Abstract
Objective:
Diagnostic assessment in giant cell arteritis (GCA) is rapidly changing as vascular imaging becomes more available. The study objective was to determine if clinical subsets of GCA have distinct profiles or reflect differential diagnostic assessment.
Methods:
Patients were recruited from an international cohort and divided into four subsets based on a temporal artery (TA) abnormality [positive TA biopsy (TAB) or halo-sign on TA ultrasound (TA-US)] and/or evidence on imaging of large-vessel (LV) involvement: 1) both TA abnormality and LV involvement [TA(+)/LV(+) GCA]; 2) TA abnormality without large-vessel involvement [TA(+)/LV(−) GCA]; 3) large-vessel involvement without TA abnormality [TA(−)/LV(+) GCA]; and 4) clinically-diagnosed GCA without large-vessel involvement or TA abnormality [TA(−)/LV(−) GCA].
Results:
941 patients with GCA were recruited from 72 international study sites. Most patients had multiple forms of diagnostic assessment, including TAB (n=704, 75%); TA-US (n=328, 35%); and LV-imaging (n=534, 57%). Assessment confirmed GCA for 66% TAB, 79% TA-US, and 40% of LV-imaging studies. GCA subsets had distinct profiles independent of diagnostic assessment strategies. TA(+)/LV(−) was the most common subset (51%) with a high burden of cranial ischemia. TA(−)/LV(−) (26%) had a high prevalence of cranial ischemia and musculoskeletal symptoms. TA(−)/LV(+) (12%) had prevalent upper extremity vascular abnormalities and a low prevalence of vision loss. TA(+)/LV(+) (11%) was older with a high prevalence of cranial ischemia, constitutional symptoms, and elevated acute phase reactants.
Conclusion:
Vascular imaging is increasingly incorporated into the diagnostic assessment of GCA and identifies clinical subsets of patients based on involvement of temporal and extracranial arteries.
Keywords: vasculitis, large-vessel vasculitis, giant cell arteritis, angiography, positron emission tomography, ultrasonography
Giant cell arteritis (GCA) is a rare form of vasculitis that causes inflammation within the medium and large arteries. GCA is a heterogeneous disease (1) in which symptoms and the extent of arterial involvement often vary between patients. GCA was traditionally thought to be confined to the cranial arteries; however, many patients with GCA also have evidence of large-vessel involvement (2–5). Patients with large-vessel involvement often present with different clinical features than patients with cranial GCA, but the extent to which patients have overlapping versus distinct cranial and extra-cranial disease is not well characterized (5–9).
A temporal artery biopsy (TAB) with mononuclear cell infiltrate or granulomatous inflammation remains the gold standard for the diagnosis of GCA. Although highly specific by definition, the sensitivity of TABs has decreased over time highlighting that many patients with GCA are diagnosed without histological evidence of arteritis (10–13). Patients suspected of having GCA are increasingly diagnosed using temporal artery ultrasound (TA-US) (14). A homogenous, hypoechoic wall thickening of the temporal artery, termed as the “halo sign”, has been proposed as a diagnostic equivalent of a positive TAB (15). Additionally, large-vessel imaging has become an increasingly common form of diagnostic assessment and clinical evaluation in GCA. A substantial percentage of patients with GCA, with and without a positive TAB, have large-vessel involvement with estimates ranging from 20-67% by angiography, 83% by 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) imaging, and 100% by autopsy (2–5, 16–18).
Patients can receive a diagnosis of GCA based on a combination of cranial and large-vessel assessment. With the expansion of diagnostic modalities in GCA, the extent to which cranial and large-vessel GCA are distinct entities or if co-occurrence is under detected is poorly understood. The objectives of this study were to 1) detail the assessment strategies currently utilized to diagnose GCA using data collected within a large, international, observational cohort; 2) determine if subsets of GCA, defined by temporal artery biopsy (TAB) and vascular imaging, are associated with distinct clinical profiles or merely reflect differential diagnostic testing.
METHODS
Patient selection
The Diagnostic and Classification Criteria in Vasculitis (DCVAS) is an international, observational cohort of patients with vasculitis (19, 20). Patients were enrolled within two years from the time of diagnosis. Only physician-submitted cases with a diagnosis of GCA confirmed by expert panel review were included in this study.
Clinical variables
Comprehensive demographic, clinical, and vascular-imaging data were collected using standardized forms. Data was available for the following clinical aspects of disease: visual changes and other symptoms of cranial ischemia; pulmonary, musculoskeletal, and constitutional symptoms; vascular abnormalities; and laboratory findings. All variables were recorded only if present by the time of diagnosis.
Vascular abnormalities
Temporal artery physical examination abnormalities were defined as tenderness, diminished or absent pulse, and/or a nodular, cord-like temporal artery. Blood pressure was recorded in both arms and blood pressure differences between the arms were categorized as less than or greater than 10 mmHg. Pulse abnormality was defined as diminished or absent pulse in either the upper limbs (subclavian, axillary, brachial, or radial arteries) or lower limbs (femoral, popliteal, posterior tibial, or dorsalis pedis arteries). Bruits in the abdominal aorta, carotid, subclavian, axillary, or renal arteries was recorded.
Laboratory findings
The maximum values of the erythrocyte sedimentary rate (ESR) and C-reactive protein (CRP) by the time of diagnosis were recorded. Anemia was defined as hemoglobin <10g/dL, hypoalbuminemia as albumin <30g/L, and thrombocytosis as platelets >500 x 109 per liter.
