Controversy and confusion surround nearly every aspect of complex regional pain syndromes (CRPS). Historically, this condition has been plagued by delayed or missed diagnosis and haphazard treatment protocols. Although the nomenclature currently recommended is new, the constellation of signs and symptoms was first recognized in American Civil War soldiers in 1864. The medical literature regarding this syndrome is difficult to interpret because of the vague nature of the condition and the inconsistent use of terminology ( Box 23-1 ).
Chronic traumatic edema
Sympathalgia (postsympathectomy pain)
Posttraumatic pain syndrome
Reflex neurovascular dystrophy
Peripheral acute trophoneurosis
Major and minor causalgia
Major and minor dystrophy
Sympathetically maintained pain
Reflex sympathetic dystrophy
Pain dysfunction syndrome
CRPS is a debilitating condition with a reported incidence ranging from 5.5 per 100,000 patients in an Olmsted County, USA community up to 26.2 per 100,000 in a national database study out of the Netherlands. This condition is precipitated by a fracture (44%), sprain (18%), elective surgery (12%), or tendon injury (6%) in more than 75% of the cases, with increased incidence in postsurgical orthopaedic patients. A recent report also noted an incidence of 8.3% after carpal tunnel release and did not correlate with the type of anesthesia used. Orthopaedic surgeons are on the front lines of diagnosing this condition and ensuring that patients begin early treatment with a coordinated team of health care professionals.
The treatment outcomes of this condition are disheartening. Sandroni et al. report that symptoms are resolved in 75% of affected persons after 1 year, but data from the Netherlands suggest that only 36% have no remaining symptoms according to the International Association for the Study of Pain (IASP) criteria. The subjects studied by Sandroni et al. had a job impairment rate of 11%, whereas data from the Netherlands showed that 31% were unable to work at all and an additional 28% required adaptations. The severity of the symptoms this condition can produce is highlighted by a review of amputation to treat persons with therapy-resistant CRPS. The authors reported a recurrence of CRPS type 1 in nearly half the patients, and only around one quarter made use of their prosthesis.
CRPS is most succinctly described as a chronic pain syndrome characterized by severe, diffuse, nondermatomal pain associated with painful responses to nonpainful stimuli (allodynia), accompanied by autonomic and trophic changes, and usually occurring after tissue injury. The pain is associated with changes in skin color, temperature changes, sudomotor dysfunction (sweating), edema, and reduced range of motion. The pain is sharp and burning and often out of proportion to the inciting event.
When treating CRPS, the clinician must understand the complex nature of this syndrome. In 1988 Amadio wrote about pain dysfunction syndrome, hypothesizing that these syndromes, now known as CRPS, are generally thought to have three primary contributing components and one secondary component. The primary components include a trigger, a personality disorder, and systemic factors that exacerbate pain. First, a local trigger commonly begins the process and should be identified and eliminated. Examples of common local triggers are painful organic conditions such as patellofemoral pain, fracture, and nerve injury. Second, psychological factors such as secondary gain issues, substance abuse, and psychiatric conditions and personality disorders are commonly present and should be identified. Psychiatric conditions such as somatization disorders, malingering, factitious injury, and conversion reactions are thought to contribute to the syndrome, but this link has been debated in the literature. Amadio concedes that although psychiatric consultation is helpful in such cases, it is often unsuccessful at resolving the issues. Third, systemic factors are thought to cause or exacerbate pain. These factors could include diabetic peripheral neuropathy, lupus erythematosus, polymyalgia rheumatica, giant cell arteritis, multiple sclerosis, ischemic heart disease, Pancoast tumors, and others.
The secondary component of this syndrome, and the hallmark of its clinical features and classification, is sympathetic dysfunction—the fourth factor of this syndrome. Sympathetic dysfunction can manifest as vasomotor, sudomotor, edematous, motor, and trophic responses. The signs and symptoms produced by this dysfunction include all of the former diagnoses of reflex sympathetic dystrophy, causalgia, traumatic dystrophies, shoulder-hand syndrome, and Sudeck atrophy. Pain is the hallmark of CRPS, and part of the original diagnostic criteria—an entity termed “Complex Regional Pain Syndrome”—has been reported that includes the sympathetic dysfunction without the associated allodynia and hyperalgesia.
All of the components can have an effect on the painful response experienced by the patient. The first three components do not involve the sympathetic nervous system and therefore are named “sympathetically independent pain” (SIP). “Sympathetically maintained pain” (SMP) is defined as pain that is sustained by sympathetic innervation or by circulating catecholamines, which stimulate the sympathetic system. Merskey and Bogduk reported that SMP is associated with CRPS at some point in the course of the disease and is defined as the component of pain that is relieved by sympathetic blockade. Any residual pain after true sympathetic blockade is termed SIP. In addition, variable amounts of SMP and SIP contribute to both types of CRPS ( Fig. 23-1 ). Differentiation of these two components of pain is clinically useful because it can influence treatment. However, the presence of autonomic dysfunction in persons with CRPS does not guarantee that all patients will respond to sympathetic blocks. SMP was introduced to explain the favorable response some patients have to sympathetic blockade. Conversely, the term SIP was introduced to explain the lack of response some patients have to sympathetic blockade. Whereas patients can display SMP or SIP characteristics, they often have contributions of both, explaining the varying response to sympathetic blockade.
