Clinical Symptoms and Associated Vascular Imaging Findings in Takayasu’s Arteritis Compared to Giant Cell Arteritis

Abstract

Objective:

To compare the presence of head, neck, and upper extremity symptoms in patients with Takayasu’s (TAK) and giant cell arteritis (GCA) and their association with vascular inflammation assessed by ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET) or arterial damage assessed by magnetic resonance angiography (MRA).

Methods:

Patients with TAK and GCA underwent clinical and imaging assessments within 24 hours, blinded to each other. Vascular inflammation was defined as arterial FDG-PET uptake greater than liver by visual assessment. Arterial damage was defined as stenosis, occlusion, or aneurysm by MRA. Clinically reported symptoms were compared to corresponding imaging findings using generalized mixed model regression. Cranial symptoms were studied in association with burden of arterial disease in the neck using ordinal regression.

Results:

Participants with TAK (n=56) and GCA (n=54) contributed data from 270 visits. Carotidynia was reported only in patients with TAK (21%) and was associated with vascular inflammation (p<0.01) but not damage (p=0.33) in the corresponding carotid artery. Posterior headache was reported in TAK (16%) and GCA (20%) but was only associated with corresponding vertebral artery inflammation and damage in GCA (p<0.01). Arm claudication was associated with subclavian artery damage (p<0.01) and inflammation (p=0.04) in TAK and with damage in GCA (p<0.01). Patients with an increased burden of damaged neck arteries were more likely to experience positional lightheadedness (p<0.01) or a major central nervous system event (p=0.01).

Conclusion:

The distribution of symptoms and association with imaging abnormalities differs in patients with TAK and GCA. These findings may help clinicians predict associated FDG-PET and MRA findings based on a specific clinical symptom.

INTRODUCTION

TAK and GCA are the two major forms of large vessel vasculitis (LVV), defined by vascular inflammation, with resultant damage of the aorta and branch arteries [1, 2]. Assessment of disease activity can be challenging in LVV, as there is a wide range of vascular symptoms that could be due to ongoing vascular inflammation, vascular damage, or both. The same symptom could be attributed to active disease or damage depending in part on the chronicity of the symptom [3].

Vascular imaging of the aorta and primary branches can be useful to assess arterial damage and inflammation in patients with LVV [4]. FDG-PET can demonstrate metabolic activity in the arterial wall and thus can be used as a potential surrogate marker for arterial inflammation [5, 6]. In contrast, magnetic resonance angiography (MRA) is useful to assess luminal damage, including stenosis, occlusion, or aneurysm [7].

Currently, there is no standardized reference and no reliable biomarkers to assess and measure disease activity in LVV. Although there are composite disease activity scores such as the Indian Takayasu Activity Score (ITAS) and Disease Extent Index-Takayasu Arteritis (DEI-Tak) [8, 9], these tools cannot precisely predict early signs of inflammation, progression of vascular disease, or end-organ vascular damage. Further complicating matters, active vascular inflammation can be detected in patients with LVV otherwise in clinical and biochemical remission. Postmortem histological studies show the significant presence of vascular inflammation in patients considered to be in remission at the time of death [10]. Additionally, persistent disease activity can be detected by FDG-PET and MRA during periods of apparent clinical remission in many patients, indicating discordance between symptoms and imaging abnormalities in LVV [11,12].

Involvement of the medium and large arteries in the head, neck, and upper extremities is common in both TAK and GCA and is associated with frequent ocular, intra- and extra-cranial manifestations including headache, lightheadedness, carotidynia, vision loss, stroke, transient ischemic attack (TIA), syncope, and upper limb claudication [13–21]. There is limited evidence comparing the presence of specific symptoms to corresponding angiographic abnormalities by both FDG-PET and MRA in LVV. Such data may inform clinicians about the relationship between specific symptoms and underlying vascular inflammation by FDG-PET or damage by MRA.

