+++
Specific Pharmacologic Adjuvants
++
Many of the antidepressant agents are useful as either primary or adjuvant therapy for neuropathic pain syndromes as well as other types of pain, such as myofascial pain syndromes. It is generally accepted that the mechanism of action of these drugs in pain is complex and may be related to reuptake inhibition of various neurotransmitters important in descending bulbospinal inhibitory pathways.14 Additional mechanisms may be important, including effects on N-methyl-d-aspartate (NMDA) excitatory amino acid receptors,15,16 sodium channel blockade,17 and even other mechanisms. Earlier studies addressed initial speculation that these agents were working as antidepressants only in pain patients with concomitant depression.18 Max et al demonstrated that amitriptyline was effective in patients with diabetic peripheral neuropathy pain both in the presence and the absence of depressed mood. Early comparisons of different types of antidepressants with selective serotonergic or nonselective reuptake inhibition of both noradrenaline and serotonin suggested that drugs with nonselective mechanisms were more effective in treating neuropathic pain states (Fig. 91-7).19 The selective serotonin reuptake inhibitors (SSRIs) in particular were generally believed to be devoid of efficacy in pain conditions, but newer studies suggest that these agents have an occasional but less pronounced role.20-22
++
++
Tricyclic antidepressants (TCAs) have a characteristic 3-ring nuclear structure and are chemically similar to phenothiazines and antihistamines, sharing side effects such as sedative and antimuscarinic actions. These side effects are perhaps the main detractors to the use of these agents in pain treatment, given their limited ability to reach effective doses. Certainly, the sedative properties of tricyclics can be used to advantage in the treatment of concomitant sleep deprivation, which frequently is present in chronic pain syndrome patients. However, many patients feel groggy or cognitively impaired even with introductory doses of these agents. Elderly patients in particular may have cardiac or urinary conditions that are exacerbated by the antimuscarinic actions of tricyclics, and the drugs may be contraindicated in some patients.23
++
Roose et al24 studied the specific SSRI paroxetine in comparison with nortriptyline in ischemic heart disease patients with depression. Approximately 60% of patients in both groups experienced improvement in depressive symptoms, but the nortriptyline group had significantly more adverse cardiac events, involving 18% of the study group. The patients in the nortriptyline group were titrated to target plasma levels over the first 3 to 7 days rather than the more typical slow ramping dose used by many physicians treating pain. Nevertheless, the development of adverse events illustrates the precautions that must be considered with the use of tricyclic agents. The prototype TCAs amitriptyline and imipramine are metabolized via mono-demethylation to the active metabolites nortriptyline and desipramine, respectively. Imipramine, similar to amitriptyline, has been successfully used for treatment of painful diabetic neuropathy.25 However, the second-generation tricyclic agents (nortriptyline and desipramine) tend to be better tolerated because of less intense antimuscarinic and sedative properties and thus are more popular with pain physicians (Table 91-5).26-28
++
++
Amitriptyline was one of the first agents shown to be effective in the treatment of postherpetic neuralgia29 and has been a prototypical pain adjuvant drug for more than 2 decades. Although the drug is effective compared with placebo, its individual effects are generally not sufficient to result in complete analgesia. Preemptive use of amitriptyline in elderly patients has been associated with a decreased incidence of pain prevalence 6 months after the diagnosis of herpes zoster. The author recommended early institution of amitriptyline in this group.30 Dosing of tricyclics for pain treatment generally begins at low evening doses and is increased gradually until limited by side effects or pain improves. Meta-analysis suggests that an average dose of approximately 75 to 100 mg/d of these agents may be required to achieve optimal results.31
++
Studies have demonstrated that amitriptyline and other tricyclic agents have important NMDA-blocking activity15,16,32,33 as well as significant and long-acting inhibitory action at the neuronal sodium channel. Renewed research interest has focused on the use of these agents for neural blockade.17,34
++
A randomized trial comparing amitriptyline with nortriptyline showed that despite the study's small sample size, there appeared to be no major differences between the 2 drugs with respect to pain scores, patient satisfaction, disability, preference, or mood. Intolerable side effects were more prevalent in the amitriptyline group.26 Although most studies have been positive, a randomized controlled trial did study the use of amitriptyline for treatment of spinal cord injury pain compared with the pharmacologically active "placebo" benztropine. The study was conducted over only 6 weeks and did not detect any difference between the groups receiving amitriptyline or benztropine.35 Recently, the role of coadministration of agents acting on different pharmacologic targets has been advocated. Raja et al27 compared the tricyclic agents nortriptyline and desipramine with the opioid drugs morphine and methadone in a randomized group of 76 patients with postherpetic neuralgia36 and reported improved analgesia in postherpetic neuralgia and diabetic neuropathy with combination therapy over either agent alone.
++
Both opioids and tricyclics had statistically significant analgesic effects relative to placebo; however, more patients (54%) preferred the opioids to the tricyclics (30%) despite similar pain reductions. Interestingly, although opioid agents are commonly thought to induce deleterious cognitive effects, the tricyclics were responsible for significant differences in hand-grooved pegboard tests as well as symbol substitution compared with the opioids.
+++
Other Antidepressants
++
Multiple studies have evaluated the potential role of non-TCAs. Previous studies had demonstrated no difference between placebo and fluoxetine.19 Based on the study by Max et al,19 it is believed that drugs with properties of nonspecific reuptake inhibition may be superior as analgesics to those with specific serotonin effects. Venlafaxine, a nonspecific reuptake inhibitor of noradrenaline, serotonin, and some actions on dopamine, has been studied in comparison with imipramine. Interestingly, venlafaxine is similar to amitriptyline in terms of reuptake inhibition pattern. In the comparison with imipramine, the authors found that both agents were effective in a painful diabetic neuropathy model.37 Another double-blind randomized trial of bupropion for treatment of neuropathic pain also was positive (see Table 91-5).38
++
In contrast to fluoxetine, duloxetine—a nonselective serotonin and norepinephrine reuptake inhibitor—has shown promise in the treatment of neuropathic pain and was approved by the US Food and Drug Administration (FDA) for the treatment of diabetic neuropathy. Duloxetine has also been shown to be safe and effective in relieving many symptoms associated with fibromyalgia independent of concurrent depression39,40 and is also FDA approved for this indication.
++
The mechanism of action of duloxetine appears to be descending inhibition, with potent inhibition of reuptake of both norepinephrine and serotonin and weak inhibition of dopamine reuptake (see Fig. 91-7).
++
Duloxetine has no major active metabolites and a long elimination half-life of approximately 12 hours. In studies of duloxetine 60 and 120 mg/d versus placebo for treatment of diabetic neuropathy, both doses were effective in reducing pain and interfering with nighttime pain. Duloxetine appears to be particularly effective for treatment of nighttime pain severity. Significant depression was not present in the study patients with diabetic neuropathy.41
++
Duloxetine therapy had been considered safe, with only the minor risk of elevated serum transaminase levels. Some cases of cholestatic jaundice and hepatitis have been reported postmarketing. Patients particularly at risk seem to be those with preexisting liver disease, with possible increased risk for further liver damage. Deleterious liver effects may be compounded in patients who consumed substantial amounts of alcohol. Gastrointestinal upset, constipation, and other side effects are common findings with use of the drug.41
++
In studies of patients with both painful physical symptoms and depression, duloxetine 60 mg/d was found to be an effective treatment for both.42
++
Evaluating the primary efficacy measure of Brief Pain Inventory average pain scores, improvements in the duloxetine patient groups averaged 25% to 50%. Side effects included nausea, dry mouth, fatigue, and decreased appetite.
