The most frequently used tests for the diagnosis of temporal arteritis include Westergren erythrocyte sedimentation rate, plasma electrophoresis, and temporal artery biopsy. Recently, interest has focused on the diagnostic utility of measuring choroidal blood flow using either newly developed methods for noninvasive measurement, such as ocular plethysmography, or timed fluorescein angiography.
Marked elevation of the erythrocyte sedimentation rate is the most characteristic laboratory abnormality in patients with temporal arteritis. Mild anemia is common, as are various degrees of leukocytosis, sometimes with a left shift. Serum protein electrophoresis is characterized by rise of "acute phase reactants" (alpha and beta globulins) in the early stages and elevated gamma globulin in the more chronic phases; both of these will act to elevate the erythrocyte sedimentation rate. Although erythrocyte sedimentation rate elevation is found at presentation in a high proportion of patients with temporal arteritis, review of large series indicates that the erythrocyte sedimentation rate is relatively normal in at least 10% of presenting patients (Goodwin 1980). A study of 167 patients with a firm clinical diagnosis of temporal arteritis confirmed that the erythrocyte sedimentation rate was less than 40 mm/hour at diagnosis in 9 patients. These 9 patients had less frequent systemic and visual symptoms than the others, and no patient with low erythrocyte sedimentation rate developed blindness (Salvarani and Hunder 2001). Therefore, a normal erythrocyte sedimentation rate does not rule out the diagnosis when other features are highly suggestive.
The Westergren method, which uses a longer tube than the Wintrobe procedure, is preferred because the result is not affected as much by anemia. The degree of erythrocyte sedimentation rate elevation in temporal arteritis and polymyalgia rheumatica is often extreme (greater than 100 mm in the first hour). Unfortunately, other conditions, some of which produce systemic symptoms similar to temporal arteritis, also elevate the erythrocyte sedimentation rate. A recent review of the literature showed that the sedimentation rate is nonspecific and has been found to be elevated in a wide variety of clinical conditions unrelated to temporal arteritis, even some noninflammatory conditions such as stroke, coronary artery disease, and prostate cancer.
Another problem is establishing the normal erythrocyte sedimentation rate range in the elderly. Review of several large series leads to the conclusion that a value greater than 40 mm (first hour) should be considered elevated in those aged 65 years and older. Values between 20 and 40 are "suspect" but far from diagnostic.
Studies have sought to find better biological markers for clinical activity of the disease and markers to differentiate temporal arteritis from other rheumatological conditions. Elevation of anticardiolipin antibodies in sera of patients with temporal arteritis has attracted considerable attention. In a prospective study of 86 patients with biopsy-positive temporal artery biopsy and 50 controls, Liozon and colleagues' group found anticardiolipin antibody in 50% of patients but only 8% of controls. The presence of antibody did not select for patients who were destined to have ischemic complications. IgG class antibodies declined with treatment and increased with exacerbations, making this a better indicator of disease activity than IgM class antibodies (Liozon et al 1995). Espinosa and colleagues found similar incidences of antiphospholipid antibodies among 86 patients with giant cell arteritis, including those with temporal arteritis and others with polymyalgia rheumatica alone or combined with temporal arteritis. Forty-four patients (67%) had ischemic events, but there was no significant correlation between the presence of antiphospholipid antibody and these events (Espinosa et al 2001).
Chakravarty and colleagues assayed anticardiolipin antibody levels at presentation and at 6 monthly intervals in 98 patients with either polymyalgia rheumatica (64 patients), temporal arteritis (12 patients), or both (22 patients), as well as 100 controls. Five of 11 patients with polymyalgia and elevated anticardiolipin developed temporal arteritis, whereas only 5 of 53 polymyalgia patients without antibody did so, a striking difference. Also, 3 of the 5 with temporal arteritis and antibodies developed severe ischemic complications, whereas none of the remainder without antibody did so (Chakravarty et al 1995). In this respect their study differed from those of Liozon and colleagues as well as of Espinosa and colleagues.
Meyer and colleagues also studied 19 patients with temporal arteritis, 16 of whom also had polymyalgia rheumatica, and 3 patients with isolated polymyalgia using an ELISA assay for anticardiolipin IgG antibodies (Meyer et al 1996). Antibodies were demonstrated in 8 patients (36%), all of whom had temporal arteritis with or without concomitant polymyalgia. Serial determinations showed that the anticardiolipin activity disappeared within a few weeks of steroid treatment. The authors suggest that presence of anticardiolipin antibody or anti-beta 2 glycoprotein I antibody in a patient with polymyalgia rheumatica should suggest the additional presence of temporal arteritis.
