Review

Gabapentin in the treatment of neuropathic painMichael I Bennett St Gemmalsquos Hospice Leeds and Karen H SimpsonSt Jameslsquos University Hospital Leeds

Abstract This paper reviews the pharmacology and clinical effectiveness of gabapentin inthe treatment of neuropathic pain Gabapentin has antihyperalgesic and antiallodynicproperties but does not have significant actions as an anti-nociceptive agent Itsmechanisms of action appear to be a complex synergy between increased GABA synthesisnon-NMDA receptor antagonism and binding to the a2d subunit of voltage dependentcalcium channels The latter action inhibits the release of excitatory neurotransmittersClinically several large randomized controlled trials have demonstrated its effectiveness inthe treatment of a variety of neuropathic pain syndromes Patients with neuropathic paincan expect a mean reduction in pain score of 205 points on an 11 point numerical ratingscale compared with a reduction of 094 points if they had taken the placebo Around 30of patients can expect to achieve more than 50 pain relief and a similar number will alsoexperience minor adverse events the most common of which are somnolence anddizziness In patients with neuropathic pain due to cancer higher response rates might beobserved with gabapentin when administered with opioids because of a synergisticinteraction Palliative Medicine 2004 18 5iexcl11

Key words Gabapentin neuropathic pain

Introduction

Drug developmentGabapentin 1-(aminomethyl) cyclohexane acetic acidwas licensed for use as an antiepileptic agent in the UK in1993 and in the USA in 1994 It was originallysynthesized as a cyclic analogue of gamma aminobutyricacid (GABA) to be used to reduce seizure frequencywhen added to conventional antiepileptic drugs GABAis a principal neurotransmitter found in inhibitoryinterneurones in the dorsal horn Like GABA gabapen-tin is lipophylic and therefore able to pass through thebloodiexclbrain barrier easily Gabapentin gained a UKlicence for the treatment of neuropathic pain in 2000 It isavailable as tablets or capsules The marketing authoriza-tion allows a maximum of 18 g per day in three divideddoses for this indication However it is sometimes usedclinically in higher doses

Rationale for antiepileptics in neuropathic painConventional antiepileptic drugs such as phenytoin andcarbamazepine have been used to treat neuropathic pain

since the 1940s1 This practice was based on observationsthat the paroxysmal pain of trigeminal neuralgia or ticdouloureux resembled epilepsy and led to the termlsquoepileptiform neuralgiarsquo2 Current evidence suggests thatthe pathophysiology of paroxysmal pain may be relatedto bursts of ectopic discharges from damaged nerves Themechanism responsible for spontaneous ectopic firing ismembrane remodelling This results in an excess ofvoltage sensitive sodium channels in the area of nervedamage at either end-bulbs3 4 or demyelinated areas5

Phenytoin and carbamazepine may block use-dependentsodium channels in preference to inactive channels67 andphenytoin has been shown to reduce ectopic discharges invitro8

Other important approaches to the control of chronicneuropathic pain have involved GABA mechanismsActivation of GABA-mediated inhibition can result inanalgesia and a number of agents have been examined inthis context Sodium valproate enhances GABA functionby an unknown mechanism but it is thought to be eitherthrough binding to the GABAA complex or enhancingGABA synthesis or release9 Anecdotal evidence existsfor sodium valproatersquos analgesic effect in the treatment oftrigeminal and post herpetic neuralgia10 11 Benzodiaze-pines such as clonazepam (that bind to GABAA

Address for correspondence Michael Bennett St GemmarsquosHospice 329 Harrogate Road Leeds LS17 6QD UKE-mail mbennettst-gemmacouk

Palliative Medicine 2004 18 5iexcl11

Arnold 2004 1011910269216304pm845ra

benzodiazepine receptor complexes to enhance GABA-mediated inhibition) have little evidence to support theirrole in neuropathic pain12 Baclofen is a GABAB agonistthat results in a presynaptic reduction of excitatoryneurotransmitter release and postsynaptic inhibition ofneuronal response13 There is good evidence that baclo-fen is effective in trigeminal neuralgia14 Vigabatrin anewer antiepileptic drug competitively inhibits GABAaminotransferase and thus enhances synaptic concentra-tions of GABA This drug may have a future role in thetreatment of neuropathic pain but no data are availableto support this at present

Antiepileptics and analgesiaThere have been two systematic reviews of randomizedcontrolled trials of antiepileptic drugs used to treat allpain types1516 These demonstrated little evidence fortheir use in acute pain but significant benefit overplacebo was seen for chronic neuropathic pain condi-tions such as diabetic neuropathy and trigeminal neur-algia

Against this background gabapentin that was thoughtto be active on GABA mechanisms was used in thetreatment of neuropathic pain from the late 1990s

Pharmacology

PharmacokineticsFollowing oral administration gabapentin is absorbedfrom the small intestine rapidly and reliably This occursvia a specific though unidentified transport mechanismthat becomes saturated at higher doses17 This has theeffect of reducing bioavailability at higher doses Thebioavailability of a 300 mg dose is approximately 60 for600 mg it is 40 and only 35 with 16 g given threetimes daily18 19 Gabapentin has no significant proteinbinding and maximum plasma concentrations (Cm ax) arereached after around 32 hours following oral ingestion18

In healthy subjects age has no effect on Cm ax or time toCm ax but a 25 increase in Cm ax was observed in femalescompared to males owing to gender differences in bodysize20 Gabapentin has a half life of 6iexcl8 hours hence theneed for three daily doses

In common with other newer antiepileptics gabapentinis excreted unchanged through the kidneys It does notinduce or inhibit hepatic enzymes It does not signifi-cantly interact with other antiepileptics17 The predomi-nant influence on gabapentin pharmacokinetics is renalfunction An inverse linear correlation exists betweencreatinine clearance and Cm ax time to Cm ax and halflife21 That is impaired renal function results in highergabapentin concentrations and a longer elimination halflife Renal function rather than age accounts for clinicallydifferent effects seen in the elderly

Site of action of gapapentinThe mechanism of action of gabapentin remains un-known Evidence suggests that it is most likely to be theresult of a complex synergy between increased GABAsynthesis non-NMDA receptor antagonism and bindingto the a2d subunit of voltage dependent calcium channelsThe latter action inhibits the release of excitatoryneurotransmitters

Although it was expected to act as a GABA agonistgabapentin does not act directly on GABA receptors22 23

nor does it influence the reuptake of GABA24 It doesincrease the synthesis of GABA from glutamate andenhances its release from astrocytes25 26

Initially gabapentin was also thought to behave as anantagonist at NMDA receptors but the situation appearsmore complex Kaneko et al 27 found that gabapentin didnot behave like an NMDA antagonist in rat models ofneuropathic pain and they concluded that it was unlikelyto act directly on NMDA receptors Furthermore Chenet al 28 have demonstrated a synergistic action between anon-NMDA receptor antagonist and gabapentin whenboth drugs were administered intrathecally This supportsthe view that gabapentin can modify the spinal transmis-sion of noxious mechanical input to spinal neurones vianon-NMDA receptor antagonism

The most important action of gabapentin appears tobe binding to the a2d subunit of voltage dependentcalcium channels29 These binding sites are located in thespinal cord with particularly high density in the super-ficial laminae of the dorsal horn The action of gaba-pentin at these sites may inhibit the release of excitatoryneurotransmitters and reduce glutamate availability atNMDA and non-NMDA receptors30 31

PharmacodynamicsTo better understand the effects of potential agents onneuropathic pain it is useful to clarify some definitionslsquoAntinociceptiversquo refers to the ability of an agent to reducethe reception of noxious stimuli at the level of the sensoryorgan eg the action of aspirin in reducing inflamma-tory activation of cutaneous nociceptors at nerve end-ings lsquoAntalgesic or analgesicrsquo refers to the action of anagent in modulating the neural pathways that transmitpain If pain is the result of increased sensitivity to apainful stimulus and treatment restores the sensitivity tonormal the effect is termed lsquoantihyperalgesicrsquo If anormally non-painful stimulus evokes pain because ofincreased sensitivity (eg lightly brushing the skin) andtreatment restores the sensitivity to normal the termlsquoantiallodynicrsquo is used

Animal studies of the action of gabapentin suggest thatit has antihyperalgesic and antiallodynic effects ratherthan antinociceptive effects32 33 Furthermore a doseiexcl

response relationship has been demonstrated for theseeffects

6 MI Bennett and KH Simpson

Formalin injected into rat paws results in an initialpainful inflammatory response (lasting around 10 min-utes) followed by a later second phase of hyperalgesia(lasting around 45iexcl60 minutes) When gabapentin hasbeen studied in this context Kaneko et al 27 havedemonstrated that it has no effect on the initial phasebut significantly reduces hyperalgesia in the late phaseThey also showed that gabapentin was about three timesas potent as an antihyperalgesic agent if it was givenbefore formalin injection rather than if it was given 7iexcl8minutes after injection This suggests that gabapentin iseffective at preventing the initial development of hyper-algesia and allodynia

Abdi et al 34 examined the effects of systemic gaba-pentin amitriptyline and lidocaine on ectopic dischargesfrom injured sensory fibres in rats Their study showedsignificant reductions in discharges after amitriptylineand more dramatically after lidocaine but gabapentinwas without effect In contrast Pan et al 35 found adose iexclresponse effect after intravenous gabapentin onectopic discharge activity from injured nerves in rats Themechanism of this peripheral effect remains unclearAlthough an action on voltage sensitive sodium channelswould fit with the known actions of other antiepilepticdrugs there is only weak evidence to support this Rocket al 36 failed to show any effect of gabapentin on sodiumchannels but Wamil and McLean37 did demonstrate aninhibitory effect in vitro

Welty et al 38 found that after gabapentin administra-tion the antihyperalgesic effects occur much morerapidly than the antiepileptic effects The antiepilepticeffect of gabapentin may rely on active cellular entry ofgabapentin into astrocytes but it appears that theantihyperalgesic effects do not rely on this process Thisindicates that there may be separate underlying mechan-isms for the action of gabapentin in epilepsy and inneuropathic pain