Theories of the Pathophysiologic Mechanism
Our understanding of the pathophysiology of CRPS continues to evolve. Although sympathetic dysfunction may play a larger role in the acute phase and sympathetic blocks have a role in limiting pain in persons with predominant SMP, the previously discussed theory that CRPS is a largely sympathetically mediated disorder has transitioned to a belief that CRPS is induced by a combination of inflammatory processes, vasomotor dysfunction, central sensitization, and cortical reorganization.
Recent clinical data from the Netherlands that incorporated the IASP diagnostic criteria has improved our understanding and provided valuable insights into the pathophysiology of CRPS. Information collected regarding a large number of patients over an 11-year span has provided new insights. These data suggest that the incidence has been underestimated in the past and is perhaps in the vicinity of 26 per 100,000 patient years. Female patients were three times more likely to be involved, the mean age was 52.7 years, and peak incidence was reported between the ages of 61 and 70 years. Upper extremity involvement was 59%, and fracture was the most common precipitating event (44%). In an analysis of comorbid conditions, this group found an association of asthma, migraines, osteoporosis, preexisting neuropathies, and a recent history (within 1 year) of menstrual cycle abnormalities. No link was suggested to psychological disorders. An interesting association with the use of angiotensin-converting enzyme (ACE) inhibitors was found; this link was more prominent in higher doses and with longer duration of use. This relationship between CRPS and use of ACE inhibitors, combined with the fact that ACE inhibitors degrade substance P and bradykinin, strengthens the theory of neurogenic inflammation in persons with CRPS (see Appendix 23-A at expertconsult.com for an expanded explanation).
Diagnostic Criteria and Clinical Presentation
CRPS is primarily a clinical diagnosis and historically has proven to be as complex as its pathophysiology. In an attempt to standardize the definition, an international symposium in 1994 proposed IASP criteria for CRPS types 1 and 2. Since then, the term CRPS has become standard nomenclature to describe syndromes formally known as reflex sympathetic dystrophy and causalgia. It has also been accepted that two subtypes exist: CRPS type 1, without a defined neurologic injury, and CRPS type 2, following a neurologic injury. However, some authors believe that many CRPS type 1 cases are in actuality CRPS type 2 cases with unrecognized nerve injury. Validation studies have shown that the diagnostic criteria are sensitive (1.00) but lack specificity (0.41). Thus the clinical use of these criteria is debatable, as reflected in a recent review of 26 studies of amputations associated with CRPS type 1, where recommended diagnostic criteria were applied in only four studies.
At a 2003 international consensus meeting, the “Budapest Criteria” were created to address these limitations, after which validation studies have suggested retained sensitivity (0.99) with improved specificity (0.68). Continued collaborative efforts have provided validation of criteria, more specific criteria for research, and a CRPS severity score, which provides both clinicians and researchers with the ability to track symptom severity and response to treatment. Standardizing nomenclature will prove invaluable for the development of successful treatment algorithms in the future.
Updated criteria must be fulfilled to make an accurate diagnosis of CRPS ( Box 23-2 ). The sole differentiating factor between the CRPS types 1 and 2 is the presence of a known nerve injury in type 2, whereas type 1 is generally attributed to a noxious event (other than nerve injury). Thus the type 1 designation is assigned when no known peripheral nerve injury exists. Electromyography/nerve conduction velocity studies may help the clinician discover occult peripheral nerve injury and thereby differentiate between type 1 and type 2 (negative electromyographic/nerve conduction velocity tests = CRPS type 1, and positive results = CRPS type 2).
Crps Type 1 (Reflex Sympathetic Dystrophy)
An initiating noxious event is present.
Spontaneous pain or allodynia/hyperalgesia occurs beyond the territory of a single peripheral nerve and is disproportionate to the inciting event.
There is or has been evidence at some time of edema, skin blood flow abnormality, or abnormal sudomotor activity in the region of the pain since the inciting event.
This diagnosis is excluded by the existence of conditions that would otherwise account for the degree of pain and dysfunction.
Crps Type 2 (Causalgia)
This syndrome follows nerve injury. It is similar in all other respects to type 1.
A more regionally confined presentation about a joint (e.g., ankle, knee, or wrist) or area (e.g., the face or eye) is associated with a noxious event.
Spontaneous pain or allodynia/hyperalgesia is usually limited to the area involved but may spread variably distal or proximal to the area, not in the territory of a dermatomal or peripheral nerve distribution.
