Published: June 2014
Polymyalgia rheumatica (PMR) and giant cell arteritis (GCA) are two related and immune-mediated inflammatory conditions that occur in the elderly. PMR coexists in 40% of patients with GCA. Similarly, 10% of PMR patients develop GCA at some point during their disease course. The relationship between PMR and GCA is further demonstrated by their preference for similar patient populations, linkage to the same human leukocyte antigen haplotypes, similar cytokine patterns in temporal artery biopsies, and similarities in anatomic involvement on positron emission tomography (PET) imaging.1-3 PMR and GCA represent two extremes of a disease spectrum.
Bruce is credited with the first description, in 1888, of PMR, which he described as "senile rheumatic gout."4 However, Barber coined the term polymyalgia rheumatica in 1957, and it has become the universally accepted name for this condition.5
PMR is characterized by proximal, symmetrical musculoskeletal pain and stiffness. Symptoms of systemic inflammation are also common. A dramatic response to low-dose corticosteroids can be a valuable diagnostic tool in patients for whom the diagnosis is uncertain. The lack of response to prednisone raises the possibility of a paraneoplastic process manifesting with proximal pain and stiffness. There is still a need for caution in patients who have a dramatic response to treatment because in some patients (about 11%) with an initial PMR-like presentation, the condition evolves into a phenotype that is more that of rheumatoid arthritis and, less often, other systemic rheumatic illnesses.6
PMR has a predilection for patients older than 50 years. The mean age at onset is 73 years, and women are affected more often than men. Its annual incidence in Olmstead County, Minnesota, a population with mostly Scandinavian heritage, is 59 per 100,000. The annual incidence of the disease increases with age.7 Whites of northern European descent have a higher incidence of disease than people of African American or Latin American descent.8,9
Much has been learned about PMR and GCA, but their cause remains unknown. Their etiology is likely multifactorial, resulting in the interplay of age, environment, and genetic susceptibility. The suggestion that PMR may be a forme fruste of GCA was first advanced in the 1950s and 1960s.5 More recently, PET scans in PMR have revealed features of vasculitis in about 1/3 of patients.10 The pathophysiology for both diseases is similar, with abnormalities of cellular immunity leading to vessel and systemic inflammation. Cytokines such as interleukin (IL)-1, IL-2, IL-6, and tumor necrosis factor (TNF)-alpha are important in the development of inflammation in both PMR and GCA.11 In PMR, however, inflammation is believed to be predominantly limited to the articular and periarticular tissues, whereas GCA is characterized by the development of frank large-vessel vasculitis. The classic histologic feature of GCA is the presence of an inflammatory cell infiltrate consisting of T cells, dendritic cells and macrophages migrating from the adventitia of a muscular artery toward the media and intima.9 If histologic examination is performed, between 16% and 20% or more of PMR patients will actually demonstrate arteritis, requiring the diagnosis to be changed to GCA.12 Interestingly, while temporal artery biopsies from true PMR patients lack the inflammatory infiltrate, they do reveal similar cytokine profiles to those with GCA, suggesting that PMR may represent a form of early, subclinical vasculitis. One key difference in cytokine profile is the presence of IFN-gamma in the arterial wall of GCA patients, suggesting that this cytokine may ultimately be necessary for the histological development of vasculitis.2
Most patients describe a subacute onset of symptoms that remain persistent over time. The most common symptom (reported in 70% to 95% of patients) is symmetrical shoulder girdle pain and stiffness. Fifty percent to 70% report neck and pelvic girdle pain. Concurrent pain in the upper arms and thighs is common and is usually worse in the morning. Shoulder and leg discomfort can lead to difficulty dressing, hair grooming, and rising from a chair. One third of patients have flulike symptoms described as fever, malaise, anorexia, or weight loss.13
Physical examination findings may reveal pain that limits active range of motion in the shoulders and hips. Passive range of motion should be normal. Despite subjective symptoms of muscle weakness, muscle strength testing should be normal unless it is affected by examination discomfort or by another condition.13 Approximately 50% of patients have been said to present with distal extremity abnormalities including swelling of the knees, wrists, or metacarpophalangeal joints. Other reported findings include soft-tissue swelling; pitting edema of the hands, ankles, and feet; and median nerve compression. However, these findings are typical of inflammatory joint disease and not PMR. The examiner should direct the evaluation along other lines in attempting to define another diagnosis. Frank synovitis of the hands or feet should suggest rheumatoid arthritis or another inflammatory arthropathy. Further laboratory and imaging may be needed to differentiate the two. (See the chapter Rheumatoid Arthritis).
