The Rheumatoid Neck: Changing Pathology Requires Altering Surgical Strategies
Rheumatoid arthritis (RA) is a relatively new disease in the Western world. The first accurate depiction of its effects on the hands can be seen in a Renoir drawing of the artist at work at the end of the 19th century.1 At the beginning of the 20th century, A.B. Garrod coined the term “rheumatoid arthritis” and his son, A.E. Garrod, reported that the disease involved the cervical spine.2 The hands, major limb joints, and cervical spine are involved by the acute or chronic systemic inflammatory disorder characterized by an erosive synovitis, ultimately producing joint destruction, deformity, and painful incapacity. However, the introduction of disease-modifying drugs that dampen the inflammatory process and the decreased use of long-term systemic steroids, once the cornerstone of treatment, have considerably reduced the disease′s impact on the joints. The change in medical management of the disease coupled with our better understanding of its long-term effects on the cervical spine by the availability of computed tomography (CT) and magnetic resonance imaging (MRI) have resulted in a major shift in surgical management. Operations once performed late in the course of myelopathy have been replaced with prophylactic operations for cervical instability to prevent myelopathy from developing.3,4 For example, although vertical translocation of the odontoid peg was common 30 years ago, now it is an extremely rare presentation, with few transoral operations performed for rheumatoid disease ( Fig. 7.1 ).
Epidemiology of Rheumatoid Arthritis
RA is two to three times more common in females than in males, with a peak age of onset in the fifth decade, although a juvenile form predominantly affects adolescent females.5 The disease affects 0.5 to 1% of the population in the West, but the incidence around the world varies considerably; RA is about five times more common in the Pima Indians of Arizona in the United States and almost unheard of in South Africa and Nigeria.6
Etiology
The appearance of the disease in Europe after the discovery of the Americas led to the hypothesis of an infectious agent, further supported by a decreasing incidence in Caucasians and an increase in Asian populations.7,8 Epstein-Barr virus, parvovirus, Proteus, Borrelia, and mycobacteria among others have been implicated as possible candidate agents6 that theoretically might cause arthritis by direct effects or the induction of an autoimmune response. The incidence of rheumatoid disease in Asians who now live in the West is twice as high compared with Asians living in their native country, suggesting that environmental factors are significant.5 Other factors include smoking, blood transfusion, and obesity,9 which increase the risk of RA, whereas a diet rich in omega-3 fatty acids (like those of Eskimos and Pacific Islanders) decreases the risk.10 Steroids have a disease-modifying effect. Multiparous women and those taking oral contraceptives are at a lower risk of developing the disease, but there is an increased risk postpartum.6,11 In 75% of pregnancies, symptoms temporarily improve, possibly due to placental steroid production.12
RA has a genetic component, and familial clustering is sometimes evident,13 although a concordance of only 12 to 15% is seen in monozygotic twin studies.14 About 40% of the genetic contribution may be accounted for by genes in the human leukocyte antigen alleles (HLA) region as well as other non-HLA genes (e.g., coding for interferons, interleukins, and tumor necrosis factors).14
Pathology
Although juvenile RA predominantly causes ankylosis, RA in adults causes joint erosion and instability. In the acute phase, synovial tissue in joints becomes hypertrophic and edematous and forms villous projections that invade joint cavities, local bone, ligaments, and cartilage. Histological examination of this synovial overgrowth, commonly known as pannus, reveals the accumulation of T cells, B cells, plasma cells, mast cells, macrophages, and natural killer cells that produce proinflammatory cytokines and adhesion molecules. However, O′Brien and colleagues showed that, in patients with cord compression at the craniovertebral junction (CVJ), much of the soft tissue mass consisted of end-stage degenerating ligaments and bone as well as pannus formation.15 Involvement of cervical bone and joints, particularly at the CVJ, can lead to structural instability and acute cord compression, but as O′Brien has shown, it is more common for the compressive mass to result from undetected or untreated subluxation causing the cord damage over a decade or more.15 A combination of movement, compression, and stretching of the cord produces edema, axonal swelling, and necrosis principally in the dorsal white matter. It is this repetitive minor trauma associated with the unstable joint—not rheumatoid vasculitis—that leads to clinical myelopathy in most patients.16 This trauma explains why many patients have asymptomatic atlantoaxial subluxation and why it takes many years of repeated minor trauma to produce symptoms of myelopathy. The therapeutic implication is that medical treatment should be directed at preventing ligamentous damage and joint instability, and surgical fixation should be performed before the onset of myelopathy if possible.