Temporal artery biopsies
Temporal artery biopsies were performed at the discretion of the physician. A positive biopsy was defined as showing active vasculitis following local review if recorded as such by the submitting physician.
Large-vessel and temporal artery imaging
Ultrasonography of the temporal arteries was performed in a subset of patients at the discretion of the submitting physician. A positive ultrasound was defined by the presence of halo sign recorded by the submitting physician.
Large-vessel involvement was assessed in a subset of patients by a combination of angiography [magnetic-resonance, computed-tomography, or catheter-based], ultrasonography, or 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) imaging. Clinical radiologists or nuclear medicine physicians at each participating institution assessed vascular imaging data. Angiographic and ultrasound damage was defined as stenosis, occlusion, or aneurysm. Wall thickness was not included in the definition of arterial damage. FDG-PET scans were assessed according to local practices (21). Large-vessel involvement was defined as having arterial damage or abnormal FDG-uptake attributed to vasculitis per the opinion of the investigator within 13 arterial territories of interest: right and left carotid, subclavian, axillary, iliofemoral, and renal arteries; mesenteric arteries; and ascending, descending, and abdominal aorta.
Subset definitions
Patients were divided into four subsets based on temporal artery biopsy (TAB), temporal-artery ultrasound (TA-US), and/or large-vessel (LV) imaging findings. Patients with evidence of both large-vessel involvement on imaging and either a positive TAB or halo-sign on TA-US were classified as TA(+)/LV(+) GCA. Patients with a TAB with definite vasculitis or halo-sign on TA-US without evidence of large-vessel involvement on imaging were classified as TA(+)/LV(−) GCA. Patients with evidence of large-vessel involvement on imaging without a positive TAB or halo sign on TA-US were classified as TA(−)/LV(+) GCA. Patients with a clinical diagnosis of GCA without evidence of large-vessel involvement on imaging, positive TAB, or halo sign on TA-US were classified as TA(−)/LV(−) GCA. Clinical differences were studied between the four subsets.
Restricted cohort
Since TAB, TA-US, and large-vessel imaging assessment was performed at the discretion of the treating physician, patients could have received different diagnostic assessments. To determine the extent that differences between the subsets were related to clinical differences versus diagnostic bias from differential assessments, analyses were conducted in a restricted cohort of patients that received both temporal-artery assessment (TAB or TA-US) and large-vessel imaging and compared to data from the whole cohort.
Statistical methods
Chi-square, Mann-Whitney, and Kruskal-Wallis tests were used, as appropriate. A p value <0.05 defined statistical significance. Nominal logistic regression was used to study the associations between the clinical decision to undergo a specific diagnostic test and clinical features of disease. All analyses were performed using Graphpad Prism V.7.0a (GraphPad, La Jolla, California, USA).
RESULTS
Subject characteristics
A total of 941 patients with GCA were included, representing 72 study sites, 26 countries, and 5 continents. Demographic information is listed in Table 1.
Table 1.
Demographics and diagnostic assessment of subsets of giant cell arteritis
| Overall n = 941 |
TA(−)/LV(−) n = 245 |
TA(+)/LV(−) n = 480 |
TA(+)/LV(+) n = 100 |
TA(−)/LV(+) n = 116 |
|
|---|---|---|---|---|---|
| Age (mean ± SD)* | 72.6 ± 9.0 | 71.0 ± 8.5 | 74.1 ± 9.2 | 74.2 ± 7.9 | 68.6 ± 8.6 |
| Sex (% female)† | 634 (67.4) | 173 (70.9) | 309 (64.2) | 64 (64.0) | 88 (75.9) |
| Continent | |||||
| Europe | 783 (83.2) | 186 (23.8) | 412 (52.6) | 92 (11.8) | 93 (11.9) |
| North America | 111 (11.8) | 38 (34.2) | 50 (45.1) | 5 (4.5) | 18 (16.2) |
| Asia | 29 (3.1) | 15 (51.7) | 7 (24.1) | 3 (10.3) | 4 (13.8) |
| Oceania | 18 (1.9) | 5 (27.8) | 12 (66.7) | 0 (0.0) | 1 (5.6) |
| Diagnostic assessment | |||||
| TAB | 705 (75.0) | 174 (71.0) | 430 (89.6) | 74 (74.0) | 27 (23.3) |
| TA-US | 328 (35.0) | 22 (9.0) | 225 (46.9) | 61 (61.0) | 20 (17.2) |
| LV imaging | 534 (57) | 90 (36.7) | 228 (47.3) | 100 (100) | 116 (100) |
| Angiography | 227 (24.1) | 54 (22.1) | 65 (13.5) | 52 (52.0) | 56 (48.3) |
| Ultrasonography | 307 (32.6) | 36 (14.8) | 160 (33.3) | 58 (58.0) | 53 (45.7) |
| FDG-PET | 175 (18.6) | 21 (8.6) | 29 (6.0) | 45 (45.0) | 80 (69.0) |
| Comprehensive | 431 (45.8) | 59 (24.1) | 227 (47.3) | 100 (100) | 45 (38.8) |
TA(−)/LV(−): patients without evidence of temporal artery or large-vessel involvement; TA(+)/LV(−): patients with temporal artery involvement; TA(+)/LV(+): patients with both temporal artery and large-vessel involvement; TA(−)/LV(+): patients with large-vessel involvement;
p<0.001;
p=0.049; TAB: temporal artery biopsy; TA-US: temporal artery ultrasound; LV imaging: large-vessel imaging; Comprehensive diagnostic assessment: patients who underwent both temporal artery and large-vessel assessments
Diagnostic assessment strategies in the DCVAS cohort
Patients received vascular assessments at the discretion of the submitting physician. Overall, 705 (75%) had a TAB, 328 (35%) patients had TA-US, 534 (57%) had large-vessel imaging, and 431 (45.8%) had both large-vessel and temporal-artery assessment (Table 1). Only 33 patients (3.5%) did not have any diagnostic testing performed beyond clinical assessment (Supplementary Table 1).
