Doi:10.1016/j.pain.2007.08.028

Sativex successfully treats neuropathic pain characterised by allodynia: A randomised, double-blind, Turo J. Nurmikko a,*, Mick G. Serpell b, Barbara Hoggart c, Peter J. Toomey d, a Division of Neurological Science, University of Liverpool, Liverpool, United Kingdom b Gartnavel General Hospital, Glasgow, United Kingdom c Solihull Hospital, Birmingham, United Kingdom d York District Hospital, York, United Kingdom f Castle Hill Hospital, Hull, United Kingdom Received 11 March 2007; received in revised form 21 August 2007; accepted 21 August 2007 Cannabinoids are known to have analgesic properties. We evaluated the effect of oro-mucosal sativex, (THC: CBD), an endo- cannabinoid system modulator, on pain and allodynia, in 125 patients with neuropathic pain of peripheral origin in a five-week,randomised, double-blind, placebo-controlled, parallel design trial. Patients remained on their existing stable analgesia. A self-titrat-ing regimen was used to optimise drug administration. Sixty-three patients were randomised to receive sativex and 62 placebo. Themean reduction in pain intensity scores (primary outcome measure) was greater in patients receiving sativex than placebo (meanadjusted scores À1.48 points vs. À0.52 points on a 0–10 Numerical Rating Scale (p = 0.004; 95% CI: À1.59, À0.32). Improvementsin Neuropathic Pain Scale composite score (p = 0.007), sleep NRS (p = 0.001), dynamic allodynia (p = 0.042), punctate allodynia(p = 0.021), Pain Disability Index (p = 0.003) and Patient’s Global Impression of Change (p < 0.001) were similarly greater on sat-ivex vs. placebo. Sedative and gastrointestinal side effects were reported more commonly by patients on active medication. Of allparticipants, 18% on sativex and 3% on placebo withdrew during the study. An open-label extension study showed that the initialpain relief was maintained without dose escalation or toxicity for 52 weeks.
Ó 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Keywords: Sativex; Cannabinoid; Peripheral neuropathic pain; Allodynia drugs, opioids and topical local anaesthetics constitutingthe first-line therapy . Despite differences in their The treatment of chronic neuropathic pain is mainly mechanism of action, these agents appear similar in pharmacological, with antidepressants, antiepileptic analgesic efficacy and tolerability. There is a well-recog-nised need for better pain relief than is currently avail-able. This study reports the effect of the administration * Corresponding author. Address: Pain Research Institute, Division of a highly standardised THC:CBD endocannabinoid of Neurological Science, University of Liverpool Clinical Sciences system modulator, sativex (SativexÒ), on the severity Centre, Lower Lane, Liverpool L9 7AL, United Kingdom. Tel.: +44 of pain and allodynia, and associated sleep disturbance, (0) 151 529 5820; fax: +44 (0) 151 529 5821.
mental distress and disability in patients with peripheral 0304-3959/$32.00 Ó 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.pain.2007.08.028 Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx neuropathic pain. Identification of cannabinoid recep- sation schedule was held by the sponsor with a copy in tors and encouraging results from preclinical and patient-specific sealed envelopes sent to the pharmacy in each clinical studies and change in the political and centre. Once the patient’s eligibility was confirmed, they were scientific scene in some countries, notably Canada, have assigned to the next sequential randomisation number within led to revived interest in cannabinoids as a therapeutic each centre. The placebo medication was identical in compo-sition, appearance, odour and taste with the study medication modality. Two controlled trials on central pain associ- but without cannabis extract. That the smell and taste of the ated with MS found short-term efficacy from them cannabinoid preparation might lead to unblinding was , whereas two other studies in which pain was averted by disguising them with addition of peppermint oil not a primary outcome measure gave conflicting results to both preparations. All medication was provided in identi- . Neuropathic pain of peripheral or mixed periph- cal amber vials, packaged and labelled by the sponsor.
eral and central origin was reported to respond to aju-lemic acid, sativex or smoked cannabis; however, treatment arms in these studies were short, between 5and 14 days .
Patients had to have a current history of unilateral periph- Sativex is derived from extracts of selected strains of eral neuropathic pain and allodynia. Further enrolment crite- cannabis plants (Cannabis sativa) which produce high ria are shown in . Concomitant analgesia wasmaintained at a stable dosage regimen for the duration of and reproducible yields of the principal active cannabi- the study. The decision to recruit was based on the patient’s noids, delta-9-tetrahydrocannabinol (THC) and canna- history. No tests for drugs of abuse potential were carried bidiol (CBD). It is administered as a spray for sub- out. Ethical approval was granted by the Local Ethics Com- mittees of the participating centres. In one centre the 100 ll spray delivers 2.7 mg of THC and 2.5 mg of approval was conditional on patients not driving during the Cannabinoids are thought to work via two types of receptors, CB1 and CB2. CB1 is widely distributed in the peripheral and central nervous system, acting as apresynaptic modulator of neurotransmitter release.