Natural History
RA initially involves the hands and feet, and large joints usually require surgery before the neck does. Wolfe found that one in four patients in the United States had a large joint arthroplasty in the first 6 years of the disease, and Casey and colleagues revealed that cervical disease required surgical treatment in patients who had two to four previous arthroplasties.17,18 These findings imply that the greater the mobility in a rheumatoid joint, the faster it will degenerate and become unstable. Hence, the cervical spine is affected commonly, particularly the atlantoaxial joint and the CVJ, with the lower spine seldom involved.
The percentage of patients with RA who develop atlanto-axial subluxation varies between series and is largely biased by the source of data collection. Few large, population-based, cohort studies are found in the literature, but from nonsurgical studies over the past decades the percentage of patients with RA with atlantoaxial subluxation varies from 14 to 73%,19,20 with an average incidence of 35% (21% horizontal and 14% vertical subluxation).21 More than 30% will have symptomatic atlantoaxial subluxation 5 to 7 years after the onset of the disease. Five percent then become myelopathic a decade later (14 to 17 years after onset).22–24 Once mye lopathy has developed, the outlook is poor with up to 50% mortality within a year. Studies by Sunahara, Hamilton, and Casey and colleagues suggest that life expectancy is ~7 years from the onset of myelopathy and, once bed-bound, death often occurs within 1 year.17,25,26 Recent studies of atlantoaxial subluxation in nonsurgical European patients, such as the Early Rheumatoid Arthritis Study (ERAS; a multicenter outpatient study), have shown a lower prevalence than before (range 12 to 33%),27–31 although the incidence of disease presenting to a surgeon remains higher as expected.32
The Changing Pattern of Rheumatoid Arthritis
The disease pattern is changing, partly as a result of improved medical treatment and possibly due to altered expression of the disease.3 Although steroids unquestionably produce relief of symptoms, it is possible that some of the neck problems described in the 1960s, 1970s, and 1980s were caused or exacerbated by steroid treatment itself. In the past few decades, there have been several advances in the development and use of disease-modifying antirheumatic drugs (DMARDs); in the 1980s, cyclosporine became available, then minocycline and leflunomide, and more recently antitumor necrosis factor (anti-TNF) agents (infliximab, etanercept, and adalimumab) and interleukin 1 (IL1) antagonists (anakinra).33,34 The anti-TNF and anti-IL1 agents have been shown to improve symptoms and signs, decrease radiographic progression, and reduce pain and fatigue.35–37 Neva and colleagues demonstrated that the use of combination DMARDs has significant advantages over single therapy in reducing the radiological incidence of atlantoaxial subluxation.38
We have recently observed a decrease in patients requiring surgery for cervical rheumatoid disease, and similar trends have been reported for large joint replacement in rheumatoid patients in California and in Rochester, Minnesota.39,40 The observed number of rheumatoid patients presenting for cervical surgery now is much lower than might have been expected based on past decades. Ten years ago, 62,700 patients in the United Kingdom and 220,000 in the United States were predicted to require cervical spinal fixation.28,41 In 2000, however, Hamilton and colleagues followed up with 3800 rheumatoid patients in western Scotland and detected only 0.7% of patients requiring cervical surgery, a 10-fold decrease in the previously estimated number.25 Looking at the prevalence of atlantoaxial subluxation and translocation in ERAS, James and colleagues suggested a 5.6% prevalence of subluxation in the first 5 years and translocation in 3.1%,42 a much lower prevalence of patients at risk in the 21st century.
