4gilr.oa

The new england journal of medicine Morphine, Gabapentin, or Their Combination Ian Gilron, M.D., Joan M. Bailey, R.N., M.Ed., Dongsheng Tu, Ph.D., Ronald R. Holden, Ph.D., Donald F. Weaver, M.D., Ph.D., b a c k g r o u n d
The available drugs to treat neuropathic pain have incomplete efficacy and dose-limit- ing adverse effects. We compared the efficacy of a combination of gabapentin and mor- icology (I.G.), the Departments of Mathe-matics and Statistics and Community phine with that of each as a single agent in patients with painful diabetic neuropathy or partment of Psychology (R.R.H.), and theDepartment of Medicine (Division of En-docrinology) (R.L.H.), Queen’s Universi- In this randomized, double-blind, active placebo–controlled, four-period crossover trial, patients received daily active placebo (lorazepam), sustained-release morphine, Chemistry (D.F.W.), Dalhousie University,Halifax, N.S. — both in Canada. Address gabapentin, and a combination of gabapentin and morphine — each given orally for reprint requests to Dr. Gilron at the Depart- five weeks. The primary outcome measure was mean daily pain intensity in patients re- ment of Anesthesiology, Queen’s Univer- ceiving a maximal tolerated dose; secondary outcomes included pain (rated according sity, 76 Stuart St., Kingston, ON K7L 2V7,Canada, or at gilroni@post.queensu.ca.
to the Short-Form McGill Pain Questionnaire), adverse effects, maximal tolerated dos-es, mood, and quality of life.
N Engl J Med 2005;352:1324-34.
Copyright 2005 Massachusetts Medical Society. Of 57 patients who underwent randomization (35 with diabetic neuropathy and 22 withpostherpetic neuralgia), 41 completed the trial. Mean daily pain (on a scale from 0 to10, with higher numbers indicating more severe pain) at a maximal tolerated dose ofthe study drug was as follows: 5.72 at baseline, 4.49 with placebo, 4.15 with gabapen-tin, 3.70 with morphine, and 3.06 with the gabapentin–morphine combination (P<0.05for the combination vs. placebo, gabapentin, and morphine). Total scores on the Short-Form McGill Pain Questionnaire (on a scale from 0 to 45, with higher numbers indicat-ing more severe pain) at a maximal tolerated dose were 14.4 with placebo, 10.7 withgabapentin, 10.7 with morphine, and 7.5 with the gabapentin–morphine combination(P<0.05 for the combination vs. placebo, gabapentin, and morphine). The maximal tol-erated doses of morphine and gabapentin were lower (P<0.05) with the combinationthan for each drug as single agent. At the maximal tolerated dose, the gabapentin–mor-phine combination resulted in a higher frequency of constipation than gabapentin alone(P<0.05) and a higher frequency of dry mouth than morphine alone (P<0.05).
c o n c l u s i o n s
Gabapentin and morphine combined achieved better analgesia at lower doses of eachdrug than either as a single agent, with constipation, sedation, and dry mouth as themost frequent adverse effects.
Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n pared the combination of gabapentin and morphine plication of cancer, diabetes mellitus, de- with each drug used as a single agent in patients with ngenerative spine disease, infection with diabetic neuropathy or postherpetic neuralgia.
the human immunodeficiency virus, the acquiredimmunodeficiency syndrome, and other infectious diseases, and it has a profound effect on quality of
life and expenditures for health care.1 Gabapentin participants
and opioids have been proposed as two of several Our trial was approved by an institutional ethics re-
first-line treatments for neuropathic pain.2 How- view board. Patients were recruited between Febru-
ever, the maximal tolerated doses of these drugs, ary 2001 and November 2003 by means of adver-
administered as single agents, reduce pain by only tisements and physician referrals and were enrolled
26 to 38 percent, owing to incomplete efficacy, dose- after giving written informed consent for participa-
limiting adverse effects, or both.3-6 The combina- tion. Patients with diabetic neuropathy had distal,
tion of mechanistically distinct analgesic agents symmetric, sensory diabetic polyneuropathy as de-
may result in additivity or synergism and may im- termined on the basis of their medical history and
prove efficacy at lower doses, with fewer side effects either an unequivocal decrease in response to pin-
than with the use of one agent alone. This strate- prick, temperature, or vibration in both feet or bi-
gy has been advocated in cases of partial treatment laterally decreased or absent ankle-jerk reflexes.