Initial dosing was under clinical supervision at the study The main target for the effects of THC, CB1, occurs site. A pre-dose 100 mm ‘‘Intoxication’’ (0 = no intoxicationand 100 = extreme intoxication) Visual Analogue Scale at many sites critical for nociception. CB2 is also acti- (VAS) was obtained and vital signs were checked. A maxi- vated by THC but in normal circumstances is found in mum of 8 sprays were administered over 2 h with Intoxica- immune cells only. However, in clinical pain the role of tion VAS and vital signs checked at regular intervals. If, the CB2 receptor may be different because following following any dose, patients scored higher than 25 mm, or tissue injury it is shown to be expressed in central ner- there were clinical concerns, e.g. the patients showing dys- vous system microglia and dorsal root ganglion cells phoria or cardiovascular changes, subsequent doses were following tissue injury CBD appears to have lim- ited affinity for either cannabinoid receptor, but in After satisfactory completion of initial dosing, patients higher doses may potentiate the effects of THC began home dose titration and were allowed a maximum dose and mediate non-cannabinoid effects by activating the of 8 sprays per 3-hour interval and a maximum of 48 sprays TRPV1 receptor . Combining the two in the same per 24 h. At the next visit (after 7–10 days) titration, compli-ance and adverse events were reviewed, and patients advised preparation is thought to lead not only to increased on how to optimise dosing for the rest of the study period.
analgesic effect but may also result in antagonism of Those patients who satisfactorily completed the trial were offered the opportunity to participate in a common open-labelextension study of sativex.
All used and unused study medication containers were returned at each visit to the research centre. Patients were withdrawn from the study if there were indications of misuse,including failure to record dosage accurately. Periodic tele- This was a 5-week multi-centre (5 centres in UK, 1 in Bel- phone monitoring was undertaken at pre-arranged times dur- gium), randomised, double-blind, placebo-controlled parallel ing home dosing to check the patient’s condition and to group study. Patients were screened to determine eligibility answer any queries. Throughout the study, allowable concom- and completed baseline diary assessments of daily pain inten- itant medications or treatments were continued to provide ade- sity and sleep disturbance scores in the 7–10 days prior to quate background analgesia at a constant dose. Any first treatment assignment. After eligibility was confirmed, medication, other than the study medication taken during patients were assigned to the next sequential randomisation number within each centre. The randomisation schedule Patients kept a diary from the screening visit until end of had a 1:1 treatment allocation ratio with randomly permuted treatment in which they recorded daily their pain and sleep blocks stratified by centre and was generated using a com- scores (on the appropriate NRS), as well as adverse events puter based pseudo-random number algorithm. The randomi- Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx Unilateral peripheral neuropathic pain and allodynia Cannabinoid use (cannabis, MarinolÒ (synthetic THC) or nabilone(synthetic cannabinoid analogue)) at least 7 days before randomisation.
Subjects were required to abstain from use of cannabis during the study Schizophrenia, psychosis, or other major psychiatric condition beyonddepression with underlying condition A history of at least 6 months duration of pain due to a clinically Concomitant severe non-neuropathic pain or the presence of cancer related neuropathic pain or from diabetes mellitus Demonstrate mechanical allodynia and impaired sensation within Known history of alcohol or substance abuse the territory of affected nerve(s) on clinical examination Patients with complex regional pain syndrome (CRPS) were Severe cardiovascular condition, poorly controlled hypertension, epilepsy, eligible if they showed evidence of peripheral nerve lesion pregnancy, lactation, significant hepatic or renal impairment A baseline severity score of at least 4 on the numerical rating scale for spontaneous pain for at least 4 of 7 days in the baseline week A stable medication regimen of analgesics for at least 2 weeks prior Terminal illness or subjects inappropriate for placebo therapies Female patients of child bearing potential and male patients whose partner was of child bearing potential had to agree to useeffective contraception Willing for his or her name to be notified to the UK Home Office Participation within a trial in the last 12 weeks surement, it was agreed that they could exercise discretion inapplying the force needed to reproduce approximately the Tests for allodynia were carried out at baseline and end of same pain as at baseline. The patients’ verbal pain score and study. The investigator recorded the most painful area within pressure used were recorded. Each punctate pain provocation the affected territory. Mechanical dynamic allodynia was test was done only once during a single visit.
assessed by stroking the skin over the affected area five timeswith a standardised brush, designed specifically for sensory testing (Senselab Brush-05, Somedic, Horby, Sweden) atP5 s intervals, and recording the pain severity on a 0–10 point The primary outcome measure was a change from baseline scale. All strokes were of the same length, minimum 2 cm.