The study objectives were: (1) to compare the occurrence of common cranial and upper extremity vascular symptoms to multimodal vascular imaging assessment in LVV; (2) to explore if there are differences in the associations between vascular symptoms and imaging abnormalities in TAK versus GCA.

METHODS

Study population

Patients were included from an ongoing prospective observational cohort study at the National Institutes of Health (NIH) in Bethesda, MD (ClinicalTrials.gov identifier: NCT02257866). Patient visits were included in the study from September 2014 through September 2018. All patients with LVV fulfilled the American College of Rheumatology (ACR) 1990 criteria for the classification of TAK [22] or modified ACR 1990 criteria for the classification of GCA [18, 23]. Patients with LVV were enrolled at various stages during the disease. Whenever possible, baseline visit imaging studies were performed during periods of clinically active disease or during remission when the patient was taking <10 mg/day of prednisone to minimize potential confounding effects of treatment.

Patient and public involvement

The authors declare no direct patient and public involvement in study design.

Clinical assessment

All patients with LVV underwent clinical assessment with categorization of symptoms one day prior to imaging assessment. Symptoms present on the day of assessment were recorded, including carotidynia, posterior neck pain, frontotemporal and posterior headache, and arm claudication. Additionally, historical presence of selected symptoms at any point in the disease course were recorded, including lightheadedness, positional lightheadedness, carotidynia, vertigo, frontotemporal and posterior headache, blurred vision, vision loss, and major CNS events defined as stroke, TIA, or syncope.

FDG-PET Imaging Assessment

PET studies were performed at each study visit within 24 hours after clinical assessment, as previously reported [12]. Patients aged <18 years underwent whole-body FDG-PET-magnetic resonance imaging (MRI) with a Siemens Biograph mMR (Siemens Medical Solutions). Patients aged ≥18 years underwent FDG-PET-computed tomography (CT) of the torso with a Siemens Biograph mCT (Siemens Medical Solutions). A nuclear medicine physician (MAA) interpreted all PET scans included in this study blinded to clinical information and MRA data. Qualitative assessment of PET activity was performed in the carotid, vertebral, and subclavian arteries. The degree of arterial uptake was visually assessed relative to liver uptake [24, 25]. Vascular inflammation on PET in a specific arterial territory was defined as arterial FDG uptake greater than the FDG uptake in the liver by visual inspection.

MRA assessment

All patients underwent three-station MRA of the aorta and primary branches at each study visit within 24 hours after clinical assessment as previously described [12]. A vascular radiologist (JM) interpreted all MRAs, blinded to clinical data and PET scan assessment. Vascular damage on MRA was defined as stenosis, occlusion or aneurysm in an arterial territory of interest. Wall thickness and edema on MRA were not included in the definition of vascular damage in this study.

Statistical analysis

Symptoms present on the day of assessment were compared to arterial involvement by imaging studies in corresponding territories. For example, left-sided carotidynia was compared to FDG uptake in the left carotid artery by PET and to left carotid damage by MRA. Performance characteristics of clinical symptoms were detailed in association with corresponding imaging findings. In the context of this study, high sensitivity means that the absence of symptoms is likely to correspond to normal imaging studies (low false negatives) and high specificity means that the presence of symptoms is likely to correspond to abnormal imaging studies (low false positives). Generalized linear mixed model regression analysis adjusting for repeated study visits, daily prednisone dose, and use of additional immunosuppressant medication was used to examine the relationship between symptoms and imaging modalities in TAK and GCA. P values < 0.05 were considered statistically significant.

Historical presence of symptoms at any point in the disease course was compared to the burden of arterial disease in the four major neck arteries (carotid and vertebral arteries). Ordinal logistic regression was performed to determine the association between each neurovascular symptom and the number of affected neck arteries (range 0 to 4 arteries).

Ethics and informed consent

All patients provided written informed consent. An institutional review board and radiation safety committee at the NIH approved the research (Protocol 14-AR-0200).