++
Milnacipran is a dual reuptake inhibitor that blocks uptake of norepinephrine 3 times greater than 5-HT (serotonin) reuptake inhibition. It is indicated for the management of fibromyalgia in the United States as well as an antidepressant in Europe. The half-life is 8 to 10 hours, and the therapeutic dose range is 100 to 200 mg daily in a twice-daily dosing schedule. It is eliminated primarily by the kidney, 55% excreted unchanged in the urine. Unlike duloxetine, it has limited hepatic metabolism. During clinical trials by Mease et al43 and Clauw et al,44 a larger proportion of patients treated with milnacipran, compared with placebo, have met the 3 criteria of improvement in pain, physical function, and global assessment., Adverse events include, but not limited, to 7 beats per minute increase in heart rate and a 3 mm Hg increase in systolic and diastolic blood pressure. Nausea is the most common side effect when initiating therapy. The practitioner should therefore follow titration recommendations to enhance compliance. Milnacipran is associated with less weight gain than TCAs.
++
Opioid use for treatment of chronic pain has increased considerably over the last 2 decades. In 1980, only 2% of patient visits for chronic musculoskeletal pain conditions resulted in opioid prescriptions. In 2000, this number more than quadrupled to 9%,45 representing a net increase of 4.6 million prescribed opioid visits compared with 20 years earlier. The increase in total musculoskeletal patient visits over the same time period was negligible, indicating that physician-prescribing practices had changed significantly. The authors45 theorized that the reasons for increased prescribing were related to advocacy by major pain organizations,46 increased pharmaceutical company direct marketing campaigns to health care consumers, national guidelines published by the Agency for Health Care Policy and Research,47 and Joint Commission on Hospital Accreditation emphasis on pain assessment and treatment.45 Chronic pain patients were increasingly referred by noninstitutionally employed ambulatory care providers to pain specialists, 34% in 1980 versus 49% in 2000.45 Fewer patients were seen for chronic headache complaints, but more were seen for extremity pain. Unfortunately, these increases in opioid prescribing were not without problems.
++
Prescription opioid mentions in the Drug Abuse Warning Network from 1994 to 2001 were analyzed.48 During that period, hydrocodone combinations increased form 9320 to 21567, an increase of 131%. Similarly, oxycodone combinations increased from 4069 to 18409, an increase of 352% that has not yet plateaued.49 The authors of a position statement noted that illicit use of opioid prescriptions had risen faster than legal use of these drugs.49 They also noted that almost no research had determined the abuse liability of different agents. Nonmedical use had increased from 1% to 3% during that period. In 2000, 2 million people used opioids for nonmedical reasons. Another study noted that opioid analgesics accounted for 9.85% of all drug abuse, an increase of 5.75% from 1997.48 Increases in illicit use may be due in part to unscrupulous physician behavior, the nature of prescribing to chronic pain patients as a group, and illegal substance diversion by some patients. Theft has risen as a new problem for the Drug Enforcement Agency monitoring illicit drug use. A total of nearly 13000 theft loss events occurred in 22 eastern US states from 2000 to 2003, including 4.4 million dose units of oxycodone alone.50
++
Use of opioids for pain syndromes has been studied, but adverse effects are troublesome. In a study of neuropathic pain patients resistant to treatment, patients received either high-strength or low-strength levorphanol capsules up to a maximum of 21 capsules per day.51 Pain intensity, quality of life, psychologic and cognitive function, and blood levels were outcome measures. Patients in the high-strength group took 11.9 capsules (89 mg/d), and those in the low-strength group took 18.3 capsules (2.1 mg/d). Pain was reduced by 36% in the high-strength group versus 21% in the low-strength group. Central poststroke neuropathic pain was least likely to improve. Despite obvious analgesic effects, higher doses produced more side effects without significant additional benefits in terms of other outcome measures. More than a fourth of patients withdrew from the study, and episodic anger and irritability were reported more by the daily group.51 Although the patients had improvement in symptoms, one wonders if the large number of withdrawals and the anger/irritability issues justify this treatment. In contrast to evidence of efficacy in some patients with neuropathic pain taking daily scheduled opioids, other pain states may not be as efficacious. In a population of patients with intractable head pain, 160 sequential patients were monitored, and 70 of these patients qualified for inclusion in the analysis. Of those reviewed, only 26% had more than 50% pain relief. Multiple problems with compliance including requests for early refills of opioids, lost prescriptions, multiple providers of opioids, and other problems occurred in a remarkable 50% of patients. The authors concluded that daily scheduled opioids were associated with low overall efficacy and an unexpectedly high incidence of maladaptive behaviors.52
++
In response to the study by Caudill-Slosberg et al,45 an editorial discussed the current state of opioid-prescribing practices and the lack of a clear evidence-driven approach to opioid therapies for chronic noncancer pain syndromes.53 Evidence-based guidelines have been offered to improve patient selection and adherence monitoring.54
++
In an effort to better manage chronic pain patients taking opioids and to avoid legal and regulatory interference, many pain treatment facilities use an opioid agreement or "contract." To determine the key attributes included in the opioid consumption contracts of major academic centers, Fishman et al55 evaluated the opioid contracts of 39 academic centers and the major care statements/prohibitions for their core meaning. The most common categories included (1) terms of treatment, (2) prohibited behavior, and (3) points of termination. Rank ordering of statements included (1) improper use of medications, (2) disciplining by termination (for missed appointments, medication abuse, etc), (3) limitations on replacement of lost medications or prescription changes, (4) informing physician of relevant information, including side effects and changes in medical condition, and (5) submission to random drug screens. Unfortunately, little evidence indicates that drug contracts improve patient compliance. Monitoring is an intensely time-consuming task that is not possible to accomplish in many clinical settings. Potentially indispensable statements on opioid treatment, including prohibited usage during pregnancy, driving automobiles, prohibition of alcohol use, and many others, are present in less than half of opioid contracts. Long-term opioid studies of chronic pain syndromes generally consist of small numbers and short duration. Short-term data limit the ability to generalize the reproducibility of opioid treatment outcomes to large groups of chronically treated patients.54,56
++
Rajagopal et al57 noted significant decreases in sex hormone production in male cancer survivors taking opioids for at least 1 year. Evidence of central hypogonadism, including decreased testosterone and reduced sexual desire, was noted in this group. This finding is similar to those in patients taking chronic intrathecal opioids.58 Immunologic studies have indicated significant impairment immunologic function caused by opioid therapies, particularly natural killer cell cytotoxicity with chronic opioid use.59,60 In a mouse model,61 the covalent agent buprenorphine did not significantly suppress immunity in contrast to fentanyl. This finding suggests that differential effects may be associated with different opioids.
