Establishing the diagnosis of giant cell arteritis (GCA) remains challenging. Although a positive temporal artery biopsy (TAB) is the only gold standard for the diagnosis of GCA, it is sometimes negative, and many clinicians choose not to always obtain a TAB in patients with suspected GCA (1,2). A number of alternative methods for the diagnosis of GCA have been suggested, including diagnostic algorithms, and noninvasive vascular imaging, such as color Doppler ultrasound (CDU) (1). This point–counter point reviews the evidence regarding the use of CDU for the diagnosis of GCA.
Point: Color Doppler Ultrasound is a Suitable Examination Method in the Management of Patients With Suspected Giant Cell Arteritis: Klara Landau, MD, and Philipp Gruber, MD
Since the seminal work of Schmidt et al in 1997 (3), many studies have examined the role of CDU in patients with suspected GCA. Controversy still exists on this topic. Here, we support the view that in most clinical scenarios, CDU is a sufficient and effective method in establishing or ruling out the diagnosis with important advantages over TAB.
Need for the correct diagnosis of GCA
GCA is a chronic systemic immune-mediated vasculitis with frequent involvement of the temporal arteries and other cranial arteries, including the vertebral, ophthalmic, and short posterior ciliary arteries. The disease most commonly affects Caucasians older than 60 years and its incidence is rising (4). Prompt diagnosis and initiation of treatment are crucial since otherwise the dreaded complication of irreversible visual loss may occur. Up to 50% of patients present with ocular symptoms. Although vision loss is the most common, other manifestations include transient visual loss and diplopia (5). If untreated, permanent visual impairment usually develops in the fellow eye within days to weeks, possibly leading to complete blindness (6,7).
Immediate and long-term use of systemic corticosteroids represents the only treatment modality that can effectively suppress the inflammatory disease process albeit side effects are inevitable. These include depression, diabetes mellitus, osteoporosis, infection, arterial hypertension, and gastrointestinal hemorrhage. Up to 90% of patients on corticosteroids will develop clinically significant side effects within 10 years of treatment (8,9).
According to the American College of Rheumatology (ACR), diagnosis of CGA can be made in a patient older than 50 years with newly acquired acute headache, painful and pulseless temporal artery, erythrocyte sedimentation rate more than 50 mm/h, and positive TAB. The diagnosis is established when 3 or more of these 5 criteria are present with a sensitivity of 93.5% and specificity of 91.2% (10). It needs to be pointed out that from neuro-ophthalmological point of view, the current ACR criteria may provide suitable inclusion criteria for clinical trials but not for managing individual patients with suspected GCA (2,11).
Although the diagnosis of GCA primarily is based on clinical criteria, TAB generally is performed to firmly establish the diagnosis and is still considered the diagnostic gold standard with a specificity of 87%–90%. It was Hall et al (12) who showed in the early 1980s that TAB is diagnostically useful. Despite the relative safety of TAB, complications include hematoma, wound dehiscence, brow ptosis (13), and facial nerve paresis (14). While the risk of more serious complications such as blindness has been discussed (15), no instance of visual loss has ever been recorded following TAB.
TAB may appear histologically normal even with obvious clinical signs (16). A frequently cited cause is the segmental involvement of the artery, referred to as “skip areas”(17). Techniques, such as contralateral biopsy, bilateral biopsy, or biopsy guided by ultrasound, have not improved the diagnostic yield of this procedure (18–21). Thus, there is a need for a diagnostic technique with a higher sensitivity than TAB and good specificity that will quickly and reliably diagnose or rule out GCA.
CDU as a diagnostic tool in patients with suspected GCA
CDU is a noninvasive method both to create images of the vessel wall and its lumen and to determine the flow characteristics within the vessel. By means of the B-mode and suitable linear array transducer of 10–15 MHz, high resolution of anatomic structure and a good delineation of blood vessels can be achieved. Combined with pulsed-wave Doppler, it is efficient in assessing the blood flow (22).
The classic sign that indicates the presence of GCA using ultrasound is termed the “halo sign,” a distinct hypoechogenic perivascular structure, with a diameter of 0.32 mm (Fig. 1). It was first described by Schmidt et al in 1995 (23). This appearance may represent perifocal arterial wall edema possibly due to increased vascular permeability caused by inflammation. Ultrasound detects the vascular wall swelling, whereas the histological findings represent inflammatory cell infiltration and granuloma formation (1). In addition, further ultrasonographic diagnostic criteria have been defined, such as stenosis and complete occlusion of the small arterial branches (3). Although this occurs in only 30% of patients, it seems to be very specific (24). Moreover, CDU allows examination of the entire length of the temporal artery and can be performed bilaterally, whereas biopsy is used to sample only one segment of the vessel.
