pathophysiology ofjoint pain ap · taglandins e2, d2, andi2, mayactivate articu-lar sensory fibres...

Annals of the Rheumatic Diseases 1996; 55: 276-283

REVIEW: PAIN Series editor: Bruce L Kidd

Pathophysiology of joint pain

Bruce L Kidd, Vanessa H Morris, Laszlo Urban

The diversity of symptoms experienced bypatients with musculoskeletal pain is in starkcontrast to the relative paucity of effectivetreatment. The intensity and character ofsymptoms vary considerably and may beindependent of the underlying disease. In somepatients, pain may occur in the absence ofdemonstrable pathology, whereas in othersobviously damaged joints may be relativelysymptom free.The past few decades have seen a move away

from the concept of a 'fixed' pain pathwaytowards acceptance of a more flexible systemcapable of a wide range of responses to variedstimuli. It is now appreciated that tissue injuryand inflammation produce prolonged changesin the function and activity of the nervoussystem. This review will focus first onperipheral events leading to sensitisation andabnormal excitation of sensory nerves, andsecond on central mechanisms resulting inenhanced transmission in the dorsal horn ofthe spinal cord and elsewhere. While manyaspects of clinical pain remain unclear, therelevance of these mechanisms to symptomsarising as a consequence of musculoskeletaldisease will be explored.

InflammationResearch Group,St Bartholomew's andLondon HospitalSchool ofMedicine,London El 2AD,United KingdomB L KiddV H MorrisSandoz Institute forMedical Research,Gower Street,London WC1E 6BN,United KingdomL UrbanCorrespondence to:Dr Bruce Kidd,Inflammation ResearchGroup,St Bartholomew's andLondon Hospital School ofMedicine,Turner Street,London El 2AD,United Kingdom.Accepted for publication12 January 1996

Peripheral sensitisationPain sensations arising from inflamed or

damaged joints are mediated in the firstinstance by primary sensory (afferent) fibreslinking peripheral receptors with second orderneurones in the spinal cord. Traditionally,these fibres have been regarded as havingrelatively fixed or static properties, and theexistence of specialised 'nociceptors' thatdetect tissue damage or potentially injuriousstimuli was first postulated by Sherrington in1900.1 Single fibre recordings have sinceconfirmed classes of afferent fibre withreceptors that respond preferentially to highintensity stimuli and generate pain whenstimulated.2

It is now reasonably clear that afferent fibres,far from being static as first assumed, are

inherently dynamic structures with an ability tovary responses to a broad range of stimuli,including noxious stimuli.3 There appears to bea critical interaction with the surroundingchemical environment whereby afferent fibresenhance or diminish their capacity to detectand respond to various stimuli. This is wellillustrated in the joint, where inflammatorymediators elicit profound changes in theresponse properties of afferents to movementand other chemical mediators.4

JOINT INNERVATION

Afferent fibres within peripheral nerves fallwithin three distinct groups: heavily myelin-ated AP (group II) fibres, thinly myelinated Ab(group III) fibres and unmyelinated C (groupIV) fibres.5 Synovial joints are innervated withall three groups, in addition to unmyelinatedsympathetic postganglionic fibres.6 It is notablethat the vast majority of articular fibres areunmyelinated, comprising sensory C fibres andsympathetic efferent fibres in nearly equalnumbers.7 Only 20% of the fibres in a typicalarticular nerve are myelinated, mainlycomprising AB fibres, with relatively few of thelarger diameter AP3 fibres.8The articular capsule of the joint receives an

extensive network of nerve fibres with free,complex, or encapsulated nerve endings. Asimilar innervation has been found in tendons,ligaments, deep fascia, and periosteum. Morerecent immunohistochemical studies haveshown that normal human synovium is alsorichly supplied with nerves.9 '0 These includefibres containing substance P and calcitoningene related peptide (CGRP), which areconsidered markers of small calibre sensoryfibres, in addition to nerves containingneuropeptide Y and its C flanking peptide,found in the majority of sympatheticneurones. I0

ARTICULAR NOCICEPTORSFollowing a comprehensive series of studiesperformed largely in the cat knee,4 Schaibleand Schmidt have identified functional groupsof articular afferent fibres on the basis of theirresponse to mechanical stimuli." The firstgroup includes low threshold afferents acti-vated by innocuous stimuli such as movementwithin the normal range, which presumablyhave a role in proprioception; the majority ofrapidly conducting AP fibres fall into this cat-egory. The second group includes afferentsactivated mainly or purely by noxious stimulisuch as movement exceeding the normal range;these are AB and C fibres and, in so far as theyrespond mainly to potentially damaging eventswithin the joints, may be classed as noci-ceptors. Finally, some fibres do not react to anymechanical stimulus applied to the normaljoint whatsoever; these have been termedmechanoinsensitive afferents or silent nocicep-tors,12 and it is estimated that in the cat kneeabout one third ofC fibres and a small percent-age ofAB fibres fall into this category.4

Several hours after induction of ex-perimental arthritis, the response pattern of

276

on June 24, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.55.5.276 on 1 M

ay 1996. Dow

nloaded from

http://ard.bmj.com/

Pathophysiology ofjoint pain

articular sensory fibres changes dramatically.'3The majority become 'sensitised' and show anincreased responsiveness to stimuli applied tothe joint. First, the high threshold sensoryfibres become sensitised to movements in thenormal range and may have resting activityin the absence of mechanical stimulation.Second, a proportion of the initially mech-anoinsensitive sensory fibres develop re-sponsiveness to mechanical stimuli. It isrelevant that the time course of these changesin articular fibres mirrors the development ofpain behaviour in awake animals.4

INFIAMMATORY MEDIATORS

Many, if not all, of the chemical mediatorsfound in inflammatory joint fluid have potentand complex effects on afferent fibres, someacting to sensitise their receptors to mechanicaland other stimuli, while others activatereceptors directly. In either case, control of thereceptor depends ultimately on the effects ofthese mediators on membrane ion channels,which may be directly coupled to membranereceptors (receptor gated ion channels) orcontrolled indirectly through intracellularsecond messengers.'4Prostaglandins-Prostanoids, particularly pros-taglandins E2, D2, and I2, may activate articu-lar sensory fibres directly, but more usuallythey sensitise fibres to mechanical stimuli andchemicals such as bradykinin4 (fig 1). Intra-articular injections of PGE2 both sensitise andactivate more than 50% of the AB and C fibresin the cat knee, but have relatively little effectin the rat knee, in which PGI2 seems to be moreimportant. 15