response, though, admittedly, in the absence of Patients with postherpetic neuralgia had had an
rigorous supportive evidence.2 Gabapentin is a eruption of herpes zoster rash not more recently
3-alkylated analogue of g-amino butyric acid, which than six months before enrollment. General crite-
modulates a d calcium-channel subunits, a mecha- ria for inclusion were daily moderate pain for three
nism thought to be important in neuropathic pain.7 months or more, an age of 18 to 89 years, a serumGabapentin analgesia is unaffected by opioid an- alanine aminotransferase or aspartate aminotrans-tagonism, and repeated administration of gaba- ferase level less than 1.2 times the normal level, apentin does not lead to analgesic tolerance.8 Fur- creatinine level less than 1.5 times the upper limitthermore, preclinical studies suggest that additive of the normal range, and sufficient language skillsinteractions may occur between gabapentin and to communicate with research staff. Exclusion cri-morphine9-11 and that opioid tolerance can be pre- teria were hypersensitivity to study medications,vented by the use of gabapentin.12 another painful condition as severe as the diabetic Common adverse effects associated with mor- neuropathy or postherpetic neuralgia, a recent my- phine include respiratory depression, sedation, nau- ocardial infarction, unstable angina or congestive
sea and vomiting, constipation, and pruritus.13 heart failure, any central neurologic disorder (in-
Common adverse effects associated with gabapen- cluding seizures), a serious mood disorder, a histo-
tin include sedation, ataxia, and dizziness.14 Except ry of serious drug or alcohol abuse, pregnancy, lac-
for sedation, most opioid-related adverse effects tation, and lack of a primary care physician.
rarely occur with gabapentin, suggesting that most
adverse effects would not “overlap” if the drug were study design
used in combination with morphine. Although se- In our single-center, four-period, crossover, ran-
dation is an effect of both drugs, it is mediated only domized trial, we compared four treatments (with
supraspinally, whereas both these drugs have been each treatment given for five weeks) — morphine
shown to have analgesic effects at supraspinal, spi- (M-Eslon, Aventis-Pharma), gabapentin (Neuron-
nal, and even peripheral sites of action.15-17 Thus, tin, Pfizer), these drugs in combination, and active
a combination of gabapentin and morphine may placebo (low-dose lorazepam). Benzodiazepines
provide more additivity for analgesia than for se- have no efficacy in neuropathic pain,20 yet their sed-
dation.
ative effects provide more effective blinding than Painful diabetic neuropathy and postherpetic those of inert placebo in trials of sedating analge- neuralgia are two neuropathic pain syndromes that sics.21,22 With the use of a balanced Latin-squarehave been investigated in mechanism-based stud- crossover design,23 patients were allocated, in aies18,19 as well as in many clinical trials of analgesic double-blind, randomized fashion, to one of fouragents. Both conditions have been shown to re- treatment sequences: morphine, placebo, gabapen-spond to opioids3,4 and to gabapentin.5,6 We com- tin, and the gabapentin–morphine combination; Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine placebo, the gabapentin–morphine combination, tained sustained-release morphine (15 mg) and graymorphine, and gabapentin; gabapentin, morphine, capsules contained lactose placebo; for treatmentthe gabapentin–morphine combination, and pla- with the gabapentin–morphine combination, bluecebo; and the gabapentin–morphine combination, capsules contained sustained-release morphine (15gabapentin, placebo, and morphine. At the com- mg) and gray capsules contained gabapentin (300mencement of the trial, a pharmacist at the Kingston mg); for the gabapentin treatment, blue capsulesGeneral Hospital in Kingston, Ontario, Canada, contained lactose placebo and gray capsules con-prepared a concealed allocation schedule randomly tained gabapentin (300 mg); for the placebo treat-assigning the four sequences, in blocks of four, to a ment, blue capsules contained lorazepam (0.2 mg)consecutive series of numbers. On enrollment, each and gray capsules contained lorazepam (0.1 mg).
patient was assigned to the next consecutive num- Among these older or smaller patients, the targetber, and the corresponding series of study medica- daily-dose ceilings were morphine at a dose of 60tions was dispensed.
mg (morphine treatment); morphine at a dose of Medications were placed in blue and gray gela- 60 mg and gabapentin at a dose of 2400 mg (treat- tin capsules by the investigational pharmacist in ment with the gabapentin–morphine combination);
order to maintain double-blind conditions. Patients gabapentin at a dose of 2400 mg (gabapentin treat-
received identical-appearing blue and gray capsules ment); and lorazepam at a dose of 1.6 mg (placebo
during each treatment regimen in accord with a treatment).
