I'm going to give you a brief overview of the pathophysiology of neuropathic pain and its treatment.

The first point I want to make is that neuropathic pain is now being defined as a disease. Nociceptive pain is some normal physiologic pain, if you like, so it's a symptom. So if you break your leg, it hurts; the disease is a broken leg, the symptom is pain.

But neuropathic pain is different. Neuropathic pain is a disease, like myocardial infarction is a disease. Myocardial infarction may be caused by smoking, or hypertension, or diabetes. Multiple different things contribute to the cause of myocardial infarction, but myocardial infarction is the disease. Similarly, neuropathic pain is a disease, and this slide shows many of the different conditions that can result in neuropathic pain. But the pain is nonprotective. It is something that persists and behaves separately as a disease itself.

There are many different types of neuropathic pain that you come across in your practices. You can see that 2 of the commonest of these are listed at the bottom there, postherpetic neuralgia and the pain associated with diabetic peripheral neuropathy. One, I think, very common form of neuropathic pain that is not included on this list â€' because it was not specifically reported in this paper â€' is radicular pain. So, I think, back pain that radiates into the leg or neck pain that radiates into the arm is neuropathic pain and is a very common form of pain that we're called on to treat.

We're not going to go a lot into the pathophysiology, but this shows the pathways of nociceptive pain. We have a receptor in the palm of the hand, as demonstrated here. A transduction occurs when the receptor is activated and sends electrical impulses up the peripheral nerve fibers; transmission is into the dorsal root where pain fibers are â€' the C fibers, the unmyelinated fibers, and the smaller myelinated fibers. They enter into the dorsal root ganglion, into the spinal cord where they ascend in the spinothalamic pathways up to the thalamus. Then there's the perception of pain in the cortex, interpretation of what that pain means, and a response to the pain, which is mediated by descending pathways. You put your hand on something hot; the message goes up to the brain; it says, "Ouch, that hurts." You then withdraw your hand from the pain. This is the pain pathway that we deal with in nociceptive pain.

In neuropathic pain there's no receptor involved. The nerve is damaged somewhere along this pathway, usually peripherally, but sometimes centrally. So, if you like, the nerves are "broken." They are setting up aberrant pain impulses and result in chronic pain.

Now there are multiple neurotransmitters and different channels that mediate the electrical activity in the afferent pathways, and many of these are at least a potential target for controlling neuropathic pain. Gamma-aminobutyric acid (GABA) is an inhibitory neurotransmitter, and one would imagine that if one can increase inhibition in the pain pathways, you might reduce pain. Tackling GABA has been a bit disappointing, I must say, but it, at least, is a potential target for neuropathic pain.

The sodium channels on the axons can also be blocked, reducing both peripheral and central sensitization so that there's less of this aberrant signal traffic. A newer mechanism that has generated a lot of interest is modulation of the calcium channels. Not calcium-channel blockade â€' like you would do with verapamil or nifedipine â€' but modulation of the calcium channel activity. So, it does not alter normal calcium traffic, but reduces abnormal increased calcium channel traffic, and by that mechanism, it results in reduced afferent signal traffic, reduced pain.

There are multiple different classes of agents that have shown to be effective in the treatment of neuropathic pain. Interestingly, you can even put stuff on the skin, like the lidocaine patch or lidocaine creams and capsaicin. Both of these agents have been studied in double-blind, placebo-controlled studies and been shown to be effective in the reduction of neuropathic pain, although capsaicin is not Food and Drug Administration (FDA)-approved for this purpose.

The opioids, in general, are effective in the treatment of neuropathic pain. Opioids are relatively more effective in the treatment of nociceptive pain, but they are effective in the treatment of neuropathic pain. They should not be used as first-line treatments, but do have an important role.

Antidepressant drugs, both of the tricyclic type and the more recently introduced serotonin-norepinephrine reuptake inhibitors (SNRIs), are also widely used and effective treatments for neuropathic pain, although only duloxetine is actually FDA approved for that purpose. You'll notice that the selective serotonin reuptake inhibitors (SSRIs), are not on this list. They're effective antidepressants, but they have no useful role in the treatment of neuropathic pain.

A wide variety of anticonvulsants or antiepileptic drugs have also been effectively used in the treatment of neuropathic pain. Carbamazepine (particularly for trigeminal neuralgia, tic douloureux) but also some of the others, and more recently gabapentin and pregabalin, are FDA approved for different pain states. In particularly intractable and difficult cases, sometimes we will actually administer drugs intrathecally. Ziconotide, or opioids, or sometimes baclofen actually can have a beneficial effect by intrathecal administration. We don't go to that very often, but it can be very useful in intractable cases.

