Added diagnostic yield of temporal artery magnetic resonance angiography in the evaluation of giant cell arteritis

Abstract

Objectives

Temporal artery magnetic resonance angiography (TAMRA) is a useful tool to investigate possible diagnoses of GCA. As acquired images also reveal other local structures, they may assist in finding alternative diagnoses when assessing for possible GCA. We sought to assess the utility of TAMRA in identifying other significant abnormalities either associated with a diagnosis of GCA or potentially mimicking a clinical presentation of GCA.

Methods

A retrospective cohort study was undertaken at St Joseph’s Healthcare in Hamilton, Ontario, Canada between February 2007 and April 2020 and included patients who underwent TAMRA for a possible diagnosis of GCA. Patient demographics, diagnosis and imaging findings were extracted, and descriptive analysis of findings was performed.

Results

We included 340 individuals who underwent TAMRA for assessment of a potential diagnosis of GCA and had clinical information available; there were 126 (37.1%) diagnoses of GCA. Fourteen (4.1%) patients had findings on TAMRA that demonstrated an alternative diagnosis, findings were predominantly in the temporomandibular joint, orbit and meninges. Eighteen (14.3%) patients with GCA had intracranial vascular changes that were demonstrative of intracranial vasculitis; one stroke was attributed to intracranial GCA.

Conclusions

TAMRA has proven utility in diagnosing GCA, and these data suggest that it also has utility in identifying alternative diagnoses to rule out the disease. Intracranial vasculitis was also seen in 14.3% of patients; the clinical impact of these findings is currently poorly understood and requires further study.

Introduction

GCA is a large vessel vasculitis that typically involves the external carotid arteries, temporal arteries and their branches. It is the most common vasculitis of older adults, with a prevalence of up to 250 per 100 000 persons aged 50 years or older [1, 2]. GCA is often described with classic symptoms of jaw claudication, headache, and visual changes, but can also present with findings that are non-specific or insufficient for the assessing clinician to make the diagnosis. As such, GCA has many mimics that are important to exclude both for the risk of vision loss with undertreated GCA and the adverse effects of glucocorticoids if they are inappropriately prescribed [3].

Multiple investigations have demonstrated utility in assisting clinicians in clarifying the diagnosis of GCA. Temporal artery biopsy (TAB) has traditionally been the standard of diagnosis over several decades, though there is increasing evidence of its imperfect sensitivity [4, 5]. More recently, several imaging modalities including ultrasound, 18-fluorodeoxyglucose positron emission topography (FDG-PET) and temporal artery magnetic resonance angiography (TAMRA) have demonstrated utility in diagnosing GCA [6]. Furthermore, TAMRA and FDG-PET visualize anatomically adjacent structures that can assist in identifying alternative diagnoses or mimics [7].

Understanding the diagnostic yield of these additional findings can be beneficial to clinicians. This study sought to demonstrate additional vasculitic and non-vasculitic TAMRA findings in patients clinically suspected of having GCA to understand their value.

Methods

We conducted a retrospective cohort study of individuals referred to a rheumatologist for assessment of a possible diagnosis of GCA who underwent TAMRA as part of their investigations to determine their underlying diagnosis at a single centre (St Joseph’s Healthcare, Hamilton, Ontario, Canada). All individuals who were able to complete all imaging sequences of the TAMRA protocol and had a known final diagnosis of GCA or not GCA were included in the study. TAMRAs were performed between February 2007 and April 2020. There were no exclusion criteria for the analysis, but individuals who had incompatible metal implants, claustrophobia or an eGFR < 30 ml/min and were thus unable to undergo gadolinium-enhanced MRI were not included in the cohort. Final diagnosis was sought from the medical records of the patients’ treating rheumatologist and was based on clinical assessment; there were no classification criteria applied nor biopsy requirements for the diagnosis to be made. Demographic data including age and sex for patients were also collected. Information concerning the clinical presentation was not collected. Approval from the Hamilton Integrated Research Ethics Board was granted for this study (approval 10935). Consent for the study was waived as individuals were passively found from a list of TAMRAs completed; information concerning the diagnosis of GCA was completed as a secondary analysis of the data of Rhéaume et al. [8] as well as a local GCA research database (ethics approval 15-403-D).