Intermittent and variable edema, skin blood flow changes, temperature change, abnormal sudomotor activity, and motor dysfunction disproportionate to the inciting event are present about the area involved.
CRPS, Complex regional pain syndrome.
Signs and symptoms from the Budapest criteria should be evaluated thoroughly. Criteria include both signs and symptoms induced by changes in sensory, vasomotor, sudomotor/edema, and motor/trophic function commonly observed in persons with CRPS. Noting each sign and symptom will allow the clinician to establish a diagnosis and calculate a severity score. A recent study of 692 patients who met the IASP criteria were evaluated for the distribution of these signs and symptoms and how symptoms changed, based on duration. These data are noted in a more detailed discussion of each sign and symptom later in this chapter.
After diagnosis of CRPS, clinicians should establish the state (hot/edematous vs. cold/atrophic). Although it has been shown that these states are not linear or progressive, the information is helpful in establishing a treatment plan based on whether neural or mechanical nociceptive foci are potentiating CRPS. This vital information should be sought and eliminated to limit long-term morbidity. Finally, it should be established whether the process is sympathetically maintained or independent.
The clinical presentation of CRPS is highly variable and difficult to characterize. Classic findings are described but are not necessarily the most common clinical findings. Delayed or failed diagnosis is common, partly related to the syndrome’s wide variability and nonspecific signs and symptoms. According to the IASP, the diagnosis of CRPS requires four elements:
An initiating noxious event or course of immobilization
Continuing pain, allodynia, or hyperalgesia disproportionate to any inciting event
Evidence, at some point, of either edema, changes in skin blood flow, or abnormal sudomotor activity
The exclusion of medical conditions that would otherwise account for the degree of pain or dysfunction
Perhaps the hallmark sensory finding of CRPS is pain that is disproportionate to the expected response for the inciting condition. The disproportionality applies to the duration of pain response, the severity of response, and the distribution of the painful area. Although pain is a perceptual event normally associated with cellular injury or death, CRPS is not associated with ongoing cellular damage in the traditional sense. In fact, persistent pain in the absence of a cellular insult is a hallmark of CRPS. Segmental ischemia as a result of arteriole-venous shunting in the cutaneous circulation may cause cell death and be partly responsible for the development of arthrofibrosis, osteopenia, abnormal function of neuroreceptors, and central pain imprinting. Within hours or days after the initial injury, the pain becomes more diffuse and unrelated to the site of the injury. As the syndrome progresses, the painful area expands in a nondermatomal distribution, instead following thermatomes, with a preponderance of distal limb involvement (i.e., stocking or glove distribution).
The character and chronology of pain provide important diagnostic clues. Patients commonly describe a burning, shooting, or deep, constant aching. de Boer et al. reported that spontaneous pain is present in 92% of patients with CRPS and that 95% report worsening pain after exercise. Patients also commonly report a classic finding in CRPS—allodynia, characterized by pain with light pressure. Frequently they are unable to tolerate the faint touch of bed sheets, clothing, or air currents. Hyperalgesia, hyperpathia, hyperesthesia, and dysesthesia also may be present, along with occasional night pain. The pain of CRPS may fluctuate and recur after overaggressive physical therapy (active and passive motion), with environmental and local temperature changes (particularly cold intolerance), in the dependent limb position, and as a sequel to emotional excitement. In addition, up to 50% of patients with chronic CRPS type 1 experience hypoesthesia and hypoalgesia of the entire half of the body or in the upper quadrant ipsilateral to the affected extremity.
A final painful manifestation of CRPS is cold intolerance, a finding that is nonspecific yet particularly sensitive in the diagnosis of SMP. Cold intolerance may first manifest as a significant painful reaction following cryotherapy used by the physical therapist to control swelling. We recommend use of the “ice test,” in which ice is applied to the involved area and the patient is questioned about characteristics of the sensation. Patients with CRPS have an intolerable “burning sensation” of the involved area, whereas application of ice to the normal limb is typically described as a “cold sensation.” Additionally, cold weather commonly precipitates recurrent flares of sympathetic dysfunction in persons with CRPS, and patients often report the need to wear socks to bed because of cold feet.
Vasomotor dysfunction is commonly manifested by skin color and temperature abnormalities. The classic skin pattern changes from acutely red, warm, and dry skin to the chronic appearance of bluish or mottled, cold, and moist skin ( Fig. 23-2 ). Approximately 80% of patients will have side-to-side differences in limb temperature averaging 3.5°C. Historical data have shown that temperature and color changes present in 91% of patients, whereas de Boer indicates that these changes occur in only 55% of patients and less frequently with prolonged duration.
