Chapter 75 Complex Regional Pain Syndrome of the Knee

Complex regional pain syndrome (CRPS) is the most current nomenclature used to describe an uncommon, yet potentially disabling condition of uncertain origin that can affect the knee. CRPS is characterized by pain disproportionate to the initial inciting event,⁵ in association with a variety of findings, including edema; changes in sensory perception, color, and temperature; and abnormal hair and nail growth in the affected limb.³¹ First recognized during the Civil War, when it was referred to as causalgia to describe intense, burning extremity pain after an injury,²⁴ it has gone through a number of name changes, including the widely recognized term reflex sympathetic dystrophy (RSD). Although many clinicians still refer to this condition as RSD, the diagnostic entity of CRPS, as published in 1994 in the Taxonomy Monograph of the International Association for the Study of Pain (IASP),²³ is believed to more accurately reflect the clinical criteria associated with this syndrome. Criteria are divided into two subtypes: type I and type II. The primary differentiation is based on whether the inciting incident includes a definable nerve injury (Box 75-1).

Box 75-1

Adapted from Merskey H, Bogduk N: Classification of chronic pain, Seattle, Wash, 1994, IASP Press.

IASP Diagnostic Criteria for Complex Regional Pain Syndrome (CRPS)

A recent retrospective study of 1280 patients who underwent total knee arthroplasty (TKA) for osteoarthritis revealed that 8 patients developed signs and symptoms consistent with the diagnosis of CRPS—an incidence of 0.7%. These patients demonstrated a slower, more arduous recovery than uncomplicated TKA patients, appeared to benefit from manipulation under anesthesia, and had less flexion at final outcome, but they were nonetheless better off after TKA than preoperatively.⁹

Physicians should consider CRPS in their differential diagnosis of intense, burning pain of the knee that cannot be attributed to another disease process because it has been well documented that early diagnosis leads to a better prognosis. This chapter focuses on hypotheses of the underlying causes and pathophysiology of CRPS, diagnostic criteria and testing, and treatment options.

Etiology and Pathophysiology

The causes and pathophysiology of CRPS remain poorly understood. No single hypothesis explains all the features of this syndrome; many mechanisms have been proposed to play a role in the pathogenesis of CRPS, including abnormal sympathetic activity, abnormal neuroendocrine or inflammatory response, genetic predisposition, central or peripheral nervous system dysfunction, stressful life events, and/or psychological dysfunction.²⁴ Trauma, even minor in severity, often precedes onset in the majority of patients. Spontaneous onset of this syndrome with no precipitating event has been described in 3% to 11% of cases,¹⁵ indicating that other contributing factors may be implicated.

CRPS has an occurrence rate of 1% to 15% in cases of peripheral nerve injury. CRPS usually occurs secondary to fracture, sprain, and trivial soft tissue injury to an extremity. Incidence after fractures and contusions ranges from 10% to 30%. Although some cases may be associated with an identifiable nerve injury, many are not. Entities that lead to the development of CRPS include operative procedures, cast or splint immobilization, and prolonged bed rest. Various reports describe an incidence ranging from 0.5% to 2% after TKA. Occurrence is more common after minor injuries and minor surgical procedures, such as arthroscopy.

New etiologic theories have been suggested, particularly as a result of German studies. Accumulating evidence suggests that genetic factors may predispose to CRPS. Researchers have noted an association between elevated levels of soluble tumor necrosis factor receptor 1 (STNF-R1) and enhanced tumor necrosis factor-alpha (TNF-α) activity in patients presenting with polyneuropathy and allodynia.³⁰

de Rooij and colleagues attempted to identify a familial form of CRPS but were unable to find a clear inheritance pattern, although their findings suggest a genetic predisposition. Patients from families with two or more affected relatives clearly developed CRPS at a younger age and demonstrated a more severe phenotype than sporadic cases.¹⁵ A later study by the same author investigated the role of the human leukocyte antigen (HLA) system and found a significant association between certain HLA alleles in 150 CRPS patients with fixed dystonia, indicating that HLA loci are indeed implicated in susceptibility to or expression of the disease.¹⁶

