Modified from Besse JL, Gadeyne S, Galand-Desme S, et al: Effect of vitamin C on prevention of complex regional pain syndrome type I in foot and ankle surgery, Foot Ankle Surg 15:179–182, 2009.

Pain is present in more than 90% of patients, but CRPS goes beyond the beneficial aspect of pain (fight-or-flight response) and is pathologic. Early diagnosis is possible but not frequent. Unilateral extremity involvement is most common. Inciting causes can include athletic injury (fracture, sprain, dislocation), fasciitis, tendinitis, surgical procedure, peripheral nerve injury, immobilization in a cast or splint, infection, deep venous thrombosis, malignancy, vasculitis, herpes zoster, polymyalgia rheumatica, and cerebral vascular accident.

CRPS almost always includes inappropriate pain.59,83 The pain is deep, spontaneous, burning, and diffuse. Pain can be triggered by a breeze, the lightest touch of a physician’s hand, or even certain emotionally laden words. Patients often describe some combination of the skin feeling raw, an electric-like shooting pain, or a deep, dull bone ache. Cutaneous and deep hyperpathia are typically found. Allodynia is present, depending on the time interval from the initiating event. The pain may be sympathetically maintained, sympathetically independent, or a combination of the two. After exercise, the discomfort usually worsens.

Autonomic changes consist of swelling (edema), abnormal vasodilation, skin warming, and changes in sweating pattern (e.g., hyperhidrosis or hypohidrosis) (Fig. 14-1). Vasoconstriction with associated trophic changes come later in the disease. Motor changes demonstrate decreased active range of motion, decreased strength, and increased physiologic tremor (dystonia) of the distal extremity. There may be difficulty in initiating motion. Sensory changes include hypoesthesia, hyperesthesia, allodynia, and anesthesia dolorosa. Trophic changes (e.g., shiny skin, nail changes, lack of hair growth, contractures, muscle atrophy, osteoporosis, and cool, pale skin) are late findings (Fig. 14-2).

Figure 14-1 Classic complex regional pain syndrome (reflex sympathetic dystrophy): swelling, redness, and temperature change. (Courtesy Mark Myerson, MD.)

Figure 14-2 Severe osteoporosis in a lateral radiograph of a 42-year-old man with complex regional pain syndrome for 6 months.

With a full-blown syndrome there is distal extremity pain and diffuse swelling; smooth, shiny skin with abnormal color and temperature is observed. Allodynia and joint pain are present. Juxtaarticular osteoporosis on radiographs and increased uptake on a three-phase bone scan may be demonstrated in the third phase. This diagnosis is excluded by the existence of conditions that would otherwise account for the degree of pain and dysfunction.

Physiology

Complex Regional Pain Syndrome Type I

A review of current literature suggests that CRPS I has a very complicated pathophysiology at several integrated levels of the entire neurologic pathway. Changes can involve the somatosensory, sympathetic, and sudomotor systems, which result in the imbalance between inhibitory and excitatory nerve function (Fig. 14-3).⁷⁰

Figure 14-3 Systemic outflow from spinal cord and course of sympathetic fibers. Heavy lines indicate preganglionic fibers; thin lines indicate postganglionic fibers. (Modified from Pick S: The autonomic nervous system, Philadelphia, 1970, JB Lippincott. Reprinted with permission of Dr. Joseph Pick.)

In the neuropathic state, peripheral areas of the nervous system not normally responsible for pain transmission are altered and now conduct pain. A-β fibers, normally pathways for touch, start to sprout and grow into the C-fiber terminals, becoming fast transmission conduits for pain.³ Nerves outside the main area of injury may be recruited, thus explaining the late development of pain in peripheral, noninjured anatomy.

With chronic central sensitization, afferent sensory information is relayed to the cerebral cortex via the thalamus. The chemicals released from the damaged tissue (histamine, substance P, bradykinins, prostaglandins, serotonin, acetylcholine) centrally and peripherally remain elevated, and neuronal plasticity develops. At this point, the structure, function, and biochemical profile of nerves actually modify.¹⁹

In CRPS, the balance of glutamate (an excitatory neurotransmitter) and γ-aminobutyric acid (GABA) (an inhibitory neurotransmitter) becomes unstable. Phosphorylation of the neuropeptide N-methyl-D-aspartate receptor (NMDA) is overactivated by the glutamate and causes increased release of calcium and potassium ions, further sensitizing the NMDA receptor. The α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors also become more active.

In addition, under these circumstances, increased substance P (a vasoactive neuropeptide causing vasodilation and extravasation) is released, and pain is processed at lower thresholds. Upregulation of sodium channels propagates nerve conduction, and the balance between inhibition and excitation is lost. This amplified response to sensory input results in allodynia.