The edema noted in persons with CRPS is generally painful and usually extraarticular, although joint effusions have been reported. Swelling is most pronounced in the acute phase of CRPS and is usually less pronounced with chronicity, a finding confirmed by de Boer et al. The soft, puffy edema observed in the acute phase is eventually replaced with tight, shiny skin that lacks normal creases. Sudomotor changes, which manifest as changes in sweating of the palmar surfaces of hands and plantar surfaces of the feet, occur in 20% of persons.
The natural history of sympathetic dysfunction has classically been divided into three sequential yet overlapping stages.
The acute stage is characterized by sympathetic hyperfunction manifested by disproportionate pain; red, warm, and dry skin; and extraarticular swelling. The classic presentation of the acute stage lasts less than 6 months.
The dystrophic stage begins after the increased sympathetic output succumbs to a period of reduced sympathetic activity, typically 3 to 9 months after onset of symptoms. This stage is characterized by cyanotic or mottled, cold, moist skin; muscle wasting; thick nails and coarse hair; and other early trophic changes.
The atrophic stage commences after a period of chronic sympathetic dysfunction. This stage is characterized by thin, tight, glossy skin, osteoporosis, and joint contractures.
The clinical usefulness of this classic staging system has been questioned because of profound variation of symptoms between patients, as well as variable chronology of an individual patient’s disease. The staging system is most useful when the clinician understands that the disease does not progress systematically through stages. Instead, the common scenario is frequent, random toggling back and forth between stages. The staging system further improves our understanding of the classic presentation as long as the clinician recognizes that exceptions not only exist but are common. For example, many patients with CRPS may present with disproportionate pain as the only symptom. The astute clinician will correctly consider the diagnosis of CRPS even in the absence of other signs of sympathetic dysfunction.
One specific exception to the classic presentation is termed “cold reflex sympathetic dystrophy,” which may have implications for prognosis. van der Laan and colleagues found that patients with CRPS type 1 who had a cold skin temperature initially were much more likely to experience severe complications, such as infection, ulceration, chronic edema, dystonia, and myoclonus.
Motor impairment is a more recently described phenomenon associated with CRPS and can present with either decreased motion at involved joints (74%), weakness (55%), dystonia (20%), or tremor (12%). In the absence of traumatic nerve injury (CRPS type 1), findings of electromyography/nerve conduction velocity studies are usually normal, suggesting that motor abnormalities are centrally mediated, presumably at the spinal cord level. Weakness of the affected limb is caused by disuse and muscle wasting, and the observation of pseudoparalysis in patients with CRPS suggests that some motor abnormalities may be psychogenic or induced by pain. Paresis and limb neglect are also observed. Motor abnormalities characterized by hyperfunction include action tremor, myoclonus, hyperreflexia, muscle spasm, and voluntary guarding. Dystonia, apraxia, and lack of coordination are findings typically associated with long-standing disease, and chronic joint stiffness can ultimately lead to fixed contractures.
Persistent or recurrent CRPS symptoms may result in trophic changes. Atrophy of the tissues probably occurs from disuse and as a direct result of sympathetic dysfunction. Tissues known to undergo atrophy in response to chronic CRPS include skin, subcutaneous fat, muscle, tendon, and bone. Thin, shiny, smooth, tight skin is the typical end result of chronic CRPS. Atrophy of subcutaneous fat is particularly evident in the digits and results in a “pencil-like” appearance. Tendon atrophy is also known to occur. The classic finding of patchy, periarticular osteoporosis is known as Sudeck atrophy, but it is relatively uncommon and nonspecific. Chronically increased blood flow to the limb leads to local hypertrichosis, coarsening of the hair, and thickening of the nails. As reported by de Boer, changes in the hair, nail, and skin occurred less frequently (approximately 25%) in more recent studies compared with historical data (40% to 60%). However, these changes increase with prolonged duration of the symptoms.
As previously discussed, CRPS is primarily a clinical diagnosis. No definite blood tests facilitate diagnosis, and objective methods used to aid in the diagnosis of CRPS are aimed primarily at assessing the degree of sympathetic dysfunction. Because these methods are fraught with high levels of inaccuracy and poor sensitivity or specificity, the diagnosis of CRPS is made primarily on clinical grounds using recent diagnostic criteria, previously detailed, followed by secondary confirmation with objective tests. Establishing and confirming the presence of sympathetic dysfunction (e.g., pain, vasomotor, and sudomotor abnormalities) is especially useful in therapeutic planning. For example, SMP is diagnosed in patients who experience improvement during or after treatment with medications that modify the sympathetic nervous system. Classically, SMP must respond to an epidural, intrathecal, lumbar plexus block, or peripheral nerve block. Rapid response to such blocks is pathognomonic for SMP. Regardless, no single pathognomonic, “gold-standard” test exists for CRPS. The diagnostic tests described in the next sections attempt to relate objective, clinical findings to facilitate the diagnosis of CRPS.