The diagnosis of PMR is based primarily on clinical features. Elevated acute phase reactants provide secondary support for the diagnosis. The erythrocyte sedimentation rate (ESR) is greater than 40 mm/hr in 90% of cases. Other laboratory findings include an elevated C-reactive protein (CRP), normocytic normochromic anemia, thrombocytosis, and elevated alkaline phosphatase. Elevation of muscle enzymes such as creatine kinase and aldolase is not a feature of PMR and should prompt consideration of an alternative diagnosis.
Imaging is not routinely employed in the diagnosis, however, PET/computerized tomography (CT) has been used recently to better understand PMR. A prospective study of 35 patients with a clinical diagnosis of PMR and negative temporal artery biopsies underwent PET prior to starting steroid therapy. Significant tracer uptake was noted in the shoulders (94%), hips (89%), and vertebral spinous processes (51%) of PMR patients. In addition, 31% had large-vessel uptake at diagnosis, again suggesting that subclinical vasculitis may be very common in these patients.10 Additional imaging modalities such as U/S and magnetic resonance imaging may be useful in ruling out other entities in the differential diagnosis such as rheumatoid arthritis, calcium pyrophosphate deposition disease, or true large-vessel vasculitis. The differential diagnosis may become confusing, however, when one considers that the average age of patients with PMR is >70 years, an age when osteoarthritis and calcium pyrophosphate deposition are common, thus increasing the likelihood that a patient with PMR may also have these comorbidities.
In 2012, the European League Against Rheumatism (EULAR) and American College of Rheumatology (ACR) joined forces to create new classification criteria. A scoring system was determined, whereby a patient over the age of 50 presenting with new onset bilateral shoulder pain and elevated ESR or CRP may be classified as having PMR if at least four clinical points or five points from clinical and ultrasound criteria are met.14 Although these criteria may serve as general guidelines for the diagnosis of PMR, it should be noted that they were developed as classification criteria for study enrollment, not for strict clinical diagnostic purposes. Most authorities agree that no single feature is necessary to diagnose PMR in all cases. The features noted in these criteria are common enough that if patients present without these symptoms or have a suboptimal response to corticosteroids, the diagnosis should be reconsidered. Conditions that can mimic PMR include malignancies, chronic infections, drug reactions, metabolic conditions (such as hypothyroidism), and other rheumatic diseases, including seronegative rheumatoid arthritis, calcium pyrophosphate disease, or polymyositis.6
The first successful use of corticosteroids in patients with PMR was reported by Kersley in 1951.5 Since that time, this treatment has remained the cornerstone of therapy for PMR. Prednisone or prednisolone are the most commonly used corticosteroids. Starting doses range from 15 mg to 20 mg per day. A dramatic response to therapy with near-total relief of symptoms should occur within 1 to 5 days. Lack of a dramatic response to corticosteroids should prompt physicians to reconsider the diagnosis. A gradual decline of the acute-phase reactants should be expected but should never be the sole gauge of therapy. After an adequate response to corticosteroids has been achieved, the initial dose should be maintained for 1 month before beginning a slow taper to the lowest effective dose. One to 2 years of treatment with corticosteroids should be expected, and a few patients require low-dose prednisone for several years.15
Disease flares during the corticosteroid taper are common and often require temporary increases in therapy. Disease flares can occur in the presence of normal acute-phase reactants. Increases in acute-phase reactants mandate an evaluation to be sure that comorbid conditions are not responsible for such changes. Isolated increases in acute-phase reactants should lead to more careful monitoring and not to reflexive increases in corticosteroid doses.16
