Diagnosis of giant cell arteritis

INTRODUCTION — Giant cell arteritis (GCA, also known as Horton disease, cranial arteritis, and temporal arteritis) is the most common of the systemic vasculitides [1]. GCA is also a classic systemic rheumatic disease of older adults; it virtually never occurs in individuals younger than 50 years of age and peaks in incidence in the seventh decade [2].

GCA mainly involves the large- and medium-sized arteries and may produce a wide spectrum of clinical manifestations [3]. At one end of the spectrum, there are patients with cranial manifestations who typically have a positive temporal artery biopsy. At the other end are patients with predominantly large vessel involvement, who frequently lack cranial manifestations, and whose symptoms are often subtle and nonspecific. These patients typically have a negative temporal artery biopsy and require large vessel imaging to confirm the diagnosis. Finally, some patients have overlapping features of cranial manifestations and large vessel disease [4,5].

The diagnosis of GCA will be reviewed here. The pathogenesis, clinical manifestations, and treatment of this disorder are discussed separately. (See "Pathogenesis of giant cell arteritis" and "Clinical manifestations of giant cell arteritis" and "Treatment of giant cell arteritis".)

INITIAL DIAGNOSTIC EVALUATION

When to suspect giant cell arteritis — The diagnosis of giant cell arteritis (GCA) should be considered in a patient over the age of 50 years who has one or more of the following symptoms or signs, particularly in the setting of an elevated erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP) (algorithm 1):

●New headache or change in characteristics of preexisting headache

●Abrupt onset of visual disturbances, especially transient/permanent monocular visual loss

●Jaw claudication

●Unexplained fever or other constitutional symptoms and signs

●Signs/symptoms of vascular abnormalities (eg, limb claudication; asymmetric blood pressures; abnormal radial pulse; vascular bruits; temporal artery abnormalities such as tenderness to palpation, decreased pulse amplitude, and presence of nodules)

A current or prior diagnosis of polymyalgia rheumatica (PMR) heightens the potential diagnostic significance of any of these findings. Because of the association between the two diseases, a diagnosis of GCA should be considered in all patients with PMR. In patients who manifest only symptoms of PMR without symptoms and signs of cranial or large vessel disease, temporal artery biopsy and large vessel imaging are not required. Nevertheless, such patients should be questioned carefully about headache, jaw or arm claudication, visual symptoms, and unusual pain in the face, throat, or tongue, any of which could suggest the possibility of GCA. A careful clinical evaluation of large vessels should include carotid and subclavian artery auscultation for the presence of bruits and bilateral blood pressure measurements. An evaluation for GCA may be indicated in selected cases of PMR that incur repeated relapses, have persistently increased acute phase reactants, and appear relatively resistant to glucocorticoids. (See "Clinical manifestations and diagnosis of polymyalgia rheumatica", section on 'Association with GCA'.)

The manifestations of GCA are protean, are variable from patient to patient, and can be transient or fluctuating. Accordingly, if the disease is suspected, both current and recent symptoms compatible with GCA should be sought. A more detailed description of the clinical manifestations of GCA is presented separately. (See "Clinical manifestations of giant cell arteritis".)

The presence of clinical manifestations unusual in GCA should prompt consideration of alternative diagnoses (eg, lymphoma or other malignancy; other vasculitides, such as granulomatosis with polyangiitis; atherosclerosis). Some of these presentations include (see 'Differential diagnosis' below):

●Adenopathy

●Pulmonary infiltrates

●Digital cyanosis, ulceration, or gangrene

●Mononeuritis multiplex

●Stroke in the distribution of the middle cerebral artery

●Evidence for glomerulitis or and/or rapidly rising serum creatinine

Evaluation — Fundamental to the diagnosis of GCA is the demonstration of typical histopathologic or imaging findings. The diagnosis should not be based upon symptoms alone (algorithm 1). The treatment of GCA with high-dose glucocorticoids can be associated with considerable potential toxicities. The commitment of an older adult to this therapy should be founded, whenever possible, on unequivocal confirmation of the diagnosis. Histopathologic evidence of GCA is most often acquired by temporal artery biopsy. It may be possible for color Doppler ultrasound (CDUS), if performed by experienced operators, to substitute for temporal artery biopsy as a diagnostic procedure in the appropriate clinical setting. (See 'Temporal artery biopsy' below and 'Ultrasound with Doppler' below.)

Though the most common presentation of GCA involves the cranial arteries, in some patients a different disease phenotype predominates in which the arteritis targets the large vessels. In these cases, imaging of the large vessels is essential for diagnosis. (See 'Patients with a nondiagnostic initial evaluation' below.)

Our diagnostic approach is generally consistent with guidelines developed by professional organizations [6-8].

Laboratory testing — Initial laboratory testing should include the following:

●Complete blood count (CBC)

●ESR and CRP levels

●Complete metabolic panel (to include serum creatinine, liver function tests, glucose)

●Urine dipstick analysis

●Serum protein electrophoresis

●Bone profile panel (to include calcium, phosphorous, albumin, total protein, alkaline phosphatase, 25-hydroxyvitamin D)

In selected patients with clinical manifestations unusual in GCA, additional tests such as chest radiograph may be helpful in excluding alternative diagnoses. (See 'Differential diagnosis' below.)