Prostaglandin induced excitation of sensorymembrane receptors leads to a sequence ofintracellular events culminating in increasedsensitivity which probably involves a stim-ulatory G protein and the cAMP second mes-senger system.16 It has long been assumed thatthe principal site of analgesic activity of non-steroidal anti-inflammatory drugs (NSAIDs) is

prostaglandins opioids sensitisation

neurogenic <

inflammation

Na+

c2+CaFigure 1 Sensitisation and activation ofprimary afferent Cfibres. Prostaglandinssensitise Cfibres, probably via a stimulatory G protein and the cAMP second messengerpathway. Opioids induce formation of inhibitory G proteins and act to inhibit this pathway.Bradykinin directly activates the Cfibre by generation ofpyruvate kinase C (PKC)activity and increasing sodium conductance. Heat and mechanical stimuli also increase ionconductance to produce activation, though exact mechanisms remain unclear. Release ofneuropeptides secondary to changes in intracellular calcium concentration stimulatesneurogenic inflammation. DAG = Diacylglycero4- PLC = phospholipase C; SP= substanceP; CGRP= calcitonin gene related peptide; NKA = neurokinin A; IP3 = 1,4,5-inositoltriphosphate.

in the periphery, where they reduce the syn-thesis of prostaglandins by inhibitingcyclo-oxygenase activity. By reducing prosta-glandins, it is reasoned, NSAIDs prevent sensi-tisation of articular sensory fibres and therebyinhibit or reduce activation by mechanical andchemical stimuli. While this serves to explainthe all purpose analgesic properties ofNSAIDs, other more central analgesic actionsfor these drugs have been proposed in recentyears. 'Bradykinin-When injected into human skin,bradykinin causes an intense sensation of acutepain. Its intra-articular application to the catknee directly activates a large proportion ofgroup III and IV fibres, in addition tosensitising a population of fibres to move-ment.'8 Although the effects of bradykinin aremediated primarily by bradykinin B2 receptorsduring acute inflammation, B1 receptors areexpressed during chronic disease and maybecome more important with time.'9 20 Ineither case, binding of the receptor is fol-lowed by G protein mediated activation ofphospholipase C (PLC) which generatestwo intracellular second messengers- 1,4,5-inositol-triphosphate (IP3) and diacyl-glycerol (DAG).2' IP3 stimulates an increase infree calcium within the cell, whereas DAGactivates protein kinase C, which plays a partin fibre activation by increasing membraneconductance, mainly to sodium ions. It shouldbe stressed that other complex pathways havebeen described and the actions of bradykininon afferent fibres remain to be fully deter-mined.22Protons-The acidic nature of inflammatoryexudates has led to the proposal that protonsmay contribute to inflammatory pain.23Application of acidic fluid causes pain inexperimental situations,24 and protons havebeen shown to have an excitatory effect onmany neurones, which may be of short orrelatively prolonged duration.25 The neuronalactivation occurs by the opening of unique ionchannels and is distinct from the non-specificeffects of protons on membranes observed inmany cell types.23 The overall contribution ofprotons to pain would be particularly markedwithin inflamed joints, given the hypoxicenvironment and extreme acidosis of in-flammatory synovial fluid.Opioids-Although endogenous opioids tradi-tionally have been thought to act principally atsites within the central nervous system, thereis now compelling evidence to suggestadditional sites of action in the periphery.Opioid receptors have been demonstrated onthe terminals of peripheral nerves26 and localapplications of opioid agonists have beenshown to reduce hyperalgesia in severalexperimental models.27 Stein et al havedemonstrated that hyperalgesia duringadjuvant induced monarthritis can be reducedin a naloxone reversible manner after stressinduced release of endogenous opioids.26Immunocytochemical studies by the samegroup have suggested that opioid peptides areproduced by many of the cell types commonlyfound within inflammatory fluids.27

277


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/

Kidd, Moms, Urban

In clinical studies, intra-articular injectionsof morphine immediately before knee artbro-scopy substantially reduced the postoperativeanalgesic requirement.28 The doses used inthese studies (1 mg) were small enough topreclude effective systemic spread and providefurther evidence for a local opioid effect. In theperiphery, opioids are believed to act via aninhibitory G protein to reduce cAMP, thusproviding a theoretical basis for the analgesicsynergy that is observed clinically betweenNSAIDs and opioids such as codeine or mor-phine'6 29 (fig 1).

OTHER INFLUENCES ON ARTICULAR NERVES

Sympathetic nervous system-Over recent years,a number of experimental studies have ex-plored the contribution of the sympatheticnervous system to persistent activation ofprimary sensory fibres and hence chronic pain.The first experimental evidence for inter-actions between sympathetic and sensoryfibres came from observations of neuronalactivity in neuromas.30 In this experimentalmodel, damaged afferent fibres sprout withinthe neuroma and, unlike undamaged afferentfibres, develop sensitivity to noradrenalinemediated by a adrenergic receptors.3' Morerecent models using subtotal resection orligature constriction have shown the expressionof adrenergic receptors on the peripheralterminals ofthose fibres that remain intact.32 Inthe context of chronic arthritis it is relevant thatprimary sensory fibres activated by prolongednoxious stimulation become sensitised tonoradrenaline and sympathetic discharge.33The overall conclusion from these and similarstudies is that, whereas under normal condi-tions primary sensory nerves do not respond tosympathetic activity, the situation undoubtedlychanges after any type of local injury, includingnerve trauma and inflammation. Under thesecircumstances, sensory fibres may bestimulated by concurrent sympathetic activity.Capsaicin-This compound is the activeingredient in hot red peppers, and has a directaction on a population of small sensory fibreswithin the joint and elsewhere. Low dosesactivate small sensory fibres to produce pain,neurogenic inflammation and, at least in skin,surrounding thermal and mechanical hyper-algesia. At these doses, capsaicin opens mem-brane ion channels to sodium and calcium ionsto produce depolarisation.34 The effect appearsto be mediated through a specific receptor andthere is some evidence linking this to the

23proton receptor. Conversely, exposure tolaroe cAncentrntion! of cmnnicin hlckq

sensory neurotoxicity which may be the resultof osmotic lysis or enzyme mediated damagewithin the cell.35 Repeated topical applicationof capsaicin to human skin causes desensitisa-tion of C fibres to noxious stimuli and loss ofneurogenic inflammation;35 these effects aredependent on the concentration of capsaicinand time of exposure, though the underlyingmechanisms remain unclear.