+++
Opioid-Induced Hyperalgesia
++
It is well understood that several agents (eg, TCAs) may provide part of their analgesic activity by augmentation of descending inhibitory pathways from the brainstem to the spinal cord.14 It also appears that just as there are descending inhibitory pathways, there also are descending facilitatory pathways. Animal models suggest that the rostral ventromedial medulla (RVM) contains populations of neurons that are characterized as either on cells or off cells,62 which may be an important reason for opioid hyperalgesia and apparent opioid-induced pain. The on cells seem to promote nociception, whereas the off cells seem to inhibit nociceptive neural firing. Morphine and other μ-agonists decrease firing of the on cells and increase activity of the off cells. Cholecystokinin (CCK) appears to be a pro-nociceptive peptide neurotransmitter that is important in decreasing morphine antinociception.63 Researchers have demonstrated that within the RVM, CCK may augment descending pain facilitation in response to ongoing opioid exposure. In an animal model, continuous morphine significantly increased thermal and tactile hypersensitivity after 3 days, and microdialysis of the RVM demonstrated a 5-fold increase in CCK in the RVM.64 Previous research had demonstrated that either a lesion of the bilateral dorsolateral funiculus (neuronal tracts used for descending modulation) or lidocaine injection into the RVM could abort evidence of opioid-induced hyperalgesia.65 Evidence indicates that descending facilitation enhances local spinal cord release of dynorphin (a potent algogenic substance), leading to enhancement of spinal afferent nociceptive receptivity. Continuous morphine infusion leads to increasing spinal dynorphin. Dynorphin then evokes release of excitatory peptides such as calcitonin gene-related peptide and substance P at dorsal root ganglia primary afferents.66 Injection of dynorphin antiserum abolishes the pro-nociceptive opioid pain facilitation.66 These studies point to a major change in the way we look at concepts such as opioid tolerance. Increased opioid requirements may be a true tolerance, that is, a rightward shift of dose-response curves, or evidence that opioid-induced descending facilitation is occurring (opioid hyperalgesia). Future research to develop compounds that either block CCK activity in the RVM or block the dynorphin-induced excitatory peptides to allow ongoing analgesia from opioid agents may be indicated.67
++
Tolerance can be characterized as the state of physiologic adaptation to a certain level of opioid dosing, recognized as declining effects of the opioid over time. This may be related to downregulation of opioid receptors or accumulation of antianalgesic substances. For example, opioid therapy may cause NMDA receptor activation and induce intracellular cascades that result in activated protein kinase C feedback inhibition on available opioid receptors.68 Tolerance is predominantly a concept relative to analgesia because a right shift in the dose-response curve occurs over time. Some side effects, such as sedation and cognitive slowing, seem to be less of an issue over time, sometimes resolving within the first 1 to 2 weeks of therapy and being less prominent than with other agents.27 Tolerance is not seen with other side effects such as urinary retention or opioid-induced constipation, which can be an overwhelming problem with higher doses of these agents. Selective opioid receptor antagonists that are unable to cross the blood–brain barrier have been developed to overcome opioid-induced constipation. Methylnaltrexone is administered subcutaneously, and alvimopan is an oral peripheral μ-opioid antagonist. Both have been shown to effectively cause laxation without reversing opioid analgesia.69 Alvimopan is an oral, peripherally acting, μ-opioid receptor antagonist for the treatment of opioid-induced bowel dysfunction.70
++
Physical dependence implies a state of requirement for receptor occupancy to prevent a physiologic withdrawal syndrome after abrupt cessation of an opioid agonist. This state of dependence occurs within days to weeks of starting the sustained delivery of a receptor-specific agonist. In contrast to withdrawal of alcohol, benzodiazepines, barbiturates, and other agents, withdrawal of opioids is rarely a life-threatening situation but nonetheless causes symptoms. Symptoms of opioid withdrawal include piloerection, nausea and vomiting, diarrhea, diaphoresis, agitation, dysphoria, nasal congestion, and seizure. Psychologic dependence is more complex and implies that the patient has an expectation of withdrawal or a fear of lack of analgesia and therefore may resist efforts to taper opioid agents.
++
Addiction can be defined as a primary chronic neurobiologic disease, with genetic, psychosocial, and environmental factors influencing its development and manifestations. It is characterized by one or more of the following behaviors: impaired control over drug use, compulsive use, continued use despite harm, and craving.72
++
Weissman and Haddox73 described a state in which patients who had been given drug doses that were inadequate to control their pain or who had abruptly stopped or were tapered too rapidly from a drug manifested what appeared to be drug-seeking behavior (consistent with addiction). However, in reality the patients simply were experiencing a normal reaction in response to pain that previously was well controlled but now had an acute exacerbation caused iatrogenically. These patients may seem to have significant knowledge of opioid agents and dosages that previously worked for them; thus they appear to be drug seekers.
++
Equianalgesic dosing tables are commonly used to guide therapy with weak or potent opioids based on equivalent doses (Table 91-6). In clinical reality, most patients respond well to opioids dosed in this manner. Sometimes a disparity is observed upon switching to another opioid based on equivalency tables, but the results are generally both safe and "close" to the right dose.
++
++
The equianalgesic dosing tables are universally used by pain specialists when switching from one drug to another. The referenced doses for these tables are based on single-dose administrations, so use of these tables for chronic dosing of opioid agents is unclear. Pereira et al74 reviewed the literature from 1966 to 1999 and determined that the literature lacked studies examining long-term dose ratios and that the published studies were not homogeneous, wide ranges existed, methadone in particular was more potent than originally thought and correlated highly with the dose of previous opioid given, the ratios might change depending on the direction of change, and discrepancies existed with respect to fentanyl and oxycodone.
++
Many researchers have postulated that use of opioid agents in the future may be determined and administered based on pharmacogenomics. An individual's specific genome likely is better suited for specific agents based on the expression of specific receptor characteristics, variations in metabolic processing, and other features. Cloning of opioid μ-type receptors gives credence to this line of reasoning.75
+++
Tamper-Resistant Formulation
++
The increased prescribing of opioids for pain has been associated with escalation of diversion and misuse/abuse of the medications. It is a challenge to balance access of these medications for appropriate use while protecting society from the consequences of abuse of the medications. Attempts to combat the abuse and diversion include increased monitoring such as Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS) System. The RADARS system collects data across the United States through surveys and Poison Control Center reports.76 Additionally, pharmaceutical companies are being asked to submit plans for opioid Risk Evaluation and Mitigation Strategies as part of the FDA drug approval process to ensure that the benefits of the medications outweigh the risks.77
++
Because many medications that are abused are altered to increase the drug delivery, tamper-resistant medications formulations are being developed. In August, 2010 a new formulation of Oxycontin was released. Although not proven to be safer or less likely to be abused the formulation makes it more challenging to crush or dissolve the tablets for nasal inhalation or intravenous injection. Embeda was a sustained-released morphine with a sequestered core of naltrexone. It has been voluntarily withdrawn from the market due to not meeting predetermined product stability benchmarks by the company. If the pellets inside the capsule are tampered with (eg, by crushing), the naltrexone released antagonizing a significant portion of the morphine. The released naltrexone also can precipitate withdrawal if the patient is opioid tolerant. When taken as directed, the naltrexone does not affect the efficacy of the opioid. These two examples highlight some of the challenges related to combating abuse and diversion of opioid medications. It is probably unlikely that all forms of abuse can be deterred (intravenous injection vs intranasal vs swallowing increased dosage of whole pills. Additionally manufacturing complexity may limit pharmaceutical company investment.78
+++
Opioids in Neuropathic Pain
++
Opioid use for many pain states is supported by randomized controlled drug studies showing efficacy. Although often characterized as an opioid-resistant type of pain, neuropathic pain syndromes may have the most hearty evidence for efficacy with opioid therapies.27,79-83 The major neuropathic pain studies generally have examined patients with postherpetic neuralgia, painful diabetic peripheral neuropathy, or central pain syndromes, including postspinal cord injury and postcerebrovascular accident pain. For example, Watson and Babul79 performed a randomized controlled trial involving 50 patients with postherpetic neuralgia and symptom duration of at least 3 months. In this study, 38 patients completed the study, which compared oxycodone tablets at a dose up to 30-mg controlled-release agent twice daily compared with placebo. Pain scores of patients showed improvement in pain relief of 2.9 points versus 1.8 points for the placebo group and a decrease in allodynia and steady and paroxysmal pain.79 Raja et al27 compared the TCA agent amitriptyline to morphine or methadone in a total of 76 patients with postherpetic neuralgia symptoms of long duration. Three treatment periods were studied over 24 weeks, with administration of opioid, TCA, and placebo. Interestingly, although both the TCA and opioid groups had improvements in study end points, 54% of patients favored opioids compared with 30% who favored the TCA. Methadone is commonly thought to be superior to other opioids for treatment of neuropathic pain because of its dual action on μ-opioid receptors as well as NMDA receptors, but this was not demonstrated in the study. In patients with intolerance to morphine, patients switched to methadone achieved doses of only 15 mg versus 91 mg for morphine. Reduction in pain intensity was less for methadone (1.2 points on a visual analogue scale compared with a pain scale reduction of 2.2 points with morphine). Another interesting finding of this study was that although most pain practitioners commonly attribute side effects such as psychomotor and cognitive slowing to opioids, patients taking the TCA actually performed worse on pegboard placement skills and other tests of psychomotor skills.27
++
A new molecule combines the effect of μ-agonism and norepinephrine reuptake inhibition, and therefore it works on ascending and descending pathways. This new drug embodies the concept of the multimechanistic approach into the treatment of acute pain. Although the long-acting formulation is not available yet, several studies have shown its efficacy in multiple chronic pain entities.84-87
++
The 75-mg dose is equivalent to oxycodone 10 mg in acute pain model. However, it achieves pain control with lesser μ-agonism. Hence the side effects of that stimulation are decreased (eg, less nausea and vomiting). It is metabolized in the liver by conjugation with inactive metabolites. The parent drug and its metabolites are eliminated via renal clearance. Time of maximal concentration is 1.25 hours, and its half-life is 4 hours.