Subsequent to these biomechanical abberations, abnormal feedback mechanisms develop. Normally dormant genes that encode ion channels, receptors, and neurotransmitters are activated,²³ causing sympathetic tone to persist inappropriately.^36,43 With chronic pain, cortical reorganization of the representation of the extremity in the central nervous system (CNS) can also take place.⁴⁹

The autonomic nervous system controls thermoregulatory and nutritional blood flow. Inappropriate arteriovenous shunting and abnormal receptor input can cause vasomotor, temperature, and trophic changes. A hot, swollen foot and a cold, ischemic, stiff foot can both be explained by this mechanism (Fig. 14-4).^42,43

Figure 14-4 Equinus deformity after arthroscopy. A, Clinical view. B, Lateral radiograph.

The study of neuropathic pain is still in its infancy, but we can say that normal regulatory mechanisms become impaired. At the periphery, nerves become sensitized, spontaneous ectopic discharges become common, and axonal sprouting and degeneration occur simultaneously. Centrally, there is increased sensitivity, disinhibition, and reorganization of synaptic connectivity. This response varies in different patients. If left untreated, irreversible anatomic and physiologic changes can occur, and symptoms can continue for years.

Complex Regional Pain Syndrome Type II

CRPS type II (causalgia) is a pain syndrome that develops after a distinct peripheral nerve injury. The trauma releases neurogenic inflammatory mediators, causing spontaneous pain associated with allodynia. The sensory deficit is usually limited to the territory of the nerve but tends to spread.

Skin blood flow and sweating abnormalities are also usually restricted to the nerve territory. Motor function deficits are explained by injury to motor axons. Swelling and trophic changes are discrete. Peripheral nerve changes are responsible for the CRPS II pathophysiologic phenomena.²

Sympathetically maintained pain (SMP) or sympathetically independent pain (SIP) may be a component of either syndrome, and symptoms of CRPS I and CRPS II may be present in the same patient. Animal research using sympathetic blocks, α-adrenergic receptor agonists, and electrical stimulation suggests that sympathetic nervous activity can be altered to produce pain. In the diabetic patient, the neuropathic painful foot has selective sympathetic denervation, but the neuropathic painless foot does not. The degree of pain is altered by the intensity and ongoing nature of the stimulus, afferent input, efferent modulation, CNS interpretation, and physiologic adaptation.³¹

Psychiatric Manifestations

Good evidence indicates that true CRPS I or II is not psychogenic. No personality disorder is specific to CRPS. Severe pain of the disease complex is the direct cause of the psychologic suffering, not the converse.9,17 One can often see anxiety, hopelessness, and depression in the patient. However, there is no question that other factors, such as fatigue, emotional state, stress, gender, and culture, have been shown to affect pain transmission.

In our litigious society, it is also imperative to determine whether legal issues are involved. Obviously, there are situations in which litigation influences the way a patient responds to treatment or documents symptoms.

CRPS is still clearly a diagnosis of exclusion, and the clinician must maintain a high index of suspicion. Such patients are often difficult to deal with and usually try one’s patience. The evaluation is intense, the patient is often anxious, frustrated, and occasionally angry. The importance of a complete history and physical examination cannot be underscored but understandably stretches the time constraints of a busy clinical practice.

Florid cases are easy to diagnose, but most cases are confusing; many patients present with only one or two of the classic symptoms involving pain of unknown cause. No two people with this disease state have the exact same disorder. Even with the assistance of a myriad of scientific tests, the physician’s role here is classic: listening, witnessing, and relieving suffering.

For CRPS type I, the test seems to be more reliable when done within 6 months of the onset of symptoms and in patients older than 50 years.7,20,28 A negative bone scan in the early phases of the disease is quite common. Despite intensive study, however, the physiologic significance of bone scanning is still unclear.

Heat, cold, and touch sensitivity can be assessed with a quantitative thermal stimulator, von Frey hairs, or electrical stimulation (Table 14-2).

Table 14-2

Simple Bedside Sensory Testing for Complex Regional Pain Syndrome Types I and II *

Symptom	Technique
Dynamic mechanical allodynia	Stroking with brush, gauze, or cotton applicator; hair movement Vibration by tuning fork
Static mechanical allodynia	Mechanical light pressure
Mechanical hyperalgesia	Manual pinprick by safety pin Calibrated stimulator
Mechanical summation	Apply dynamic stimuli every 2-3 sec, 3-6 times
Heat allodynia	Contact with objects kept immersed in 40°-42° C (104°-107.6° F) water
Heat hyperalgesia	Contact with objects kept immersed in 45°-47° C (113°-116.6° F) water (feels weakly painful to the clinician)
Cold allodynia	Contact with room-temperature metal object, with refrigerated object, and with coolants

^*Compare the opposite side for each test.

From Gracely RH, Price DD, Roberts WJ, et al: In Jánig W, Stanton-Hicks M, editors: Reflex sympathetic dystrophy: a reappraisal. Progress in pain research and management, vol 6, Seattle, 1996, International Association for the Study of Pain, pp 151.