CLINICAL CONSEQUENCESDifferent mediators released either by in-flammatory cells or from the terminals ofsympathetic nerves in response to inflamma-tion or tissue damage produce varying effectson articular receptors and afferent fibres. Thediverse nature of joint innervation ensures that,at least in the early stages of disease, the patternof sensitisation and activation in the peripheryis of pivotal importance in determining thecharacter and magnitude of articularsymptoms. Spontaneous activity in highthreshold afferent fibres may well produce painsensations at rest. Similarly, the developmentof mechanosensitivity in previously insensitiveafferent fibres provides a convincing explana-tion of pain on joint movement-the so-called'incident' pain.Treatment regimens designed to reduce

receptor sensitisation and abnormal afferentfibre activation undoubtedly represent a potentmethod for reducing joint pain. To datethese have depended largely on inhibitionof prostaglandin synthesis, but while thisapproach has proved successful, it is notwithout considerable morbidity. As Perkinsand Dray will report later in this series of painreviews, alternative approaches based on aknowledge of the pathophysiological changesthat underlie joint pain have enabled thedevelopment of new analgesic treatmentsincluding bradykinin and cyclo-oxygenase IIinhibitors, in addition to the use of capsaicinanalogues to reduce small sensory fibre activityselectively.36

Neurogenic inflammationBasic clinical observation indicates that sen-sory nerves not only signal potential oractual tissue damage, but also have an activerole in inflammation. Stimulation of un-myelinated sensory fibres produces a localinflammatory response (neurogenic inflamma-tion) characterised in the skin by the familiarweal and flare reaction.37 An 'axon reflex' hasbeen proposed whereby activation of sensoryfibres after tissue injury results not only inimpulse transmission to the central nervoussystem, but also in reverse transmissionthrough the extensive network of peripheralnerve fibres known to terminate in close proxi-mity to blood vessels and mast cells.38Neurogenic inflammation is now understood

to be mediated through biologically activepeptides.38 These are synthesised within smallto medium sized dorsal root ganglion cellbodies and are transported via unmyelinatedsensory fibres to peripheral tissues and toqvn2ntir terminniq within the inPerfiAinl

One of the commonest peptides is substance P,but others including neurokinin A, CGRP,vasoactive intestinal polypeptide, and somato-statin have been reported.Neurogenic vasodilatation-A number of clinicalstudies have provided evidence for alteredneurogenic inflammatory responses in mus-culoskeletal disease. Capsaicin, as describedearlier has a selective action on small diameter

278


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/


unmyelinated afferent fibres to produce painand axon reflex vasodilatation. Although usedroutinely for a number of years to investigateperipheral sensory function in neuropathic dis-orders, this property has only recently beenused to study musculoskeletal disorders.Helme and colleagues have shown that cap-saicin induced skin flares in patients with spinalpain secondary to degenerative disease are sig-nificantly smaller than those with pain of non-organic origin.40 In contrast, a second study bythe same group reported increased skin flaresin patients with chronic pain syndromescharacterised by muscle tender points.4' Amore recent study assessing sensory and auto-nomic axon reflex responses in rheumatoidarthritis has shown capsaicin induced axon

over the forearms) of patients with rheumatoidarthritis, compared with responses in agematched normal subjects.42 The results suggestselective upregulation of sensory axon reflexactivity in patients with rheumatoid arthritis,but whether this occurs as a result of localmediators or changes in central controlremains unclear.Substance P-This is an 11 amino acid peptidebelonging to the tachykinin family that share acommon sequence of six amino acids at thecarboxyl terminal.43 A number of studies haveshown increased mRNA concentrations ofpreprotachykinin A, the precursor for sub-stance P, in dorsal root ganglia after induction-.c 44 45 _ .. :

ganglia and sciatic nerve have been shown toincrease by 40 to 70% after induction ofexperimental adjuvant induced arthritis.4445Somewhat paradoxically, concentrations ofsubstance P in inflamed synovium have beenreported to be reduced during the acutestages of short term inflammation models.44This has been taken to reflect rapid deg-radation of peptides by degradative enzymesknown to be present in high concentrationswithin inflamed tissues. The same explanationmay also serve to explain the wide variationin concentrations of substance P reportedfrom synovial fluid removed from arthriticjoints.46Within the joint antidromic (reversed)

stimulation of unmyelinated articular nervescauses plasma extravasation.47 This effectappears to be mediated by substance P, asit is completely blocked by prior intra-articular administration of a substance Pantagonist. Intra-articular injections of sub-stance P produce vasodilatation andplasma extravasation,48 49 while in vitrostudies show mast cell degranulation, secre-tion of PGE2 and collagenase from synovio-cytes, and stimulation of interleukin- 1(IL-l)-like activity from macrophages.29 Arole in activation of the immune system hasalso been suggested. In studies of acute jointinflammation, substance P antagonists sig-nificantly inhibit the early articularresponses to inflammatory agents such ascarrageenin.50

Nerve growth factor-This is a major regulatorof substance P gene expression in adult sensorynerves,5' and may have an important rolein the maintenance of inflammatory pain.52Nerve growth factor (NGF) is a memberof the neurotrophin family of peptides and isproduced by a range of cell types inresponse to a number of different cyto-kines, including IL-I B and tumour necro-sis factor cx.53 Concentrations are increasedduring inflammation,54 and NGF has beenidentified in synovial fluid of patientswith chronic arthritis.55 In the nervoussystem, NGF binds to a high affinity trkAreceptor, where it is intemalised and retro-gradely transported to the cell body. Systemicadministration ofNGF neutralising antibodies

behavioural sensitivity, and development oflong term changes within the spinal cord.53 56Disease outcome-A number of experimentalstudies have examined the contribution ofspecific neural components to the expressionand outcome of chronic arthritis. In 1983,Colpaert et al reported that capsaicin treatmentof adult rats, using doses sufficient to depletesubstance P stores, markedly reduced pawswelling and prevented weight loss duringadjuvant arthritis.57 This beneficial effectoccurred irrespective of whether capsaicin wasgiven before or after the onset of inflammation.More recent studies by Levine et al haveconfirmed and extended these observations by

neurogenic influence in the production ofsymmetrical joint disease has been suggestedfollowing the observation that inflammationor trauma in one joint can result in aneurogenically mediated response in thecontralateral joint." 6Although neurogenic inflammation has long

been thought to be a purely peripheralphenomenon, recent studies suggest that itmay be under central control.62 63 Spinallyadministered sensory receptor antagonistssignificantly reduce joint swelling in kaolin andcarrageenan induced arthritis administeredbefore62 or after the onset of disease.63 A similareffect has been demonstrated after dorsalrhizotomy.64

Central sensitisationSensitisation of primary afferent fibres allowsneurogenic responses to be extensively modi-fied according to the requirements of localtissues. A similar process also occurs in spinalneurones (central sensitisation), where it mayhave far reaching consequences for symptomsand signs arising not only from damaged orinflamed tissues, but also from tissues notdirectly involved in the primary pathology. Themost obvious musculoskeletal examples arereferred pain and the observation that appar-ently normal tissues around inflamed ordamaged joints are often painful and tenderdespite having no apparent involvement in theunderlying disease.