Almost 20 years ago, Puechal et al (25,26) showed that the use of Doppler ultrasound in temporal, facial and ophthalmic arteries has a sensitivity of 77% and a specificity of 80% for the detection of GCA. In the ground-breaking study of Schmidt et al (3) in 1997, it was shown that CDU can be used as an effective technique to establish the diagnosis of GCA using the halo sign. In their prospective, controlled study of 30 patients with GCA, 37 patients with polymyalgia rheumatica, 15 patients with other disorders and 30 control subjects, Schmidt et al (3) obtained a sensitivity of 73% and specificity of 100% compared with the clinical presentation and a sensitivity of 76% and specificity of 92% compared with TAB. Other vascular features such as stenosis or occlusion of temporal arterial branches showed similar sensitivity and specificity. Combining all ultrasound features (halo sign, stenosis, and occlusion) improved the sensitivity markedly (73% to 93% vs clinical presentation; 76% to 95% vs TAB) while only mildly decreasing the specificity (100% to 93% vs clinical presentation; 92% to 85% vs TAB) (3).
Subsequent reports have yielded similar results. The prospective study of Romera-Villegas et al (27) included 68 patients and established a sensitivity of 95.4% and specificity of 91.3% (27). In a smaller study of 26 patients, Murgatroyd et al (28) found a sensitivity of 86% and a specificity of 68% compared with TAB. Another prospective study of 86 patients revealed a sensitivity of only 40% and a specificity of 79% compared with biopsy (29). LeSar et al (30) found in a prospective study of 32 patients, a sensitivity of 85.7% and a specificity of 92.0% compared with histological findings. A retrospective study of 85 patients, which compared CDU with histopathology, showed a sensitivity of 44.4% with a specificity of 90% (31). Habib et al (32) compared the diagnostic accuracy of the halo sign with clinical findings and TAB and calculated a sensitivity of 81% and specificity of 88%. Karahaliou et al (33) used CDU as the sole diagnostic method in suspected cases of GCA and showed that a bilateral presence of the halo sign increases the specificity to 100%. They concluded when the halo sign is bilaterally present, TAB is not necessary.
CDU-guided TAB increased the diagnostic yield compared with TAB alone. Alberts and Mosen (34) conducted a retrospective study of 290 patients suspected of having GCA. Comparing medical decision and outcome after CDU-negative vs TAB-negative results, they concluded that bilateral CDU is comparable with TAB and TAB should only be reserved for cases in which CDU results were equivocal. Pfenninger et al (31) published a retrospective study of 85 patients with suspected GCA and also concluded that given the high positive predictive value of CDU, temporal artery biopsy should be performed in CDU-negative cases. In a prospective study of 182 patients with potential GCA, Stammler et al (35) studied the value of CDU in relation to clinical pretest probability. In the group with high clinical pretest probability, the halo sign was found in 83% of patients. They concluded that in patients with high positive clinical pretest probability and negative halo sign, temporal artery biopsy is not needed. In only 33% of their patients was a biopsy performed to establish or exclude the diagnosis. In a prospective study by Nesher et al (36) of 69 patients suspected of having GCA, sensitivity of 86% and specificity of 78% was found for CDU compared with clinical findings. A high negative predictive value was established, whereas the positive predictive value was low. They concluded that TAB could be omitted only in patients with negative CDU results.
Meta-analyses of studies on the value of CDU in GCA
The first-ever meta-analysis involving 2,036 patients from 23 studies concluded that ultrasound with the halo sign is an accurate diagnostic test for GCA with an overall sensitivity of 69% and overall specificity of 82% compared with TAB (37). Two recently conducted meta-analyses found similar results. One of them included 17 studies with a total of 998 patients and documented an overall sensitivity of 73% and a specificity of 83% (38). The other, comprised 8 studies with a total of 575 patients, found a sensitivity of 75% and a specificity of 98% (39). This study confirmed that with a positive halo sign bilaterally, the specificity rose to 100%.