RESULTS

Demographic characteristics of the study population

A total of 110 patients with LVV were recruited into the study. There were 56 patients with TAK and 54 patients with GCA. A total of 105 FDG-PET and 102 MRA were performed in TAK, and 134 FDG-PET and 129 MRA were performed in GCA. Among patients with GCA, 56% had a positive temporal artery biopsy and 76% had involvement of the large arteries by FDG-PET or MRA. Baseline demographic information of the study population is shown in Table 1. Seventy nine out of 110 (72%) patients were studied during clinically active disease, and 70/110 (64%) patients were taking prednisone <10 mg/day at the baseline visit.

Table 1.

Demographic characteristics of the study population

Diagnosis	TAK	GCA
Number of subjects	56	54
Number of visits	123	147
Number of visits per patient
1	26 (46%)	22 (40 %)
2–3	23 (41%)	16 (30 %)
>3	7 (13%)	16 (30 %)
Age, years (mean ± SD)	33.2 ± 12.1	69.9 ± 8.9
Female (%)	45 (80.4%)	44 (78.6%)
Disease duration, years (mean ± SD)	10.6 ± 9.8	2.9 ± 2.4
Prednisone (%)	33 (59%)	44 (81%)
Daily prednisone dose (mg, mean ± SD)	7.4 ± 12.2	9.0 ± 14.1
Other medications, Methotrexate	27 (48%)	27 (50%)
Tocilizumab	8 (14%)	21 (39%)
Infliximab	17 (30%)	1 (2%)
Azathioprine	9 (16%)	3 (6%)
Other	20 (36%)	6 (11%)

TAK, Takayasu’s arteritis; GCA, giant cell arteritis; SD, standard deviation

Frequency of clinical symptoms in patients with TAK and GCA

The occurrence of symptoms in patients with TAK and GCA is compared in Table 2. No significant differences in the frequency of symptoms were observed between patients with TAK and GCA for 6 of the 11 symptoms of interest. There was significantly more carotidynia (21 vs 0%, p<0.01), lightheadedness (30 vs 9%, p<0.01), positional lightheadedness (29 vs 5%, p<0.01), major CNS events (25 vs 9%, p=0.04), and arm claudication (52 vs 28%, p=0.01) in patients with TAK compared to GCA. The most common symptom in patients with TAK was arm claudication (52%) and in patients with GCA was blurred vision (37%).

Table 2.

Frequency of clinical symptoms in patients with TAK and GCA

Symptom	TAK (n=56)	GCA (n=54)	P-value
Carotidynia	12 (21%)	0 (0%)	<0.01
Lightheadedness	17 (30%)	5 (9%)	<0.01
Positional lightheadedness	16 (29%)	3 (5%)	<0.01
Posterior neck pain	4 (7%)	10 (18%)	0.09
Frontotemporal headache	14 (25%)	17 (31%)	0.53
Posterior headache	9 (16%)	11 (20%)	0.63
Vertigo	3 (5%)	5 (9%)	0.48
Blurred vision	18 (32%)	20 (37%)	0.69
Vision loss	6 (11%)	4 (7%)	0.74
Major CNS event	14 (25%)	5 (9%)	0.04
Arm claudication	29 (52%)	15 (28%)	0.01

TAK, Takayasu’s arteritis; GCA, giant cell arteritis; major CNS event = central nervous system event defined as stroke, transient ischemic attack, or syncope.

Prevalence of cephalic and upper extremity arterial involvement in TAK and GCA

The prevalence of arterial disease detected by FDG-PET or by MRA was compared between patients with TAK and GCA (Table 3). In general, there was more vascular inflammation in patients with GCA and more vascular damage in patients with TAK. In the carotid arteries, there was more PET activity in patients with GCA compared to TAK (52 vs 31%, p=0.05) and more arterial damage in patient with TAK compared to GCA (69 vs 33%, p<0.01). In the vertebral arteries, there was more PET activity in patients with GCA compared to TAK (19 vs 2%, p=0.05), but vertebral artery damage was not significantly different in patients with TAK compared to GCA (31 vs 17%, p=0.17). In the subclavian arteries, there was more PET activity in patients with GCA compared to TAK (50 vs 28%, p=0.03), but subclavian artery damage was not significantly different in patients with TAK compared to GCA (63 vs 52%, p=0.32).