Diagnostic assessment strategies by demographics
The use of the various diagnostic assessments differed between patients based on demographics, including age and sex. Male patients were more likely to undergo a TAB compared to female patients (Male: 81.4% vs Female: 71.6%; p=0.001). In logistic regression models, male patients were 34% more likely to have a TAB compared to female patients independent of symptoms of cranial ischemia, age at diagnosis, temporal-artery abnormality on physical examination, and large-vessel involvement on imaging (Supplementary Table 2). There was no difference by sex in the rate of TA-US (Male: 36.8% vs Female: 33.9%, p=0.38) or large-vessel imaging (Male: 57.7% vs Female: 56.3%, p=0.70). Male and female patients were equally likely to have a positive TAB (Male: 61.6% vs Female: 64.3%, p=0.51) and evidence of large-vessel involvement on imaging (Male: 36.2% vs Female: 42.6%, p=0.16). Male patients were more likely to have a halo-sign on TA-US than female patients (Male: 85.0% vs 75.4%, p=0.05).
Patients who underwent a TAB were older (73.4 years vs 70.5 years, p<0.0001) and patients who had large-vessel imaging were younger (71.9 years vs 73.8 years, p=0.0009). There was no difference in age between patients that did and did not have a TA-US (72.8 years vs 72.6 years, p=0.71).
Diagnostic assessment by continent
Diagnostic assessment differed between patients based on the continent they were seen. The rate of TAB did not differ between continents (North America: 82.8%, Europe: 73.1%, Asia: 82.8%, or Oceania defined as the islands of the Central and South Pacific including Australia: 88.9%) (Figure 1). Patients from European centers were more likely to have underwent TA-US (North America: 2.7%, Europe: 41.0%, Asia: 13.8%, Oceania: 0.0%). Large-vessel imaging was performed in about half of patients from North American, European, and Asian centers, but less commonly done in Oceanian centers (North America: 58.6%, Europe: 57.7%, Asia: 51.7%, Oceania: 11.1%).
Figure 1. Diagnostic assessment of giant cell arteritis by continent.
Percent of patients with giant cell arteritis who underwent a temporal artery biopsy (TAB), temporal artery ultrasound (TA-US), or large-vessel imaging (LV-imaging) from North American, European, Asian, and Oceanian study centers.
Patients with multiple forms of diagnostic assessment
Most patients had multiple forms of diagnostic assessment. The majority of patients who underwent a TAB had additional vascular imaging regardless of the TAB result (TAB positive: 57.2%; TAB negative: 56.2%) (Supplementary Table 1). Vascular imaging was performed in 86.1% of patients who did not have a TAB and confirmed disease in 70.0% of these patients (Supplementary Table 1, Figure 2). TA-US was rarely used alone as a form of vascular assessment. Only 1.1% of patients had TA-US without another form of vascular assessment compared to 10.9% of patients that had only LV-imaging and 32.3% of patients that had only TAB.
Figure 2. Diagnostic assessment of giant cell arteritis by temporal artery biopsy results.
Large-vessel imaging (LV-imaging) and temporal artery ultrasound (TA-US) results stratified by whether a temporal-artery biopsy (TAB) was diagnostic for giant cell arteritis (Positive TAB), non-diagnostic for giant cell arteritis (Negative TAB), or was not performed (Did not have TAB). A positive TA-US was defined by presence of a halo-sign on TA-US. Positive LV-imaging was defined by evidence of vasculitic involvement of the aorta or branch arteries on angiography, ultrasonography, or FDG-PET imaging.
The results of TAB were not associated with large-vessel imaging or TA-US results (Figure 2). Patients with a negative or positive TAB were equally likely to have evidence of large-vessel involvement on imaging (Positive TAB: 31% vs Negative TAB: 27%, p=0.54). Similarly, patients with a negative or positive TAB were equally likely to have a halo sign on TA-US (Positive TAB: 82% vs Negative TAB: 76%, p=0.38).
Clinical subsets in giant cell arteritis in the whole cohort
Subsets of patients were defined by cranial and large-vessel involvement. Among the whole cohort, patients were characterized into one of four subsets as follows: 480 (51.0%) patients as TA(+)/LV(−) GCA; 245 (26.0%) as TA(−)/LV(−) GCA; 116 (12.3%) as TA(−)/LV(+) GCA; and 100 (10.6%) as TA(+)/LV(+) GCA. Demographics and assessment strategies in the whole cohort for each subset are provided in Table 1.