Plain radiographs and computed tomography (CT) are not specific. Plain radiographs can show patchy demineralization at 3 weeks, usually more than would be expected from disuse alone. Magnetic resonance imaging (MRI) can show bone marrow stress reaction earlier than the bone scan (Fig. 14-9).

Figure 14-9 A and B, Underlying talar contusion on magnetic resonance image rules out complex regional pain syndrome.

Attempts to block α-adrenergic receptor activity are not entirely reliable or reproducible and may be fraught with the problems of placebo effect and false-positive results. Despite these limitations, response to the block indicates that sympathetically maintained pain is involved. Use of blocks is therefore important for establishing the diagnosis of CRPS and planning treatment (Table 14-3).

Table 14-3

Techniques for Blocking the Sympathetic Nervous System

The IASP and others no longer regard differential spinal blockade to be reliable. At best, it can be considered a screening test of sympathetic function.⁷¹ Nerve blocks must be evaluated in terms of the entire clinical picture (e.g., decreased pain, temperature increase of 3° F or more, venous engorgement, change in skin color). Therefore sympatholysis, whether pharmacologic or by nerve conduction block, is critical in disease determination. Pain conditions without any response to sympathetic block are defined, by contrast, as sympathetically independent pain states.

Sudomotor asymmetry (resting or evoked) is thought to be significant. The quantitative sudomotor axon reflex test assesses for sweating abnormalities.

Isolated cold stress testing (cold stressor test), combined with thermographic imaging, can measure total digital (great toe) blood flow and microvascular perfusion (Fig. 14-10). The procedure combines a stressor (refrigerated environment) with a measurement technique and can provide accurate assessment of autonomic and vasomotor stability and the effectiveness of sympathetic blocks.^28,44 Limb surface temperature differential of 1° F or more is thought to be significant (Fig. 14-11).^4,25,51,70

Figure 14-10 Equipment for isolated cold stress testing.

Figure 14-11 Thermogram showing differential blood flow.

Laser Doppler flowmetry, with appropriate stressors, can monitor background segmental vasomotor control. It is fast, noninvasive, painless, and relatively free of artifact interference.

Electromyographic (EMG) and nerve conduction velocity (NCV) testing should be done together if an underlying discrete nerve lesion is suspected. This is critical to the diagnosis of CRPS II. Unfortunately, insurers might rule out this diagnosis because results of these studies are negative. These tests measure large-fiber disease. The critical distinction to make is that much of neuropathic pain is small-fiber (C-fiber)–mediated pain and that small-fiber abnormalities are not well tested with EMG. Therefore EMG/NCV testing has some internal inconsistencies (Fig. 14-12).

Figure 14-12 Electromyography and nerve conduction velocity testing should be done together if an underlying discrete nerve lesion is suspected.

Quantitative sensory testing is used to objectively quantify perception thresholds. Precise reproducible stimuli allow comparisons between symptomatic and asymptomatic areas. Skin biopsies are being performed at selected laboratories around the country that actually show damage to the neural elements themselves in affected patients. Their clinical role is unclear.

Conservative Treatment

The main goal of treatment in CRPS is to quench the fires of the exaggerated symptoms and normalize the functioning of the extremity. Early recognition and urgent, assertive treatment are critical. Undertreatment of this pain still persists despite intense public and clinical education programs to teach about the proper use of analgesics. Estimates are that half of the patients untreated in the first year have significant residual effects and risk chronicity.

Initially, CRPS I and CRPS II can manifest with the same symptoms. Physical therapy, sympathetic blocks, pharmacologic agents, and repeated careful clinical examinations are the foundations of treatment. A clinician treating this condition must take control of care management, beware of the patient’s tendency to doctor shop, look out for many different physicians prescribing at the same time, and look carefully at the literature. Much of it is based on studies in failed populations.

Physical Therapy

If CRPS is diagnosed early, nonsteroidal antiinflammatory drugs (NSAIDs) should be started, mild analgesics given, and intense physical therapy initiated (desensitization; heat; whirlpool, gentle progressive range of motion). The goal of pain reduction is to facilitate function and improve the quality of life. These patients usually have intolerance to cold, and moist heat is effective.

When using physiotherapeutic maneuvers, it is best to minimize joint movement in the affected region to reduce the mechanoreceptor barrage and its increase in perceived pain.⁷⁰ One should begin with gentle gliding exercises. One should emphasize to the therapists, even by including it in the prescription, that forcing patients beyond their tolerance can worsen the syndrome.

Recreational therapy should also be part of the program. Have the patient take part in previously lost or new pleasurable activity. This increases mobility, while the socialization helps work on the often-associated depression.