double-dummy design. Blue capsules were admin-
istered twice daily, and gray capsules three times protocol
daily. For the morphine treatment, blue capsules Patients completed a baseline diary in which they
contained sustained-release morphine (30 mg) and rated the intensity of pain three times a day for sev-
gray capsules contained lactose placebo; for the en consecutive days after discontinuing treatment
treatment with the gabapentin–morphine combi- with previously prescribed opioids or gabapentin.
nation, blue capsules contained sustained-release A similar daily pain diary was kept throughout the
morphine (15 mg) and gray capsules contained gab- study. Nonopioid drugs other than gabapentin were
apentin (300 mg); for the gabapentin treatment, permitted at a steady dose throughout the trial;
blue capsules contained lactose placebo and gray procedural pain therapies (e.g., nerve blocks) were
capsules contained gabapentin (400 mg); and for forbidden. The schedule of dose escalation for each
the placebo treatment, blue capsules contained lor- set of capsules was identical in each treatment peri-
azepam (0.2 mg) and gray capsules contained lor- od. During the first three weeks of each five-week
azepam (0.1 mg).
treatment period, the dose was escalated toward a The target daily-dose ceilings were morphine maximal tolerated dose or the target ceiling dose, at a dose of 120 mg (morphine treatment), mor- whichever was reached first. In the fourth week ofphine at a dose of 60 mg and gabapentin at a dose each period, each patient received his or her maxi-of 2400 mg (gabapentin–morphine combination mal tolerated dose for the particular treatment.
treatment), gabapentin at a dose of 3200 mg (gab- During the fifth week of each treatment period, pa-apentin treatment), and lorazepam at a dose of tients underwent a four-day dose tapering and a1.6 mg (placebo treatment). To facilitate the grad- three-day complete washout.
ual titration, for some doses there was a slightly A research nurse telephoned patients twice uneven distribution of the number of capsules with- weekly to evaluate adverse effects and guide drugin a day — for example, in the morning, one blue titration. With each increase in the dose of studycapsule and one gray capsule might be adminis- medications, adverse effects were rated (as mild,tered; at midday, two gray capsules; and at bedtime, moderate, or severe), and patients were asked wheth-one blue capsule and two gray capsules.
er they could tolerate the increased dose for an- In the expectation that older patients and those other two to three days. If they could, this dose was with smaller body size might tolerate a less steep continued in the expectation that side effects woulddose titration with smaller increments than patients diminish. If, however, the side effects were intoler-no more than 60 years of age or with larger body able or did not diminish, the study medicationssize, adjustments were made for those more than were decreased and an increase was attempted one60 years of age and those weighing less than 60 kg: more time, at the next telephone call. If this next in-for the morphine treatment, blue capsules con- crease again resulted in intolerable side effects, the Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n study drugs were decreased to the level of the previ- On the basis of previous estimates of variance,3 andous dose, which was defined as the maximal tol- accounting for five pairwise comparisons (i.e., pla-erated dose. Patients were given docusate sodium cebo vs. each treatment plus treatment with the(100 to 300 mg per day) as prophylaxis against con- gabapentin–morphine combination vs. each singlestipation. Those in whom constipation developed agent), we calculated that 40 patients would be re-during any treatment period were given sennosides quired to provide the study with 80 percent power(17 to 34 mg twice daily) in addition.
to detect (with a two-sided alpha of 0.05) a meandifference in pain intensity at week 4 among the o u t c o m e m e a s u r e s
treatments that was equivalent to 1 point on a scale The primary outcome measure was the mean in- from 0 to 10.29 On the basis of previous dropouttensity of pain (on a scale from 0 to 10, with 0 indi- rates22 of approximately 10 percent during treat-cating no pain and 10 indicating “the worst pain ment periods of four to six weeks, we calculated thatimaginable”), which patients rated three times dai- if 58 patients were enrolled, 40 would complete allly. Ratings were averaged over seven days in which four treatment periods. Patients who completed atthe patients were receiving the maximal tolerated least two treatment periods (allowing for one pair-dose of the study drug. Secondary outcomes were wise comparison) were included in the efficacyadverse effects, including major adverse events; analysis. Patients who received at least one dose ofpain assessed according to responses on the Short- any study medication were included in the analysisForm McGill Pain Questionnaire (on a scale from of adverse effects. 