++
With our enhanced understanding of the development of both peripheral and central sensitization (windup) and the role of excitatory amino acids acting at NMDA receptors, blockade of this central hyperalgesic response has assumed greater therapeutic importance. Likewise, use of NMDA blockers for opioid tolerance and opioid-induced hyperalgesia is a reasonable therapy. Currently available NMDA blockers include ketamine, methadone, dextromethorphan, amantadine, magnesium, and memantine.
++
Methadone, an opioid first synthesized in the mid-twentieth century, has become the agent most commonly used for management of long-term opioid addiction. Methadone is highly lipophilic and has high oral bioavailability. Unlike many modern-era formulated long-acting continuous relief opioids, it is inherently long acting. The drug is a racemic mixture88; the d-isomer reverses morphine tolerance and blocks development of hyperalgesia thought to be mediated by NMDA receptor activation.89 Methadone is useful for both the treatment of neuropathic pain and the management of cancer pain, predominantly as a second-line agent when other opioids become ineffective. The equianalgesic dosing ratio seems to increase with increasing opioid doses and may exceed 10:1 compared with morphine.90 Dosing paradigms to change from other opioids91 have been designed to more effectively load the drug and prevent respiratory depression secondary to methadone's prolonged pharmacologic actions. Possible side effects at high doses include QT-interval prolongation and a propensity to sudden cardiac death.92 Patients receiving higher doses may be reasonably monitored for QT-interval prolongation.
++
Ketamine, a noncompetitive antagonist at the NMDA receptor, has been found efficacious in the treatment of primary neuropathic pain and hyperalgesic states in both intravenous (IV) and oral applications. When used orally, the drug appears to work primarily through its metabolite norketamine.93 Psychomimetic effects appear to be most limiting, although the deleterious side effects were minimized at lower doses (40-60 mg/d) and when coupled with oral benzodiazepine therapy. Ketamine therapy has been attempted as intermittent IV infusion for chronic neuropathic pain states such as postherpetic neuralgia94 and for posttraumatic pain,95 with some apparent efficacy. In a small study comparing ketamine with magnesium (another NMDA antagonist), ketamine administered IV at 0.3 mg/kg/h resulted in significant improvement in a group of patients with neuropathic pain, but magnesium did not reach significance.96 Other NMDA agents have been used with inconclusive results. A double-blind placebo-controlled trial of the agent memantine in postherpetic neuralgia patients did not show separation from the placebo group in terms of allodynia, spontaneous pain, or hyperalgesia; both groups improved. The authors concluded that the agent was ineffective in reducing postherpetic pain.97 Dextromethorphan also has been studied, with inconclusive results. In a randomized double-blind controlled trial of 2 doses of dextromethorphan over a 10-day treatment period, no clinically significant difference between any of the tested outcome measures was observed.98 Amantadine, better known as an agent for parkinsonism or influenza prophylaxis, is a noncompetitive NMDA antagonist and has been well tolerated from a side-effect profile. IV infusion of amantadine was evaluated for treatment of surgical neuropathic pain (postmastectomy, postthoracotomy, postherniorrhaphy). Amantadine reduced both mean pain intensity and hyperalgesic "windup" pain on a statistically significant basis.99
++
Overall, the NMDA antagonists appear to have therapeutic potential; further well-designed randomized trials with larger numbers of patients are necessary to form conclusions. At this time, methadone and ketamine appear to have better evidence supporting their use.
+++
Clinical Rationale for Antiepileptic Drugs
++
In neuropathic pain, changes in sodium channel expression lead to alterations and accumulation in sensory nerves and nociceptive neural pathways. Voltage-gated sodium channels have critical functions in the neuronal transmission of action potentials. They also play an important role in the genesis of neuropathic pain via an increase in neuronal excitability. For the development of neuropathic pain, complete expression of NaV1.8 seems to be necessary.100 It appears that injured axons are functionally degraded and do not play a role in neuropathic pain. However, the role of neighboring C-fibers that are not injured appears to expand, with C-fibers playing an important role in the development of aberrant pain transmission (Fig. 91-8).
++
++
Thus redistribution of the tetrodotoxin-resistant sodium channels NaV1.8 and NaV1.9 to peripheral sites of injury is critical in the development of neuropathic pain. These sodium channel subunits congregate near the site of injury and result in spontaneous neural ectopic firing. Likewise, in addition to the NaV1.8 and NaV1.9 sodium channels, β3 is found in increased concentration at the site of nerve injury. β3 messenger RNA is found in increased concentration within these neurons. Neuropathic pain and epilepsy have functional similarities. In animal models of limbic epilepsy, sodium channel expression and upregulation of NaV1.3 were increased, and changes in β subunits occurred. Thus the ability pharmacologically to block these redistributed, uninjured, unmyelinated C-fiber sodium channels may be the most useful approach to therapy for neuropathic pain caused by partial nerve injuries.
++
Calcium channels may be altered in neuropathic pain states. A commonly used neural constriction injury animal model shows that, after peripheral ligation injury, the α1δ1 segment of the calcium channel is upregulated in dorsal ganglion neuropathy. This change is associated with the production of tactile allodynia (Table 91-7).101
++
++
Evidence indicates that neuronal-type voltage-gated calcium channels play a role in neuropathic pain induced by peripheral nerve injury. Agents have been developed for intrathecal use acting via the ω-conopeptide, which is known to inhibit N-type–gated calcium channels.102 Gabapentin suppresses allodynia. Dorsal root ganglion α2δ upregulation likely plays a significant role in the neuroplasticity of pain expression after peripheral nerve injuries and contributes to allodynia development. Peripheral but not central axonal injury significantly upregulates the expression of dorsal root ganglion α2δ calcium channel subunits. These precede the onset of allodynia. Evidence also indicates that messenger RNA upregulation occurs after spinal nerve ligation. In a rat model, the L5-L6 dorsal root ganglions ipsilateral to nerve injury showed a 6- to 7-fold increase of α2δ expression at 7 days. Likewise, as recovery from tactile allodynia occurs, the upregulation in dorsal root ganglion α2δ subunits diminishes.101
+++
Specific Anticonvulsant Agents
++
Carbamazepine is a tricyclic compound chemically similar to imipramine. It is effective for treatment of trigeminal neuralgia and has long been the drug of choice.103 It also has been found useful for treatment of bipolar disorder and seizures.104 No studies have compared imipramine with other recently released anticonvulsants for treatment of trigeminal neuralgia. The drug acts chiefly by blocking sodium channels and is thought to be most useful for treatment of shooting or lancinating types of pain. In experimental neuromas, peripherally located ectopic discharges are suppressed within the typical ranges of serum concentrations.105 In patients, dosing generally begins at 100-200 mg/d and can be titrated very slowly (4-8 wk) up to 1200 mg/d. Significant side effects can occur, including agranulocytosis, aplastic anemia, hepatic function abnormalities, and Stevens-Johnson syndrome. The chief drawback to use of carbamazepine is poor tolerability in elderly patients, often related to diplopia, sedation, and ataxia. In particular, combining carbamazepine with structurally similar agents (such as TCAs) may enhance sedating side effects.104 Drug interactions may occur due to carbamazepine induction of hepatic microsomal enzymes. Monitoring of carbamazepine should include baseline serum hepatic function tests, complete blood counts, and serum sodium, with particular screening for complications in the first 4 to 6 months of therapy. Carbamazepine is commonly used for treatment of diabetic peripheral neuropathy and other neuropathic pain conditions.104,107 Table 91-8 compares various anticonvulsant therapies for neuropathic pain, migraine headache, and trigeminal neuralgia.