Further advantages of CDU
Beyond its use as a diagnostic tool, CDU can also be used to monitor treatment. Schmidt et al (3) demonstrated that the halo sign remained in the same location on follow-up examination and resolved after 10 to 16 days of corticosteroid administration. The range for this to occur was 7–56 days. Similar observations were made in later studies in which the halo sign disappeared with steroid therapy in 2–4 weeks, correlating with clinical and laboratory findings (32). A retrospective study, comparing the sensitivity of magnetic resonance imaging (MRI) and CDU in GCA patients on corticosteroids, demonstrated that the sensitivity, both of the halo sign and the specific MRI findings, drops significantly within days of treatment (40). In single case reports, when corticosteroid therapy was tapered and the patient experienced a relapse of GCA, the halo sign was shown to reappear (33,40).
The halo sign also can be found in other cranial arteries that are affected by GCA. A prospective study of 78 patients that compared the halo sign with clinical presentation and with biopsy in occipital arteries revealed a sensitivity of 65% and specificity of 100% compared with the histological findings (mostly taken from the temporal artery) (41). The halo sign was also described by Garcia-Garcia et al (42) in vertebral arteries. Their prospective study included 1,237 patients with stroke and demonstrated that the halo sign in vertebral arteries assisted in the diagnosis of GCA as a cause of stroke.
The possibility of examining several different arteries in a single session represents another major advantage of CDU over temporal artery biopsy. Moreover, the result of a CDU examination is available immediately without a need to wait for several days until a histological diagnosis is provided by the pathologist.
Note of caution
Admittedly, CDU is a sensitive and very specific test, but it remains highly examiner dependent. This was addressed by Maldini et al (43) in a retrospective study of 77 patients seen over 7 years in a single institution in which a sensitivity of 17% was found, whereas specificity was comparable with previous studies. The low sensitivity was probably due to interoperator variability because the detection of the crucial halo sign is dependent on the experience of the examiner.
To make CDU more operator independent, Aschwanden et al (44) introduced the “compression sign,” which should be methodologically more robust than the halo sign. A positive compression sign was defined as visibility of the temporal artery upon transducer-imposed compression of the artery. In their study of 80 patients, the compression sign proved to have the same sensitivity of 79% and specificity of 100% in comparison with the halo sign (44). Further studies are required to validate this observation.
In the setting of a typical clinical presentation and in the hands of a skilled examiner, CDU is a noninvasive, sensitive, and highly specific examination technique that helps to confirm the diagnosis of GCA and may assist in monitoring the effect of treatment. Several arteries can be examined in the same session and an immediate result is provided to the clinician to help guide the patient’s management. In patients with a low pretest probability of having GCA, normal CDU examination helps to rule out the disease and renders TAB unnecessary. We advocate that CDU should be increasingly implemented as a diagnostic and monitoring technique when caring for patients with suspected GCA.
Counter Point: All Patients With Presumed Giant Cell Arteritis Should Have a Temporal Artery Biopsy: Peter J. Savino, MD
I agree that accurately making the diagnosis of giant cell arteritis (GCA) is critical. Failing to make the diagnosis in a patient who has GCA can result in bilateral, untreatable, permanent blindness. On the other hand, making the diagnosis of GCA in a patient who does not have the disease, results in an elderly patient being subjected to high dose, long-term toxic corticosteroid therapy. Therefore, it behooves us to utilize the tests for GCA that have the highest sensitivity and specificity.
Diagnosis of GCA
A positive temporal artery biopsy (TAB) is near 100% specific for GCA. There are rare conditions where other conditions (herpes zoster ophthalmicus) might be confused with GCA but a positive TAB is diagnostic of GCA. The sensitivity of a TAB is a more difficult to state unequivocally. The best study regarding this question, I believe, is a mathematical evaluation of the sensitivity of TAB by Niederkohr and Levin (45). They found that the highest sensitivity for accurately diagnosing GCA was using TAB (sensitivity 98.7%).
Therefore, the question must be posed: if the TAB is negative for GCA, how does one make the diagnosis of “biopsy negative GCA”? In fact, does this entity exist? Various strategies have been advocated to substitute for TAB to diagnose GCA. The 5 criteria (age older than 50 years, onset of new headache, erythrocyte sedimentation rate greater than 50 mm/h, clinical abnormalities of the superficial temporal artery, and positive TAB) endorsed by the American College of Rheumatology (ACR) is most quoted (10). However, a prospective study of TAB vs the ACR criteria has shown that the sensitivity of the ACR vs a positive TAB was 72.7% and the specificity was 61.2% (2). This would seem to eliminate the ACR criteria as a reliable guide to diagnose GCA.
