Giant Cell Arteritis and Polymyalgia Rheumatica: 2016 Update

Giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) are both more common among people of North European decent than among Mediterranean people. Women are 2–3 times more commonly affected. Giant cell arteritis and PMR are extremely rare before age 50 years. Polymyalgia rheumatica may be “isolated” or associated with GCA. There is increased expression of inflammatory cytokines in temporal arteries of PMR patients, without overt histological evidence of arteritis. One-third of “isolated” PMR patients have vascular uptake in positron emission tomography (PET) scans, suggesting clinically unrecognized, “hidden” GCA. Typical manifestations of GCA are headache, tenderness over temporal arteries, jaw claudication, PMR, acute vision loss, and low-grade fever. Bilateral aching of the shoulders with morning stiffness is typical for PMR. In both conditions sedimentation rate and C-reactive protein are elevated, and anemia and thrombocytosis may occur. Color duplex ultrasonography of the temporal arteries may aid in GCA diagnosis. Temporal artery biopsy showing vasculitis, often with giant cells, confirms GCA diagnosis. In cases with negative biopsy one must rely on the clinical presentation and laboratory abnormalities. The diagnosis of PMR is made primarily on clinical grounds. Other conditions that may mimic GCA or PMR must be excluded. Glucocorticoids are the treatment of choice for both conditions. Prompt treatment is crucial in GCA, to prevent irreversible complications of acute vision loss and stroke. Addition of low-dose aspirin may further prevent these complications. The average duration of treatment is 2–3 years, but some patients require a prolonged course of treatment, and some may develop disease-related or treatment-related complications. No steroid-sparing agent has been proven to be widely effective thus far, but some promising therapeutic agents are currently being studied.


INT RODUCTION
Giant cell arteritis (GCA) is a form of v asculitis that inv olv es the major branches of the aorta with predilection for the v ertebral, subclav ian, and the ex tracranial branches of the carotid arteries, including the temporal arteries. The aorta and other large and medium-sized arteries may also be inv olv ed. Giant cell arteritis develops in individuals older than 50 y ears of age, and its incidence increases with age. It is considered to be the most common ty pe of v asculitis in this age group.

EPIDEMIOLOGY
Giant cell arteritis is more common among North Europeans, reaching an annual incidence rate above 20 cases per 100,000 population at risk (age >50). It is less common among Mediterranean people. In Israel, the incidence is similar to other Mediterranean countries. [1][2] The incidence of GCA seems to hav e been decreasing in the last two decades. [1][2][3][4] Women are affected 2-3 times more commonly than men.
Poly myalgia rheumatica (PMR) patients share many epidemiologic and pathogenetic features with GCA . Like GCA, PMR dev elops in patients older than 50 y ears and is more common in women. Similar to GCA, the highest annual incidence rates were observed in Northern Europe, 50 -100 per 100,000 population older than 50 y ears. It is unknown whether PMR is just an ex pression of an underlying GCA. More likely , it seems that both are a result of an unknown causative factor (or factors), sometimes ex pressed as PMR, sometimes as GCA, and sometimes as a combination of both conditions. There is a wide range of the reported frequency of PMR in GCA patients (1 7 %-66%), and it is estimated that 1 0% of "isolated" PMR cases dev elop GCA sy mptoms. 5 Clinically , the two conditions may present together but may sometimes be separated by long interv als, and either one may present first.

ET IOLOGY AND PATHOGENESIS
Etiology is unknown. Sev eral studies implicated infectious agents such as varicella-zoster virus, 6 but DNA sequencing of temporal artery biopsy specimens from GCA cases showed no ev idence of prev iously suspected pathogens. 7 The pathogenesis of GCA has been ex tensiv ely studied but is still not fully understood. I t is considered to be a T celldependent disease. Upon dendritic cell activation in the adv entitia by an unknown antigen, CD4 T cells are recruited and polarized into Th1 and Th17 lines, producing interferon (IFN)-γ and interleukin (IL)-1 7 , respectiv ely , as their main cy tokines. The production of cy tokines and activ ation of macrophages and v ascular smooth muscle cells induce sy stemic manifestations, v ascular remodeling, and local ischemic manifestations. 8-10 Macrophages, often forming giant cells, are the major source of cy tokines, growth factors, and metalloproteinases.
There is ev idence for increased ex pression of inflammatory cytokines in the temporal arteries of PMR patients, without overt histological evidence of arteritis. There is local expression of IL-1, IL-2, IL-6, and transforming growth factor (TGF)-beta. 11 Howev er, in contrast with GCA, T cells producing IFN-γ are not attracted to the vessel wall. This lack of IFNγ ex pression in temporal arteries from PMR patients suggests that its production is crucial to the development of ov ert vasculitis and that a Th1 response may be suppressed in PMR. Using positron emission tomography (PET) scans, increased uptake was documented in thoracic blood vessels in one-third of apparently "isolated" PMR patients, suggestiv e of inflammation in these v essels, 1 2,1 3 or clinically unrecognized, "hidden" GCA.