These are data that were presented a couple of years ago at the American Diabetes Association looking at pain in diabetic peripheral neuropathy. It shows the drugs that are commonly used in the management of neuropathic pain. The first 3 on the list, the nonsteroidals and opioids, particularly the short-acting opioids, I would regard as largely inappropriate management for neuropathic pain. So, it's only when you get down to here where these are the drugs we should be using. Yet, you can see that only 5% to 10% of patients are receiving these drugs. About 25% receive some antiepileptic drug.

This shows us, I think, very nicely that, in fact, most of the treatment for neuropathic pain that is going on is actually inappropriate treatment. Opioids are appropriate; but if you do use them, you should be using the long-acting, not the short-acting. Yet, you can see that fewer than 5% of prescribing physicians are using the long-acting opioids. So, clearly, we need to change the prescribing patterns of physicians in the management of at least 1 pain state, diabetic peripheral neuropathy.

Which of the following FDA-approved drugs is indicated for the pain associated with both postherpetic neuralgia and the pain associated with diabetic neuropathy? So, which is approved for both conditions?

The correct answer is pregabalin. It is the only one that has been approved for both conditions. Duloxetine has been approved for diabetic peripheral neuropathy; lidocaine patch has been approved for postherpetic neuralgia; carbamazepine has been approved only for tic douloureux or trigeminal neuralgia. Gabapentin has only been approved for postherpetic neuralgia. So, pregabalin is the one that has been approved for both conditions.

This just shows again, that for diabetic peripheral neuropathy, 2 of those drugs have been FDA approved: duloxetine and pregabalin. For postherpetic neuralgia 3 of them have been approved: lidocaine, gabapentin, and pregabalin.

Let's go over these briefly.

Let's look at the topical agents first. The lidocaine patch is a 5% impregnation into a patch that you place on the skin. It's thought to act by blocking sodium channels, and because the sodium channels set up the afferent signal traffic, ectopic nerve impulses are suppressed.

Capsaicin, which is the hot pepper oil, acts differently. It actually depletes substance P in the skin, and substance P is a neurotransmitter that actually mediates pain. It also blocks the vanilloid receptor centrally. So, it has a very different mechanism of action. It's not approved for any pain condition, but has perhaps a small role.

Here are some of the data from the lidocaine studies showing that, if you watch the patients without a patch or anything on, just watching a patient makes them temporarily better, but it doesn't last. If you put on a patch that has no active ingredient, again, you get this increased placebo effect, but it quickly wears off. Where you put on the active patch, there is a sustained benefit in reducing the pain. It is perhaps most useful, in my opinion, in that it provides a mechanical barrier. The symptom of allodynia -- the intense painful perception that you get from what would normally be a nonpainful stimulus (the light stroking of the clothes on the skin, or the bedclothes on your feet at night) â€' can be extremely painful in patients with neuropathy. And the patch blocks that.

The downside is that you're supposed to take it off every 12 hours. If you have allodynia, tearing a sticky patch off your skin twice a day is not particularly comfortable. But it does have an important role and probably a bigger role in postherpetic neuralgia than in other neuropathic pain states.

The opioids, I've said, do have a role in the management of neuropathic pain.

They act on the mu-opioid receptor primarily. So, for the traditional opioids, that is their site of action. Tramadol has a dual action that blocks norepinephrine uptake, as well as acting at that mu-opioid receptor. Although not FDA approved, these drugs are effective.

Here are some of the data from one of the opioids, oxycodone, the controlled-release form, looking at pain intensity on your left and also the average weekly pain score on the right, and there's a highly significant reduction in the level of pain, different types of pain â€' steady pain, brief pain. It doesn't matter. There is a consistent improvement in pain with this and other opioids. On the left it shows diabetic polyneuropathy, and on the right it shows postherpetic neuralgia.

The problem with opioids is side effects, and very few people can tolerate particularly high doses. Nausea is common; constipation is common. Down the list, there are a lot of different side effects that you have and that limits their utility and, of course, the concern of habituation.

What about the antidepressant drugs?

The 2 that are useful are the tricyclic antidepressants, which have a broad mechanism of action inhibiting both serotonin and norepinephrine uptake, but they also have major anticholinergic effects, which are the dose-limiting side effects. The chemists have modified these and produced more selective norepinephrine reuptake inhibitors. They affect both serotonin and epinephrine but have much less anticholinergic effect. There's no zero anticholinergic effect, but it's greatly reduced and it makes them easier to use.