0 to 45, with higher numbers indicating more se- Before the analyses were performed, we defined vere pain)24; pain-related interference assessed ac- the intensity of pain as an average of the scores forcording to the Brief Pain Inventory (on a scale from pain in the patient’s diary for week 4 in each treat-0 to 10, with 0 indicating no interference and 10 in- ment period if no more than 50 percent of the scoresdicating complete interference)25; mood assessed were missing. If more than 50 percent of the scoresaccording to the Beck Depression Inventory (on a were missing, the mean daily score for pain intensi-scale from 0 to 63, with higher numbers indicat- ty was considered to be missing. A linear mixeding more severe depression)26; health status as- model30 in which the drug regimen, sequence, treat-sessed according to the Medical Outcomes Study ment period, and the first-order carryover term (the36-Item Short-Form General Health Survey (SF-36 carrying over of treatment to the next treatment pe-Health Survey) (on a scale from 0 to 100, with high- riod — i.e., from period A to period B but not to pe-er numbers indicating better health-related quali- riod C or D)30 were the fixed effects and the patientty of life)27; mental status assessed according to the (nested in the sequence of treatment periods) wasMini–Mental State Examination (on a scale from 0 to the random effect was fitted with the pain-intensity30, with lower numbers indicating impaired men- data. If the carryover effect was not significant, thental status)28; and global pain relief assessed in re- a reduced model that excluded the carryover termsponse to questions from the research nurse on the was refitted. The least-square means and associatedfollowing scale: pain worse, no relief, slight relief, standard errors estimated from the initial model ormoderate relief, a lot of relief, or complete relief. the reduced model were calculated for each treat-The assessments were made at baseline and during ment. We used Fisher’s least-significant-differenceeach treatment period at the point when patients method to calculate multiple comparisons amongwere receiving the maximal tolerated dose. A “blind- treatments.31 According to this method, the globaling” questionnaire to assess which treatments the difference among all treatments was first tested inpatients and the research nurse thought the patients the model. Only when this test was significant at thewere receiving was completed by both patients and 0.05 level were pairwise comparisons made, also atthe nurse when the patients were taking the maxi- the 0.05 level, with the use of the estimated contrastmal tolerated dose of the assigned study drug.
from the initial or the reduced model (i.e., the gab-apentin–morphine combination vs. gabapentin as a n a l y s i s
a single agent, the combination vs. morphine as a The preplanned primary analysis of the primary single agent, the combination vs. placebo, gabapen-outcome was to compare patients’ mean scores for tin vs. placebo, and morphine vs. placebo).
pain while taking the maximal tolerated dose of the As a sensitivity analysis, the level of change in assigned drug during week 4 across all treatments. the intensity of pain during each treatment period Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine was calculated as the difference between the score three treatment periods. No patients were exclud-for pain at baseline (the mean of the last three days ed from the analysis because of missing data. Week-of the pretrial baseline before treatment period A ly averages of daily pain scores for each treatmentstarted or at washout for treatment periods B, C, sequence are shown in Figure 2A. The primary analy-and D) and the scores for pain during treatment sis showed no significant main effects of either se-(the mean of the last three days while the patient quence or treatment period, but the effects of drugwas taking the maximal tolerated dose). The per- treatment (P<0.001) and carryover (P=0.04) werecent change was calculated as the change in pain statistically significant. An exploratory analysis ofdivided by the score for pain at baseline times 100 all differences in pairwise carryover effects showedpercent. These estimates were analyzed in the lin- a difference only between morphine and placeboear mixed model. Secondary continuous outcome (P=0.005); the effect of morphine is more likelymeasures were analyzed in a similar fashion. Data to carry over to the next treatment period than thaton proportions were analyzed with the use of Fish- of placebo. er’s exact test, with the use of Fisher’s least-signifi- In the linear mixed model, all treatment con- cant-difference method for multiple comparisons.32 trasts were adjusted for all observed carryover ef-A preplanned interim analysis without stopping fects. Mean pain intensity on a scale from 0 to 10rules was performed and reviewed approximately at baseline and at the maximal tolerated dose washalfway through the trial.
calculated as follows: mean (±SE) at baseline, 5.72± All reported P values are two-sided. All statisti- 0.23; with placebo, 4.49±0.34; gabapentin, 4.15± cal analyses were conducted with the use of SAS 0.33; morphine, 3.70±0.34; and the gabapentin–software (version 8.0, SAS Institute). All authors morphine combination, 3.06±0.33 (Fig. 2B). Painvouch for the veracity and completeness of the re- treated with the combination was rated lower thanported data, and all authors contributed to various pain treated with morphine alone (P=0.04), gaba-aspects of the trial design, data gathering and analy- pentin alone (P<0.001), or placebo (P<0.001). Thesis, and preparation of the manuscript.
analysis of the percent change in pain intensity in-dicated greater reduction of pain with the use ofthe gabapentin–morphine combination than with placebo (20.4 percent greater reduction, P=0.03).
s u b j e c t s
Other comparisons were not significant.