Blood studies such as white blood cell count, erythrocyte sedimentation rate, and C-reactive peptide are all indicative of cellular inflammatory processes yet are historically negative in persons with CRPS. This disease is thought to be induced by a neurogenic inflammatory response, and thus white blood cell count, erythrocyte sedimentation rate, C-reactive peptide, and other studies, including rheumatoid factor and antinuclear antibody titers, are more useful in ruling out other inflammatory conditions.
Objective assessment of virtually any extremity pain begins with plain radiographs. Standard radiographic series of the affected area typically reveal soft tissue swelling and osteoporosis. Sudeck atrophy is classically described as patchy periarticular osteoporosis of the long bones, whereas the osteoporosis seen in the small bones of the hand and foot is more diffuse. Subchondral bone margins are usually retained. These radiographic findings are nonspecific, may be subtle, and generally indicate chronicity of disease. Comparison films of the contralateral extremity are recommended. These findings can be seen as early as 2 weeks after symptoms begin and can be used in follow-up evaluation to monitor progression of the disease. However, findings of radiographs will be normal in 30% of patients.
The use of a triple-phase bone scan is highly controversial. The scan consists of images obtained seconds (arterial phase), minutes (soft tissue phase), and hours (bone phase) after the intravenous injection of a radionuclide tracer. Increased periarticular uptake of tracer in the soft tissue and bone phase images are the classic bone scan findings in the acute stage of CRPS as a result of both increased blood flow and bone turnover in early stages of CRPS. The acute, increased tracer uptake is replaced by reduced tracer in the dystrophic and atrophic stages of CRPS. Whereas the sensitivity and specificity of the triple-phase bone scan have been reported to be as high as 96% and 97%, respectively, these numbers drop significantly 6 months after the onset of disease. Conversely, a small percentage of patients who have bone scans early in the course of disease show an abnormally decreased tracer uptake. This finding may account for the wide variability in the literature with regard to reported sensitivity rates (60% to 100%) and specificity rates (80% to 98%). (See Appendix 23-A at expertconsult.com for an expanded explanation.)
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) is generally not used to confirm the diagnosis of CRPS. Instead, MRI is used either to rule out other pathologic processes, narrow the differential diagnoses, or determine the inciting event that either triggered CRPS or continues to function as a “trigger point.” Several nonspecific findings of CRPS have been described on MRI. Characteristic findings in the acute phase include thickening of the skin, soft tissue swelling that enhances with administration of contrast, and intraarticular effusions. Over time these findings abate, and chronic changes include muscle atrophy and thinning of skin. Bone marrow edema, once thought to be a characteristic feature, is usually not seen as part of CRPS.
Paravertebral Sympathetic Ganglion Blockade
The diagnosis of SMP, as a component of CRPS, is made by the degree of the patient’s response to sympatholytic procedures. Several methods of achieving diagnostic sympathetic blockade are available. The most effective method and the “gold standard” in diagnosis of SMP is paravertebral sympathetic ganglion block. Under fluoroscopic control, the lumbar paravertebral sympathetic ganglion (for lower extremity CRPS) or the stellate ganglion (for upper extremity CRPS) is injected with a local anesthetic. The accuracy of the injection is assessed by evaluating vasomotor and sudomotor functions (e.g., skin color, skin temperature, and sweating) in the affected limb. Thus skin temperature changes must be monitored and recorded. An adequate block is accomplished when local skin temperature approaches core body temperature. In addition, the induction of Horner syndrome (i.e., ipsilateral ptosis, miosis, and anhidrosis) is required to confirm an adequate stellate ganglion blockade. Inadvertent spread of local anesthetic to sensory nerve roots is screened by a carefully documented sensory examination. Partial or complete pain relief with intact sensation indicates the presence of SMP.
Several limitations of this technique must be recognized by clinicians. First, the technique is dependent on the accuracy of the injection’s location, and the result can be influenced by the selected anesthetic agent. Furthermore, systemic uptake of the anesthetic may also adversely bias results. Lastly, placebo responses may be as high as 33% because of the high expectations of the patient and physician. Despite these limitations, the paravertebral block remains the diagnostic procedure of choice to objectively quantify the degree of SMP. Pain that remains after successful paravertebral sympathetic block is by definition SIP, and in cases of chronic CRPS, the majority of the patient’s pain may be SIP. Therefore the diagnosis of CRPS cannot be excluded solely on the basis of the paravertebral block.
