Topical lidocaine for the treatment of postherpetic neuralgia

Abstract

Background

This is an update of the original Cochrane review published in Issue 2, 2007. The cause of postherpetic neuralgia is damage to peripheral neurons, dorsal root ganglia, and the dorsal horn of the spinal cord, secondary to herpes zoster infection (shingles). In postherpetic neuralgia, peripheral neurons discharge spontaneously and have lowered activation thresholds, and exhibit an exaggerated response to stimuli. Topical lidocaine dampens peripheral nociceptor sensitisation and central nervous system hyperexcitability, and may benefit patients with postherpetic neuralgia.

Objectives

To examine efficacy and safety of topical lidocaine in the treatment of postherpetic neuralgia.

Search methods

We searched the Cochrane Pain, Palliative and Supportive Care Group Trials Register, CENTRAL, MEDLINE, EMBASE, LILACS, SIGLE, Citation Index, the reference lists of all eligible trials, key textbooks, and previous systematic reviews. Last search conducted April 2011.

Selection criteria

Randomised or quasi‐randomised trials comparing topical applications of lidocaine in patients of all ages with postherpetic neuralgia (pain persisting at the site of shingles at least one month after the onset of the acute rash).

Data collection and analysis

Two review authors extracted data, and a third checked them.

Main results

In the original review three studies involving 182 topical lidocaine treated participants and 132 control participants were included. Two studies gave data on pain relief, and the remaining study provided data on secondary outcome measures. The largest study published as an abstract compared topical lidocaine patch to a placebo patch and accounted for 150 of the 314 participants (48%).

A meta‐analysis combining two studies identified a significant difference between topical lidocaine and control groups for the primary outcome measure: a mean improvement in pain relief according to a pain relief scale. Topical lidocaine relieved pain better than placebo (P = 0.003).

There was a statistical difference between the groups for the secondary outcome measure of mean VAS score reduction (P = 0.03), but this was only for a single small study. There were a similar number of adverse skin reactions in both treatment and placebo groups.

The highest recorded blood lidocaine concentration varied between 59 ng/ml and 431 ng/ml between studies. The latter figure is high and the authors of the study suggest that the sample had been contaminated during the assay procedure.

Authors' conclusions

Since the last version of this review in Issue 2, 2007 no new studies have been found and the results therefore remain the same. There is still insufficient evidence to recommend topical lidocaine as a first‐line agent in the treatment of postherpetic neuralgia with allodynia. Further research should be undertaken on the efficacy of topical lidocaine for other chronic neuropathic pain disorders, and also to compare different classes of drugs (e.g. topical anaesthetic applications versus anti‐epileptic drugs).

Topical lidocaine (a local anaesthetic) for the treatment of postherpetic neuralgia (nerve pain)

Topical lidocaine may benefit some patients on an individual basis though there is stronger evidence for the use of other drugs. Postherpetic neuralgia is a long‐lasting pain disorder that causes pain from stimuli that are not normally painful. Local anaesthetics (such as lidocaine) can reduce the sensation of pain that is transmitted through nerves, and allow pain relief in patients with postherpetic neuralgia. This review found three small studies involving 182 topical lidocaine treated participants and 132 control participants using lidocaine for patients with postherpetic neuralgia. Two studies provided data on pain relief amongst patients with postherpetic neuralgia, and they showed some improvement in pain when topical lidocaine was compared to a placebo. No comparison was made with other medications that are in current use for the treatment of postherpetic neuralgia. The side effects of topical lidocaine are very minimal, but include skin problems (such as irritation and redness). We are unable to recommend the use of topical lidocaine as a first‐line treatment for postherpetic neuralgia at this stage. Further studies are needed to compare topical lidocaine to other medications in the treatment of postherpetic neuralgia.

Background

Description of the condition

This review is an update of a previously published review in the Cochrane Database of Systematic Reviews (Issue 2, 2007) on 'Topical lidocaine for the treatment of postherpetic neuralgia' (Khaliq 2007); an update search has been run in both July 2008 and April 2011. Postherpetic neuralgia is a chronic pain syndrome, being the most common complication of acute herpes zoster infection. After varicella zoster infection (chickenpox), the virus remains dormant in dorsal root ganglia and reactivation causes shingles. Herpes zoster (shingles) begins as a pain in a dermatomal distribution. Typically this is in one or two adjacent dermatomes (commonly in the thoracic, cervical or ophthalmic dermatomes) followed by malaise and fever a few days later. A rash consisting of papules and vesicles arises in the same dermatome. The pain of postherpetic neuralgia may last for a few months or up to a lifetime. One of the most widely used definitions of postherpetic neuralgia is "pain persisting, or recurring, at the site of shingles at least one month after the onset of the acute rash" (MacDonald 2000).