Clinical subsets in giant cell arteritis in the restricted cohort
Clinical subsets of GCA were characterized in a restricted cohort of patients who had assessment of both the temporal arteries and the large arteries. Out of these 431 patients (46% of the whole cohort), 340 (79%) patients had TAB, 258 (60%) patients had TA-US, and all patients had large-vessel imaging. Assessment of both the temporal and large arteries was more common in patients who were older, male, and presented with fever, weight loss, syncope, scalp tenderness, arm claudication, upper extremity pulse abnormality, or elevated acute phase reactants (p<0.01). All four clinical subgroups were again observed in the restricted cohort, including 59 (14%) patients with TA(−)/LV(−) GCA, 227 (53%) patients with TA(+)/LV(−) GCA, 100 (23%) patients with TA(+)/LV(+) GCA, and 45 (10%) patients with TA(−)/LV(+) GCA (Table 2).
Table 2.
Clinical associations in GCA patients with both temporal-artery and large-vessel assessment
| TA(−)/LV(−) n = 59 |
TA(+)/LV(−) n = 227 |
TA(+)/LV(+) n = 100 |
TA(−)/LV(+) n = 45 |
P-value | |
|---|---|---|---|---|---|
| Age (mean ± SD) | 69.34 ± 7.54 | 73.67 ± 7.98 | 74.17 ± 7.93 | 70.31 ± 9.32 | <0.001 |
| Female sex (%) | 41 (69.49) | 142 (62.56) | 64 (64.0) | 28 (62.2) | 0.794 |
| Diagnostic assessment | |||||
| Positive TAB | 0/54 | 151/185 | 68/74 | 0/27 | <0.001 |
| Halo sign on TA-US | 0/14 | 148/163 | 55/61 | 0/20 | <0.001 |
| Temporal artery abnormality | |||||
| Temporal artery abnormality | 20 (33.90) | 145 (63.88) | 51 (51.0) | 7 (15.56) | <0.001 |
| Visual changes | 19 (32.2) | 82 (36.1) | 25 (25.0) | 7 (15.6) | 0.03 |
| Amaurosis fugax | 7 (11.86) | 37 (16.30) | 6 (6.00) | 4 (8.89) | 0.06 |
| Sudden visual loss | 9 (15.25) | 34 (14.98) | 9 (9.0) | 1 (2.22) | 0.06 |
| Blurred vision | 8 (13.56) | 38 (16.74) | 16 (16.0) | 5 (11.11) | 0.77 |
| Ischemic cranial symptoms | 53 (89.8) | 201 (88.6) | 78 (78.0) | 24 (53.3) | <0.001 |
| Jaw claudication | 25 (42.37) | 125 (55.07) | 42 (42.0) | 10 (22.22) | <0.001 |
| Scalp tenderness | 22 (37.29) | 68 (29.96) | 21 (21.0) | 10 (22.22) | 0.11 |
| Headache | 49 (83.05) | 178 (78.41) | 69 (69.0) | 20 (44.44) | <0.001 |
| Pulmonary symptoms | 6 (10.2) | 24 (10.6) | 16 (16.0) | 9 (20.0) | 0.22 |
| Dyspnea | 4 (6.78) | 10 (4.41) | 9 (9.0) | 2 (4.44) | 0.40 |
| Non-productive cough | 3 (5.08) | 16 (7.05) | 8 (8.0) | 8 (17.78) | 0.08 |
| Musculoskeletal symptoms | 28 (47.5) | 88 (38.8) | 45 (45.0) | 22 (49.9) | 0.41 |
| Morning stiffness | |||||
| >1 hour | 16 (27.12) | 34 (14.98) | 14 (14.00) | 7 (15.56) | 0.12 |
| Neck or torso | 8 (13.56) | 19 (8.37) | 8 (8.00) | 8 (17.78) | 0.17 |
| Hip and thighs | 12 (20.34) | 33 (14.54) | 15 (15.0) | 6 (13.33) | 0.71 |
| Muscle tenderness | 11 (18.64) | 17 (7.49) | 13 (13.00) | 1 (2.22) | 0.01 |
| Myalgia | 17 (28.81) | 67 (29.52) | 33 (33.0) | 12 (26.67) | 0.87 |
| Vascular abnormalities | 24 (40.7) | 58 (25.6) | 37 (37.0) | 18 (40.0) | 0.03 |
| Arm claudication | 5 (8.47) | 1 (0.44) | 1 (1.0) | 6 (13.33) | <0.001 |
| Leg claudication | 8 (13.56) | 6 (2.64) | 4 (4.0) | 2 (4.44) | 0.005 |
| Any bruit | 9 (15.25) | 18 (7.93) | 16 (16.0) | 8 (17.78) | 0.06 |
| Pulse abnormality | |||||
| Upper limb | 2 (3.39) | 3 (1.32) | 4 (4.0) | 8 (17.78) | <0.001 |
| Lower limb | 4 (6.78) | 20 (8.81) | 15 (15.0) | 4 (8.89) | 0.27 |
| Absent blood pressure | 0 (0.0) | 0 (0.0) | 1 (1.0) | 1 (2.2) | 0.17 |
| Blood pressure difference | 14 (23.7) | 31 (13.7) | 15 (15.0) | 11 (24.4) | 0.12 |
| Constitutional symptoms | 33 (55.9) | 117 (51.5) | 68 (68.0) | 28 (62.2) | 0.04 |
| Night sweats | 12 (20.34) | 52 (22.91) | 29 (29.00) | 13 (18.89) | 0.49 |
| Fever (>38C) | 15 (25.42) | 45 (19.82) | 30 (30.0) | 13 (28.89) | 0.19 |
| Weight loss | 21 (35.59) | 80 (35.24) | 52 (52.0) | 20 (44.44) | 0.03 |
| Laboratory findings | 24 (40.7) | 66 (29.1) | 38 (38.0) | 15 (33.3) | 0.23 |
| ESR (mm/hr) | 75.8 + 34 | 69.2+31 | 80.9+32 | 80.6+31 | 0.009 |
| CRP (mg/L) | 51.7 (22-150) | 68 (34-128) | 75.9 (46-139) | 50 (26-90) | 0.06 |
| Anemia | 11 (18.64) | 25 (11.01) | 22 (22.0) | 15 (33.3) | 0.001 |
| Hypoalbuminemia | 12 (20.34) | 30 (13.22) | 24 (24.00) | 7 (15.56) | 0.10 |
| Thrombocytosis | 26 (27.12) | 35 (15.42) | 25 (25.0) | 11 (24.44) | 0.07 |
Subset diagnosed by temporal artery biopsy or ultrasound in the restricted cohort
Patients with TA(+)/LV(−) GCA had a high prevalence of temporal-artery abnormalities on physical examination, vision changes, and other symptoms of cranial ischemia, including amaurosis fugax, sudden ongoing visual loss, jaw claudication, scalp tenderness, and new-onset headache. Patients with TA(+)/LV(−) GCA were older and had a low prevalence of vascular abnormalities, laboratory abnormalities, and constitutional and musculoskeletal symptoms compared to other subsets (Table 2).