Table 3.

Frequency of cephalic and upper extremity arterial disease detected by FDG-PET or MRA in patients with TAK and GCA.

Artery	Imaging study	TAK	GCA	P-value
Carotid	FDG-PET	17/54 (31%)	28/54 (52%)	0.05
	MRA	35/51 (69%)	17/52 (33%)	<0.01
Vertebral	FDG-PET	1/54 (2%)	10/54 (19%)	<0.01
	MRA	16/52 (31%)	9/52 (17%)	0.17
Subclavian	FDG-PET	15/54 (28%)	27/54(50%)	0.03
	MRA	32/51 (63%)	27/52(52%)	0.32

TAK, Takayasu’s arteritis; GCA, giant cell arteritis; FDG-PET, 18F-fluorodeoxyglucose positron emission tomography; MRA, magnetic resonance angiography

Comparison of specific vascular symptoms to imaging studies in TAK and GCA

Carotidynia and posterior neck pain

Carotidynia on the day of evaluation was reported during 12 out of 123 study visits (8%) in patients with TAK. Carotidynia was not reported by any patient with GCA. Carotidynia was more strongly associated with inflammation of the carotid artery by FDG-PET (p<0.01) than carotid artery damage by MRA (p= 0.33) in patients with TAK (Table 4, Figure 1A and 1B). Sensitivity was low for the association of carotidynia with FDG-PET or MRA abnormalities (27% and 11% respectively) indicating that an absence of carotidynia could still be associated with imaging abnormalities in the carotid artery, particularly on MRA compared to FDG-PET. Specificity was high for both FDG-PET and MRA (96% and 95% respectively) indicating that the presence of carotidynia was strongly associated with corresponding carotid artery abnormalities on both FDG-PET and MRA.

Table 4.

Comparison of specific vascular symptoms to imaging studies in patients with TAK and GCA

Symptom	Artery	LVV	Image	TN	TP	FN	FP	P- value	Sensitivity	Specificity
Carotidynia	Carotid	TAK	PET	166	10	27	7	<0.01	27% (14–44)%	96% (92–98)%
			MRA	90	12	95	5	0.33	11% (6–19)%	95% (88–98)%
		GCA	PET	No patients with GCA had carotidynia
			MRA
Posterior Neck pain	Vertebral	TAK	PET	208	0	1	3	1.00	0% (0–97)%	98% (96–100)%
			MRA	159	0	42	3	0.78	0% (0–8)%	98% (95–100)%
		GCA	PET	244	3	22	1	<0.01	12% (3–31)%	100% (98–100)%
			MRA	229	0	25	4	0.51	0% (0–14)%	98% (96–100)%
Posterior Headache	Vertebral	TAK	PET	206	0	1	4	1.00	0% (0–97)%	98% (95–99)%
			MRA	159	1	40	3	0.79	2% (0–13)%	98% (95–100)%
		GCA	PET	238	3	21	4	<0.01	12% (3–32)%	98% (96–99)%
			MRA	229	3	22	4	<0.01	12% (2–31)%	98% (96–99)%
FrontotemporalH eadache	Carotid	TAK	PET	165	4	33	8	0.33	11% (3–25)%	95% (91–98)%
			MRA	92	7	99	3	0.74	7% (3–13)%	97% (91–99)%
		GCA	PET	149	7	63	18	0.20	10% (4–19)%	89% (83–93)%
			MRA	163	11	73	13	0.21	13% (7–22)%	93% (88–96)%
Arm claudication	Subclavian	TAK	PET	142	8	19	42	0.04	29% (14–50)%	77% (70–83%)
			MRA	89	42	60	9	<0.01	41% (31–51)%	91% (83–96)%
		GCA	PET	129	25	81	32	0.57	23% (16–33)%	80% (73–86)%
			MRA	126	49	75	10	<0.01	40% (31–49)%	93% (87–96)%

