Where does it hurt?

These are the “four little words” that Eli Lilly (in ubiquitous ads) is encouraging us to ask our depressed patients. The idea, evidently, is that if we psychiatrists get into the habit of evaluating somatic pain symptoms, we will discover them in many of our patients, and we will adjust our treatment plans accordingly. Lilly, of course, is hoping that our plan will include a prescription for duloxetine, but any drug company would be foolish not to jump on such a bandwagon: there are a lot of people out there who hurt.

The proper use of antidepressants (ADs) for pain in psychiatric patients is both complex and controversial, and Dr. Scott Fishman weighs in with his own views in this month's interview. In this article, we'll cover some of the relevant research and end with some common sense recommendations. We'll also touch, somewhat lightly, on the use of anticonvulsants for pain, focusing on those meds psychiatrists are most likely to use.

To begin with, how good is the evidence that ADs work for "straight" pain-that is, pain in the absence of depression? Well, the evidence is surprisingly strong. Hundreds of studies have been published, along with several metaanalyses. Fishbain, for example, did a meta-analysis of 93 placebo-controlled trials of various ADs for various pain syndromes. A majority of studies found tricyclics more effective than placebo for headache, fibromyalgia, chronic low back pain, arthritis pain, and miscellaneous acute pain syndromes. SSRIs largely flunked in this analysis, however (Ann Med 2000; 32:305-316).

Other reviews have also concluded that SSRIs are only marginally effective pain meds, but what about SNRIs, such as Effexor (venlafaxine) and the soonto- be-released Cymbalta (duloxetine)? Of these two, Effexor is the only one with published efficacy data for pain without depression; it was shown to be as effective as imipramine for painful polyneuropathy (Neurology 2003; 60:1284-1289), and showed promise in uncontrolled trials for migraine and tension headaches (Headache 2000; 40:572- 80), and for fibromyalgia (Ann Pharmacother 2003; 37:1561-5).

When it comes to comorbid depression and pain, however, Cymbalta has plenty of controlled data to recommend it (see TCR Jan 2004 for a review). As mentioned in that review, the methodology used in the Cymbalta studies included an outcome measure--the Visual Analog Scale for pain (VAS)--that has rarely been used in trials of other ADs. Thus, whether Cymbalta is actually more effective for "achy depression" than other ADs is still an open question.

Indeed, Effexor has recently entered the VAS fray with an uncontrolled study in which patients with depression and chronic pain showed improvement in both Ham D and VAS scores over the course of a year (Am J Ther 2003; 10:318-23). And we would guess that most TCR readers would have little trouble recalling depressed, somatizing patients in their practices in whom standard SSRIs have apparently quelled both physical and emotional anguish.

What about the anticonvulsants for psychiatric patients in pain? Of course, the only approved psychiatric indication for anticonvulsants is the treatment of bipolar disorder, an approval shared by Depakote (valproic acid) and Lamictal (lamotrigine). In addition, Tegretol (carbamazepine) is widely considered effective for bipolar disorder. All three of these medications are decent analgesics, and are approved for some painful conditions.

Depakote ER is FDA-approved for migraine prophylaxis, and is started at 500 mg QD for the first week, then increased to 1000 mg QD. As you can see, these doses are somewhat lower than what we are accustomed to prescribing for our bipolar patients. It's irksome and odd that Depakote ER is approved for migraine but not bipolar disorder, whereas standard Depakote is approved for bipolar but not migraine, but experienced clinicians won't let these relics of corporate marketing strategies get in the way of good patient care.

Tegretol is approved for one pain condition (trigeminal neuralgia), and considered effective for most types of neuropathic pain. Lamictal is not quite as well studied as Tegretol, but is also considered effective for neuropathic pain (Drugs 2000; 60:1029-52). Of these two, most psychiatrists would prefer to prescribe Lamictal, given the lack of drug interactions or need for blood monitoring. True, it may rarely cause Stevens-Johnson syndrome, but so can Tegretol (see TCR 1:9).

One of the most effective anticonvulsants for pain is Neurontin (gabapentin), which unfortunately is almost useless psychiatrically other than as a very expensive hypnotic (see TCR 1:3 for a review of these studies). But if you feel moved to treat your patients's neuropathic pain, Neurontin is a good bet because it's so safe and widely effective (Clin Ther 2003; 25:81-104). Just be forewarned that it has to be dosed fairly high for analgesic efficacy, in the range of 1800-3600 mg QD.

Of course, the list of novel anticonvulsants hardly ends here. The newer darlings, all of which have been used in psychiatry, are Topamax (topiramate), Trileptal (oxcarbazepine), Gabitril (tiagabine), Zonegram (zonisamide), Keppra (levetiracetam), and the soonto- be-approved pregabilin. See TCR 1:9 for coverage of Topamax and Trileptal, and stay tuned--we’ll get to all of them eventually!

TCR VERDICT
Pain: There’s plenty we can do with the meds we know.