Laboratory data can aid in the evaluation of GCA and its differential diagnosis, but they are not specific and cannot be relied on as definitive evidence for or against a possible diagnosis of GCA.

Although the ESR and/or CRP are almost always high in GCA, they are not specific for GCA. Their role is more important to rule out GCA when they are normal rather than to confirm GCA when they are elevated. Only 4 percent of patients with GCA have both the ESR and CRP in the normal range [9]. Even if the CRP is more sensitive than the ESR, an elevated ESR with a normal CRP can also infrequently be observed [10]. Thus, we suggest performing both the ESR and CRP at the initial evaluation. (See "Clinical manifestations of giant cell arteritis", section on 'Erythrocyte sedimentation rate and C-reactive protein'.)

Other laboratory indices that reflect systemic inflammation include a normochromic normocytic anemia, thrombocytosis, as well as a reduced albumin and elevated alpha2 fraction on serum protein electrophoresis and a raised fibrinogen. Liver function tests can be abnormal in one-third to one-fourth of patients. In such cases, it is usually the alkaline phosphatase that is (often mildly) elevated. (See "Clinical manifestations of giant cell arteritis", section on 'Other laboratory abnormalities'.)

Assessment of the temporal artery — A suspected diagnosis of GCA should be confirmed by temporal artery biopsy or temporal artery CDUS. (See 'Ultrasound with Doppler' below.)

However, scheduling the temporal artery biopsy or CDUS should not delay initiation of treatment, and the appropriate dose of glucocorticoids should be started when GCA is suspected. (See "Treatment of giant cell arteritis", section on 'Glucocorticoids'.)

Temporal artery biopsy — In clinical practice, we recommend an initial unilateral temporal artery biopsy with an in vivo length of at least 1 cm. The diagnostic sensitivity of a unilateral temporal artery of adequate length that has been subjected to careful histologic evaluation is acceptably high (see 'Diagnostic accuracy' below). Proceeding with the contralateral biopsy is appropriate if the unilateral biopsy and imaging for temporal arteritis and large vessel vasculitis are negative, the clinical suspicion is high, and additional evidence for cranial GCA is sought.

We start glucocorticoid treatment as soon as the diagnosis of GCA is suspected, and we perform temporal artery biopsy within two to four weeks of treatment. Beyond four weeks of high-dose glucocorticoid exposure, we perform temporal artery biopsy only in selected cases that require diagnostic confirmation of GCA. The dosing of glucocorticoids for the treatment of GCA is discussed separately. (See 'Diagnostic accuracy' below and "Treatment of giant cell arteritis", section on 'No visual loss at diagnosis'.)

Temporal artery biopsy procedure — Temporal artery biopsies are obtained on an outpatient basis under local anesthesia in a surgical procedure suite or operating room. The procedure is brief and of low risk. The technique of temporal artery biopsy is discussed in detail elsewhere. (See "Temporal artery biopsy technique".)

Diagnostic accuracy — The temporal artery biopsy is not a perfect test. The sensitivity of the temporal artery biopsy is difficult to estimate because there is no definitive diagnostic test for GCA. False negative results occur in up to 44 percent of patients with an established diagnosis of GCA [11,12]. In addition, several factors can affect the diagnostic sensitivity of temporal artery biopsy including the discontinuous character of the histopathologic changes (skip lesions), bilaterality of the biopsy procedure, length of specimens, and number of sections evaluated [13,14]. Glucocorticoid treatment may also reduce the sensitivity of temporal artery biopsy. Finally, the temporal arteries may be spared by the inflammatory process, especially in patients with large vessel involvement, in whom temporal artery biopsy findings are negative in a substantial number of cases [4]. (See 'Evaluation for large vessel giant cell arteritis' below.)

●Diagnostic sensitivity – The reported sensitivity of the temporal artery biopsy in most studies ranges from 50 to 95 percent, depending on the prevalence of GCA in the evaluated population; the clinical threshold for considering the procedure; and differences in temporal artery biopsy sampling, processing, or interpretation. In a large meta-analysis including 3092 patients, the diagnostic sensitivity of the temporal artery biopsy was found to be 77.3 percent (95% CI 71.8-81.9 percent) [15], providing indirect evidence that temporal artery biopsy is not less sensitive than temporal artery imaging.

●Usefulness of bilateral biopsies – The extent to which the sensitivity of the temporal artery biopsy is improved by the performance of bilateral biopsies remains a matter of debate. The range of reported discordance rates of bilateral temporal artery biopsies (positive on one side, negative on the other) has varied widely [16-20]. A study of 1935 bilateral temporal artery biopsies, performed either sequentially or simultaneously, collected from 11 older reports, found a mean discordance rate (calculated from the total numbers of positive and negative biopsies) of 5.5 percent (95% CI 3.2-7.8) [21]. In 2 studies of simultaneously performed biopsies of 132 and 250 patients, discordance rates of 9.8 and 4.4 percent, respectively, were reported [22,23]. The increased yield of contralateral biopsy for the diagnosis is thus in the range of 5 percent.

