The second half of the 20th century saw the introduction of a range of therapeutic agents that were shown to have an antidepressant effect. Among these agents were those with a tricyclic chemical structure, which led to their classification as tricyclic antidepressants (TCAs). Even before their introduction into clinical practice, the concept of a link between depression and pain was obvious, and the possibility that this link was causal encouraged the use of antidepressants for patients who exhibited features of both pain and depression.
In 1962, Kuipers reported a case series in which the TCA imipramine was used in patients with “nonarticular rheumatism” and in whom 60% to 70% experienced pain relief. Similarly, Scott reported a double-blind trial in patients with rheumatoid arthritis, osteoarthritis, and ankylosing spondylitis in whom imipramine provided significantly more pain relief than placebo did. In both these reports it was postulated that the pain relief produced is secondary to mood elevation rather than an intrinsic analgesic effect of the antidepressant. It is now recognized that the pain relief apparent with the use of antidepressants can be independent of any alteration in mood caused by the drug, although it has been noted with, for example, doxepin treatment that a reduction in pain is intimately associated with a reduction in depression.
Therefore, the focus of this chapter is on the potential pain-reducing capability of drugs otherwise associated with the treatment of depression. However, pain rarely exists in isolation, and any muscle relaxation, mood enhancement, or improvement in the quality and duration of sleep, all of which are potential effects of antidepressant use, is often a welcome accompaniment of any pain relief that is produced.
Dharmshaktu and coworkers reviewed the efficacy of antidepressants as analgesics, and Häuser and colleagues performed a systematic review and meta-analysis of the role of antidepressants in the management of fibromyalgia (FM).
Classification of Antidepressants
Antidepressants are currently classified partly on the basis of their chemical structure and partly according to their primary in vivo effects ( Box 38.1 ).
Monoamine Oxidase Inhibitors
- •
Harmaline
- •
Iproclozide
- •
Iproniazid
- •
Isocarboxazid
- •
Moclobemide
- •
Nialamide
- •
Selegiline
- •
Toloxatone
- •
Tranylcypromine
Dopamine Reuptake Inhibitors
- •
Amineptine
- •
Bupropion
Serotonin-Norepinephrine Reuptake Inhibitors
- •
Duloxetine
- •
Milnacipran
- •
Nefazodone
- •
Venlafaxine
Selective Serotonin Reuptake Inhibitors
- •
Alaproclate
- •
Citalopram
- •
Escitalopram
- •
Etoperidone
- •
Fluoxetine
- •
Fluvoxamine
- •
Paroxetine
- •
Sertraline
- •
Zimeldine
Selective Serotonin Reuptake Enhancer
- •
Tianeptine
Tricyclic Antidepressants
- •
Amitriptyline
- •
Clomipramine
- •
Desipramine
- •
Dothiepin
- •
Doxepin
- •
Imipramine
- •
Iprindole
- •
Lofepramine
- •
Nortriptyline
- •
Opipramol
- •
Protriptyline
- •
Trimipramine
Tetracyclic Antidepressants
- •
Amoxapine
- •
Maprotiline
- •
Mianserin
- •
Mirtazapine
Tricyclic Antidepressants
The structures of some TCAs are shown in Figure 38.1 .
Analgesic Mechanism of Action
In 1987, Max and colleagues demonstrated that TCAs possess analgesic effects independent of their effects on moods. Later, it was found that the analgesic effects of TCAs tend to occur more rapidly (a week or less after initiating TCA therapy), at lower serum blood levels, and at lower doses than those used for antidepressive effects. It is now clear that the TCAs have a number of diverse effects that contribute to their analgesic effect ( Table 38.1 ). The extent to which each individual TCA exerts these effects differs, which may account for differences in the effectiveness and propensity to cause side effects when members of this class of drugs are used. As noted later in this chapter, not all the proposed modes of action of TCAs are the result of central effects, with a number of possible peripheral actions now becoming apparent.