++
++
Oxcarbazepine is an antiepileptic drug chemically similar to carbamazepine. The drug acts at voltage-gated sodium and calcium channels to block epileptiform or ectopic discharges. The major clinical advantage to carbamazepine may be the requirement for less frequent serum monitoring.
++
Although carbamazepine may induce agranulocytosis, hyponatremia, and hepatic function abnormalities, it appears that hyponatremia is the only reason for ongoing generalized serum monitoring of this agent. An open-label, 9-week trial of 30 patients with diabetic painful neuropathy demonstrated improvement of patients' visual analogue pain scores of 29 points (48.3%), with side effects of dizziness and drowsiness. Other studies have demonstrated improvements in pain with use of oxcarbazepine.108-110 Additional randomized controlled studies with oxcarbazepine are needed to predict the effects of this agent reliably in various pain syndromes.
++
Topiramate is a substituted monosaccharide agent that is significantly different from other anticonvulsant drugs. Topiramate has 3 main mechanisms of action. Like other anticonvulsants, the drug is a sodium channel blocker, with some action at voltage-gated calcium channels. It can potentiate γ-aminobutyric acid (GABA)-mediated inhibition at separate sites from the benzodiazepines. Interest also has focused on its action at the kainate and α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptors, where it can decrease excitatory amino acid (glutamate) neurotransmission. Topiramate has found its greatest use in headache therapy as a preventive agent for migraine headache. Large trials have demonstrated significant decreases in migraine headache prevalence when the drug has been used as a preventive agent.111,112
++
Three randomized controlled trials of topiramate for painful diabetic neuropathy did not support a statistically significant difference between topiramate and placebo in visual analogue scores or any of the secondary efficacy end points.113 However, pain reductions were numerically greater in the topiramate study groups in 2 of the studies. Patients were allowed rescue analgesia throughout this study, and it was believed that this may have confounded the results. In another randomized controlled trial, patients with painful diabetic neuropathy were found to have evidence of analgesia with topiramate. In this study, rescue analgesia was only permitted for a 6-week period. The major findings of the study were that pain visual analogue scores decreased from 68 to 46.2 in the topiramate groups versus a decrease from 69 to 54 in the placebo group (p < 0.003). Fifty percent of the topiramate patients versus 34% of the placebo patients had more than a 30% response.114 The other notable finding was that topiramate reduced body weight significantly versus placebo. The weight loss had no adverse effect on serum glycemic control of diabetic patients. Topiramate has been used in diet clinics to aid weight loss and has been used for essential tremor, binge eating disorders, and bipolar disorder.104 In a study of patients taking topiramate for epilepsy, weight reduction was maintained in 86% of patients despite a return of caloric intake to normal baseline levels. Obese patients appeared to develop reductions in weight of greater magnitude than patients who were not obese.115 Thus it appears that topiramate is potentially an ideal agent for patients with obesity, diabetes, and painful neuropathy or migraine.
++
Topiramate is generally started at 25 mg at night and ramped higher every 1 to 2 weeks on a twice-daily schedule to a target dose of 200 to 400 mg/d. Patient side effects are frequent and include cognitive slowing, somnolence, nervousness, fatigue, dizziness, and diarrhea.104 As with all anticonvulsants, female patients with the potential for pregnancy should be using birth control because of the risks of teratogenicity and altered effectiveness of birth control pills.104
++
Gabapentin is a GABA analogue that binds to the α2-δ1 ligand of voltage-sensitive neuronal calcium channels.116 There is no evidence that it functions through GABA-mediated actions despite its structural similarity with GABA. Gabapentin is approved by the FDA for adjunctive therapy of tonic-clonic and partial seizures and for postherpetic neuralgia at doses of 1800 mg or more. Early case series showed the drug had some efficacy in other neuropathic pain syndromes, including complex regional pain syndrome.117 Subsequently, several excellent-quality randomized studies demonstrated the efficacy of gabapentin for a variety of neuropathic pain syndromes (Tables 91-8 and 91-9).118-123
++
++
Backonja et al117 found that gabapentin was effective for the symptomatic treatment of neuropathic pain in patients with diabetes mellitus. The study compared 70 patients who received gabapentin and 65 who completed the study in the placebo group. Using an intent-to-treat analysis, gabapentin-treated patients had significantly lower mean daily pain scores (p < 0.001) compared with placebo patients. Pain scores decreased in the active gabapentin group from 6.4 at baseline to 3.9 compared with baseline scores of 6.5 versus 5.1 at end in the placebo group. Additionally, secondary outcome measures were significantly better in the gabapentin group, including the Short-Form 36 (SF-36) quality-of-life questionnaire and profile of mood states. Similar to other anticonvulsants, gabapentin should be titrated slowly to maximum doses of 3600 mg/d, usually dosed on a 3 to 4 times per day schedule secondary to the short half-life of 5 to 8 hours.104 Gabapentin is not metabolized, and, unlike carbamazepine, it does not induce hepatic enzymatic function. It is eliminated by the kidneys, and dosage should be reduced according to creatinine clearance in renally impaired patients. Adverse events include somnolence, peripheral edema, memory loss (usually high doses), dry mouth, dizziness, and nausea.
++
Pregabalin is an agent that causes inhibition at the neuronal calcium ion channel. This mechanism of action and its molecular structure are similar to gabapentin, which has perhaps become the first-line broad-spectrum agent used for many painful neuropathic syndromes. Pregabalin has been found to bind selectively at the α2δ1 subunit, an auxiliary protein commonly associated with voltage-gated calcium channels.116
++
Pregabalin binding causes reduction of calcium influx and neurotransmitter release, including substance P, glutamate, and noradrenaline.124
++
In a randomized double-blind, controlled trial, patients received placebo or 150 mg or 300 mg daily of pregabalin, and primary end points included decreased mean pain scores, health-related quality of life, and SF-36 health survey scores. Overall, patients with postherpetic neuralgia had significantly lower pain scores and less sleep interference.125 In another group of patients with painful diabetic peripheral neuropathy, a placebo-controlled trial demonstrated significant improvements in pain scores and sleep interference.126 Pregabalin may be better tolerated than gabapentin, and less dose titration is necessary. Pregabalin has been reported to cause symptoms of somnolence, dizziness, and peripheral edema, which are similar to side effects reported with gabapentin.126 Drug dosage can be started at 50 mg 2 or 3 times daily and increased to 100 mg 3 times daily by 1 week. The drug has been scheduled for both diabetic peripheral neuropathy and postherpetic neuralgia by the FDA. Dosages should be reduced for renal impairment based on creatinine clearance similar to gabapentin.
++
Levetiracetam is a novel oral antiepileptic drug with structural features unrelated to those of other antiepileptic drugs. Most antiepileptic drugs are screened for inhibition of epileptiform activity by maximal electroshock or pentylenetetrazol seizure induction. Levetiracetam antiepileptic action seems to occur via inhibition of neuronal hypersynchronization.127 Levetiracetam appears to have minimal interactions with other agents. It causes occasional increased drowsiness and cognitive dysfunction and may place patients at risk for driving or other activities exacerbated by cognitive dysfunction or altered wakefulness. It has a low protein binding (<10%) and is not hepatically metabolized.104 Small case series have described 3 patients who achieved partial or complete relief of pain related to sensory motor peripheral neuropathy.128 At this time, levetiracetam-controlled studies are limited, and no comparison studies exist for other more commonly used antiepileptic drugs.