●Effect of biopsy length – The effect of biopsy length on the diagnostic yield of temporal artery biopsy has also been debated. Because of the issue of skip areas, longer biopsy specimens have traditionally been recommended, taking into account that there is a shrinkage rate of at least 10 percent in the length of the temporal artery sample after formalin fixation [24]. One retrospective study of 966 biopsies found that positive biopsies had a median postfixation length that was statistically longer than negative biopsies (1.2 versus 1.0 cm) [25]. Another study of 1520 temporal artery biopsies concluded that specimens of shorter lengths were also of diagnostic value [26]. In that study, statistical analysis suggested that the change point in the odds ratio for finding a positive temporal artery biopsy occurred at a biopsy length of 0.5 cm [26]. Similar results were observed in a retrospective study of 694 temporal artery biopsies [14]. A postfixation temporal artery biopsy specimen length of 5 mm was sufficient to make a histologic diagnosis of GCA.

Because the arterial specimen contracts after removal, surgeons should aim for a prefixation surgical specimen length of 7 to 10 mm in order to maximize the diagnostic accuracy of temporal artery biopsy. In order not to miss inflammatory changes, at least three further sections at deeper levels should be evaluated in all negative temporal artery biopsy specimens [14].

●Effect of glucocorticoids – Resolution of the inflammatory infiltrate of GCA occurs slowly, and histopathologic evidence will be evident for at least a month after glucocorticoid therapy has been instituted. In one series, the rates of positive biopsies were similar in patients who were treated with glucocorticoid therapy for less than two weeks compared with those treated for more than two weeks [27]. Another study observed a drop-off in the yield of a positive temporal artery with continued treatment: a positive biopsy was found in 78 percent of patients with suspected GCA who had less than two weeks of treatment; in 65 percent with treatment of two to four weeks' duration; and in 40 percent treated for more than four weeks [28]. Another large series, however, reported that characteristic histopathologic changes of GCA were seen as frequently in biopsies obtained after four weeks of glucocorticoid therapy as after one to three days of therapy [29]. In some patients, the temporal artery biopsy can remain diagnostic for weeks and months after the start of treatment [29]. Another longitudinal histopathologic study reported that vasculitis may still be demonstrated on repeated biopsies in 70 percent at 3 months, 75 percent at 6 months, and 44 percent at 9 and 12 months [30]. (See "Treatment of giant cell arteritis", section on 'Initial management'.)

In clinical practice, temporal artery biopsies performed for evaluation of patients with suspected GCA are positive in 25 to 35 percent of cases [14].

CDUS-guided temporal artery biopsy does not increase the positive yield of temporal artery biopsy [31] but is useful for locating the artery in preparation for the biopsy procedure.

Interpretation of histopathologic findings — The classic histologic picture of GCA is a transmural inflammatory infiltrate comprised of lymphocytes, macrophages, and, in approximately 75 percent of cases, giant cells (picture 1 and picture 2). The lesion frequently has a "concentric rings" appearance, with a thicker inflammatory band surrounding the external elastic lamina and a thinner inflammatory band along the internal elastic lamina (picture 3). Generally, the two bands are focally interconnected, but the interposed media is relatively spared, and only in the most severe cases is it variably destroyed by inflammation (picture 4). A peculiar laminar necrosis, consisting of a band of acellular eosinophilic material sometimes bordered by palisading histiocytes along the internal elastic lamina, is present in approximately 25 percent of cases (picture 5). Fibrinoid necrosis is extremely rare, and its presence should prompt consideration for the possibility of an alternative diagnosis (ie, one of the systemic necrotizing vasculitides such as granulomatosis with polyangiitis) (picture 6) [32-34].

In some patients, the inflammation spares the media and is restricted to the periadventitial small vessels (picture 7), to the adventitial vasa vasorum (picture 8), or to the adventitia (picture 9) [32]. Whether these forms of restricted inflammation carry the same diagnostic and prognostic importance as the typical transmural findings is still unclear. Because a temporal artery biopsy showing restricted inflammation has low sensitivity and specificity for GCA [35], the fact that a substantial proportion of patients may have this pattern of limited evidence of inflammation emphasizes the need for diagnosis and treatment based on clinical judgment [36].

In the absence of mural inflammation, the other structural changes (such as focal mediointimal scar, medial attenuation, intimal hyperplasia, fragmentation of inner elastic lamina, calcification, adventitial fibrosis, and neoangiogenesis) found in the wall of temporal arteries, and often referred to as "healed" or quiescent arteritis, are not specific or useful for the diagnosis of GCA (picture 10) [13]. We suggest that the terms "healed" or "quiescent arteritis" should best be avoided, and the definition of a positive temporal artery biopsy be restricted to the biopsies showing inflammation.

Rarely, temporal artery biopsy will disclose pathology other than GCA, such as systemic necrotizing vasculitis, antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis, polyarteritis nodosa, amyloidosis, or lymphoma [37,38].

Ultrasound with Doppler — CDUS of the head, neck, and upper extremities can serve as a diagnostic surrogate for temporal artery biopsy when performed by clinicians skilled in this technique. In the absence of extensive experience with this technology in routine clinical practice, however, temporal artery biopsy remains an essential diagnostic measure for the evaluation of suspected GCA. (See 'Temporal artery biopsy' above.)

We start glucocorticoids when the diagnosis of GCA is suspected. Temporal artery CDUS should be performed when the diagnosis is suspected, before or as early as possible after initiation of therapy, ideally within one week, because treatment with glucocorticoids rapidly reduces the sensitivity of CDUS for GCA diagnosis [39]. The dosing of glucocorticoids for the treatment of GCA is discussed separately. (See "Treatment of giant cell arteritis", section on 'No visual loss at diagnosis'.)

