Intracranial Involvement in Takayasu’s Arteritis

Takayasu’s arteritis (TAK) is a large-vessel vasculitis that targets the aorta and its major branches. Although extracranial vascular involvement is uniformly present in this disease, the frequency of intracranial involvement in TAK has not been well studied. We retrospectively reviewed the clinical and imaging records of patients diagnosed with TAK at a single Canadian university medical centre to determine the prevalence of intracranial vascular involvement. Intracranial vascular and non-vascular findings were described, and a review of the literature was performed. Of 20 patients with TAK, 12 had vascular neuroimaging completed. Intracranial vascular lesions were identified in 4 patients (33.3% of those with imaging available, 20% of all patients). The frequency of intracranial vessel involvement in TAK may be more common than appreciated. Imaging of both the intra- and extra-cranial vessels should be considered in these young patients.


Introduction
Takayasu's arteritis (TAK) is a large-vessel vasculitis of unknown etiology, which, by definition, affects younger persons, and results in aneurysms, stenoses, and occlusions of the aorta and its major branch vessels. With an estimated annual incidence of 2.6 per million in North America [1], the clinical presentation of TAK is heterogenous and insidious. Some patients are asymptomatic, whereas others present with severe manifestations of endorgan ischemia. Among the breadth of symptoms that patients may experience, neurologic manifestations such as headache, light-headedness, vision changes, transient ischemic attack and stroke are common, estimated to be present in between 42% and 80% of TAK cases [2][3][4][5][6]. Neurologic manifestations in TAK most often develop as a result of vascular injury to the extracranial carotid, vertebral, or subclavian arteries.
Intracranial vascular involvement in TAK is considered uncommon; however, it has not been well studied. Existing literature suggests that intracranial vascular disease may affect between 5% and 50% of TAK patients [5][6][7][8][9][10][11][12]. This wide range is likely attributed to the differences in study designs, imaging modalities utilized, and the limited availability of neuroimaging studies among asymptomatic patients.
In this study, we aimed to determine the prevalence of intracranial vascular lesions in TAK patients at a single Canadian university centre, describe the most frequently observed lesions, and review the existing literature regarding the frequency of intracranial vascular disease in TAK.

Materials and Methods
The clinical records of patients diagnosed with TAK by rheumatologists at the University of Alberta, Edmonton, Canada, and seen between 2012 and 2018 were retrospectively reviewed. The electronic medical record database was searched by diagnostic code; then, identified records underwent complete chart review. For study inclusion, patients were required to have a confirmed clinical diagnosis of TAK by their treating rheumatologist, an

Results
The initial diagnostic code search identified 24 possible patients in the electronic database. Upon complete chart review, 4 were excluded due to the presence of alternate diagnosis, resulting in 20 included TAK patients in this study. All patients had a clinical diagnosis of TAK by their rheumatologist, and all had imaging evidence of large-vessel vasculitis involving the aorta and/or its major branches, as defined above. All patients were assessed by their rheumatologist more than once, with a median number of 9.5 clinical assessments (range 2-47). All patients met the 2012 Chapel Hill Consensus Conference nomenclature definition of Takayasu's arteritis [15], and 12 of 20 also met the 1990 American College of Rheumatology (ACR) criteria for TAK [13]. Of the 20 TAK patients identified, 18 (90%) were female, with an average age at diagnosis of 29.5 years (range 13-48).
Neurologic symptoms were present in ten (50%) TAK patients, of which eight patients (40%) had neuroimaging. Table 1 details complete baseline patient demographics and presenting symptoms. Symptoms were present for a median of 2 months prior to diagnosis (range 0 to 36 months).