The number of cases of herpes zoster increases sharply with age and only about 5% of cases occur in children under 15 years (Donahue 1995). As a result, the main group of patients likely to develop postherpetic neuralgia after acute herpes zoster infection are people aged over 50 (Cluff 1998; Whitley 1998). Herpes zoster also occurs more frequently in immunosuppressed patients, and hence these people are also more likely to suffer from postherpetic neuralgia compared to their immunocompetent counterparts (Lojeski 2000).

The annual incidence of herpes zoster varies in different studies from 140 to 480 per 100,000 of the population (Chidiac 2001; Donahue 1995). In relation to this, the annual incidence of postherpetic neuralgia has been reported as 11 cases per 100,000 of the population in London, UK (MacDonald 2000). The lifetime prevalence of postherpetic neuralgia is 0.7 cases per 1000 population (MacDonald 2000).

Postherpetic neuralgia probably has multiple distinct underlying pain mechanisms. One of the most common symptoms is touch allodynia. The pain of postherpetic neuralgia is a neuropathic pain that is commonly precipitated by movement (mechanical allodynia) or thermal change (warm or cold allodynia). It results from injury of the peripheral nerves and altered central nervous system signal processing during the attack of shingles (Wall 1991). During shingles a patient's skin becomes inflamed, and the dorsal root ganglia also reveals signs of inflammation, haemorrhagic necrosis and neuronal loss (Watson 1991). Inflammation in peripheral nerves may persist for weeks to months, and usually leads to demyelination (Watson 1991).

Central nervous changes are also evident during shingles and these include acute degeneration of the dorsal horn of the spinal cord and involvement of spinal cord segments at levels adjacent to affected skin (Watson 1991). In patients who have had postherpetic neuralgia, autopsy reveals atrophy of the dorsal horn. This is not found in those patients who did not have postherpetic neuralgia (Watson 1991). It is in the dorsal horn that primary afferent nerve fibres release L‐glutamate and substance P to excite second‐order neurones of the spinothalamic system. This transmission is inhibited by local interneurones and by descending brainstem pathways, which contain serotonin and norepinephrine.

Following an injury, the peripheral neurons discharge spontaneously and have lowered activation thresholds. Axonal regrowth produces nerve sprouts that are also prone to unprovoked discharge. These regenerating sprouts are also hypersensitive to mechanical stimuli. This excessive activity in the peripheral nervous system is thought to lead to hyperexcitability of the dorsal horn, which in turn leads to exaggerated responses in the central nervous system.

Description of the intervention

A wide range of treatments for postherpetic neuralgia can be identified in the medical literature. Agents that are known to have been used in the treatment of postherpetic neuralgia include opioids such as: oxycodone (Pappagallo 1994; Watson 1998a), capsaicin cream (Frucht‐Pery 1997), antidepressants (such as amitriptyline and nortriptyline) (Graff‐Radford 2000; Watson 1998b), and antiepileptic drugs. A number of randomised controlled trials (RCTS) have shown the antiepileptic gabapentin to be very successful for treating postherpetic neuralgia (Filadora 1999; Rice 2001; Rowbotham 1998b). In addition, local anaesthetics such as lidocaine have been used to treat patients with postherpetic neuralgia (Comer 2000; Galer 1997; Rowbotham 1998a; Rowbotham 1989). Novel treatments have been used, such as electrical spinal cord stimulation (Harke 2002; Meglio 1989; Tseng 2000) but generally no RCTs have been carried out on such treatments, and evidence for their effectiveness is therefore limited.

Clinicians and patients are therefore faced with an array of therapeutic options for the treatment of postherpetic neuralgia, and combined with the fact that postherpetic neuralgia is a difficult condition to treat, patients may suffer debilitating symptoms and inadequate pain relief whilst undergoing different treatment options. It is therefore important that systematic reviews of the evidence on postherpetic neuralgia treatments are undertaken. At the time of writing (2007), the review authors are aware of the following systematic reviews that are relevant to the treatment of postherpetic neuralgia:

• Treatments for postherpetic neuralgia (Volmink 1996).