on a numerical rating scale (NRS) of mean intensity of global Each dynamic allodynia score was calculated as the average neuropathic pain, where 0 = ‘‘No Pain’’ and 10 = ‘‘Worst Pos- sible Pain’’. Secondary measures included the composite score Punctate allodynia was measured using an in-house built calculated from the Neuropathic Pain Scale (NPS) tests pressure algometer comprising a strain gauge connected to a for mechanical allodynia, a four-step verbal rating scale for metal filament with a diameter of 1 mm and blunt tip at base- sleep disturbance (see below), the Pain Disability Index line and end of study. The filament was manually directed (PDI) , the Patient Global Impression of Change (PGIC) against the skin at an angle of 90° and a steadily increasing of both pain and allodynia, and the General Health Question- pressure applied until the patient verbally indicated that they naire (GHQ-12) Possible cognitive decline was assessed perceived pain (punctate pressure pain threshold). A contralat- using the Brief Repeatable Battery of Neuropsychological tests eral mirror image site was used as control to identify any sys- (BRB-N) . Information regarding the frequency of adminis- temic effect from the trial drugs, as well as to introduce the tration of the medication was recorded by the subjects in their method to the patient before performing the test on the allo- diary. Adverse events were collected at each clinic visit, and dynic site. This control site was checked for evidence of local haematology, clinical chemistry and ECG monitored at the injury, scar, rash or neurological deficit. During each session the normal contralateral side was tested first. Once the patientindicated that the sensation of pressure had turned into pain, the algometer was removed and the pressure reading (in grams)recorded. The same method was used for allodynic sites.
The sample size calculation was based on an expected SD of In addition, patients were asked to verbally rate the inten- 1.8 for the pain intensity score, estimated from several studies sity of the pain elicited, choosing a number between 0 (no pain) on peripheral neuropathic pain. To detect a difference between and 10 (most intense pain imaginable). The investigators were treatment groups of 1.0 on a 0–10 (11-point) NRS with 80% aware of the previous punctate allodynia threshold and could power and a 5% level of significance, 52 evaluable subjects use it as guidance. Because some investigators expressed con- per group were required. A dropout rate of 15% was antici- cern at using a rigid threshold as a targe for the second mea- pated, bringing the total number of patients needed to 120.
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx The primary analysis for the primary and secondary end- 2 = once, 3 = twice, 4 = more than twice. The scores for this points was performed on the intention-to-treat (ITT) popula- ‘‘Sleep Disturbance NRS’’ were obtained at baseline and tion. The neuropathic pain intensity NRS score at baseline weekly thereafter until the end of trial. Statistical comparisons was defined as the mean of all diary entries from Day À7 to were performed in the same way as the primary outcome mea- Day À1 and, for the end of treatment score, the mean of all sure. The PGIC was compared between treatments using Fish- diary entries during the last 7 days in the study, or the last 3 days in the event of withdrawal. The NRS scores were summa-rised by treatment group for baseline, each week and end of treatment. The change in NRS pain scores was comparedbetween treatment groups using analysis of covariance, the A total of 141 patients were assessed for eligibility, 16 model including treatment and trial centre as factors and base- (11%) of whom failed to meet the eligibility criteria.
line pain severity as a covariate. From this analysis the Sixty-three subjects were randomised to sativex and 62 adjusted treatment means, treatment difference and 95% Con- to placebo At all participating centres, the ran- fidence Interval (95% CI) for the treatment difference werecalculated.
domisation led to a complete balance between treatment The total scores for all questionnaires (NPS, PDI, GHQ- allocations. Baseline demographic details for both groups 12), as well as 0–10 NRS ratings of punctate and mechanical are shown in The treatment groups were well allodynia, were obtained at baseline and end of the 5-week matched for age, duration of neuropathic pain, distribu- trial. Sleep disturbance was measured by asking the subjects tion of diagnostic pain subgroups, height, weight and to indicate the number of times they woke in previous nights for history of previous cannabis use. The diagnosis was due to symptoms on a four category scale where 1 = none, based on existing clinical, imaging and neurophysiological 4 did not meet entry criteria3 abnormal lab result3 withdrew consent2 administrative reason4 other Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx eral neuropathics were similar across the two groups ). The background use of concomitant analgesic medication was high in both groups. The most frequently reported medication was opioids, being taken by 74% of the placebo group and 63% of the sativex group. Other fre- quently used background medications were tricyclic anti- depressants, antiepileptic drugs, and NSAIDs Thirteen sativex patients (21%) failed to complete the study; 11 withdrew because of side effects, 1 due to patient non-compliance and one due to lack of efficacy.
Seven patients (11%) on placebo failed to complete the study, 2 because of adverse effects and 5 because of lack of effect. All randomised patients were included in the ITT analysis. For the per-protocol (PP) analysis, there were 47 patients on sativex and 56 on placebo. Protocol violations were due to failure to meet the stringent time window set for the final visits (12 patients on sativex, 2 on placebo), use of prohibited medication (6 on sativex, two of whom also failed to meet the final visit time win- dow, and 2 on placebo) or violation of inclusion/exclu- sion criteria (0 on sativex, 2 on placebo). One patient in each group had their data censored because of use of At baseline, the mean intensity of reported pain scores (SD) on NRS was in the severe range with no dif- ference between the sativex and placebo groups 7.3 (1.4) and 7.2 (1.5), respectively (At the end of treat- ment, the sativex group demonstrated an adjusted mean change in NRS score of À1.48 points (a 22% reduction) while the change for the placebo group was À0.52 points (an 8% reduction) The estimated treatment Morphine, methadone, oxycodone, pethidine.