Gabapentin has been shown to have a superior anti-allodynic profile than morphine or amitriptyline in workby Field et al 39 These authors induced static anddynamic allodynia in rats through streptozocin injectioninto the hind paw Amitriptyline and morphine dosedependently blocked static allodynia but not dynamicallodynia Gabapentin blocked both types of allodynia ina dose dependent manner between 10 and 100 mgkg

These studies suggest that gabapentin has significantantihyperalgesic and antiallodynic effects It suggests thatthese are mediated predominantly at a spinal or suprasp-inal level through stabilization of dorsal horn neuronesand to a lesser extent at a peripheral level

Interaction with opioidsAlthough there is no evidence to support a direct effect ofgabapentin on opioid pathways animal studies havepointed to an indirect effect Rocha et al 40 found raised

levels of extracellular opioid peptide in rat brainsfollowing a single dose of gabapentin Shimoyama etal 41 demonstrated a synergistic effect between morphineand gabapentin that was significantly greater than theeffects of either drug alone in a rat tail flick latency testto radiant heat

Clinical studies

Numerous case studies have been published describingthe effects of gabapentin in a variety of neuropathic painconditions Examples include complex regional painsyndromes (CRPS) deafferentation neuropathies centralpost stroke pain and pain associated with multiplesclerosis42 iexcl 45 In general treatment with gabapentinwas successful in reducing pain and the doses used werebetween 900 and 1200 mg in three divided dosesFollowing the apparent success of gabapentin describedin case reports randomized placebo controlled trials wereundertaken in various defined neuropathic pain condi-tions

Post herpetic neuralgiaTwo large randomized controlled trials have been con-ducted on the use of gabapentin in the treatment of postherpetic neuralgia The first study recruited 229 patientsfrom North American pain services in whom doses weretitrated up to 3600 mg of gabapentin or placebo over afour week period46 Treatment was maintained at max-imum dose for a further four weeks concomitantantidepressants used as coanalgesics were continued iftherapy was stabilized and remained constant Around80 of patients in each arm completed the study medianage was 73 and 74 years in the gabapentin and placebogroups respectively There were no significant differencesin other medication taken during the study 72 took noother medication 12 took tricyclic antidepressants and19iexcl27 took opioids

In those that received gabapentin the average dailypain score was significantly reduced from 63 to 42points on an 11-point numerical rating scale In thisgroup at least 83 received 2400 mg and 65 received atleast 3600 mg of gabapentin This compares with areduction from 65 to 60 in those that received placebo(PB0001) Other outcome measurements includingquality of life sleep interference and mood also showedsignificant improvement with gabapentin The numbersof withdrawals in each group were similar but patientstaking gabapentin reported more somnolence dizzinessand ataxia than those taking placebo The authorsconcluded that gabapentin was safe and efficacious atthese doses in a predominantly elderly population

The second and more recent study recruited 334patients from UK pain clinics using similar methodol-

Gabapentin in neuropathic pain 7

ogy47 Patients were randomized to receive 1800 mggabapentin or 2400 mg gabapentin or placebo Therationale for the study was to examine whether lowerdoses would be as efficacious as higher doses but wouldhave a lower incidence of side effects This trial excludedpatients who had previously failed to respond to gaba-pentin Patients taking active treatment were titrated upto 1800 mg gabapentin over two weeks and then up to2400 mg by week 3 for those randomized to receive thehigher dose All doses remained stable between weeks 3iexcl

7 Median age of patients recruited was around 75 yearsin all three groups Average daily pain score showedsignificant improvements after both 1800 and 2400 mggabapentin 65 to 43 and 65 to 42 respectively on an11 point numerical rating scale (PB0001) In compar-ison the reduction for the placebo group was 64 to 53points This translates to an 188 reduction in daily painscore using gabapentin instead of placebo These differ-ences were apparent from the end of the first week oftreatment Between 32 and 34 of patients takinggabapentin achieved more than 50 pain relief comparedto only 14 taking placebo Predominant side effects inthe treatment arms were dizziness (31iexcl33) and somno-lence (174iexcl204) Between 13 and 17 of patientstaking gabapentin withdrew because of side effects

Diabetic neuropathyAnother large randomized placebo controlled study ofgabapentin was conducted in patients with diabeticneuropathy48 attending pain clinics in the USA Of the165 patients recruited 84 were randomized to receivegabapentin and 81 to receive placebo The methodologyand dosing schedule was similar to that used in the trialby Rowbotham et al 46 Two thirds of the patients takinggabapentin were titrated to 3600 mg daily

This study also showed a significant reduction inaverage daily pain score for the gabapentin group from64 to 39 points on a numerical rating scale Thiscompares with a reduction from 65 to 51 in the placebogroup Quality of life measures sleep and mood allimproved significantly in those taking gabapentin and47 of the gabapentin group achieved a lsquomuch ormoderately improvedrsquo rank on a global change scorecompared with 25 of those receiving placebo Althoughthe median age in this study was 53 years (20 yearsyounger than the post herpetic neuralgia trials) thefrequency of side effects was similar Significant differ-ences between gabapentin and placebo were observed fordizziness (238 versus 49) and somnolence (226 versus62)

Mixed neuropathic painDespite the apparent efficacy of gabapentin in postherpetic neuralgia and diabetic neuropathy clinicalpractice is characterized by patients with a variety of

neuropathic pains These patients are often defined bytheir pain symptoms and signs rather than discretediseases To address this issue a further clinical trial ofgabapentin was undertaken that recruited 307 patientswith various neuropathic pain syndromes (excludingcancer) from pain clinics in the UK and Ireland49

Eligible patients were those with mean daily pain scoresof at least 4 on an 11 point numerical rating scale priorto entry Patients were also required to have at least 2 ofthe following symptoms and signs allodynia burningpain shooting pain hyperalgesia The median age ofrecruited patients was around 57 years in both groups

In this study patients were randomized to receiveplacebo or to be titrated up to receive 2400 mg ofgabapentin over five weeks to be followed by a three weekperiod where doses were stable By week 5 in thegabapentin arm 101 patients received 2400 mg 19received 1800 mg and 27 received 900 mg daily Themean weekly pain scores at week 8 showed significantdifferences for gabapentin In this group pain scores werereduced from 71 to 56 points on an 11 point numericalrating scale compared with a reduction from 73 to 63 inthe placebo group (PB0001) This difference wassignificant at the end of week 1 and remained similarfor most of the weeks throughout the study period Nosignificant difference was seen in the number of patientswho achieved more than 50 pain relief between the twogroups (21 for gabapentin and 14 for placebo)Interestingly significant improvements in burning painand hyperalgesia were seen but this was not observed forallodynia or shooting pain when gabapentin was com-pared with placebo Significant improvements were alsoseen after gabapentin in various quality of life measuresparticularly social and emotional functioning The mostfrequent side effects with gabapentin compared withplacebo were dizziness (242 versus 79) and somno-lence (144 versus 53)

Cancer neuropathic painTo date there have been no randomized trials ofgabapentin in patients with neuropathic pain due tocancer One open-label trial has been reported in which22 patients with cancer whose neuropathic pain was notcontrolled by opioids were given gabapentin as lsquoadd-onrsquotreatment50 The dose of gabapentin was titrated overthree to seven days up to a mean daily dose of around1004 mg (range 600iexcl1800 mg) Opioid doses wereunchanged throughout the study period Compared tobaseline scores mean global pain scores on a numericalrating scale were reduced from 64 to 32 points at thefinal assessment that took place 7iexcl14 days later Symp-toms of burning pain shooting pain and allodynia wereall also significantly reduced compared to baselineOverall no additional side effects were attributed to

8 MI Bennett and KH Simpson

gabapentin treatment apart from one case of sedationand one case of dizziness

This study was commented upon in a letter byChandler and Williams51 who argued that gabapentinhas a much lower incidence of efficacy in this contextbased on their experience They reported on a retro-spective review of 20 patients with neuropathic pain dueto cancer who were treated with gabapentin in addition totheir usual analgesia Only nine patients reported sig-nificant improvement and six others discontinued gaba-pentin because of side effects In the remaining fivepatients efficacy could not be established Furtheranalysis of the data in the Caraceni et al study50 showsthat 13 of 22 (59) patients achieved more than 50 painrelief compared to 45 (assuming lsquosignificant reliefrsquo isequivalent) in Chandler and Williamsrsquos series Both ofthese series compare favourably to larger randomizedcontrolled trials4749 which were performed in presum-ably fitter patients where between 21 and 34 achievedthis degree of pain relief

Comparative studies with other drugsA small number of studies exist that directly compare theeffectiveness of gabapentin against other analgesics in thetreatment of neuropathic pain Dallocchio et al 52

randomized 25 patients with diabetic neuropathy toreceive amitriptyline (mean dose 53 mg daily) or gaba-pentin (mean dose 1785 mg daily) in an open label trialDoses were titrated over four weeks and maintained foreight weeks Pain scores were measured on a five pointscale (0 frac34none 4frac34excruciating) Baseline pain scores inthe gabapentin group were reduced by a mean of 19(29iexcl10) and those in the amitriptyline group werereduced be a mean of 13 (275iexcl15) PB0026 Sig-nificant differences in the reduction in paraesthesiascores also favoured gabapentin Adverse events weremore frequent in the amitriptyline group and theselimited dose escalation according to the authors

However an earlier double blind randomized cross-over trial had not found any difference in efficacybetween the two drugs in patients with painful diabeticneuropathy53 In this study 21 of the 25 patients enrolledcompleted the six week trial of each drug with a one weekwashout period in between drug dosing (mean doses were59 mg amitriptyline and 1565 mg gabapentin) Pain scorediaries showed no statistical differences for patientstaking either treatment 11 (52) taking amitriptylineexperienced moderate or greater pain relief compared to14 (67) of those taking gabapentin

Eckhardt et al 54 have compared morphine gabapentinand combined treatment with both drugs in a doubleblind placebo controlled four way crossover study Thiswas conducted in 12 volunteers whose pain tolerance andpain threshold were measured during forearm immersionin cold water under experimental conditions Although

this was not a study of clinical neuropathic pain theresults are interesting Neither gabapentin nor morphinehad any significant effect on pain threshold but mor-phine had a significant effect on pain tolerance Im-portantly the combination of the two drugs significantlyraised pain threshold and tolerance (above that ofmorphine alone) This synergistic effect had been pre-viously demonstrated in rats41 It also supports thefindings of Caraceni et al 50 suggesting that gabapentinaugments the action of morphine via an unexplainedmechanism