Subset diagnosed by clinical symptoms without confirmatory histology or imaging findings in the restricted cohort
Patients with TA(−)/LV(−) GCA were most similar to patients with TA(+)/LV(−) GCA (Table 2, Figure 3). When compared to TA(+)/LV(−) GCA, patients with TA(−)/LV(−) had a similarly high prevalence of vision changes and other symptoms of cranial ischemia, including sudden ongoing visual loss, jaw claudication, scalp tenderness, and headache. Patients with TA(−)/LV(−) had the highest prevalence of musculoskeletal symptoms including morning stiffness greater than one-hour, morning stiffness in the hips and thighs, and muscle tenderness. Despite normal LV assessment by imaging, a proportion of patients with TA(−)/LV(−) had leg claudication (14%), arterial bruits (15%), and blood pressure differences between the arms >10mmHg (24%).
Figure 3. Clinical characteristics of subsets of giant cell arteritis in the restricted cohort.
Differences are shown in the prevalence of (A) ischemic cranial symptoms, (B) vascular abnormalities, (C) constitutional symptoms, and (D) musculoskeletal symptoms for each giant cell arteritis subset. TA(−)/LV(−): patients without evidence of temporal artery or large-vessel involvement; TA(+)/LV(−): patients with evidence of temporal artery involvement without evidence of large-vessel involvement; TA(−)/LV(+): patients with evidence of large-vessel involvement without evidence of temporal artery involvement; TA(+)/LV(+): patients with evidence of both temporal artery and large-vessel involvement.
Subset diagnosed by large-vessel imaging abnormalities in the restricted cohort
Patients with TA(−)/LV(+) GCA had significantly fewer symptoms of cranial ischemia and temporal artery abnormalities compared to the other subsets (Table 2, Figure 3). Patients with TA(−)/LV(+) GCA had a high prevalence of vascular symptoms (i.e. arm claudication) and examination findings (i.e. vascular bruit, upper limb pulse abnormality, blood pressure difference between arms >10mmHg), pulmonary symptoms (i.e. non-productive cough), and constitutional symptoms (i.e. fever, weight loss).
Subset diagnosed by temporal-artery and large-vessel imaging abnormalities in the restricted cohort
The clinical profile of TA(+)/LV(+) GCA had features similar to TA(+)/LV(−) GCA and TA(−)/LV(+) GCA, as well as features that were not seen in either subset (Figure 3, Table 2). Patients with TA(+)/LV(+) GCA had a high prevalence of temporal artery abnormalities and ischemic cranial symptoms, significantly higher than TA(−)/LV(+) GCA but lower than TA(+)/LV(−) GCA, including jaw claudication, scalp tenderness, and new onset headache. Although the overall prevalence of symptoms of cranial ischemia was similar between TA(+)/LV(+) GCA and TA(+)/LV(−) GCA, patients with TA(+)/LV(+) GCA were less likely to have visual changes including amaurosis fugax and sudden ongoing visual loss. Patients with TA(+)/LV(+) GCA and TA(−)/LV(+) GCA had a similarly high prevalence of vascular abnormalities including bruit and lower-limb pulse abnormality, and constitutional symptoms (i.e. night sweats, fever, weight loss) (Figure 3, Table 2). However, TA(+)/LV(+) GCA had significantly less vascular symptoms and examination abnormalities of the upper limbs compared to TA(−)/LV(+) GCA, including arm claudication and pulse abnormalities. Patients with TA(+)/LV(+) GCA had a high prevalence of laboratory abnormalities, including elevated acute-phase reactants (Table 2).