279


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/

Kidd, Morris, Urban

HYPERALGESIA AND REFERRED PAINHyperalgesia is characterised by a reduction ofthe threshold for pain and an increasedsensation of pain following a noxious stimulus.Damage to a joint, an area of skin, or othertissue produces tenderness and pain not onlyat the site of injury (primary hyperalgesia), butalso in the surrounding area (secondary hyper-algesia).65 Although Lewis first postulated inthe 1940s that secondary hyperalgesia wasproduced by sensitisation of peripheralneurones,65 later work by others arguedstrongly against this.6"8 Studies performedprimarily in human skin showed that, while aninitial noxious stimulus from the periphery isrequired for induction of hyperalgesia andpain, continuing abnormal inputs from theperiphery are not required for mainten-ance.67-69 The observations suggested theexistence of central mechanisms acting toenhance and modify the responsiveness of thespinal cord to both normal and abnormal inputfrom the periphery.The classical studies by Lewis and Kell-

gren,70 and later by Hardy et a166 showed thatlocal injections of hypertonic saline into inter-spinous ligaments produced hyperalgesia andreferred pain in the abdomen and elsewhere.These effects often occur in areas not sharingthe same dermatome and may spread to sitesof previous injury or disease;66 71 knee painreferred from an osteoarthritic hip is anobvious example. The fact that hyperalgesiaand pain spread to areas far removed from theinjured region argues strongly against a localprocess and has provided further evidence fora central mechanism.70 71

,,10.. Pathological

C

4..: (FD.B /1*,7

NMDA \NK

C fibreMA, SP)

J/Ii,E|

nociception hyperalgesiaallodynia

Figure 2 Activation ofdorsal horn neurones by selective sensory input. NORMALCONDITIONS: Myelinatedfibres are activated by innocuous stimuli (touch, light pressureetc) and release excitatory amino acids (EAA) in the spinal cord which excite non-NMDAreceptors (A). In contrast, Cfibres are activated by noxious stimuli (pinch, heavy pressureetc) and release EAAs and neurokinins (substance P (SP)) which excites neurokinin(NK), NMDA, and non-NMDA receptors (B), resulting in painful sensations.PATHOLOGICAL CONDITIONS (eg inflammation): Cfibres are spontaneously active andmaintain neurokinin andNMDA receptor activation to produce sustained hyperexcitabilityand consequent primary hyperalgesia. Because of the sustained background spinalhypersensitivity, inputsfrom other unmyelinated Cfibres or large myelinatedfibres may alsoproduce depolansation producing secondary hyperalgesia and allodymia, respectively.(Activefibres are represented by solhd lines, inactivefibres by dotted lines, respectively.Filled shapes represent activated receptors on dorsal horn cells; empty ones are inactive.)

WIND UP AND SPINAL HYPEREXCITABILITYA major experimental breakthrough in thediscovery of a central component to pain wasthe observation of C fibre dependent increasesof excitability in dorsal horn neurones. Theterm 'wind up' was coined by Mendell andWall to describe changes of activity in dorsalhorn neurones whereby sustained or repetitiveinputs from C fibres lead to a progressiveincrease in the discharge produced by furtherstimuli.72 Subsequently, Woolf showed thatacute peripheral injury produced prolongedchanges in spinal cord excitability thatoutlasted the duration of the noxiousstimulus.73 Once wind up and central hyper-excitability are established, sensory processingwithin the spinal cord is substantially modifiedand may produce new sensory modalities suchas allodynia, whereby normally innocuousstimuli are perceived as being painful (fig 2).Neurokinins-The most important requirementfor wind up is that it develops only afterrepetitive stimulation of C fibres terminatingprimarily in the superficial layer (lamina I) ofthe dorsal horn. Contrary to expectations,substantia gelatinosa (lamina II) neurones withC fibre input do not develop wind up.74Interestingly, these latter neurones do notexpress neurokinin receptors and it is nowaccepted that wind up occurs only in cells thatexpress neurokinin receptors and areinnervated by substance P containing afferentC fibres.75The principal neurotransmitters within the

spinal cord are the excitatory amino acids,which include glutamine and aspartate. Sincethe discovery that neurokinins (for examplesubstance P, neurokinin A) and excitatoryamino acids coexist in afferent C fibres,76 ithas been suggested that these transmittersmay interact at postsynaptic neurones. Thereis much evidence that substance P hasmodulatory functions in the spinal dorsalhorn,75 77 78 and it is generally agreed thatsustained activation of high threshold C fibresis required for these functions. Duggan et alhave shown that substance P is released inincreased quantitites during periods of C fibreactivation,79 and neurokinin-1 receptor antag-onists have proved to be blockers of spinalhyperexcitability evoked by intradermal cap-saicin or by acute arthritis in the primate.80 81Consistent with these observations, neuro-kinin- 1 receptor antagonists proved to beanalgesics in animal models of inflammatorypain.82The NMDA receptor-At the heart of wind upand central sensitisation lies the N-methyl-D-aspartate (NMDA) receptor. This receptorbelongs to the family of ionotropic glutamatereceptors that are coupled to ion channels andare activated by endogenous excitatory aminoacids (fig 3). Their activation evokes longlasting membrane depolarisation and en-hanced membrane excitability. Blocking theNMDA receptor blocks the developmentof wind up after electrical stimulation83and hyperactivity of spinal neurones afterinjection of formalin84 or acute inflammationof the knee joint.85 In keeping with this, it has

Normalmy.lnated

primary afferent(EAA)