Corticosteroids are the cornerstone of therapy, but they are not without side effects. Most patients have at least one relapse as corticosteroids are tapered, and adverse events occur in almost every patient. The role of immunosuppressive agents other than corticosteroids in PMR is controversial. Methotrexate has been studied in two randomized, double-blind, controlled trials. Van der Veen and colleagues reported that patients randomized to take oral methotrexate (10 mg/week) had the same number of relapses and received the same total cumulative prednisone dose compared with patients who received placebo.17 Caporali and colleagues reported that patients randomized to oral methotrexate (10 mg/week) for 48 weeks had fewer relapses and required lower cumulative prednisone doses than patients who took placebo.18 However, further review of the patients who received methotrexate revealed that they had the same number of relapses while they were taking prednisone as did patients who received placebo. Furthermore, the number of corticosteroid-related adverse events was equal in both treatment groups, and the total cumulative prednisone dose reduction achieved by taking methotrexate in place of placebo equaled only about 1 mg/day.19
TNF-alpha, a cytokine produced by macrophages and T-lymphocytes, appears to play a significant role in the inflammatory process of PMR and GCA. In one pilot study, 3 mg/kg of intravenous infliximab was administered as adjunctive therapy to patients on corticosteroids. This therapy allowed 12 months of remission in three of four patients treated with the drug.20 These results were not confirmed, however, in a subsequent double-blind, randomized, placebo-controlled study by the same author. Patients who received 3 mg/kg of infliximab at the same dosing intervals as used for rheumatoid arthritis over 22 weeks experienced the same number of weeks in remission as those who received placebo. No difference was seen in the duration of corticosteroid therapy or in the total number of patients who were able to discontinue corticosteroids between the groups.21
Most recently, treatment with IL-6 blockade has been gaining favor for patients with relapsing disease. Studies have documented elevated serum levels of IL-6 in patients with PMR, and these levels appear to correlate with disease activity, inflammatory markers, and response to prednisone.11 Accordingly, use of tocilizumab, a humanized monoclonal antibody against the IL-6 receptor, has been tried and reported in four patients with pure PMR, with promising results.22–25 At doses of 8 mg/kg once monthly, three of four patients entered clinical remission, and two achieved remission without concomitant use of prednisone.23,25 Interestingly, one patient did not have a clinical response, despite complete normalization of inflammatory markers, with tocilizumab, and remission was subsequently achieved with reinstitution of glucocorticoids.24 This case highlights the concern that when IL-6 is blocked for therapeutic reasons, inflammatory markers may no longer be reliable for assessment of disease activity. At this point, further study of tocilizumab in PMR is warranted.
Adequate treatment with corticosteroids allows most patients to enter clinical remission. However relapses occur in at least 30% to 50% of patients when prednisone is tapered.26–28 Factors associated with an increased risk of relapse include persistently elevated levels of IL-6 and CRP despite treatment29 and a rapid rate of prednisone withdrawal.26 Patients who have PMR need continued follow up to monitor for drug-related toxicities and for possible demonstration of underlying GCA. The development of a new headache or visual changes should prompt immediate medical evaluation for GCA and institution of higher doses of corticosteroids. Bilateral upper- and lower-extremity blood pressures should be obtained periodically. Differences between contralateral extremity pressures of 10 mm Hg or more should be considered abnormal. Bruits over carotid, subclavian, or femoral arteries may be due to either atheromatous disease and/or GCA, and require further evaluation through vascular imaging.