Meta-analysesInformation from meta-analyses and systematic reviewsallows comparisons of effectiveness of drugs in variousneuropathic pain conditions These are expressed asNNT (number needed to treat) and NNH (numberneeded to harm)1655 56 In post herpetic neuralgia theNNT for gabapentin was 32 (CI 24iexcl50) whichcompares with 23 (CI 17iexcl33) for tricyclic antidepres-sants and 25 (CI 16iexcl51) for oxycodone In diabeticneuropathy the NNTs are gabapentin 38 (24iexcl87)tricyclic antidepressants 30 (23iexcl43) and carbamazepine23 (16iexcl38) Using the calculation for NNT suggestedby Cook and Sackett57 and the data for mixed neuro-pathic pain syndromes49 the NNT for gabapentin in thiscontext is 14

In diabetic neuropathy studies the NNH for minorevents is 25 (20iexcl32) for gabapentin 37 (24iexcl78) forcarbamazepine and 28 (20iexcl47) for tricyclic antidepres-sants The NNH for major events (drug related with-drawal from study) is 20iexcl40 this is broadly similar fortricyclic antidepressants and anticonvulsants (excludinggabapentin) In relation to gabapentin three of the fourmajor trials in neuropathic pain demonstrate no differ-ences between study withdrawals in those taking placebocompared with those taking active treatment4648 49 Inthe fourth study47 63 subjects withdrew whilst takingplacebo compared to 13 and 176 of those takinggabapentin 1800 and 2400 mg respectively these resultswere not assessed statistically It has been argued thatearlier studies of treatments in neuropathic pain (thathave contributed to meta-analyses) may have been lessable to detect and report adverse events This would leadto an overestimation of NNH for older anticonvulsantsand tricyclic antidepressants (ie they appear bettertolerated than is the case in clinical practice) whencompared with more recent studies of gabapentin

Summary

Gabapentin appears to be an effective treatment forneuropathic pain acting centrally to reduce hyperalgesiaand allodynia but it does not have significant actions as

Gabapentin in neuropathic pain 9

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

benzodiazepine receptor complexes to enhance GABA-mediated inhibition) have little evidence to support theirrole in neuropathic pain12 Baclofen is a GABAB agonistthat results in a presynaptic reduction of excitatoryneurotransmitter release and postsynaptic inhibition ofneuronal response13 There is good evidence that baclo-fen is effective in trigeminal neuralgia14 Vigabatrin anewer antiepileptic drug competitively inhibits GABAaminotransferase and thus enhances synaptic concentra-tions of GABA This drug may have a future role in thetreatment of neuropathic pain but no data are availableto support this at present

Antiepileptics and analgesiaThere have been two systematic reviews of randomizedcontrolled trials of antiepileptic drugs used to treat allpain types1516 These demonstrated little evidence fortheir use in acute pain but significant benefit overplacebo was seen for chronic neuropathic pain condi-tions such as diabetic neuropathy and trigeminal neur-algia

Against this background gabapentin that was thoughtto be active on GABA mechanisms was used in thetreatment of neuropathic pain from the late 1990s

Pharmacology

PharmacokineticsFollowing oral administration gabapentin is absorbedfrom the small intestine rapidly and reliably This occursvia a specific though unidentified transport mechanismthat becomes saturated at higher doses17 This has theeffect of reducing bioavailability at higher doses Thebioavailability of a 300 mg dose is approximately 60 for600 mg it is 40 and only 35 with 16 g given threetimes daily18 19 Gabapentin has no significant proteinbinding and maximum plasma concentrations (Cm ax) arereached after around 32 hours following oral ingestion18

In healthy subjects age has no effect on Cm ax or time toCm ax but a 25 increase in Cm ax was observed in femalescompared to males owing to gender differences in bodysize20 Gabapentin has a half life of 6iexcl8 hours hence theneed for three daily doses

In common with other newer antiepileptics gabapentinis excreted unchanged through the kidneys It does notinduce or inhibit hepatic enzymes It does not signifi-cantly interact with other antiepileptics17 The predomi-nant influence on gabapentin pharmacokinetics is renalfunction An inverse linear correlation exists betweencreatinine clearance and Cm ax time to Cm ax and halflife21 That is impaired renal function results in highergabapentin concentrations and a longer elimination halflife Renal function rather than age accounts for clinicallydifferent effects seen in the elderly

Site of action of gapapentinThe mechanism of action of gabapentin remains un-known Evidence suggests that it is most likely to be theresult of a complex synergy between increased GABAsynthesis non-NMDA receptor antagonism and bindingto the a2d subunit of voltage dependent calcium channelsThe latter action inhibits the release of excitatoryneurotransmitters

Although it was expected to act as a GABA agonistgabapentin does not act directly on GABA receptors22 23

nor does it influence the reuptake of GABA24 It doesincrease the synthesis of GABA from glutamate andenhances its release from astrocytes25 26

Initially gabapentin was also thought to behave as anantagonist at NMDA receptors but the situation appearsmore complex Kaneko et al 27 found that gabapentin didnot behave like an NMDA antagonist in rat models ofneuropathic pain and they concluded that it was unlikelyto act directly on NMDA receptors Furthermore Chenet al 28 have demonstrated a synergistic action between anon-NMDA receptor antagonist and gabapentin whenboth drugs were administered intrathecally This supportsthe view that gabapentin can modify the spinal transmis-sion of noxious mechanical input to spinal neurones vianon-NMDA receptor antagonism

The most important action of gabapentin appears tobe binding to the a2d subunit of voltage dependentcalcium channels29 These binding sites are located in thespinal cord with particularly high density in the super-ficial laminae of the dorsal horn The action of gaba-pentin at these sites may inhibit the release of excitatoryneurotransmitters and reduce glutamate availability atNMDA and non-NMDA receptors30 31

PharmacodynamicsTo better understand the effects of potential agents onneuropathic pain it is useful to clarify some definitionslsquoAntinociceptiversquo refers to the ability of an agent to reducethe reception of noxious stimuli at the level of the sensoryorgan eg the action of aspirin in reducing inflamma-tory activation of cutaneous nociceptors at nerve end-ings lsquoAntalgesic or analgesicrsquo refers to the action of anagent in modulating the neural pathways that transmitpain If pain is the result of increased sensitivity to apainful stimulus and treatment restores the sensitivity tonormal the effect is termed lsquoantihyperalgesicrsquo If anormally non-painful stimulus evokes pain because ofincreased sensitivity (eg lightly brushing the skin) andtreatment restores the sensitivity to normal the termlsquoantiallodynicrsquo is used

Animal studies of the action of gabapentin suggest thatit has antihyperalgesic and antiallodynic effects ratherthan antinociceptive effects32 33 Furthermore a doseiexcl

response relationship has been demonstrated for theseeffects

6 MI Bennett and KH Simpson

Formalin injected into rat paws results in an initialpainful inflammatory response (lasting around 10 min-utes) followed by a later second phase of hyperalgesia(lasting around 45iexcl60 minutes) When gabapentin hasbeen studied in this context Kaneko et al 27 havedemonstrated that it has no effect on the initial phasebut significantly reduces hyperalgesia in the late phaseThey also showed that gabapentin was about three timesas potent as an antihyperalgesic agent if it was givenbefore formalin injection rather than if it was given 7iexcl8minutes after injection This suggests that gabapentin iseffective at preventing the initial development of hyper-algesia and allodynia

Abdi et al 34 examined the effects of systemic gaba-pentin amitriptyline and lidocaine on ectopic dischargesfrom injured sensory fibres in rats Their study showedsignificant reductions in discharges after amitriptylineand more dramatically after lidocaine but gabapentinwas without effect In contrast Pan et al 35 found adose iexclresponse effect after intravenous gabapentin onectopic discharge activity from injured nerves in rats Themechanism of this peripheral effect remains unclearAlthough an action on voltage sensitive sodium channelswould fit with the known actions of other antiepilepticdrugs there is only weak evidence to support this Rocket al 36 failed to show any effect of gabapentin on sodiumchannels but Wamil and McLean37 did demonstrate aninhibitory effect in vitro

Welty et al 38 found that after gabapentin administra-tion the antihyperalgesic effects occur much morerapidly than the antiepileptic effects The antiepilepticeffect of gabapentin may rely on active cellular entry ofgabapentin into astrocytes but it appears that theantihyperalgesic effects do not rely on this process Thisindicates that there may be separate underlying mechan-isms for the action of gabapentin in epilepsy and inneuropathic pain

Gabapentin has been shown to have a superior anti-allodynic profile than morphine or amitriptyline in workby Field et al 39 These authors induced static anddynamic allodynia in rats through streptozocin injectioninto the hind paw Amitriptyline and morphine dosedependently blocked static allodynia but not dynamicallodynia Gabapentin blocked both types of allodynia ina dose dependent manner between 10 and 100 mgkg

These studies suggest that gabapentin has significantantihyperalgesic and antiallodynic effects It suggests thatthese are mediated predominantly at a spinal or suprasp-inal level through stabilization of dorsal horn neuronesand to a lesser extent at a peripheral level

Interaction with opioidsAlthough there is no evidence to support a direct effect ofgabapentin on opioid pathways animal studies havepointed to an indirect effect Rocha et al 40 found raised

levels of extracellular opioid peptide in rat brainsfollowing a single dose of gabapentin Shimoyama etal 41 demonstrated a synergistic effect between morphineand gabapentin that was significantly greater than theeffects of either drug alone in a rat tail flick latency testto radiant heat

Clinical studies

Numerous case studies have been published describingthe effects of gabapentin in a variety of neuropathic painconditions Examples include complex regional painsyndromes (CRPS) deafferentation neuropathies centralpost stroke pain and pain associated with multiplesclerosis42 iexcl 45 In general treatment with gabapentinwas successful in reducing pain and the doses used werebetween 900 and 1200 mg in three divided dosesFollowing the apparent success of gabapentin describedin case reports randomized placebo controlled trials wereundertaken in various defined neuropathic pain condi-tions