Sorbi and colleagues looked for degradation products of intercellular matrix and elastin in serum samples and in tissue from temporal artery biopsy from 12 patients with temporal arteritis and 12 healthy controls. Gelatinase activity, and specifically matrix metalloproteinase 9 levels, was "substantially elevated" in sera from the patients but not the controls. Matrix metalloproteinase 9 RNA was found in the media of inflamed vascular segments. It was suggested that further work should be done to determine if this could be made into a clinically useful marker of disease activity (Sorbi et al 1996).
Hayreh and colleagues reexamined the sensitivity and specificity of the erythrocyte sedimentation rate and C-reactive protein levels in diagnosis of temporal arteritis (Hayreh et al 1997). The values for these 2 tests were compared between a group of 106 patients with positive temporal artery biopsy and another group of 247 patients with negative biopsy. These authors concluded that the laboratory criteria most strongly suggestive of temporal arteritis included C-reactive protein above 2.45 mg/dl and erythrocyte sedimentation rate of 47 mm/hr or more, in that order. C-reactive protein was more sensitive (100%) than erythrocyte sedimentation rate (92%) for detection of temporal arteritis. The combination of C-reactive protein and erythrocyte sedimentation rate above the cutoff values gave the best specificity (97%).
Comparing a group of 121 biopsy proven temporal arteritis patients with 287 patients with nonarteritic anterior ischemic optic neuropathy, Costello and colleagues found that in addition to high erythrocyte sedimentation rate and C-reactive protein, there was significant elevation of platelet count and white blood count in the arteritis patients compared with the nonarteritic patients (Costello et al 2004). Erythrocyte sedimentation rate was a better predictor of temporal arteritis than platelet count and the combination of erythrocyte sedimentation rate and platelet count was a better predictor than erythrocyte sedimentation rate alone.
Demonstration of the inflammatory granulomatous lesion of temporal arteritis has made temporal artery biopsy a mainstay in establishing the diagnosis with assurance. The typical lesion is a granulomatous inflammatory reaction concentrated at the innermost part of the vascular media near zones of fragmented and reduplicated internal elastic lamina with adjacent intimal proliferation, sometimes to the point of occluding the lumen. Multinucleated giant cells are classically present but are not necessary for the diagnosis.
The philosophy behind obtaining a temporal artery biopsy requires some special comments. There are many cases of active temporal arteritis in which an entire temporal artery or segments of the artery are not involved; these were called "skip lesions" by Klein and colleagues (Klein et al 1976). To compensate for the patchy nature of the pathologic process, these authors suggested that the length of the biopsy segment should measure at least 2.5 cm, and if frozen section of samples from that specimen are all normal, that the other temporal artery be biopsied at the same surgical session. On the other hand, Chakrabarty and colleagues report that in a reexamination of 172 temporal artery biopsy specimens, only 1 of the 132 initially normal cases, and 2 of 14 diagnosed with periarterial lymphocytic infiltration, revealed giant cell arteritis after examining the tissue at multiple levels. The authors concluded that "routinely examining a temporal artery biopsy at multiple levels does not increase the diagnostic yield of the test, although selective further examination may be indicated in some cases" (Chakrabarty and Franks 2000).
Among 5 series on temporal artery biopsy, the average rate of positive result for temporal arteritis was 19%, ranging from 11% to 27% in various series. These series contrast sharply with the large epidemiologic studies in which the biopsy positivity rate exceeds 80% (Allison and Gallagher 1984; Lie and members and consultants of the American College of Rheumatology Subcommittee on Classification of Vasculitis 1990; Baldursson et al 1994). One obvious factor that would influence the percentage of positive biopsies is the clinical rigor with which candidates were chosen for biopsy, and this factor is usually not thoroughly described in the clinical reports.
Another factor would be the potential effect of corticosteroid treatment on the histologic characteristics of the biopsy specimen. Allison and Gallagher presented evidence suggesting that corticosteroid treatment prior to biopsy results in a significant reduction in the rate of positive results. Of the patients biopsied before steroid treatment, 82% (50 of 61) were positive, but this figure fell to 59% (30 of 51) with 1 week or less of prebiopsy treatment, and to only 20% (4 of 20, 2 of whom had only "healed vasculitis") when prebiopsy steroid treatment exceeded 1 week.