Neuroimaging Results
Neuroimaging was performed in 15 of 20 patients (75%); however, images were unavailable to review in 2 patients, and 1 patient had non-vascular neuroimaging only, leaving 12 TAK patients with vascular neuroimaging studies available for review. Eight patients had vascular neuroimaging completed at the time of diagnosis, although in three patients, the first study was delayed for more than 1 year after initial TAK diagnosis (median 0, range 0-24 months). At the time of vascular neuroimaging, 8 of 12 patients (66.7%) had neurologic symptoms, including stroke (4, 33.3%), vision changes (3, 25.0%), headache (2, 16.7%), or light-headedness (2, 16.7%). Intracranial vascular lesions were identified in four patients (33.3% of those with vascular neuroimaging available, 20% of all patients). Three patients (15%) had involvement of the middle cerebral arteries, two patients (10%) had involvement the intracranial portion of the internal carotid arteries,

Neuroimaging Results
Neuroimaging was performed in 15 of 20 patients (75%); however, images were unavailable to review in 2 patients, and 1 patient had non-vascular neuroimaging only, leaving 12 TAK patients with vascular neuroimaging studies available for review. Eight patients had vascular neuroimaging completed at the time of diagnosis, although in three patients, the first study was delayed for more than 1 year after initial TAK diagnosis (median 0, range 0-24 months). At the time of vascular neuroimaging, 8 of 12 patients (66.7%) had neurologic symptoms, including stroke (4, 33.3%), vision changes (3, 25.0%), headache (2, 16.7%), or light-headedness (2, 16.7%). Intracranial vascular lesions were identified in four patients (33.3% of those with vascular neuroimaging available, 20% of all patients). Three patients (15%) had involvement of the middle cerebral arteries, two patients (10%) had involvement the intracranial portion of the internal carotid arteries, and a single patient (5%) had involvement of the anterior cerebral artery. The clinical and radiographic details of all 12 patients with vascular neuroimaging studies available are summarized in Table 2.

Comparison of TAK Patients with and without Intracranial Vascular Disease
TAK patients with and without intracranial vascular disease were compared with respect to age at diagnosis, frequencies of smoking (current or former), hypertension, the presence of carotid artery involvement, c-reactive protein (as a surrogate for disease activity), and Numano classification (see Table 3). All patients with intracranial vascular lesions had involvement of the extracranial carotid arteries and great arch vessels (Numano classification I, IIA, or V). No statistically significant differences were observed. A detailed description of the four patients with intracranial vascular lesions is found below.

Case 1
Patient 1 was a 30-year-old female who presented with new-onset hypertension, and after medical attention it was found that their left brachial pulse was absent. The patient also reported a several-year history of left arm claudication, occasional headaches, and scalp tenderness. Inflammatory markers were elevated, with ESR 28 mm/h, and CRP 21.3 mg/L. CT-A of the chest, abdomen, and pelvis revealed a long segment (7 cm) of left subclavian stenosis/occlusion, and diffuse stenosis to the left common carotid artery. A diagnosis of TAK was performed, and treatment with prednisone, methotrexate, folic acid, and ASA were initiated. Despite treatment, inflammatory markers remained persistently elevated (CRP 21 mg/L), left arm claudication worsened, and the patient reported new episodes of slowed thinking and light-headedness. A CT-A of the head and neck arteries, followed by the MR brain with and without enhancement, revealed high-grade stenosis of the left cavernous carotid artery (see Figure 2) with mild multi-segmental stenoses in the skull base portion of the left internal carotid artery, and irregularities of the bilateral MCAs and left ACA. Diffuse vessel wall enhancement was noted. Due to worsening symptoms, elevated inflammatory markers, and intracranial vessel involvement, prednisone was increased, and infliximab 5 mg/kg was added. The patient's symptoms improved, serial imaging stabilized, and brain perfusion study was normal. Follow-up imaging one year later revealed stable lesions with the ongoing diffuse narrowing of the left internal carotid artery and circumferential enhancement, as well as ongoing stenoses of left M1 and left A1 branches, but no new vascular lesions.