The question often arises: what if the biopsy is negative but GCA is still suspected on clinical grounds? There is no information in the literature to answer this query. I suspect most neuro-ophthalmologists would continue treating the patient with systemic corticosteroids but for a shorter period than if the patient had a positive TAB. I have always stopped the corticosteroids if the biopsy is negative and have never had a patient lose vision in a pattern suggestive of GCA after discontinuing treatment.
Studies on CDU as a diagnostic tool in patients with suspected GCA
Ultrasonography of the superficial temporal artery had been used to guide the surgeon to the area of the superficial temporal artery to biopsy. In 1997, Schmidt et al (3) proposed that CDU could substitute for TAB to diagnose GCA. They described the results of duplex ultrasonography performed on patients from January 1994 to October 1996. Thirty patients were diagnosed with GCA. However, only 21 had biopsy confirmed disease. The authors stated that 22 (73%) of these 30 patients had a positive ultrasonographic finding indicated by a dark halo around the lumen of the temporal artery. They also stated that 24 patients (80%) had stenosis or occlusions of the temporal artery segments and that 28 patients (93%) had stenosis, occlusions, or a halo. They further indicated that none of the patients “without temporal arteritis” had halos identified. One of the major problems with this study is that almost one-third of the patients said to have GCA had negative biopsies, and additional criteria, other than positive temporal artery biopsies (TABs), were used to make that diagnosis. On the basis of their data, the authors stated that CDU could replace TAB in the diagnosis of GCA and, therefore, biopsies need no longer be done. However, in an accompanying editorial, Hunder and Weyand (46) correctly (in my opinion) indicated that ultrasonography may not be sensitive enough in the more difficult patients who present with no definite credible signs of GCA on physical examination. They also indicated that the study did not prove the likelihood of ultrasonography diagnosing GCA better than TAB.
Meta-analyses of studies on the value of CDU in GCA
Although many ultrasonography articles have appeared, the number of patients in each study has been small. Meta-analyses addressing this subject have appeared. Karassa et al (37) performed a meta-analysis published in 2005. Twenty-three studies involving 2,036 patients were included in this analysis. It was found that the weighted sensitivity and specificity of the halo sign was 69% and 82%, respectively, compared with TAB and 55% and 94%, respectively, compared with the ACR criteria. They indicated that the best use of Doppler was when the pretest probability of GCA was low (10%) in that the ultrasound was normal.
A meta-analysis in 2010 by Ball et al (38) using 17 eligible studies containing 998 patients found that the halo sign on CDU, compared with TAB, had a 75% sensitivity and 83% specificity. Despite the low sensitivity and specificity, these authors recommended that Doppler ultrasonography should replace TAB and that the surgery should be reserved only for patients with negative scans. It should be noted, however, in this meta-analysis that patients were included who were diagnosed as having GCA with a negative TAB on the basis of the ACR criteria.
An earlier study addressed the issue of “biopsy negative GCA” by using ultrasonography and comparing the results only with patients having positive temporal artery biopsies (28). Although the numbers were small, with only 7 patients having positive TABs, 6 were identified accurately on ultrasonography. However, 6 patients were found to have false-positive findings on ultrasonography. This provided a sensitivity of 86% and a specificity of 68% and a positive predictive value of 50% for the use of ultrasound in the diagnosis of GCA. The authors indicated that they could not recommend changing from the routine practice of TAB to diagnose GCA.
Note of caution
Major caution has to be used when evaluating studies that claim to compare favorably or unfavorably with TAB. In almost all studies, it is stated that a positive TAB is the “gold standard” for the diagnosis for GCA. Yet, in almost all of the ultrasonography studies, patients have been included as having GCA when they have negative TABs. Even with this manipulation, the ultrasound, which is an indirect method of diagnosing GCA, does not compare favorably with TAB.
Finally, the arguments regarding any tests to replace TAB be that ultrasonography, laboratory testing, MRI scanning, or PET scanning involves, to a large extent, circular reasoning. Although patients have negative TABs and a positive TAB is said to be the “gold standard,” almost all these articles continue to include patients as having GCA by using other than the biopsy criteria. This type of circular reasoning cannot lead to an accurate evaluation of any of these modalities. Ultimately, the “gold standard” should be respected and that standard, TAB, should be performed in all patients suspected of having GCA.