Chmelewski and colleagues analyzed 98 patients who underwent temporal artery biopsy and found that only 30 (31%) were positive, but that the percentage that were positive did not decline with the duration of previous corticosteroid therapy (Chmelewski et al 1992). Six of 16 (37%) patients still had positive biopsies after more than 2 weeks on corticosteroids. The dosage at time of biopsy and before was not specified, but for the entire series the average starting prednisone dose was 68 mg/dL.
Achkar and colleagues assessed the effect of corticosteroid treatment on biopsy positivity among 535 patients biopsied at the Mayo Clinic between 1988 and 1991, taking into account the dosage and duration of treatment (Achkar et al 1994). The patients who had received corticosteroid treatment for 1 to 7 days prior to biopsy had a somewhat higher positivity rate (46 of 107, 43%) than those who had not received treatment (89 of 286, 31%). More of the treated patients had classic symptoms and signs, such as tender or pulseless temporal artery or jaw claudication, than the untreated patients, probably indicating that they had more intense disease activity. Also, the positivity rate was higher for those who had been treated for less than 7 days (46 of 107, 43%) than for those treated for 8 to 14 days (3 of 10, 30%) or longer (9 of 32, 28%), which suggests either that the duration of corticosteroid treatment is important for its effect on the histopathology or that the perceived urgency for treatment and rapid biopsy (and, therefore, the disease intensity) was highest for those biopsied right away. The percentage of positive biopsies that were atypical (having no giant cells or location of inflammation in the adventitia rather than the media) rose strikingly with the duration of corticosteroid prebiopsy treatment. An earlier study of the same patient cohort from the Mayo Clinic looked at follow-up data for a median of 6 years and determined that only 9% of the biopsy-negative cases were later treated for temporal arteritis, and none of this group had ischemic complications (Hall et al 1983). It should be kept in mind, however, that the standard practice at the Mayo Clinic is to biopsy artery segments longer than 2.5 cm as well as the opposite temporal artery if the frozen sections of the first artery are all negative. This intensive approach probably increases the reliability of a negative biopsy, but is probably not matched in most institutions. A smaller study found no difference in the percent of positive biopsies with duration of corticosteroid treatment at the time of biopsy, at least up to 4 weeks: 82% positive overall, 86% positive biopsied 4 or more weeks after treatment started (Ray-Chaudhuri et al 2002).
Chmelewski and colleagues also tabulated the fate of their 68 patients with negative temporal artery biopsies and found that their presenting symptoms and signs could not be distinguished from those of patients whose biopsies were positive. The eventual diagnosis was still temporal arteritis in 14 of 68 (21%), whereas the other patients eventually carried a wide array of other diagnoses, the most frequent being primary neurologic disorders (22%), polymyalgia rheumatica (21%), or other rheumatologic syndromes (15%) (Chmelewski et al 1992). Duhaut found that the clinical course and other parameters of temporal arteritis were generally more severe in patients with biopsy positivity as compared with biopsy negative cases (Duhaut et al 1999b).
Occasionally, a temporal artery biopsy specimen shows small-vessel vasculitis surrounding a spared temporal artery and the significance of this is unclear. When the clinical picture is typical for temporal arteritis and the histology of the vasculitis in the smaller arteries is typical of the disorder including the presence of giant cells, this small vessel involvement is probably as good evidence of temporal arteritis as is involvement of the main vessel. This means that the small branches attached to the temporal artery specimen should be looked at histologically and not discarded from the specimen (Chiu et al 2004).
Esteban and colleagues studied 28 patients who had temporal artery biopsy showing this pattern of small-vessel vasculitis and found that 13 of them probably had temporal arteritis on analysis of the entire clinical picture. Three of them had systemic necrotizing vasculitis in these involved small vessels and demonstrated significantly more fibrinoid necrosis than the others. The authors looked for small-vessel vasculitis in 30 temporal artery specimens from patients with biopsy-proven giant cell arteritis. No difference was found in the pattern of small vessel involvement surrounding histologically spared temporal or involved arteries among patients with clinically or histologically diagnosed temporal arteritis (Esteban et al 2001). Another study found that perivascular inflammation in the absence of arteritis in temporal artery biopsy specimens did not increase the likelihood of having temporal arteritis over patients with no inflammation on biopsy (Corcoran et al 2001).