| Effect | Mode of Action |
|---|---|
| Serotonergic | Interferes with serotonin reuptake Alters serotonin binding to receptors |
| Noradrenergic | Interacts with α 2 adrenoreceptors |
| Opioidergic | Modifies opioid receptor densities Increases opioid levels in some brain areas |
| N -Methyl- d -aspartate (NMDA) receptor | Binds to the NMDA receptor complex Alters NMDA binding characteristics |
| Adenosine receptor | Inhibits adenosine uptake |
| Sodium channel | Blocks sodium channels |
| Calcium channel | Increases densities of L-type calcium channels |
| Other receptors | Inhibits histaminic, cholinergic, muscarinic, and nicotinic receptors |
Serotonergic Effect
The presence of a descending bulbospinal inhibitory influence on spinal neural activity has been well defined in animal models of antinociception. When 5-hydroxytryptamine (5-HT) antagonists are administered, the antinociceptive effect of TCAs is inhibited. Similarly, when central 5-HT systems are depleted with the use of p -chlorophenylalanine, the antinociceptive effects of TCAs are again reduced. Some tricyclics interfere with serotonin reuptake into nerve terminals. In addition, some TCAs alter binding of serotonin to receptors on neural tissue. Although this evidence exists in the animal literature, the contribution of 5-HT to the antinociceptive effects of TCAs and its role in the anti-hyperalgesic or anti-allodynic properties of TCAs in humans have not been established.
Noradrenergic Effect
In a similar fashion to the serotonergic effect of TCAs, the descending bulbospinal noradrenergic inhibitory influence is thought to be important in their analgesic effect. Depletion of central norepinephrine systems with α-methyl p -tyrosine inhibits the antinociceptive actions of TCAs, and α-adrenoreceptor antagonists also have the same effect. Specifically, when phentolamine, a nonspecific α 1 – and α 2 -adrenoreceptor antagonist, is administered with a TCA, antinociception is inhibited. However, when the α 1 -adrenoreceptor antagonist prazosin is coadministered with the TCA amitriptyline in mice, antinociception is observed. Conversely, when amitriptyline is coadministered with the α 2 -adrenoreceptor antagonist RX821002, antinociception is observed, thus suggesting that TCAs derive at least part of their antinociceptive effect by interacting with α 2 adrenoreceptors rather than α 1 receptors.
Opioidergic Effect
The noradrenergic effects of TCAs are thought to be the primary mechanism of action of this drug class, but other indirect actions may also be important. When clomipramine is administered to rats in the formalin test, the opioid antagonist naloxone can completely antagonize the antinociceptive effect of that TCA. Similarly, administration of the delta (δ)-opioid antagonist naltrindole with antidepressants shifts the antinociceptive dose-response curves to the right, thus suggesting inhibition of the antinociceptive effects of TCAs, whereas administration of the enkephalin catabolism inhibitor acetorphan with the antidepressants dothiepin, amitriptyline, or sibutramine enhances their antinociceptive effects. Chronic antidepressant administration can modify opioid receptor densities and increase opioid levels in certain brain regions. The alterations seen with opioid receptor antagonists may represent a direct action of TCAs on opioid receptors; however, this is unlikely given the lack of TCA affinity in opioid receptor binding assays.
N -Methyl- d -Aspartate Receptor Effect
Reynolds and Miller observed that desmethylimipramine and imipramine both prevent the Ca 2+ influx into cultured cortical neurons of the rat produced by N -methyl- d -aspartate (NMDA). They also noted that other TCAs had a similar but less intense effect. Others have observed that antidepressants bind to the NMDA receptor complex and that chronic administration of antidepressants alters NMDA binding characteristics. There is considerable debate on the importance of an NMDA effect in regard to the analgesic effect of antidepressants.
