Diagnosing temporal arteritis: duplex vs. biopsy
From Kaiser Permanente Northwest, Vancouver, USA
Address correspondence to Dr M.S. Alberts, Vancouver Medical Office, 2211 East Mill Plain Blvd, Vancouver, Washington 98661, USA. email: michael.s.alberts{at}kp.org
Received 4 May 2007 and in revised form 6 August 2007
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Temporal artery biopsy is the traditionally-accepted method of diagnosing temporal arteritis, but is of limited sensitivity.
Aim: To compare the clinical decisions made after negative temporal artery biopsy vs. negative temporal artery duplex, and the effects on patient outcomes.
Design: Retrospective analysis.
Methods: Of 290 patients suspected of having temporal arteritis, 147 underwent bilateral temporal artery duplex with a negative result, and 143 underwent unilateral temporal artery biopsy with a negative result. These groups were compared. Dependent measures included the proportion of patients in each group whose steroids were discontinued by their primary care doctor after either negative test, and the difference in the number of alternative diagnoses considered after a negative test. The incidence of blindness in each group was also compared, as a measure of adverse outcomes. Patients were then stratified by pre-test probability of having the disease, and compared using the same measures.
Results: Equivalent proportions of patients in the two groups had steroids discontinued after a negative test result, even when further stratified into risk groups by the probability of having temporal arteritis. No differences in adverse outcomes or number of alternative diagnoses considered were noted between groups.
Discussion: In clinical practice, bilateral temporal artery duplex served the same function as biopsy, but without subjecting patients to the potential morbidity of a surgical procedure. Temporal artery biopsy could be reserved only for situations where the duplex result is inconsistent with the clinical picture, and the biopsy result, if different from the duplex result, might influence the treatment decision.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Temporal artery biopsy has been the traditionally-accepted method of diagnosing temporal arteritis. Its sensitivity, however, has been questioned. This is primarily due to the nature of the disease which is characterized by skip lesions, and the limited portion of the temporal artery amenable to biopsy.1,2 Additionally, the procedure of temporal artery biopsy, though largely well tolerated and with minimal risk, has been associated with some morbidity.3 For these reasons, it has remained a useful tool in diagnosis, but never accepted as a solitary indicator.
Clinical assessment has been proposed as the closest equivalent to a reference standard. The American College of Rheumatology (ACR) developed research criteria in 1990 as a tool for the classification of temporal arteritis.4 In this system, biopsy is one of five final equally-weighted clinical and demographic criteria for the diagnosis of the disease, thereby allowing a positive diagnosis even with a negative biopsy. The presence of any three of the five final criteria, yields a sensitivity of 93.5% and a specificity of 91.2% when compared against patients who were clinically classified as having the disease. The authors of the ACR classification noted that 15/211 patients who met the criteria for temporal arteritis had negative biopsies (a false negative rate of 7.1%).
Although temporal artery biopsy has remained a reference criteria for diagnosis when positive, its limitations have motivated some to explore other diagnostic modalities. Schmidt and colleagues were the first to demonstrate the association between a periarterial inflammatory halo on duplex and positive histology from temporal artery biopsy.5 Others have noted that when either an inflammatory halo (defined as an hypoechoic or anechoic region surrounding the perfused lumen of the temporal artery or branches for a discrete region), segmental arterial stenosis, or both were present, sensitivity, specificity, positive predictive value, and negative predictive value compared with histology were 100%, 80%, 58.3% and 100%, respectively.6 This high sensitivity favours duplex as a beneficial screening evaluation, particularly when compared with the limited sensitivity of biopsy. At present, duplex is used primarily as an adjunct to biopsy in the diagnostic process, rather than as a replacement.7
What is not clear, however, is how the information derived from duplex is being used by clinicians. To date, no published study has specifically evaluated clinician decision-making and associated clinical outcomes when duplex replaces temporal artery biopsy as the initial diagnostic measure for suspected temporal arteritis. Our purpose was to record actual clinical behaviour and subsequent patient outcomes in such an algorithm. The assumption was that using duplex as an initial technique can only be valuable if: (i) clinicians accepted it as a reliable alternative to biopsy as an initial measure, and (ii) adverse outcomes (e.g. blindness) were not more frequent. If these two criteria were met, duplex was performing a similar function to biopsy as an initial measure with less morbidity and cost. Our intent was to observe clinical decisions made after a negative diagnostic study (either duplex or biopsy), and to observe whether the type of study done made a difference in treatment decisions. We also wanted to ascertain whether any negative outcomes (blindness) occurred when clinicians chose to discontinue steroids after either negative study (i.e. did they miss a true diagnosis of temporal arteritis by falsely relying on a negative result).
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In September 2003, the diagnostic algorithm for patients suspected of having temporal arteritis was changed by joint agreement of the Departments of Rheumatology, General Surgery and Radiology in the Kaiser Permanente Northwest Region, a large managed-care organization serving Northwest Oregon and Southwest Washington in the US. All patients suspected of having temporal arteritis who were referred to the Department of General Surgery for temporal artery biopsy, underwent bilateral colour flow duplex of their temporal arteries prior to consideration of biopsy. The exam was conducted by a technician using a 5–15 Mhz linear probe (depending on the depth of the vessel), and read by a vascular surgeon experienced in non-invasive vascular studies of extracranial arteries.