Nonpharmacologic Options

Transcutaneous Nerve Stimulation

A transcutaneous nerve stimulator (TNS) can help reduce pain and allow the patient to progress, especially in patients in whom the pain is limited to one dermatome. This has been shown to release endorphins (naturally occurring morphine-like substances that relieve pain) and can block nerves with benign impulses, preventing their use for pain transfer (Fig. 14-13).

Figure 14-13 Transcutaneous nerve stimulation unit for complex regional pain syndrome.

Acupuncture

Alternative medicine in the form of acupuncture is receiving new interest. It has been part of Chinese medicine for 2500 years and is based on the general theory that patterns of energy (Qi, pronounced “chee”) flow through the body and are essential for health. All body parts are divided into yin and yang. In the human body, tendons and bones are yin and the skin is yang. Harmony and good health occur when the yin and yang are in perfect balance. Disruptions of this balance cause disease by changing the flow of Qi, blood, and body fluids.⁴⁷

Acupuncture points are located where the Qi moves to the surface of the body (361 points). Insertion of the needles at the appropriate points causes a stimulus to travel to the spinal cord, releasing enkephalin and dynorphin (central neuropeptide endorphins) that attenuate pain transmission (Fig. 14-14). According to the gate control theory of pain, the stimulation of large-fiber transmission during treatment “closes the gate” to pain by overwhelming the nerve’s transmission capability. Stimulation is controlled by the length and depth of insertion, the shape of the needle, and needle stimulation (heat, electricity, vibration, and rotation).

Figure 14-14 Acupuncture on the right foot to help treat allodynia resulting from tatoo removal treatments.

Electroacupuncture has provided relief of pain, vasodilation, and erythema, even in some refractory cases. Treatments are usually done at low frequencies (less than 10 Hz) for 20 minutes. Unfortunately, the relief from acupuncture is usually transient, and reimbursement for the procedure limits access to it.⁵³

Nerve Blocks

In adults with suspected CRPS, nerve blocks should be given early. These syndromes can partially respond to sympathetic nerve blockade. Increased limb temperature and at least 50% relief in pain after the block are good signs (Fig. 14-15).

Figure 14-15 In adults with suspected complex regional pain syndrome, nerve blocks should be given early. These syndromes can partially respond to sympathetic nerve blockade. Increased limb temperature and at least 50% relief in pain after the block are good signs.

Neural blockade can provide sympathetic pain relief much longer than pharmacologic agents can. If associated motion loss and trophic changes are present, physical therapy, especially during the pain-free period provided during the block, is essential. Blocks may be used as often as every other day for 2 weeks to achieve long-lasting effect.7,46

Other Methods

Trained psychologists use relaxation therapy, biofeedback, and cognitive-behavioral modification with guided imagery for selected patients. Acupuncture and these other options should not be reserved for when standard methods fail; they should also be used in conjunction with other treatments.

Pharmacologic Agents *

Opioids

The use of opioids (compounds with morphine-like characteristics) for the treatment of neuropathic pain is no longer controversial and should be part of your armamentarium. Neuropathic pain is not opiate insensitive, and it was suggested at the Fourth World Congress of The World Institute of Pain, Budapest, 2007 that these drugs be titrated for therapeutic efficacy versus side effects. Opiates have the side effects of nausea, constipation, respiratory depression, urinary retention, orthostatic hypotension, confusion, and sedation, but these can be managed clinically. Opioids also inhibit rapid eye movement sleep, which can disturb sleep patterns.¹⁸

Lowering the dose of opioids whenever possible is advisable. However, modern pain management recognizes that there is no ceiling to the therapeutic dose of opioids. As stated by Tennant,⁷² “There is no upper limit on the amount or number of different opioids prescribed for pain relief as long as the administration of the medicine does not produce physiologic abnormalities as diagnosed by face-to-face, hands on physical examination.”

Current clinical practice encourages the use of opiates to cover all of the painful time as opposed to on an as-needed basis. There should be fixed doses of long-acting medication (e.g., transdermal fentanyl patch, MS Contin, OxyContin), with the ability to supplement with additional short-acting medicine for breakthrough pain (oral transmucosal fentanyl, oxycodone, hydromorphone, or hydrocodone).⁶⁴ In the elderly, long-acting opiates may be more efficacious than tricyclics.

When prescribing opioids, you must become comfortable with the clear clinical differences among tolerance, addiction, and dependency. Tolerance, which may be mediated through the NMDA receptor, implies that the patient has gotten used to the medication and needs a higher dose to produce the same effect. Addiction is an irrational dependence on drugs, which requires psychiatric care. Physical dependence means that the patient has to be titrated slowly off the medication to avoid withdrawal. Warnings of inappropriate opioid use include the following:

Impaired control over drug use

Refusal to participate in a medication taper

Reports that nothing but a specific opioid works

Preference for short-acting versus long-acting opioids

Use of multiple prescribers and pharmacies

Not taking medicine as prescribed

Loss of medication more than once

Use of nonprescribed psychoactive drugs in addition to prescribed medication

Because of legal ramifications, it is important to carefully document in the patient’s chart your intentions in the use and administration of these substances.