Reports from imaging studies were reviewed for the presence or absence of findings of cranial GCA within the scalp vessels, findings suggestive of arteritis within the intracranial vessels, and other findings unrelated to arteritis. Details of the TAMRA protocol were previously published; a TAMRA was considered to have signs of vasculitis when there was evidence of wall thickening, post-contrast vessel wall enhancement and/or perivascular fat stranding [8]. Where there was evidence of intracranial vasculitis, the images were re-reviewed at the time of study data collection to ensure that reporting and results were consistent across the 13 years in which data were gathered, as recent publications suggest physiological vasa vasorum wall enhancement may be more frequent than previously thought [9]. If older images were thought to represent non-vasculitic lesions (e.g. physiological or atherosclerotic disease), the revised interpretation was used within this analysis.

Non-GCA diagnostic findings were considered explanatory by the assessing radiologist if it was thought that they may or may not be the cause of the patient’s symptoms or could result in a change in the patient’s management or follow-up. Findings were considered non-explanatory if they were thought to be unlikely to contribute to the patient’s presentation or did not require changes in patient follow-up or management. Non-vasculitic findings were categorized as intracranial vascular (including aneurysms and malformations), brain parenchymal, meningeal, non-brain neurological (neurological findings occurring outside of the brain, brainstem or cerebellum), orbital, sinus and musculoskeletal/temporomandibular joint (TMJ) findings.

Descriptive analysis was performed using mean and standard deviation as well as median and interquartile range where appropriate. Student’s t-test and the chi-square test were used for continuous and binary variables as appropriate. Statistical analysis was completed using SAS On Demand, reporting was completed in accordance with the STROBE statement, and P < 0.05 was considered significant by convention [10, 11].

Results

Cohort characteristics

Our cohort included 424 individuals who underwent TAMRA for a diagnosis of GCA, of whom 340 (80.2%) had complete data. One hundred and twenty-six (37.1%) were diagnosed with GCA. TAMRA was found to be 68.3% (95% CI: 63.3, 73.6%) sensitive and 94.4% (91.9, 97.3%) specific for a diagnosis of GCA. Those with GCA were older (P < 0.01); there was no difference in the proportion of individuals who were female (P = 0.26). For the 214 patients whose final clinical diagnosis was not GCA, a clear alternative diagnosis was available for 84 patients (39.3%).

Cranial involvement in GCA

Stratified by final diagnosis, 86 (68.3%) individuals with GCA were found to have superficial cranial involvement on TAMRA (Table 1). Evidence of intracranial involvement was also seen in 18 individuals (14.3% of all cases of GCA and 20.9% of all TAMRA-positive cases); there were no individuals with intracranial involvement who did not have arteritis in the scalp arteries. The intracranial arteries most often involved were the petrous and cavernous segments of the internal carotid artery (n = 16) and the intradural segment of the vertebral artery (n = 6) (Fig. 1). One patient had multiple infarctions that were attributed to active arteritis. Stenosis without enhancement or areas of eccentric enhancement were attributed to atherosclerosis, rather than arteritis, and were seen in two patients without GCA.

There were 12 patients whose TAMRA was reported as having findings suggestive GCA but who had an alternative final diagnosis. One of these patients had MRI findings suggestive of GCA as well as inflammatory pseudotumour of left orbit, the patient’s final diagnosis was optic perineuritis/pseudotumour. Two cases were diagnosed as PMR and there were no data indicating whether or not they had additional imaging that would suggest a diagnosis of extracranial large vessel vasculitis. Alternative diagnostic data were not available for the other nine TAMRAs with findings of arteritis that were ultimately deemed to not have GCA.

Non-arteritic TAMRA findings

Non-arteritic vascular findings included 26 aneurysms (P = 0.90 for those with and without GCA), four cases of atherosclerotic disease, two arterio-venous fistulae, and one case of non-arteritic vessel occlusion. None of these findings were felt to be explanatory for the clinical presentation. The most common non-vascular findings were seen in the brain parenchyma and sinuses. Parenchymal changes were generally microvascular or post-ischaemic changes; the only findings that were considered explanatory were two meningiomas. Twenty-seven cases of sinus abnormalities were documented; the most common changes of mucosal thickening, polyps and mucus retention cysts were seen in 24 cases, of which 14 were associated with non-GCA diagnoses and 10 with cases of GCA (P = 0.63). A case of mastoiditis and a sinotemporal inflammatory pseudotumour provided explanatory diagnoses in two patients, and a non-explanatory cystic lesion thought consistent with a Thornwaldt mass was seen in a third. One diagnosis of sinusitis was associated with a final clinical diagnosis of possible anti-neutrophil cytoplasmic antibody-associated vasculitis. Non-brain neurological abnormalities were postsurgical changes after trigeminal nerve decompression and a facial nerve schwannoma.