I A.4nonproceDA

proprioception

280


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/


3pIUotin cUpJIeU NMDA receptor ion channelreceptor

,<X1>OG

phosphory-lation Ca2+

Activation of second messenger systemsModulation of protein synthesis

Figure 3 Interaction between neurokinin-1 (NK-1) andNMDA receptors. Activationofa G protein coupled receptor such as NK1 modulates magnesium ion binding ofNMDAreceptors through protein kinase C (PKC). Magnesium ions are releasedfrom thephosphorylatedNMDA receptor and enable binding of excitatory amino acids. Activationof the NMDA receptor complex produces accumulation of intracellular calcium, with furtherconsequences.

been shown that sensitisation of spinalneurones during acute arthritis is criticallydependent on activation of NMDAreceptors.86 87

Cellular mechanisms-NMDA receptors are

activated by glutamate and aspartate, whichare released in abundance within the spinalcord from both myelinated and unmyelinatedprimary afferent fibres and also from dorsalhorn neurones. It is notable, however, thatrelease of these neurotransmitters from non-

neurokinin-containing A fibres does not evokewind up in the dorsal horn.88 The crucial roleof neurokinins in modulation of NMDAreceptor/ion channel properties has beenstudied by several groups.75 80 89 It is est-ablished that neurokinin receptor activation bysubstance P and neurokinin A receptorenhances the activity ofNMDA receptors bothon single dorsal horn cells89 and in isolatedspinal cord.78 The interaction takes placethrough the activation of protein kinase C75 89which can phosphorylate the NMDA receptorand change its Mg2" binding kinetics. Undernormal circumstances Mg2" binding blocksthe NMDA receptor, but the alteration ofMg2" binding kinetics allows the release ofMg2" from the receptor and permits gluta-mate induced activation and subsequentdepolarisation of the cell membrane75 (fig 3).

OTHER MEDIATORS

A number of endogenous mediators, includingprostaglandins, nitric oxide, opioids, andadrenergic agonists, have been shown toinfluence the excitability of spinal neurones.7'

It has been suggested that prostaglandins not

only contribute to the peripheral component ofhyperalgesia, but also have an importantcentral role.'7 90 91 The basis of the argumentis that, in rats, intrathecal NSAIDs reduce painrelated behaviour evoked by the spinal actionof substance P and NMDA90 or injection of

formalin into the hindpaw.9' Glutamate releaseevoked by injection of formalin into thehindpaw is blocked by NSAIDs applied eitherintrathecally or systemically, with the potencyof intrathecal application being two orders ofmagnitudes greater.92 It is also suggested thatspinal NSAIDs act primarily through blockadeof cyclo-oxygenase as the inhibitory effect isstereospecific, such that S(+)-ibuprofen, whichblocks cyclo-oxygenase, is active in theexperimental model, while R(-)-ibuprofen(with no cyclo-oxygenase effect) is not. Takentogether, the findings imply that NSAIDs havea direct antinociceptive effect in the spinal cordor at supraspinal level. It must be stressed,however, that the issue remains contro-versial.93 94

In addition, several classes of receptorselective agents act spinally to alter nociceptiveprocessing. Alpha2 adrenergic and pu opioidreceptor agonists produce analgesia by pre-synaptic inhibition of C fibre neurotransmitterrelease and postsynaptic hyperpolarisation ofsecond order neurones.7" Coadministration ofintrathecal morphine and selected a2 agonistsor NSAIDs results in substantial analgesicsynergy" and highlights a role for combinationtreatments in clinical settings.

CLINICAL CONSEQUENCESUnderstanding of the precise mechanism ofreceptor interaction in spinal neurones is farfrom complete; however, its importance in thedevelopment and maintenance of spinalhyperexcitability is now generally agreed.75 95-98The current model provides an explanation foraltered sensations arising from peripheralinjury or similar pathological stimuli. Once Cfibre activity 'unmasks' receptors in spinalneurones, neurotransmitters released fromboth C and A fibres may produce increasedNMDA receptor activation and maintaincentral hyperexcitability. Under these condi-tions, both C and A fibres become able toevoke pain (fig 3).

Dorsal horn neurones receive inputs fromvarious afferent fibres with receptive fields thatare not necessarily restricted to one area oreven to the same tissue.97 As an example of this,dorsal horn neurones with receptive fields inthe knee also receive inputs from adjacentstructures including skin and muscle, and fromremotes sites near the ankle and even thecontralateral limb.99 l'o Under normal cir-cumstances these inputs remain suppressed.The development of central hyperexcitability,however, allows dorsal horn neurones torespond in an abnormal and exaggerated way.96The net effect is that responses to normalstimuli are increased, the size of the receptivefields is expanded, and the threshold foractivation by novel inputs is reduced.96Clinically, this results in enhanced pain at thesite of injury (primary hyperalgesia) and devel-opment of pain and tenderness in normaltissues both adjacent to (secondary hyper-algesia) and removed from (referred pain) theprimary site.98 Innocuous mechanical stimulisuch as light pressure on the skin may produce

281

r"- r% re%,;n ~%sI r% lor


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/

Kidd, Moms, Urban

pain as a result of input from mechanoceptiveA fibres (allodynia).

Increasing awareness of central mechanismsinvolved in pain has stimulated the search forand development of new analgesic treatmentregimens. In particular, attention is beingfocused on neurokinin and NMDA receptorinhibitors, though other targets are also underinvestigation.36 It is to be hoped that novelagents delivered alone or in combination withperipherally acting drugs will ultimatelyprovide safer and more effective analgesia foruse in that large group of long sufferingpatients with musculoskeletal disease.

1 McMahon S B, Koltzenburg M. Novel classes ofnociceptors: beyond Sherrington. Trends Neurosci 1990;13: 199-201.

2 Raja S N, Meyer R A, Campbell J N. Peripheralmechanisms of somatic pain. Anesthesiology 1988; 68:571-90.

3 Rang H P, Bevan S, Dray A. Nociceptive peripheralneurones: cellular properties. In: Wall P D, Melzac M,eds. Textbook of pain, 3rd Edn. Edinburgh: ChurchillLivingstone, 1994; 57-78.

4 Schaible H, Grubb B D. Afferent and spinal mechanismsof joint pain. Pain 1993; 55: 5-54.

5 Lynn B. The fibre composition of cutaneous nerves andthe classification and response properties of cutaneousafferents, with particular reference to nociception. PainReviews 1994; 1: 172-83.

6 Freeman M A R, Wyke B. The innervation of the kneejoint. An anatomical and histological study in the cat.Anat 1967; 101: 505-32.