CDUS has a resolution of 0.1 mm and thus can visualize the temporal arteries and other small extracranial arteries. In GCA, the temporal arteries show a circumferential dark area about the vascular lumen, thought to represent mural edema, dubbed the "halo sign" [40,41]. The halo, which measures 0.3 to 2.0 mm, is hypoechogenic, not anechogenic (image 1). The presence of bilateral halo signs of the temporal arteries is highly specific for GCA [42]. The "compression sign" also has high specificity for the diagnosis; it refers to the persisting visibility of the halo during compression of the vessel lumen by the ultrasound probe [43]. Stenoses and occlusions can also be seen but are much less specific and thus of limited diagnostic value.

CDUS has several advantages: It is noninvasive, provides simultaneous image acquisition and interpretation, and involves no exposure to radiation. Other cranial arteries (eg, the facial, occipital, and vertebral arteries) can also be visualized. Scanning of the axillary and subclavian arteries can show evidence of large vessel vasculitis, which is missed with diagnostic studies limited to the cranial arteries, thereby augmenting the potential yield of CDUS as a diagnostic test (image 2) [44-46]. (See 'Patients with a nondiagnostic initial evaluation' below.)

Published studies of the sensitivity and specificity of CDUS for the diagnosis of GCA reveal a range of results. A 2010 meta-analysis of 285 patients studied with both CDUS and temporal artery biopsy found a weighted sensitivity of the halo sign compared with the gold standard of a positive temporal artery (an imperfect standard itself, as discussed above) of 68 percent (95% CI 0.61-0.74) and a specificity of 91 percent (95% CI 0.88-0.94) [42]. Similar results were obtained in a 2005 meta-analysis of earlier data, in which the weighted sensitivity and specificity of the halo sign compared with biopsy were 69 percent (95% CI 0.57-0.79) and 82 percent (95% CI 0.75-0.87), respectively [47]. A 2016 review of 10 studies, comprising 696 patients, found specificity ranged from 78 to 100 percent but a wider extent of sensitivity, from 55 to 100 percent [48]. Other reports have consistently described higher figures for sensitivity and specificity; in these centers, CDUS has superseded temporal artery biopsy as the initial diagnostic procedure of choice [49,50].

The heterogeneous conclusions about the utility of CDUS likely reflect the operator-dependent nature of the procedure and may also result from variability of equipment, probe settings, technique, and clinical context. Standardization of these factors may lead to more widespread use of CDUS for the diagnosis of GCA.

Other tests with limited roles

Other temporal artery imaging — High-resolution magnetic resonance imaging (MRI) with MR angiography (MRA) can visualize the temporal arteries and, when used with contrast, demonstrate mural edema [51]. A systematic review and meta-analysis comparing MRI with a clinical diagnosis of cranial GCA reported a pooled sensitivity and specificity of 73 and 88 percent, respectively [52]. When temporal artery biopsy was used as the reference standard, MRI yielded a pooled sensitivity and specificity of 93 and 81 percent, respectively. Sensitivity decreases after more than five days of glucocorticoid treatment [53]. Though promising, widespread adoption of this technology may be constrained by the need for contrast and by considerations of cost.

Positron emission tomography (PET), computed tomography (CT), CT with angiography (CTA), and conventional MRA lack sufficient spatial resolution to permit visualization of the temporal artery. PET scanning of the cranial arteries is also obscured by high fluorodeoxyglucose (FDG) uptake in the brain. However, higher-resolution PET/CT scanners employing time-of-flight (TOF) imaging could have improved detectability, but further study is needed [54].

Biopsy of other arteries — We generally do not biopsy arteries other than the temporal arteries. Although GCA is also referred to as "temporal" arteritis, the temporal arteries are biopsied in this disease not because they are the focus of GCA, but because they are the blood vessels most easily accessible for diagnosis. Arteritic involvement tends to be widespread and can affect multiple extracranial vessels and other branches of the external carotid arteries. On rare occasions, the occipital or facial arteries can be biopsied in the presence of signs of inflammatory involvement. (See "Temporal artery biopsy technique", section on 'Alternative biopsy sites'.)

PATIENTS WITH A NONDIAGNOSTIC INITIAL EVALUATION — Additional testing may be warranted for patients in whom the initial evaluation for cranial arteritis with temporal artery biopsy and/or color Doppler ultrasound (CDUS) is negative, but the clinical suspicion for giant cell arteritis (GCA) remains high (algorithm 1).

Reasons for a negative workup — An initially negative evaluation for suspected GCA may have several explanations:

●The patient has GCA, but temporal artery biopsy or CDUS failed to detect the area of vascular pathology, possibly due to skip areas and the segmental nature of arteritis. (See 'Diagnostic accuracy' above.)

●The patient has GCA but has a disease phenotype not associated with cranial arteritis. The most common form of GCA involves the cranial arteries, but in some cases, the disease is confined to the large vessels; in this phenotype, temporal artery biopsies are more often negative than positive [4]. (See "Clinical manifestations of giant cell arteritis", section on 'Large vessel involvement' and 'Clinical features of large vessel giant cell arteritis' below.)