GCA is a vasculitis characterized by granulomatous inflammation of the medium and large arteries. Inflammation is seen most commonly in the extracranial branches of the carotid arteries, resulting in characteristic headache and scalp tenderness, and in other primary branch vessels of the aortic arch. Less often, internal branches of the carotid are affected, most notably the ophthalmic and posterior ciliary arteries, where stenosis or occlusion may cause loss of vision. At least one of five patients will develop clinically apparent large-vessel inflammation leading to branch vessel (e.g., most often subclavian) stenosis or aortic aneurysm.30 Imaging studies, especially PET scans, suggest 85% of patients have large-vessel disease, and one excellent post-mortem study indicates that 100% of patients with GCA have large-vessel involvement.29 GCA is the most common vasculitis in Caucasians over the age of 50. It has a predilection for people of northern European heritage, and women are affected at least twice as often as men.9
In the United States, GCA affects about 18 out of 100,000 people older than 50 years. The epidemiologic-demographic characteristics of GCA are similar to those of PMR. The incidence of GCA is much higher in the northern latitudes, with a mean age at onset of 74 years. GCA and PMR likely represent extremes of a disease spectrum, with many patients presenting with features of both diseases. Approximately 40% of GCA patients have concurrent features of PMR at some point during their disease course.31
The cause of GCA is unknown, but vessel wall inflammation is believed to be predominantly cell-mediated, rather than autoantibody-induced. Dendritic cells residing in the adventitia become activated by an unknown antigen, and signal for T lymphocytes to enter the vessel wall via the vasa vasorum. Activated T cells differentiate and clonally expand, producing IFN-gamma, which results in macrophages infiltration.32 Macrophages infiltrate all layers of the arterial wall, secreting a broad range of proinflammatory cytokines including IL-1, IL-6, and TNF-alpha; growth factors like platelet-derived growth factor and vascular endothelial growth factor; and matrix metalloproteinases. Subsequent panarterial inflammation leads to arterial damage, resulting in either intimal proliferation, and ultimately luminal narrowing or aneurysmal dilatation, most commonly affecting the thoracic ascending aorta (Figure 1). Luminal narrowing is ultimately responsible for ischemic events (loss of vision, stroke, and jaw and limb claudication), while aortic aneurysms may result in dissection or aortic valve insufficiency. Improved understanding of how these cytokines participate in inflammation, and determination of the antigen(s) that trigger initial dendritic cell activation will lead to better targeted therapies in both GCA and PMR.33
Headache is the most common symptom in GCA and occurs in 63% to 87% of patients. Systemic symptoms including fever, weight loss, and myalgias occur in 50% of patients.34 Other symptoms include scalp pain, jaw pain while chewing, and arm or leg claudication. Vision loss, the most dreaded complication of GCA, occurs in more than 30% of patients.35 Anterior ischemic optic neuropathy is the most common cause of blindness. Twenty-seven percent of patients develop either an aortic aneurysm or large artery stenosis at some point during their disease course. Six percent of patients who develop an aortic aneurysm present with symptoms consistent with a dissection of the aneurysm.30 Extremity claudication occurs when aortic branch vessels, such as the subclavian artery, become critically narrow. Stroke and vision loss can occur without any preceding symptoms; however, patients can present with insidious nonspecific symptoms before the diagnosis of GCA is made. Forty percent of patients present with symptoms not considered classic for GCA. These symptoms can include cough, throat pain, or tongue pain.
A thorough physical examination may reveal a prominent, tender temporal artery. Evaluation of the artery may reveal a decreased pulse and a nodular appearance. Asymmetrical extremity blood pressures or pulses, bruits over subclavian or carotid arteries, or a murmur of aortic insufficiency suggests aortic or primary aortic branch involvement.
