Library
APS Bulletin • Volume 19, Number 3, 2009
Clinical Practice: Innovations in Practice
Debra Gordon, MS RN FAAN, Department Editor
Antidepressant Medications for Painful Conditions: Another Translational Success Story
By Jane Martinsons, Staff Writer, and Liz Newman, Managing Editor
Editor’s Note: The current article is yet another one in our series highlighting successful translational research/application in pain. The application of antidepressant medications to painful conditions is a clear example of translational research at its creative best. This translation started some 30 years ago with the off-label application of tricyclic antidepressants. This was followed with the establishment of an evidence basis for this application and for some second and third generation antidepressants, leading to the current development of some newer antidepressant-based formularies specifically for primary use with painful conditions.
Experts in the field Edward C. Covington, MD, W. Michael Hooten, MD, and Miroslav “Misha” Backonja, MD, contributed their thoughts about the applications of antidepressant medications.
What role do antidepressants play in treating painful disorders?
Dr. Covington: In recent years, we’ve learned that spine pain probably had as much or more to do with central sensitization as with nociception at the level of the disc and facets. It’s useful to know that duloxetine also works in that model, and it suggests that all of the other SNRIs and tricyclics might also be useful for treating spine pain even though they haven’t been studied in an effective way.
Dr. Covington: About 32 years ago, tricyclic antidepressants started to be used fairly widely in treatment of neuropathic pain. In those days, there was a tendency for neurologists to use primarily carbamazepine (Tegretol ), and for psychiatrists to use primarily the tricylic amitriptyline—so both were using tricyclics to treat neuropathic pain. In Europe, they were using a lot of clomipramine for that purpose.
There was a flood of early studies that tried to identify which drugs work for which pains. A large number of well-controlled studies clearly demonstrated that these drugs were useful for not only neuropathic pain, but also fibromyalgia, migraine prophylaxis, and other pain such as vulvodynia and functional bowel pain. What these painful conditions have in common is that they result primarily from a process of neurological sensitization rather than from acute nociception.
The tricyclics were largely eclipsed in psychiatry by selective serotonin reuptake inhibitors (SSRIs). In the pain field, we were very impatient for these drugs to reach the market because the tricyclics are ‘dirty drugs’ with a lot of side effects. They are frequently associated with orthostasis, especially in the elderly. They prolong the QT interval and cause cardiac arrhythmias. Many of them are anticholinergic and cause serious problems with dry mouth, blurred vision, and constipation. They were problematic enough that we were excited to have something without so many side effects. We eagerly anticipated the SSRIs only to find when they arrived that they were very poor analgesics. It seems that the more selective an antidepressant is for serotonin, the less likely it is to be a good analgesic. Some SSRIs have a mild analgesic effect, but in human and animal models they are clearly far inferior to the tricyclic antidepressants.
Next came serotonin norepinephrine reuptake inhibitors (SNRIs), of which venlafaxine was the first. Venlafaxine was shown to have some efficacy in migraine prophylaxis and neuropathic pain, and it has been used extensively off-label for treatment of fibromyalgia and for some other visceral hyperalgesias. Clinically, many pain specialists thought that it was a very good analgesic in these situations, but the manufacturer never supported sufficient studies to demonstrate this scientifically. [A recent addition is desvenlafaxine, or Pristiq, a metabolite of venlafaxine. As far as I know, it hasn’t been studied in pain, but I’m seeing some people prescribe it. It’s considerably more expensive than venlafaxine and, as far I know, offers no advantage other than improved kinetics and interactions.]
The next drug that came out was duloxetine (Cymbalta), which was indicated for fibromyalgia but also widely used for such conditions as migraine prophylaxis and other neuropathic pains. Recently, a European Journal of Neurology study (Skljarevski et al, 2009) found it to be effective in treating spine pain, which is interesting because nearly all pain clinics spend a preponderance of their time treating spine pain or spine-related pain. Also, a study with maprotiline, a rarely used antidepressant, actually controlled for the presence of depression and neuropathic pain and showed that it was effective for treating axial spine pain.
In recent years, we’ve learned that spine pain probably had as much or more to do with central sensitization as with nociception at the level of the disc and facets. It’s useful to know that duloxetine also works in that model, and it suggests that all of the other SNRIs and tricyclics might also be useful for treating spine pain even though they haven’t been studied in an effective way.
The latest antidepressant marketed in the United States for pain indications is milnacipran (Savella), another SNRI. Interestingly, they did not seek an indication for depression, but only for fibromyalgia, and they got it. It’s a well-established antidepressant in Europe. “Milnacipran is a unique, dual norepinephrine/serotonin-reuptake inhibitor that has been used primarily for depression outside of the U.S. over the past 10 years” (Dellwo, About.com, 2009).
Pain specialists have always treated conditions that were not sufficiently common to warrant the expense for a pharmaceutical manufacturer to seek an indication. I don’t anticipate that we’ll see a company seeking an indication for vulvodynia, for example, even though many people suffer from it and these drugs have the potential to relieve it. It’s usually not a good business decision to fund analgesic studies for uncommon conditions.