●The patient does not have GCA.

Evaluation for large vessel giant cell arteritis — When the diagnosis of GCA is still suspected in a patient who has had a negative temporal artery biopsy and/or CDUS, the possibility of large vessel involvement should be considered. The diagnostic procedure of choice for large vessel GCA is an advanced imaging study of the aortic vascular tree.

Large vessel GCA denotes involvement of the aorta and its first-order branches, especially the subclavian arteries. Large vessel involvement in GCA is common and, depending on the methodology used, can be demonstrated in 30 to 80 percent of cases [45,55-58]. Importantly, the temporal artery biopsy is negative in less than one-half of patients with large vessel GCA [4]. Because diagnostic biopsy of the aorta and large arteries is not feasible, imaging is crucial for the diagnosis of the phenotype of large vessel GCA.

Clinical features of large vessel giant cell arteritis — Clues to the possible presence of large vessel GCA include limb claudication, asymmetric blood pressures, abnormal radial pulse, or vascular bruits. These clinical features, however, are relatively uncommon, especially at disease onset, and more often the diagnosis becomes suspect in the presence of constitutional symptoms and signs, especially if occurring in the context of what appears to be glucocorticoid-resistant polymyalgia rheumatica (PMR). The first evidence for large vessel GCA can also arise from advanced imaging exams performed in the course of evaluation for infection or malignancy or, on occasion, from histopathologic review of an aorta resected during aneurysm repair or of other surgical specimens [59]. (See "Clinical manifestations of giant cell arteritis", section on 'Physical examination findings' and "Clinical manifestations of giant cell arteritis", section on 'Large vessel involvement'.)

Compared with cranial GCA, patients with large vessel GCA are younger at diagnosis (68 versus 76 years old) and have fewer cranial symptoms (eg, headaches and jaw claudication) and more symptoms of upper extremity vascular claudication [4]. (See "Clinical manifestations of giant cell arteritis", section on 'Large vessel involvement'.)

Imaging modalities — Imaging modalities for diagnosing large vessel GCA include CT or CT with angiography (CTA), MRI or MR angiography (MRA), and fluorodeoxyglucose (FDG) positron emission tomography (PET) or PET with CT of the aorta and its branches. Large vessel GCA can also be identified by CDUS of the epiaortic vessels (eg, carotid, subclavian, and axillary arteries). Conventional angiography is largely of historic interest.

Both CTA and MRA involve contrast administration and are tailored to evaluate vessels with high-resolution multiplanar image. In the absence of controlled, prospective studies, it is not possible to recommend MRA over CTA or vice versa. The clinician's selection of a given procedure will depend on the availability of local resources (technologic, professional, and fiscal).

Numerous case reports and series, especially those using CT and PET, demonstrate the value of these modalities for identifying large vessel GCA [56,60-62].

CDUS can visualize the axillary, subclavian, and carotid arteries, but because of overlying bony structures, it cannot image the intrathoracic aorta. CDUS has lower sensitivity for the detection of abdominal aorta involvement compared with PET as a reference standard [63]. In this setting, the sensitivity of CDUS is approximately 20 percent lower than that of PET, which is able to visualize all arteries, including the aorta [63]. CT and CTA can demonstrate concentric mural thickening, continuous over long segments, which is a characteristic of large vessel inflammation (image 3). MRI and MRA can detect wall thickening and, on T2 sequences, the presence of mural edema (image 4). Enhanced MRI is superior to T2 sequences for the demonstration of arterial wall edema [64]. CTA and MRA can also identify stenoses, dilatation, and aneurysms. PET detects increased uptake of FDG by metabolically active cells, either malignant or inflammatory, and as such, can reveal sites of large vessel vasculitis. The combination of PET with CT enhances the delineation of vascular anatomy. With or without CT, PET can be useful for the detection of large vessel GCA presenting with fever of unknown origin or with nonlocalizing systemic symptoms and signs (image 5).

There are caveats for all the imaging techniques. The differentiation of mural inflammation from mural fibrosis or remodeling on MRI/MRA, CT/CTA, and PET can be difficult. Results can be affected by concurrent glucocorticoid treatment [65]. The interpretation of findings on PET scanning remains largely qualitative, as standards for quantitative analysis have not been codified. Exposure to ionizing radiation, the need for intravenous contrast, and the economic costs of the various modalities, apart from CDUS, are also considerations.

DIAGNOSIS

Patients with a positive biopsy or imaging — The diagnosis of giant cell arteritis (GCA) should not be based on clinical presentation alone but also requires histopathologic proof or evidence from imaging exams. Temporal artery biopsy is the cardinal diagnostic procedure. The typical histopathologic finding is a non-necrotizing panarteritis.

Skilled operators can use color Doppler ultrasound (CDUS) of the head, neck, and upper extremities as a substitute for the temporal artery biopsy as an initial diagnostic procedure. In active arteritis, a hypoechoic area about the lumen of affected vessels, called the halo sign, can be demonstrated. (See 'Temporal artery biopsy' above and 'Ultrasound with Doppler' above.)

While the typical presentation of GCA involves the cranial arteries, in other patients, the symptoms and signs arise from large vessel involvement, referred to as large vessel GCA. The diagnosis of large vessel GCA is established on the basis of imaging of the aorta and large arteries. (See 'Evaluation for large vessel giant cell arteritis' above.)

