Joint Bone Spine 73 (2006) 599–605 http://france.elsevier.com/direct/BONSOI/
Clinical-state-of-the-art
Polymyalgia rheumatica: diagnosis and treatment Martin Soubrier*, Jean-Jacques Dubost, Jen-Michel Ristori Service de Rhumatologie, CHU de Clermont-Ferrand, place Henri-Dunant, BP 69, 63003 Clermont-Ferrand cedex 1, France Received 1 August 2006; accepted 6 September 2006 Available online 12 October 2006
Abstract Polymyalgia rheumatica (PMR) typically manifests as inflammatory pain in the shoulder and/or pelvic girdles in a patient over 50 years of age. This condition was long underrecognized and therefore underdiagnosed. Today, however, overdiagnosis may occur. Physicians must be aware that many conditions may simulate PMR, including diseases that carry a grim prognosis or require urgent treatment. PMR may be the first manifestation of giant cell arteritis, and a painstaking search for other signs is mandatory. PMR may inaugurate other rheumatologic diseases such as rheumatoid arthritis, RS3PE syndrome, spondyloarthropathy, systemic lupus erythematosus (SLE), myopathy, vasculitis, and chondrocalcinosis. Finally, PMR may be the first manifestation of an endocrine disorder, a malignancy, or an infection. Failure to respond to glucocorticoid therapy should suggest giant cell arteritis, malignant disease, or infection. Ultrasonography may assist in the diagnosis by showing bilateral subdeltoid bursitis. Glucocorticoids are the mainstay of the treatment of PMR. Although the optimal starting dosage and tapering schedule are not agreed on, a low starting dosage and slow tapering may decrease the relapse rate. Methotrexate is probably useful when glucocorticoid dependency develops. In contrast, TNF-α antagonists are probably ineffective. © 2006 Elsevier Masson SAS. All rights reserved. Keywords: PMR; Giant cell arteritis; RS3PE; Rheumatoid arthritis; Methotrexate
1. Introduction Polymyalgia rheumatica (PMR) is characterized by inflammatory pain and stiffness of the shoulder and/or pelvic girdles accompanied with laboratory evidence of severe inflammation in a patient older than 50 years of age [1–4]. Although this independent entity carries a good prognosis, it can occur as a manifestation of a number of diseases, some of which are serious or require immediate treatment [1–4]. The distinction is often impossible to achieve given the absence of pathognomonic signs, and nosological confusion probably contributes substantially to the differences in reported manifestations and treatments of PMR. 2. Epidemiology The lack of universally accepted classification criteria for PMR complicates the interpretation of epidemiological data. * Corresponding
author. E-mail address:
[email protected] (M. Soubrier).
1297-319X/$ - see front matter © 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2006.09.005
The disease is exceedingly rare before 50 years of age. Its prevalence after 50 years of age has been estimated at 1 case per 133 population [5]. The incidence increases with age [3,6]. PMR is more common in women than in men in all affected age groups, although the difference may be smaller in older populations [6]. A North-to-South gradient has been reported. In Norway, the annual rate after 50 years of age was 11.2/10,000 between 1987 and 1994 [7]. Annual rates of 8.4/10,000 in the UK and 1.9/10,000 in Spain have been reported [6,8]. These geographic variations may be ascribable to genetic factors. Supporting this hypothesis is the high incidence of PMR in Minnesota, where a large proportion of the population is of North European descent [9]. Until recently, the incidence of PMR was thought to be stable [9]. However, a recent study from the UK showed an increase in the incidence of the disease from 6.9/10,000 patient-years in 1990 to 9.3/10,000 patient-years in 2001 [6]. 3. Etiology Seasonal variations in the incidence of PMR have been reported, suggesting a role for an infectious agent [6,10–13].