Identification of alternative diagnosis

Explanatory findings on TAMRA that were clearly associated with diagnoses other than GCA were less common (Fig. 2). Four cases of TMJ synovitis were seen in individuals without GCA: two patients had an alternative diagnosis of PMR, and two did not have a known final diagnosis. Four cases of pachymeningitis were all associated with non-GCA diagnoses. Finally, in the orbit, a case of orbital inflammatory syndrome as well as a case of optic neuritis/inflammatory pseudotumour were also indicative of a non-GCA diagnosis. Collectively, explanatory findings that were suggestive of alternative diagnoses were seen in 14 individuals or 4.1% of TAMRAs performed.

Discussion

It has been well recognized that the addition of imaging modalities when assessing for GCA has both improved diagnostic accuracy of the disease and lowered the need for temporal artery biopsy to confirm the disease [12, 13]. This study demonstrates that TAMRA provides additional diagnostic yield: the concurrent assessment of anatomically adjacent structures for mimics of GCA that may otherwise go unrecognized. Furthermore, these results suggest that there may be a larger burden of intracranial GCA than previously recognized, which merits further exploration of its potential impact on therapeutic decision-making and patient follow-up.

The 14 patients in this cohort who were found to have alternative diagnoses all demonstrated radiological findings that can readily present with syndromes of headache, scalp tenderness and elevated inflammatory markers (e.g. sinusitis, mastoiditis, inflammatory pseudotumour, TMJ synovitis, pachymeningitis). While visual changes were not as likely with diagnoses including pachymeningitis and mastoiditis, clinicians who assess patients with GCA often seek to diagnose and treat the disease before there are many visual manifestations that may portend more serious consequences [14]. The analysis of Koster et al. supports these findings, though it was found that mimics had lower inflammatory markers and often more atypical features [3]. These diagnoses also markedly change the trajectory of care where therapy for GCA may be harmful (in cases of infectious aetiologies), and different diagnostic assessment, specialist involvement and follow-up may be required.

Several potentially non-explanatory findings were seen in addition to findings of arteritis. Studies of MRIs of asymptomatic individuals demonstrated aneurysms and benign tumours in 1.8% and 1.6% of individuals across all ages, respectively [15]. Higher rates of both were seen within this cohort, but the average age was older; however, these suggest that the additional 4.1% of cases with alternative explanatory findings are MRI findings that would not be otherwise expected. While there were diagnoses of sinusitis within both clinical groups, the frequency of non-specific sinus changes is common on MRI, with the prevalence of abnormalities in other studies of MRIs found to be as high as 44% [16, 17]. In the 17 patients with a non-GCA diagnosis who were found to have sinus changes, there were insufficient data available to analyse the total impact of the radiological finding on the final diagnosis. A subset of five of these patients, however, had marked mucosal enhancement and more severe sinus changes ipsilateral to the symptomatic side, and a non-GCA diagnosis was made, suggesting these were explanatory findings.

An unexpected finding within this cohort was the high frequency (14.3%) of intracranial arterial involvement in individuals who were ultimately diagnosed with GCA. It is important to contextualize these findings with recent literature demonstrating that lesions previously thought to be inflammatory may be artefactual due to non-pathological contrast enhancement of the vasa vasorum [9]. This study also provides the first large cohort estimates of the prevalence of these findings in a population being assessed for GCA. Previous reports of intracranial involvement of GCA have been limited to case reports/series or based on patients with GCA who present with clear neurological changes. The frequency of these findings is not well established but may potentially be as high as 40%, and their prognostic significance is unclear [18–21].

In this study, vessel wall enhancement seen in cases of GCA was thought to represent intracranial arteritis due to the presence of concentric wall enhancement, demonstrable wall thickening, the magnitude of enhancement and length of disease (Fig. 2). Further, in almost all cases with intracranial vascular involvement, there were multifocal inflammatory vascular changes intracranially, near the dural penetrance, and within several other visualized vessels including the ophthalmic arteries and external carotid artery branches—superficial temporal, internal maxillary, occipital and middle meningeal. The most affected intracranial vascular segments were the horizontal petrous portion of the ICA, the cavernous segments of the ICA, the supraclinoid ICA and the proximal fourth segment of the vertebral arteries.