b Tramadol, codeine, dihydrocodeine, dextropropoxyphene.
data. Aetiologies varied from post-infectious to trau- matic, vascular and idiopathic. In nearly one-half of Adjusted Mean Pain NRS Scores (95% CI)
patients the cause was posttraumatic and involved a single nerve or nerve branch (focal nerve lesion) while in one-fifth the lesion was at cervical, brachial or lumbosacral Fig. 2. Reduction of global neuropathic pain NRS scores in the two plexus level or involved several nerves (peripheral neurop- groups during the trial. First-week: home-titration; subsequent fourweeks: maintenance therapy. Weekly mean pain scores were obtained athy); in this group the original cause was either inflam- from pain diaries. End-point scores were obtained from diary entries mation or diffuse trauma and remained frequently during the last 7 days, or last 3 days in case of withdrawal, for the ITT unknown. The locations of focal nerve lesions and periph- analysis. Error bars represent 95% confidence intervals.
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx difference of À0.96 points was statistically significant in Sleep Disturbance
favour of sativex (p = 0.004; 95% CI: À1.59, À0.32).
The improvement in pain over placebo was evident from the second week after self-titration and was maintaineduntil the end of the study (). On sativex, 26% of patients had at least a 30% reduction in pain score and20% of patients had at least a 50% reduction in painscore, compared with 15% and 8% of patients on pla-cebo; the NNT (50%) and NNT (30%) calculated from these figures were 8.5 and 8.6, respectively. Analysis of the PP population also showed a significant treatment difference of À1.42 points in favour of sativex (p < 0.001; 95% CI: À2.10, À0.74).
Adjusted Mean Sleep Scores (95% CI) 1
Fig. 3. Reduction is sleep disturbance scores in the two groups during All questionnaire-based measures of pain and pain- the trial. For details, see text, and legend for .
related co-morbidity improved significantly more inpatients randomised to sativex than placebo (mean (SD) difference between the contralateral site NPS composite score in the sativex group decreased signif- and allodynic site: 127 (78) g). The severity of the allo- icantly more than in the placebo group. Sleep disturbance dynia within the affected area was comparable between also decreased early on and improvement was maintained both sativex and placebo groups for pressure needed until the end of the study ). Of the seven functional to elicit pain (68.8 (47.7) g vs. 83.0 (77.4) g) and for areas assessed in the PDI, only sexual activity failed to the level of pain generated by the stimulus itself (7.3 show a substantial improvement on sativex ( (1.7) vs. 7.4 (2.1)). At the end of study, there was no evi-dence of a change in the punctate pain threshold at the contralateral control site, irrespective of whether thepatients were on sativex or placebo (treatment difference 11.1 g in favour of placebo; p = 0.3). At the allodynic All patients recruited into the study showed dynamic site, the placebo group reported unchanged punctate allodynia. There was no difference in detected mean pain pressure thresholds at end of study (83.0 (77.4) g (SD) allodynia pain scores between the two groups at vs. 85.8 (68.9) g) with no change in pain levels (7.4 baseline (5.4 (2.7) vs. 5.0 (3.4)). At the end of treatment, (2.1) and 7.2 (2.2)). In the sativex group, the threshold the mean reduction of dynamic allodynia was 20% in the levels increased from 68.8 (47.7) g to 86.2 (73.2) g but sativex group, and 5% in the placebo group, with an esti- not significantly compared to the placebo group mated mean treatment difference of À0.82 (p = 0.042; (p = 0.14). Despite this increase of applied punctate 95% CI: À1.6, À0.03)) in favour of sativex. NNT for pressure there was a notable decrease in the allodynia 30% reduction in the allodynia score was 9.2 and for pain scores (baseline: 7.3 (1.7) vs. end of treatment: 6.2 (2.6)). The estimated treatment difference of À0.87 wasin favour of sativex (p = 0.021; 95% CI: À1.62, À0.13), giving an NNT (30%) of 5.9 and NNT (50%) of 13.4.
At baseline, all randomised patients except one on Inspection of punctate allodynia data revealed that in sativex showed punctate allodynia with clearly reduced some cases the pressure applied with the algometer to thresholds in the affected area vs. contralateral control the allodynic site had changed considerably between Table 3Summary of the results of the secondary efficacy end-points (ITT analysis) a All treatment comparisons in favour of sativex.
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx pre-treatment and post-treatment. When a sensitivity both treatment groups, with no tendency to increasing analysis was carried out in patients in whom the investi- dose over the duration of the study. The number of gators applied a similar degree of force (<5% greater) to sprays used daily in the placebo group was higher than the allodynic area on the second testing occasion (44 in the sativex group Over the study period, patients on sativex and 45 on placebo), there was a sig- patients randomised to sativex used a mean (SD) of nificantly larger reduction of allodynia pain in the sati- 10.9 (6.8) sprays daily compared with 19.0 (8.3) by vex group than the placebo group leading to a treatment difference of À0.94 (p = 0.046; 95% CI:À1.85, À0.02) in line with the ITT analysis.