• Anticonvulsant drugs for acute and chronic pain (Wiffen 2002).

• Systematic lidocaine and oral analogues for neuropathic pain (Tremont‐Lukats 2005).

How the intervention might work

Lidocaine is an amide‐type local anaesthetic agent that acts by stabilising neuronal membranes. It impairs membrane permeability to sodium, which in turn blocks impulse propagation, and thus dampens both peripheral nociceptor sensitisation, and eventually central nervous system (CNS) hyperexcitability. It also suppresses neuronal discharge in A delta and C fibres. Regenerating nerve fibres have an accumulation of sodium channels. When lidocaine binds to such sodium channels it initiates an 'inactive state' from which normal activation is unable to occur. Lidocaine reduces the frequency rather than the duration of sodium channel opening. In a small dose it inhibits ectopic discharges, although it does not disrupt normal neuronal function. Lidocaine also suppresses spontaneous impulse generation from dorsal root ganglia, where the postherpetic neuralgia virus remains dormant after initial infection by varicella zoster (chickenpox). In a study by Wasner 2005 postherpetic neuralgia patients were divided into those with impaired nociceptor function and those in whom it was intact. Patients with impairment of nociceptor function had significantly greater pain reduction when lidocaine was used compared to placebo. Patients with preserved nociceptor function demonstrated no significant pain relief.

Possible adverse reactions to topical lidocaine are usually limited to skin reactions, including a transient paleness, redness and oedema. Systemic adverse reactions are rare and may result from high plasma levels due to excessive dosage or rapid absorption. Such reactions involve excitation or depression of the CNS, as well as depression of the cardiovascular system. CNS reactions may be characterised by drowsiness, dizziness, convulsions and respiratory arrest. Cardiovascular reactions may be characterised by hypotension, bradycardia and cardiac arrest.

Why it is important to do this review

Currently, no systematic review of RCTs of local anaesthetic treatments for postherpetic neuralgia has been prepared, and it is thus the object of this review to assess the effectiveness of topical lidocaine.

Objectives

To review systematically all randomised and quasi‐randomised trials of the use of topical lidocaine, and to examine its efficacy and safety in the treatment of postherpetic neuralgia.

Methods

Criteria for considering studies for this review

Types of studies

All randomised and quasi‐randomised trials that compare the use of topical lidocaine in the treatment of post herpetic neuralgia, with placebo or with any other active treatment. Quasi‐randomised trials are those in which treatment allocation is intended to be random but might be biased (e.g. due to alternate allocation, allocation by date of birth, etc).

Types of participants

Participants of any age who fulfil the criteria which approximate to the definition of postherpetic neuralgia posed by MacDonald 2000, that is: pain persisting, or recurring, at the site of shingles at least one month after the onset of the acute rash.

Types of interventions

The interventions included are all topical applications of lidocaine, such as patch and gel preparations. Studies that include the use of systemic lidocaine will not be included, as there is already a review on the use of systemic anaesthetic agents (Tremont‐Lukats 2005). Studies will be included where the topical lidocaine is administered in any setting (for example, in hospitals, or in primary healthcare centres), and by any person (for example, doctors, nurses or the patient themselves).

Types of outcome measures

Primary outcomes

The primary outcome measure is the mean improvement in patients' reports of pain relief measured by a categorical scale similar to that employed by Rowbotham 1996a, or a similar scale reporting a change in pain. This scale has the following steps:

0 = the pain is 'worse' 1 = 'no' pain relief 2 = 'slight' pain relief 3 = 'moderate' pain relief 4 = 'a lot' of pain relief 5 = 'complete' relief of pain

As the scale is designed to measure change, the baseline pre‐application rating is assumed to be 'no' relief of pain (score 1). A measure of the pain relief has been selected as the primary outcome measure on the ground that a patient‐centred measure would be more relevant in what is a subjective symptom.

Secondary outcomes

The secondary outcome measures are:

1. mean reduction in visual analogue scales (VAS) scores at any time after randomisation;

2. highest recorded blood lidocaine level at any time between four hours and 30 days;

3. the proportion of participants with one or more adverse skin reactions, such as redness, blanching, or allergic reaction, due to patch or gel application.

Search methods for identification of studies

Electronic searches

To identify studies for inclusion in this review, we developed detailed search strategies which could be used in a variety of electronic databases. The search strategies were based on the search strategy developed for MEDLINE but revised appropriately for each database.