Differential Spinal and Epidural Blockade
Physicians not skilled in the technique of paravertebral block may substitute a more common interventional method for the diagnosis of SMP—that is, spinal or epidural blockade. In this common technique, a spinal or epidural puncture is performed and variable concentrations of an anesthetic agent are injected over time. Initially, a saline placebo may be injected. The patient is subsequently queried about symptoms, which are recorded. Increasing concentrations of the anesthetic agent are then injected, and after each injection, the patient’s symptoms are recorded. Low concentrations should penetrate the least myelinated sympathetic nerve fibers and relieve SMP. A moderate concentration of anesthetic agent will additionally block moderately myelinated sensory fibers, and a high concentration adds motor blockade. The limitations of this method result from a variable amount of sensory blockade even at low anesthetic concentrations. As a result, attributing the pain relief solely to sympathetic blockade may be inaccurate. Thus the more sympathetic-specific, paravertebral ganglion block is the preferred technique.
Regional Intravenous Sympathetic Blockade
Regional intravenous sympathetic blockade is another means of confirming clinical diagnosis, and the blocking effect is sometimes used for treatment as well. Despite not being available in the United States, guanethidine was frequently used for this technique; however, reserpine and bretylium blocks have also been described. The common denominator of these medications is their mechanism of action. The drugs are absorbed into sympathetic terminals where they deplete norepinephrine for up to 2 days by stimulating its release and inhibiting its reuptake. This mechanism explains the initial burning pain on injection experienced by many patients, indicating the presence of SMP. Significant pain relief lasting 2 weeks to 6 months may occur and constitutes a positive test result. Little or no pain relief for less than 5 days favors the diagnosis of SIP. Central ephapses (in the dorsal root ganglion) as a source of SMP cannot be excluded on the basis of tourniquet-controlled regional alpha blockade. Intravenous administration of guanethidine is not currently approved for this use in the United States, but it has been used extensively in other countries for more than 25 years.
Conflicting studies exist in the literature, and the usefulness of regional intravenous sympathetic blockade is hotly debated. We currently do not recommend use of intravenous regional sympathetic blockade for numerous reasons. First, the specificity of these medications for sympathetic functions is undetermined. Second, local anesthetics are commonly injected in conjunction with these medications and may confuse results. Third, injection of these medications is often painful for patients and often poorly tolerated. Fourth, the use of a tourniquet on a limb with profound sympathetic dysfunction is generally unfavorable. Paravertebral ganglion injections are our preferred method of achieving prolonged sympathetic blockade and have the added advantage of blocking central ephapses for a more complete sympatholytic effect.
Treatment: Principles and Methods
Many treatments have been proposed, and yet few are supported by solid data. A conspicuous deficiency of blinded, randomized, placebo-controlled clinical trials evaluating the efficacy of specific treatments exists in the CRPS literature. Many of the current recommendations are based on the art more than on the science of medicine. The relative benefit of oral medications compared with widely used treatments of physical therapy, nerve blocks, sympathectomy, intraspinally administered drugs, and neuromodulatory therapies remains uncertain. Significant placebo effects have been observed in up to 33% of patients with CRPS, emphasizing the need for placebo control. Uncontrolled trials must be interpreted with extreme caution for this reason. Furthermore, because the diagnostic criteria of CRPS have only recently become standardized, comparison between studies is virtually impossible. Therefore current treatment of persons with CRPS is largely empirical.
In spite of these limitations, several general statements about treatment principles and specific treatment methods can be made with confidence. Perhaps the cardinal principles are early recognition and expeditious treatment. Ideally, treatment should begin within 2 to 3 weeks of the onset of symptoms. Several authors have emphasized that long-term prognosis is critically dependent on the amount of time from disease onset to the beginning of treatment. Delay of treatment beyond 6 months from onset is associated with a poor long-term prognosis. More than 80% of patients diagnosed and treated within 1 year of the onset of CRPS experience significant improvement of their symptoms. However, if treatment is initiated after 1 year, permanent functional impairment occurs in 50% of patients.
The diagnosis of CRPS mandates a full and thorough investigation in search of a persistent, painful focus. Any painful focus will aggravate and perpetuate SMP. Elimination of the painful focus is a critical step in the treatment of CRPS. Although correction of a painful focus alone will not always eliminate CRPS, SMP generally will not resolve without elimination of the painful lesion. During sympathetic blockade, encourage the patient to think about the character and location of residual pain while the block is working. When the burning, nonspecific pain is abolished, an underlying pain may be unmasked. The patient will often describe a specific location and may point to a location, such as the medial joint line rather than the entire medial knee.
Additionally, the clinician should remember that MRI and other diagnostic tests may miss either small or large lesions. Correctable lesions such as meniscal tears, chondral lesions, infections, painful neuromas, and many others that were not previously revealed by MRI are often discovered by arthroscopy. Surgery generally is not recommended for patients with active CRPS because of the risk of worsening pain. However, in the senior author’s experience, patients with chronic refractory CRPS whose complaints suggest a painful nonneurologic condition frequently require arthroscopy to correct the painful focus that exacerbates CRPS. Eliminating the focus of CRPS can eradicate the condition.