When all subjects were analysed together, there was a strong correlation between the intensity of punctate Fifty-seven (91%) patients in the sativex group expe- allodynia, dynamic allodynia and spontaneous pain at rienced at least one adverse event (AE) during the course baseline and end of study, with similar strong correla- of the study compared with 48 (77%) patients in the pla- tions between the three parameters for change in scores cebo group. The most frequent AEs were central ner- (punctate allodynia vs. dynamic allodynia, r = 0.526, vous system related or gastrointestinal. Most were observed at onset of treatment, and in the majority described as mild. However, 6 (10%) patients on sativex p < 0.001). Inspection of sativex and placebo groups reported several gastrointestinal AEs (nausea, vomiting separately showed that similar significant correlations diarrhoea, constipation) with none on placebo reporting were present, except for change in dynamic allodynia the same. Severe symptoms suggesting involvement of in the placebo group (r = 0.065, p = 0.61).
the nervous system were reported with sativex in 7 Results of the secondary efficacy end-points are sum- (11%) and placebo 5 (8%) cases. All reported gastroin- marised in Thirty-two (51.6%) patients taking testinal AEs combined irrespective of their severity were sativex compared to 12 (19.3%) taking placebo consid- more common in the sativex group (31/63 (49%) than in ered their primary condition to be very much, much or the placebo group (20/62 (32%), p = 0.003, Fisher’s minimally improved (p < 0.001, Fisher’s exact test).
exact test), whereas the nervous system AEs (33/63 vs.
The odds ratio for achieving a better response on sativex 23/62, p > 0.10) were not. One case of severe psychiatric than placebo, calculated from a logistic regression of the AE was recorded on both groups (with sativex, emo- data, was 3.55 (95% CI: À7.61, À1.72) in favour of sat- tional stress associated with paranoid thinking and with ivex. There was no difference between groups in the placebo, confusion) and 6 further mild-to-moderate ones in the sativex group as opposed to 3 in the placebogroup; these were mainly mood related. AEs seen in 3 or more subjects are shown in for all AEs and forthose considered possibly related to treatment.
The mean (SD) number of sprays taken during the In the sativex group, 11 (18%) patients withdrew due first week of dose titration for sativex and placebo was to an AE compared with 2 (3%) in the placebo group.
7.3 (3.5) and 10.9 (3.9), respectively. From the second There was one transient ischaemic attack in the sativex week onwards, the dose frequency remained stable in group rated as a serious adverse event (SAE) and Table 4Summary of exposure to study medicine (number of sprays per day based on patient diary entries) Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx Table 5Treatment emergent adverse events (AEs) experienced by 3 or more subjects ($ 5%) receiving sativex compared with placebo and the % of subjectwho withdrew due to these AEs Number (%) of patients who withdrew due to AE considered unrelated to study treatment. Oral discom- to that used in the randomisation phase. Patients were fort, other than dryness of mouth, occurred in 8 (13%) reviewed initially at 4 weeks thereafter every 8 weeks.
patients taking sativex and 11 (18%) taking placebo The duration of participation in the extension trial and was usually reported as mild. One patient on sativex ranged from 1 to 871 days. By study closure, 56 (63%) had transient mucosal ulcerations but leukoplakia was patients had been withdrawn; 18 patients due to adverse not observed. No significant haematological or bio- effects, 16 due to lack of efficacy, 15 due to withdrawal of chemical abnormalities were encountered in laboratory consent, 7 for other reasons. The mean (SD) duration of the participation of withdrawn patients was 135 (147) The Brief Repeatable Battery of Neuropsychological days. An LOCF analysis involving 76 patients carried Tests (BRB-N) was given to 85 patients (43 randomised out at 52 weeks demonstrated a mean decrease of pain to sativex and 42 to placebo). No difference was seen NRS from the baseline of 7.3 (1.4) to 5.9 (2.4), i.e., similar between groups assessed for cognitive function with this to that seen in the randomised trial. The daily number of method at the beginning and end of treatment ( sprays did not increase appreciably during this period (N Intoxication scores (SD) remained low throughout (SD) 10.2 (6.0) at the end of the re-titration vs. 12.2 (7.6) the study, peaking after the self-titration week at 8.0 at 52 weeks). Two episodes of serious adverse effects were (15.4) for sativex and 3.0 (7.9) for placebo on a 0–100 reported (urticaria with eyelid oedema and an event of scale, respectively. Five patients on sativex and 2 somnolence, dysarthria and weakness) both leading to patients on placebo scored more than 40/100 during withdrawal of the patient in question from the study.
This study demonstrates that sativex is effective in the At the end of their 5-week trial period, each patient relief of peripheral neuropathic pain when given in addi- was offered the chance to enter an open-label extension tion to existing medication. Greater than 30% improve- study. Of the 125 subjects eligible, a total 89 (71%) of the ment in pain intensity, generally considered as clinically patients accepted the offer. They subsequently under- meaningful was reported by 26% of subjects receiv- went re-titration of sativex from zero, in a way identical ing sativex, compared with 15% of patients taking Table 6Psychomotor function during the trial shown as adjusted mean change from baseline in the BRB-N for each treatment group No difference between groups (positive difference denotes better function on sativex and negative on placebo).