TN = True Negative; TP = True Positive; FN = False Negative; FP = False Positive, PET, positron emission tomography; MRA, magnetic resonance angiography; LVV, large vessel vasculitis; TAK, Takayasu’s arteritis; GCA, giant cell arteritis. P values were derived by mixed model regression adjusting for repeated measures, daily prednisone dose, and use of additional immunosuppressant medication (yes/no).

Figure 1.

Clinical images from patients with large-vessel vasculitis. A patient with Takayasu’s arteritis who complained of left sided carotidynia had corresponding vascular inflammation by FDG-PET (A) without vascular damage by MRA (B). Fused FDG-PET and angiography images from a patient with giant cell arteritis who complained of a left sided posterior headache demonstrate increased FDG uptake (red) and throughout a stenotic left vertebral artery (white arrows) (C).

Posterior neck pain was more commonly reported by patients with GCA than TAK [10 (18%) vs 4 (7%), p=0.09]. Posterior neck pain contributed to 6% of the total visits in patients with GCA and significantly associated with inflammation of the vertebral artery by FDG-PET in patients with GCA (p<0.01) but not TAK (p=1.00) (Table 4). Posterior neck pain was not significantly associated with vertebral damage by MRA in either patients with GCA or TAK (p=0.78 and p=0.51 respectively). Sensitivity was low (range 0–12%) for the association of posterior neck pain with FDG-PET or MRA abnormalities in patients with TAK and GCA. Specificity was excellent (range 98–100%) for the association between posterior neck pain and both FDG-PET and MRA in patients with TAK and GCA, indicating the presence of posterior neck pain was strongly associated with corresponding vertebral artery abnormalities on both FDG-PET and MRA.

Posterior and frontotemporal headache

Posterior headache was reported in 5% of the total visits of patients with GCA and was significantly associated with both vertebral artery damage (p<0.01) and PET activity (p<0.01) in patients with GCA (Figure 1C). No association of posterior headache with vertebral PET activity (p=1.00) or damage (p=0.79) was observed in patients with TAK (Table 4). Sensitivity was low (range 0–12%) and specificity was high (98%) for the association between posterior headache and imaging abnormalities by FDG-PET or MRA in patients with TAK and GCA.

Frontotemporal headache was present in 7 out of the 123 total visits (5.7%) of patients with TAK and 15 out of the 147 total visits (10.2%) of patients with GCA. Frontotemporal headache was not associated with either carotid PET activity or damage in patients with TAK or GCA (Table 4). Sensitivity was low (range 7–13%) and specificity was high (range 89–97%) for the association between frontotemporal headache and FDG-PET or MRA findings in patients with TAK and GCA.

Arm claudication

Arm claudication at the time of clinical assessment was a highly prevalent symptom, present in 37% and 23% of the total visits in patients with TAK and GCA respectively. Arm claudication was more strongly associated with damage to the subclavian arteries by MRA (p<0.01) than inflammation by FDG-PET (p=0.04) in patients with TAK (Table 4). Similarly, in patients with GCA, arm claudication was significantly associated with damage to the subclavian arteries by MRA (p<0.01) but was not associated with PET activity (p=0.57). Sensitivity was moderate (range 23–40%) and specificity was high (range 77–93%) for the association between arm claudication and abnormalities by FDG-PET or MRA in patients with TAK and GCA. Presence of arm claudication had higher sensitivity in association with MRA findings compared to PET findings in both diseases.