++
Lamotrigine is thought to act similarly to phenytoin in decreasing rapid firing of neurons. It causes use-dependent inhibition of voltage-gated sodium channels.104 Several studies have demonstrated the efficacy of lamotrigine in a variety of pain syndromes, but because of side effects it is generally relegated to second-line therapy.
++
In a randomized controlled study of 59 patients with painful diabetic neuropathy, lamotrigine was titrated from 25 mg/d up to 400 mg/d over a 6-week period compared with similar titration of placebo. The lamotrigine group improved compared with the placebo group, and the authors concluded lamotrigine was both safe and effective.129 Lamotrigine also has been studied in HIV-associated neuropathy, with a small improvement seen compared with the placebo group. In this study, the titration schedule was very slow and similar to that used in other studies, beginning at 25 mg/d up to 300 mg/d over 7 weeks. Despite the gradual titration, the dropout rate was high, with 5 of the 13 dropouts from a sample of 42 patients citing rash.130 Among a group of 14 patients with refractory trigeminal neuralgia, those given lamotrigine improved significantly better than those on placebo. Although the patients already were taking either phenytoin or carbamazepine, these drugs were continued during the study. The one patient who withdrew from the study did so during the placebo arm of this crossover design.131
++
In a randomized double-blind, placebo-controlled crossover study, lamotrigine was given to 30 patients with spinal cord injury pain syndromes. The drug was titrated from an initial 25 mg/d dosage up to 400 mg/d over a 9-week period, compared with a 9-week placebo titration. There was a 2-week washout period between each arm of the study. Twenty-two patients completed the trial. Daily numerical pain scale in the group treated with lamotrigine was reduced from 6.4 ± 0.1 to 4.2 ± 0.1, and numerical pain scales decreased from 6.5 ± 0.1 to 5.3 ± 0.1 in the control group. The difference between groups was statistically significant (p < 0.001 at lamotrigine doses of 200-300 and 400 mg). The primary finding of this study was that, in patients with incomplete spinal cord injuries, lamotrigine significantly reduced the pain at the same level of injury or below the level of injury. Patients who had evidence of central sensitization (windup pain) and brush-evoked allodynia were more likely to have a positive response to lamotrigine. The drug appeared to have no effect in patients with complete spinal cord injuries. The drug was generally well tolerated.132
++
Lamotrigine most notably has been associated with the development of Stevens-Johnson syndrome (a severe exfoliating dermatitis that can be a dermatologic emergency). Pediatric patients may be most prone to these dermatologic complications, which may be seen in up to 1% to 2% of patients in that age group.104 Significant evidence now indicates that lamotrigine has a role in the treatment of various neuropathic pain syndromes, but side effects (particularly rash) must be closely monitored, and patients must adhere to a slow titration schedule.
++
Tiagabine has not been studied in a randomized fashion for treatment of pain syndromes. Tiagabine is an inhibitor of the GABA transporter protein GAT-1. It also acts as a GABA reuptake inhibitor, thus increasing available GABA within neuronal synapses and surrounding glia.104 An open-label study of tiagabine compared with gabapentin in 91 patients with chronic pain suggested that both drugs were efficacious.133 Notably, the effects of tiagabine on sleep quality in this study suggest that it may be useful for patients who have both pain and sleep disturbances. Further randomized studies will clarify its role in pain management.
++
Phenytoin has been used for several decades for treatment of epilepsy, cardiac arrhythmias, and various pain syndromes. Phenytoin alters Na+, Ca2+, and K+ conductance and inhibits repetitive high-frequency discharges.104 Phenytoin was used in a study of 40 patients with painful diabetic neuropathy who received either the active drug 100 mg 3 times per day or placebo for 2 weeks, followed by a 1-week washout period before crossing over to the other group. In group 1, 70% of the phenytoin patients versus 25% of the placebo patients had decreased pain. In group 2, similarly 78% of the phenytoin patients had improved scores pain compared with 28% of the placebo patients. Nearly a fourth of the phenytoin patients had complete relief.134 Although phenytoin is cost effective, it does have significant side effects, including gingival hyperplasia, hirsutism, and rash. Long-term use can result in peripheral neuropathy.104 One advantage of phenytoin is its availability in IV form. A study of IV infusion of 15 mg/kg phenytoin in patients with neuropathic pain resulted in significant improvement of several features of altered nerve conduction (shooting pain, burning pain, sensitivity, and numbness). The investigator concluded that IV phenytoin could be useful for acute flares of pain (Table 91-10).135
++
++
During a 1-week open-phase segment of a pilot trial studying the analgesic efficacy of a topical formulation of 1% amitriptyline and 0.5% ketamine, 11 patients had a significant analgesic effect for a variety of different neuropathic pain syndromes, including postherpetic neuralgia and painful diabetic peripheral neuropathy. In a follow-up of this initial pilot trial, 92 patients with mixed neuropathic pain syndromes, including postherpetic neuralgia, postsurgical, posttraumatic neuropathic pain, and diabetic neuropathy with evidence of allodynia, hyperalgesia, and pinprick hypesthesia, were randomly assigned to receive placebo, 2% amitriptyline, or 1% ketamine, combined with a primary outcome measure of change in average daily pain intensity.136 Unfortunately, a reduction in pain score of 1.1 to 1.5 units was observed in all groups, and there were no differences between the groups. Blood concentrations showed no significant systemic absorption. Other studies have shown that higher doses may be required to produce significant analgesia.137 This abstract detailed a comparison of high-dose amitriptyline-ketamine (4% amitriptyline/2% ketamine) cream for 1 week. Of the 129 subjects who responded, 118 patients were randomized to receive either high dose (4% amitriptyline/2% ketamine) or low dose (2% amitriptyline/1% ketamine) cream, or a placebo cream for 2 additional weeks. After 3 weeks of treatment, results showed the high-dose group had average daily pain intensity that decreased from 6.5 to 3.28 versus the placebo pain intensity score of 4.34. There was no difference between the placebo and the low-dose group. The percentage of subjects with 30% reduction was superior to placebo for patient sleep quality and global satisfaction. Plasma levels were detectable in less than 10% of subjects and were well below therapeutic levels. Topical cream was well tolerated. This finding may imply that, compared with the study by Lynch et al,136 higher concentrations of amitriptyline-ketamine cream are required to produce analgesia for mixed neuropathic pain syndromes.
++
Older topical agents for treatment of neuropathic pain include capsaicin. A multicenter double-blind, vehicle control study was conducted by the capsaicin study group.138 The trial was conducted over 8 weeks, and either capsaicin or vehicle cream was applied to painful areas 4 times per day. Pain intensity and relief were recorded every 2 weeks using the physician's global evaluation as well as visual analogue pain scales. A total of 252 patients were studied. Capsaicin showed 69.5% versus 53.4% improvement according to the physician's global evaluation scale, and 38.1% versus 27.4% decrease in pain intensity (see Tables 91-9 and 91-10). Other studies have also demonstrated efficacy for chronic postherpetic neuralgia.139 As in previous studies, initial burning at the application site was noted. The mechanism of action of capsaicin is depletion of substance P at peripheral thermal receptors, a process that can take several weeks. Over the past decade, capsaicin has been used less frequently by pain physicians secondary to poor tolerance of the drug when used by patients with postherpetic neuralgia and other peripheral neuropathic conditions.