For the MEDLINE search, the subject search was run with the Cochrane Highly Sensitive Search Strategy (CHSSS) for identifying randomised trials in MEDLINE: sensitivity maximising version (2008 revision) as referenced in Chapter 6.4.11.1 of The Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.0 (Higgins 2008), in order to retrieve trial reports that utilised RCT or controlled clinical trial (CCT) methodologies. The subject search used a combination of controlled vocabulary and free text terms. These can be found in Appendix 1. The search was run in November 2006 for the original review and subsequent searches were run in July 2008 and April 2011 for this update.

We searched the following databases:

• Cochrane Pain, Palliative & Supportive Care Register (up to November 2006);

• The Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 4, 2011);

• MEDLINE (January 1966 to 24th April 2011);

• EMBASE (January 1980 to 24th April 2011);

• LILACS (January 1982 to 24th April 2011);

• SIGLE for conference proceedings (January 1980 to 24th April 2011);

• Citation Index (January 1980 to 24th April 2011).

Searching other resources

In addition, we searched the reference lists of all eligible studies, key textbooks, and previous systematic reviews for additional studies. We also wrote to authors of identified studies to try to obtain data from unpublished studies.

The search attempted to identify all relevant studies irrespective of language. Non‐English papers were assessed and, if necessary, translated with the assistance of a native speaker.

Data collection and analysis

Study selection

All three review authors checked the titles and abstracts of the articles identified by the search. The full text of all potentially relevant studies was obtained and the three review authors independently decided upon inclusion. Disagreements were resolved by discussion.

Quality assessment

Using the information in the full text of each included study, the three review authors independently assessed the methodological quality of each study. The assessment of methodological quality took into account:

• allocation concealment,

• patient blinding,

• observer blinding,

• explicit inclusion criteria,

• explicit outcome criteria,

• how studies dealt with the baseline differences of the experimental groups, and

• completeness of follow up.

Disagreements were resolved by discussion. Allocation concealment was also graded according to the criteria presented in Cochrane's review writing software, RevMan 4.2 where A is adequate, B is unclear, C is inadequate and D meant 'not done'.

Data extraction

Two review authors extracted data from each included study using a specially designed data extraction form. If there were any disagreements between the two review authors, then the third review author was consulted.

The following data items were extracted:

• participants' characteristics: postherpetic neuralgia diagnosis, age range and gender;

• details of the intervention: dosing regimen, method of administration, duration of treatment;

• details of the control: placebo, other active treatment, or other intervention;

• outcomes: pain relief, adverse events and effects;

• study design: methods of randomisation, parallel/cross‐over design, whether studies were single centre or multi‐centre;

• study withdrawals, protocol violators, drop‐outs.

Data analysis

To make it possible to perform the analysis in RevMan, we treated cross‐over trials as parallel group trials. This is an approximation but, if anything, reduces the apparent significance of the results.

We expressed results as relative risks (RR) with 95% confidence intervals (CIs) and risk differences (RDs) with 95% CI for dichotomous outcome measures, and weighted mean difference (WMDs) with 95% CI for continuous outcomes. We then calculated weighted mean differences between the treatment effects (using a 'fixed‐effect' approach) across trials using the statistics in the software package, RevMan Analyses, in RevMan version 4.2.

We tested the results for statistical heterogeneity across trials (using the chi‐squared test). If statistical heterogeneity was found we undertook sensitivity analyses by repeating the calculations after omitting the trials which had low scores on individual quality items. If there was still any unexplained heterogeneity the study results were combined using 'random‐effects' methods.

We undertook statistical analysis to obtain numbers‐needed‐to‐treat‐to‐benefit (NNT) data, wherever this was appropriate.

We defined subgroups of interest in advance:

1. time from onset of pain to start of treatment (one month or less after onset, more than one month after onset);

2. duration of treatment (up to and including 30 days, more than 30 days).

With regards to adverse events, we collated data and grouped it as either adverse skin reactions, adverse cardiovascular reactions, or adverse central nervous system (CNS) reactions.

Results

Description of studies

No new studies were found for this update (April 2011) nor the previous update (July 2008) and the results found in the original review still stand. For study characteristics: see the 'Characteristics of included studies' and 'Characteristics of excluded studies' tables.