Clinical subsets of GCA in the whole versus restricted cohort
A summary of the clinical subsets observed in the restricted cohort is provided in Figure 4. Overall, the clinical associations of the subsets of GCA in the whole cohort were consistent with the restricted cohort with only minor differences (Supplementary Table 3). There was no difference in prevalence of the TA(+)/LV(−) GCA (51% vs 53%, p=0.57) and TA(−)/LV(+) GCA (12% vs 10%, p=0.31) subgroups in the whole cohort compared to the restricted cohort. Patients in the restricted cohort were less likely to have TA(−)/LV(−) GCA (14% vs 26%, p<0.0001) and more likely to have TA(+)/LV(+) GCA (23% vs 11%, p<0.0001). Unlike the restricted cohort, differences in sex were observed between the clinical subsets using data from the whole cohort (Table 1). Patients with TA(−)/LV(+) were more likely to be female and patients with TA(+)/LV(−) and TA(+)/LV(+) were more likely to be male. Patients with TA(−)/LV(−) GCA in the whole cohort had a significantly lower prevalence of leg claudication (Restricted Cohort: 13.6% vs Whole Cohort: 4.9%; p<0.0001) and arterial bruit (Restricted Cohort: 15.3% vs Whole Cohort: 7.0%; p=0.04) compared to TA(−)/LV(−) GCA patients in the restricted cohort (Supplementary Table 3).
Figure 4. Clinical profile of patients with giant cell arteritis by clinical subset in the restricted cohort.
Some of the defining clinical features of each giant cell arteritis clinical subset are displayed within the Venn diagram. TA(−)/LV(−): patients without evidence of temporal artery or large-vessel involvement; TA(+)/LV(−): patients with evidence of temporal artery involvement without evidence of large-vessel involvement; TA(−)/LV(+): patients with evidence of large-vessel involvement without evidence of temporal artery involvement; TA(+)/LV(+): patients with evidence of both temporal artery and large-vessel involvement.
Discussion
Diagnostic assessment in GCA is rapidly changing as non-invasive vascular imaging techniques become increasingly available to categorize the extent of arterial involvement. Data from this study demonstrate how clinicians across the world use different diagnostic assessment strategies to assess arterial involvement in GCA. Although TAB is still the most common form of diagnostic assessment, non-invasive techniques such as ultrasonography, angiography, and PET imaging are increasingly used to assess disease in GCA. Patients are now more frequently recognized who have isolated or overlapping cranial and large-vessel involvement. This study also details the characteristics of clinical subsets based on documented involvement of the cranial or large arteries. Besides differences in the extent of vascular involvement, patients in these subsets have different clinical profiles that likely reflect underlying biological differences. These findings should inform clinicians on the clinical variability among patients with GCA and aid in researching more homogenous patient populations.
Temporal artery biopsy remains the diagnostic gold standard for GCA; however, the sensitivity of TAB has declined in recent years (22). Decreasing sensitivity is due in part to the growing practice of using vascular imaging as a diagnostic surrogate to biopsy to detect arterial involvement in and beyond the temporal arteries. Three quarters of patients with a physician-confirmed diagnosis of GCA in the DCVAS cohort underwent a TAB to diagnose GCA, but a third of these biopsies were not diagnostic. Of those with a non-diagnostic TAB, 32.6% had evidence of vascular involvement by vascular imaging and 67.4% were diagnosed based on clinical symptoms alone. Vascular imaging was used as an alternative method to diagnose GCA in a quarter of patients who did not have a TAB.
Vascular imaging techniques are being used to complement TAB and to capture the full extent of arterial disease. Over half of the patients who had a TAB had additional vascular imaging, independent of TAB results. Approximately 50% of patients with a TAB had large-vessel imaging and about 30% had a TA-US. Regional practices strongly influenced diagnostic assessment strategies, as TA-US was performed mostly at European centers. Patients who had a TA-US also typically had a concomitant TAB, although recent recommendations suggest that TA-US may be a diagnostic surrogate to TAB in patients in whom there is a high clinical suspicion for GCA (23). Patients with a negative or positive TAB were equally likely to have a halo-sign on TA-US, suggesting that TA-US and TAB may identify different aspects of disease.
The increased use of both vascular imaging in both the assessment and evaluation of patients with suspected GCA has seemingly expanded the definition of GCA with more patients now being given this diagnosis. This evolution of clinical practice may result in not only a larger population of people given the diagnosis of GCA, but also a change in the relative prevalence of different clinical characteristics of what is labeled as GCA.
Clinical subsets in GCA have been proposed to account for the significant clinical heterogeneity seen among patients with GCA. Patients with large-vessel involvement often present with different clinical features than patients with cranial GCA, but the extent to which patients have overlapping versus distinct cranial and extra-cranial disease is not well characterized. In this study, clinical subsets were created a priori based on documented cranial and large-vessel involvement. Different clinical profiles were observed in patients with GCA based on whether or not disease was observed in the cranial or large arteries. Restriction of analyses to a subset of patients who had comprehensive temporal artery and large-vessel assessment confirmed that these clinical differences were not simply the product of differential diagnostic testing. However, some bias in diagnostic assessment was also observed. Women were less likely than men to undergo a biopsy of the temporal artery to confirm a diagnosis of GCA independent of differences in the presenting features of disease.
TA(+)/LV(−) GCA represents patients with a traditional GCA clinical profile. These patients had a high burden of symptoms of cranial ischemia and visual changes, and had limited evidence of large-vessel involvement. Patients with isolated large-vessel involvement differed substantially from patients with isolated temporal-artery involvement in their demographics and presenting clinical features. Patients with TA(−)/LV(+) GCA had a high burden of upper extremity vascular abnormalities, constitutional, and pulmonary symptoms.