Adenosine Receptor Effect
Adenosine is known to produce analgesia, and antidepressants inhibit the uptake of adenosine in neuronal preparations. The antinociceptive effect of antidepressants is inhibited by adenosine receptor antagonists. Adenosine receptors have both peripheral and central representation (see later discussion).
Sodium Channel Effect
Sodium channel blockade may contribute to the analgesic efficacy of antidepressants. Amitriptyline appears to be the most potent TCA in its ability to block sodium channels, with doxepin and imipramine following (all were superior to bupivacaine) and then desipramine (less effective than bupivacaine), and nortriptyline was one of the least effective TCAs in blocking sodium channels. Sudoh and associates concluded that N -methyl doxepin is a potent Na + channel blocker and a long-acting local anesthetic for rat sciatic nerve blockade.
Although amitriptyline is more potent than bupivacaine in a subcutaneous infiltration model and in an intrathecal administration model in rats and sheep, when amitriptyline was evaluated for ulnar nerve blockade in healthy human volunteers, it was found to be less effective than bupivacaine, contrary to the results of a large number of animal studies. This may be due to the thicker nerve sheaths present in humans than in rats, which presents more of a barrier for amitriptyline to penetrate into the nerve. Local anesthetics and TCAs both bind more tightly to the inactivated state of the sodium channel. Thus, neural blockade with amitriptyline (like local anesthetics) is use dependent.
Potency is extremely difficult to assess because of differences in agents, different routes of administration, different environments, different species, different sodium channels, and different measurements.
Rats differ from sheep and humans, topical administration differs from intrathecal and perineural application, and motor function differs from proprioception and nociception. However, in an effort to present a very rough idea of local anesthetic potency, N -phenylethyl amitriptyline is about 50 times as potent as lidocaine, amitriptyline is about 8 times as potent as lidocaine, and bupivacaine is about 4 times as potent as lidocaine. However, N -phenylethyl amitriptyline appears to have a narrow therapeutic index. N -Methyl amitriptyline is similar to amitriptyline but has a much longer duration of action. Furthermore, N -methyl amitriptyline appears to exhibit significant differential blockade (i.e., selective block of a specific [pain-transmitting] nerve fiber group), greater than that achieved with amitriptyline, bupivacaine, and lidocaine in sheep.
Since amitriptyline blocks persistently open sodium channels at low plasma concentrations, because of its significantly longer half-life, it may have utility in the perioperative period. Lidocaine has been demonstrated to be effective for chronic neuropathic pain states and has also been shown to diminish acute postoperative pain, as well as facilitate return of normal bowel function ; preclinical evidence also seems to support this notion. Amitriptyline demonstrates antinociceptive effects in experimental mouse models of pain states. Additionally, amitriptyline in combination with morphine had better analgesic effects than did morphine alone.
Calcium Channel Effect
Although acute treatment with an antidepressant has no observable effect on calcium channels, chronic treatment with citalopram and chlorprothixene (but not imipramine) increases the density of L-type calcium channels. It also has an antinociceptive effect, with this effect being nullified by administration of nifedipine.
Other Effects
Antidepressants also interact with histaminergic, cholinergic muscarinic, and cholinergic nicotinic receptors in an inhibitory manner. However, individual antidepressant agents may differ markedly in their potency at different receptors. These interactions may contribute to the side effects of the antidepressants (e.g., dry mouth, sedation, urinary retention). These important side effects are discussed in the following sections.
Animal Studies of Antinociceptive Effects
Because of the many proposed pharmacologic effects of TCAs, one would expect them to have an antinociceptive effect. A number of studies have confirmed that this class of antidepressants does have this property. For example, Abdel-Salam and colleagues have shown that antidepressants, including those in the TCA class, display antinociceptive properties in a rat tail electrical stimulation assay, and others have shown that chronic administration decreases self-mutilation in the rat autotomy test, produces an antinociceptive effect in the formalin test, and has similar action in a hot plate test.