We'll go through the tricyclics first, again reminding you that these are not FDA-approved agents for this, but they are still useful. The weakness of most of these studies reported here is the small numbers. For example, the imipramine study was a tiny number, but it was double blind and placebo controlled. The doses that were needed were quite high; many patients are not going to be able to tolerate that, but patients did achieve significant pain relief. You can go through them one at a time. Amitriptyline, again, small numbers, good trial design, high doses, but significant pain relief. The same with desipramine, and we could go through nortriptyline and imipramine and a whole lot of them. These are just some selected examples. This is for diabetic polyneuropathy.

For postherpetic neuralgia, similar data using slightly different agents â€' amitriptyline, nortriptyline, and desipramine. Again, all resulted in a significant reduction in pain compared to placebo. But, again, look at the doses that are being used, 150 mg a day of amitriptyline. I don't know about your practice, but in mine, I don't have many patients who can tolerate those kinds of doses. Patients who have neuropathic pain are usually elderly, they usually have comorbidities, and trying to use these kinds of doses of amitriptyline is fraught with risk.

Here are some of the risks that we come across. The more serious ones are things like cardiac arrhythmias, but also more annoying ones, like sexual dysfunction, constipation, so on, cognitive problems, particularly in the elderly. You can see that the different tricyclic antidepressants have these effects to different degrees. I argue that there is no role for the use of amitriptyline in the treatment of neuropathic pain â€' not because of lack of efficacy, but because of high anticholinergic side-effect profile. I personally tend to use nortriptyline, but others are also effective and have a relatively acceptable side-effect profile.

On to the newer drugs. This is duloxetine, and these were some studies done recently in diabetic polyneuropathy. You can see that different dose levels were used, and the numbers are good â€' over 400 patients for the first study on the left and over 300 for the second study on the right. You can see the different doses used, and you can see a sort of dose-response curve. The top curve is the placebo. We always get a placebo response in patients who enter into pain studies, but certainly a significant reduction in pain compared to placebo in both of these studies.

Side effects are much less frequent than you see in the opioid studies or in the tricyclic studies, but still a problem. The major side effect that I've experienced has been nausea with these patients, and it can be quite difficult and dose limiting. You can see that there are some anticholinergic effects remaining, with constipation, dry mouth, erectile dysfunction; but they're much less common than with the tricyclics.

Remember that there's a general boxed warning for all antidepressants that suicidal ideation may increase in children and adolescents. We're not treating many children/adolescents for neuropathic pain, but just be aware of that possibility. Some more recent data have indicated that duloxetine may not be a good drug in patients who have compromised hepatic function.

Although it's not FDA approved, another SNRI is also being studied in neuropathic pain and diabetic neuropathy and has shown to be effective, at least at the higher doses. You can see that the lower dose of, I think, 75 mg a day does not deviate significantly from placebo, but in doses of 150 mg to 225 mg a day, there is significant deviation. Again, the side effects are listed on the right. Headache is a problem, but it's usually relatively mild.

One problem to recognize is this strange split personality. Some patients get sleepy on the drug, but others get insomnia. So, it's important to administer the dose depending on whether they get sleepy or insomnia. If they get sleepy, give it at night; if they get insomnia, give it in the morning.

What about the anticonvulsants?

There are several studies of these drugs also. Tell me which of the following drugs is not an anticonvulsant â€' valproate, pregabalin, amoxapine, gabapentin, or oxcarbazepine?

Pregabalin is an anticonvulsant, and it is approved for treatment of epilepsy; but amoxapine is not. I have to shamefacedly confess that I had to ask what amoxapine is. It made me feel better that nobody else knew what it was either. But, the other 4 are definitely anticonvulsant drugs.

So, different mechanisms of action for the anticonvulsant drugs. Many of them have sodium-channel blockade â€' in the middle panel, valproate, lamotrigine, and so on are sodium-channel blockers. Oxcarbazepine is a sodium-channel blocker, but it also modulates one type of calcium channel. Gabapentin, pregabalin â€' the modulator is a different type. They're voltage-gated calcium channel, and that reduces the excessive passage of calcium into the nerve terminal, so reducing synaptic activity in the pain fibers.

Here are some data from some of these studies. Oxcarbazepine, again, not FDA approved, but it does show biologically and statistically significant reduction in pain over a relatively short-duration study, just 4 months. Side effects, again, are a problem. We hear this over and over, that one of the problems in treating neuropathic pain is not so much that the drugs are not effective. It is that the drugs are effective only at doses that create a lot of side effects. You can see the side effects listed here. I find dizziness to be the most disabling side effect in patients being treated with oxcarbazepine.