+++
Intravenous Analgesics
+++
Intravenous Lidocaine Therapy
++
IV lidocaine has been used for the last 2 decades for the control of predominantly neuropathic pain syndromes. IV lidocaine is generally given via a 1-mg/kg bolus and infusion of 4 mg/kg over 30 to 60 minutes. IV lidocaine for phantom and stump pain was compared with morphine infusion, but lidocaine IV infusion was effective only for stump pain.83 In an animal model, systemic lidocaine suppressed aberrant impulses at the dorsal root ganglia and at sites of experimentally induced nerve injuries. Minimal affect was seen on normal sensory conduction. The median effective dose was lower at dorsal root ganglia sites.140 Local anesthetics administered systemically also depress evoked unmyelinated C-fiber activity in the dorsal cord.141 In another study comparing the effects of systemic lidocaine with morphine, both agents relieved the pain of postherpetic neuralgia.82
++
In a study by Finnerup et al,142 IV lidocaine was found to reduce spontaneous pain significantly in all patients either with or without evoked pain. There was no difference in the number of responders. Both spinal cord injury pain at and below the level of injury as well as brush-evoked dysesthesia were relieved. The authors concluded that agents with systemic sodium channel blocking properties may be treatment options for spinal cord injury pain. IV lidocaine also has been studied for painful chronic diabetic neuropathy pain, with positive results.143,144 Mexiletine, an oral class IB antiarrhythmic drug, also has been used for pain relief and is thought to have oral lidocaine-like activity. Galer et al145 studied a small group of patients receiving one of 2 IV lidocaine doses. Although the higher dose of IV lidocaine was more analgesic, both doses were highly correlated with mexiletine responsiveness. The authors concluded that IV lidocaine could be used as a predictor of ultimate response to mexiletine. Previous studies have demonstrated that oral mexiletine is helpful for painful diabetic neuropathy, particularly in reducing symptoms such as burning and stabbing.146 Mexiletine can often be limited by intolerance to the drug, particularly abdominal symptoms (nausea) and nervousness/tremor in some patients. Dosing should begin with 150 mg twice daily or less, with a goal of at least 10 mg/kg (see Table 91-10).
++
Few trials have examined the combination of different antinociceptive agents either from the same class (ie, antiepileptic drugs) or from multiple classes (pairing opioids, antiepileptic drugs, and antidepressant agents). Although most of these medications have numbers needed to treat of 3 or greater,31 therefore necessitating a multimodal analgesic approach, the rationale for combination therapies has not been adequately studied. A study by Gilron et al81 evaluated patients with diabetic neuropathy or postherpetic neuralgia, comparing gabapentin, morphine, or the combination of both drugs versus placebo. The study demonstrated a reduction in mean daily pain from 5.72 at baseline to 3.06 with the gabapentin-morphine combination. Both agents were analgesic. At the maximal tolerated doses, the combination was well tolerated but did produce higher frequencies of dry mouth and constipation for gabapentin and morphine, respectively. The authors concluded that there was merit to additional trials of this type comparing combinations of antinociceptive agents as a multimodal approach to pain control. Number-needed-to-treat evaluations can be important in determining the relative effects of different pharmacologic treatments. In a review, Sindrup and Jensen31 reported numbers needed to treat of various agents in different classes. The comparison is useful in choosing an agent(s) that has high efficacy and is safe, based on numbers-needed-to-harm analysis (see Table 91-10).
+++
Skeletal Muscle Relaxants
++
Skeletal muscle relaxants include a disparate group of medications with multiple modes of action. Benzodiazepines include drugs such as diazepam and clonazepam. These agents act to depress the central nervous system via the chloride channel component of the inhibitory neurotransmitter GABA receptor. Benzodiazepines have broad effects on patients, including functioning as anxiolytics, sedatives, and relaxants. Diazepam, as a prototypical benzodiazepine, has rapid onset of action with a prolonged half-life, high lipophilicity, and high protein binding.147 Although generally not considered potent analgesics in their oral/parenteral forms, research has focused on their use as intrathecal drugs where they are quite analgesic.148 Clonazepam has perhaps the most robust evidence for conventional oral or topical forms for conditions such as burning mouth pain and other facial neuropathic pain syndromes/trigeminal neuralgia.149 Clonazepam is perhaps the agent of choice as an inhibitor of myoclonic jerking that can be frequently seen in high-dose opioid patients.150
++
Tizanidine, an α2-adrenergic agonist chemically, has been used for painful spasm and headache prophylaxis.151 The drug initially was marketed for control of spasticity in association with brain injury152 and has been found useful in controlling spastic hypotonia as measured by Ashworth scale improvement.
++
Baclofen, a GABAB agonist used for spasticity, has found limited usefulness for the control of neuropathic pain. Three subtypes of the GABA receptor have been described: GABAA, GABAB, and GABAC. GABAB receptors are primarily located in the lamina I-III area of the dorsal spinal cord.153 A small trial of oral baclofen given to 15 patients compared the effects of the racemic mixture of baclofen versus the levorotatory form alone and noted greater improvement in patients who received the levorotatory form than in patients treated with the mixed form.154 A comprehensive review of intrathecal baclofen use indicates several reports of analgesic effects of baclofen in animals and humans.153
++
More conventionally used oral muscle relaxants include agents such as cyclobenzaprine, methocarbamol, orphenadrine, metaxalone, and carisoprodol. Of these agents, cyclobenzaprine has perhaps the best evidence for efficacy in pain control. Chiefly, cyclobenzaprine was compared with ibuprofen as therapy for acute neck and back pain in a randomized trial.155 In this study, the combination of ibuprofen with cyclobenzaprine provided no advantage over cyclobenzaprine alone. A review of the use of 3 commonly prescribed muscle relaxant agents—carisoprodol, cyclobenzaprine, and metaxalone—demonstrated that these agents represent 45% of all prescriptions for musculoskeletal pain. The review concluded that all were effective, some based on good randomized trial evidence, particularly for cyclobenzaprine. Unfortunately, although carisoprodol represented 13.3% of all prescriptions, its usefulness is significantly limited by its abuse potential.156 Carisoprodol blocks descending reticular formation interneuronal activity and is metabolized to meprobamate. Meprobamate is a potent anxiolytic and sedative agent with a high propensity for physical and psychologic dependency, which ultimately led to the parent drug's removal from the market. Meprobamate is similar in function to barbiturates.157 The danger of the effects of carisoprodol on driving have been detailed, with more of the impaired drivers manifesting higher blood carisoprodol levels. Impairment appeared to mimic that seen with benzodiazepine use but with several other manifestations, including horizontal nystagmus, hand tremor, and involuntary movements.158
+++
Nonsteroidal Anti-Inflammatory Drugs
++
NSAIDs are used on a daily basis by millions of people and on an occasional basis by many more. The proportion of patients using NSAIDs is nearly 4-fold higher in an older age group compared with a younger group.159 In a study of 3154 elderly patients in Italy, 24.7% were taking NSAIDs, a third chronically.160 As the percentage of baby-boomer Americans older than 65 years advances in the next 2 decades, the number of patients taking these agents likely will continue to expand. NSAIDs are commonly and appropriately used for many acute pain syndromes, including traumatically induced pain, surgical pain, pain from sprains and strains, temporomandibular joint pain, some headaches, menstrual cramping/dysmenorrhea, and some chronic conditions such as rheumatoid arthritis, seronegative arthritides, osteoarthritis, and other painful states. The ultimate goal of therapy with NSAIDs is to decrease inflammation, harness any ongoing tissue destruction, prevent peripheral sensitization of nociceptors by locally released inflammatory factors (bradykinin, histamine, substance P; Fig. 91-9), provide analgesia during restorative healing, and to do so without causing concomitant gastrointestinal injury, renal effects, or cardiovascular thrombotic events. The major mechanism of NSAID action occurs by arresting the action of COX enzyme breakdown of arachidonic acid precursors. The COX enzyme is involved in the synthesis of cyclic endoperoxides from these arachidonic acid precursors to prostaglandins (Fig. 91-10). Prostaglandins have varied effects, often as proinflammatory substances, as well as generation of a gastric mucous membrane protective barrier and modulation of renal plasma flow and electrolyte balance. Two isoforms of the cyclo-oxygenase (COX) enzyme exist: COX-1 and COX-2. The COX-1 enzyme normally is present or constitutive and functions in homeostasis, as in the gastric mucosa. One of the main side effects of prolonged NSAID use is gastric ulceration related to inhibition of COX-1 activity. The COX-2 form of the enzyme is constitutive in the central nervous system and renal tissues but is inducible at sites of pain and inflammation. Efforts to block the inducible form of the enzyme at sites of peripheral inflammation (ie, COX-2) led to the introduction of COX-2-selective NSAIDs in the last decade.161
++
++
++
The COX-2-specific agents had demonstrated lower risks of gastric ulceration compared with nonspecific COX agents, but all COX inhibitors have a propensity to cause gastrointestinal bleeding depending on dose and duration of therapy. Platelet aggregation is permanently affected with aspirin, which irreversibly acetylates platelet COX. The older nonacetylated salicylates, such as salsalate and magnesium choline salicylate, as well as the selective COX-2 blockers, appear to have less effect on platelet function.161 Thus some NSAIDs and aspirin, in particular, may react adversely with other anticoagulant agents (eg, warfarin).