While it is possible that the proximal vertebral artery wall enhancement could be simple vasa vasorum enhancement near the dural penetration, in our cases the degree of wall thickening was greater than those previously suggested as physiological (Fig. 2) [9]. The presence of multiple other inflamed arteries further increased the probability that the vertebral arteries demonstrated true intracranial vasculitic changes. None of the cases in our series had definite involvement of the anterior, middle or posterior cerebral arteries. Most cases manifested with wall thickening without significant stenosis, but four cases had stenosis associated with the vasculitic changes. Only one patient with intracranial vascular wall enhancement manifested with infarcts with the caveat that diffusion weighted imaging and complete parenchymal imaging were not performed.

A cohort published by Siemonsen et al. demonstrated GCA in 21 of 25 (84%) cases (compared with 37.1% within cohort) and higher rates of intracranial arteritis; intradural internal carotid involvement was seen in 10 (40%) patients [21]. Differences between their study and the current study likely account for the higher prevalence of intracranial arteritis. The Siemonsen cohort only included those who were felt to be likely to have GCA, whereas this cohort imaged people across a spectrum of probabilities of disease and there may have been fewer vessels involved. Siemonsen et al. also used 3D post-contrast images with the ability for multiplanar reformats. The protocol used within this study used high resolution 2D images with 4 mm-thick axial images, which could complicate assessment of the exact cranio-caudal extent of wall enhancement and its proximity to the area of dural penetration. These changes would render the protocol within this study less sensitive and suggest an underestimate of the prevalence of intracranial disease. Both studies also found a similar distribution of arterial involvement, with intracranial internal carotid artery involvement being the most common finding, followed by vertebral artery changes.

Parenchymal findings were rare in both this cohort and the patients characterized by Siemonsen et al. [21]. Given that the incidence of stroke in addition to GCA can carry additional morbidity, it is important to understand which patients are at risk of stroke as well as whether different therapeutic approaches are needed to manage intracranial disease. It has been suggested that IL-6 receptor inhibitors, a treatment that is commonly used as an adjunct to glucocorticoids, may be ineffective in treating intracranial disease, and other compounds that cross the blood–brain barrier may be needed [19, 20, 22].

This study has several clinical and radiological limitations. Clinically, given that the data for final diagnoses were only available on a limited basis, the incomplete data concerning alternative diagnoses limited the analysis of the value of alternative findings; often the clinician would indicate that a diagnosis was not GCA without posting a plausible alternative diagnosis. This likely led to under-reporting of the value of TAMRA in finding diagnostic alternatives such as sinusitis. Radiologically, imaging studies were only assessed by a single reader rather than two independent readers (though the reader was well-experienced in reading TAMRA) [8]. Furthermore, the TAMRA protocol used to evaluate patients was not tailored to comprehensively image brain parenchyma as diffusion-weighted, FLAIR, and T2-weighted imaging were not routinely performed in patients. This would, again, lead to under-reporting of potential intracranial involvement of GCA as well as neurological sequelae. Finally, there was no dedicated timeframe in which TAMRA needed to be done from symptom onset or evaluation, and as such some changes may have improved with interval immunosuppressive therapy and further affected detection of abnormalities.

The use of TAMRA in evaluating individuals for a possible diagnosis of GCA provides additional diagnostic yield by demonstrating alternative diagnoses in 4.1% of patients who would present with a similar clinical presentation to GCA. Further, 14.3% of cases of GCA demonstrated intracranial vasculitis. These findings indicate that more widespread usage of TAMRA can provide broader and more accurate diagnostic information in evaluating patients for GCA and that wider deployment will permit detection of intracranial vasculitis, which may be of therapeutic and prognostic significance for these patients.

Data availability statement

The de-identified data underlying this article, but not original images, are available upon reasonable request to the corresponding author.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: M.J., R.R., S.R. have no disclosures to report. S.G. has previously received educational grants from Roche. N.K. has been supported for clinical trials by Sanofi and AbbVie; he has received drug but no funding supports from BMS; he has received travel funding from Astra Zeneca; he has participated in advisory boards and employee presentations with Roche.

References