Sodium valproate, again, a randomized placebo-controlled study, but with small numbers of patients (fewer than 50) in both cases; but, again, a significant deviation from placebo in terms of pain, both in diabetic peripheral neuropathy and in postherpetic neuralgia.

The side effects (nausea and vomiting) are mainly a problem in studies because you usually have a forced escalation, and you're trying to push the dose up too fast. In practice, if you escalate the dose more slowly, you tend not to run into these problems.

This is the "biggie," that neuropathic pain is a chronic condition, and you're going to have to treat it for a long time. Weight gain is a real big problem in patients taking valproate. Weight gain is bad in most of us, and it's particularly bad in diabetics who, of course, constitute a significant proportion of the patients you're treating with neuropathic pain. People like me don't worry too much about hair loss -- I've got so little as it is â€' but it is another side effect to be aware of with valproate. Women find that particularly bothersome.

Lamotrigine has also been studied and shown to be effective. I liked the study design actually. What they did was to administer lamotrigine over a period of time, and you can see the usual placebo response; but as the dose was escalated, you can see that the lamotrigine deviated significantly from placebo. At this point, they stopped treatment. I think, it really helps you say this was a true biological response because, once they stopped treatment, the lamotrigine patients came back up to placebo again.

The dizziness, again, is the big dose-limiting side effect. You don't have to worry in adults, too much, about the skin problems. Be aware that Stevens-Johnson syndrome may occur, but really that's much more a problem in children than in adults.

Topiramate was a big disappointment, because topiramate has so many different mechanisms of action. It is a GABA, acts at the GABA receptor, and it's a sodium-channel blocker. It does all sorts of good things â€' or we think they should be good things â€' in the management of neuropathic pain; but it has been studied in 3 separate trials and in each of them shown to be ineffective. In fact, the closest one to reaching statistical significance, at a P value of 0.096, favored placebo. So, topiramate really doesn't have a role in the management of neuropathic pain â€' unless you regard the migraine state as neuropathic, which is sort of trendy these days. It's good for migraine, but not for the types of neuropathic pain that we're talking about.

Gabapentin has been around for a while now; more than 10 years of experience with it. This is a study in postherpetic neuralgia from the United Kingdom showing, again, this statistically and biologically significant reduction in the level of neuropathic pain. No difference between the 2 doses used in the study, 1800 and 2400 mg a day. The recommendation is that you treat these patients with doses up to 1800 mg a day. Large studies, over 300 patients â€' a good study design.

Somnolence and dizziness were by far the commonest symptoms, but, again, this is a study. In studies, there's forced escalation; you increase the dose according to a preordained schedule. In your practice, you increase these drugs very slowly; by doing that, you'll run into those symptoms much less. Patients will still feel chronically tired though, and it's important to be aware of that.

The more recently introduced pregabalin has been available in Europe for about a year and a half, and has been available in North America for about 6 months. Again, it has a similar mode of action to gabapentin. It acts by modulating the calcium channel, by binding to the alpha-2-delta part of the receptor. Here is a study in diabetic peripheral neuropathy (on the left) and in postherpetic neuralgia (on the right).

The lower doses in this study did not deviate at all from placebo. I will tell you that it does not reflect my experience. I'm finding that 75 mg a day, as a single dose administered at night, is very useful for the treatment of night pain, which is such a big problem in patients with neuropathic pain. But, the higher doses work better, and you sort of see a dose response curve. Certainly, the lower doses coincide, but the higher dose does give a slightly greater response.

Side effects are very similar to gabapentin. About 25% of patients develop dizziness or somnolence, but that can be minimized by slow escalation. Recognize that patients on both gabapentin and pregabalin can get edema; it scares a lot of people, but it's totally benign and generally does not need any treatment. Certainly, I've never had to withdraw a drug because of it. It does not need to be treated with diuretics. It's not a sign of heart failure or kidney failure. We don't exactly know what its mechanism is, but do be aware of it.

Not on this list is also a worrying side effect that we need some more information about, and that is, some weight gain did occur in some of the studies. That needs further study.

I think this is an important slide. Many of you, I'm sure, have used gabapentin, but perhaps you're just starting to use pregabalin. This was always the problem with gabapentin â€' that as you increase the dose, you get progressively less absorption. At low doses of gabapentin, you're absorbing about 60% of the drug; but at the more typical doses that we use, you're only absorbing about 30% of it. I think that we have to recognize that, as we increase the dose of gabapentin, we're not actually increasing the serum level proportionately; whereas, over the range of doses studied for pregabalin, it has got exactly linear kinetics, and about 90% of pregabalin is absorbed at all dose levels. So, that makes it an easier drug to use than gabapentin.