Post herpetic neuralgiaTwo large randomized controlled trials have been con-ducted on the use of gabapentin in the treatment of postherpetic neuralgia The first study recruited 229 patientsfrom North American pain services in whom doses weretitrated up to 3600 mg of gabapentin or placebo over afour week period46 Treatment was maintained at max-imum dose for a further four weeks concomitantantidepressants used as coanalgesics were continued iftherapy was stabilized and remained constant Around80 of patients in each arm completed the study medianage was 73 and 74 years in the gabapentin and placebogroups respectively There were no significant differencesin other medication taken during the study 72 took noother medication 12 took tricyclic antidepressants and19iexcl27 took opioids

In those that received gabapentin the average dailypain score was significantly reduced from 63 to 42points on an 11-point numerical rating scale In thisgroup at least 83 received 2400 mg and 65 received atleast 3600 mg of gabapentin This compares with areduction from 65 to 60 in those that received placebo(PB0001) Other outcome measurements includingquality of life sleep interference and mood also showedsignificant improvement with gabapentin The numbersof withdrawals in each group were similar but patientstaking gabapentin reported more somnolence dizzinessand ataxia than those taking placebo The authorsconcluded that gabapentin was safe and efficacious atthese doses in a predominantly elderly population

The second and more recent study recruited 334patients from UK pain clinics using similar methodol-

Gabapentin in neuropathic pain 7

ogy47 Patients were randomized to receive 1800 mggabapentin or 2400 mg gabapentin or placebo Therationale for the study was to examine whether lowerdoses would be as efficacious as higher doses but wouldhave a lower incidence of side effects This trial excludedpatients who had previously failed to respond to gaba-pentin Patients taking active treatment were titrated upto 1800 mg gabapentin over two weeks and then up to2400 mg by week 3 for those randomized to receive thehigher dose All doses remained stable between weeks 3iexcl

7 Median age of patients recruited was around 75 yearsin all three groups Average daily pain score showedsignificant improvements after both 1800 and 2400 mggabapentin 65 to 43 and 65 to 42 respectively on an11 point numerical rating scale (PB0001) In compar-ison the reduction for the placebo group was 64 to 53points This translates to an 188 reduction in daily painscore using gabapentin instead of placebo These differ-ences were apparent from the end of the first week oftreatment Between 32 and 34 of patients takinggabapentin achieved more than 50 pain relief comparedto only 14 taking placebo Predominant side effects inthe treatment arms were dizziness (31iexcl33) and somno-lence (174iexcl204) Between 13 and 17 of patientstaking gabapentin withdrew because of side effects

Diabetic neuropathyAnother large randomized placebo controlled study ofgabapentin was conducted in patients with diabeticneuropathy48 attending pain clinics in the USA Of the165 patients recruited 84 were randomized to receivegabapentin and 81 to receive placebo The methodologyand dosing schedule was similar to that used in the trialby Rowbotham et al 46 Two thirds of the patients takinggabapentin were titrated to 3600 mg daily

This study also showed a significant reduction inaverage daily pain score for the gabapentin group from64 to 39 points on a numerical rating scale Thiscompares with a reduction from 65 to 51 in the placebogroup Quality of life measures sleep and mood allimproved significantly in those taking gabapentin and47 of the gabapentin group achieved a lsquomuch ormoderately improvedrsquo rank on a global change scorecompared with 25 of those receiving placebo Althoughthe median age in this study was 53 years (20 yearsyounger than the post herpetic neuralgia trials) thefrequency of side effects was similar Significant differ-ences between gabapentin and placebo were observed fordizziness (238 versus 49) and somnolence (226 versus62)

Mixed neuropathic painDespite the apparent efficacy of gabapentin in postherpetic neuralgia and diabetic neuropathy clinicalpractice is characterized by patients with a variety of

neuropathic pains These patients are often defined bytheir pain symptoms and signs rather than discretediseases To address this issue a further clinical trial ofgabapentin was undertaken that recruited 307 patientswith various neuropathic pain syndromes (excludingcancer) from pain clinics in the UK and Ireland49

Eligible patients were those with mean daily pain scoresof at least 4 on an 11 point numerical rating scale priorto entry Patients were also required to have at least 2 ofthe following symptoms and signs allodynia burningpain shooting pain hyperalgesia The median age ofrecruited patients was around 57 years in both groups

In this study patients were randomized to receiveplacebo or to be titrated up to receive 2400 mg ofgabapentin over five weeks to be followed by a three weekperiod where doses were stable By week 5 in thegabapentin arm 101 patients received 2400 mg 19received 1800 mg and 27 received 900 mg daily Themean weekly pain scores at week 8 showed significantdifferences for gabapentin In this group pain scores werereduced from 71 to 56 points on an 11 point numericalrating scale compared with a reduction from 73 to 63 inthe placebo group (PB0001) This difference wassignificant at the end of week 1 and remained similarfor most of the weeks throughout the study period Nosignificant difference was seen in the number of patientswho achieved more than 50 pain relief between the twogroups (21 for gabapentin and 14 for placebo)Interestingly significant improvements in burning painand hyperalgesia were seen but this was not observed forallodynia or shooting pain when gabapentin was com-pared with placebo Significant improvements were alsoseen after gabapentin in various quality of life measuresparticularly social and emotional functioning The mostfrequent side effects with gabapentin compared withplacebo were dizziness (242 versus 79) and somno-lence (144 versus 53)

Cancer neuropathic painTo date there have been no randomized trials ofgabapentin in patients with neuropathic pain due tocancer One open-label trial has been reported in which22 patients with cancer whose neuropathic pain was notcontrolled by opioids were given gabapentin as lsquoadd-onrsquotreatment50 The dose of gabapentin was titrated overthree to seven days up to a mean daily dose of around1004 mg (range 600iexcl1800 mg) Opioid doses wereunchanged throughout the study period Compared tobaseline scores mean global pain scores on a numericalrating scale were reduced from 64 to 32 points at thefinal assessment that took place 7iexcl14 days later Symp-toms of burning pain shooting pain and allodynia wereall also significantly reduced compared to baselineOverall no additional side effects were attributed to

8 MI Bennett and KH Simpson

gabapentin treatment apart from one case of sedationand one case of dizziness

This study was commented upon in a letter byChandler and Williams51 who argued that gabapentinhas a much lower incidence of efficacy in this contextbased on their experience They reported on a retro-spective review of 20 patients with neuropathic pain dueto cancer who were treated with gabapentin in addition totheir usual analgesia Only nine patients reported sig-nificant improvement and six others discontinued gaba-pentin because of side effects In the remaining fivepatients efficacy could not be established Furtheranalysis of the data in the Caraceni et al study50 showsthat 13 of 22 (59) patients achieved more than 50 painrelief compared to 45 (assuming lsquosignificant reliefrsquo isequivalent) in Chandler and Williamsrsquos series Both ofthese series compare favourably to larger randomizedcontrolled trials4749 which were performed in presum-ably fitter patients where between 21 and 34 achievedthis degree of pain relief

Comparative studies with other drugsA small number of studies exist that directly compare theeffectiveness of gabapentin against other analgesics in thetreatment of neuropathic pain Dallocchio et al 52

randomized 25 patients with diabetic neuropathy toreceive amitriptyline (mean dose 53 mg daily) or gaba-pentin (mean dose 1785 mg daily) in an open label trialDoses were titrated over four weeks and maintained foreight weeks Pain scores were measured on a five pointscale (0 frac34none 4frac34excruciating) Baseline pain scores inthe gabapentin group were reduced by a mean of 19(29iexcl10) and those in the amitriptyline group werereduced be a mean of 13 (275iexcl15) PB0026 Sig-nificant differences in the reduction in paraesthesiascores also favoured gabapentin Adverse events weremore frequent in the amitriptyline group and theselimited dose escalation according to the authors

However an earlier double blind randomized cross-over trial had not found any difference in efficacybetween the two drugs in patients with painful diabeticneuropathy53 In this study 21 of the 25 patients enrolledcompleted the six week trial of each drug with a one weekwashout period in between drug dosing (mean doses were59 mg amitriptyline and 1565 mg gabapentin) Pain scorediaries showed no statistical differences for patientstaking either treatment 11 (52) taking amitriptylineexperienced moderate or greater pain relief compared to14 (67) of those taking gabapentin

Eckhardt et al 54 have compared morphine gabapentinand combined treatment with both drugs in a doubleblind placebo controlled four way crossover study Thiswas conducted in 12 volunteers whose pain tolerance andpain threshold were measured during forearm immersionin cold water under experimental conditions Although

this was not a study of clinical neuropathic pain theresults are interesting Neither gabapentin nor morphinehad any significant effect on pain threshold but mor-phine had a significant effect on pain tolerance Im-portantly the combination of the two drugs significantlyraised pain threshold and tolerance (above that ofmorphine alone) This synergistic effect had been pre-viously demonstrated in rats41 It also supports thefindings of Caraceni et al 50 suggesting that gabapentinaugments the action of morphine via an unexplainedmechanism

Meta-analysesInformation from meta-analyses and systematic reviewsallows comparisons of effectiveness of drugs in variousneuropathic pain conditions These are expressed asNNT (number needed to treat) and NNH (numberneeded to harm)1655 56 In post herpetic neuralgia theNNT for gabapentin was 32 (CI 24iexcl50) whichcompares with 23 (CI 17iexcl33) for tricyclic antidepres-sants and 25 (CI 16iexcl51) for oxycodone In diabeticneuropathy the NNTs are gabapentin 38 (24iexcl87)tricyclic antidepressants 30 (23iexcl43) and carbamazepine23 (16iexcl38) Using the calculation for NNT suggestedby Cook and Sackett57 and the data for mixed neuro-pathic pain syndromes49 the NNT for gabapentin in thiscontext is 14

In diabetic neuropathy studies the NNH for minorevents is 25 (20iexcl32) for gabapentin 37 (24iexcl78) forcarbamazepine and 28 (20iexcl47) for tricyclic antidepres-sants The NNH for major events (drug related with-drawal from study) is 20iexcl40 this is broadly similar fortricyclic antidepressants and anticonvulsants (excludinggabapentin) In relation to gabapentin three of the fourmajor trials in neuropathic pain demonstrate no differ-ences between study withdrawals in those taking placebocompared with those taking active treatment4648 49 Inthe fourth study47 63 subjects withdrew whilst takingplacebo compared to 13 and 176 of those takinggabapentin 1800 and 2400 mg respectively these resultswere not assessed statistically It has been argued thatearlier studies of treatments in neuropathic pain (thathave contributed to meta-analyses) may have been lessable to detect and report adverse events This would leadto an overestimation of NNH for older anticonvulsantsand tricyclic antidepressants (ie they appear bettertolerated than is the case in clinical practice) whencompared with more recent studies of gabapentin