A search of the Cochrane Pain, Palliative and Supportive Care Group register, MEDLINE, EMBASE, LILACS and CINAHL revealed over 800 titles. Of these, nine studies were relevant to this review. We excluded six studies; two studies did not have or describe random allocation (Iseki 2000; Tajti 1999); one study was single‐blinded and did not have random allocation (Tamakawa 1998); one study contained results that were in a duplicate publication (Galer 2002); one study studied postherpetic neuralgia with other types of neuralgia, and separate data were unavailable (Meier 2003). One further study (Galer 1999) was an enriched enrolment study, and was therefore not included in the meta‐analysis.

We included two randomised, double‐blind controlled clinical trials: Rowbotham 1995 and Rowbotham 1996a. One further randomised study appeared as an abstract (Rowbotham 1996b) but missing data for this study was obtained from the US Food and Drugs Administration authority (FDA 1999).

The treatment sequence design was a crossover in two of the three trials (Rowbotham 1995; Rowbotham 1996a).

The three eligible trials included 182 participants treated with topical lidocaine and 132 control participants. The largest trial (which was published as an abstract) was of a topical lidocaine patch versus a placebo patch (Rowbotham 1996b) and accounted for 150 of the 314 participants (48%).

Two of the trials were single centre studies (Rowbotham 1995; Rowbotham 1996a) and the remaining one a multicentre study (Rowbotham 1996b).

Results of the search

All included studies compared topical lidocaine to placebo. One of the three trials used lidocaine gel versus a vehicle gel (Rowbotham 1995), while the remainder used lidocaine patches. All the lidocaine concentrations that were used, whether for the gel preparation or the patches, were 5%. The treatment time in one study was 28 days (Rowbotham 1996b). It was between eight and 24 hours in the remaining two studies.

All three trials analysed blood samples from patients to assess blood lidocaine levels (Rowbotham 1995; Rowbotham 1996a; Rowbotham 1996b).

It is important to note that all the included studies had allodynia as an inclusion criterion. It is not known from any of the studies whether topical lidocaine relieves pain in patients who do not have allodynia.

Risk of bias in included studies

See Methodological quality scores' table in 'Additional Table 2'.

Table 1.

Methodological Quality Scores

Study	Patient blinding	Observer blinding	Allocation conceal.	Inclusion criteria	Explicit outcomes	Baseline differences	Follow up
Rowbotham 1995	A	A	A	A	B	A	A
Rowbotham 1996a	A	A	D	A	B	A	A
Rowbotham 1996b	A	A	A	A	A	A	A

The methodological quality assessment took into account allocation concealment, patient blinding, observer blinding, explicit outcome criteria, explicit inclusion criteria, how studies dealt with baseline differences and completeness of follow‐up. We graded these items using the following grades: A: adequate, B: unclear, C: inadequate. For the explicit outcome criteria, we used the following grades: A: explicit outcome criteria, with a clear primary outcome, B: explicit outcome criteria but no clear primary outcome, C: not stated.

In all three studies participants were randomly assigned to receive either topical lidocaine or an identical appearing placebo. In the crossover design studies (Rowbotham 1995; Rowbotham 1996a) the participants were randomised as to which treatment they would receive first (i.e. placebo or topical lidocaine).

In one of the three studies, allocation concealment was not mentioned (Rowbotham 1996a). In the remaining studies (Rowbotham 1995 and Rowbotham 1996b) allocation concealment was considered to be adequate.

Patient and observer blinding was intended in all studies, although none of the studies stated the effectiveness of blinding. All the studies used explicit inclusion criteria, and had used the internationally accepted definition of postherpetic neuralgia being a pain that lasts greater than one month after the disappearance of the rash of shingles.

Two of the studies had stated outcome criteria, but had not specified which of these outcomes was their primary outcome (Rowbotham 1995; Rowbotham 1996a).

All studies considered baseline clinical features and these were found to be very similar in all. Follow‐up was complete in all studies for the intended follow up period with the greatest period being one month.

We were able to obtain data for our primary outcome measure (pain relief according to a pain relief score, such as the six‐point pain relief scale) for two of the studies (Rowbotham 1996a; Rowbotham 1996b). All studies used one of our secondary outcome measures for assessing pain relief: mean reduction in VAS (Rowbotham 1995; Rowbotham 1996a; Rowbotham 1996b).