Some have claimed that finding evidence of scarring and other changes interpreted as residua of healed arteritis in a temporal artery specimen that does not show any active inflammation can be used as evidence of temporal arteritis; others have argued that these changes are nonspecific and may reflect arteriosclerosis. Cox and colleagues carried out a retrospective study of 47 temporal artery biopsy specimens and compared them with 10 control temporal arteries from autopsies on patients who had no clinical evidence of vasculitis. They found that apart from inflammation, no histologic findings were specific for temporal arteritis. In other words, structural changes other than inflammation do not allow reliable differentiation between healed or quiescent temporal arteritis and arteriosclerosis (Cox and Gilks 2001).
A large survey of U.S. specialists in the fields of ophthalmic plastic and reconstructive Surgery (n=127), neuro-ophthalmology (n=119), rheumatology (n=799), and other (n=28) was carried out in 2010 and reported in 2013 (Schallhorn et al 2013). Sixty-six percent of respondents favored initial unilateral temporal artery biopsy, 18% bilateral biopsy in all cases, and 16% unilateral or bilateral biopsy depending on the degree of clinical suspicion. Rheumatologists were 4.5 times more likely to favor initial bilateral biopsy than the other 2 groups. Most believed that biopsy results were not rendered falsely negative with immunosuppressive treatment for up to 2 weeks.
Each physician must deal with negative biopsy results based on the adequacy of the specimen submitted and the intensity with which the biopsy material is studied in his or her own institution. Also, the strength of the clinical presentation must be weighed, and if the signs and symptoms are typical and fairly specific (jaw claudication, tender arteries) then patients should not be denied prompt treatment based on negative biopsy results.
Gillanders and colleagues introduced the idea of using selective angiography of the temporal artery in the diagnosis of temporal arteritis (Gillanders 1969). The study was used to identify abnormal-appearing segments of the vessels both as an indication of the diagnosis and as a guide to which segments should yield a positive biopsy. Fifteen patients with temporal arteritis or polymyalgia rheumatica were studied using temporal artery biopsy and angiography. The vascular lumen in temporal arteritis was characterized by long alternating stenotic and dilated segments having smooth transitions between the two. Other segments showed more abrupt transitions to different caliber, giving a more beaded appearance. The biopsy was positive in 5 patients, 2 of whom also had positive angiograms. Remarkably, the angiogram was considered characteristic of temporal arteritis in 2 patients with negative biopsy of the same vessel.
Elliott and colleagues at the Mayo Clinic carried out a more extensive study (Elliott et al 1972). As a preliminary, the authors studied 100 normal angiograms and, when possible, correlated the findings with autopsy injections of the vessels. These studies were compared with angiograms on 34 patients with temporal arteritis or polymyalgia rheumatica, all of whom had the vessel biopsied. Seven of the angiograms were considered diagnostic for arteritis, whereas the biopsies were positive in only 5 of these. Gillanders' earlier finding of an occasional positive biopsy from angiographically normal segments was corroborated in this study: in 1 patient the angiographic abnormality was distal to the site from which a positive biopsy was obtained. Moncada and colleagues reported an essentially similar experience, finding 5 abnormal temporal artery angiograms in 20 patients with polymyalgia rheumatica, only 1 of which was histologically abnormal (Delecoeuillerie et al 1988).
All of the workers who reported on the use of temporal artery angiography to diagnose temporal arteritis concurred that the characteristic gentle alternations between stenotic and dilated segments are not seen with any frequency in atherosclerotic vessels, which may, however, look beaded and irregular. There have been no recent reports of invasive angiography in the diagnosis of temporal arteritis, probably because sophisticated noninvasive methods have since emerged to quantify flow velocity and direction in both the external carotid and internal carotid systems.