600
M. Soubrier et al. / Joint Bone Spine 73 (2006) 599–605
Several organisms have been incriminated, including adenovirus, respiratory syncytial virus, type 1 parainfluenza virus, parvovirus B19, Mycoplasma pneumoniae, and Chlamydia pneumoniae [10–13]. Other studies found no evidence of seasonal variations or associations with infectious agents [14,15]. The ethnic distribution and reports of familial cases suggest a role for genetic factors. The HLA DRB1*04 and DRB1*01 alleles are associated with PMR and perhaps also with greater severity of the disease [16,17]. Genetic polymorphisms for intercellular adhesion molecules (ICAM-1), TNF-α, and the IL-1 receptor antagonist may influence the susceptibility to PMR [3]. Inappropriate secretion of serum cortisol, ACTH, and dehydroepiandrosterone sulfate has been documented, suggesting a pathogenic role for adrenal gland dysfunction [18– 20]. 4. Clinical manifestations Inflammatory pain with morning stiffness for longer than 1 hour in the shoulder girdle (70–95% of patients) and/or pelvic girdle (50–70%) is the typical presentation. Pain for at least 1 month is required for the diagnosis. The cervical or lumbar spine may be affected also. The pain may be confined to one side at first but rapidly becomes bilateral. One-third of patients have constitutional symptoms such as low-grade fever, asthenia, anorexia, and weight loss. The marked functional impairment contrasts with the paucity of the physical findings. Examination of the shoulders fails to disclose objective evidence of inflammation. The active range of motion of the shoulders may be diminished, most notably in the morning. Passive range of motion is occasionally reduced. Muscle strength is normal. Peripheral manifestations seem common [21]. Thus, 79 (45%) of 177 patients followed up prospectively in Italy exhibited peripheral manifestations [21]. Asymmetric polyarthritis without joint erosions was noted in 45 patients (25%), carpal tunnel syndrome in 24 (14%), and swelling of the hands in (12%). These peripheral manifestations coincided with girdle involvement in 69% of cases and occurred during follow-up in the absence of girdle symptoms in 30% of cases, indicating a relapse. Peripheral arthritis may develop, most notably in women with a history of multiple relapses over several years of treatment. Peripheral edema predominantly affects older patients treated with low-dose glucocorticoid therapy for a short period. 5. Laboratory tests Erythrocyte sedimentation rate (ESR) elevation to more than 40 mm/h is a major diagnostic criterion for PMR. However, the value is normal in 7–20% of patients [22–25]. C-reactive protein (CRP) elevation with a normal ESR has been reported [22]. IL-6 elevation is common [26,27], and presence of this abnormality despite treatment indicates an increased risk of relapse [26,27]. However, the IL-6 assay is not widely available. Cholestasis without jaundice is found in 20–30% of cases
[3]. Tests are negative for rheumatoid factors and antinuclear antibodies. Muscle enzyme levels are normal. 6. Imaging studies Studies using ultrasonography and magnetic imaging resonance (MRI) have established that bursitis and synovitis are very common in patients with PMR [28,29]. Acromiodeltoid bursitis and subdeltoid bursitis were present in 55 of 57 patients in one study [29], compared to only 25 of 114 controls (46 with rheumatoid arthritis, 21 with spondyloarthropathies, six with connective tissue diseases, 35 with osteoarthritis, and six with fibromyalgia). Bursitis was present in 12 of the 46 patients with rheumatoid arthritis and 7 of the 21 patients with spondyloarthropathies. Of the 55 patients with bursitis and PMR, 53 had bilateral bursitis, compared to only 1 of the 25 controls with bursitis. Bilateral bursitis had 92.9% sensitivity, 99.1% specificity, and 98.1% positive predictive value for PMR [29]. In another study [30], however, bilateral bursitis was noted in only 35 (70%) of 50 patients with PMR, compared to 22 (44%) of 50 patients with rheumatoid arthritis. MRI studies have shown tenosynovitis in patients with peripheral edema [31]. 7. Diagnostic criteria Diagnostic criteria for the diagnosis of PMR are empirical. The first criteria were developed by Bird et al. in 1979 (Table 1) [32]. However, they fail to include pelvic girdle involvement, a prompt response to glucocorticoid therapy, or exclusion of other diagnoses. Among the many other criteria sets [33], those developed by Hunder and Healey are widely used in the US and UK (Table 1) [34,35]. The only difference between them is that a prompt response to glucocorticoid therapy is included in the Healey set [3]. Not surprisingly, sensitivity is highest for the Bird criteria set (99.5%), followed by the Hunder set [33]. Table 1 A: Criteria for PMR developed by Bird and colleagues: three of the following seven features are required; B: criteria for PMR developed by Healey and colleagues: all six criteria are required A 1° 2° 3° 4° 5° 6° 7°
Bilateral shoulder pain and/or stiffness Bilateral upper arm tenderness Onset of illness within 2 weeks ESR >40 mm/h Morning stiffness >1 hour Age >65 years Depression or weight loss or both
B 1° Pain for at least 1 month at two or more of the following sites: shoulders, pelvic girdle, and cervical spine 2° Morning stiffness >1 hour 3° Age >50 years 4° ESR >40 mm/h 5° Exclusion of other diagnoses 6° Prompt and marked response to glucocorticoid therapy in a dosage