So what does the future hold?

One of the ways of looking at how effective a drug is, is to see what proportion of patients, what number of patients, get at least a 50% reduction in their pain level. That's what this shows: the percentage of patients who achieved at least a 50% reduction in the level of their pain. This slide is for postherpetic neuralgia. On the left is pregabalin, which shows that about over 60% of patients on pregabalin achieve a greater than 50% reduction in their pain level compared to placebo. Now, that's a very powerful treatment effect.

You can see that the gabapentin gives a very similar reduction in pain level compared to placebo. Tramadol, an opioid but also with norepinephrine reuptake inhibiting properties, does have a benefit; but it's only about a 30% reduction, only about 30% of patients achieve this reduction in pain level.

This is for diabetic polyneuropathy, looking at pregabalin and duloxetine compared to oxcarbazepine. Remember oxcarbazepine is not approved for this purpose. Again, about 60% of patients achieve at least a 50% reduction in pain level compared to placebo â€' for pregabalin. Not quite so impressive for duloxetine: about 40% interestingly at the lower doses, and only about 20% at the higher doses. But, if one looks at oxcarbazepine, again, about 50% of patients achieve at least a 50% reduction in the level of their pain.

So, this is a newer way of looking at the data. Not new data, but it's just a different way of looking at the data. This is something that your patients want to know: "How much is my pain going to get better?", "What are my chances of getting some benefit from this?"

When I was in medical school, polypharmacy was a "4-letter word," but today we realize that you must use polypharmacy if you're going to successfully manage neuropathic pain. This is one study looking at this, looking at a combination of morphine and gabapentin in the management of pain in both conditions â€' diabetic peripheral neuropathy and postherpetic neuralgia.

You can see on the left, that the pain level at baseline was about 6. That's pretty severe pain. There's the usual placebo response, with a drop in the pain level to about 4.5 with the administration of placebo. When gabapentin was used by itself, there was a further reduction in pain; when morphine was used by itself, an even further reduction in pain. But, if you use the 2 together, it was even better. The problem is in that bottom right-hand box. Look at those side effects â€' 40% or so got constipation and sedation. The problem with using these drugs in combination is the adverse-effect profile.

This is a complicated slide, but I'm going to really simplify it for you. This is looking at patients who were refractory to treatment with multiple other drugs. All patients had at least 3 different drugs. They had gabapentin, they had tricyclic antidepressants, and they had a third agent; so they had to have failed those to get into the study. Here we have the pain level when they were put onto pregabalin. You can see that it was reasonably effective in reducing pain scores. Then the drug was stopped, and they had a drug holiday (DH stands for drug holiday). Their pain popped back up again.

Repeatedly, when the drug was reintroduced and then withdrawn, the patients did get excellent pain relief, almost down into the mild range. So, even if your patients have proved to be refractory to other forms of neuropathic pain treatment, do still consider using pregabalin. It seems to be more potent than the other available drugs.

Then just in the last minute or so to change gear, postherpetic neuralgia, of course, is a terrible scourge. The question is: well, if we don't have to treat postherpetic neuralgia with these drugs, wouldn't that be better? This is a study looking at the zoster vaccine. This is vaccinating old people, not babies; these were people at least 60 years of age. Look at the size of the study. More than 38,000 people were vaccinated against the zoster virus.

Then they looked at the cumulative incidence of postherpetic neuralgia over subsequent years. Now, the cumulative incidence is still very low, but there's a striking difference â€' two-thirds reduction in the incidence of postherpetic neuralgia in patients who had the vaccine compared to those who received the placebo. That has to be weighed against the cost and the side effects of the vaccination. I think that we need more information before making a recommendation because, again, these cumulative incidences are still very low; but it may be that this is going to become an important treatment.

So to finish, these are the agents available to you â€' the pharmacologic agents available to you for the treatment of neuropathic pain. You may use topical drugs, either creams or patches; opioids are effective, but should not be used as a first-line treatment; antidepressants, particularly the SNRIs, but not the SSRIs; tricyclics are great in terms of efficacy, but have a lot of side effects; and then a variety of anticonvulsant drugs, including pregabalin.

Just to finish off, it's important to recognize that there are a lot of drugs we haven't talked about that are not useful for the treatment of neuropathic pain. The SSRIs as I mentioned, nonsteroidals, cyclooxygenase-2 inhibitors, and so on, down the list. These are drugs you should not be using in the treatment of neuropathic pain.