We’re on relatively sound scientific ground when we use tricyclics for treatment of most of these conditions. Yet most of us prefer not to use tricyclics as drugs of first choice because they’re likely to lead to either hazardous or uncomfortable side effects. So we face the dilemma of using drugs of demonstrable efficacy that have risks, or choosing drugs that are similar mechanistically for which studies have not been done, and relying on clinical judgment to assess their efficacy.
Dr. Hooten: The newer agents, serotonin norepinephrine reuptake inhibitors (SNRIs) such as duloxetine, for instance, have similar effects as the tricyclics but with possibly fewer side effects. These medications might be more easily tolerated.
Dr. Hooten: The tricyclics, such as nortriptyline and amitriptyline, have been the most widely used medications for neuropathic pain, including diabetic peripheral neuropathy and other neuropathic pain conditions. They are very effective medications, and they have good efficacy. However, they are limited by adverse side effects, including problems due to anticholinergics as well as weight gain. That’s important considering that a lot of folks with neuropathic pain are older individuals who may be more susceptible, or at least more sensitive, to those adverse side effects.
The newer agents, serotonin norepinephrine reuptake inhibitors (SNRIs) such as duloxetine, for instance, have similar effects as the tricyclics but with possibly fewer side effects. These medications might be more easily tolerated.
Tricyclics might be beneficial for fibromyalgia; it’s very clear that amitriptyline specifically has modest efficacy. SNRIs, such as milnacipran or Savella, may show some efficacy for fibromyalgia.
Dr. Backonja: Tricyclics are an older class of drugs and they were systematically studied about 20–30 years ago, which was many years before opioids were systemically studied for chronic pain disorders and that started only in the 1990s. In the early 1980s, a wide range of studies were conducted and the efficacy of tricyclics was demonstrated for relief of pain, which was independent of the effect of mood, and this was best demonstrated in the case of neuropathic pain and headaches. These observations were the basis of which other antidepressants were studied for treatment of neuropathic pain and headaches, and later other pain, such as muscoskeletal low back pain, and other painful disorders, such as fibromyalgia.
How strong is the scientific evidence for using antidepressants to treat neuropathic pain?
Dr. Hooten: There is very strong evidence [based on] multiple randomized placebo-controlled trials that tricyclics, at least, are very effective for neuropathic pain. Because SNRIs are newer drugs there are fewer trials available, yet those that have been done show strong efficacy. For back pain, there is moderate evidence with tricyclics, and the the majority of trials have been positive. Again, the treatment effect is modest, as it is for duloxetine. The SNRI data for low back pain is just emerging, and so there still needs to be more work there.
Dr. Backonja: Regarding the efficacy of tricyclics for relief of neuropathic pain, it’s as strong as, if not stronger than anything else for chronic pain. The number of clinical trials is pretty large. One of the caveats, however, is that earlier studies that served as the foundation were smaller and conducted in single medical centers. Now the standard is to conduct large, multicenter studies.
Dr. Covington: I know of five randomized trials showing that tricyclics treat fibromyalgia, 11 randomized trials in functional GI disorders, and probably 10 or 11 controlled trials in back pain. So even though we don’t have indications for these, actually the scientific evidence is very, very strong.
And as far as best practice protocols, are they available, and if so, what and where?
Dr. Backonja: The uses of tricyclics for treatment of neuropathic pain has been addressed in a number of published treatment guidelines, probably too many to start identifying one by one. Their role is fairly standard by now.
Dr. Hooten: In terms of practice guidelines, the Institute of Clinical Systems Improvement (ICSI), a large conglomerate based in Minneapolis, is developing algorithms and guidelines for a variety of scenarios, from asthma or coronary disease to chronic pain. Their chronic pain guidelines strongly support the use of tricyclics. They even lay out some algorithms about how to use them.
Dr. Covington: I don’t know of any guidelines specifically targeted on the use of antidepressants for pain; however, numerous guidelines for chronic pain in general, rheumatologic pain, neuropathic pain, back pain, and others, do address antidepressants as part of their recommendations. These include the Institute for Clinical Systems Improvement (ICSI). Assessment and management of chronic pain (2008); Chou et al, Diagnosis and Treatment of Low Back Pain: A Joint Clinical Practice Guideline from the American College of Physicians and the American Pain Society, published in the Annals of Internal Medicine (2007); Serge Perrot et al, Guidelines for the use of antidepressants in painful rheumatic conditions (from CEDR (Cercle d’Etude de la Douleur en Rhumatologie) a specic interest group of the French Society of Rheumatology), European Journal of Pain (2006); and EFNS guidelines on pharmacological treatment of neuropathic pain by Attal et al, European Journal of Neurology (2006).