Rebuttal: Klara Landau, MD
Dr. Savino poses the crucial question: What does “gold standard” mean in the context of using TAB for GCA diagnosis if false-negative biopsies do occur? How do we determine that a particular patient does have GCA despite a negative biopsy report? The elegant mathematical approach to this problem by Niederkohr and Levin (45) using data from studies with bilateral biopsies comes as close to a definitive albeit theoretical answer as it gets. Their result of 98.7% sensitivity also shows that the “gold standard” contains fewer karats of gold than the “pure 24” that we would wish for.
Let’s face the fact that performing a TAB in a routine clinical setting is not as straightforward as it may seem: patients are not keen to have it done and surgeons often are not eager to perform the procedure. A considerable number of patients will either not have a biopsy done at all or it will stay inconclusive for a variety of reasons, including specimen too small, vein being biopsied instead of artery, performed too long after beginning treatment, and lack of expertise by the pathologist. Dr. Savino points out that in the study by Schmidt et al (3), 30 patients were diagnosed with GCA despite the fact that only 21 were biopsy positive. What about the remaining nine? In only 4 patients (and not in all 9) TAB was negative, but we do not know when in the course of symptoms it was performed. Three patients refused to have a biopsy and in 2 the result was inconclusive.
Dr. Savino’s main criticism points to the “circular reasoning” in recommending CDU instead of TAB for diagnosing temporal arteritis because the results published in studies on the use of CDU include patients who did not have their disease proven by a TAB. But if only patients with a positive TAB were included in such studies, it would be impossible to prove any test more sensitive than the 100% of positive biopsies!
During the Christmas holiday of 2011, my 92-year-old mother developed symptoms highly suggestive of GCA, including headache, jaw claudication, and transient visual loss, as well as elevated erythrocyte sedimentation rate (ESR) and C-reactive protein. I immediately prescribed systemic steroids and asked my colleagues to perform CDU on her temporal arteries. Admittedly, no halos, stenoses, or occlusions were found. I did not opt for a TAB that I considered too stressful for her in view of the immediate relief of signs and symptoms that she experienced after starting steroid treatment. I suppose it was too early for ultrasound or biopsy to have yielded a positive result—as “waiting increases the likelihood of positive biopsy findings, but it also increases the chance of a vascular complication” as addressed by Hunder and Weyand in their editorial in 1997 (46). But the CDU was far less disruptive than the hassle of a TAB during the holiday season would have been for all parties involved. Over a period of 4 months, steroids were tapered and my mother’s symptoms never reappeared. This personal experience exemplified the immense gap between the awareness of evidence-based medicine on one side and the decision making based on an individual patient’s situation on the other.
I agree that a TAB should be performed in patients in whom GCA is possible but equivocal and histological diagnosis will be very helpful to either adhere to a treatment regimen if the biopsy is positive or taper the treatment off rapidly when it is negative. I do not think that TAB must be performed when clinical presentation is highly suggestive of GCA and CDU supports the diagnosis. Similarly, TAB can be omitted in patients with low clinical probability of GCA and normal CDU findings obtained by a skilled examiner.
The editorial by Hunder and Weyand (46) in the New England Journal of Medicine was not quite as critical as described by Dr. Savino. Admittedly, Schmidt et al (3) may have overstated their conclusions because 16 years later, the use of CDU to diagnose GCA remains controversial. Nevertheless, CDU is a powerful method that is increasingly being used as a diagnostic and monitoring technique in patients with suspected GCA, and rightly so.
Rebuttal: Peter J. Savino, MD
Drs. Landau and Gruber have cited numerous articles from the ultrasound literature to support the use of CDU instead of TAB to diagnose GCA. However, a reading of these articles shows that they do not attain the levels of sensitivity and specificity of TAB in diagnosing GCA and therefore, I believe, establishes the point that CDU should not be used in lieu of TAB. A close reading of the cited article shows several additional reasons why they are not “proof of concept”:
- The range of sensitivities and specificities varies greatly among these publications. One of the reasons is identified by my coauthors themselves when they write that the test “remains highly examiner dependent.” This statement alone should disqualify CDU from replacing TAB at this point in time. Another significant problem is that patients with negative TABs were used in most of these publications to label a patient as having GCA. Patients with negative TABs but with positive CDU studies were accepted as having GCA on clinical grounds. How is it possible for CDU to show indirect positive findings of histologic changes from GCA in the superficial temporal artery when the direct histopathological examination fails to do so? To be credible, a newly proposed test must be compared with the “gold standard” test before the new test can be proposed as a substitute for standard test. Including so-called biopsy-negative GCA as proof of the validity of CDU is less than convincing.