Case 1
Patient 1 was a 30-year-old female who presented with new-onset hypertension, and after medical attention it was found that their left brachial pulse was absent. The patient also reported a several-year history of left arm claudication, occasional headaches, and scalp tenderness. Inflammatory markers were elevated, with ESR 28 mm/h, and CRP 21.3 mg/L. CT-A of the chest, abdomen, and pelvis revealed a long segment (7 cm) of left subclavian stenosis/occlusion, and diffuse stenosis to the left common carotid artery. A diagnosis of TAK was performed, and treatment with prednisone, methotrexate, folic acid, and ASA were initiated. Despite treatment, inflammatory markers remained persistently elevated (CRP 21 mg/L), left arm claudication worsened, and the patient reported new episodes of slowed thinking and light-headedness. A CT-A of the head and neck arteries, followed by the MR brain with and without enhancement, revealed high-grade stenosis of the left cavernous carotid artery (see Figure 2) with mild multi-segmental stenoses in the skull base portion of the left internal carotid artery, and irregularities of the bilateral MCAs and left ACA. Diffuse vessel wall enhancement was noted. Due to worsening symptoms, elevated inflammatory markers, and intracranial vessel involvement, prednisone was increased, and infliximab 5 mg/kg was added. The patient's symptoms improved, serial imaging stabilized, and brain perfusion study was normal. Follow-up imaging one year later revealed stable lesions with the ongoing diffuse narrowing of the left internal carotid artery and circumferential enhancement, as well as ongoing stenoses of left M1 and left A1 branches, but no new vascular lesions.

Case 2
Patient 2 was a 48-year-old female who presented with dysarthria and left arm weakness, and was found to have a right MCA ischemic stroke. Past medical history was significant for hypertension, dyslipidemia, and previous smoking history. Inflammatory

Case 2
Patient 2 was a 48-year-old female who presented with dysarthria and left arm weakness, and was found to have a right MCA ischemic stroke. Past medical history was significant for hypertension, dyslipidemia, and previous smoking history. Inflammatory markers were normal at the time of diagnosis. CTA of the head and neck revealed long segments of circumferential thickening of bilateral common, internal, and external carotid arteries with lumenal irregularities. Significant narrowing of the right internal carotid artery extended up into the cavernous portion of internal carotid artery. A small right frontoparietal MCA embolic infarct was also noted. Thoracic aorta and subclavian arteries appeared normal on MRA; however, the abdominal aortic wall was thickened and enhanced with gadolinium contrast. A diagnosis of TAK was made by her treating rheumatologist. She was treated with prednisone taper, followed by mycofenolate mofetil upon diagnosis. Follow-up MRA 1 month later revealed improved wall enhancement, suggesting a response to treatment, and repeat imaging several years later revealed stability of these lesions.

Case 3
Patient 3 was a 29-year-old female who presented with acute onset sensory loss of the right face, arm, and leg, and was found to have left MCA ischemic stroke. The patient denied any preceding constitutional or vascular symptoms. She had a past medical history of psoriatic arthritis and ulcerative colitis. Inflammatory markers at presentation were elevated, with an ESR 58 mm/h and CRP of 12.7 mg/L. CTA and MRA of the head and neck vessels, followed by conventional angiogram of the brain, were performed. Imaging studies revealed wall thickening of the aortic arch and proximal great vessels, mild narrowing in both distal common carotid arteries, severe stenosis of the left subclavian artery, and moderate stenosis of the right subclavian artery. Intracranially, the left MCA was occluded. She was diagnosed with TAK by rheumatology, and treated with pulse steroids, followed by subcutaneous weekly methotrexate. Subsequent follow-up showed clinical stability and improvements in the radiographic appearance of the subclavian artery stenosis.