Similar to PMR, no serologic test is diagnostic for GCA. Diagnosis is based on clinical symptoms, with support from the presence of abnormal acute-phase reactants. More than 90% of patients have an increased ESR, and elevations in CRP, alkaline phosphatase, and platelets are frequently seen. It should be noted, however, that inflammatory markers may be normal in up to 5% of patients with GCA, and therefore cannot be used to rule out disease if clinical suspicion is high36,37.
Temporal artery biopsy is considered the standard diagnostic test for GCA. The sensitivity of biopsy, in series from medical practitioners, in detecting GCA is about 50%.38 The yield of biopsy is a function of pretest probability, which might explain why some ophthalmology and other series, in which visual abnormalities were common, have yields as high as 80%.39 Biopsy of the contralateral temporal artery adds very little to the sensitivity of the test.40
Imaging of the vessel lumen with CT angiography (CTA) or magnetic resonance arteriography (MRA) may reveal aortic or arterial branches with stenoses or aneurysms (such as those seen in Figure 2) in up to 67.5% of patients at the time of diagnosis.41 The subclavian arteries, carotid arteries, and ascending aorta are the most commonly affected areas. Other arterial branches such as the mesenteric and renal arteries can also be affected. Vascular PET scans, with or without concomitant CT utilize 18F-fluorodeoxyglucose to portray areas of active inflammation and may reveal large-vessel uptake in between 50% and 80% of GCA patients at diagnosis.42,43 Furthermore, aortic uptake has been shown in one study to correlate with increasing vessel diameter at follow up.44 While PET appears to be a more sensitive marker for disease than biopsy, its diagnostic specificity and utility in the setting of glucocorticoid use are still in the process of being defined.45
Corticosteroids are the drug of choice for the treatment of GCA. They quickly reduce symptoms and decrease risk of visual complications from 60% to 14%.46 Therapy with corticosteroids should start when GCA is first suspected. Waiting to start corticosteroids until after a temporal artery biopsy could result in irreversible loss of vision.
The optimal initial dose of prednisone is unclear, but most authorities agree that the initial dose should be between 40 mg and 60 mg/day. Doses of at least 60 mg are preferred when presenting features include ophthalmic or neurologic complications. Some authorities advocate intravenous methylprednisolone at doses of 1,000 mg a day for 3 to 5 days in patients presenting with amaurosis fugax or blindness.46–48 The initial oral dose of corticosteroids should continue for 1 month before taper is considered. Many tapering schedules exist, but few have been studied in clinical trials. One general rule is to taper by 10% to 20% every 2 weeks.9 Treatment duration is different for each patient. Long-term therapy is required in most patients, and it is not unusual for corticosteroid therapy to extend beyond 4 years.49–51
The efficacy of methotrexate in granulomatosis and polyangiitis (GPA/Wegener's granulomatosis) and Takayasu's arteritis has led to its use in three randomized, double-blind, placebo-controlled trials for GCA. Jover and colleagues reported that patients randomized to methotrexate doses of 10 mg/week had fewer relapses and lower cumulative steroid doses compared with those who received placebo. However, a difference in relapse rates was only noted after 1 year of disease. There was no advantage to methotrexate in the first year. In addition, patients who received methotrexate had the same number of steroid-related side effects as those who received placebo.52 Hoffman and colleagues conducted the only multicenter, randomized, double-blind, placebo-controlled trial of methotrexate at doses of 15 mg each week. Their conclusions did not support the use of methotrexate as adjunctive therapy with corticosteroids. Patients who were randomized to receive methotrexate did not have reduced disease activity, cumulative corticosteroid doses, or corticosteroid-related toxicities.53 Spiera and colleagues reported similar results.54 The use of adjunctive methotrexate as a steroid-sparing agent in patients with GCA remains a controversial issue.