Eighty-six patients underwent screening assess-
ment in the research clinic. Of these, 29 were ex- secondary outcomes
cluded, 57 underwent randomization, and a total The mean maximal tolerated dose of morphine
of 16 withdrew during the treatment periods — (Fig. 2C) was 45.3±3.9 mg as a single agent, as com-
13 before completing the second treatment peri- pared with 34.4±2.6 mg of morphine in the gaba-
od (period B), and 3 because of adverse effects but pentin–morphine combination (P<0.05). The mean
after completing at least two treatment periods maximal tolerated dose of gabapentin (Fig. 2C) was
(these 3 were therefore included in the efficacy 2207±89 mg as a single agent, as compared with
analysis). Forty-one patients completed the trial 1705 ± 83 mg in the combination (P<0.05). The
(Fig. 1). Table 1 lists the demographic and base- mean maximal tolerated dose of lorazepam (active
line characteristics of patients who underwent ran- placebo) was 1.38±0.05 mg.
domization; these characteristics were balanced
Patients’ total scores in response to the Short- Form McGill Pain Questionnaire (Table 2) whenreceiving the gabapentin–morphine combination p r i m a r y o u t c o m e
were lower than when receiving placebo (P<0.05), Among the patients included in the efficacy analy- gabapentin as a single agent (P<0.05), or mor-sis, two assigned to the treatment sequence of gab- phine as a single agent (P<0.05). Patients’ scoresapentin, morphine, the gabapentin–morphine com- for pain-related interference with mood when re-bination, and placebo withdrew after completing ceiving the gabapentin–morphine combinationtwo treatment periods and one assigned to the treat- were lower than when receiving placebo (P<0.001)ment sequence of morphine, placebo, gabapentin, or morphine (P=0.03), and scores for pain-relatedand the combination withdrew after completing interference with general activity, normal work, Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n Figure 1. Enrollment, Randomization, Withdrawals,
and Completion of the Four Treatment Periods.
Two of the three patients included as having withdrawn after treatment period B withdrew before week 4. Combi-nation denotes the gabapentin–morphine combination.
Declined to participate, 13Excluded, 6Inadequate pain, 4Competing pain, 2 sleep, and enjoyment of life were significantly low- er when patients were receiving any of the active treatments than when receiving placebo (Table 2). According to responses to the SF-36 Health Survey, the gabapentin–morphine combination was associated with higher scores for vitality (P=0.007) and social functioning (P=0.004) than was place- bo and higher scores than morphine for vitality(P=0.03) and social functioning (P=0.04). All ac-tive treatments were associated with significantlyhigher scores for the domains of “role-physical” (problems with work or other daily activities as a result of physical health) and bodily pain and for mental health than was placebo (Table 2). All treat- ments were associated with significantly lowerscores on the Beck Depression Inventory (Table 2)than was placebo.
Table 3 lists adverse effects reported by patients during dose titration (weeks 1 through 3) and at the maximal tolerated dose. At the maximal toler- ated dose, the gabapentin–morphine combination was associated with a higher frequency of constipa- tion than gabapentin (P=0.006) but not morphine,and with a higher frequency of dry mouth than mor-phine (P=0.03) but not gabapentin (Table 3).
The numbers of patients who completed a given treatment and reported at least moderate pain re- lief at the maximal tolerated dose were as follows: placebo, 13 (31 percent, P<0.05 for the compari- son with all treatments); gabapentin as a single agent, 27 (61 percent); morphine as a single agent,35 (80 percent); and the gabapentin–morphinecombination, 32 (78 percent). The mean scores onthe Mini–Mental State Examination were 28.9±0.3 when receiving placebo, 28.8±0.3 when receiving gabapentin as a single agent, 29.0±0.2 when re- ceiving morphine as a single agent, and 29.0±0.3 According to responses to the blinding question- naire, the numbers of correct guesses by patients with regard to their treatment assignment were Diabetic neuropathy, 25Postherpetic neuralgia, 16 25 (66 percent) among those receiving placebo, 16(42 percent) among those receiving gabapentin, 16 Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine percent) for gabapentin, 14 (33 percent) for mor- Table 1. Demographic and Baseline Characteristics of the Patients.*
phine, and 21 (53 percent) for the combination.
Patients
Patients
with Diabetic
with Postherpetic
Neuropathy
Neuralgia
Characteristic
Our results suggest that treatment of neuropathic pain with the combination of gabapentin and mor- phine results in less pain than treatment with ei- ther gabapentin or morphine as a single agent, as indicated by patients’ pain intensity (the primary outcome) and responses on the Short-Form McGill Pain Questionnaire (a secondary outcome). The ab- sence of statistically significant differences in the percent change in pain level between baseline and treatment with the combination and between base- line and treatment with the single agents may have been due to the study’s insufficient statistical power with respect to this sensitivity analysis. Among patients receiving the maximal tolerated dose of a study drug, the frequency of adverse effects was sim- ilar among all treatments, except that patients re- ceiving the gabapentin–morphine combination had a higher frequency of constipation than those re- ceiving gabapentin alone and a higher frequency of dry mouth than those receiving morphine alone.
