Takayasu arteritis (TAK) is a chronic inflammatory vasculitis that predominantly affects the aorta and its major branches, as well as the coronary and pulmonary arteries. TAK commonly affects young female in East Asia [1]. TAK is characterized by the initiation of inflammation at the vasa vasorum and around the medial-adventitial junction, with perivascular cuffing and mononuclear infiltration composed mainly of cluster of differentiation (CD)4+/CD8+ lymphocytes, macrophages, and giant cells. The progression of inflammation to pan-arteritis leads to arterial stenosis or aneurysms, predominantly in the chronic phase [2,3]. TAK is primarily managed with pharmacotherapy using corticosteroids or immunosuppressive agents. However, in advanced stages, surgical or endovascular interventions are necessary to address complications, including organ ischemia, uncontrolled renovascular hypertension, and arterial aneurysms [4].

Owing to its low incidence, previous studies have often analyzed treatment outcomes, including all cardiac lesions in patients with TAK, or reported TAK along with other forms of vasculitis, including giant cell arteritis and microscopic polyangiitis, in patients with middle aortic syndrome [5-8]. As TAK predominantly affects young female with a longer life expectancy than those with atherosclerotic disease, assessing specific long-term outcomes after revascularization for TAK is important [9,10].

Therefore, in this study, we reviewed the perioperative data of patients with TAK who required revascularization and aimed to investigate the patency of both endovascular and surgical approaches. Additionally, we explored the treatment and clinical outcomes of revascularization and compared early and late complications between patients receiving endovascular and surgical treatments.

1) Study population

We retrospectively reviewed a database comprising 73 cases of extra-cardiac revascularization in patients with TAK who underwent surgery or endovascular interventions at Seoul National University Hospital between 1983 and 2020. Data were collected from the medical records of each patient. These data included demographic information, preoperative symptoms, inflammatory markers, imaging findings, details of the surgical or endovascular procedures, postoperative complications, and follow-up imaging findings. The patients were categorized as “surgery” or “endovascular” based on the initial revascularization modality.

This study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Seoul National University Hospital (IRB No. 1410-038-616). The need for informed consent was waived due to the retrospective nature of this study.

2) Diagnosis

We included patients who met the diagnostic criteria for TAK defined during each period: the 1984 and 1990 American College of Rheumatology (ACR) criteria, and the 2018 and 2022 ACR/European League Against Rheumatism recommendations.

The diagnosis of TAK was confirmed in all patients through imaging studies, including assessments of vascular stenosis, occlusion, and dilatation/aneurysm on contrast-enhanced computed tomography angiography (CTA), magnetic resonance angiography (MRA), or digital subtraction angiography (DSA). Arterial lesions were divided into six categories: supra-aortic branches, aorta, renal artery, simultaneous aorta and renal arteries, visceral branches, and iliofemoral arteries. In the iliofemoral artery category, none of the patients had isolated iliofemoral lesions, and all patients had lesions involving the heart, aorta, or major branches.

To assess TAK disease activity, the erythrocyte sedimentation rate, C-reactive protein, and fibrinogen levels were used as biological markers of inflammation, along with imaging studies. Interventions are usually avoided during the active stage unless an emergency involving end-organ ischemia or rupture develops.

3) Procedures

Endovascular intervention (n=40, 54.8%) was the treatment of choice for short, stenotic, and ostial lesions, whereas surgery (n=33, 45.2%) was the preferred treatment for diffuse and complex pathologies. However, the first treatment modality was determined by the managing surgeon. Endovascular repair was performed under local anesthesia in all patients, except for children who were unable to cooperate. The primary revascularization procedure was percutaneous transluminal angioplasty (PTA). In cases where repeat balloon angioplasty was technically unsuccessful or complications, such as dissection, occurred during PTA, stenting was performed.

Surgical interventions were performed under general anesthesia in patients who did not respond to medical therapy or had diffuse multiple vascular stenoses, occlusions, or aneurysms with unsuccessful endovascular treatment (Fig. 1). Usually, saphenous vein grafts were used for visceral and renal arterial revascularization, whereas polytetrafluoroethylene or Dacron grafts were used for supra-aortic or aortic revascularization.

Figure 1. Distribution of initial vascular procedures based on anatomical classification and intervention type. BAP, balloon angioplasty; Lt, left; CCA, common carotid artery; CIA, common iliac artery; IIA, internal iliac artery.