Case 4
Patient 4 was a 13-year-old female presenting with acute onset left hemiparesis, aphasia, and vision changes, on a background history of 2 months of headaches. On examination, no pulses were detectable in the left arm. Inflammatory markers were normal, but initial non-vascular imaging with CT revealed a right frontal intracerebral hematoma and left parietal infarcts. Multiple assays, including CT, MR, CT-A, MR-A, and conventional angiography, were performed. Severe luminal stenoses approaching 70% were noted in the left common carotid artery, left carotid bifurcation, right common carotid, and proximal right subclavian arteries. The left subclavian artery was completely occluded. Stenoses of the right MCA, affecting M1 and M2 branches, were also noted. TAK was diagnosed by a pediatric rheumatology. She was treated with prednisone, aspirin, and azathioprine. Symptoms improved, and lesions stabilized on follow-up imaging.
In this small series of 20 TAK patients, neurologic symptoms were present in 50% and intracranial vascular abnormalities suggestive of intracranial vasculitis were noted in 20% overall, or 33.3% of those with dedicated neurovascular imaging, suggesting that intracranial vascular involvement may be present in a significant minority of TAK patients. Of note, all four of our patients with intracranial vascular disease had neurologic symptoms, and stroke was present in three of four (75%) at presentation. Our findings are broadly in-keeping with the existing literature, which suggests that intracranial involvement in TAK may be identified in approximately 25% of patients in whom such imaging is performed. The frequency of intracranial involvement varies considerably in individual studies (from 4.9% to 41.9%), likely as a result of study design, reporting bias, patient selection, and the type of imaging modality performed. Significant variability in the clinical presentation of TAK has also been observed across different geographic regions [31] and ethnicities [32], suggesting another possible source of differences. In our cohort, we were unable to assess the role of ethnicity due to inconsistent documentation in medical records.
The baseline demographics of this TAK patient cohort are consistent with those described in the literature [33]. Stroke or TIA were observed in 27.3% of our TAK patients, compared to 9-20% in recent reports [2,4,[17][18][19][20]. Similar to other studies, in our patients, the intracranial arteries most likely to be affected were the middle cerebral arteries, followed by the intracranial portions of the internal carotid arteries [13,34]. In our patients, intracranial vascular lesions were identified at the time of first dedicated vascular neuroimaging, which was usually performed early in the disease course. Given small patient numbers, we were not able to identify any statistically significant risk factors for intracranial artery involvement; however, we note that all four (100%) of our TAK patients with intracranial vascular lesions also had involvement of the great arch vessels (Numano types I, IIa, and V), and more specifically, all four had involvement of the extracranial carotid arteries. This raises the hypothesis that TAK patients with extracranial carotid artery involvement may be more likely than those without to develop intracranial vascular disease; however, further study is needed.
This study is limited by the retrospective nature and small sample size. The majority of patients who had complete neuroimaging had concurrent neurologic symptoms, potentially falsely increasing the prevalence of intracranial disease found. Additionally, in most of our patients, conventional angiograms were not performed; therefore, the involvement of smaller intracranial arteries could have been missed.
Currently, many vasculitis experts advocate for serial large-vessel imaging of the aorta and major branch vessels in TAK patients, in order to detect new lesions indicative of ongoing disease activity that may be clinically silent by history or physical exam [16]. There is a lack of evidence regarding the value of routine imaging of the intracranial vessels in this disease, however. In the present study, intracranial vasculitic lesions were identified in one in three patients in whom such imaging was performed, and were typically present at the time of first neuroimaging, suggesting that intracranial involvement may be more common than previously suspected. Intracranial vascular lesions were only observed in those patients who had stenoses of the extra-cranial carotid arteries. Obtaining both intra-and extra-cranial vascular imaging in TAK patients at baseline may provide a more complete assessment of the extent of patients' disease and provide an opportunity for monitoring disease progression. Given the potentially serious neurologic consequences, gaining an improved understanding of the true frequency of intracranial involvement in TAK will be important to help inform recommendations for disease monitoring in these young patients. Prospective studies are needed. Informed Consent Statement: Patient consent was waived due to the retrospective nature of this study, deemed to pose minimal risk to patients.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no relevant conflict of interest.