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx placebo. At recruitment, all our patients were either shows adequate sensitivity to change in longitudinal non-responders to several conventional neuropathic studies in manifest depression The role of the analgesics, or were in severe pain despite taking appro- endocannabinoid system in the regulation of anxiety priate therapy. Considering the refractory nature of and mood disorders still remains unclear, and both their pain, and that patients remained on their existing CB1 agonists and antagonists have been shown to pos- analgesia, the improvement of the ongoing pain in those sess either anxiolytic or anxiogenic effects as well as on the active drug is encouraging. Further evidence for variable effects on mood . It is possible that the efficacy of sativex comes from improvement in GHQ-12 cannot detect modest changes in a population mechanical dynamic and punctate allodynia pain, sleep such as ours scoring just above the mean of the general and disability demonstrated in this study. Reduction in population Alternatively, the above paradoxical systematically measured mechanical allodynia is not effects of THC, or the ability of CBD to block some commonly reported in controlled trials on neuropathic of the psychomimetic effects of THC, may explain pain and usually only seen in single dose studies or following other than oral administration, and failure The self-titration schedule used in this study was cho- sen for several reasons. Previous studies indicated no reliable data converting reduction in allodynia scores that individual subjects have a variable threshold to the to clinically meaningful improvement, the NNT values known pharmacodynamic effects of sativex. A self-titra- presented should be interpreted with caution.
tion regimen permitted individual patients to optimise In comparison with pain relief reported from other their dose on the basis of their own efficacy and tolera- cannabis-related clinical trials, sativex in our group of bility response. Both experimental and human volunteer patients demonstrated a greater difference over placebo studies suggest that tolerance to some of the side effects (0.96, 95% CI À1.59, À0.32) than in patients with plexus of cannabis occurs within days of its repeated adminis- avulsion (treatment difference À0.58, 95% CI À0.98, tration A self-titration regimen allows for this À0.18) but somewhat less than in patients with central to occur, further optimising the therapeutic response.
pain due to MS (À1.25; 95% CI À2.11, 0.39) .
There appears to be substantial between-patient vari- The treatment difference reported for dronabinol in ability in the pharmacokinetics of THC and other can- MS patients deprived of concomitant analgesic medica- tion was 0.6 (95% CI À1.8, 0) while that for smoked can- implementation of a fixed-dose regimen is likely to yield nabis in painful HIV neuropathy was approximately the same as in the present study (as extrapolated from the The mean number of sprays taken daily by the sativex reported median 18% treatment difference in pain relief group remained stable during the course of the study from mean baseline scores of 53 and 54/100) Dif- despite patients having the freedom to determine their ferences in patient populations, numbers of withdraw- own dosing, indicating that tolerance did not develop at als, concomitant medications, trial designs and trial least over the 4-week stable treatment period of this study.
durations probably explain a great deal of these varying The dose titration regimen used was usually successful in results. Interestingly, the two other cannabinoid trials in providing the optimal therapeutic level for individual which evoked pain was assessed, albeit in a limited fash- patients. This conclusion is endorsed by the observation ion, also report some benefit in line with the present that those patients who took part in the open-label exten- sion study did not increase the number of daily sprays Our reason for maintaining existing analgesia was during the first 52 weeks of open-label treatment while based on both ethical and clinical considerations. A apparently maintaining the initial analgesic effect.
number of treatments that have shown efficacy in While the therapeutic effects of cannabis have often peripheral neuropathic pain are in widespread use in been attributed to THC, the second major constituent accordance with existing guidelines . Depriving a of the trial medication, CBD has been shown to have patient from such therapies during a placebo-controlled effects which may be additive to those of THC in pain trial could not be ethically justified. Clinical practice is relief in animal models, and also to have the potential also moving toward combination therapies due to the to ameliorate some of the psychoactive effects of THC realisation that in chronic neuropathic pain multiple . This interaction between the two components may permit subjects to tolerate mean daily doses of The lack of GHQ-12 to show any change during the more than 27 mg THC. This dose is in excess of those present study is in line with virtually all other cannab- used in other controlled studies of THC, and may inoid trials in which the psychosocial domain was explored, irrespective of the measure used (GHQ-30, The adverse events reported by the patients were mostly gastrointestinal, central nervous system related or topical. While reported gastrointestinal AEs were of anxiety, depression and social dysfunction and more common in the sativex group, central nervous sys- Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx tem AEs were not; and, importantly, objective measure- [2] Attal N, Cruccu G, Haanpa¨a¨ M, Hansson P, Jensen TS, ment of psychomotor performance did not vary across Nurmikko T, et al. EFNS guidelines on pharmacological treat-ment of neuropathic pain. Eur J Neurol 2006;11:1153–69.
the two groups. In general, the number of patients who [3] Attal N, Rouaud J, Brasseur L, Chauvin M, Bouhassira D.
withdrew is similar to those reported in well-known large Systemic lidocaine in pain due to peripheral nerve injury and trials of other drugs used in neuropathic pain That predictors of response. Neurology 2004;62:218–25.