7 Langford L A, Schmidt R F. Afferent and efferent axons

in the medial and posterior articular nerves of the cat.Anat Rec 1983; 206: 71-8.

8 Heppleman B, Heuss C, Schmidt R F. Fibre sizedistribution of myelinated and unmyelinated axons in themedial and posterior articular nerves of the cat's kneejoint. Somatosens Res 1988; 5: 267-75.

9 Gronblad M, Konntinen Y T, Korkala 0, Liesi P,Hukkanen M, Polak J. Neuropeptides in the synovium ofpatients with rheumatoid arthritis and osteoarthritis.Rheumatol 1988; 15: 1807-10.

10 Mapp P I, Kidd B L, Gibson S J, et al. Substance P-,calcitonin gene-related peptide- and C-flanking peptide ofneuropeptide Y-immunoreactive fibres are present in thenormal synovium but depleted in patients withrheumatoid arthritis. Neuroscience 1990; 37: 143-53.

11 Schaible H, Schmidt R F. Responses of fine medialarticular nerve afferents to passive movements of kneejoint. JNeurophysiol 1983; 49: 1118-26.

12 Schaible H, Schmidt R F. Time course of mechano-sensitivity changes in articular afferents during a develop-ing experimental arthritis. Neurophysiol 1988; 60:2180-95.

13 Grigg P, Schaible H, Schmidt R F. Mechanical sensitivityofgroup III and IV afferents from posterior articular nervein normal and inflamed cat knee. JNeurophysiol 1986; 55:635-43.

14 Rang H P, Bevan S, Dray A. Chemical activation ofnociceptive peripheral neurones. Br Med Bull 1991; 47:534-48.

15 Birrell G J, McQueen D S, Iggo A, Coleman R A, GrubbB D. PGI2-induced activation and sensitization ofarticular mechanociceptors. Neurosci Lett 1991; 124: 5-8.

16 Taiwo Y 0, Bjerknes L, Goetzl E, Levine J D. Mediationof primary afferent hyperalgesia by the cAMP secondmessenger system. Neuroscience 1989; 32: 577-80.

17 Malmberg A B, Yaksh T L. Pharmacology of the spinalaction of ketorolac, morphine, ST-91, U50488H, andL-PIA on the formalin test and an isobolographic analysisof the NSAID interaction. Anesthesiology 1993; 79:270-81.

18 Kanaka R, Schaible H, Schmidt R F. Activation of finearticular afferent units by bradykinin. Brain Res 1985;327: 81-90.

19 Davis A J, Perkins M N. Induction of BI receptors in vivoin a model of persistent inflammatory hyperalgesia in therat. Neuropharmacology 1994; 33: 127-33.

20 Cruwys S C, Garrett N E, Perkins M N, Blake D R, KiddB L. The role of bradykinin B, receptors in themaintenance of intra-articular plasma extravasation inchronic antigen-induced arthritis. Br Pharmacol 1994;113: 940-4.

21 Thayer S A, Pemey T M, MillerR J. Regulation ofcalciumhomeostasis in sensory neurones by bradyvinin. Neurosci

1988; 8: 4089-97.22 Dray A, Perkins M. Bradykinin and inflammatory pain.

Trends Neurosci 1993; 13: 99-104.23 Bevan S, Geppetti P. Protons: small stimulants of

capsaicin-sensitive sensory nerves. Trends Neurmsci 1994;17: 509-12.

24 SteenK H, Reeh P W, Anton F, HandwerkerH 0. Protonsselectively induce long lasting excitation and sensitizationto mechanical stimulation of nociceptors in rat skin, invitro. JNeurosci 1992; 12: 86-95.

25 Kessler W, Kirchoff C, Reeh P W, Handwerker H 0.Excitation of cutaneous afferent nerve endings in vitroby a combination of inflammatory mediators andconditioning effect of substance P. Exp Brain Res 1992;91: 467-76.

26 Stein C, Hassan A H S, Prezewlocki R, Gramsch C,Peter K, Herz A. Opioids from immunocytes interactwith receptors on sensory nerves to inhibit nocicep-tion in inflammation. Proc Nad Acad Sci USA 1990; 87:5935-9.

27 Stein C. Peripheral mechanisms of opioid analgesia. AnesthAnalg 1993; 76: 182-91.

28 Stein C, Comisel K, Haimerl E, et al. Analgesic effect ofintra-articular morphine after arthroscopic knee surgery.NEnglJMed 1991; 325: 1123-6.

29 Levine J D, Fields H L, Basbaum A L. Peptides and theprimary afferent nociceptor. J Neurosci 1993; 13:2273-86.

30 Wall P D, Gutnick M. Ongoing activity in peripheralnerves: the physiology and pharmacology of impulsesoriginating from a neuroma. Exp Neurol 1974; 43:580-93.

31 Devor M. Nerve pathophysiology and mechanisms of painin causalgia.J Auton Nerv Syst 1983; 7: 371-84.

32 Bennet G J. The role of the sympathetic nervous systemin painful peripheral neuropathy. Pain 1991; 45: 221-3.

33 Hu S, Zhu J. Sympathetic facilitation of sustaineddischarges of polymodal nociceptors. Pain 1989; 38:85-90.

34 Wood J N, Wmter J, James I F, Rang H P, Yeats J, BevanS J. Capsaicin-induced ion fluxes in dorsal root ganglioncells in culture. J Neurosci 1988; 8: 3208-20.

35 Lynn B. Capsaicin: actions on nociceptive C-fibres andtherapeutic potential. Pain 1990; 41: 61-9.

36 Perkins M, Dray A. Novel pharmacological strategies foranalgesia. Ann Rheum Dis 1996; 55: In press.

37 Bayliss W M. On the origin from the spinal cord of thevasodilator fibres of the hind limb and on the nature ofthese fibres. JPhysiol (Lond) 1901; 26: 173-209.

38 Holzer P. Local effector functions of capsaicin-sensitivesensory nerve endings: involvement of tachykinins,calcitonin gene-related peptide and other neuropeptides.Neuroscience 1988; 24: 739-68.

39 Foreman J C. Peptides and neurogenic inflammation. BrMed Bull 1987; 43: 386-400.

40 LeVasseur S A, Gibson S J, Helme R D. The measurementof capsaicin-sensitive sensory nerve fibre function inelderly patients with pain. Pain 1990; 41: 19-25.