Effects of interventions

Primary outcome measure: mean improvement in pain relief

In the largest study (Rowbotham 1996b ‐ an abstract) the mean improvement was marginally greater in the lidocaine patch group compared with the vehicle patch, weighted mean difference 0.50 (95% CI 0.12 to 0.88). This outcome was reported in one of the smaller studies, albeit a slightly greater difference between active and control groups (Rowbotham 1996a). The combined weighted mean difference for these two studies was 0.42 (95% CI 0.14 to 0.69), with a P‐value of 0.003. In the remaining study (Rowbotham 1995) this outcome measure was not reported, therefore these could not be included in the meta‐analysis for this outcome. Interestingly, the missing data obtained for Rowbotham 1996b abstract (from FDA 1999), showed that there was no statistical difference between participants receiving topical lidocaine or placebo in pain relief, except in participants who were treated for three to four weeks. Even in this group of participants, there was only a slight improvement in pain relief using a pain relief score, but no difference when using a VAS score.

Secondary outcome measures

1. Mean VAS score reduction

Mean VAS scores (but not VAS score reduction) were published for the largest study (Rowbotham 1996b) in which the mean VAS score was 37.7 (SD 19.9) in the lidocaine patch group compared with 40.8 (SD 19.9) in the control group. The difference between the groups was not significant (P = 0.437). A VAS score reduction was only reported in one trial (Rowbotham 1996a), with a significant result (P = 0.03). Similar outcome measures were not reported in the remaining study.

2. Highest recorded blood lidocaine concentration between four hours and 14 days

All studies reported this outcome measure, with the highest recorded blood lidocaine concentrations ranging from low values of 59 ng/ml (Rowbotham 1995) and 104 ng/ml (Rowbotham 1996a), to the highest of 431 ng/ml (Rowbotham 1996b).

3. Number of participants with one or more adverse skin reactions, such as redness, blanching or irritation, due to lidocaine application

All studies except for the largest reported this outcome measure. Rowbotham 1995 reported 12 adverse reactions in both groups. Rowbotham 1996a reported three adverse reactions: two in the active group and one in the placebo group. None of the participants dropped out from any study due to adverse skin reactions. It should be remembered that the adverse skin reactions may simply be due to the use of a patch, as opposed the lidocaine itself.

There were no reported cardiovascular, respiratory or neurological adverse reactions.

These outcome measures showed no heterogeneity, therefore sensitivity analyses according to any component of methodological quality would not have altered the conclusion of the review and were not performed.

Discussion

Our search strategy yielded three studies suitable for this systematic review, with a total of 182 participants treated with topical lidocaine and 132 control participants. This is a relatively small number in relation to other trials that have looked at treatments for postherpetic neuralgia, such as gabapentin. We feel this constitutes a firm foundation for a call for further RCTs ‐ ideally comparing topical lidocaine with other types of treatment that are currently used for postherpetic neuralgia.

There are a number of limitations of the included studies. Firstly, only a small number of studies on topical lidocaine have been performed. Secondly, different outcome measures have been used between the various studies, and so there is very little data to combine efficacy. One of the studies included was only published as an abstract (Rowbotham 1996b), and the actual paper has not been through a peer review and remains unpublished.

Overall this review was only able to identify two published and one unpublished trial, which between them only show modest efficacy on pain relief. In fact, the large unpublished study (Rowbotham 1996b) had no statistical difference between VAS scores, furthermore, all the studies have been written by the same first author. One of the strengths of the Rowbotham 1995 study is that it has been controlled for systemic effect.

Regarding the results of the meta‐analysis, our primary outcome measure was an improvement in patients' reports of pain relief according to a pain relief scale. For this outcome we were able to undertake a meta‐analysis using only two of the studies (Rowbotham 1996a; Rowbotham 1996b), because the remaining study did not provide relevant data. The meta‐analysis revealed that topical lidocaine was better than placebo for pain relief (P = 0.003).

Our first secondary outcome measure was mean VAS score reduction. All three studies used VAS scores as one of their outcome measures (Rowbotham 1995; Rowbotham 1996a; Rowbotham 1996b). Only one of the studies actually provided data on mean values and standard deviations for VAS score reductions (Rowbotham 1996a). In a second study, data on mean values and standard deviations was missing (Rowbotham 1995), although the authors comment that the mean VAS score reduction was greater in topical lidocaine than in placebo (P = 0.024). For us to comment any further on this result, we would need to obtain the actual mean and standard deviation values, add the data into a meta‐analysis, and check heterogeneity. In the third study (Rowbotham 1996b) actual VAS scores after treatment were provided, as opposed to a VAS score reduction. This could not be included in a meta‐analysis for our secondary outcome measure. For the only study that remained (Rowbotham 1996a), there was a statistically significant difference in VAS score reduction between topical lidocaine and placebo treatments (P = 0.03). However, it has to be borne in mind that this was a very small study.