Summary

Gabapentin appears to be an effective treatment forneuropathic pain acting centrally to reduce hyperalgesiaand allodynia but it does not have significant actions as

Gabapentin in neuropathic pain 9

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

Formalin injected into rat paws results in an initialpainful inflammatory response (lasting around 10 min-utes) followed by a later second phase of hyperalgesia(lasting around 45iexcl60 minutes) When gabapentin hasbeen studied in this context Kaneko et al 27 havedemonstrated that it has no effect on the initial phasebut significantly reduces hyperalgesia in the late phaseThey also showed that gabapentin was about three timesas potent as an antihyperalgesic agent if it was givenbefore formalin injection rather than if it was given 7iexcl8minutes after injection This suggests that gabapentin iseffective at preventing the initial development of hyper-algesia and allodynia

Abdi et al 34 examined the effects of systemic gaba-pentin amitriptyline and lidocaine on ectopic dischargesfrom injured sensory fibres in rats Their study showedsignificant reductions in discharges after amitriptylineand more dramatically after lidocaine but gabapentinwas without effect In contrast Pan et al 35 found adose iexclresponse effect after intravenous gabapentin onectopic discharge activity from injured nerves in rats Themechanism of this peripheral effect remains unclearAlthough an action on voltage sensitive sodium channelswould fit with the known actions of other antiepilepticdrugs there is only weak evidence to support this Rocket al 36 failed to show any effect of gabapentin on sodiumchannels but Wamil and McLean37 did demonstrate aninhibitory effect in vitro

Welty et al 38 found that after gabapentin administra-tion the antihyperalgesic effects occur much morerapidly than the antiepileptic effects The antiepilepticeffect of gabapentin may rely on active cellular entry ofgabapentin into astrocytes but it appears that theantihyperalgesic effects do not rely on this process Thisindicates that there may be separate underlying mechan-isms for the action of gabapentin in epilepsy and inneuropathic pain

Gabapentin has been shown to have a superior anti-allodynic profile than morphine or amitriptyline in workby Field et al 39 These authors induced static anddynamic allodynia in rats through streptozocin injectioninto the hind paw Amitriptyline and morphine dosedependently blocked static allodynia but not dynamicallodynia Gabapentin blocked both types of allodynia ina dose dependent manner between 10 and 100 mgkg

These studies suggest that gabapentin has significantantihyperalgesic and antiallodynic effects It suggests thatthese are mediated predominantly at a spinal or suprasp-inal level through stabilization of dorsal horn neuronesand to a lesser extent at a peripheral level

Interaction with opioidsAlthough there is no evidence to support a direct effect ofgabapentin on opioid pathways animal studies havepointed to an indirect effect Rocha et al 40 found raised

levels of extracellular opioid peptide in rat brainsfollowing a single dose of gabapentin Shimoyama etal 41 demonstrated a synergistic effect between morphineand gabapentin that was significantly greater than theeffects of either drug alone in a rat tail flick latency testto radiant heat

Clinical studies

Numerous case studies have been published describingthe effects of gabapentin in a variety of neuropathic painconditions Examples include complex regional painsyndromes (CRPS) deafferentation neuropathies centralpost stroke pain and pain associated with multiplesclerosis42 iexcl 45 In general treatment with gabapentinwas successful in reducing pain and the doses used werebetween 900 and 1200 mg in three divided dosesFollowing the apparent success of gabapentin describedin case reports randomized placebo controlled trials wereundertaken in various defined neuropathic pain condi-tions

Post herpetic neuralgiaTwo large randomized controlled trials have been con-ducted on the use of gabapentin in the treatment of postherpetic neuralgia The first study recruited 229 patientsfrom North American pain services in whom doses weretitrated up to 3600 mg of gabapentin or placebo over afour week period46 Treatment was maintained at max-imum dose for a further four weeks concomitantantidepressants used as coanalgesics were continued iftherapy was stabilized and remained constant Around80 of patients in each arm completed the study medianage was 73 and 74 years in the gabapentin and placebogroups respectively There were no significant differencesin other medication taken during the study 72 took noother medication 12 took tricyclic antidepressants and19iexcl27 took opioids

In those that received gabapentin the average dailypain score was significantly reduced from 63 to 42points on an 11-point numerical rating scale In thisgroup at least 83 received 2400 mg and 65 received atleast 3600 mg of gabapentin This compares with areduction from 65 to 60 in those that received placebo(PB0001) Other outcome measurements includingquality of life sleep interference and mood also showedsignificant improvement with gabapentin The numbersof withdrawals in each group were similar but patientstaking gabapentin reported more somnolence dizzinessand ataxia than those taking placebo The authorsconcluded that gabapentin was safe and efficacious atthese doses in a predominantly elderly population

The second and more recent study recruited 334patients from UK pain clinics using similar methodol-

Gabapentin in neuropathic pain 7

ogy47 Patients were randomized to receive 1800 mggabapentin or 2400 mg gabapentin or placebo Therationale for the study was to examine whether lowerdoses would be as efficacious as higher doses but wouldhave a lower incidence of side effects This trial excludedpatients who had previously failed to respond to gaba-pentin Patients taking active treatment were titrated upto 1800 mg gabapentin over two weeks and then up to2400 mg by week 3 for those randomized to receive thehigher dose All doses remained stable between weeks 3iexcl

7 Median age of patients recruited was around 75 yearsin all three groups Average daily pain score showedsignificant improvements after both 1800 and 2400 mggabapentin 65 to 43 and 65 to 42 respectively on an11 point numerical rating scale (PB0001) In compar-ison the reduction for the placebo group was 64 to 53points This translates to an 188 reduction in daily painscore using gabapentin instead of placebo These differ-ences were apparent from the end of the first week oftreatment Between 32 and 34 of patients takinggabapentin achieved more than 50 pain relief comparedto only 14 taking placebo Predominant side effects inthe treatment arms were dizziness (31iexcl33) and somno-lence (174iexcl204) Between 13 and 17 of patientstaking gabapentin withdrew because of side effects

Diabetic neuropathyAnother large randomized placebo controlled study ofgabapentin was conducted in patients with diabeticneuropathy48 attending pain clinics in the USA Of the165 patients recruited 84 were randomized to receivegabapentin and 81 to receive placebo The methodologyand dosing schedule was similar to that used in the trialby Rowbotham et al 46 Two thirds of the patients takinggabapentin were titrated to 3600 mg daily

This study also showed a significant reduction inaverage daily pain score for the gabapentin group from64 to 39 points on a numerical rating scale Thiscompares with a reduction from 65 to 51 in the placebogroup Quality of life measures sleep and mood allimproved significantly in those taking gabapentin and47 of the gabapentin group achieved a lsquomuch ormoderately improvedrsquo rank on a global change scorecompared with 25 of those receiving placebo Althoughthe median age in this study was 53 years (20 yearsyounger than the post herpetic neuralgia trials) thefrequency of side effects was similar Significant differ-ences between gabapentin and placebo were observed fordizziness (238 versus 49) and somnolence (226 versus62)

Mixed neuropathic painDespite the apparent efficacy of gabapentin in postherpetic neuralgia and diabetic neuropathy clinicalpractice is characterized by patients with a variety of

neuropathic pains These patients are often defined bytheir pain symptoms and signs rather than discretediseases To address this issue a further clinical trial ofgabapentin was undertaken that recruited 307 patientswith various neuropathic pain syndromes (excludingcancer) from pain clinics in the UK and Ireland49

Eligible patients were those with mean daily pain scoresof at least 4 on an 11 point numerical rating scale priorto entry Patients were also required to have at least 2 ofthe following symptoms and signs allodynia burningpain shooting pain hyperalgesia The median age ofrecruited patients was around 57 years in both groups

In this study patients were randomized to receiveplacebo or to be titrated up to receive 2400 mg ofgabapentin over five weeks to be followed by a three weekperiod where doses were stable By week 5 in thegabapentin arm 101 patients received 2400 mg 19received 1800 mg and 27 received 900 mg daily Themean weekly pain scores at week 8 showed significantdifferences for gabapentin In this group pain scores werereduced from 71 to 56 points on an 11 point numericalrating scale compared with a reduction from 73 to 63 inthe placebo group (PB0001) This difference wassignificant at the end of week 1 and remained similarfor most of the weeks throughout the study period Nosignificant difference was seen in the number of patientswho achieved more than 50 pain relief between the twogroups (21 for gabapentin and 14 for placebo)Interestingly significant improvements in burning painand hyperalgesia were seen but this was not observed forallodynia or shooting pain when gabapentin was com-pared with placebo Significant improvements were alsoseen after gabapentin in various quality of life measuresparticularly social and emotional functioning The mostfrequent side effects with gabapentin compared withplacebo were dizziness (242 versus 79) and somno-lence (144 versus 53)

Cancer neuropathic painTo date there have been no randomized trials ofgabapentin in patients with neuropathic pain due tocancer One open-label trial has been reported in which22 patients with cancer whose neuropathic pain was notcontrolled by opioids were given gabapentin as lsquoadd-onrsquotreatment50 The dose of gabapentin was titrated overthree to seven days up to a mean daily dose of around1004 mg (range 600iexcl1800 mg) Opioid doses wereunchanged throughout the study period Compared tobaseline scores mean global pain scores on a numericalrating scale were reduced from 64 to 32 points at thefinal assessment that took place 7iexcl14 days later Symp-toms of burning pain shooting pain and allodynia wereall also significantly reduced compared to baselineOverall no additional side effects were attributed to