Patients with negative biopsy and imaging — Occasionally clinicians are faced with the diagnostic challenge of a negative workup, which should include negative temporal artery biopsy or biopsies and, if indicated, imaging of the large vessels. In this situation, the clinician can choose to either:

●Conclude that the patient does not have GCA and pursue alternative diagnoses, or

●Make a clinical diagnosis of GCA and treat accordingly

Clinical improvement following a brief trial of high-dose glucocorticoids is nonspecific and should not be relied upon for establishing the diagnosis of GCA. Glucocorticoids can ameliorate a variety of symptoms, including malaise, headache, and fevers, as well as lower or even normalize the acute phase reactants. Thus, treatment with high-dose glucocorticoid therapy of individuals with negative temporal artery biopsies and large vessel imaging must be carefully considered and reserved for selected patients with a classic clinical presentation of GCA and no other plausible alternative diagnoses. Such patients are generally treated in the same way as those with positive workups. (See "Treatment of giant cell arteritis".)

Classification criteria — Classification criteria for GCA should not be used to establish a clinical diagnosis [66]. They were created to differentiate GCA from other forms of vasculitis in research studies but lack sufficient sensitivity and specificity to be used in diagnosing individual patients. However, these criteria may be useful to remind the clinician of factors that are relevant to making a diagnosis of GCA.

The most widely used criteria were formulated by the American College of Rheumatology (ACR) in 1990 [66]. These criteria were updated by the 2022 ACR/European Alliance of Associations for Rheumatology (EULAR) classification criteria for giant cell arteritis, which uses a weighted algorithm that includes clinical, laboratory, imaging, and biopsy criteria (table 1). The 2022 ACR/EULAR classification criteria were constructed in part to identify patients who present with disease that predominantly affects the large vessels, such as the aorta and axillary arteries [67]. (See "Overview of and approach to the vasculitides in adults", section on 'Classification criteria'.)

DIFFERENTIAL DIAGNOSIS — Special considerations in the differential diagnosis of giant GCA are discussed below.

●Other systemic vasculitides – Vasculitic disorders other than GCA can present with unexplained fever, anemia, or other constitutional symptoms and signs, as well as elevated acute phase reactants (eg, erythrocyte sedimentation rate [ESR] and C-reactive protein [CRP]). However, most of the other vasculitides have other distinguishing features.

•Takayasu arteritis – The histopathologic and radiographic findings of GCA and Takayasu arteritis can be indistinguishable. These two disorders are differentiated primarily by differences in the age of the affected patients: GCA almost never occurs in individuals younger than 50, while Takayasu arteritis typically begins before the age of 40 (and frequently at much younger ages). In addition, the clinical expressions of the two diseases differ. As examples, renovascular hypertension due to renal artery stenosis does not occur in GCA, and sight loss due to anterior ischemic optic neuropathy and polymyalgia rheumatica (PMR) are unusual in Takayasu arteritis [68]. (See "Clinical features and diagnosis of Takayasu arteritis".)

•Small and medium vessel vasculitides – Certain small and medium vessel vasculitides can mimic GCA through their shared tendency to cause systemic symptoms. There is, however, generally little difficulty differentiating GCA from microscopic polyangiitis, granulomatosis with polyangiitis, and polyarteritis nodosa because of the different vascular distributions of these diseases, their usually distinctive histopathology, and their patterns of organ involvement. In occasional cases, however, these forms of necrotizing vasculitis involve the temporal artery [69]. (See "Granulomatosis with polyangiitis and microscopic polyangiitis: Clinical manifestations and diagnosis" and "Clinical manifestations and diagnosis of polyarteritis nodosa in adults".)

•Primary angiitis of the central nervous system – Stroke is uncommon in GCA, and it is generally caused by the inflammatory involvement of the extracranial tract of the vertebral or internal carotid arteries. Although the histopathology of primary angiitis of the central nervous system can be indistinguishable from that of GCA, cranial ischemic events in GCA are only exceptionally identified as associated with inflammatory involvement of the intracranial vessels [70-73]. (See "Primary angiitis of the central nervous system in adults" and "Clinical manifestations of giant cell arteritis", section on 'Central nervous system involvement'.)

●Idiopathic aortitis – Aortitis of unknown cause can affect the ascending aorta with a histopathology similar to that of GCA and Takayasu arteritis, but without the characteristic clinical stigmata of these diseases. The discovery of aortitis is nearly always unexpected, and usually comes to attention on histopathologic inspection of a surgically resected aneurysm (image 6A-B) or as an incidental finding on an imaging procedure. The histopathology of the resected aneurysm is similar to that of GCA, including the presence of giant cells. Aortitis can also occur as a complication of a variety of infections, sarcoidosis, Cogan syndrome, and other systemic rheumatic diseases, including Takayasu arteritis, relapsing polychondritis, spondyloarthropathy, and Behçet syndrome [74]. Lymphoplasmacytic aortitis characteristic of immunoglobulin G4 (IgG4) disease has also been reported [75]. (See "Clinical features and diagnosis of Takayasu arteritis" and "Clinical manifestations of relapsing polychondritis", section on 'Clinical manifestations' and "Overview of the clinical manifestations and classification of spondyloarthritis", section on 'Major extra-musculoskeletal features' and "Cogan syndrome", section on 'Systemic vasculitis' and "Clinical manifestations and diagnosis of Behçet syndrome", section on 'Vascular disease'.)

