecular-weight polysaccharide, has recently shown no effect in OA . Glucosamine and chrondroitin sulphate have enjoyed striking popularity for the treatment of OA and enjoyed favourable early reports, but a recent, large-scale trial failed to show benefit over placebo .