Ardid and Guilbaud confirmed that both acute and chronic administration of TCAs (e.g., clomipramine, amitriptyline, desipramine) has an antinociceptive effect in a rat mononeuropathy model. This effect on neuropathic pain has been substantiated in other studies.
Some of the TCAs also seem to possess anti-inflammatory effects. For example, when imipramine is administered on a chronic basis to rats that are then exposed to carrageenan, which induces intense inflammation, the local inflammatory response normally observed is significantly reduced. Similarly, clomipramine reduces carrageenan-induced skin inflammation in a dose-dependent fashion, as well as decreases the prostaglandin E 2 –like biologic and immunologic activity and substance P concentration in the inflammatory exudate. When both amitriptyline and imipramine are administered on a chronic basis to rats with adjuvant-induced arthritis, behavioral tests suggest that they both induce antinociception.
Human Experimental Pain
Even with a single 100-mg dose of imipramine, verbal pain ratings and the amplitude of somatosensory evoked cerebral potentials in response to suprathreshold intradermal electrical stimuli are reduced significantly, more than in subjects taking placebo. Similarly, a single oral dose of desipramine has been shown to increase subjective pain thresholds and the nociceptive withdrawal reflex threshold in response to percutaneous electrical stimulation of the sural nerve. Poulsen and associates examined the effect of a single oral dose of 100 mg of imipramine on pain detection and tolerance thresholds to heat and pressure, thresholds of the quadriceps femoris muscle withdrawal reflex to single and repeated electrical stimulation of the sural nerve, and continuous pain ratings during the cold pressor test in 12 healthy volunteers. They found that imipramine significantly increases pain thresholds to heat and pressure, as well as the pain tolerance threshold and reflex threshold to single electrical stimulation. Pain ratings during the cold pressor test and pain detection thresholds in response to heat and pressure were unaltered. These studies have suggested that TCAs can have a differential hypoanalgesic effect in different human experimental pain tests.
Tricyclic Antidepressants in Clinical Pain Management
Historically, TCAs were used for human pain management before their modes of action as analgesics were elucidated. The fact that they can reduce pain and independently elevate mood, as well as normalize sleep patterns and cause muscle relaxation, is an additional potential benefit of their use. In no human field of use is the evidence for an analgesic effect of TCAs greater than in neuropathic pain conditions. A significant body of evidence underpins the use of TCAs for a number of specific neuropathic pain conditions, and because the features of neuropathic pain are not dependent on the causal disease or neural irritation, it is widely accepted that the evidence for analgesia in specific conditions is strong enough to allow uniform use for any condition manifesting the symptoms of neuropathic pain.
Post-herpetic Neuralgia
A prototypical neuropathic pain condition involving neural irritation and destruction makes post-herpetic neuralgia (PHN) a particularly difficult condition to treat. With established PHN, palliation rather than cure is the only prospect. Perhaps in no other condition have TCAs made such an impact. Evidence has suggested that amitriptyline, nortriptyline, and desipramine are among the TCAs that can usefully alleviate the suffering associated with PHN. As an example of potential efficacy, Watson and coworkers reported “good to excellent” pain relief in 16 of 24 patients studied, and Max and coauthors reported that 47% of 58 patients studied in their randomized controlled trial obtained “moderate or greater” pain relief with amitriptyline. Interestingly, in a study comparing amitriptyline with the tetracyclic antidepressant maprotiline (which has a predominantly noradrenergic effect), Watson and associates noted that “amitriptyline relieves some patients with postherpetic neuralgia. Many patients suffer side effects and better therapies are necessary.” Incidentally, the pain relief produced by amitriptyline was greater than that apparent after maprotiline.
Watson and coworkers compared the effect of nortriptyline and amitriptyline and found both to have an analgesic effect but that nortriptyline was associated with fewer side effects. Kishore-Kumar and colleagues examined the effect of desipramine on PHN and confirmed an analgesic effect. They stated that “other antidepressants—notably amitriptyline—are known to ameliorate postherpetic neuralgia, but those agents are often toxic.” Almost 2 decades after this study, amitriptyline is still considered a frontline agent and arguably the preferential first therapeutic agent for the treatment of PHN.