In a population-based study from Denmark of 1210 surgical resections involving the thoracic aorta, aortitis was found in 6.1 percent of cases [76]. This result approximates retrospective reports of cases of aortitis from other surgical centers [77-80]. Of the patients with aortitis in the Danish study, a history of "connective tissue disease" was found in 27 percent of patients and was specifically linked to temporal arteritis in 8 percent of patients. A retrospective study of aortitis found at surgery identified a "rheumatologic history" in a similar proportion (20 percent) of patients [80].

Evaluation of cases of unexpected aortitis requires a diligent search for associated disease, especially GCA and Takayasu arteritis, and careful imaging of the entire arterial tree. For patients who unexpectedly come to clinical attention following surgery for an aneurysm, measurement of acute phase reactants immediately postoperation is not a reliable indicator of active disease, as the ESR and CRP are likely to be significantly elevated as a result of the procedure. If the diagnostic workup for cases of unexpected aortitis is unrevealing, the possibility of idiopathic aortitis can be considered [81]. In approximately 20 percent of cases, systemic rheumatic disease can be identified (classic GCA, spondyloarthropathy, Behçet syndrome, and others), for which treatment is then pursued as appropriate [76,80].

Though idiopathic aortitis may remain restricted to the ascending thoracic aorta, vigilant follow-up imaging studies are essential, as new aneurysms or sites of involvement can ensue [82,83]. In these situations, it can be appropriate to withhold treatment (ie, glucocorticoids) as some patients will not manifest subsequent disease activity, but attentive clinical follow-up and repeat imaging studies are essential.

●VEXAS syndrome – Vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome can occur in older males with symptoms that can mimic GCA, especially treatment-refractory disease. Presence of neutrophilic dermatosis and hematologic abnormalities including myeloma, myelodysplastic syndrome, or monoclonal gammopathy, findings of characteristic vacuoles in myeloid precursor cells on bone marrow aspirate, and positive gene testing for UBA1 mutations support the VEXAS diagnosis [84]. (See "Autoinflammatory diseases mediated by NFkB and/or aberrant TNF activity", section on 'Vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome'.)

●Nonarteritic anterior ischemic optic neuropathy – Nonarteritic anterior ischemic optic neuropathy (NAAION) presents with sudden monocular visual loss and, when occurring in an older adult, will prompt concern for underlying GCA [85]. Risk factors include hypertension, diabetes mellitus, and the use of certain drugs such as sildenafil. A critical observation is a small, crowded optic nerve head and small physiologic cup, resulting in a small cup-to-disc ratio. These findings can be difficult to discern if the optic disc is edematous but can be appreciated in the contralateral eye. Other clinical features of GCA and laboratory markers of inflammation, such as an elevated ESR and/or CRP, are generally absent in NAAION. (See "Nonarteritic anterior ischemic optic neuropathy: Clinical features and diagnosis" and "Nonarteritic anterior ischemic optic neuropathy: Epidemiology, pathogenesis, and etiologies" and "Treatment of male sexual dysfunction", section on 'Visual effects'.)

●Infection – The occurrence of fever always mandates consideration for infection. Endocarditis and other infectious diseases can be accompanied by myalgias, arthralgias, headache, and elevations of the ESR and CRP, and thus mimic GCA. Blood cultures should be considered in all such cases.

POSTDIAGNOSTIC IMAGING EVALUATION IN PATIENTS WITH CRANIAL ARTERITIS — In clinical practice, we routinely evaluate for large vessel involvement in all patients with newly diagnosed cranial giant cell arteritis (GCA) by performing color Doppler ultrasound (CDUS) of the epiaortic vessels (eg, carotid, subclavian, and axillary arteries). Although the American College of Rheumatology (ACR) conditionally suggests that such patients be evaluated specifically for aortic involvement with either MR angiography (MRA) or CT with angiography (CTA), this approach may be associated with a higher cost or additional radiation and is of uncertain benefit. To date, there are no long-term large cohort studies that have examined the utility of advanced imaging in predicting outcomes of large vessel disease or assessing response to treatment.

Aortitis is a feature of large vessel involvement in GCA (picture 11). Prospective imaging studies have demonstrated aortitis at the time of initial diagnosis of biopsy-proven GCA in 45 percent of patients evaluated by CT and 65 percent of patients evaluated by CTA and approximately 50 percent of patients evaluated by PET [55,56,86]. Furthermore, patients with intense aortic inflammation at PET/CT appear to be at increased risk of aortic dilatation [87]. (See "Clinical manifestations of giant cell arteritis", section on 'Large vessel involvement'.)

Aortitis can lead to development of aortic aneurysms in GCA. Reports on the incidence of such aneurysms are mixed, in large part because of heterogeneity in methods of data collection and ascertainment. A meta-analysis of nine studies found that thoracic aortic aneurysms were detected in 2 to 8 percent of GCA patients [88].