PCIG scores favoured sativex over placebo suggests that [4] Backonja MA, Beydoun A, Edwards KR, Schwartz SL, Fonseca subjective pain relief, reduced disability and improved V, Hes M, et al. Gabapentin for the symptomatic treatment ofpainful neuropathy in patients with diabetes mellitus: a random- sleep overrode the negative impact of AEs.
ized controlled trial. JAMA 1998;280:1831–6.
There was no formal assessment of whether unblin- [5] Banks MH. Validation of the General Health Questionnaire in a ding might have taken place. The psychotrophic effects young community sample. Psychol Med 1983;13:349–53.
of cannabis are well known to the public, and 20% of [6] Berman JS, Symonds C, Birch R. Efficacy of cannabis based the participants in the present trial had previous expo- medicinal extracts for relief of central neuropathic pain from sure to cannabis. A post-hoc analysis found that previ- brachial plexus avulsion: results of a randomized controlled trial.
Pain 2004;112:299–306.
ous use of cannabis was not predictive of the change [7] Bever Jr CT, Grattan L, Panitch PH, Johnson KP. The Brief in mean pain scores. Classical psychotrophic effects of Repeatable Battery of Neuropsychological Tests for Multiple cannabis were reported by relatively few patients. The Sclerosis: a preliminary serial study. Mult Scler 1995;1:165–9.
intoxication scores were marginally higher in the sativex [8] Costa B, Giagnoni G, Franke C, Trovate AE, Colleani M.
group, and psychometric tests (BRB-N) remained Vanilloid TRPV1 receptor mediates the antihyperalgesic effect ofthe nonpsychoactive cannabinoid, cannabidiol, in a rat model of unchanged during the trial. It is therefore unlikely that acute inflammation. Br J Pharmacol 2004;143:247–50.
a significant number of those on sativex would have cor- [9] Farrar JT, Young JP, La Moreaux L, Werth JL, Poole RM.
rectly guessed they were on active medication unless Clinical importance of changes in chronic pain intensity measured they deliberately overdosed. From returned trial medica- on an 11-point numerical pain rating scale. Pain 2000;94:149–58.
tion it was concluded that such practice did not take place.
[10] Galer BS, Jensen MP. Development and preliminary validation of a pain measure specific to neuropathic pain: the Neuropathic Pain Patients taking placebo may have concluded that they were taking inactive substance, given that they used a rel- [11] Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, atively high number of sprays. However, the majority of Gureje O, et al. The validity of two versions of the GHQ in the patients took less than the highest allowable dosage. Also, WHO study of mental illness in general health care. Psychol Med only 5 (8%) of the placebo group withdrew for lack of effi- cacy, suggesting that no significant unblinding took place.
[12] Gilron I, Bailey JM, Tu D, Holder R, Weaver DF, Houlder RL.
Morphine, gabapentin, or their combination for neuropathic pain.
We conclude that the results from this study indicate that sativex has a positive broad spectrum therapeutic [13] Guy GW, Flint ME. A single centre, placebo-controlled, four effect in neuropathic pain, when used in addition to exist- period, crossover, tolerability study assessing pharmacokinetic ing analgesic medication. The emergence of a highly stan- effects, pharmacokinetic characteristics and cognitive profiles of dardised, uniform preparation of THC:CBD should allow as single dose of three formulations of cannabis based medicineextracts (CBMEs) plus a two period tolerability study compar- for further studies which better define the role for cannab- ing pharmacodynamic effects and pharmacokinetic characteris- inoids in the treatment of neuropathic pain syndromes.
tics of a single dose of a cannabis based medicine extract givenvia two administration routes. J Cannab Ther 2003;3:35–77.
[14] Guy GW, Robson P. A Phase I, a double blind, three-way crossover study to assess the pharmacokinetic profile of cannabisbased medicine extract (CBME) administered sublingually in GW Pharma acted as the sponsor of the study, pro- variant cannabinoid ratios in normal healthy male volunteers. J vided the medication, participated in the study design, co-ordinated the study between centres and carried out [15] Herzberg U, Eliav E, Bennett GJ, Kopin IJ. The analgesic effects of R(+)-WIN 55,212-2 mesylate, a high affinity cannabinoid the first set of analyses. The analyses were verified by an agonist, in a rat model of neuropathic pain. Neurosci Lett independent statistician. The principal investigator had full access to all the data and carried out further confirma- [16] Iskedijan M, Bereza B, Gordon A, Piwko C, Einarson TR. Meta- tory analyses. All authors contributed to the study design, analysis of cannabis based treatments for neuropathic and collection of the data and interpretation of the results. The multiple sclerosis pain. Curr Med Res Opin 2007;23:17–24.