As compared with morphine, the combination was associated with less pain-related interference with mood and with higher scores for vitality and social The maximal tolerated doses of morphine and gabapentin were significantly lower with the com- bination than in treatment with each as a single agent, which suggests an additive interaction. A previous study in healthy subjects suggested that the addition of morphine to gabapentin resulted in higher serum concentrations of gabapentin than are seen with gabapentin alone.10 Although our trial did not involve pharmacokinetic characterizations, this previously observed pharmacokinetic interaction may, in part, explain the differences between treat-ment with the combination and treatment with gab- * Plus–minus values are means ±SD. TCA denotes tricyclic antidepressant, SSRI selective serotonin-reuptake inhibitor, and NSAID nonsteroidal antiin- Unlike laboratory studies of analgesic agents in † Race was determined on the basis of hospital registration data.
combination in which such techniques as isobolo- ‡ Pain was measured on a scale from 0 to 10, with 0 indicating no pain and 10 “the graphic analysis are used, our trial was not designedto distinguish between additivity and synergism.
Although a drug combination may demonstrate (44 percent) among those receiving morphine, and synergism, it is crucial also to characterize adverse8 (25 percent) among those receiving the combi- drug interactions (i.e., whether there is also addi-nation. The numbers of correct guesses by the re- tivity or synergism for adverse effects). Specifically,search nurse with regard to patients’ treatment as- the gabapentin–morphine combination should pro-signments were 29 (71 percent) for placebo, 18 (43 duce fewer adverse effects than either drug as a sin- Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n A Mean Weekly Pain
Score for Pain Intensity
Treatment Period
B Mean Daily Pain
Maximal Tolerated Dose
Dose (mg)
Score for Pain Intensity
Baseline
Gabapentin
Morphine
Morphine
Gabapentin
Combination
Figure 2. Mean Daily Pain and Maximal Tolerated Doses of the Study Drugs.
Panel A shows weekly mean scores for daily pain intensity throughout the trial for each of the four treatment sequences. The numbers of pa-tients include 41 who completed the trial and 3 who completed at least two treatment periods. Scores for pain intensity were rated on a scale from 0 to 10, with higher numbers indicating more severe pain. A, B, C, and D denote the four treatment periods, and numbers 1 through 4 the weeks within each period; in week 5, patients underwent a four-day dose tapering (T) and a three-day complete washout (W). Secondary outcomes were assessed with the use of the Short-Form McGill Pain Questionnaire, Beck Depression Inventory, Brief Pain Inventory, Medical Outcomes Study 36-Item Short-Form General Health Survey, global pain relief, and Mini–Mental State Examination, which were administered at baseline and again in week 4 of each treatment period. Panel B shows the mean (±SE) daily pain scores during week 4 at the maximal tol-erated dose of each regimen. Baseline pain intensity is included for comparison only and was not used in the efficacy analysis. P<0.001 for the comparison of gabapentin as a single agent with the gabapentin–morphine combination; P=0.01 for the comparison of morphine as a single agent with placebo; P=0.04 for the comparison of morphine as a single agent with the combination; and P<0.001 for the comparison of the combination with placebo. Panel C shows the mean (±SE) maximal tolerated doses of gabapentin and morphine administered as single agents as compared with them in combination. P<0.05 for the comparison of the gabapentin–morphine combination with gabapentin as a single agent, and P<0.05 for the comparison of the combination with morphine as a single agent.
Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine Table 2. Mean (±SE) Scores on the Short-Form McGill Pain Questionnaire, Brief Pain Inventory, Medical Outcomes Study
36-Item Short-Form General Health Survey (SF-36 Health Survey), and Beck Depression Inventory.*
Mean Score
* Scores for present pain intensity are on a scale from 0 to 3, with higher numbers indicating more severe pain. Total scores on the Short-Form McGill Pain Questionnaire are on a scale from 0 to 45, with higher numbers indicating more severe pain. Scores on the Brief Pain Inventory are on a scale from 0 to 10, with 0 indicating pain that does not interfere with the activity specified and 10 indicating pain that completely interferes. Scores on the SF-36 Health Survey are on a scale from 0 to 100, with higher numbers indicating better health status; in this survey, “role-physical” denotes problems with work or other daily activities as a result of physical health and “role-emotional” problems with work or other daily ac-tivities as a result of emotional problems. Scores on the Beck Depression Inventory are on a scale from 0 to 63, with high-er numbers indicating more severe depression.