- We are in agreement that systemic corticosteroid therapy must be instituted in suspected GCA before confirmatory testing is performed. The histopathological changes of GCA are evident on TAB for 2 weeks at a minimum and usually longer. CDU, however, shows a marked reduction of the diagnostic sensitivity to 50% within 2–4 days (40).
- Another argument used against TAB is the presence of “skip lesions.” The length of these normal areas in a positive TAB is small so that obtaining a biopsy length of 1–2 cm and sampling various areas of the specimen makes this fact moot.
- Likewise the often-cited “serious complications” of TAB in the nonsurgical literature. Admittedly, the world’s worst surgeon may experience more than a minor complication with TAB; however, in competent hands, TAB is, for all intents and purposes, devoid of “serious” complications.
- The authors state that CDU in GCA “can be used for treatment monitoring.” How exactly would this be done? Should corticosteroid therapy be halted or decreased when the repeat CDU improves? Since we know this may occur in less than 1 week, this suggestion is potentially dangerous.
- The authors admit that CDU cannot distinguish between GCA and “different types of inflammatory changes in the vessel wall.” They state: “This is a minor drawback in comparison with TAB as GCA is by far the most important and most frequently encountered entity that is being diagnosed using TAB.” This would result in patients with all other forms of vasculitis being treated for months with large amounts of corticosteroids even if that is not the appropriate therapy for that specific disorder.
Finally, I am not aware of the medicolegal climate in other countries but the United States is a litigious society. A physician’s best defense, in any matter, is that he or she has adhered to the “standard of care.” Abandoning the “gold standard” of TAB for an indirect diagnostic modality in the diagnosis of GCA would be problematic at best.
Conclusion: Valérie Biousse, MD, and Andrew G. Lee, MD
The sensitivity, specificity, and utility of CDU for diagnosing GCA continue to evolve and improve. As with any clinical decision, however, the pretest likelihood of disease (i.e., clinical suspicion of disease) is a much more powerful predictor of disease than the test per se and this applies even to the gold standard of TAB. For example, in an 80-year-old woman with new onset headache, no light perception vision due to pallid disc edema, and jaw claudication, the finding of a normal ESR and C-reactive protein and even a negative TAB would not dissuade us from treating for GCA. Perhaps the question is not whether CDU can replace TAB in the diagnosis of GCA but instead whether CDU could be a sufficiently sensitive and specific adjunctive tool in specific cases to make a diagnosis without a confirmatory TAB. If the clinical findings are highly suspicious for GCA (as in the patient example above) and in the setting of a positive CDU, then the posttest likelihood of disease might be high enough to justify steroid treatment and be sufficient for medical and medicolegal purposes without a TAB. Likewise, in a patient with a very low clinical suspicion for GCA (e.g., 40-year-old man with ESR of 100 mm/h) in whom we might be loath to perform an essentially non-diagnostic TAB, a negative CDU might be very reassuring to the patient and attending physicians.
1. Melson MR, Weyand CM, Newman NJ, Biousse V. The diagnosis of giant cell arteritis. Rev Neurol Dis. 2007;4:128–142.
2. Murchison AP, Gilbert ME, Bilyk JR, Eagle RC Jr, Pueyo V, Sergott RC, Savino PJ. Validity of the American College of Rheumatology criteria for the diagnosis of giant cell arteritis. Am J Ophthalmol. 2012;154:722–729.
3. Schmidt WA, Kraft HE, Vorpahl K, Volker L, Gromnica-Ihle EJ. Color duplex ultrasonography in the diagnosis of temporal arteritis. N Engl J Med. 1997;337:1336–1342.
4. Hunder GG. Epidemiology of giant-cell arteritis. Cleve Clin J Med. 2002;69(suppl 2):SII79–SII82.
5. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifestations of giant cell arteritis. Am J Ophthalmol. 1998;125:509–520.
6. Jonasson F, Cullen JF, Elton RA. Temporal arteritis. A 14-year epidemiological, clinical and prognostic study. Scott Med J. 1979;24:111–117.
7. Meli B, Landau K, Gloor BP. [The bane of giant cell arteritis from an ophthalmological viewpoint]. Schweiz Med Wochenschr. 1996;126:1821–1828.