8 MI Bennett and KH Simpson

gabapentin treatment apart from one case of sedationand one case of dizziness

This study was commented upon in a letter byChandler and Williams51 who argued that gabapentinhas a much lower incidence of efficacy in this contextbased on their experience They reported on a retro-spective review of 20 patients with neuropathic pain dueto cancer who were treated with gabapentin in addition totheir usual analgesia Only nine patients reported sig-nificant improvement and six others discontinued gaba-pentin because of side effects In the remaining fivepatients efficacy could not be established Furtheranalysis of the data in the Caraceni et al study50 showsthat 13 of 22 (59) patients achieved more than 50 painrelief compared to 45 (assuming lsquosignificant reliefrsquo isequivalent) in Chandler and Williamsrsquos series Both ofthese series compare favourably to larger randomizedcontrolled trials4749 which were performed in presum-ably fitter patients where between 21 and 34 achievedthis degree of pain relief

Comparative studies with other drugsA small number of studies exist that directly compare theeffectiveness of gabapentin against other analgesics in thetreatment of neuropathic pain Dallocchio et al 52

randomized 25 patients with diabetic neuropathy toreceive amitriptyline (mean dose 53 mg daily) or gaba-pentin (mean dose 1785 mg daily) in an open label trialDoses were titrated over four weeks and maintained foreight weeks Pain scores were measured on a five pointscale (0 frac34none 4frac34excruciating) Baseline pain scores inthe gabapentin group were reduced by a mean of 19(29iexcl10) and those in the amitriptyline group werereduced be a mean of 13 (275iexcl15) PB0026 Sig-nificant differences in the reduction in paraesthesiascores also favoured gabapentin Adverse events weremore frequent in the amitriptyline group and theselimited dose escalation according to the authors

However an earlier double blind randomized cross-over trial had not found any difference in efficacybetween the two drugs in patients with painful diabeticneuropathy53 In this study 21 of the 25 patients enrolledcompleted the six week trial of each drug with a one weekwashout period in between drug dosing (mean doses were59 mg amitriptyline and 1565 mg gabapentin) Pain scorediaries showed no statistical differences for patientstaking either treatment 11 (52) taking amitriptylineexperienced moderate or greater pain relief compared to14 (67) of those taking gabapentin

Eckhardt et al 54 have compared morphine gabapentinand combined treatment with both drugs in a doubleblind placebo controlled four way crossover study Thiswas conducted in 12 volunteers whose pain tolerance andpain threshold were measured during forearm immersionin cold water under experimental conditions Although

this was not a study of clinical neuropathic pain theresults are interesting Neither gabapentin nor morphinehad any significant effect on pain threshold but mor-phine had a significant effect on pain tolerance Im-portantly the combination of the two drugs significantlyraised pain threshold and tolerance (above that ofmorphine alone) This synergistic effect had been pre-viously demonstrated in rats41 It also supports thefindings of Caraceni et al 50 suggesting that gabapentinaugments the action of morphine via an unexplainedmechanism

Meta-analysesInformation from meta-analyses and systematic reviewsallows comparisons of effectiveness of drugs in variousneuropathic pain conditions These are expressed asNNT (number needed to treat) and NNH (numberneeded to harm)1655 56 In post herpetic neuralgia theNNT for gabapentin was 32 (CI 24iexcl50) whichcompares with 23 (CI 17iexcl33) for tricyclic antidepres-sants and 25 (CI 16iexcl51) for oxycodone In diabeticneuropathy the NNTs are gabapentin 38 (24iexcl87)tricyclic antidepressants 30 (23iexcl43) and carbamazepine23 (16iexcl38) Using the calculation for NNT suggestedby Cook and Sackett57 and the data for mixed neuro-pathic pain syndromes49 the NNT for gabapentin in thiscontext is 14

In diabetic neuropathy studies the NNH for minorevents is 25 (20iexcl32) for gabapentin 37 (24iexcl78) forcarbamazepine and 28 (20iexcl47) for tricyclic antidepres-sants The NNH for major events (drug related with-drawal from study) is 20iexcl40 this is broadly similar fortricyclic antidepressants and anticonvulsants (excludinggabapentin) In relation to gabapentin three of the fourmajor trials in neuropathic pain demonstrate no differ-ences between study withdrawals in those taking placebocompared with those taking active treatment4648 49 Inthe fourth study47 63 subjects withdrew whilst takingplacebo compared to 13 and 176 of those takinggabapentin 1800 and 2400 mg respectively these resultswere not assessed statistically It has been argued thatearlier studies of treatments in neuropathic pain (thathave contributed to meta-analyses) may have been lessable to detect and report adverse events This would leadto an overestimation of NNH for older anticonvulsantsand tricyclic antidepressants (ie they appear bettertolerated than is the case in clinical practice) whencompared with more recent studies of gabapentin

Summary

Gabapentin appears to be an effective treatment forneuropathic pain acting centrally to reduce hyperalgesiaand allodynia but it does not have significant actions as

Gabapentin in neuropathic pain 9

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

ogy47 Patients were randomized to receive 1800 mggabapentin or 2400 mg gabapentin or placebo Therationale for the study was to examine whether lowerdoses would be as efficacious as higher doses but wouldhave a lower incidence of side effects This trial excludedpatients who had previously failed to respond to gaba-pentin Patients taking active treatment were titrated upto 1800 mg gabapentin over two weeks and then up to2400 mg by week 3 for those randomized to receive thehigher dose All doses remained stable between weeks 3iexcl

7 Median age of patients recruited was around 75 yearsin all three groups Average daily pain score showedsignificant improvements after both 1800 and 2400 mggabapentin 65 to 43 and 65 to 42 respectively on an11 point numerical rating scale (PB0001) In compar-ison the reduction for the placebo group was 64 to 53points This translates to an 188 reduction in daily painscore using gabapentin instead of placebo These differ-ences were apparent from the end of the first week oftreatment Between 32 and 34 of patients takinggabapentin achieved more than 50 pain relief comparedto only 14 taking placebo Predominant side effects inthe treatment arms were dizziness (31iexcl33) and somno-lence (174iexcl204) Between 13 and 17 of patientstaking gabapentin withdrew because of side effects

Diabetic neuropathyAnother large randomized placebo controlled study ofgabapentin was conducted in patients with diabeticneuropathy48 attending pain clinics in the USA Of the165 patients recruited 84 were randomized to receivegabapentin and 81 to receive placebo The methodologyand dosing schedule was similar to that used in the trialby Rowbotham et al 46 Two thirds of the patients takinggabapentin were titrated to 3600 mg daily

This study also showed a significant reduction inaverage daily pain score for the gabapentin group from64 to 39 points on a numerical rating scale Thiscompares with a reduction from 65 to 51 in the placebogroup Quality of life measures sleep and mood allimproved significantly in those taking gabapentin and47 of the gabapentin group achieved a lsquomuch ormoderately improvedrsquo rank on a global change scorecompared with 25 of those receiving placebo Althoughthe median age in this study was 53 years (20 yearsyounger than the post herpetic neuralgia trials) thefrequency of side effects was similar Significant differ-ences between gabapentin and placebo were observed fordizziness (238 versus 49) and somnolence (226 versus62)

Mixed neuropathic painDespite the apparent efficacy of gabapentin in postherpetic neuralgia and diabetic neuropathy clinicalpractice is characterized by patients with a variety of

neuropathic pains These patients are often defined bytheir pain symptoms and signs rather than discretediseases To address this issue a further clinical trial ofgabapentin was undertaken that recruited 307 patientswith various neuropathic pain syndromes (excludingcancer) from pain clinics in the UK and Ireland49

Eligible patients were those with mean daily pain scoresof at least 4 on an 11 point numerical rating scale priorto entry Patients were also required to have at least 2 ofthe following symptoms and signs allodynia burningpain shooting pain hyperalgesia The median age ofrecruited patients was around 57 years in both groups

In this study patients were randomized to receiveplacebo or to be titrated up to receive 2400 mg ofgabapentin over five weeks to be followed by a three weekperiod where doses were stable By week 5 in thegabapentin arm 101 patients received 2400 mg 19received 1800 mg and 27 received 900 mg daily Themean weekly pain scores at week 8 showed significantdifferences for gabapentin In this group pain scores werereduced from 71 to 56 points on an 11 point numericalrating scale compared with a reduction from 73 to 63 inthe placebo group (PB0001) This difference wassignificant at the end of week 1 and remained similarfor most of the weeks throughout the study period Nosignificant difference was seen in the number of patientswho achieved more than 50 pain relief between the twogroups (21 for gabapentin and 14 for placebo)Interestingly significant improvements in burning painand hyperalgesia were seen but this was not observed forallodynia or shooting pain when gabapentin was com-pared with placebo Significant improvements were alsoseen after gabapentin in various quality of life measuresparticularly social and emotional functioning The mostfrequent side effects with gabapentin compared withplacebo were dizziness (242 versus 79) and somno-lence (144 versus 53)

Cancer neuropathic painTo date there have been no randomized trials ofgabapentin in patients with neuropathic pain due tocancer One open-label trial has been reported in which22 patients with cancer whose neuropathic pain was notcontrolled by opioids were given gabapentin as lsquoadd-onrsquotreatment50 The dose of gabapentin was titrated overthree to seven days up to a mean daily dose of around1004 mg (range 600iexcl1800 mg) Opioid doses wereunchanged throughout the study period Compared tobaseline scores mean global pain scores on a numericalrating scale were reduced from 64 to 32 points at thefinal assessment that took place 7iexcl14 days later Symp-toms of burning pain shooting pain and allodynia wereall also significantly reduced compared to baselineOverall no additional side effects were attributed to

8 MI Bennett and KH Simpson

gabapentin treatment apart from one case of sedationand one case of dizziness

This study was commented upon in a letter byChandler and Williams51 who argued that gabapentinhas a much lower incidence of efficacy in this contextbased on their experience They reported on a retro-spective review of 20 patients with neuropathic pain dueto cancer who were treated with gabapentin in addition totheir usual analgesia Only nine patients reported sig-nificant improvement and six others discontinued gaba-pentin because of side effects In the remaining fivepatients efficacy could not be established Furtheranalysis of the data in the Caraceni et al study50 showsthat 13 of 22 (59) patients achieved more than 50 painrelief compared to 45 (assuming lsquosignificant reliefrsquo isequivalent) in Chandler and Williamsrsquos series Both ofthese series compare favourably to larger randomizedcontrolled trials4749 which were performed in presum-ably fitter patients where between 21 and 34 achievedthis degree of pain relief