[17] Jensen TS. Anticonvulsants in neuropathic pain: rationale and editorial content of this paper is of the authors, as is the clinical evidence. Eur J Pain 2002;6(Suppl A):618.
final decision to submit for publication.
[18] Jones RT. Cardiovascular system effects of marijuana. J Clin [19] Karst M, Salim K, Burstein S, Conrad I, Hog C, Schneider U.
Analgesic effect of the synthetic cannabinoid CT-3 on chronicneuropathic pain. JAMA 2003;290:1757–62.
[20] Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI.
[1] Abrams DI, Jay CA, Shade SB, Vizoso RN, Reda H, Press S, Structure of cannabinoid receptor and functional expression of et al. Cannabis in painful HIV-associated sensory neuropathy. A the cloned cDNA. Nature 1990;346:561–4.
randomised placebo-controlled trial. Neurology 2007;68:515–21.
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028 T.J. Nurmikko et al. / Pain xxx (2007) xxx–xxx [21] Meier T, Wassner G, Faust M, Kuntzner T, Ochsner F, Hueppe [31] Tait RC, Chibnall JT, Krause S. The Pain Disability Index: M, et al. Efficacy of lidocaine patch 5% in the treatment of focal psychometric properties. Pain 1990;40:171–82.
peripheral neuropathic pain syndromes: a randomized, double- [32] Varvel SA, Wiley JL, Yang R, Bridgen DT, Lomg K, Lichtman blind, placebo-controlled trial. Pain 2003;106:151–8.
AH, et al. Interactions between THC and cannabidiol in mouse [22] Nahas GG, Schwartz IW, Adamec JM, Manger WM. Tolerance of delta 9 THC in the spontaneously hypertensive rat. Proc Soc [33] Wade DT, Makela P, Robson P, House H, Bateman C. Do [23] Nikolasjen L, Gotrup H Kristensen AG, Jensen TS. Memantine cannabis-based medicinal extracts have general or specific effects (a N-methyl-D-aspartate receptor antagonist) in the treatment of in multiple sclerosis? A double-blind, randomised, placebo- neuropathic pain after amputation or surgery: a randomized, controlled study on 160 patients. Mult Scler 2004;10:434–41.
double-blind, cross-over study. Anesth Analg 2000;91:960–6.
[34] Wallace M, Braun J, Schulters G. Postdelivery of alfentanil and [24] Pevalin DJ. Multiple applications of the GHQ-12 in a general ketamine has no effect on intradermal capsaicin induced pain and population sample: an investigation of long-term retest effects. Soc hyperalgesia. Clin J Pain 2002;18:373–9.
Psychiatry Psychiatr Epidemiol 2000;35:508–12.
[35] Wallace MS, Magnusson S, Ridgeway B. Efficacy of oral [25] Rice AS, Maton S. Gabapentin in postherpetic neuralgia: a mexiletine for neuropathic pain: a double-blind, placebo-con- [26] Rog DJ, Nurmikko TJ, Friede T, Young CA. Randomized [36] Wallace MS, Rowbotham MC, Katz NP, Dworkin DH, Dotson Controlled Trial of Cannabis Based Medicine in Central Neuro- RM, Galer BS, et al. A randomized, double-blind, placebo- pathic Pain due to Multiple Sclerosis. Neurology 2005;65:812–9.
controlled trial of glycine antagonist in neuropathic pain. Neu- [27] Russo EB, McPartland JM. Cannabis is more than simply [37] Werneke U, Goldberg DP, Yalsin I, Ustun BT. The stability of the factor structure of the General Health Questionnaire. Psychol [28] Sagar DR, Kelly S, Millns PJ, O’Shaughnessey CT, Kendall DA, Chapman V. Inhibitory effects of CB1 and CB2 receptor agonists [38] Witkin JM, Tzavara ET, Nomikos GG. A role for cannabinoid on responses of DRG neurons and dorsal horn neurons in CB1 receptors in mood and anxiety disorders. Behav Pharmacol neuropathic rats. Eur J Neurosci 2005;22:371–9.
[29] Serpell MG. Neuropathic Pain Study Group. Gabapentin in [39] Woolf CJ. Dissecting out mechanisms responsible for peripheral neuropathic pain syndrome: a double-blind, placebo-controlled neuropathic pain: implications for diagnosis and therapy. Life Sci [30] Svendsen KB, Jensen TS, Bach FW. Does the cannabinoid [40] Zajisec J, Fox P, Sanders H, Wright D, Vickery J, Nunn A, et al.
dronabinol reduce central pain in multiple sclerosis? Randomised Cannabinoids for treatment of spasticity and other symptoms double blind placebo controlled crossover trial. Br Med J related to multiple sclerosis (CAMS study): multicentre random- ised placebo-controlled trial. Lancet 2003;363:1517–26.
Please cite this article in press as: Nurmikko TJ et al., Sativex successfully treats neuropathic pain characterised ., Pain (2007),doi:10.1016/j.pain.2007.08.028

Source: http://www.archives.hempembassy.net/hempe/resources/Nurmikko%20Sativex%20NP%20Allodynia%20Pain%202007.pdf