4) Primary and secondary outcomes

The primary outcomes were primary patency (PP) and secondary patency (SP) based on the procedure. Patency status was defined according to the reporting standards of the Society for Vascular Surgery [11]. Secondary outcomes included mortality and early and late complications. Restenosis was defined as narrowing of ≥50% of the target artery diameter, and re-occlusion was defined as the absence of flow in the treated segment, as confirmed by CTA. Technical failure was defined as persistent luminal stenosis ≥50% within the target vessel. Complications were classified as early if they occurred within the first postoperative month.

5) Follow-up

All patients who underwent surgery after 2000 underwent postoperative imaging before discharge, followed by imaging surveillance at 1 year postoperatively. Subsequently, the follow-up interval was extended when the patient’s condition stabilized. Blood pressure (BP) and pulse palpation were recorded at each visit, and the ankle-brachial index (ABI) was assessed annually. All patients were prescribed one or two antiplatelet agents, with or without corticosteroids or immunosuppressive agents.

Patients who underwent endovascular treatment were subjected to imaging surveillance in case of symptom recurrence; if the symptoms correlated with the imaging findings, re-intervention was performed.

6) Statistical analyses

Continuous variables are presented as means±standard deviations or median values with interquartile ranges (IQRs) after normality testing, and categorical variables are presented as percentages. Comparisons between the surgical and endovascular groups were performed using Pearson chi-square and Fisher exact tests. Cumulative patency rates were estimated using Kaplan–Meier analysis. All analyses were performed using IBM SPSS Statistics ver. 26.0 (IBM Corp.) and R software (ver. 4.3.1, R Foundation for Statistical Computing). A P-value of <0.05 was considered significant.

1) Patient characteristics

Between November 1983 and October 2020, 73 revascularization procedures were performed in 56 of the 545 patients diagnosed with TAK. The mean follow-up duration was 147.7±111.6 months. Patient demographics are summarized in Table 1. All patients were of Korean ethnicity, with a predominance of females (n=42, 75.0%). The mean ages at symptom onset, diagnosis, and intervention were 27.6±13.7, 30.3±14.3, and 39.0±20.4 years, respectively. The most common comorbidities were hypertension (n=41, 73.2%) and heart failure with ejection fraction <40% (n=12, 21.4%).

Table 1 . Baseline characteristics of patients.

Parameter	Patient (n=56)	Parameter	Patient (n=56)
Sex, female	42 (75.0)	Clinical presentation
Age at onset (y)	27.6±13.7	Uncontrolled BP	36 (64.3)
Age at diagnosis (y)	30.3±14.3	Claudication	29 (50.0)
Age at intervention (y)	39.0±20.4	Bruit	24 (42.9)
Comorbidities		Headache	22 (39.3)
Hypertension	41 (73.2)	Myalgia	22 (39.3)
Heart failure	12 (21.4)	Pulselessness	21 (37.5)
Dyslipidemia	10 (17.9)	Dizziness	17 (30.4)
Stroke	8 (14.3)	U/E BP difference	16 (28.6)
Angina	5 (8.9)	Chest pain	13 (23.2)
Diabetes mellitus	5 (8.9)	Abdominal pain	5 (8.9)
TIA	2 (3.6)	Amaurosis fugax	4 (7.1)
Medication		Carotidynia	4 (7.1)
Corticosteroid	15 (26.8)	Fever	2 (3.6)
Methotrexate	4 (7.1)
Cyclosporine	4 (7.1)	Inflammatory marker
Azathioprine	1 (1.8)	ESR (mm/h)	14.0 (7.0-29.0)
Aspirin	25 (44.6)	CRP (mg/dL)	0.2 (0.1-1.1)
Clopidogrel	9 (16.1)	Fibrinogen (mg/dL)	324 (276-399)
Anticoagulant	6 (10.7)
Statin	11 (19.6)

Values are presented as number (%), means+standard deviation or median (interquartile range) after normality testing..

TIA, transient ischemic attack; BP, blood pressure; U/E, upper extremity; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate..

The most common symptom observed before the procedure was uncontrolled BP (n=36, 49.3%), followed by upper- or lower-extremity claudication (n=29, 39.7%). Systemic symptoms, including myalgia and fever, and symptoms related to specific pathologies, such as bruit, headache, dizziness, pulselessness, and differences in BP between the upper extremities, were observed. At the time of revascularization, single or dual antiplatelet agents were administered to 28 (38.4%) patients, corticosteroids to 15 (20.5%) patients, and immunosuppressants to 10 (13.7%) patients. Eighteen (24.7%) interventions were performed during the active phase, which was characterized by elevated levels of inflammatory markers.