Kraft, using newly developed color Doppler techniques to study 10 patients with temporal arteritis, 8 with polymyalgia rheumatica and 23 controls, reported that the superficial temporal artery showed a characteristic hypoechoic halo around the lumen of an often stenosed or occluded segment in the patients but in none of the controls (Kraft et al 1996). The authors comment that the sensitivity and specificity of this finding has yet to be determined, and that study of more patients is needed. Lauwerys and colleagues used color Doppler sonography to study 11 patients with temporal arteritis, 21 with polymyalgia rheumatica, and 32 controls (Lauwerys et al 1997). They found significant reduction in the peak velocity at both the proximal and distal segments of the superficial temporal arteries of temporal arteritis patients as compared with polymyalgia rheumatica patients and controls. They found hypoechoic temporal artery thickening in only 2 patients. Follow-up of 6 patients with temporal arteritis under treatment showed significant recovery of peak velocity at the distal temporal artery site. A prospective color duplex study of 30 patients with temporal arteritis, 37 with polymyalgia rheumatica, 15 with final diagnoses other than temporal arteritis or polymyalgia rheumatica, and 30 age- and sex-matched controls found that 28 of the 30 (92%) temporal arteritis patients had superficial temporal artery stenoses, occlusions, or dark halo (edematous artery?), whereas none of the 82 patients without temporal arteritis had these findings. The authors found the dark halo to be the most specific of the abnormalities. Also using color Doppler techniques, Vecsei and colleagues found abnormal flow characteristics in the central retinal artery, ophthalmic artery, and lateral short posterior ciliary arteries during symptom-free intervals in patients with amaurosis fugax and giant cell arteritis, but normal flow in these vessels between attacks in nonarteritic cases of amaurosis fugax (Vecsei et al 1999). Currently the largest study of duplex ultrasonography of the temporal arteries examined 86 patients suspected of having temporal arteritis or polymyalgia rheumatica. The dark (hypoechoic) halo sign was present in only 40% and showed a specificity of 79% for correct diagnosis of biopsy proven temporal arteritis. Temporal artery abnormalities on physical examination had higher sensitivity and specificity (Salvarani et al 2002). Similar conclusions about the value of the dark halo sign were reached in another study of 69 patients (Nesher et al 2002). These findings indicate the potential usefulness of color Doppler investigation in elderly patients with amaurosis fugax.
Studies have demonstrated reduced ocular and choroidal perfusion in patients with temporal arteritis, even before ischemic visual loss has occurred, and it has been suggested that this type of measurement is a reliable diagnostic marker for the arteritic form of retinal and optic nerve ischemia. Bosley and colleagues used ocular plethysmography to measure the pulse amplitudes in the fellow-eyes of patients with anterior ischemic optic neuropathy (Bosley et al 1989). Pulse amplitudes were significantly lower in the fellow-eyes of 9 patients with biopsy-documented temporal arteritis and anterior ischemic optic neuropathy, as compared with 112 fellow-eyes in patients with nonarteritic, presumed atherosclerotic anterior ischemic optic neuropathy as well as fellow-eyes in another 9 biopsy-negative patients suspected to have temporal arteritis. In the authors' study group the test had a sensitivity of 100% and a specificity of 93.4%, with diagnostic accuracy of 93.9% in temporal arteritis.
Slavin and Barondes reported on 3 patients who presented with visual loss in 1 eye but had a normal fundus exam. All 3 patients had significantly delayed choroidal filling on fluorescein angiography. Only 1 patient had symptoms of temporal arteritis, but the other 2 had diagnostic-positive temporal artery biopsy (Slavin and Barondes 1994). Siatkowski and colleagues used fluorescein angiography to study 35 patients with anterior ischemic optic neuropathy, of whom 16 (46%) had biopsy-positive temporal arteritis as well. These authors found that the arteritic group of patients had higher erythrocyte sedimentation rates, larger physiologic cup-to-disc diameter ratios, and significantly delayed fluorescein dye appearance and choroidal filling times as compared with the patients with nonarteritic, presumed atherosclerotic anterior ischemic optic neuropathy (Siatkowski et al 1993). In a similar study, Mack and colleagues performed timed fluorescein angiography on 13 patients with temporal arteritis (biopsy-positive) and visual symptoms, of whom 11 had anterior ischemic optic neuropathy, and 2 had only transient visual loss (Mack et al 1991). They also studied a group of 33 patients with nonarteritic anterior ischemic optic neuropathy and 23 eyes from 23 age-matched normal subjects. The arteritic group had significant delay of filling time (mean=69 seconds), as compared with the nonarteritic patients with anterior ischemic optic neuropathy (mean=5.8 seconds) and the control eyes (mean=5.5 seconds). Indocyanine green ocular angiography, which shows choroidal circulation preferentially, was not found to increase the diagnostic precision over standard fluorescein angiography (Valmaggia et al 1999).