Group 1 comprised the first 147 patients who had a negative bilateral duplex exam after beginning the new algorithm. A negative study was defined as the absence of any abnormality noted on duplex. Specifically, no inflammatory halo, segmental stenosis (velocity of a single segment double the velocity of a prior segment), or absence of flow (occlusion) was noted. All duplex exams were conducted by vascular radiology technicians skilled in the procedure. No patient had received more than 48 h of steroid treatment prior to the duplex.
The comparison group (group 2, n = 143) comprised historical control patients referred for temporal artery biopsy for the three years prior to initiation of the new duplex algorithm (June 2000 to September 2003). All patients referred for evaluation of temporal arteritis were considered. Each patient underwent unilateral biopsy of the temporal artery on the side most symptomatic. Only those with a negative result were included. None went on to contralateral biopsy, as none was requested. No patient had been treated with steroids for more than 48 hs. One patient experienced an episode of angina during the procedure, requiring brief hospitalization. One scalp haematoma requiring drainage occurred. No other complications were noted.
All comparisons were between the two groups.
Clinician confidence
Clinician confidence in a negative result was measured by comparing the number of patients whose steroids were discontinued after either negative study (duplex or biopsy). The mean number of alternative diagnoses considered after a negative study (either duplex or biopsy) was also compared as a way to indirectly detect differences in clinician decision making between groups. The hypotheses were that, if clinicians lacked confidence in the test result (i.e. considered it a false negative), they would continue steroid treatment and have fewer alternative explanations for the clinical picture.
The Pearson 
2 test was used for non-continuous variables; ANOVA was used for continuous variables. The number of patients who subsequently experienced the most adverse outcome of a missed diagnosis (blindness) was compared between the groups. Follow-up averaged 2 years for group 1, and 5.2 years for group 2 (historical controls).
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The demographic, clinical, and laboratory characteristics of the two groups are displayed in Table 1. Equivalency of group membership was demonstrated. Age and gender, presence of headache (and whether the headache was unilateral), presence of jaw claudication, any visual disturbance, and mean erythrocyte sedimentation rate were roughly equivalent between groups. Only the presence of any temporal artery abnormality on physical exam (tenderness, pulselessness, swelling) was more frequently noted in the historical controls, and the difference was only just significant (p = 0.050). These results support the assumption that the two groups, other than being separated as historical controls by roughly three years, were essentially equivalent.
|
At the time of referral, more patients in the biopsy group (historical controls) were begun on steroids (79% vs. 62.6%, p = 0.002, Table 2). However, of those patients who had steroid treatment initiated, an equivalent number of patients’ steroids were discontinued after a negative study in the duplex and biopsy groups (90% vs. 86.5%). This suggested that the type of negative diagnostic test was not an independent factor in the decision to discontinue steroids. Clinicians continued steroids in 10% of the patients in group 1 and 13.5% of the patients in group 2, suggesting that either the negative test result was not considered accurate, or that the patient could have a different disease treatable with steroids (i.e. polymyalgia rheumatica).
|
With regard to the number of alternative diagnoses considered by the primary care doctor after a negative study, no differences were found between groups (Table 3). There were no cases of blindness noted in either group.
|
A few differences emerged when the two groups were subdivided by pre-test probability of having the disease (risk groups, Table 4). Low risk groups included patients with 1–2 criteria, medium risk patients had 3–4 criteria, high risk patients had 5–6 criteria (age greater than 50, presence of unilateral headache, jaw claudication, visual disturbance, temporal artery abnormality on exam, or elevated sedimentation rate). More historical controls (Group 2 patients) in the medium risk category were begun on steroids at the time of referral (76 vs. 62, p = 0.016, Table 5), perhaps suggesting greater clinical confidence in the diagnosis prior to the diagnostic test ordered. However, risk group stratification did not alter the proportion of patients tapered from steroids after a negative duplex or biopsy result (Table 6), demonstrating equivalent confidence in either diagnostic test.
|
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Our results suggest that in clinical practice a negative duplex is considered valid and is therefore used in the same way as a negative biopsy. Equivalent numbers of patients’ steroids were discontinued in each group, without significant adverse outcome in either. The usefulness (and perceived validity) of either diagnostic test for clinicians was, therefore, equivalent. There was no increase in adverse outcomes when duplex was used instead of biopsy, suggesting no missed positive diagnoses in either group possibly resulting in untreated arteritis. We made no attempt here to evaluate the sensitivity or specificity of either test (this has been debated elsewhere8,9).