The finding of abundant TNF-alpha in GCA arteries (Figure 3) has led investigators to study TNF-alpha inhibition as a potential disease modulator in GCA. Hoffman and colleagues reported results of a multicenter, randomized, double-blind, placebo-controlled trial of intravenous infliximab as adjunctive therapy to corticosteroids in patients with newly diagnosed GCA.55 The study was stopped at 22 weeks when infliximab was shown to not improve durability of remission or reduce cumulative steroid doses. The results of this study suggest that although TNF is found in abundance in affected vessels, it might not play a critical role in the pathogenesis of GCA. Perhaps other mediators might play more important roles in disease propagation.
Although the evidence is conflicting,56 some studies have reported elevated levels of IL-6 in the serum and tissues of active GCA patients, making IL-6 inhibition of therapeutic interest in this disease. Small case series and case reports have been published and reviewed describing use of tocilizumab in 25 GCA patients to date.23,57–61 All reported patients have shown good clinical responses to tocilizumab, and no serious adverse events requiring drug discontinuation were described. Autopsy data from a single patient who died of an unrelated cause while in apparent clinical and biochemical remission with tocilizumab did, however, show ongoing, active vasculitis affecting the aorta and first-order branches.23 In addition, three of four described patients who discontinued tocilizumab therapy subsequently relapsed, indicating that this drug may be effective for controlling symptoms of disease but is not curative. The ideal dose and duration of therapy for GCA are not known, The use of tocilizumab in GCA is currently being formally evaluated in a multicenter, randomized, double-blind, placebo-controlled trial (GiACTA).62 Although not yet studied prospectively, a retrospective review of 175 GCA patients by Nesher and colleagues observed that those who took low-dose ASA (100 mg/day) were five times less likely to present with visual complications or stroke.63 A second report by Lee and colleagues showed similar results in patients on antiplatelet or anticoagulant therapy.64 Sixteen percent of GCA patients taking ASA, warfarin, or clopidogrel developed an ischemic event compared with 48% of patients not on this therapy (P <0.0005). Bleeding complications were not increased in the patients on antiplatelet or anticoagulant therapy. Low-dose ASA should be considered in all patients with GCA who have no contraindications for its use.
Some studies have found that the overall mortality in patients with GCA is similar to that of age- and gender-matched controls, while others have noted an increase in mortality, particularly from cardiovascular events. There is agreement, however, that patients with GCA have an increased risk of death from the complications of aortic aneurysms, as compared with the general population. Thoracic aneurysms are 17 times more likely to occur in GCA patients and can occur at any time during the disease course. Fifty percent of patients with aortic aneurysms experience dissection or rupture of their aneurysm.30,65–67 Cost-benefit data are not available; however, consideration should be given to periodic imaging of the large vessels. Because the risks of aortic catastrophes are well documented, we recommend at minimum regular, careful auscultation for aortic valve murmurs or bruits, and a yearly chest x-ray. Abnormalities should be followed by MRA or CTA to determine the extent and progression of the aortic lesion, as well as other large-vessel involvement. Size of the lesion, hemodynamic consequences, and change over time per sequential imaging determine the need for surgical interventions.
Nearly every patient treated with long-term corticosteroids develops complications related to therapy. Sixty percent develop severe adverse events such as corticosteroid-induced diabetes, avascular necrosis, glaucoma, or vertebral fractures.68 Corticosteroid-induced osteoporosis is a well-known complication of long-term steroid use. Therapies such as calcium, vitamin D, and antiresorptive agents should be considered in every patient who will receive corticosteroids for more than 3 months. Screening for osteoporosis with bone densitometry at the induction of corticosteroid therapy (and at regular intervals) is as important an intervention as prescribing steroids to prevent blindness.
PMR and GCA are autoimmune inflammatory diseases of the periarticular tissues and blood vessels, and most likely represent two extremes of the same disease spectrum. These diagnoses should be considered in persons over the age of 50 who present with systemic features, myalgias, headaches or symptoms of claudication with elevated inflammatory markers. At present, prednisone and anti-platelet agents remain the cornerstone of therapy; however, ongoing research is focused on the discovery of safer and more effective treatment strategies.