Although analgesics are often used when treating patients with CRPS I, and their use is described in various treatment protocols and guidelines, the scientific support for their administration to patients with CRPS I is very limited.

One placebo-controlled randomized clinical trial (n = 43) was found that investigated the effects of sustained-release oral morphine on patients who had previously been treated with epidural spinal cord electrical stimulation (ESES).³⁴ No significant differences were found between the extent of pain reduction and the average time for ESES to become effective. On average, the morphine group reported 20 side effects per day, compared to 2 per day in the placebo group. An uncontrolled study with nine CRPS I and II patients evaluated the effects of continuous infusion of morphine in the axillary plexus after stellate blockade.⁵ Significant pain reduction at rest and at movement and increased grip strength were found. However, the steady-state morphine concentrations were lower than the minimally effective analgesic concentration. There is little information on weak- and strong-acting opioids in patients with CRPS I. Systematic reviews on their use for neuropathic pain have found tramadol to be effective.²²

Positive short-term effects have also been reported for strong-acting opioids administered for neuropathic and musculoskeletal pain.³⁵

Atypical Opioids

Atypical opioids such as tramadol, an opioid receptor agonist, are clearly different from other opioids and are only minimally inhibited by naloxone. Tramadol inhibits the uptake of serotonin and norepinephrine and has been shown to be effective in some patients with neuropathic pain. It has a low potential for abuse, but it should be avoided in patients with a tendency to abuse drugs.⁷

Paracetamol

The use of paracetamol is described in the context of an adjuvant pain protocol in a study about the efficacy of free radical scavengers in treating CRPS I (n = 146).⁶³

No studies were found evaluating paracetamol as a stand-alone treatment for CRPS I. There is no current evidence that paracetamol is effective in treating pain in CRPS I patients.

Nonsteroidal Antiinflammatory Drugs (NSAIDs)

Sixty-one CRPS I patients were retrospectively evaluated with respect to the effects of 60 mg of ketorolac administered by means of a regional intravenous blockade.¹⁴ Twenty-six percent of patients had a complete response, 42% had a partial response, and 31% had no response. Patients with allodynia had significantly less response to the treatment. Conflicting data have been published with regard to the use of NSAIDs in patients with neuropathic pain.⁵⁸

Antidepressants

Antidepressants may be used to relieve pain by blocking presynaptic uptake of norepinephrine and serotonin. They can help lessen the global intensity of symptoms, enabling the discovery of an underlying neurologic problem (e.g., tarsal tunnel, neuroma). However, chronic pain patients might already be taking multiple medications that have significant side effects. One has to carefully watch for oversedation when adding an antidepressant to such a regimen.

In considering the use of antidepressants, four major issues must be considered:

1. Major depression and chronic pain are common conditions, and they frequently overlap.

2. Antidepressants can improve symptoms of major depression, regardless of the presence or absence of comorbid pain (although pain can reduce the chances of optimal recovery).

3. Antidepressants improve pain symptoms, regardless of the presence or absence of comorbid major depression.

4. Chronic pain and major depression have a shared neurobiology and appear to have a shared neuroanatomy (in the brain and spinal column) and neurochemistry (norepinephrine and serotonin), with similar hypothalamic-pituitary-adrenal (HPA) axis, autonomic nervous system (ANS), and inflammatory cytokine disturbances.

Tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) have more benign side-effect profiles than classic analgesics and offer the added benefit of often restoring sleep.25,40,55 TCAs, especially amitriptyline (Elavil), work by blocking norepinephrine and serotonin reuptake, thus making more neurotransmitter available. They might also block the NMDA receptor. Amitriptyline also is a sodium channel blocker and imparts a local anesthetic-like quality.

By blocking α-adrenergic and histamine receptors, TCAs can cause postural hypotension and sedation. Their anticholinergic properties can cause dry mouth, constipation, blurry vision, and urinary retention. Desipramine and nortriptyline seem to be better tolerated than amitriptyline. However, all three seem to have a similar clinical effect. Optimal doses have not yet been established, but it is best to start at lower doses and increase until analgesia is achieved or side effects become intolerable. In the elderly, blood levels should be checked.

Selective serotonin reuptake inhibitors (SSRIs) slow the absorption of the neurotransmitter serotonin. They do not have significant cardiac effects and are generally safer than tricyclics. They can cause sedation, agitation, insomnia, and sexual dysfunction.

Multiple-action agents can often be helpful when TCAs and SSRIs fail, because these agents often affect more than one receptor or neurotransmitter system. Their exact biochemical effect on pain is not completely understood. Complications can include weight gain and sexual dysfunction.