A subset of patients with GCA have overlapping cranial and large-vessel abnormalities leading to speculation that the extent of arterial involvement may be under-detected in patients with isolated cranial or large-vessel involvement. More of these patients were identified in the restricted cohort where assessment of both the temporal and large arteries was performed. However, in the DCVAS cohort, patients with overlapping cranial and large-vessel involvement had unique clinical associations, confirming that patients with overlapping disease represent an independent subgroup that is not merely due to under detection of arterial involvement in the other subgroups. Patients with overlapping disease were older with a high prevalence of symptoms of cranial ischemia, similar to patients with isolated cranial disease, and had a high prevalence of vascular abnormalities, bruits, and constitutional symptoms, similar to patients with isolated large-vessel disease. Patients with overlapping disease had a high prevalence of laboratory abnormalities and constitutional symptoms. Similar to prior studies, patients with large-vessel involvement, including those with overlapping disease, were less likely than patients without large-vessel involvement to have visual changes (6).
In the DCVAS cohort, almost all patients had some form of arterial assessment, but a quarter of patients did not have diagnostic confirmation of involvement of either the cranial or extra-cranial arteries. The clinical profile of these patients was nearly indistinguishable from patients with TA(+)/LV(−) GCA. These patients also had an increased burden of polymyalgia rheumatica (PMR)-like symptoms and had vascular examination abnormalities and symptoms associated with the large arteries in absence of corresponding imaging findings. Even in the face of negative diagnostic assessment, these patients were diagnosed with GCA, most likely due to the suggestive pattern of symptoms of cranial ischemia and PMR-like symptoms.
This study has potential limitations to consider. Diagnostic and imaging assessment was not standardized across the cohort. Temporal artery or large-vessel involvement may have been under detected. However, when analyses were repeated in the restricted cohort, the same pattern of clinical symptoms was observed in the four subsets. Additionally, standardized definitions for a diagnostic TAB or halo sign on TA-US were not used and were defined at the discretion of the submitting physician. Large-vessel imaging was not standardized in the DCVAS cohort. Different rates of large-artery involvement in GCA have been described based on modality, and there is no gold standard for assessing large-vessel involvement in GCA. Large-vessel involvement may be overestimated due to difficulty distinguishing between atherosclerosis and vasculitis on imaging, particularly in the lower extremities. Treatment data was not available in the DCVAS cohort, and use of glucocorticoids in particular may have impacted the performance characteristics of diagnostic testing. There is circularity in subgrouping based on documented arterial involvement since vascular assessment may be driven by the presenting clinical symptoms. However, in the restricted cohort, where patients had cranial and large-vessel assessment, all four subgroups were observed with a similar pattern of clinical symptoms. The majority of participating centers in the DCVAS cohort are tertiary, research hospitals; therefore, diagnostic assessment may not be fully representative of general practice. Long-term outcome data was not available in the DCVAS cohort and might reveal other signs and symptoms of GCA and/or increased prevalence of involvement of the large arteries at later stages of disease.
With the widespread adoption of vascular imaging into clinical practice, clinical variability is increasingly appreciated among patients with GCA. Previous studies have shown that patients with GCA can differ in their systemic inflammatory response, extent of arterial involvement, and treatment response (6) and suggest that there are distinct subgroups of GCA with potentially divergent disease etiology. Identifying subgroups may lead to stratified clinical decision making and enable research into differences in disease pathology. In this study, subsets of patients with GCA have distinct clinical profiles based on involvement of the cranial or large arteries. The extent to which differences in biologic mechanisms of disease underlie these subsets is unknown. Prospective studies to assess potential differences in disease risk factors, response to treatment, and long-term outcomes among subsets of patients with GCA are warranted. Comprehensive vascular assessment of the cranial and large arteries should be considered in the diagnostic assessment of all patients with suspected GCA.
Supplementary Material
Acknowledgments
Financial supports and conflicts disclosure:
This study was supported by the Intramural Research Program at the National Institute of Arthritis and Musculoskeletal and Skin Diseases. The DCVAS study is sponsored and coordinated from the University of Oxford with funding support from the National Institute for Health Research Musculoskeletal Biomedical Research Unit, the University of Oxford, the Vasculitis Foundation, and the American College of Rheumatology and European League against Rheumatism. Further information about the study can be found at http://www.dcvas.org.