Data on the frequency of aortic aneurysm in GCA patients compared with the general population are also mixed. A previous population study found patients with GCA to be 17.3 times more likely to develop thoracic aortic aneurysm than an unmatched control population [89]. A large cohort study from the United Kingdom of 6999 GCA patients, as defined by International Classification of Diseases (ICD)-10 codes, and 41,994 controls, matched for age and sex, found a twofold increased relative risk of aortic aneurysm in GCA patients [90]. The latter study identified smoking and male sex as independent predictors for aortic aneurysm.

The rate of actual complications of aortic aneurysms in GCA (ie, dissection and rupture) is also not easily quantified, again owing to methodologic heterogeneity and small numbers of patients, but based on reported series, it lies between 2 and 6 percent [88].

Finally, it is uncertain whether treatment influences the clinical course of diagnosed aortitis, including the development of aneurysms and their outcomes.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Giant cell arteritis and polymyalgia rheumatica".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Vasculitis (Beyond the Basics)" and "Patient education: Polymyalgia rheumatica and giant cell arteritis (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●When to suspect giant cell arteritis – The diagnosis of giant cell arteritis (GCA, also known as Horton disease, cranial arteritis, and temporal arteritis) should be considered in a patient over the age of 50 years who has one or more of the following, particularly in the setting of an elevated erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP) (algorithm 1) (see 'When to suspect giant cell arteritis' above):

•New headache or change in characteristics of preexisting headache

•Abrupt onset of visual disturbances, especially transient/permanent monocular visual loss

•Jaw claudication

•Unexplained fever or other constitutional symptoms and signs

•Signs/symptoms of vascular abnormalities (eg, limb claudication; asymmetric blood pressures; abnormal radial pulse; vascular bruits; temporal artery abnormalities such as tenderness to palpation, decreased pulse amplitude, and presence of nodules)

Any of these findings are of special concern in the context of a current or previous diagnosis of polymyalgia rheumatica (PMR) because of the association between GCA and PMR. (See "Clinical manifestations of giant cell arteritis", section on 'Association with polymyalgia rheumatica'.)

●Initial diagnostic evaluation

•Laboratory data – Laboratory data can aid in the evaluation of GCA and its differential diagnosis, but they are not specific and cannot be relied on as definitive evidence for or against a possible diagnosis of GCA. Initial laboratory testing should include the ESR and CRP levels, which are almost always high in GCA. However, normal acute phase reactants do not exclude the diagnosis of GCA. (See 'Laboratory testing' above.)

•Assessment of the temporal artery – A suspected diagnosis of GCA should be confirmed by temporal artery biopsy or temporal artery color Doppler ultrasound (CDUS). (See 'Temporal artery biopsy' above and 'Ultrasound with Doppler' above.)

Scheduling the temporal artery biopsy or CDUS should not delay initiation of treatment with the appropriate dose of glucocorticoids in a patient with a high likelihood of GCA, since delay can put the patient at risk for complications, particularly sight loss. (See "Treatment of giant cell arteritis" and "Treatment of giant cell arteritis", section on 'No visual loss at diagnosis'.)

●Evaluation for large vessel GCA in patients with a nondiagnostic initial workup – When the diagnosis of GCA is still suspected in a patient who has had a negative temporal artery biopsy and/or CDUS, the possibility of large vessel involvement should be considered. The diagnostic procedure of choice for suspected large vessel GCA is an advanced imaging study of the aorta and/or its branches. CT or CT with angiography (CTA), MRI or MR angiography (MRA), and positron emission tomography (PET) or PET with CT are useful for the identification of large vessel GCA. CDUS of the epiaortic vessels (eg, carotid, subclavian, and axillary arteries) can also be used to diagnose large vessel vasculitis. (See 'Imaging modalities' above.)

●Diagnosis – The diagnosis of GCA requires histopathologic proof or evidence from imaging exams. Occasionally, clinicians are faced with the diagnostic challenge of a negative workup, which should include negative temporal artery biopsy or biopsies and imaging of the large vessels. In this situation, the clinician can choose to either (see 'Diagnosis' above):

•Conclude that the patient does not have GCA and pursue alternative diagnoses, or

•Make a clinical diagnosis of GCA and treat accordingly

Clinical improvement following a brief trial of high-dose glucocorticoids is nonspecific and should not be relied upon for establishing the diagnosis of GCA. Treatment with high-dose glucocorticoid therapy of individuals with negative temporal artery biopsies and large vessel imaging must be carefully considered and reserved for selected patients with a classic clinical presentation of GCA and no other plausible alternative diagnoses. (See 'Patients with negative biopsy and imaging' above.)

●Differential diagnosis – The differential diagnosis of GCA includes other vasculitides (eg, Takayasu arteritis, small- and medium-sized vessel vasculitides, primary angiitis of the central nervous system), nonarteritic anterior ischemic optic neuropathy (NAAION), and infection. (See 'Differential diagnosis' above.)

●Postdiagnostic imaging evaluation in patients with cranial arteritis – We routinely evaluate for large vessel involvement in all patients with newly diagnosed cranial GCA by performing CDUS of the epiaortic vessels (eg, carotid, subclavian, and axillary arteries). MRA or CTA is conditionally suggested by the American College of Rheumatology (ACR) for evaluating for aortic involvement. (See 'Postdiagnostic imaging evaluation in patients with cranial arteritis' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Gene Hunder, MD, and William P Docken, MD, who contributed to earlier versions of this topic review.