41 Helme R D, Littlejohn G 0, Weinstein C. Neurogenic flareresponses in chronic rheumatic pain syndromes. Clin ExpNeurol 1987; 23: 91-4.

42 Jolliffe V A, Anand P, Kidd B L. Assessment of cutaneoussensory and autonomic axon reflexes in rheumatoidarthritis. Ann Rheum Dis 1995; 54: 251-5.

43 Pemow B. Substance P. Pharmacol Rev 1983; 35: 85-141.44 Donaldson L F, Harmar A J, McQueen D S, Secki J R.

Increased expression of preprotachykinin, calcitoningene-related peptide, but not vasoactive intestinal peptidemessenger RNA in dorsal root ganglia during thedevelopment ofadjuvant monarthritis in the rat. Brain ResMol Brain Res 1992; 16: 143-9.

45 Garrett N E, Kidd B L, Cruwys S C, Tomlinson D R.Changes in preprotachykinin mRNA expression andsubstance P levels in dorsal root ganglia of monoarthriticrats: Comparison with changes in synovial substance Plevels. Brain Res 1995; 675: 203-7.

46 Mapp P I, Kidd B L. Substance P in rheumatic disease.Semin Arthritis Rheum 1994; 6 (suppl 3): 3-9.

47 Ferrell W R, Russell N J W. Extravasation in the kneeinduced by antidromic stimulation of articular C fibreafferents of the anaesthetised cat. J Physiol (Lond) 1986;379: 407-16.

48 Lam F Y, Ferrell W R. Effects of interactions of naturally-occurring neuropeptides on blood flow in the rat kneejoint. BrJ7Pharmacol 1993; 108: 694-9.

49 Cruwys S C, Kidd B L, Mapp P I, Walsh D A, BlakeD R. The effects of calcitonin gene-related peptide onformation of intra-articular oedema by inflammatorymediators. BrJPharmacol 1992; 107: 116-9.

50 Lam F Y, Ferrell W R. Inhibition of carrageenan inducedinflammation in the rat knee joint by substance Pantagonist. Ann Rheum Dis 1989; 48: 928-32.

51 Lindsay R M, Harmar A J. Nerve growth factor regulatesexpression of neuropeptide genes in adult sensoryneurones. Nature 1989; 337: 362-4.

52 Lewin G R, Mendell L M. Nerve growth factor andnociception. Trends Neurosci 1993; 16: 353-8.

53 Woolf C J, Safieh-Garabedian B, Ma Q P, Crilly P,Wimters J. Nerve growth factor contributes to thegeneration of inflammatory sensory hypersensitivity.Neuroscience 1994; 62: 327-31.

54 Donnerer J, Schuligoi R, Stein C. Increased content andtransport of substance P and calcitonin gene-relatedpeptide in sensory nerves innervating inflamed tissue:evidence for a regulatory function of nerve growth factorin vivo. Neuroscience 1992; 49: 693-8.

55 Aloe 1o Turveri M A, Carcassi U, Levi-Montalcini R.Nerve growth factor in synovial fluid of patients withchronic arthritis. Arthritis Rheum 1992; 35: 351-5.

56 McMahon S B, Bennett D L, Priestley J V, Shelton D L.The biological effects of endogenous nerve growth factoron adult sensory neurons revealed by a trkA-IgG fusionmolecule. Nature Medicine 1995; 1: 774-80.

57 Colpaert F C, Donnerer J, Lembeck F. Effects of capsaicinon inflammation and on the substance P content of

282


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/


nervous tissues in rats with adjuvant arthritis. Life Sci1983; 32: 1827-34.

58 Levine J D, Dardick S J, Roizen M F, Helms C.Contribution of sensory afferents and sympatheticefferents to joint injury in experimental arthritis. .7Neurosci 1986; 6: 3423-9.

59 Cruwys S C, Garrett N E, Kidd B L. Sensory denervationwith capsaicin attenuates inflammation and nociceptionin arthritic rats. Neurosci Lett 1995; 193: 205-7.

60 Levine D J, Dardick S J, Roizen M F, Basbaum A I,Scipio E. Reflex neurogenic inflammation. Contributionof the peripheral nervous system to spatially remoteinflammatory responses that follow injury. J Neurosci1985; 5:1380-6.

61 Kidd B L, Cruwys S C, Garrett N E, Mapp P I, JolliffeV A, Blake D R. Neurogenic influences on contralateralresponses during experimental rat monarthritis. Brain Res1995; 688: 72-6.

62 Sluka K A, Westlund K N. Centrally administered non-NMDA but not NMDA receptor antagonists blockperipheral knee joint inflammation. Pain 1993; 55:217-25.

63 Sluka K A, Jordan H H, Westlund K N. Reduction in jointswelling and hyperalgesia following post-treatment with anon-NMDA receptor antagonist. Pain 1994; 59: 95-100.

64 Sluka K A, Lawand N B, Westlund K N. Jointinflammation is reduced by dorsal rhizotomy and not bysympathectomy or spinal cord transection. Ann RheumDis 1994; 53: 309-14.

65 Lewis T. Pain. New York: MacMillan, 1942.66 Hardy J D, Wolff H G, Goodell H. Experimental evidence

on the nature of cutaneous hyperalgesia. Jf Clin Invest1950; 29: 15-140.

67 Baumann T K, Simone D A, Shain C N, LaMotte R H.Neurogenic hyperalgesia: the search for the primarycutaneous afferent fibers that contribute to capsaicin-induced pain and hyperalgesia. Jf Neurophysiol 1991; 66:212-27.

68 LaMotte R H, Lundberg L E, Torebjork H E. Pain,hyperalgesia and activity in nociceptive C units in humansafter intradermal injection of capsaicin. Jf Physiol (Lond)1992; 448: 749-64.

69 Torejork H E, Lundeberg L, LaMotte R H. Centralchanges in processing of mechanoreceptive input incapsaicin-induced secondary hyperalgesia in humans. JPhysiol (Lond) 1992; 448: 765-80.

70 Lewis T, Kellgren J H. Observations relating to referredpain, viscero-motor reflexes and other associatedphenomena. Clin Sci 1939; 4: 47-71.

71 Coderre T J, Katz J, Vaccarino A L, Melzack R.Contribution of central neuroplasticity to pathologicalpain: review of clinical and experimental evidence. Pain1993; 52: 259-85.

72 Mendell L M, Wall P D. Responses of single dorsal cordcells to peripheral cutaneous unmyelinated fibres. Nature1965; 206: 97-9.