Joelson and colleagues reported on the MRI findings in a patient with biopsy proven temporal arteritis. They described multifocal dural enhancement, which was characterized histologically by perivascular inflammatory infiltrates and enhancement of the temporalis muscle (Joelson et al 2000). Lee and colleagues described MRI optic nerve gadolinium enhancement in 3 patients with anterior ischemic optic neuropathy caused by temporal arteritis and commented that nonarteritic anterior ischemic optic neuropathy has not been reported with gadolinium enhancement on MRI (Lee et al 1999). Liu and Chesnutt also reported 2 cases of biopsy-positive giant cell arteritis with MRI-documented perineural enhancement of both optic nerves (Liu and Chesnutt 2013). The authors cited the earlier Lee paper and a publication by Morgenstern (Morgenstern et al 2003) indicating that pathologic examination of an optic nerve from a patient with giant cell arteritis showed an inflammatory infiltrate in the small perineural meningeal blood vessels, including infiltration with giant cells.
In a patient with biopsy-proven temporal arteritis, fludeoxyglucose[F18] PET showed striking uptake of fludeoxyglucose[F18] in the walls of the entire aorta and left main coronary artery as well as the subclavian, carotid, and common iliac arteries on both sides. These findings normalized after 2 weeks treatment with corticosteroids (Turlakow et al 2001).
Ultrasound and PET can be used to demonstrate arteritis in the large arteries of the chest and neck in patients with temporal arteritis. Both procedures agree on the distribution of these changes in patients with large-vessel giant cell arteritis and these changes are most often clinically silent. Hypoechoic halo around arteries indicating mural edema is the most useful ultrasound feature (Schmidt and Blockmans 2005; Habib et al 2012).
Habib and colleagues, using color Doppler ultrasound, studied 32 patients suspected of temporal arteritis of which 16 were eventually diagnosed as having temporal arteritis with positive temporal artery biopsy. Of the 16 with biopsy-proven disease, the color Doppler exam showed hypoechoic dark halos surrounding the perfused lumen of the temporal artery in 13 patients (81%). The finding was unilateral in 7 patients and bilateral in 6 patients. Halos were detected in 2 patients outside the temporal arteritis group, 1 with polymyalgia rheumatica and the other with Takayasu arteritis. None of 30 age- and gender-matched control subjects had halos detected (Habib et al 2012).
Because interpretation of F FDG-PET requires subjective judgment and experience, Hautzel and colleagues sought to demonstrate an objective measure that could with sensitivity distinguish significant F-fluorodeoxyglucose uptake in the aorta by quantifying the relationship between aorta and liver uptake in a group of 23 patients suspected of having temporal arteritis and in a control group. They found that a “receiver operating characteristic”-based cutoff ratio of 1.0 led to a sensitivity of 88.9%, specificity of 95.1%, and accuracy of 94.4%. This aorta-to-liver ratio applied to the control group resulted in specificity of 95.6%, thus, providing a reliable, investigator-independent indicator of giant cell arteritic involvement of the aorta (Hautzel et al 2008).
Another study of 46 patients with temporal arteritis showed that those with F-fluorodeoxyglucose uptake in the aorta had significantly larger diameter ascending, descending, and thoracic aorta at CT scan done a mean of 46 months later (Blockmans et al 2008).
Another addition to the armamentarium for diagnosing vascular inflammation is 3-Tesla MRA, which can demonstrate not only segmental areas of lumen narrowing, but wall thickening and enhancement that are characteristic of large vessel vasculitis (Bley et al 2005).
Among 59 patients with giant cell arteritis who had both high-resolution MRI and color-coded duplex sonography of the temporal arteries within 10 days of each other and in the first 2 weeks of corticosteroid treatment, investigators in Hamburg found rapidly declining sensitivity of both procedures to demonstrate arteritic changes in the vessels during just the first few days of corticosteroid treatment. Temporal artery biopsy was done in 41 patients and was positive in 24 of these (59%). Comparing duplex exam/MRI against biopsy results in these 24 patients, the sensitivity was 92%/90% after 0 to 1 day of steroid treatment, 80%/78% after 2 to 4 days of treatment and 50%/64% among patients who had received corticosteroids for more than 4 days when studied. The results were very similar when comparing MRI and duplex sensitivity against final diagnosis in the larger group of 59 patients (Hauenstein et al 2012).