Painful Diabetic Neuropathy
Again, strong evidence exists for pain relief with TCAs in patients with painful diabetic neuropathy (PDN). Amitriptyline, desipramine, clomipramine, imipramine, and nortriptyline have all been shown to have an analgesic effect in patients with PDN.
In terms of comparative efficacy, Max and associates found that desipramine and amitriptyline are equally efficacious, whereas Sindrup and coworkers found that clomipramine tends to produce better pain relief than desipramine does. When the dose-response relationship is considered, Sindrup and colleagues found that imipramine is associated with such a relationship. Although a dose-response relationship is also noted when clomipramine is used, this does not seem to be the case with desipramine.
Painful Mononeuropathy and Polyneuropathy
Some evidence exists for an analgesic effect of clomipramine in the treatment of painful mononeuropathy and polyneuropathy pain. Langohr and associates compared treatment with clomipramine and acetylsalicylic acid in a blinded crossover study and were able to show a greater analgesic effect during the clomipramine treatment phase.
Pain Associated with Spinal Cord Injury
Not all studies examining the effect of TCAs on neuropathic pain have produced a positive result. Cardenas and coworkers studied 84 patients with pain from a spinal cord injury (SCI) who were randomized to receive amitriptyline or an active placebo, benztropine mesylate. No significant differences in measured pain parameters were found between the treatment groups or when comparing pretreatment and treatment periods. This evidence is in contrast to that presented by others who have suggested a beneficial effect of TCAs on SCI pain, although their evidence is based on case reports rather than blinded, placebo-controlled trials. In contrast to the negative response in studies of SCI pain, Leijon and Boivie reported a useful analgesic effect when amitriptyline is used in patients with central post-stroke pain.
Fibromyalgia
Although the use of TCAs in patients with FM is widespread, consideration of the evidence supporting their use is difficult, largely because FM is a complex disorder with a spectrum of symptoms and signs. From a broad perspective, there can be little doubt that antidepressants do improve the symptoms of FM in some patients. O’Malley and colleagues undertook a meta-analysis of studies that examined the effect of antidepressants in patients with FM. They calculated that the odds ratio for improvement with antidepressant therapy was 4.2 (95% confidence interval, 2.6 to 6.8). They concluded that antidepressant therapy has a positive effect on sleep, fatigue, pain, and well-being, but not on trigger points. They also found that in only one of the five studies that measured depression scores was there a correlation between improvement in symptoms and depression scores. When TCAs are specifically considered, Arnold and associates concluded from their meta-analysis that TCAs produce the largest improvement in sleep quality, with modest improvement found in measures of stiffness and tenderness.
In terms of the number of patients who can improve with antidepressant treatment, Carette and coworkers found that after 1 month of treatment of FM, 21% were improved (as opposed to 0% with placebo) and that after 6 months of treatment, the proportion had increased to 36% and 19%, respectively.
Osteoarthritis
Historically, an analgesic effect of TCAs was first noted in patients with joint pain. Limited evidence has suggested that TCAs can reduce joint pain caused by osteoarthritis.
Low Back Pain
Although a diagnosis of low back pain is extremely wide and nonspecific and encompasses a wide range of different problems, it does represent a significant clinical problem for many practitioners. A single report of a randomized, controlled trial involving the use of doxepin in patients with low back pain suggested that it can reduce pain and decrease indices of depression.
Cancer-Related Neuropathic Pain
TCAs seem to exert an analgesic effect on a range of painful conditions, but Mercadante and colleagues reported that amitriptyline failed to produce any pain relief in 16 patients with advanced cancer who had features of neuropathic pain. However, their study numbers were small, the neuropathic pain may not have been present in isolation given the diagnosis of cancer, and any neuropathic pain arising in association with cancer could have been emanating from a diverse number of neural structures irritated by tumor deposits.