Cymbalta, an SSRI, has been approved by the Food and Drug Administration (FDA) for both the treatment of depression and the management of pain associated with peripheral neuropathy. Cymbalta targets two chemicals, serotonin and norepinephrine, that are thought to play a role in how the brain and body affect mood and pain. Dosage for neuropathic pain is 60 mg PO QD, starting at 30 mg PO QD × 1 wk and may increase to 120 mg QD. It is not recommended in patients with impaired liver function.

Serotonin-norepinephrine reuptake inhibitors (SNRIs) are a new class of antidepressants. Like TCAs, they seem to be more effective than SSRIs for treating neuropathic pain because they also inhibit reuptake of both norepinephrine and serotonin. Venlafaxine is as effective against painful polyneuropathies (including painful diabetic neuropathy) as imipramine (the reference TCA), and both are significantly better than placebo. Like the TCAs, the SNRIs, duloxetine (Cymbalta) seem to confer benefits independent of their antidepressant effects. Side effects include sedation, confusion, hypertension, and withdrawal syndrome.

Monoamine oxidase inhibitor (MAOI) use should be reserved for experienced clinicians in special situations, because their complications are potentially fatal (Box 14-2).

Systemic Steroids

Oral or intravenous systemic steroids decrease capillary permeability and stabilize the basement membrane. Tapered oral methylprednisolone (Medrol Dosepak) can be effective early. With regional intravenous (IV) blocks (methylprednisolone sodium succinate and lidocaine HCl), good results were reported in 20 of 24 patients when blocks were started within 6 months of symptoms. Overall, results are mixed.⁷³

Corticosteroids have been used in open trials (n = 64-69)⁴⁵ and in one controlled trial (n = 23)¹² to treat CRPS I, all of limited methodologic quality. All the studies found corticosteroids to have a very pronounced beneficial effect. Corticosteroids may have a positive effect on CRPS I, but little is known as to the duration and dosage.

Oral Sympatholytics

Oral sympatholytics are poorly tolerated. Fatigue, sedation, hypotension, and sexual dysfunction are major problems.²⁰

Bisphosphonates

Three placebo-controlled studies have been carried out to date.²⁶ One study (n = 20) involved administration of alendronate 3 days in a row.¹ Another study evaluated the efficacy of clodronate (n = 32).⁷⁷ In a third study, treatment comprised alendronate (40 mg: this dose is four times as high as that given for osteoporosis) administered to 40 CRPS I patients. In the three studies, the parameters in the group of patients treated with bisphosphonates improved significantly more than in the placebo group.

It has been suggested, based on this data, that bisphosphonates have a beneficial effect on the signs of inflammation in patients with CRPS I. At present, little is known about optimum dosage, frequency, and duration of treatment.

Calcium Channel Blockers

Calcium channel blockers slow the movement of calcium ions into cells, leading to inhibition of the excitation–contraction coupling. With diminished flow, smooth muscle relaxes, peripheral blood flow increases, and discharge from hypersensitive scars can decrease. Nifedipine (Procardia) 10 to 30 mg three times daily or diltiazem (Cardizem) 30 mg four times daily (can be increased to 60 to 90 mg four times daily) can be effective. Calcium channel blockers are also helpful in the treatment of the Raynaud phenomenon.

Two studies of moderate quality and size investigated the effect of nifedipine and phenoxybenzamine in treating CRPS I.48,49,57

One retrospective study with 59 patients reports that nifedipine (20 mg per day) or phenoxybenzamine (up to 120 mg/day) are most effective for CRPS I in the acute phase.⁵⁷ Both studies are primarily descriptive and the outcomes are subjective, failing to describe the nature of the improvement in patients’ conditions. There are indications that calcium channel blockers have some effect in the acute phase of CRPS I. Although they improve blood circulation, they also cause side effects, such as a drop in blood pressure and headache.

Anticonvulsants

Anticonvulsants suppress pathologic electrical discharges in the CNS and peripheral nervous system. They raise the threshold for nerve activation and block sodium channeling. Some also facilitate γ-aminobutyric acid (GABA, an inhibitory neurotransmitter) activity. Side effects include CNS depression, drowsiness, and edema. Gabapentin (Neurontin) is the most commonly used, but lamotrigine (Lamictal), oxcarbazepine (Trileptal), phenytoin (Dilantin), and carbamazepine (Tegretol) are also reported in the literature.

Gabapentin is incorporated in a carrier protein, and there is insignificant absorption (Physicians’ Desk Reference [PDR]). Interactions with other drugs are insignificant, and doses up to 3600 mg/day are well tolerated. Off-label uses of 4000 to 6000 mg/day are reported in the literature to be safe and effective.