† P<0.05 for the comparison with placebo.
‡ P<0.05 for the comparison with gabapentin and morphine in combination.
gle agent. Thus, a clinically useful combination may has a therapeutic profile superior to that of eitherhave an additive or even less than additive analge- drug as a single agent.
sic interaction, provided that the adverse effects In addition to evaluating combination therapy, show even less additivity in the combination.33 Data this trial replicates evidence from previous stud-from the present trial that indicate superior efficacy ies of the efficacy of opioids in neuropathic pain.
without greatly increased adverse effects suggest Analgesia with morphine was associated withthat a combination of gabapentin and morphine mood improvement and reduced pain-related in- Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n Table 3. Adverse Effects.*
Adverse Effect
Dose Titration, Wk 1–3
At Maximal Tolerated Dose, Wk 4
* Data are reported only for moderate-to-severe adverse effects with an incidence greater than 5 percent for any treatment.
† P<0.05 for the comparison with the gabapentin–morphine combination.
‡ P<0.05 for the comparison with placebo. terference with patients’ activity, work, sleep, and Unlike previous studies, the present trial used enjoyment of life, as well as improvement in men- an active placebo (low-dose lorazepam) that mim-tal health and in the domains of “role-physical” ics the adverse effects of the active treatments with-and bodily pain assessed with the use of the SF-36 out producing pharmacologic analgesia. Data fromHealth Survey. In light of previous trials that showed the blinding questionnaire indicate that approxi-efficacy of gabapentin as a single agent, it is sur- mately one third of the patients guessed that theyprising that gabapentin did not produce signifi- were receiving an active drug while they were re-cantly better results than placebo with regard to the ceiving placebo. Such guesses may have led toprimary outcome of this trial.
higher expectations during treatment with the ac- Despite this result, patients’ total scores for the tive placebo 34 and may have resulted in lower self- intensity of pain in response to the Short-Form assessments of pain intensity than might have beenMcGill Pain Questionnaire were significantly low- reported with the use of an inert placebo and con-er while they were receiving gabapentin than while sequently decreased the difference between treat-receiving placebo, as was also the case for pain- ment with gabapentin or placebo. Another possi-related interference with activity, mood, walking, ble explanation is that maximal tolerated doseswork, sleep, and enjoyment of life. Furthermore, were slightly lower than those reached in previousgabapentin was associated with improvements in trials of gabapentin.5,6 Nevertheless, the results ofmood and in almost all domains of the SF-36 Health this trial unequivocally show that gabapentin sig-Survey. These discrepancies leave us with an equiv- nificantly enhances the efficacy of morphine.
ocal result with regard to the analgesic efficacy of This trial suggests superior efficacy of a com- bination of gabapentin and morphine in the treat- Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved. g a b a p e n t i n a n d m o r p h i n e c o m b i n e d f o r n e u r o p a t h i c p a i n ment of neuropathic pain. Given the potential ben- The study medication gabapentin (Neurontin) was provided by efits (e.g., improved efficacy and fewer adverse Pfizer, and morphine (M-Eslon) was provided by Aventis-Pharma.
These companies had no other input into the study. Dr. Gilron re- effects) and drawbacks (e.g., adverse drug interac- ports having served on paid advisory boards for Pfizer during thetions) of any drug combination, trials are needed to past two years. Dr. Houlden reports having received grant supportcompare other analgesic combinations with their for research from Pfizer and Aventis-Pharma for other studies dur- We are indebted to Drs. Alan Jackson and Tanveer Towheed for Supported by a grant (MCT-38149) from the Canadian Institutes thoughtful comments on the manuscript.
of Health Research (CIHR). Dr. Gilron is the recipient of a CIHRNew Investigator Award.
r e f e r e n c e s
in healthy volunteers. Anesth Analg 2000; over trial of intravenous fentanyl in neuro- pact on quality of life of postherpetic neural- gia and painful diabetic neuropathy. Clin J 11. Matthews EA, Dickenson AH. A combi-
22. Gilron I, Booher SL, Rowan JS, Smoller
MS, Max MB. A randomized, controlled trial ates enhanced inhibitory effects on dorsal MC, et al. Advances in neuropathic pain: di- horn neuronal responses in a rat model of neuralgias. Neurology 2000;55:964-71.