2) Diagnosis and the nature of lesions on imaging study

All patients underwent preoperative imaging of the aorta and its branches. TAK lesions were confirmed in 45 (61.6%) cases using CTA. In 24 (32.9%) lesions, DSA was performed, and in 4 (5.5%) lesions, MRA was performed. Before 2000, anatomical assessments were predominantly performed using DSA (67.7%). Thereafter, most patients (88.1%) were evaluated using multi-slice thin-section CTA (Table 2).

Table 2 . Diagnostic modality for Takayasu arteritis.

Imaging study	Total (n=73)	1983-2000 (n=31)	2001-2020 (n=42)
Computed tomography angiography	45 (61.6)	8 (25.8)	37 (88.1)
Digital subtraction angiography	24 (32.9)	21 (67.7)	3 (7.1)
Magnetic resonance angiography	4 (5.5)	2 (6.5)	2 (4.8)

Values are presented as number (%)..

Among the anatomical lesions confirmed through imaging studies, the renal artery was the most affected (n=23, 31.5%), followed by the supra-aortic branches (innominate, subclavian, vertebral, and carotid arteries [n=19, 26.0%]) and the aorta (n=14, 19.2%). Concurrent pathologies involving both the aorta and renal artery were observed in 8 (11.0%) cases. One (1.4%) patient had lesions in the visceral branches of both the celiac artery and the superior mesenteric artery (SMA). All eight patients with iliofemoral artery involvement also had vasculitis of other larger vessels (five patients underwent cardiac surgery, and three patients had aortic pathologies), meeting the diagnostic criteria for TAK (Table 3).

Table 3 . Anatomical classification and initial procedure for Takayasu arteritis.

Anatomical classification			Initial procedure
Location	Total (n=73)		Endovascular (n=40)	Surgery (n=33)
Renal artery	23 (31.5)		15 (65.2)	8 (34.8)
Aorta	14 (19.2)		7 (50.0)	7 (50.0)
Aorta+renal artery	8 (11.0)		0	8 (100.0)
Supra-aortic branches	19 (26.0)		15 (78.9)	4 (21.1)
Iliofemoral artery	8 (11.0)		3 (37.5)	5 (62.5)
Celiac axis+SMA	1 (1.4)		0	1 (100.0)

Values are presented as number (%)..

SMA, superior mesenteric artery..

Among the lesions identified in the imaging studies, stenosis (n=56, 76.7%) was the most common, followed by combined stenosis and aneurysms (n=8, 11.0%). Occlusion and aneurysm alone occurred in five (6.8%) and four (5.5%) patients, respectively. Balloon angioplasty was the primary endovascular intervention performed. Stent placement was performed in cases of complications during balloon dilatation or recurrent lesions. For surgical treatment, bypass to the distal artery without a lesion or segmental resection and interposition graft was performed. Before 2000, 7 (9.6%) cases of renal autotransplantation or unilateral nephrectomy were performed for renal artery lesions associated with uncontrolled BP.

3) Treatment outcomes

The 5-year freedom from all-cause death was 95.9%. One patient who underwent axillobifemoral bypass for aortoiliac steno-occlusion developed a perigraft lymphocele 9 months postoperatively. After simple excision of the perigraft hygroma under general anesthesia, visceral perfusion via the collaterals was compromised, causing pneumatosis intestinalis and portal vein gas, resulting in multiorgan failure and death. Two patients died due to acute respiratory distress syndrome: one following a lung biopsy and the other due to pneumonia at 26 and 35 months postoperatively.

Fig. 2 shows the 20-year PP and SP in the endovascular and surgical groups. During the 20-year follow-up period, significant differences were observed in PP between the groups (P<0.001). The 5-, 10-, and 20-year PP rates in the endovascular vs. surgical groups were 42.2% vs. 84.4%, 31.7% vs. 75.7%, and 17.0% vs. 59.0%, respectively. The 5-, 10-, and 20-year SP rates in the endovascular vs. surgical groups were 76.3% vs. 87.9%, 72.3% vs. 79.7%, and 55.1% vs. 79.7%, respectively. Although the surgery group had better SP than the endovascular group, the difference was not statistically significant (P=0.230). In particular, significant rates of PP and SP loss were observed during the first 3 years after revascularization (Fig. 2).

Figure 2. (A) Kaplan–Meier survival curves of the 20-year primary patency and (B) secondary patency for the endovascular and surgery groups in patients who underwent revascularization for Takayasu arteritis.

We performed a subgroup analysis to evaluate the patency outcomes of the different intervention modalities for the renal artery, which was the most frequently revascularized lesion. This analysis included cases with isolated renal artery lesions and those involving both the renal artery and aorta. We compared the PP and SP rates between patients initially treated with an endovascular approach (n=15) and those treated with surgical intervention (n=16). The PP rate was significantly higher in patients who underwent surgery as the initial treatment (P=0.023). Although the SP rate tended to be higher in the surgery group, the difference was not statistically significant (P=0.097) (Supplementary Fig. 1).