Pregabalin (Lyrica), a new α2δ ligand (similar to gabapentin), seems to offer the benefit of more reliable dosing with fewer side effects. A calcium channel modulator, it is the latest treatment for peripheral neuropathic pain. The dose range is lower than for gabapentin: 150 to 600 mg/day divided into two or three doses. Like most drugs, it has side effects; the most commonly reported adverse reactions are dizziness and drowsiness.

Two placebo-controlled, randomized studies have been found that examined the use of gabapentin in neuropathic pain patients. The first study ⁶⁹ (n = 307) shows that gabapentin causes a modest but significant reduction in neuropathic pain symptoms 8 weeks after the start of treatment. It is unclear what this means for the CRPS I patients, who made up 28% of the sample population.

In the second study (n = 58), a moderate effect on pain was found, but no significant reductions in other sensory abnormalities were found.⁷⁵ Dizziness, sleepiness, and fatigue occurred significantly more often in patients taking gabapentin than in patients taking placebo.

There are indications that gabapentin administered at doses of 600 to 1800 mg every 24 hours in the first 8 weeks can cause some reduction in pain symptoms suffered by patients with CRPS I.

There is limited evidence that gabapentin reduces sensory abnormalities such as hyperaesthesia and allodynia. The long-term effects of gabapentin on patients with CRPS I are not known.

Calcitonin

The effects of calcitonin have been evaluated in two meta-analyses and two systematic reviews. The meta-analysis carried out by Kingery et al ⁴¹ reports conflicting findings as to the effects of calcitonin. In contrast, the meta-analysis carried out by Perez et al⁶² points to calcitonin having a positive effect on pain on average, and the review carried out by Forouzanfar et al²⁶ also describes positive results for calcium-regulating drugs (including calcitonin) administered to CRPS I patients. There is conflicting evidence with respect to the efficacy of calcitonin for treatment of CRPS I.

Topical Treatments

Pain cream is recommended for treatment of superficial pain. Capsaicin, a chili pepper derivative, which depletes substance P, must be applied topically with gloves. A dose of 0.075% cream should be applied to the affected area four times a day, using only enough cream to penetrate into the skin. Doses up to 7.5% are now being tried experimentally.

EMLA is a eutectic mixture of lidocaine and prilocaine in cream form and is approved by the FDA for neuropathic pain.

Lidoderm, a 5% lidocaine patch, is applied directly over the painful site. Systemic effects are unlikely, and the primary problem is application-site sensitivity. It is FDA approved for 12 hours on and 12 hours off. However, off-label use (increased time on) is common. Quality of life has been often shown to improve after 4 weeks of treatment (Fig. 14-16).

Figure 14-16 Lidoderm, a 5% lidocaine patch, is applied directly over the painful site. Systemic effects are unlikely. The primary problem is application-site sensitivity.

Local Anesthetics

One quasi-experimental study with seven patients evaluated the effects of lumbar and stellate blockades with lidocaine/bupivacaine compared with placebo (follow-up 2-2.5 weeks). No significant differences were found between active and placebo treatment with respect to initial peak pain reduction.⁶⁵

An uncontrolled open study investigated the long-term effects (mean follow-up, 32 months; range, 7-48 months) of epidural administration of bupivacaine to 14 patients with CRPS I in the knee.¹⁵ Treatment was continued with continuous administration of a narcotic. No pain control data were described; however, 11 patients were seen to have a complete improvement of CRPS I symptoms at the end of the follow-up.

New Treatments

Current literature suggests that some new treatments are just over the horizon. NMDA receptor–site blockers are being developed. Botulinum toxin might inhibit the release of glutamate. Pain and bone density can improve after intravenous administration of pamidronate, alendronate, or clodronate. Magnetic stimulation of the contralateral brain cortex can decrease pain.⁷¹ Inhalation of cannabinoids (marijuana), much touted in the news for their pain-relieving qualities, does not relieve neuropathic pain, and the side effects of somnolence, fatigue, hypotension, and memory disorder make them unlikely to improve CRPS symptoms.⁴

Free-Radical Scavengers

Dimethylsulphoxide (DMSO), a free-radical scavenger, has been tried in patients with CRPS I. A prospective crossover study ³⁰ with 20 patients found a positive effect of DMSO on the function of the affected limb. In addition, in 26 CRPS I patients, DMSO was found to be significantly more effective than the conventional regional ismelin block²⁷ in reducing pain. A randomized double-blind trial conducted with 32 CRPS I patients⁸⁶ also showed that five times daily use of topical DMSO cream provided significantly better results on CRPS I symptoms than placebo after 2 months of treatment. In general, DMSO generates lower (direct and indirect) costs than N-acetylcysteine (another available free-radical scavenger). Subgroup analysis indicates that N-acetylcysteine is more cost effective in patients with a cold form of CRPS I than DMSO. The opposite holds for warm forms of CRPS I.⁷⁶

Robert J. Schwartzman, MD, a neurologist at Drexel University College of Medicine, is experimenting with a more radical approach. Cooperating with a group of anesthesiologists in Germany, he is treating patients who have refractory pain by injecting them with ketamine at doses high enough to induce coma for 5 days. Fifty percent improvement has been reported.³²

The effects of a subanesthetic ketamine infusion (10 mg/hr up to 15-50 mg/hr) was assessed in a retrospective study of 33 patients with CRPS I or II.¹⁶ Twelve patients experienced a relapse and had a second course of infusions, 3 patients had a third course, by which pain disappeared completely in 83% of patients. The average duration of pain reduction (data of 20 patients) was 9.4 months. The side effects were intoxication, hallucinations, dizziness, nausea, light-headedness, and blurred vision.