Our next secondary outcome measure was blood lidocaine concentrations (in ng/ml). The highest recorded value in any study was 431 ng/ml (Rowbotham 1996b). The author of this study states that this result was not in agreement with results from any other patient, and perhaps the sample had been contaminated during the assay procedure. The next highest blood lidocaine concentration in Rowbotham 1996b was 205 ng/ml. Excluding the potentially contaminated sample in Rowbotham 1996b, the highest value in any study would have been 205 ng/ml.

It is interesting to note that estimates of 'toxic' systemic lidocaine concentrations range between 400 ng/ml and 1000 ng/ml. According to these estimates, a concentration of 431 ng/ml would have been high enough to pose a risk of systemic toxicity (which can lead to hypotension and bradycardia, which may in turn lead to cardiac arrest).

Our final secondary outcome measure was the number of participants reporting adverse skin reactions. All studies (except Rowbotham 1996b) reported these data. Similar numbers of adverse skin reactions are reported in all studies, between topical lidocaine and placebo groups. This would suggest that the vehicle (for example, patch), rather than the active components (lidocaine), were the cause of these reactions.

There is clinical evidence to show that the following drug groups are also effective in relief of postherpetic neuralgia: antidepressants, antiepileptics, opioids and capsaicin. However, almost all studies we discovered compared the active drug against a placebo. We did not find any studies that compared one class of drug to another class (for example, antiepileptics versus opioids, etc.), and it is therefore not possible to determine which drug group is the most effective in the treatment of postherpetic neuralgia.

The lidocaine patch (Lidoderm), has been approved recently by the US Food and Drug Administration for the treatment of postherpetic neuralgia (FDA 1999). The 5% lidocaine patch is the only topical anaesthetic agent that has received FDA approval for the treatment of a neuropathic pain condition, in particular postherpetic neuralgia (FDA 1999). It is a topical agent that is easy to apply, with limited side effects. There are studies to suggest that the lidocaine patch is effective in other chronic neuropathic pain conditions other than postherpetic neuralgia.

In all the studies we reviewed, patients' reports of pain were assessed using subjective tools (such as the VAS scale and six point pain relief scale). In chronic pain, unlike postoperative pain, using such subjective tools in isolation may not represent the best method for pain relief assessment. It needs to be remembered that chronic pain can lead to varying degrees of disability in patients, affecting their daily activities, and therefore this should also be assessed when studying treatments for chronic pain (Elliot 1999).

Authors' conclusions

Since the last version of this review (Issue 2, 2007) no new studies were found. There is still insufficient evidence to recommend topical lidocaine as first line therapy in post‐herpetic neuralgia with allodynia. All studies were by the same first author and this further limits any evidence of efficacy. Topical lidocaine may benefit some patients on an individual basis, but at present other drugs (such as gabapentin) have far stronger evidence for their use.

RCTs need to compare topical lidocaine with other drug types (e.g. antiepileptics and antidepressants). Further research should be undertaken to check for the efficacy of topical lidocaine in the treatment of other chronic neuropathic pain disorders.

Appendices

Appendix 1. MEDLINE search strategy

1. randomized controlled trial.pt. 2. controlled clinical trial.pt. 3. randomized.ab. 4. placebo.ab. 5. drug therapy.fs. 6. randomly.ab. 7. trial.ab. 8. groups.ab. 9. or/1‐8 10. humans.sh. 11. 9 and 10 12. lidocaine 13. lignocaine 14. Anesthesia, Local/ or Anesthetics, Local/ 15. or/12‐14 16. Postherpetic 17. Post‐herpetic 18. Neuralgia 19. or/16‐18 20. 11 and 15 and 19

Data and analyses

Comparison 1.

Topical lidocaine versus placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Improvement in pain relief according to a pain relief scale	2	220	Mean Difference (IV, Fixed, 95% CI)	0.42 [0.14, 0.69]
2 Improvement in visual analog scale rating	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3 Highest recorded blood lidocaine concentration			Other data	No numeric data

Analysis 1.1.