Treatment often requires a multidisciplinary team approach. Patients often must seek the cooperative advice of an orthopaedic surgeon, an anesthesiologist or pain management specialist, a physical or occupational therapist, a psychologist or psychiatrist, a neurologist or physiatrist, and a primary care physician. An open line of communication between health care providers is essential to provide the most effective and efficient care. The primary care physician, orthopaedic surgeon, or pain management specialist will often serve as the “team manager” to help coordinate the treatment program. Nevertheless, for patients with recurrent exacerbations or flares of CRPS, the patient’s direct access to the pain management specialist improves efficiency and reduces the patient’s frustration.
It is imperative to consistently remain the patient’s advocate. Personality conflicts between patients and health care providers must be disregarded, and confrontations are almost always counterproductive. Patients must believe that their physicians, nurses, and therapists are on their side to lend credibility to the prescribed treatment programs. Therefore the development of rapport and establishment of a therapeutic alliance between physicians and patients are pivotal to successful outcomes. A sound physician-patient relationship can also help eliminate the “doctor shopping” tendency displayed by many patients.
Because specific treatment protocols have not been validated, good clinical judgment must be used to individualize treatment on the basis of the patient’s signs and symptoms. Therapeutic decisions must be made individually on the basis of all of the available medical information. The myriad of available therapeutic options reported in the literature are a testimony to the fact that only a few choices are supported by conclusive, scientific evidence. Treatment techniques include physical therapy, medications, psychotherapy, surgical sympathectomy, neuromodulation, surgical intervention, and patient education ( Fig. 23-3 ).
Several medical, interventional, and surgical treatment options are intended to improve patients’ symptoms with the purpose of facilitating an active physical therapy program. Physical therapy, with or without occupational therapy, is generally regarded as the mainstay of CRPS treatment. Occasionally, physical therapy alone can reverse CRPS. Controlled trials demonstrating the efficacy of physical therapy are scant, yet several guiding principles must be adhered to while designing therapy programs for patients with CRPS. These principles are based on science, experience, and the recommendations of consensus panels.
To enhance outcomes, confrontations between aggressive therapists and resistant patients should be avoided, and physical therapy programs should never induce intolerable pain. The program must work within the limits of a patient’s pain tolerance, supplemented by medications and blocks. The goals of any therapy program in the treatment of CRPS are prevention and treatment of pain, swelling, stiffness, weakness, and disuse with the end goal of restoration of function. A four-step program emphasizing gradual progressive return to function has been outlined.
The first step involves the development of a patient-therapist treatment alliance. By remaining a strong advocate for the patient in all matters, the therapist will gain the confidence and trust of the patient, lending credibility to the treatment program.
The second step involves motivation of the patient and desensitization and mobilization of the affected limb. Desensitization is accomplished via high- and low-frequency vibration, gentle textured massage, contrast baths using temperatures within the pain-free range, and transcutaneous electrical nerve stimulation. Movement phobias must be overcome. Motion exercises in persons with CRPS are best accomplished through active and active-assisted exercises. Active motion, elevation, massage, and compression, if tolerated, will help reduce edema. Passive motion absolves the patient of “limb control” and is often poorly tolerated. Passive stretching exercises are thought to increase sympathetic output and can thereby aggravate pain, swelling, and sympathetic dysfunction. Techniques of immobilization (splints and casts) are generally counterproductive and are avoided when possible.
The third step involves strengthening and stress loading. Isometric muscle strengthening is recommended to minimize unnecessary motion. The patient is gradually progressed to isotonic strengthening. Stress loading is encouraged to prevent disuse and restore functionality. Upper and lower extremity load-bearing activities are implemented gradually through the use of weight-carrying or water exercises. Postural and balance training of bilateral extremities are accomplished with the use of biomechanical ankle platform system boards. General aerobic conditioning helps maintain range of motion, strength, and balance.
The fourth step emphasizes return to normal function. Patients are enrolled in vocational rehabilitation with work hardening programs. Eventually, a functional capacity evaluation is performed and job modifications are implemented as needed. Patients are encouraged to return to work, school, sports, or other daily activities. This program is supplemented by sympathetic blocks and medications whenever appropriate to allow the patient to make steady progress without intolerable pain.
A contemporary physical therapy modality, “pain exposure” physical therapy, has gained popularity in the treatment of CRPS type 1 in the past few years. Pain exposure physical therapy has shown some promise in the treatment of patients with long-standing CRPS type 1 and consists of a progressive-loading exercise program and management of pain-avoidance behavior without the use of CRPS-specific medications or analgesics. It is based on the premise that a progressive loading exercise program can reduce peripheral and central sensitization and restore local autonomic deregulation and cortical representation in persons with CRPS type 1 . Emphasis is placed on cognitive-behavioral aspects to motivate the patient to increase the activity and utilization of the affected limb, set clear functional goals, and increase self-confidence and independence. A recent study of 20 patients indicated significant improvement in CRPS type 1 signs and symptoms, including pain intensity and muscle strength and improvement in functional metrics (i.e., visual analog scale [VAS]), disability index, and walking speed).