CLINICAL FEATURES
The signs and sy mptoms of GCA may be classified into four subsets: signs and sy mptoms of cranial arteritis, ex tracranial arteritis, systemic symptoms, and PMR. Patients may develop any combination of these manifestations. In most instances symptoms dev elop gradually over a period of several weeks, but onset may sometimes be abrupt.
Headache is a frequent presenting symptom and is ty pically felt over one or both temporal areas. The temporal arteries may seem prominent and tender.
Generalized or occipital headaches are not uncommon. The headache may be continuous or paroxy smal. Pain in the jaw during mastication, facial pain, and scalp tenderness may also be present. Rarely , GCA may lead to segmental scalp necrosis or tongue infarction.
Neurologic manifestations are uncommon. Strokes occur in 3%-7 % of GCA patients. 14 Involvement of the vertebro-basilar system is relatively more common in GCA -related strokes than in atherosclerotic strokes. 1 4,1 5 It is important to note that strokes may still develop after glucocorticoid therapy is begun. 14,16,17 Neuropsychiatric manifestations and peripheral neuropathies are uncommon.
Arteritic anterior ischemic optic neuropathy (A-AION), dev eloping in 5%-1 5% of the patients, is the leading cause of blindness in GCA, sometimes being the presenting manifestation. 18, 19 Arteritic AION is unilateral in most cases, but v isual loss in the other ey e may develop; A -AION results from v asculitis of the posterior ciliary arteries, branches of the ophthalmic artery, which supply the optic nerv e head and the choroid. Less common causes of v isual loss are central retinal artery occlusion, posterior ischemic optic neuropathy, and cortical blindness. V ision loss in GCA is sudden. Howev er, sev eral studies reported that 50% or more of GCA patients with irrev ersible v isual loss had premonitory v isual sy mptoms, such as blurry vision, amaurosis fugax, v isual hallucinations, or diplo pia. 1 4,1 8 Amaurosis fugax is the most ominous sign of impending v isual loss. It should be considered a medical emergency in GCA, as prompt treatment with glucocorticoids and low-dose aspirin may prevent the dev elopment of irrev ersible blindness. Jaw claudication is also associated with increased risk of dev eloping vision loss. 20 Following v ision loss, the v isual outcome is poor, ev en in patients treated with high doses of steroids. Only 5% had some improvement of both v isual acuity and central v isual field. 21 V estibulo-auditory manifestations are quite common in prospectiv e studies. 22 Sy mptoms are unilateral or bilateral hearing loss, v ertigo, and tinnitus. Onset is usually insidious. A total of 89% of GCA patients had abnormal v estibular tests, 64% had subjective hearing impairment, 52% had v ertigo, and 50% had tinnitus. These were reversible in most cases following steroid therapy .
Signs of occlusive changes in large arteries of the chest and ex tremities are uncommon. The clinical findings are those of the aortic arch sy ndrome, indistinguishable from those of Takay asu arteritis: arm claudication, Ray naud's phenomenon, absent or decreased pulses, and bruits ov er the inv olv ed arteries. Clinical signs of aortitis are uncommon. 23 Howev er, this condition is under-diagnosed, as sy mptoms are frequently mild, non-specific, and insidious. It is estimated that GCA patients have a 2fold increased risk of aortic aneurysm. 24 Sy mptoms suggestive of inv olvement of the coronary, mesenteric, and lower ex tremity arteries are rare.
Sy stemic manifestations (fever, malaise, fatigue, anorex ia, and weight loss) occur quite often, in 30%-60% of the patients. In some cases, these may be the only symptoms of GCA. Fever is typically lowgrade, and temperatures rarely ex ceed 39°C. The ty pical symptoms of PMR are aching of the shoulder girdle and neck, associated with morning stiffness. The hip girdle may also be inv olv ed. These sy mptoms are probably related to inflammation of the glenohumeral and hip joints, and the subacromial, subdeltoid, and trochanteric bursae. One -third of PMR patients hav e systemic manifestations such as low-grade fever, malaise, and anorexia, but these are milder than in GCA. A subgroup of PMR patients presents with arthralgia or synovitis of peripheral joints, predominantly hands and knees. 25 Some present with sy mmetric sy nov itis with pitting edema, compatible with the relapsing seronegative sy mmetric sy nov itis with pitting edema (RS3PE) sy ndrome. 26 It is difficult at times to distinguish between these patients and patients with elderlyonset rheumatoid arthritis (RA). 27