On a personal note, the American Psychiatric Association and the United Kingdom’s National Institute for Health and Clinical Excellence have issued guidelines that state that benzodiazepines are really not first-choice drugs for treating long-term anxiety. I think this is probably even truer in chronic pain patients who are likely to be taking opioids anyway. When you combine benzodiazepines with opioids, you risk impaired cognition, habituation, loss of efficacy, poor motor coordination, etc. As it turns out, a lot of the drugs that we use as adjuvant analgesics are extremely effective tranquilizers; they just aren’t advertised for that. There are 30-year-old studies, for example, showing that several of the tricyclics—for example, doxepin—are as effective as Valium for treating anxiety, and not addicting, and excellent choices to use in patients who have a comorbid addictive disorder. This represents a significant portion of patients with chronic pain. A number of studies demonstrate that some anti-epileptic drugs also have strong anxiolytic effects. So we have drugs that demonstrably improve function, reduce anxiety, and also improve sleep, while at the same time reducing many pains. We can use these as an option to the benzodiazepine tranquilizers. It’s problematic that there is a very large number of patients with chronic pain who are taking opioids, plus benzodiazepines, and then having to take stimulants so that they can maintain consciousness. It makes much more sense to use adjuvant analgesics that have anxiolytic and sedative properties to treat people’s anxiety and insomnia, rather than adding on tranquilizers to their opioids.
What are future trends for the next 5 or 10 years?
Dr. Backonja: SNRIs are newer antidepressants that have a dual mode of action, as opposed to SSRIs that have a single mode of action. SNRIs have primarily demonstrated efficacy in the treatment of a limited number of disorders, primarily neuropathic pain in fibromyalgia. However, a number of studies presented at meetings on duloxetine actually demonstrate efficacy in treatment of low-back pain as well. It pretty much follows the traditional previous observation of efficacy and treatment of the other chronic painful disorders. It’s worth noting that one of the advantages of having medications like duloxetine on board for pain management is a number of frequently associated comorbidities, such as mood and sleep disorders, can be treated as well.
I’m hoping to see wider studies that demonstrate or confirm some of the other advantages of SNRI in treating headache disorders, including migraine headaches or chronic daily headaches. However, there are really no good studies providing evidence that these drugs fit within a realm of multi-drug therapy, otherwise known as combination therapy. Also, a question remains of how to identify patients who will be responders. Given that 40% or 50% of patients respond to a clinically-relevant degree of pain relief, the question is can we identify those patients up front and maximize their treatment, rather than exposing patients to side effects with potentially no benefit?
Dr. Backonja: Given that 40% or 50% of patients respond to a clinically relevant degree of pain relief, the question is can we identify those patients up front and maximize their treatment, rather than exposing patients to side effects with potentially no benefit?
Dr. Covington: I’m not anticipating major developments in the area of antidepressants for pain; minor changes are likely, such as fast-dissolving, longer-acting, extended release preparations. Basically, we’ve got drugs that range from purely serotonergic to almost purely noradrenergic, and so we’ve got the spectrum fairly well covered. I would expect to see changes from the standpoint of convenience, but I don’t expect to see major developments in the SNRIs that will enable us to treat people more effectively than we do now.
One needed change is not so much new drugs as new practice patterns. In our practice we see a lot of patients who have failed several treatments. It is surprising that most of the people with hypersensitivity/neuropathic/ hyperalgesic sorts of pains have had thorough and appropriately administered trials of anti-epileptic drugs, but they’ve either not had trials of antidepressants or the [trials] have been too little, brief, and poorly done. This may relate to stigma—that clinicians are more comfortable telling the patient that an anti-epileptic, rather than an antidepressant, helps with pain. Either the patient doesn’t like being made to ‘sound psychiatric’ or the provider doesn’t want the hassle of assuring the patient that we want to use this because it treats pain effectively in animal models.
I think antidepressants are underused in treating these syndromes. We know that well-controlled studies with the anti-epileptic drugs demonstrate pain reduction, but they never demonstrate pain elimination. Often, a drug considered very successful reduces mean pain from 7 down to 5 on a 0–10 scale. Most of us, however, don’t want to live with a pain of five, if we have a choice. By using both an anti-epileptic and an antidepressant, the patient may obtain a greater degree of pain reduction without increased side effects. I don’t think people do enough of that. I think we’re all holding our breath for new developments outside the SRNI family. We’re looking for calcitonin gene related peptide antagonists, glial inhibitors, and drugs that modify a host of other transmitters and receptors.
Edward C. Covington, MD, founded the Chronic Pain Rehabilitation Program at the Cleveland Clinic in 1979 and has served as its director since that time. He developed a hospital pain consultation service for the diagnosis and management of problematic acute, chronic and malignant pain.
W. Michael Hooten, MD, is assistant professor of anesthesiology at the Mayo Clinic in Rochester, MN. He works in the departments of anesthesiology, pain medicine, psychiatry and psychology, and in the Pain Rehabilitation Center.
Miroslav Backonja, MD, is a neurologist with special interests in pain physiology and the treatment of pain syndromes. Dr. Backonja is associate professor of neurology and anesthesiology at the University of Wisconsin Medical School.