Increasing evidence suggests involvement of the immune system in CRPS.⁴ One study demonstrated that 30% to 40% of CRPS patients have surface-binding autoantibodies against an inducible autonomic nervous system autoantigen.²⁸ Peripheral mechanical, chemical, or physical trauma is thought to be a trigger acting on nociceptive nerve endings, leading to rapid release of cytokines and neutrophils, which activate and sensitize primary and likely spinal afferents. This results in increased secretion of neuropeptides that mediate neurogenic inflammation into the affected body region. These neuropeptides, such as substance P³⁸ and calcitonin gene-related protein (CGRP), induce plasma protein extravasation, recruit immune cells,⁴⁶ and lead to the release of proinflammatory cytokines such as TNF-α, interleukin (IL)-1, and IL-6.¹ Chronic release of these neuropeptides is hypothesized to be responsible for clinically evident inflammatory CRPS symptoms. Further argument to implicate an immune origin is provided by the association of CRPS with different HLA alleles, because many autoimmune diseases have similar distinct HLA associations.²⁸ Additionally, speculation continues regarding the involvement of reactive oxygen species in triggering or sustaining CRPS.¹¹

Stanton-Hicks and others note a female predominance in the range of 60% to 80% of all cases.⁴⁵ Reported female-to-male ratios vary from 2 : 1 to 4 : 1. Because CRPS is clearly more prevalent in women, de Mos et al investigated the association between cumulative and current exposure to estrogens and the risk of developing this syndrome. Peak incidence of occurrence was found to be between the fifth and seventh decades, suggesting increased risk after menopause with a concomitant decrease in endogenous estrogen levels.¹⁴ It is interesting to note that although the outcome of a recent study did not reveal an association between CRPS onset and cumulative endogenous estrogen exposure or current oral contraceptive or hormone replacement therapy use, study findings did reveal decreased risk of CRPS during pregnancy and increased risk immediately after pregnancy. This observation correlates with findings when estrogen was studied in other inflammatory diseases, such as rheumatoid arthritis and multiple sclerosis. In addition to a similar profound female predominance in these disorders, they share clinical features with CRPS, including inflammatory signs, pain, and functional impairment. The prevalence of CRPS in a cohort of MS patients, for example, was high compared with the estimated prevalence in the general population.⁴²

de Mos and colleagues attempted to identify risk factors and thereby potential underlying disease mechanisms by investigating the medical history of CRPS patients before the onset of their illness. Their study revealed that CRPS patients were likely to report preexisting disorders of asthma, migraine, osteoporosis, menstrual cycle–related problems, and neuropathies. Investigators did not find an association with psychological factors,¹³ although a recent study conducted in The Netherlands to evaluate psychological features of CRPS patients with dystonia did reveal that early traumatic experiences (87%) might be a predisposing factor.³⁵

The pathophysiologic mechanism of CRPS indeed appears to include input from multiple systems, including somatic and visceral sensory systems, central endogenous control systems, the sympathetic nervous system, the somatomotor system, and the neuroendocrine system. Continued research will likely identify the specific mechanism or mechanisms that must be targeted to better diagnose and treat this disease.

Diagnosis

CRPS is generally regarded as a systemic condition involving both central and peripheral components of the neuraxis, as well as interactions between the immune and nervous systems. Schwartzman states in a review of the natural history of the disease that the signs and symptoms of CRPS patients can be divided into four distinct subgroups: (1) abnormalities in pain processing; (2) skin color and temperature changes; (3) edema, vasomotor, and sudomotor abnormalities; and (4) motor dysfunction and trophic changes.⁴⁰

To date, the IASP criteria have been followed to formulate a diagnosis of CRPS. The Budapest Consensus Panel has proposed revised diagnostic criteria that reflect international efforts to further improve the accuracy of diagnosing this syndrome (Box 75-2). In patients meeting criteria in two of four sign categories and three of four symptom categories, diagnostic sensitivity was 0.85 and specificity 0.69; if all symptom categories were met, sensitivity decreased to 0.70 and specificity increased to 0.94. This resulted in the greatest probability of an accurate diagnosis for both CRPS and non-CRPS patients of approximately 80% and 90%, respectively.²³ Although CRPS type I and type II subtypes were retained intact, a third diagnostic subtype, CRPS-NOS (not otherwise specified), was proposed to identify the 15% of patients previously diagnosed by IASP standards who would not fully meet the new clinical criteria, but whose signs and symptoms could not be better explained by another diagnosis.²²

Box 75-2

From Harden R, Bruehl S, Stanton-Hicks M, et al: Proposed new diagnostic criteria for complex regional pain syndrome. Pain Med 8:326–331, 2007.

Proposed Clinical Diagnostic Criteria for Complex Regional Pain Syndrome (CRPS)