DIAGNOSIS
The diagnosis of PMR and GCA is made primarily on clinical grounds and is bolstered by laboratory ev idence of an acute-phase reaction, most commonly elev ated levels of erythrocyte sedimentation rate (ESR) and C-reactiv e protein (CRP), as well as anemia of inflammation and thrombocytosis. Various conditions can mimic GCA and PMR, 28 and should be considered in the differential diagnosis (Table 1 ). A small group of patients may prese nt with ty pical features of the disease but without elev ation of the ESR. 29 The only test that confirms the diagnosis of GCA is a temporal artery biopsy, showing vasculitis with mononuclear cell infiltrates, often with giant cells (Figure 1 ). Giant cell arteritis affects the v essels in segments, therefore areas of v asculitis may be missed and the histological examination is normal in 1 0%-30% of GCA patients ("biopsy -negativ e GCA"). 30 A temporal artery biopsy length larger than 5 mm is associated with increasing diagnostic yield, but the optimal length is probably higher. 31 Biopsy ing temporal arteries on both sides may also increase the diagnostic y ield. 32 It is preferable to perform the biopsy as soon as possible, but the specimen may show signs of arteritis even after 2-4 weeks of treatment. 33 Some imaging modalities may aid in the diagnosis of GCA. Among those, color duplex ultrasonography of the temporal arteries is more commonly used. A peri-luminal hypo-echoic halo (Figure 2), probably representing v essel-wall edema, was suggested to be highly sensitive and specific for GCA. 34 Howev er, results are operator-dependent and vary considerably among studies ex amining the diagnostic v alue of color duplex ultrasonography . 35 High-resolution contrast-enhanced magnetic resonance imaging (MRI) of the temporal arteries also enables evaluation of possible inflammation of the v essel wall. Preliminary results show high sensitiv ity of this imaging modality , 36 with diagnostic y ield comparable to color duplex ultrasonography. 37 Angiography of the aortic arch and its branches may serv e to diagnose large-vessel inv olv ement. Noninv asive modalities, such as PET scans, may also be employed to detect large-vessel inv olvement in the chest, neck, and abdomen. 38 The specificity of PET for GCA diagnosis is considered to be v ery high (~1 00%), but sensitivity is lower (~66%). It may also serv e to evaluate response to treatment and GCA disease ex acerbations.
Serological tests were not helpful in diagnosing GCA. Autoantibodies were not consistently found in GCA, although plasma cells can be found in the adv entitia in 7 %-24% of temporal artery biopsies from patients with GCA. 39 The ex ception was antiphospholipid antibodies (APLA), which were found in 30%-80% of GCA cases. [40][41][42][43] Antibodies to lamin C were found in one-third of GCA patients and in none of the controls. 44 Autoantibodies to a human ferritin peptide (the heavy chain N-terminal) were found in 92% of 36 patients with GCA and/or PMR. 45,46 In addition, 89% had antibodies to bacterial ferritin peptide of Staphylococcus epidermidis. Anti-ferritin antibodies were found in much lower rates in disease controls. Following these reports, another group of researchers reported their experience with anti-ferritin antibodies in GCA. 47 They found a test sensitiv ity of 82% in biopsy -positiv e  Image shows intense trans-mural inflammatory infiltrate, multi-nucleated giant cells, intimal hyperplasia, and severe narrowing of the lumen.
GCA patients. Anti-ferritin antibodies were found in 34% of disease controls and 3% of healthy controls. Upon further testing, these antibodies may prov e useful as a diagnostic marker of GCA.
There are no independent v alidating criteria to determine whether GCA is present when a temporal artery biopsy is negative. The American College of Rheumatology (ACR) criteria for classification of GCA 48 may assist in GCA diagnosis ( Table 2). Howev er, those classification criteria serve mainly to classify GCA within the group of v asculitides. Their purpose was to differentiate GCA from other types of v asculitis, and not to differentiate GCA from other disease conditions. It is highly important to realize that these are not diagnostic criteria. Their v alidity as criteria for GCA diagnosis has been questioned. 49 Such classification criteria do not work well when used for diagnosing indiv idual cases. 50 The final diagnosis should be based on all clinical, laboratory, imaging, and histological findings.
The diagnosis of PMR is made primarily on clinical grounds and laboratory evidence of an acute phase reaction. 28,51 There is no single diagnostic test for PMR, but sets of diagnostic and classification criteria have been suggested by sev eral groups of inv estigators. Recently, prov isional classification criteria (Table 3) were published as a collaborative initiativ e of the European League Against Rheumatism (EULAR) and the ACR. 52 With these criteria, the required score has 68% sensitiv ity and 7 8% specificity for discriminating PMR from comparison patients. The positive predictive v alue is 69%, and the negativ e predictive value 7 7%. These new classification criteria need further v alidation.
There is a wide range of the reported frequency of GCA in patients presenting with PMR. One diagnostic option is routinely to biopsy the temporal arteries in all PMR patients. In this approach, the chance of missing GCA is small, but the frequency of positiv e biopsies is v ery low. 53 The more accepted strategy is to biopsy only those patients who hav e sy mptoms suggestive of GCA. With this approach the results are likely to v ary according to the expertise of the ex amining physician. 54 There have been some attempts to develop guidelines for performing temporal artery biopsy in patients with PMR. 55,56 Patients with PMR who were younger than 7 0 years  At least three of the five parameters must be present, which yields a sensitivity of 93% and a specificity of 91%, in relation to controls with other vasculitides.
of age, and with no features of cranial v asculitis (such as a new headache, jaw claudication, abnormalities of the temporal arteries on examination, or amaurosis fugax), were unlikely to have a positiv e biopsy . Older patients with cranial vasculitis findings were more likely to hav e a positiv e biopsy . Sev ere degrees of anemia and thrombocy tosis are suggestive of GCA in patients presenting with PMR sy mptoms. Also, poor clinical response to low-dose prednisone (1 5-20 mg/day) and persistent abnormalities in laboratory parameters of inflammation are also suggestiv e of GCA in patients presenting with "isolated" PMR. In such cases ultrasonography and biopsy of the temporal arteries should be performed to rule out GCA.