Association of cephalic and neck symptoms with burden of arterial neck disease in patients with LVV

The association between specific clinical symptoms and the number of affected neck arteries was assessed. Stronger associations were observed between presence of clinical symptoms and burden of arterial disease by MRA rather than by FDG-PET (Figure 2A and 2B). Patients with LVV who had increased number of damaged neck arteries detected by MRA were significantly more likely to experience lightheadedness (OR=2.61, p=0.04), positional lightheadedness (OR=3.51, p<0.01), or a major CNS event (OR=3.23, p=0.01) at some point in their disease course. Patients with LVV who had increased number of inflamed neck arteries detected by FDG-PET were only significantly more likely to experience posterior headache (OR=2.84, p=0.03) at some point in the course of their disease.

Figure 2.

Odds ratio plots comparing the presence of cranial symptoms at any point during the disease course to the cumulative burden of vascular damage by MRA (A) or vascular inflammation by FDG-PET (B) in the four major neck arteries. Associations of each symptom with the burden of neck vessel disease in the carotid and vertebral arteries are displayed. Odds ratios >1 indicate that a specific symptom is associated with increased odds of vascular disease affecting increasingly more of the neck arteries. VL, vision loss; BV, blurred vision; PHA, posterior headache; FHA, frontotemporal headache; CNS, central nervous system event defined as stroke, TIA or syncope; CD, carotidynia; PLH, positional lightheadedness; LH, lightheadedness; MRA, magnetic resonance angiography; FDG-PET, fluorodeoxyglucose positron emission tomography.

DISCUSSION

Distinct associations between clinical symptoms and corresponding imaging pathology by FDG-PET and MRA were observed in a prospective, longitudinal observational cohort of patients with LVV. Certain vascular symptoms on the day of clinical and imaging assessment were more closely aligned with abnormal FDG-PET activity, angiographic damage, both, or neither. Some of these associations differed significantly between patients with TAK and GCA. Increased burden of neck arterial disease was associated with an increased likelihood of major CNS events. Clinical symptoms were not sensitive markers of underlying vascular pathology but were specific when present. Vascular imaging should be considered in the management of these patients since reliance on the presence of clinical symptoms may not be sensitive to detect vascular pathology within an acceptable window to prevent or minimize damage.

FDG-PET and MRA can be used to assess different aspects of vascular pathology, and imaging findings on these modalities correlate differently to specific symptoms. Symptoms of vascular pain, e.g. carotidynia, were significantly associated with FDG-PET abnormalities rather than angiographic damage and can therefore be considered more likely to reflect vascular inflammation. In contrast, arm claudication was only weakly associated with PET activity in patients with TAK but was strongly associated with angiographic damage in both patients with TAK and GCA. Symptoms of carotidynia could therefore be considered a stronger indicator of corresponding active vascular inflammation as compared to symptoms of limb claudication, which more strongly reflected vascular damage.

There were similarities and differences in the clinical and vascular imaging associations observed in patients with TAK compared to GCA. Symptoms related to carotid artery involvement aligned to carotid artery imaging findings only in patients with TAK, while symptoms related to vertebral artery involvement aligned to corresponding vertebral artery imaging findings only in patients with GCA. Posterior neck pain and posterior headache was associated with vertebral artery imaging abnormalities in patients with GCA. While posterior headaches in the occipital region are uncommon in patients with GCA [26, 27], this study emphasizes that presence of a posterior headache should alert the clinician to the likelihood of associated vascular inflammation and damage in the corresponding vertebral artery.

In contrast to posterior headache, frontotemporal headache was not associated with disease activity or damage of the carotid artery in TAK nor GCA. While frontotemporal headaches frequently occur in patients with TAK [2, 28], and are a cardinal feature of GCA [26], headaches in this region may reflect inflammation in smaller branches of cranial arteries, rather than the corresponding larger arteries of the neck. Additionally, most patients in this study were studied several years into the course of disease, when frontotemporal headaches may be less specific for vasculitis compared to a potentially stronger association at the time of diagnosis.