8. Proven A, Gabriel SE, Orces C, O'Fallon WM, Hunder GG. Glucocorticoid therapy in giant cell arteritis: duration and adverse outcomes. Arthritis Rheum. 2003;49:703–708.
9. Meola DC, Fierz A, Tschopp A, Landau K. [Corticosteroids in giant cell arteritis: primum nil nocere?]. Klin Monbl Augenheilkd. 2006;223:379–381.
10. Hunder GG, Bloch DA, Michel BA, Stevens MB, Arend WP, Calabrese LH, Edworthy SM, Fauci AS, Leavitt RY, Lie JT, Lightfoot RW Jr, Masi AT, McShane DJ, Mills JA, Wallace SL, Zvaifler NJ. The American College of Rheumatology 1990 criteria for the classification of giant cell arteritis. Arthritis Rheum. 1990;33:1122–1128.
11. Rao JK, Allen NB, Pincus T. Limitations of the 1990 American College of Rheumatology classification criteria in the diagnosis of vasculitis. Ann Intern Med. 1998;129:345–352.
12. Hall S, Persellin S, Lie JT, O'Brien PC, Kurland LT, Hunder GG. The therapeutic impact of temporal artery biopsy. Lancet. 1983;2:1217–1220.
13. Murchison AP, Bilyk JR. Brow ptosis after temporal artery biopsy: incidence and associations. Ophthalmology. 2012;119:2637–2642.
14. Yoon MK, Horton JC, McCulley TJ. Facial nerve injury: a complication of superficial temporal artery biopsy. Am J Ophthalmol. 2011;152:251–255 e251.
15. Vollrath-Junger C, Gloor B. [Why perform Doppler sonography before every biopsy of the temporal artery?]. Klin Monbl Augenheilkd. 1989;195:169–171.
16. Townes DE, Blodi FC. The diagnostic value of temporal-artery biopsy. Trans Am Ophthalmol Soc. 1968;66:33–44.
17. Albert DM, Ruchman MC, Keltner JL. Skip areas in temporal arteritis. Arch Ophthalmol. 1976;94:2072–2077.
18. Boyev LR, Miller NR, Green WR. Efficacy of unilateral versus bilateral temporal artery biopsies for the diagnosis of giant cell arteritis. Am J Ophthalmol. 1999;128:211–215.
19. Breuer GS, Nesher G, Nesher R. Rate of discordant findings in bilateral temporal artery biopsy to diagnose giant cell arteritis. J Rheumatol. 2009;36:794–796.
20. Danesh-Meyer HV, Savino PJ, Eagle RC Jr., Kubis KC, Sergott RC. Low diagnostic yield with second biopsies in suspected giant cell arteritis. J Neuroophthalmol. 2000;20:213–215.
21. Pless M, Rizzo JF III, Lamkin JC, Lessell S. Concordance of bilateral temporal artery biopsy in giant cell arteritis. J Neuroophthalmol. 2000;20:216–218.
22. Venz S, Hosten N, Nordwald K, Lemke AJ, Schroder R, Bock JC, Hartmann CF, Felix R. [Use of high resolution color Doppler sonography in diagnosis of temporal arteritis]. Rofo. 1998;169:605–608.
23. Schmidt WA, Kraft HE, Volker L, Vorpahl K, Gromnica-Ihle EJ. Colour Doppler sonography to diagnose temporal arteritis. Lancet. 1995;345:866.
24. Schmidt WA, Gromnica-Ihle E. What is the best approach to diagnosing large-vessel vasculitis? Best Pract Res Clin Rheumatol. 2005;19:223–242.
25. Puechal X, Chauveau M, Hilliquin P, Perrot S, Job-Deslandre C, Menkes CJ. Superficial temporal Doppler flow studies in suspected giant cell arteritis: validation of a diagnostic score. Arthritis Rheum. 1994;37:S409 (abstract).
26. Puechal X, Chauveau M, Menkes CJ. Temporal Doppler-flow studies for suspected giant-cell arteritis. Lancet. 1995;345:1437–1438.
27. Romera-Villegas A, Vila-Coll R, Poca-Dias V, Cairols-Castellote MA. The role of color duplex sonography in the diagnosis of giant cell arteritis. J Ultrasound Med. 2004;23:1493–1498.
28. Murgatroyd H, Nimmo M, Evans A, MacEwen C. The use of ultrasound as an aid in the diagnosis of giant cell arteritis: a pilot study comparing histological features with ultrasound findings. Eye (Lond). 2003;17:415–419.