73 Woolf C J. Evidence for a central component of postinjurypain hypersensitivity. Nature 1986; 308: 686-8.

74 Yoshimura M, Jessell T M. Primary afferent-evokedsynaptic responses and slow potential generation in ratsubstantia gelatinosa in vitro. Jf Neurophysiol 1989; 62:96-108.

75 Urban L, Thompson S W N, Nagy I, Dray A.Hyperexcitability in the spinal dorsal hom: cooperation ofneuropeptides and excitatory amino acids. In: Urban L,ed. Cellular mechanisms of sensory processing. Berlin:Springer, 1994; 379-99.

76 Debiasi S, Rustioni A. Glutamate and substance P coexistin primary afferent terminals in the superficial laminae ofspinal cord. Prod NadtAcad Sci USA 1988; 85: 7820-4.

77 Coderre T J, Melzack R. The contribution of excitatoryamino acids to central sensitization and persistentnociception after formalin-induced tissue injury. JNeurosci 1992; 40: 3665-70.

78 Urban L, Naeem S, Patel I A, Dray A. Tachykinin inducedregulation of excitatory amino acid responses in the ratspinal cord in vitro. Neurosci Lett 1994; 168: 185-8.

79 Duggan A W, Hendry I A, Mortom C R, Hutchison W D,Zhao ZQ. Cutaneous stimuli releasing immunoreactivesubstance P in the dorsal hom of the cat. Brain Res 1988;451: 261-73.

80 Dougherty P M, Palecek J, Paleckova V, Willis W D.Neurokinin 1 and 2 antagonists attenuate the responses

and NKI antagonists prevent the sensitization of primatespinothalamic tract neurons after intradennal capsaicin.JNeurophysiol 1994; 72: 1464-75.

81 Rees H, Sluka K A, Tsuruoka M, Chen P S, Willis W D.The effects ofNK1 and NK2 receptor antagonists on thesensitization of STF cells following acute inflammation inthe anaesthetized primate. J Physiol (Lond) 1995; 483:125P.

82 Yamamoto T, Yaksh T L. Stereospecific effects of a non-peptidic NK1 selective antagonist, CP-96,345: antinoci-ception in the absence of motor dysfunction. Life Sci1991; 49: 1955-63.

83 Thompson S W N, Gerber G, Sivilotti L G, Woolf C J.Long duration ventral root potentials in the neonatal ratspinal cord in vitro; the effects of ionotropic andmetabotropic excitatory amino acide receptor antagonists.Brain Res 1992; 595: 87-97.

84 Haley J E, Sullivan A F, Dickenson A H. Evidence forspinal N-methyl-D-aspartate receptor involvement inprolonged chemical nociception in the rat. Brain Res1990; 518: 218-26.

85 Schaible H G, Grubb B D, Neugebauer V, Oppmann M.The effects ofNMDA antagonists on neuronal activity incat spinal cord evoked by acute inflammation in the kneejoint.J Neurosci 1991; 3: 981-91.

86 Neugebauer V, Luicke T, and Schaible H-G. Differentialeffects of N-methyl-D-aspartate (NMDA) and non-NMDA receptor antagonists on the responses of ratspinal neurons with joint input. Neurosci Len 1993; 155:29-32.

87 Neugebauer V, Locke T, Schaible H-G. N-methyl-D-aspartate (NMDA) and non-NMDA receptorantagonists block the hyperexcitability of dorsal hornneurons during development of acute arthritis in rat'sknee joint.J Neurophysiol 1993; 70: 1365-77.

88 Thompson S W N, Dray A, Urban L. Injury-inducedplasticity of spinal reflex activity: NK1 neurokininreceptor activation and enhanced A- and C-fibermediated responses in the rat spinal cord in vitro. JNeurosci 1994; 14: 3672-87.

89 Rusin K I, Ryu P D, Randic M. Modulation of excitatoryamino acid responses in rat dorsal horn neurones bytachykinins. J Neurophysiol 1992; 68: 265-86.

90 Malmberg A B, Yaksh T L. Hyperalgesia mediated byspinal glutamate or SP receptor blocked by spinalcyclooxygenase inhibition. Science 1992; 257: 1276-9.

91 Malmberg A B, Yaksh T L. Antinociceptive actions ofspinal nonsteroidal anti-inflammatory agents on theformalin test in the rat. 7 Pharmacol Exp Ther 1992; 263:136-46.

92 Malmberg A B, Yaksh T L. Cyclooxygenase inhibition andthe spinal release of prostaglandin E2 and amino acidsevoked by paw formalin injection: A microdialysis studyin unanesthetised rats. J7Neurosci 1995; 15: 2768-76.

93 Chapman V, Dickenson A H. The spinal and peripheralroles of bradykinin and prostaglandins in nociceptiveprocessing in the rat. Eur J Pharmacol 1992; 219:427-33.

94 Brune K, Beck W S, Geisslinger G, Menzel-Soglowek S,Peskar B M, Peskar B A. Aspirin-like drugs may blockpain independently of prostaglandin synthesis inhibition.Experientia 1991; 47: 257-61.

95 McMahon S B, Lewin G R, Wall P D. Centralhyperexcitability triggered by noxious inputs. Curr OpinNeurobiol 1993; 3: 602-10.

96 Woolf C J. Generation of acute pain: Central mechanisms.Br Med Bul 1991; 47: 523-33.

97 Woolf C J, King A E. Subthreshold components of thecutaneous mechanoreceptive fields of dorsal hornneurones in the rat lumbar spinal cord. J Neurophysiol1989; 62: 907-16.

98 Morris V H, Cruwys S C, Kidd B L. Evidence for a centralcomponent to pain in rheumatoid arthritis. Br JRheumatol 1996. In press.

99 Schaible H, Schmidt R F, Willis W D. Convergent inputsfrom articular, cutaneous and muscle receptors ontoascending tract cells in the cat spinal cord. Exp Brain Res1987; 66: 479-88.

100 Grubb B D, Stiller R U, Schaible H G. Dynamic changesin the receptive field properties of spinal cord neuroneswith ankle input in rats with unilateral adjuvant-inducedinflammation in the ankle region. Exp Brain Res 1992; 92:441-52.

283


http://ard.bmj.com

/A

nn Rheum


ay 1996. Dow

nloaded from

http://ard.bmj.com/

pathophysiology ofjoint pain ap · taglandins e2, d2, andi2, mayactivate articu-lar sensory fibres...

Documents