Comparison 1 Topical lidocaine versus placebo, Outcome 1 Improvement in pain relief according to a pain relief scale.

Analysis 1.2.

Comparison 1 Topical lidocaine versus placebo, Outcome 2 Improvement in visual analog scale rating.

Analysis 1.3.

Comparison 1 Topical lidocaine versus placebo, Outcome 3 Highest recorded blood lidocaine concentration.

Highest recorded blood lidocaine concentration

Study
Rowbotham 1995	59
Rowbotham 1996a	104
Rowbotham 1996b	431

What's new

Date	Event	Description
16 November 2016	Amended	Update added; see Published notes.

History

Protocol first published: Issue 3, 2004 Review first published: Issue 2, 2007

Date	Event	Description
28 October 2013	Amended	This review has been withdrawn following the development of a new protocol. See Published notes.
26 July 2013	Amended	This review is being updated following a new protocol in preparation and will be withdrawn when the new review is published.
2 June 2011	New search has been performed	Review updated. Search performed. No new studies found. No changes to conclusions.
2 July 2008	New search has been performed	Review updated. Search performed. No new studies found. No changes to conclusions.
2 July 2008	Amended	Converted to new review format.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Rowbotham 1995

Methods	Double blind crossover design RCT
Participants	47 patients in control group and 47 in the placebo group, of any age, with postherpetic neuralgia greater than one months duration
Interventions	5% lidocaine gel (or vehicle gel) applied on areas of painful or non‐painful skin, for eight to 24 hours
Outcomes	VAS scores, blood lidocaine levels and allodynia severity
Notes	Single centre conducted in USA
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Low risk	A ‐ Adequate

Rowbotham 1996a

Methods	Double blind crossover design four‐session RCT
Participants	35 patients in the control group and 35 in the placebo group, of any age, with postherpetic neuralgia greater than one months duration
Interventions	1) Observation only, 2) 5% lidocaine patch, 3) Vehicle patch, and 4) 5% lidocaine patch (for a second time). Each intervention lasting between eight and 24 hours
Outcomes	VAS scores (primary outcome), six‐point pain relief scale improvement, and blood lidocaine levels
Notes	Single centre conducted in USA
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	B ‐ Unclear

Rowbotham 1996b

Methods	Double blind parallel group RCT
Participants	150 patients older than 21 years, with postherpetic neuralgia greater than one months duration
Interventions	5% lidocaine patches (or vehicle patches) covering a maximum area of 420 cm², applied up to a maximum of three times daily, for up to 28 days
Outcomes	VAS scores (primary outcome), six‐point pain relief scale improvement, and sensory skin testing for allodynia. Also blood lidocaine levels and side effects on a 28 point checklist
Notes	Multicentre conducted in USA
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Low risk	A ‐ Adequate

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Galer 1999	Study was an enriched enrolment study and therefore results could not be included in a meta‐analysis
Galer 2002	Results of this study were contained in a duplicate publication, containing more information with means and standard deviations
Iseki 2000	No random allocation
Meier 2003	Trial contained patients with all types of neuropathic pain, not specifically postherpetic neuralgia
Tajti 1999	No random allocation
Tamakawa 1998	Single blinded study without random allocation

Contributions of authors

All three review authors contributed equally to the first and subsequent drafts of the review. All three review authors agreed upon the final version of the text.

Sources of support

Internal sources

• Guy's, King's & St. Thomas' School of Medicine, King's College London, UK.

• Mayday University Hospital (Mayday Healthcare NHS Trust), UK.

• Kettering General Hospital, Kettering, UK.

• Princess Royal University Hospital (Bromley Hospitals NHS Trust), Farnborough, UK.

External sources

• No sources of support supplied

Declarations of interest

None known

Notes

The original author team are unable to complete the updates for this review. A new protocol is in preparation, with the revised title 'Topical lidocaine for neuropathic pain and fibromyalgia'. Publication of the protocol is anticipated in 2014.

The new full review on topical lidocaine for neuropathic pain has been published. Derry S, Wiffen PJ, Moore RA, Quinlan J. Topical lidocaine for neuropathic pain in adults. Cochrane Database of Systematic Reviews 2014, Issue 7. Art. No.: CD010958. DOI: 10.1002/14651858.CD010958.pub2.

Withdrawn from publication for reasons stated in the review