Vitamin C

Recent investigations have suggested a role for vitamin C in the prevention of CRPS. In a double-blind, prospective, multicenter trial, 427 wrist fractures were randomly allocated to treatment with placebo or treatment with 200, 500, or 1500 mg of vitamin C daily for 50 days.⁸⁵

The effect of gender, age, fracture type, and cast-related complaints on the occurrence of complex regional pain syndrome was analyzed. The prevalence of complex regional pain syndrome was 2.4% in the vitamin C group and 10.1% in the placebo group (P = .002). Early cast-related complaints predicted the development of CRPS, and based on these results, the authors recommended a daily dose of vitamin C 500 mg for 50 days to prevent CRPS. Additional investigations by the same author have yielded similar results,⁸⁴ including those which suggest a benefit from vitamin C 2 days before surgery and for 50 days after surgery for trapeziometacarpal arthritis. Thirty four patients were administered 500 mg of vitamin C and followed for a minimum of 44 months. Contrary to another published report, which demonstrated a 13% CRPS rate with this procedure, there were no cases of CRPS in this study.⁸⁴

Pain Therapies

If patients fail to respond in 1 to 2 months, they should be referred to a multidisciplinary pain center (anesthesia, orthopaedics, psychiatry, psychopharmacology, neurosurgery, neurology, physical therapy). Hospitalization may be required.

Implantable spinal electrodes for dorsal column stimulation to control intractable pain should be considered once a plateau in improvement with sympathetic blocks has been reached or in refractory cases. Based on the gate theory of pain control, a percutaneous route for implantation with low morbidity is now well developed. Long-term results are good, and late failures are primarily based on technical problems. Permanent placement should be preceded by a trial with a temporary electrode.⁶⁰

There are numerous studies investigating these devices. Patients with chronic refractory CRPS I were randomly allocated to spinal cord stimulation (SCS) plus physiotherapy or physiotherapy alone. Trial stimulation proved successful in 24 of the 36 patients; only these patients underwent a procedure to implant a permanent SCS device. Pain intensity was reduced by 2.4 cm on a visual analogue scale after 6 months in the group receiving spinal cord stimulation plus physiotherapy compared with a group receiving only physiotherapy.³⁷ At 2 years follow-up, pain decrease in the SCS group was 2.1 cm more than pain decrease observed in the physiotherapy group.³⁹ Quality of life improved only in the patients with an implanted system; function remained unchanged. Two retrospective cohort studies have investigated effects of SCS on pain relief (n = 23-31).^11,38

All studies relate to carefully selected patients with refractory CRPS I. There is no scientific evidence for SCS being effective in acute CRPS I. Complications requiring further surgery do occur in 25% to 50% of patients.⁷⁴

Overall, it appears that spinal cord stimulation administered to CRPS I patients who are carefully selected and undergo successful trial stimulation experience long-term pain reduction and improvements in their quality of life but do not seem to see improved function.

Chemical or surgical lumbar sympathectomy works best in the foot and ankle. When sympathetic blocks are successful but the pain keeps returning, radiofrequency or cryoprobe lesioning could provide long-term relief. Surgical resection creates an identifiable interruption.

The efficacy of surgical sympathectomy was addressed in a systematic review,⁵⁰ based on analysis of retrospective cohort, including 7 to 73 individuals. All the studies report a clear reduction in pain resulting from sympathectomy, whereby the extent of pain relief declines over time. Long-term follow-up studies (>1 year) indicate that the chance of success is greatest if treatment is given within 3 months after the initial trauma.⁶⁸

Chemical destruction (alcohol, phenol) of a nerve, despite excellent justification, can also have drawbacks. Further obliteration of the nerve in the background of a disease state that already affects nerves can paradoxically lead to more pain. Radiofrequency ablation, by using generated local temperatures, can be more selective in just destroying pain fibers (C fibers) while leaving the nonpain portion of the nerve alone.

Recurrence of SMP can occur secondary to upregulation of distal receptors, producing increased sensitivity to catecholamines. In some cases, postsympathectomy neuralgia in the thigh can be distressing. Sympathectomy for multiple extremities leads to problems with thermal regulation and with bowel or bladder control.²⁹