Recently some favorable outcomes have been reported in the management of CRPS in the pediatric population with use of intensive inpatient rehabilitation programs. These programs focus on intensive inpatient physical and occupational therapy combined with psychological counseling, recreational and art therapy, conditioning, and stress management. Significant improvement in pain and functional restoration was reported at a median of 2 months from the start of treatment, and maintenance or improvement in gains was observed at long-term follow-up.
Medications for Symptomatic Relief
Many different medications are used to treat persons with CRPS. However, their use is based on anecdotal data and only a few medications have been validated by scientific evaluation. The perfect combination of medications for each patient is based on a trial-and-error approach. The correct dosages must be titrated to allow maximal benefit with minimal adverse effects ( Table 23-1 ). Medications that relieve symptoms are intended to increase the patient’s comfort and facilitate the implementation of a physical therapy program. Medications discussed in this chapter are thought to provide some benefit to patients with CRPS, although their use must be individualized. Randomized controlled trials (RCTs) for medical treatment of CRPS have been performed for bisphosphonates, adrenergic active drugs, and steroids, and treatment aimed at alpha-adrenergic receptors and the sympathetic nervous system have not been proven effective in RCTs. Other medications (e.g., gabapentin, tricyclic antidepressants, and opioids) have proven effective in RCTs for neuropathic pain induced by conditions other than CRPS. The evidence to date is in favor of use of these medications for CRPS, even though RCTs have not been performed for this specific condition.
|Drug||Mechanism of Action||Use||Adverse Effects|
100-300 mg qhs or tid
|Blocks peripheral alpha receptors; decreases sympathetic tone||Causalgias||Hypotension|
|Prazosin (Minipress) |
1 mg bid to tid
Decreased sexual function
|Terazosin (Hytrin)||May have fewer side effects|
|Clonidine (Catapres)||α 2 agonist; blocks α 1 transmission peripherally and centrally||Pain relief can be obtained with clonidine patches||Hypotension|
|Beta Blockade *|
|Propranolol (Inderal) |
Up to 320 mg/day
|β-adrenergic response resynaptically (centrally)||Hypotension |
Aggravation of asthma and cardiac arrhythmias
|Atenolol (Tenormin)||Sensitivity to catecholamines |
Resultant cardiac arrhythmias and myocardial infarction
|Trazodone (Desyrel) |
50 mg tid
|Blocks serotonin receptors; activates descending pain-inhibitory fibers||Most chronic pain syndromes||Multiple|
|Desipramine (Norpramin) |
|Myocardial ischemia |
|Doxepin (Sinequan) |
10-25 mg tid
|Amitriptyline (Elavil) |
25 mg tid
|Calcium Channel Blockers|
|Nifedipine (Procardia) |
10-30 mg tid
|Relaxes smooth muscle; increases peripheral blood flow; decreases discharges from ephaptic scars||Peripheral nerve injury, especially in decreased blood flow states||Headaches |
|Diltiazem (Cardizem) |
30 mg qid
Increase to 60-90 mg qid
|Carbamazepine (Tegretol) |
100 mg po bid up to 800 mg/day
|Suppresses pathologic electrical discharges in CNS and PNS||Ephapses burning pain; sharp discharges||CNS depression |
Many other adverse effects
|Clonazepam (Klonopin) |
Tricycline 0.5 mg bid to tid
|Valproic Acid (Depakene)|
|Salmon calcitonin (Miacalcin) |
1 puff daily
|Inhibits bone resorption; direction of action on SMP unknown||SMP||Nasal irritation|
|Bisphosphonates (alendronate IV)|
|Vitamin C |
No consensus on dosage recommendation 500-1500 mg/day
|Scavenge hydroxyl and superoxide radicals produced in free-radical reactions that are associated with tissue damage; protect the vascular endothelium and stop the progression of vascular permeability after burns, limiting fluid and protein leakage and sequestering circulating cells in the capillaries||Prophylactic modality||Increased risk for hemolysis in persons deficient in glucose-6-phosphatase dehydrogenase |
Higher risk for kidney stone formation
Dosage is dependent on body weight
|NMDA antagonist||Effective in treating allodynia in patients with postherpetic neuralgia, chronic posttraumatic pain and chronic neuropathic pain||Sedation |
|Narcotics and Benzodiazepines|
|Exogenous sources of narcotics and benzodiazepines cause decreases in endorphins and endobenzodiazepines||In the brain stem and limbic systems, long-term use of these agents may lead to drug dependence, depression, and increased pain|
|Cyclooxygenase inhibitor||Decreased peripheral inflammation mediated by prostanoids; spinal prostanoids facilitate substance P and glutamate pain fiber transmission; central (supraspinal effects) present mechanism is unclear|