Positive biopsy is, by definition, the gold standard of diagnosis of temporal arteritis when it is positive. However, with poor sensitivity when using clinical diagnosis as the comparison reference (false negative rates reported as high as 44%10), biopsy is, in actual practice, rarely the factor that establishes the diagnosis of temporal arteritis, although it continues to be widely advocated. Careful clinical evaluation must also be used. It has been postulated that clinical evaluation may be the most accurate diagnostic technique, and that information from biopsy adds to diagnostic yield only in a minority of cases.11
Consideration of pre-test clinical suspicion did not have any influence on the clinicians’ use of test results. As expected, the majority of patients referred for evaluation were in the intermediate risk category (Table 4). In this study, even when patients were stratified by risk group (number of clinical criteria), steroids were discontinued after either type of negative study in the same proportion of patients. A majority of patients’ steroids were stopped, suggesting that test results from both studies were trusted. As expected, a smaller proportion of patients’ steroids were discontinued in the medium and high risk groups, compared to in the low risk group.
Based on these results, whether to continue to use biopsy at all remains a relevant question. Others have suggested that, despite the continued enthusiasm for biopsy, it may in fact play an insignificant role in the decision to treat or to discontinue treatment.12 In our study group, biopsy was of equivalent value to duplex, but exposed the patient to additional potential morbidity.
Based on these results, it may be that biopsy should be used only in very limited circumstances when reliable duplex is also available, rather than routinely. There may be times when biopsy will, in fact, influence the decision to treat. Our opinion, after observing the experience of clinicians here, incorporates some variation of the recommendations made recently by Karassa et al.13 in their meta analysis. All patients suspected of having temporal arteritis should receive bilateral duplex evaluation by an experienced radiology team. Low probability patients with a positive duplex probably need a biopsy to demonstrate that the duplex was a true positive, in order to justify long-term steroid treatment. In this situation, the area of abnormality can be specifically marked and targeted for biopsy. A negative duplex in this group should be considered justification to stop steroid treatment. High probability patients should be treated, and no test is required prior to treatment. Negative studies in such patients should not necessarily preclude treatment. However, there may be a subgroup of patients with high clinical suspicion who do not have the disease, and these patients would be receiving steroids unnecessarily. For them, the risk of long-term steroids vs. the risk of blindness must be individually assessed by the clinician.
For intermediate probability patients, a duplex result in the setting of an experienced radiology team, should provide enough evidence to determine whether to discontinue steroids, and most of these patients should not need further evaluation with biopsy. No adverse outcomes were experienced in this study by intermediate probability patients whose steroids were stopped after a negative exam. Positive duplex is enough evidence to continue treatment.
In summary, we recommend bilateral temporal artery duplex as the initial assessment in all patients for whom the clinical picture is unclear (low and intermediate clinical probability) followed by biopsy only when a positive duplex result is inconsistent with the clinical picture. Temporal artery biopsy may remain useful, but only in limited circumstances.
|
Acknowledgments |
|---|
The authors would like to acknowledge Sharise LaValley and Mary Wellwood for their invaluable assistance.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Taylor-Gjevre R, Vo M, Shukla D, Resch L. Temporal artery biopsy for giant cell Arteritis. J Rheum (2005) 32:1279–82.
2. Allsop CJ, Gallagher PJ. Temporal artery biopsy in giant-cell arteritis. A reappraisal. Am J Surg Path (1981) 5:317–23.[Web of Science][Medline]
3. Romera-Billegas 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) 23:1493–8.
4. Hunder GG, Bloch DA, Michel BA, et al. The American College of Rheumatology 1990 criteria for the classification of giant cell arteritis. Arthritis Rheum (1990) 33:1122–8.[Web of Science][Medline]
5. Schmidt W, Kraft H, Vorpahl K, Volker L, Gromnica-Ihle E. Color duplex ultrasonography in the diagnosis of temporal arteritis. N Engl J Med (1997) 337:1336–42.
6. LeSar CJ, Meier GH, DeMasi RJ, et al. The utility of color duplex ultrasonography in the diagnosis of temporal arteritis. J Vasc Surg (2002) 36:1154–60.[CrossRef][Web of Science][Medline]
7. Butteriss DFA, Clarke L, Dayan M, Birchall D. Use of colour duplex ultrasound to diagnose giant cell arteritis in a case of visual loss of uncertain aetiology. Br J of Radiol (2004) 77:607–9.[CrossRef]
8. Salvarani C, Silingardi M, Ghirarduzzi A, et al. Is duplex ultrasonography useful for the diagnosis of giant-cell arteritis? Ann Int Med (2002) 137:232–8.
9. Pfadenhauer K, Weber H. Duplex sonography of the Temporal and Occipital artery in the diagnosis of temporal arteritis. A prospective study. J Rheum (2003) 30:2177–81.
10. Ashton-Key MR, Gallagher PJ. False-negative temporal artery biopsy [letter]. Am J Surg Pathol (1992) 16:634–5.[CrossRef][Web of Science][Medline]
11. Varma D, O’Neill D. Quantification of the role of temporal artery biopsy in diagnosing clinically suspected giant cell arteritis. Eye (2004) 18:384–8.[CrossRef][Web of Science][Medline]
12. Lenton J, Donnelly R, Nash JR. Does temporal artery biopsy influence the management of temporal arteritis. Q J Med (2006) 99:33–6.[Web of Science]
13. Karassa FB, Miltiadis IM, Schmidt WA, Ioannidis JPA. Meta-analysis: Test performance of ultrasonography for Giant-Cell arteritis. Ann Int Med (2005) 142:359–69.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||