Presence of clinical symptoms aligned with corresponding imaging abnormalities by both FDG-PET and MRA. High specificity was observed between clinical symptoms and corresponding vascular imaging findings. In the context of this study, high specificity indicated that when a patient reported a specific head and neck symptom, there was often a corresponding vascular imaging abnormality. Thus, new symptoms of head, neck and upper limbs should prompt consideration of vascular imaging in order to confirm new inflammation or damage in corresponding arterial territories.

However, these results also suggest that the absence of clinical symptoms does not necessarily rule out underlying imaging pathology. Low sensitivity was observed between clinical symptoms and both FDG-PET and MRA, indicating that imaging abnormalities were frequently detected in the absence of corresponding symptoms. In the context of FDG-PET, these findings align with several prior studies that demonstrated that subclinical vascular inflammation is relatively common in patients with LVV [29–33]. In the context of angiography, presence of vascular damage in the absence of accompanying clinical symptoms underscores the importance of angiography for categorizing disease extent. Progression of vascular damage is often insidious and accompanied by compensatory collateral circulation [34–36], which may explain how patients with LVV can sometimes develop profound vascular damage in the absence of ischemic symptoms.

The current study has some limitations. First, most patients were studied during later phases of disease and not necessarily at time of diagnosis. Many patients were on treatment for vasculitis which may have weakened the associations between symptoms and imaging findings, in particular the FDG-PET results. Although statistical analysis adjusted for differences in glucocorticoid use, residual confounding from treatment effect is possible as only daily rather than cumulative glucocorticoid exposure was considered. More vascular inflammation was observed in patients with GCA compared to TAK which could be due to biologic differences, shorter disease duration in the patients with GCA, or differences in concomitant atherosclerosis. A large, international study also reported more FDG-PET activity in GCA and more angiographic damage in TAK in data collected at the time of diagnosis [37]. Symptoms were defined as present or absent on the day of assessment, but the duration of those symptoms was not considered. New or worsening claudication may have a different relationship to imaging findings compared to persistent claudication.

The study has several important strengths. MRA and FDG-PET were performed on the same day within 24 hours of clinical assessment. Imaging assessments were performed by central readers independent of clinical assessment. This study does not dictate how angiography and FDG-PET should be used in a clinical setting. Rather, these findings may help clinicians predict imaging pathology in specific vascular territories based on patient-reported symptoms and may inform which type of imaging modality would be the most useful to obtain in certain clinical scenarios, recognizing that additional sequences to detect wall morphology may augment the ability of MR-based assessments to detect vascular inflammation in addition to luminal damage [12]. Additionally, these findings may facilitate the development and refinement of disease activity indices in LVV for use in future clinical research studies as clinical features that are more strongly associated with vascular inflammation than damage may be preferentially considered.

TAK and GCA are complex diseases that pose clinical management challenges, particularly at later stages of disease when accurate assessment of disease activity can be difficult. Findings from this study support the concept that clinical assessment should be integrated with imaging assessment in order to facilitate clinical care and research in these conditions.

KEY MESSAGES.

What is already known about this subject?

• Angiography is useful to detect vascular damage, and 18F-fluorodeoxyglucose positron emission tomography is useful to detect vascular inflammation in patients with large-vessel vasculitis.

What does this study add?

• This study details the relationships between clinical features commonly reported by patients with Takayasu’s arteritis and giant cell arteritis with corresponding vascular imaging findings by MRA and FDG-PET.

• Presence of specific symptoms such as carotidynia more closely align with vascular inflammation by FDG-PET and presence of other symptoms such as claudication are more tightly linked to vascular damage by MRA.

• Absence of clinical symptoms does not rule out corresponding imaging pathology in patients with large-vessel vasculitis.

How might this impact on clinical practice or future developments?

• These findings will enable clinicians to predict imaging pathology based on patient-reported symptoms and will help researchers develop and refine disease activity indices in LVV.