29. Salvarani C, Silingardi M, Ghirarduzzi A, Lo Scocco G, Macchioni P, Bajocchi G, Vinceti M, Cantini F, Iori I, Boiardi L. Is duplex ultrasonography useful for the diagnosis of giant-cell arteritis? Ann Intern Med. 2002;137:232–238.
30. LeSar CJ, Meier GH, DeMasi RJ, Sood J, Nelms CR, Carter KA, Gayle RG, Parent FN, Marcinczyk MJ. The utility of color duplex ultrasonography in the diagnosis of temporal arteritis. J Vasc Surg. 2002;36:1154–1160.
31. Pfenninger L, Horst A, Stuckmann G, Flury R, Sturmer J. Comparison of histopathological findings with duplex sonography of the temporal arteries in suspected giant cell arteritis. Klin Monbl Augenheilkd. 2012;229:369–373.
32. Habib HM, Essa AA, Hassan AA. Color duplex ultrasonography of temporal arteries: role in diagnosis and follow-up of suspected cases of temporal arteritis. Clin Rheumatol. 2012;31:231–237.
33. Karahaliou M, Vaiopoulos G, Papaspyrou S, Kanakis MA, Revenas K, Sfikakis PP. Colour duplex sonography of temporal arteries before decision for biopsy: a prospective study in 55 patients with suspected giant cell arteritis. Arthritis Res Ther. 2006;8:R116.
34. Alberts MS, Mosen DM. Diagnosing temporal arteritis: duplex vs. biopsy. QJM. 2007;100:785–789.
35. Stammler F, Grau C, Schnabel A. [Value of colour Doppler ultrasonography in relation to clinical pretest probability in giant cell (temporal) arteritis]. Dtsch Med Wochenschr. 2009;134:2109–2115.
36. Nesher G, Shemesh D, Mates M, Sonnenblick M, Abramowitz HB. The predictive value of the halo sign in color Doppler ultrasonography of the temporal arteries for diagnosing giant cell arteritis. J Rheumatol. 2002;29:1224–1226.
37. Karassa FB, Matsagas MI, Schmidt WA, Ioannidis JP. Meta-analysis: test performance of ultrasonography for giant-cell arteritis. Ann Intern Med. 2005;142:359–369.
38. Ball EL, Walsh SR, Tang TY, Gohil R, Clarke JM. Role of ultrasonography in the diagnosis of temporal arteritis. Br J Surg. 2010;97:1765–1771.
39. Arida A, Kyprianou M, Kanakis M, Sfikakis PP. The diagnostic value of ultrasonography-derived edema of the temporal artery wall in giant cell arteritis: a second meta-analysis. BMC Musculoskelet Disord. 2010;11:44.
40. Hauenstein C, Reinhard M, Geiger J, Markl M, Hetzel A, Treszl A, Vaith P, Bley TA. Effects of early corticosteroid treatment on magnetic resonance imaging and ultrasonography findings in giant cell arteritis. Rheumatology (Oxford). 2012;51:1999–2003.
41. Pfadenhauer K, Weber H. Giant cell arteritis of the occipital arteries—a prospective color coded duplex sonography study in 78 patients. J Neurol. 2003;250:844–849.
42. Garcia-Garcia J, Ayo-Martin O, Argandona-Palacios L, Segura T. Vertebral artery halo sign in patients with stroke: a key clue for the prompt diagnosis of giant cell arteritis. Stroke. 2011; 42:3287–3290.
43. Maldini C, Depinay-Dhellemmes C, Tra TT, Chauveau M, Allanore Y, Gossec L, Terrasse G, Guillevin L, Coste J, Mahr A. Limited value of temporal artery ultrasonography examinations for diagnosis of giant cell arteritis: analysis of 77 subjects. J Rheumatol. 2010;37:2326–2330.
44. Aschwanden M, Daikeler T, Kesten F, Baldi T, Benz D, Tyndall A, Imfeld S, Staub D, Hess C, Jaeger KA. Temporal artery compression sign–a novel ultrasound finding for the diagnosis of giant cell arteritis. Ultraschall Med. 2013;34:47–50.
45. Niederkohr RD, Levin LA. A Bayesian analysis of the true sensitivity of a temporal artery biopsy. Invest Ophthalmol Vis Sci. 2007;48:675–680.
46. Hunder GG, Weyand CM. Sonography of giant cell arteritis. N Engl J Med. 1997;337:1385–1386.