Human Immunodeficiency Virus–Related Sensory Neuropathy
TCAs do not seem to be effective in relieving this condition.
Comparative Studies
Morello and associates studied 28 patients with PDN in a crossover study comparing the effect of amitriptyline and gabapentin. They found that the pain relief produced by the TCA is of similar magnitude and quality to that obtained with gabapentin.
Overall Effectiveness
One way of displaying the potential efficacy of any agent is to consider the number needed to treat (NNT; Table 38.2 ). In terms of analgesic medication, this represents the number of patients who need to take the treatment to obtain a 50% or greater reduction in their pain.
| Condition | NNT |
|---|---|
| Painful diabetic neuropathy | 3.0 3.4 1.3 |
| Post-herpetic neuralgia | 2.3 2.1 2.2 |
| Atypical facial pain | 2.8 |
| Central pain | 1.7 |
Selective Serotonin Reuptake Inhibitors
TCAs have analgesic potential for various pain conditions and a diverse range of pharmacologic actions, but these actions can also increase their propensity to cause side effects. It was hoped that with the advent of antidepressants with more specific modes of action, analgesia would still be associated with their use and the potential to produce side effects would be reduced.
When the antinociceptive effect of selective serotonin reuptake inhibitors (SSRIs; Fig. 38.2 ) were examined in a mouse hot plate pain test, fluvoxamine induced a dose-dependent antinociceptive effect, whereas fluoxetine and citalopram induced only a weak antinociceptive effect. Escitalopram failed to elicit any antinociceptive effect. The antinociceptive effect of these three SSRIs was not blocked by the opioid antagonist naloxone. In contrast, again using a mouse hot plate test, paroxetine produced an antinociceptive effect that was inhibited by naloxone, thus suggesting that this SSRI may act not only via its serotonergic effect but also via an interaction with the opioidergic system. In the same study, paroxetine-induced antinociception was inhibited by the 5-HT 3 antagonist ondansetron, but not by the 5-HT 2 receptor antagonist ketanserin.
When considering the overall results from studies, it has been calculated that the NNT for one patient to obtain a 50% reduction in pain is 5 for paroxetine and 15.3 for fluoxetine. This leads to the conclusion that evidence for the effectiveness of SSRIs in pain management is limited, at best.
Selective Serotonin Reuptake Inhibitors and Human Pain
Painful Diabetic Neuropathy
A number of studies have examined the effect of SSRIs on PDN. Sindrup and colleagues compared the effects of paroxetine and imipramine. Paroxetine did produce pain relief, but less than that obtained with imipramine. On the other hand, use of paroxetine was associated with fewer side effects than occur with imipramine. Max and associates compared the effect of amitriptyline, desipramine, fluoxetine, and placebo in patients with PDN. When subjects were considered in terms of the percentage of those who derived moderate or greater pain relief, the results were 74% in those receiving amitriptyline and 61%, 48%, and 41%, respectively, in those receiving desipramine, fluoxetine, and placebo. Citalopram has also been studied and found to be relatively effective with few side effects.
Fibromyalgia
A number of studies have suggested that SSRIs have little effect on FM pain. In one, Norregaard and associates studied 22 patients with FM and compared the effect of citalopram with that of placebo. After 8 weeks of treatment (4 weeks taking placebo, 4 weeks taking citalopram), no changes were observed in any pain parameter measured, nor in depression scores. Similarly, Anderberg and coworkers found no difference with citalopram treatment when the results were analyzed on an intent-to-treat basis, but there were reductions in pain and well-being scores in those completing the study.