Comparative studies with other drugsA small number of studies exist that directly compare theeffectiveness of gabapentin against other analgesics in thetreatment of neuropathic pain Dallocchio et al 52

randomized 25 patients with diabetic neuropathy toreceive amitriptyline (mean dose 53 mg daily) or gaba-pentin (mean dose 1785 mg daily) in an open label trialDoses were titrated over four weeks and maintained foreight weeks Pain scores were measured on a five pointscale (0 frac34none 4frac34excruciating) Baseline pain scores inthe gabapentin group were reduced by a mean of 19(29iexcl10) and those in the amitriptyline group werereduced be a mean of 13 (275iexcl15) PB0026 Sig-nificant differences in the reduction in paraesthesiascores also favoured gabapentin Adverse events weremore frequent in the amitriptyline group and theselimited dose escalation according to the authors

However an earlier double blind randomized cross-over trial had not found any difference in efficacybetween the two drugs in patients with painful diabeticneuropathy53 In this study 21 of the 25 patients enrolledcompleted the six week trial of each drug with a one weekwashout period in between drug dosing (mean doses were59 mg amitriptyline and 1565 mg gabapentin) Pain scorediaries showed no statistical differences for patientstaking either treatment 11 (52) taking amitriptylineexperienced moderate or greater pain relief compared to14 (67) of those taking gabapentin

Eckhardt et al 54 have compared morphine gabapentinand combined treatment with both drugs in a doubleblind placebo controlled four way crossover study Thiswas conducted in 12 volunteers whose pain tolerance andpain threshold were measured during forearm immersionin cold water under experimental conditions Although

this was not a study of clinical neuropathic pain theresults are interesting Neither gabapentin nor morphinehad any significant effect on pain threshold but mor-phine had a significant effect on pain tolerance Im-portantly the combination of the two drugs significantlyraised pain threshold and tolerance (above that ofmorphine alone) This synergistic effect had been pre-viously demonstrated in rats41 It also supports thefindings of Caraceni et al 50 suggesting that gabapentinaugments the action of morphine via an unexplainedmechanism

Meta-analysesInformation from meta-analyses and systematic reviewsallows comparisons of effectiveness of drugs in variousneuropathic pain conditions These are expressed asNNT (number needed to treat) and NNH (numberneeded to harm)1655 56 In post herpetic neuralgia theNNT for gabapentin was 32 (CI 24iexcl50) whichcompares with 23 (CI 17iexcl33) for tricyclic antidepres-sants and 25 (CI 16iexcl51) for oxycodone In diabeticneuropathy the NNTs are gabapentin 38 (24iexcl87)tricyclic antidepressants 30 (23iexcl43) and carbamazepine23 (16iexcl38) Using the calculation for NNT suggestedby Cook and Sackett57 and the data for mixed neuro-pathic pain syndromes49 the NNT for gabapentin in thiscontext is 14

In diabetic neuropathy studies the NNH for minorevents is 25 (20iexcl32) for gabapentin 37 (24iexcl78) forcarbamazepine and 28 (20iexcl47) for tricyclic antidepres-sants The NNH for major events (drug related with-drawal from study) is 20iexcl40 this is broadly similar fortricyclic antidepressants and anticonvulsants (excludinggabapentin) In relation to gabapentin three of the fourmajor trials in neuropathic pain demonstrate no differ-ences between study withdrawals in those taking placebocompared with those taking active treatment4648 49 Inthe fourth study47 63 subjects withdrew whilst takingplacebo compared to 13 and 176 of those takinggabapentin 1800 and 2400 mg respectively these resultswere not assessed statistically It has been argued thatearlier studies of treatments in neuropathic pain (thathave contributed to meta-analyses) may have been lessable to detect and report adverse events This would leadto an overestimation of NNH for older anticonvulsantsand tricyclic antidepressants (ie they appear bettertolerated than is the case in clinical practice) whencompared with more recent studies of gabapentin

Summary

Gabapentin appears to be an effective treatment forneuropathic pain acting centrally to reduce hyperalgesiaand allodynia but it does not have significant actions as

Gabapentin in neuropathic pain 9

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

gabapentin treatment apart from one case of sedationand one case of dizziness

This study was commented upon in a letter byChandler and Williams51 who argued that gabapentinhas a much lower incidence of efficacy in this contextbased on their experience They reported on a retro-spective review of 20 patients with neuropathic pain dueto cancer who were treated with gabapentin in addition totheir usual analgesia Only nine patients reported sig-nificant improvement and six others discontinued gaba-pentin because of side effects In the remaining fivepatients efficacy could not be established Furtheranalysis of the data in the Caraceni et al study50 showsthat 13 of 22 (59) patients achieved more than 50 painrelief compared to 45 (assuming lsquosignificant reliefrsquo isequivalent) in Chandler and Williamsrsquos series Both ofthese series compare favourably to larger randomizedcontrolled trials4749 which were performed in presum-ably fitter patients where between 21 and 34 achievedthis degree of pain relief

Comparative studies with other drugsA small number of studies exist that directly compare theeffectiveness of gabapentin against other analgesics in thetreatment of neuropathic pain Dallocchio et al 52

randomized 25 patients with diabetic neuropathy toreceive amitriptyline (mean dose 53 mg daily) or gaba-pentin (mean dose 1785 mg daily) in an open label trialDoses were titrated over four weeks and maintained foreight weeks Pain scores were measured on a five pointscale (0 frac34none 4frac34excruciating) Baseline pain scores inthe gabapentin group were reduced by a mean of 19(29iexcl10) and those in the amitriptyline group werereduced be a mean of 13 (275iexcl15) PB0026 Sig-nificant differences in the reduction in paraesthesiascores also favoured gabapentin Adverse events weremore frequent in the amitriptyline group and theselimited dose escalation according to the authors

However an earlier double blind randomized cross-over trial had not found any difference in efficacybetween the two drugs in patients with painful diabeticneuropathy53 In this study 21 of the 25 patients enrolledcompleted the six week trial of each drug with a one weekwashout period in between drug dosing (mean doses were59 mg amitriptyline and 1565 mg gabapentin) Pain scorediaries showed no statistical differences for patientstaking either treatment 11 (52) taking amitriptylineexperienced moderate or greater pain relief compared to14 (67) of those taking gabapentin

Eckhardt et al 54 have compared morphine gabapentinand combined treatment with both drugs in a doubleblind placebo controlled four way crossover study Thiswas conducted in 12 volunteers whose pain tolerance andpain threshold were measured during forearm immersionin cold water under experimental conditions Although

this was not a study of clinical neuropathic pain theresults are interesting Neither gabapentin nor morphinehad any significant effect on pain threshold but mor-phine had a significant effect on pain tolerance Im-portantly the combination of the two drugs significantlyraised pain threshold and tolerance (above that ofmorphine alone) This synergistic effect had been pre-viously demonstrated in rats41 It also supports thefindings of Caraceni et al 50 suggesting that gabapentinaugments the action of morphine via an unexplainedmechanism

Meta-analysesInformation from meta-analyses and systematic reviewsallows comparisons of effectiveness of drugs in variousneuropathic pain conditions These are expressed asNNT (number needed to treat) and NNH (numberneeded to harm)1655 56 In post herpetic neuralgia theNNT for gabapentin was 32 (CI 24iexcl50) whichcompares with 23 (CI 17iexcl33) for tricyclic antidepres-sants and 25 (CI 16iexcl51) for oxycodone In diabeticneuropathy the NNTs are gabapentin 38 (24iexcl87)tricyclic antidepressants 30 (23iexcl43) and carbamazepine23 (16iexcl38) Using the calculation for NNT suggestedby Cook and Sackett57 and the data for mixed neuro-pathic pain syndromes49 the NNT for gabapentin in thiscontext is 14

In diabetic neuropathy studies the NNH for minorevents is 25 (20iexcl32) for gabapentin 37 (24iexcl78) forcarbamazepine and 28 (20iexcl47) for tricyclic antidepres-sants The NNH for major events (drug related with-drawal from study) is 20iexcl40 this is broadly similar fortricyclic antidepressants and anticonvulsants (excludinggabapentin) In relation to gabapentin three of the fourmajor trials in neuropathic pain demonstrate no differ-ences between study withdrawals in those taking placebocompared with those taking active treatment4648 49 Inthe fourth study47 63 subjects withdrew whilst takingplacebo compared to 13 and 176 of those takinggabapentin 1800 and 2400 mg respectively these resultswere not assessed statistically It has been argued thatearlier studies of treatments in neuropathic pain (thathave contributed to meta-analyses) may have been lessable to detect and report adverse events This would leadto an overestimation of NNH for older anticonvulsantsand tricyclic antidepressants (ie they appear bettertolerated than is the case in clinical practice) whencompared with more recent studies of gabapentin

Summary

Gabapentin appears to be an effective treatment forneuropathic pain acting centrally to reduce hyperalgesiaand allodynia but it does not have significant actions as

Gabapentin in neuropathic pain 9

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

an anti-nociceptive agent Based on large clinical trials ithas similar (though not superior) efficacy to tricyclicantidepressants and carbamazepine but it may be bettertolerated particularly in relation to major harms Evi-dence suggests that when receiving treatment for diabeticneuropathy and post herpetic neuralgia with gabapentinaround 30 of patients can expect to achieve more than50 pain relief The mean reduction in pain score forthose taking gabapentin is 205 points on an 11 pointnumerical rating scale compared with a reduction of 094points for those taking placebo when the results of fourmajor trials are combined46 iexcl 49 Around 30 of patientstreated with gabapentin will experience minor adverseevents the most common of which are somnolence anddizziness In neuropathic pain due to cancer evidencesuggests that higher response rates might be observedwith gabapentin when administered with opioids How-ever the incidence of adverse effects in this patientpopulation has not been assessed and would need to bebalanced against any advantage of this combination

References

1 Blom S Trigeminal neuralgia its treatment with a newanticonvulsant drug Lancet 1962 i 839 iexcl40

2 Swerdlow M Anticonvulsants in the therapy of neur-algic pain Pain Clin 1986 1 9iexcl19

3 Lambet A Laduron P Mourre C Jacomet Y LazdunskiM Axonal transport of the voltage dependent Naraquo

channel protein identi ed by its tetrodotoxin binding sitein rat sciatic nerves Brain Res 1985 345 153 iexcl58