T HERAPY , DISEASE COURSE, AND PROGNOSIS
Glucocorticoids are the mainstay of treatment for both PMR and GCA. 57 , 58 In PMR, the starting recommended dose is 1 5 -20 mg/day . Sy mptoms ty pically begin to abate within 1 -3 day s, but a few patients may still hav e some pain and stiffness sev eral weeks after initiation of treatment. After 2-4 weeks, following improv ement of the clinical features of the disease together with normalization of the inflammatory markers, the dose of glucocorticoids can be tapered gradually. Relapses occur in about one-half of the patients, with response to increasing the dose for several weeks. Tapering of the dose should then be resumed.
The duration of treatment for PMR v aries from 6 months to several years, and the majority of patients dev elop steroid-related side effects. Methotrex ate was ev aluated as a steroid-sparing agent with mixed results. Addition of methotrexate can be considered in relapsing disease, in patients who are regarded at high risk for dev eloping adv erse ev ents, or in patients ex periencing steroid-related adverse effects. 59 Poly myalgia rheumatica is thought to have a benign course, with a v ariable degree of treatment-related morbidity .
In GCA, the initial dose is 40 -60 mg/day for most cases. 58 Patients with v ascular complications such as v ision loss (transient or permanent), diplopia, transient ischemic attacks, or stroke may be treated initially with higher doses such as 500 -1 ,000 mg/day of intravenous methylprednisolone for three consecutive days, in an attempt to prevent additional ischemic complications.
Rapid improvement of all clinical manifestations following treatment initiation is characteristic. Prompt treatment is crucial in GCA to prev ent irrev ersible ischemic complications, such as acute v ision loss or stroke. Thus therapy should not be delay ed pending temporal artery biopsy , which should be performed as soon as possible.