Although in general the analgesic effects of SSRI are limited and inconsistent, when higher doses are used, the analgesic effects may be somewhat better. Arnold and colleagues showed that patients with FM who received fluoxetine at doses of 45 mg/day or higher had significant improvement in the Fibromyalgia Impact Questionnaire (FIQ) total score, as well as in FIQ pain, fatigue, and depression scores. This makes intuitive sense since at higher doses most SSRI agents will inhibit the reuptake of norepinephrine somewhat. Nakajima and coworkers presented data suggesting that an increase in norepinephrine in the spinal cord plays an important role in the anti-hyperalgesic effects of not only norepinephrine reuptake inhibitors but also SSRIs.
Serotonin-Norepinephrine Reuptake Inhibitors
Serotonin-norepinephrine reuptake inhibitors (SNRIs; Fig. 38.3 ) selectively block the reuptake of norepinephrine and serotonin (5-HT). Milnacipran blocks 5-HT and norepinephrine reuptake with equal affinity, whereas duloxetine has a 10-fold selectivity for 5-HT and venlafaxine a 30-fold selectivity for 5-HT.
Animal Pain Studies
The probable effects of drugs in humans can often be predicted by testing them on specific animal pain models. However, there may be species differences in response; in human clinical practice, mixed pain states are common, whereas animal models are based on specific pain types.
When formalin is applied to animal paws, a two-stage response is observed that can be measured electrophysiologically or by behavioral observation. Both duloxetine (an SNRI) and citalopram (an SSRI) attenuate the second phase of the formalin response. When compared with venlafaxine and milnacipran, duloxetine attenuates the second phase of this test more significantly.
In the tail flick test (a test of acute nociception), duloxetine has minimal effect, whereas in the hot plate test, some antinociceptive response with duloxetine is observed. In the chronic nerve constriction injury model of neuropathic pain, venlafaxine and duloxetine both have a significant antinociceptive effect.
Clinical Pain Management
Duloxetine is the first antidepressant to have a specific pain indication in the United States—treatment of PDN. SNRIs may also be useful for other conditions.
Duloxetine and venlafaxine are antidepressants with both serotonergic and noradrenergic reuptake–inhibiting properties (SNRIs). In the treatment of PDN, duloxetine has been demonstrated to be more efficacious than placebo at doses of 60 and 120 mg/day, although the higher dose appears to be associated with similar efficacy but greater side effects. The side effect profile of duloxetine seems to be more favorable than that of TCAs, especially with respect to anticholinergic and cardiac side effects. Nausea is one of the more common side effects but can be reduced by lowering the dose. In many patients, nausea is self-limited and resolves within the first several weeks of use. Duloxetine has been extensively studied in patients with PDN, FM, musculoskeletal back pain, and osteoarthritis and has been approved by the Food and Drug Administration (FDA) for all four indications. Venlafaxine has been effective in the treatment of PDN and other polyneuropathies except for PHN. A small subset of patients demonstrated abnormalities in cardiac conduction; thus, precautions should be taken in patients with a history of cardiac disease. Venlafaxine should be tapered rather than abruptly discontinued because of the potential for a withdrawal syndrome. At doses lower than 150 mg/day, venlafaxine behaves more like an SSRI; at doses above 150 mg, it behaves more like an SNRI agent. Therefore, pain relief is more likely to occur with doses of 150 mg/day or greater. This point should be taken into consideration when prescribing this medication for analgesic purposes. Venlafaxine is not currently approved by the FDA for any pain indication.
Milnacipran is the SNRI (also referred to as an NSRI) with the most balanced activity on inhibition of norepinephrine and 5-HT reuptake. Milnacipran and its metabolites are eliminated primarily by renal excretion, with approximately 55% of milnacipran excreted unchanged in urine, 19% as a carbamoyl- O -glucuronide conjugate, 8% as N -desethyl milnacipran, and the remainder of the administered dose as other minor metabolites, all of which are inactive.