Evaluation
In the past, the outcome of surgery was reported in terms of operative mortality and major complications. Surgery was performed often in patients with end-stage disease, who did not have an option—quality of life and less serious complications were overlooked. Now that surgery is aimed at prevention of myelopathy, deformity, and treatment of pain, it is essential that internationally accepted objective measures of function and quality of life are used by spine surgeons.
Clinical Assessment
Clinical grading tools have been devised for various spinal conditions, but these are not always suitable for evaluation of rheumatoid myelopathy. The first universally accepted grading, the Ranawat classification ( Table 7.1 ), is simple to apply but is a blunt tool that does not distinguish small differences in the severe stages of myelopathy.43 The move toward using more accurate measures of quality of life rather than physician-reported neurological scales allows comparative health-care studies (e.g., the “short forms” of SF36 and SF12 questionnaires and the EuroQol EQ-5D tool for quality of life and cost effectiveness analysis).44–46
Radiological Assessment: Cervical X-rays
Despite a trend in the 1980s and early 1990s to establish radiological criteria for surgery, such methods are now relegated to the archives of history ( Fig. 7.2 ) and are not particularly useful for clinical decision making since the advent of CT and MRI ( Fig. 7.3 ). Although cervical radiographs were the key to surgical decision making in the past, they are now largely for diagnosis of the at-risk patient rather than directly influencing the management plan. Previously, lively debates around the degree of acceptable subluxation and when to perform fixation47–53 have been put in perspective by White and Panjabi′s demonstration that gross movements occurred when the transverse and apical alar ligament complex was damaged.54 Thus, significant instability should be defined by the point at which these ligaments fail mechanically, and it is incorrect to define a continuous gradation of instability depending on the degree of radiological subluxation. Another fallacy of plain radiographs is that a previously unstable atlantoaxial joint that develops vertical translocation may appear stable. The stability in these cases might be compared with the stability of an impacted fractured neck of the femur. Further evaluation should include MRI and CT scanning to determine the presence of neural compression, soft tissue masses, and malalignment and to estimate bone quality.
Computed Tomography Scans
Although the anterior atlantodental interval on a CT scan is useful in the diagnosis of atlantoaxial subluxation, the posterior atlantodental interval (PADI) is of more practical use and is significant if less than 14 mm. PADIs of less than 10 mm correlate well with MRI findings of significant cord compression but, like plain radiographs, assessment of the PADI does not take into account the effect of local soft tissue or pannus that is not seen, and significant brainstem compression by vertical translocation can be underestimated.55 CT scans of the CVJ are useful for assessing the degree of joint erosion, bone density, and presence of degenerative cysts and allow an appreciation of the likelihood of good screw purchase.
Magnetic Resonance Imaging
MRI scanning is effective in revealing neural compression by soft tissue masses that is not seen by X-ray or CT and is essential to demonstrate the full extent of brainstem compression caused by vertical translocation. On sagittal scans in flexion, a spinal cord diameter of less than 6 mm was taken as the criterion for diagnosing significant compression in one study.56 Flexion MRI may diagnose significant cord compression in an additional 26% of patients compared with neutral position MRI.56 However, MRI findings do not necessarily correlate with clinical findings because myelopathy is usually the result of years of repetitive trauma; therefore, cord compression in itself is not directly associated with clinical signs.16 Because of this discrepancy, Hamilton and colleagues highlighted the importance of making decisions on the basis of clinical and not radiological findings.25
Related posts:
Primary Osseous and Metastatic Neoplasms of the Craniovertebral Junction
Congenital Malformations of the Craniovertebral Junction
Management of Vertebral Artery Dissections and Vascular Insufficiency
Traumatic Injuries of the Craniovertebral Junction
Management of Intramedullary Lesions of the Cervicomedullary Junction and High Cervical Spinal Cord
Aneurysms of the Craniovertebral Junction
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