++
Unfortunately, the apparent effects of increased cardiovascular risk have led to the withdrawal from the US market of 2 COX-2 selective drugs (rofecoxib and valdecoxib) and have derailed the ultimate availability of an IV COX-2 form of valdecoxib (parecoxib). Celecoxib, which is much less selective for the COX-2 isoform than rofecoxib and valdecoxib, remains available. It is thought that, to some extent, all NSAIDs as a class may have deleterious effects on the cardiovascular system. Therefore, use of these agents for prolonged periods, particularly by chronic pain patients, should be done with caution and with appropriate balancing of risks versus benefits. An initial study investigated the risks of rofecoxib compared with naproxen was performed (Vioxx Gastrointestinal Outcomes Research [VIGOR]).162 In the VIGOR study, 8000 patients with rheumatoid arthritis were treated with rofecoxib 50 mg/d or naproxen 1000 mg/d. A statistically significant increase in nonfatal myocardial infarctions was noted in the rofecoxib patients compared with the naproxen patients. In the Celecoxib Long-Term Arthritis Safety Study (CLASS),163 20% to 22% of patients were commonly receiving low-dose aspirin in addition to celecoxib. Early recommendations that concomitant low-dose aspirin therapy along with the COX-2 agent might be protective against thrombotic events were discussed with practitioners.
++
In the Adenomatous Polyp Prevention on Vioxx (APPROVe) trial, 2586 patients with a history of colorectal adenoma were studied because it appeared that NSAIDs might have beneficial effects in primary prevention. Comparing a 25-mg dose of rofecoxib with placebo, 46 patients in the rofecoxib group and only 26 patients in the placebo group had confirmed thrombotic events, which included myocardial coronary syndromes, cerebrovascular events, and cardiac/pulmonary failures. The risks became apparent at 18 months of therapy.164 Nussmeier et al165 studied the use of perioperative valdecoxib in coronary bypass patients and found a similar result. As for rofecoxib, valdecoxib appeared to increase the incidence of acute thrombotic events compared with placebo therapy, even though these patients already had diseased vasculature.
++
NSAIDs as a rule are weak acids and highly protein bound. Particular care is mandatory when prescribing these agents to older debilitated or cachectic patients who are critically ill, are intravascularly volume depleted, or have protein-deficient nutritional states. NSAIDs are metabolized by either the CYP 2C or CYP 3A cytochrome P450 enzymes and are excreted in the urine. It is generally believed that NSAIDs are prescribed according to their specific chemical class. Many of the more popular analgesic agents are propionic acids (eg, ibuprofen, ketoprofen, oxaprozin, naproxen), but multiple classes exist. No evidence indicates that one particular agent or class is superior in specific patients if comparable doses are given. Some clinicians believe that a failed drug trial from one class necessitates a trial with another group, but this has not been demonstrated.166 Patients with a history of other gastrointestinal problems, who are nicotine or alcohol users, who are taking concomitant oral steroids, or who are taking warfarin or other anticoagulant substances are least likely to do well. The wisdom of long-term NSAID use as a chronic pain treatment appears to be suspect considering the high degree of adverse drug side effects.
++
An association has been reported between long-term administration of NSAIDs (>180 d) and increased radiographic evidence of osteoarthritis progression. A large group of 1695 patients with either hip or knee osteoarthritis was studied. Radiographs from baseline to a mean of 6.6 years were reviewed. The authors concluded that chronic use of diclofenac, a generally well-tolerated and popular NSAID, was associated with a 2.4-fold (hip) or 3.2-fold (knee) increased risk of osteoarthritis progression compared with short-term use.167
++
Topical NSAIDs may have a therapeutic effect in certain well-localized situations. A systemic review and meta-analysis of topical NSAIDs for musculoskeletal pain concluded that although the number needed to treat was approximately 4.7 to achieve one patient with 50% or more pain relief, the topical agents were well tolerated with few side effects. Interestingly, 3 trials in the meta-analysis demonstrated no difference between oral and topical NSAIDs.168
++
In some cases, acetaminophen may be a safer alternative than NSAIDs. Even though millions of patients take acetaminophen daily, the drug's mechanism of analgesia is not clear. Previously, acetaminophen was considered a weak nonspecific COX inhibitor with little anti-inflammatory action and predominantly central versus peripheral effects, but this may not be correct. An emerging concept of its mechanism of action is that acetaminophen has a COX-3 inhibition profile. COX-3 is expressed in the cerebral cortex and heart in humans and effectively and selectively blocked by acetaminophen as well as some NSAIDs.169 A more recent proposal is that the acetaminophen functions via supraspinal serotonergic mechanisms (Table 91-11).170
++
++
Early studies comparing acetaminophen with NSAIDs have been somewhat surprising, particularly given the fact that acetaminophen's actions are primarily central. Bradley et al171 compared acetaminophen 4 g/d with ibuprofen up to 2.4 g/d in a randomized double-blind experiment of 182 patients with knee osteoarthritis. The patients had improved rest pain and less disability independent of which agent they given, and the authors concluded that presumptive evidence of synovial inflammation does not necessarily predict which agent will provide a greater response. In another study of 382 patients comparing rofecoxib, celecoxib, and acetaminophen, patient functional end points including measures such as walking pain, rest pain, and morning stiffness improved in all groups, but only significantly for rofecoxib. In terms of average 6-week functional disability score, only the 25-mg rofecoxib group was statistically different than acetaminophen. No cardiac events were noted, and adverse events were generally limited.172 Given that rofecoxib is no longer clinically available, these results indicate that practitioners should be wary in potentially higher-risk groups and strongly consider acetaminophen.
++
Acetaminophen-induced hepatic failure has been thought to be a risk for patients with existing hepatic disease as well as those ingesting toxic levels of the drug. Patients ingesting 2 oz or more of alcohol per day should have dose reductions to a maximum of 2.5 g/d acetaminophen.166 Risk of analgesic nephropathy is possible with acetaminophen but is a rare occurrence. Evidence suggests that even in patients with hepatic disease, although the serum half-life of acetaminophen may be prolonged, glutathione stores are not sufficiently depleted to critical levels at recommended doses.173 Overall, according to American Pain Society guidelines, the favorable risk-to-benefit ratio of acetaminophen warrants ongoing use if efficacy is present during the first few weeks at doses of at least 2 to 3 g/d.166