Desvenlafaxine, or O -desmethylvenlafaxine, is the major active metabolite of the SNRI venlafaxine. Like venlafaxine, desvenlafaxine selectively inhibits neuronal uptake of serotonin and norepinephrine and has little affinity for muscarinic, cholinergic, histaminergic H 1 , and α 1 -adrenergic receptors. Desvenlafaxine has been shown to be active in preclinical in vitro and in vivo models used to predict antidepressant efficacy ; however, there are no robust trials of desvenlafaxine for pain relief. Desvenlafaxine succinate is well absorbed following oral administration, with a mean terminal-phase elimination half-life of approximately 9 to 11 hours.
Few published data are available regarding milnacipran, which has been approved by the FDA in the United States only for FM. In a small randomized study, patients treated with milnacipran were shown to have a greater reduction in pain than were placebo-treated patients with FM. In a published case report, milnacipran was reported to successfully manage trigeminal neuralgia in a 64-year-old patient. The SNRIs have differing activity in both the serotonergic and noradrenergic systems. Table 38.3 compares effects on the serotonergic system relative to the noradrenergic system for multiple SNRIs.
| Agent | Selectivity Potency Ratio (5-HT/NE) |
|---|---|
| Venlafaxine (Effexor) | 30:1 |
| Desvenlafaxine (Pristiq) | 14:1 |
| Duloxetine (Cymbalta) | 5-10:1 |
| Milnacipran (Savella) | 1:1.6-3 |
Painful Diabetic Neuropathy
It is well established that duloxetine reduces the pain associated with diabetic neuropathy. Bearing in mind the relatively high risk for side effects when TCAs are used for the treatment of PDN, dropout rates from studies are as low as 12% with the use of duloxetine, and long-term studies examining the use of this drug for up to 52 weeks have shown that it has a favorable safety profile when taken over this prolonged period. Duloxetine treatment has been associated with modest adverse changes in glycemia in patients with diabetic peripheral neuropathic pain, but this does not have an effect on the significant improvement in pain observed with duloxetine treatment.
Fibromyalgia
Traditionally, TCAs have been used for the treatment of FM. However, their side effect profiles often reduce compliance. Conversely, the SSRIs have a more acceptable side effect profile but are relatively ineffective. The SNRIs, however, combine a relatively low risk for side effects with a relatively high chance of alleviating symptoms.
In a number of large studies, duloxetine has been found to be efficacious, not only in terms of pain reduction but also for many of the other problematic complaints associated with this condition. For example, in a study of 207 subjects, Arnold and colleagues found that duloxetine significantly reduces pain, number of tender points, and stiffness scores while significantly increasing the tender point pain threshold when compared with placebo. Furthermore, measures of quality of life were improved by active treatment. In an even larger study of 354 patients with FM, Arnold and associates confirmed their previous findings and also showed that the beneficial effects of duloxetine therapy are independent of its effect on mood.
These positive effects with duloxetine therapy seem to be reproduced by other SNRIs, with milnacipran and venlafaxine also having been shown to have a positive effect. In Vitton and coworkers’ 123 study of the use of milnacipran in subjects with FM, of the 125 enrolled in the study, 37% reported at least a 50% reduction in pain intensity (as opposed to 14% in the placebo group).
Norepinephrine Reuptake Inhibitors
Reboxetine is a selective norepinephrine reuptake inhibitor (NRI; also known as NARI). Reboxetine has little or no affinity for serotonin and dopamine uptake sites or for muscarinic and histaminergic receptors and is no less effective for depression than the TCAs and SSRIs. Limited clinical data point to reboxetine being a poorly effective stand-alone analgesic.
Tetracyclic Antidepressants
Limited evidence for an analgesic effect of tetracyclic antidepressants ( Fig. 38.4 ) exists. When amitriptyline is compared with the tetracyclic antidepressant maprotiline in patients with PHN, even though maprotiline displays analgesic properties, those of amitriptyline are more pronounced. In animal nociceptive models, mirtazapine exhibits antinociceptive properties with evidence of an antinociceptive effect in the hot plate chronic nerve constriction model of neuropathic pain and in the second phase of the formalin response.