4 Devor M Govrin-Lippman R Angelides K Naraquo

channel immunolocalisation in peripheral mammalianaxons and changes following nerve injury and neuromaformation J Neuroscience 1993 13 1976 iexcl92

5 Moll C Mourre C Lazdunski M Ulrich J Increase ofsodium channels in demyelinated lesions of multiplesclerosis Brain Res 1991 556 311iexcl16

6 Catteral WA Common modes of drug action on sodiumchannels local anaesthetics antiarrhythmics and antic-onvulsants Trends Pharmacol Sci 1987 8 57iexcl65

7 McQuay HJ Pharmacological treatment of neuralgicand neuropathic pain Cancer Surv 1988 7 141 iexcl59

8 Yaari Y Devor M Phenytoin suppresses spontaneousectopic discharge in rat sciatic nerve neuromas NeurosciLett 1985 58 117 iexcl22

9 Budd K Sodium valproate in the treatment of pain InChadwick D ed Fourth international symposium onsodium vaplroate and epilepsy London Royal Societyof Medicine 1989 213 iexcl16

10 Raftery H The management of post herpetic neuralgiausing sodium valproate and amitriptyline J Irish MedAssoc 1979 72 399 iexcl401

11 Peiris JB Perera GL Devendra SV Lionel ND Sodiumvalproate in trigeminal neuralgia Med J Aust 1980 2278

12 Reddy S Patt R The benzodiazepines as adjuvantanalgesics J Pain Symptom Manage 1994 9 510iexcl14

13 Fromm GH Baclofen as an adjuvant analgesic J PainSymptom Manage 1994 9 500 iexcl509

14 Fromm GH Terrence CF Chattha AS Baclofen in thetreatment of trigeminal neuralgia double blind studyand long term follow-up Ann Neurol 1984 15 240 iexcl44

15 McQuay H Carroll D Jadad A Wiffen P Moore AAnticonvulsant drugs for management of pain a sys-tematic review BMJ 1995 311 1047 iexcl52

16 Wiffen P Collins S McQuay H Corroll D Jadad AMoore A Anticonvulsant drugs for acute and chronicpain Cochrane Database System Rev 2000 2CD001133

17 Elwes RD Binnie CD Clinical pharmacokinetics of thenewer antiepileptic drugs Lamotrigine vigabatrin ga-bapentin and oxcarbazepine Clin Pharmacokinet 199630 403 iexcl15

18 Vollmer KO Anhut H Thomann P Wagner F JahnckenD Pharmacokinetic model and absolute bioavailabilityof the new anticonvulsant gabapentin Adv Epileptol1989 17 209 iexcl11

19 Turck D Vollmer KO Bockbrader H Sedman A Doselinearity of the new anticonvulsant gabapentin aftermultiple oral doses Eur J Clin Pharmacol 1989 36(suppl) A310

20 Boyd RA Turck D Abel RB Sedman AJ BockbraderHN Effects of age and gender on single dose pharma-cokinetics of gabapentin Epilepsia 1999 40 474 iexcl9

21 Blum RA Comstock TJ Sica DA Schultz RW Keller EReetze P Bockbrader H Pharmacokinetics of gapapen-tin in subjects with various degrees of renal impairmentClin Pharmacol Ther 1994 56 154iexcl59

22 MacDonald RL Olsen RW GABAA recpetor channelsAnn Rev Neurosci 1994 17 569iexcl602

23 Taylor CP Gabapentin mechansims of action In LevyRH Mattson RH Meldrum BS eds Antiepileptic drugs New York Raven Press 1995 829iexcl41

24 Suzdak PD Jansen JA A review of preclinical pharma-cology of tiagabine a potent and selective anticonvulsantGABA uptake inhibitor Epilepsia 1995 36 612 iexcl26

25 Fichter N Taylor CP Feuerstein TJ Nipecotate-inducedGABA release from slices of rat caudate-putameneffects of gapapentin NiexclS Arch Pharmacol 1996 34535

26 Kocsis JD Honmou O Gabapentin increases GABA-induced depolarisation in rat neonatal optic nerveNeurosci Lett 1994 169 181 iexcl84

27 Kaneko M Mestre C Sanchez EH Hammond DLIntrathecally administered gabapentin inhibits formalin-evoked nociception and the expression of Fos-likeimmunoreactivity in the spinal cord of the rat JPharmacol Exp Ther 2000 292 743 iexcl51

28 Chen S-R Eisenach JC McCaslin PP Pan H-LSynergistic effect between intrathecal non-NMDA an-tagonist and gabapentin on allodynia induced by spinalnerve ligation in rats Anesthesiology 2000 92 500 iexcl506

29 Gee NS Brown JP Dissanayake VU Offord J ThurlowR Woodruff GN The novel anticonvulsant drug gaba-

10 MI Bennett and KH Simpson

pentin binds to the a2d subunit of a calcium channel JBiol Chem 1996 271 5768 iexcl76

30 Field MJ Oles RJ Lewis AS McCleary S Hughes JSingh L Gabapentin (Neurontin) and S-(raquo)-3-isobutyl-gaba represent a novel class of selective antihyperalgesicagents Br J Pharmacol 1997 121 1513 iexcl22

31 Shimoyama M Shimoyama N Hori Y Gabapentinaffects glutamatergic excitatory neurotransmission inthe rat dorsal horn Pain 2000 85 405iexcl14

32 Dixit R Bhargava VK Kaur N Antinociceptive effect ofgabapentin and clinical pharmacology Methods FindExp Clin Pharmacol 1999 21 481 iexcl82

33 Yoon MH Yaksh TL Evaluation of interaction betweengabapentin and ibuprofen on the formalin test in ratsAnesthesiology 1999 91 1006 iexcl13

34 Abdi S Lee DH Chung JM The antiallodynic effects ofamitriptyline gabapentin and lidocaine in a rat model ofneuropathic pain Anesth Analg 1998 871360 iexcl66

35 Pan HL Eisenach JC Chen SR Gabapentin suppressesectopic nerve discharges and reverses allodynia inneuropathic rats J Pharmacol Exp Ther 1999 2881026iexcl30

36 Rock DM Kelly KM Macdonald RL Gabapentinactions on ligand- and voltage-gated responses incultured rodent neurons Epilepsy Res 1993 16 89iexcl98

37 Wamil AW McLean MJ Limitation by gabapentin ofhigh frequency action potential ring by mouse centralneurons in cell culture Epilepsy Res 1994 17 1iexcl11

38 Welty DF Shielke GP Vartanian MG Taylor CPGabapentin anticonvulsant action in rats disequilibriumwith peak drug concentrations in plasma and brainmicrodialysate Epilepsy Res 1993 16 175 iexcl81

39 Field MJ McCleary S Hughes J Singh L Gabapentinand pregabalin but not morphine and amitriptylineblock both static and dynamic components of mechan-ical allodynia induced by streptozocin in the rat Pain1999 80 391 iexcl98

40 Rocha L Ondarza-Rovira R Maidment NT Gabapen-tin modi es extracellular opioid content in amygdala amicrodialysis study Epilepsy Res 1999 35 13iexcl20

41 Shimoyama M Shimoyama N Inturissi CE Elliott KJGabapentin enhances the antinociceptive effects ofspinal morphine in the rat tail ick test Pain 1997 72375 iexcl82

42 Rosner H Rubin L Kestenbaum A Gabapentin ad-junctive therapy in neuropathic pain states Clin J Pain1996 13 251 iexcl55

43 Mellick GA Mellick LB Re ex sympathetic dystrophytreated with gabapentin Arch Phys Med Rehabil 199778 98iexcl105

44 Hays H Woodroffe MA Using gabapentin to treatneuropathic pain Can Fam Physician 1999 45 2109 iexcl12

45 Holtrom N Gabapentin for treatment of thalamic painsyndrome Palliat Med 2000 14 167

46 Rowbotham M Harden N Stacey B Bernstein PMagnus-Miller L Gabapentin for the treatment ofpostherpetic neuralgia JAMA 1998 280 1837iexcl42

47 Rice AS Maton S Postherpetic Neuralgia Study GroupGabapentin in postherpetic neuralgia a randomiseddouble blind placebo controlled study Pain 2001 94215iexcl24

48 Backonja M Beydoun A Edwards KR Schwartz SLFonseca V Hes M LaMoreaux L Garofalo E Gaba-pentin for the symptomatic treatment of painful neuro-pathy in patients with diabetes mellitus JAMA 1998280 1831 iexcl36

49 Serpell M Neuropathic Pain Study Group Gabapentinin neuropathic pain syndromes a randomised double-blind placebo-controlled trial Pain 2002 99 557 iexcl66

50 Caraceni A Zecca E Martini C de Conno F Gaba-pentin as an adjuvant to opioid analgesia for neuro-pathic cancer pain J Pain Symptom Manage 1999 17441iexcl45

51 Chandler A Williams JE Gabapentin an adjuvanttreatment for neuropathic pain in a cancer hospital JPain Symptom Manage 2000 20 82iexcl83

52 Dallocchio C Buffa C Mazzarello P Chiroli S Gaba-pentin versus amitriptyline in painful diabetic neuro-pathy an open label pilot study J Pain SymptomManage 2000 20 280 iexcl85

53 Morello CM Leckband SG Stoner CP Moorhouse DFSahagian GA Randomised double-blind study compar-ing the ef cacy of gabapentin with amitriptyline ondiabetic peripheral neuropathy pain Arch Intern Med1999 159 1931 iexcl37

54 Eckhardt K Ammon S Hofmann U Riebe A GugelerN Mikus G Gabapentin enhances the analgesic effect ofmorphine in healthy volunteers Anesth Analg 2000 91185iexcl91

55 Sindrup SH Jensen TS Ef cacy of pharmacologicaltreatments of neuropathic pain an update and effectrelated to mechanism of drug action Pain 1999 83389iexcl400

56 McQuay H Tramer M Nye BA Carroll D Wiffen PMoore RA A systematic review of antidepressants inneuropathic pain Pain 1996 68 217 iexcl27

57 Cook RJ Sackett DL The number needed to treat aclinically useful measure of treatment effect BMJ 1995310 452 iexcl54

Gabapentin in neuropathic pain 11