23. Pocock SJ. Clinical trials: a practical
approach. Chichester, England: John Wiley, 12. Gilron I, Biederman J, Jhamandas K,
24. Melzack R. The short-form McGill Pain
lo M, et al. Opioids versus antidepressants antinociceptive morphine tolerance in the rat in postherpetic neuralgia: a randomized, pla- paw-pressure and tail-flick tests. Anesthesi- 25. Cleeland CS, Ryan KM. Pain assess-
cebo-controlled trial. Neurology 2002;59: ment: global use of the Brief Pain Inventory.
13. Myoshi HR, Leckband SG. Systemic
Ann Acad Med Singapore 1994;23:129-38.
opioid analgesics. In: Joeser JD, Butler SH, 26. Beck AT, Steer RA. Beck Depression In-
Gordon A, Eisenhoffer J. Controlled-release Chapman CR, Turk DC, eds. Bonica’s man- ventory manual. San Antonio, Tex.: Psycho- oxycodone relieves neuropathic pain: a ran- agement of pain. 3rd ed. Philadelphia: Lip- domized controlled trial in painful diabetic pincott Williams & Wilkins, 2001.
27. Ware JE, Snow KK, Kosinski MA, Gan-
14. Marson AG, Kadir ZA, Hutton JL, Chad-
dek MS. SF-36 Health Survey manual and in- wick DW. The new antiepileptic drugs: a sys- terpretation guide. Boston: Nimrod Press, tematic review of their efficacy and tolerabil- for the treatment of postherpetic neuralgia: 28. Folstein MF, Folstein SE, McHugh PR.
a randomized controlled trial. JAMA 1998; 15. Dickenson AH. Where and how opioids
“Mini-mental state”: a practical method for grading the cognitive state of patients for TS, eds. Progress in pain research and man- the clinician. J Psychiatr Res 1975;12:189-98.
al. Gabapentin for the symptomatic treat- agement. Vol. 2. Seattle: IASP Press, 1994.
29. Farrar JT, Young JP Jr, LaMoreaux L,
ment of painful neuropathy in patients with 16. Singh L, Field MJ, Ferris P, et al. The an-
Werth JL, Poole RM. Clinical importance of diabetes mellitus: a randomized controlled changes in chronic pain intensity measured possesses anxiolytic-like and antinocicep- on an 11-point numerical pain rating scale.
Luo ZD, Calcutt NA, Higuera ES, et al.
tive actions that are reversed by D-serine. Psy- Injury type-specific calcium channel alpha 2 chopharmacology (Berl) 1996;127:1-9.
30. Jones B, Kenward MG. Design and analy-
delta-1 subunit up-regulation in rat neuro- 17. Carlton SM, Zhou S. Attenuation of for-
sis of cross-over trials. 2nd ed. Boca Raton, pathic pain models correlates with antiallo- malin-induced nociceptive behaviors follow- dynic effects of gabapentin. J Pharmacol Exp ing local peripheral injection of gabapentin.
31. Hochberg Y, Tamhane AC. Multiple com-
parison procedures, New York: Wiley, 1987.
18. Polydefkis M, Griffin JW, McArthur J.
32. Agresti A. Categorical data analysis, New
Hughes J, Singh L. Gabapentin (neurontin) New insights into diabetic polyneuropathy.
and S-(+)-3-isobutylgaba represent a novel 33. Max MB. Combining opioids with oth-
class of selective antihyperalgesic agents. Br 19. Fields HL, Rowbotham M, Baron R. Pos-
er drugs: challenges in clinical trial design.
therpetic neuralgia: irritable nociceptors and deafferentation. Neurobiol Dis 1998;5:209- eds. Progress in pain research and manage- ment. Vol. 2. Seattle: IASP Press, 1994.
antinociceptive effects of spinal morphine 20. Max MB, Schafer SC, Culnane M, Smol-
34. Turner JA, Jensen MP, Warms CA, Carde-
in the rat tail-flick test. Pain 1997;72:375- ler B, Dubner R, Gracely RH. Amitriptyline, nas DD. Blinding effectiveness and associa- but not lorazepam, relieves postherpetic neu- tion of pretreatment expectations with pain 10. Eckhardt K, Ammon S, Hofmann U,
21. Dellemijn PL, Vanneste JA. Randomised
controlled trial. Pain 2002;99:91-9.
enhances the analgesic effect of morphine double-blind active-placebo-controlled cross- Copyright 2005 Massachusetts Medical Society. Downloaded from www.nejm.org at Hinari Phase 2 sites on June 10, 2006 . Copyright 2005 Massachusetts Medical Society. All rights reserved.

Source: http://archivo.bitacoramedica.com/wp-content/uploads/2007/03/Morphine.pdf