Clinical Criteria
M orning stiffness duration >45 min 2 Hip pain or limited range of motion 1 Absence of rheumatoid factor or anti-citrullinated peptide antibodies 2 Absence of other joint pain 1

Ultrasound Criteria
At least one shoulder with subdeltoid bursitis and/or biceps tenosynovitis and/or glenohumeral synovitis (either posterior or axillary), and at least one hip with synovitis and/or trochanteric bursitis 1 Both shoulders with subdeltoid bursitis, biceps tenosynovitis, or glenohumeral synovitis 1 Required criteria for all cases: (1) age 50 or older; (2) bilateral shoulder pain; (3) elevated ESR and/or CRP.
With only clinical criteria, a score of >4 is required. With combined clinical and ultrasound criteria, a score of >5 is required. The required score has 68% sensitivity and 78% specificity for discriminating PM R from comparison patients. The positive predictive value is 69%, and the negative predictive value 77%.
After 2-4 weeks, following improvement of the clinical features of the disease together with normalization of ESR and CRP, the dose of glucocorticoids can be tapered, with close monitoring for recurrence of sy mptoms. During follow-up, lev els of ESR and CRP do not alway s correlate with disease activ ity . Elev ation of these acute-phase reactants while the patient is asy mptomatic is not an absolute indication to increase the dose of prednisone. I n such cases, it is preferable to slow the rate of dosage tapering and continue to watch closely for recurrence of sy mptoms. On the other hand, the dose should be increased if sy mptoms recur, even when the ESR or CRP remain within the normal range. The av erage duration of treatment is 2 -3 y ears. Relapses are ex perienced by about half of GCA patients. Most relapses are mild, but some patients may still dev elop v ision loss or stroke during the course of corticosteroid treatment or after discontinuation of therapy. Addition of low-dose aspirin (100 mg/d) has been shown to decrease the rate of v ision loss and stroke during the course of the disease, probably mediated by its anti-platelet effect. 60,61 Some GCA patients with thoracic or peripheral v ascular inv olv ement may benefit from surgical interv entions, such as balloon angioplasty or stents for v ascular stenosis, and prostheses for aortic aneury sms. Aortic complications may occur early or late in the course of GCA, sometimes after the completion of drug treatment. In most cases, aortic structural damage was not associated with persistence of detectable disease activity. 62 Indications for regular imaging of the aorta in GCA hav e not y et been determined. 63 Indiv idual cases vary greatly, therefore the exact doses and the duration of treatment should be adjusted to the needs of the individual patient, considering both disease manifestations and glucocorticoid adv erse effects. Patients with strong initial sy stemic inflammatory response tend to hav e a prolonged disease course with more flares, requiring higher cumulativ e steroid doses. 64 No steroidsparing agent has been proven to be highly effective thus far, but a few reports suggest some beneficial effects of methotrexate and cyclophosphamide. 65,66 These medications may be prescribed in certain situations, such as resistant cases, or in cases at high risk of glucocorticoid-related side effects. Recently, sev eral case studies reported beneficial effects of tocilizumab, an IL-6 receptor antagonist. A randomized prospectiv e study of 30 patients reported recently that patients given a combination of tocili-zumab and prednisolone were able to discontinue prednisolone 1 2 weeks earlier than patients giv en prednisolone only . 67 It is apparent that GCA may result in fatal complications. The major causes of mortality hav e been v ascular: stroke, coronary artery ev ents, rupture of thoracic aortic aneury sms, and aortic dissection. In addition, there is increased occurrence of sev ere infections. 68 The impact on the ov erall prognosis is controversial. 2,3,68-70 Some epidemiologic studies reported increased mortality, mostly during the first y ear after diagnosis, while others found that the ov erall life ex pectancy among GCA patients was essentially identical to that of the general population.

CONCLUSIONS
Giant cell arteritis is considered to be the most common type of primary v aculitis in the elderly. It may cause severe morbidity such as acute loss of v ision and stroke, but prompt diagnosis and proper treatment would prevent such ischemic episodes in most cases. It is unknown whether PMR is just an ex pression of an underly ing GCA. More likely , it seems that both are a result of an unknown causative factor (or factors), sometimes expressed as PMR, sometimes as GCA, and sometimes as a combination. Clinically , the two conditions may present together but may sometimes be separated by long interv als, and either one may present first. Treatment-related morbidity is a concern in both conditions, but several promising therapeutic agents are currently being studied.