References
- 1.Buttgereit F, Dejaco C, Matteson EL, Dasgupta B. Polymyalgia Rheumatica and Giant Cell Arteritis: A Systematic Review. JAMA. 2016;315(22):2442–58. [DOI] [PubMed] [Google Scholar]
- 2.Garcia-Martinez A, Hernandez-Rodriguez J, Arguis P, Paredes P, Segarra M, Lozano E, et al. Development of aortic aneurysm/dilatation during the followup of patients with giant cell arteritis: a cross-sectional screening of fifty-four prospectively followed patients. Arthritis Rheum. 2008;59(3):422–30. [DOI] [PubMed] [Google Scholar]
- 3.Nuenninghoff DM, Hunder GG, Christianson TJ, McClelland RL, Matteson EL. Incidence and predictors of large-artery complication (aortic aneurysm, aortic dissection, and/or large-artery stenosis) in patients with giant cell arteritis: a population-based study over 50 years. Arthritis Rheum. 2003;48(12):3522–31. [DOI] [PubMed] [Google Scholar]
- 4.Cid MC, Prieto-Gonzalez S, Arguis P, Espigol-Frigole G, Butjosa M, Hernandez-Rodriguez J, et al. The spectrum of vascular involvement in giant-cell arteritis: clinical consequences of detrimental vascular remodelling at different sites. APMIS Suppl. 2009(127):10–20. [DOI] [PubMed] [Google Scholar]
- 5.Prieto-Gonzalez S, Arguis P, Garcia-Martinez A, Espigol-Frigole G, Tavera-Bahillo I, Butjosa M, et al. Large vessel involvement in biopsy-proven giant cell arteritis: prospective study in 40 newly diagnosed patients using CT angiography. Ann Rheum Dis. 2012;71(7):1170–6. [DOI] [PubMed] [Google Scholar]
- 6.van der Geest KSM, Sandovici M, van Sleen Y, Sanders JS, Bos NA, Abdulahad WH, et al. Review: What Is the Current Evidence for Disease Subsets in Giant Cell Arteritis? Arthritis Rheumatol. 2018;70(9):1366–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Muratore F, Kermani TA, Crowson CS, Green AB, Salvarani C, Matteson EL, et al. Large-vessel giant cell arteritis: a cohort study. Rheumatology (Oxford). 2015;54(3):463–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.de Boysson H, Liozon E, Lambert M, Dumont A, Boutemy J, Maigne G, et al. Giant-Cell Arteritis: Do We Treat Patients with Large-Vessel Involvement Differently? Am J Med. 2017;130(8):992–5. [DOI] [PubMed] [Google Scholar]
- 9.de Boysson H, Liozon E, Ly KH, Dumont A, Delmas C, Aouba A. The different clinical patterns of giant cell arteritis. Clin Exp Rheumatol. 2019;37 Suppl 117(2):57–60. [PubMed] [Google Scholar]
- 10.Luqmani R, Lee E, Singh S, Gillett M, Schmidt WA, Bradburn M, et al. The Role of Ultrasound Compared to Biopsy of Temporal Arteries in the Diagnosis and Treatment of Giant Cell Arteritis (TABUL): a diagnostic accuracy and cost-effectiveness study. Health Technol Assess. 2016;20(90):1–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hunder GG. Giant cell arteritis and polymyalgia rheumatica. Med Clin North Am. 1997;81(1):195–219. [DOI] [PubMed] [Google Scholar]
- 12.Hedges TR 3rd, Gieger GL, Albert DM. The clinical value of negative temporal artery biopsy specimens. Arch Ophthalmol. 1983;101(8):1251–4. [DOI] [PubMed] [Google Scholar]
- 13.Roth AM, Milsow L, Keltner JL. The ultimate diagnoses of patients undergoing temporal artery biopsies. Arch Ophthalmol. 1984;102(6):901–3. [DOI] [PubMed] [Google Scholar]
- 14.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(6):1224–6. [PubMed] [Google Scholar]
- 15.Chrysidis S, Duftner C, Dejaco C, Schafer VS, Ramiro S, Carrara G, et al. Definitions and reliability assessment of elementary ultrasound lesions in giant cell arteritis: a study from the OMERACT Large Vessel Vasculitis Ultrasound Working Group. RMD Open. 2018;4(1):e000598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Blockmans D, de Ceuninck L, Vanderschueren S, Knockaert D, Mortelmans L, Bobbaers H. Repetitive 18F-fluorodeoxyglucose positron emission tomography in giant cell arteritis: a prospective study of 35 patients. Arthritis Rheum. 2006;55(1):131–7. [DOI] [PubMed] [Google Scholar]
- 17.Ostberg G Morphological changes in the large arteries in polymyalgia arteritica. Acta Med Scand Suppl. 1972;533:135–59. [PubMed] [Google Scholar]
- 18.Ostberg G An arteritis with special reference to polymyalgia arteritica. Acta Pathol Microbiol Scand Suppl. 1973;237:Suppl 237:1–59. [PubMed] [Google Scholar]
- 19.Luqmani RA, Suppiah R, Grayson PC, Merkel PA, Watts R. Nomenclature and classification of vasculitis - update on the ACR/EULAR diagnosis and classification of vasculitis study (DCVAS). Clin Exp Immunol. 2011;164 Suppl 1:11–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Craven A, Robson J, Ponte C, Grayson PC, Suppiah R, Judge A, et al. ACR/EULAR-endorsed study to develop Diagnostic and Classification Criteria for Vasculitis (DCVAS). Clin Exp Nephrol. 2013;17(5):619–21. [DOI] [PubMed] [Google Scholar]
- 21.Slart R, Writing g, Reviewer g, Members of EC, Members of EI, Inflammation, et al. FDG-PET/CT(A) imaging in large vessel vasculitis and polymyalgia rheumatica: joint procedural recommendation of the EANM, SNMMI, and the PET Interest Group (PIG), and endorsed by the ASNC. Eur J Nucl Med Mol Imaging. 2018;45(7):1250–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rubenstein E, Maldini C, Gonzalez-Chiappe S, Chevret S, Mahr A. Sensitivity of temporal artery biopsy in the diagnosis of giant cell arteritis: a systematic review and meta-analysis. Rheumatology (Oxford). 2019;Sep14 [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
- 23.Dejaco C, Ramiro S, Duftner C, Besson FL, Bley TA, et al. EULAR recommendations for the use of imaging in large vessel vasculitis in clinical practice. Ann Rheum Dis. 2018;77(5):636–43. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.