The overall technical procedural success rate was 97.3%. Two cases of procedural failure developed with endovascular approaches as the initial treatment. Regarding early complications, three (7.5%) cases of endovascular procedures developed arterial dissection, identified either during the procedure or on final angiography. In two cases, immediate stent insertion was performed. In one patient who developed dissection during PTA of the vertebral artery, the procedure was concluded without further intervention because the vertebral artery was significantly tortuous, and the dissection flap was not large. The patient was prescribed lifelong antiplatelet therapy.

Two patients in the surgery group developed early postoperative complications. One patient, who underwent multiple visceral bypasses from the supraceliac aorta to the common hepatic artery, splenic artery, and SMA because of visceral aneurysms, developed thrombosis of the SMA graft on follow-up CTA. However, reconstitution of the distal SMA was confirmed, and no additional treatment was administered, except for dual antiplatelet therapy. The other patient underwent supraceliac-to-distal abdominal aortic bypass using a Dacron graft. Five days postoperatively, the patient experienced right leg discomfort, and CTA revealed right common iliac artery dissection from the distal anastomotic site. The patient subsequently underwent kissing stent placement in the bilateral common iliac arteries, which relieved the symptoms. No procedural mortality was observed.

The most common late complication was restenosis (endovascular, n=22 [55.0%] vs. surgery, n=4 [12.1%]; P<0.001) with a significant intergroup difference. Other late complications included thrombosis, dissection, and pseudoaneurysms, with no significant intergroup differences (Table 4).

Table 4 . Early and late complications of revascularization for Takayasu arteritis.

	Early complication				Late complication
	Endovascular (n=40)	Surgery (n=33)	P-value		Endovascular (n=40)	Surgery (n=33)	P-value
Restenosis	-	-	-		22 (55.0)	4 (12.1)	<0.001
Thrombosis	-	1 (3.0)	0.466		9 (22.5)	6 (18.2)	0.772
Dissection	3 (7.5)	1 (3.0)	0.618		1 (2.5)	2 (6.11)	0.595
Pseudoaneurysm	-	-	-		-	1 (3.0)	0.466
Bleeding	-	-	-		-	2 (6.1)	0.213
Stroke	2 (5.0)	-	0.498		-	-	-
Lymphocele	-	-	-		-	1 (3.0)	0.466
Pneumonia	-	-	-		-	1 (3.0)	0.466
Death	-	-	-		-	1 (3.0)	0.466

Values are presented as number (%)..

-, not available..

4) Clinical outcomes

Regarding the 36 cases of uncontrolled BP, the mean systolic BP was 182±25 mmHg and diastolic BP was 103±20 mmHg at the time of diagnosis. Of these, 66.7% (n=24) were treated with 1-5 antihypertensive medications. At discharge, the average systolic and diastolic BPs were 136.38±25.86 mmHg and 80.76±17.82 mmHg, respectively. At discharge, patients were administered 0-4 antihypertensive medications. At the most recent follow-up, the average systolic and diastolic BPs were 130±14 mmHg and 76±13 mmHg, respectively (Fig. 3).

Figure 3. Systolic and diastolic BPs at diagnosis, discharge, and final follow-up in patients who underwent revascularization for Takayasu arteritis. DBP, diastolic blood pressure; SBP, systolic blood pressure.

In twenty-nine cases (39.7%), the patients presented with symptoms of lower extremity claudication, and the median preoperative ABI was 0.64/0.43 (IQR, 0.45-0.93/0.43-0.62). The post-procedural ABI was 0.90/1.02 (IQR, 0.76-1.41/0.87-1.33), and the last follow-up ABI was 0.83/0.88 (IQR, 0.72-1.08/0.79-1.14).

Our study, based on long-term data from a large cohort, compared the PP and SP of surgical and endovascular treatments for extra-cardiac revascularization in patients with TAK. Furthermore, this study identified the postoperative complications and clinical outcomes associated with both revascularization approaches. Before 2000, DSA was primarily used for diagnosis; however, CTA has since replaced DSA. Similarly, kidney autotransplantation or nephrectomy was performed for uncontrolled renovascular hypertension if stenosis recurred after PTA or surgical angioplasty. Nevertheless, these surgeries are no longer performed because of the emergence of various devices and technical developments.

The diagnosis and treatment of vascular lesions in TAK have evolved over time, as evidenced by changes in the guidelines. According to the 2021 ACR/Vasculitis Foundation Guidelines for Giant Cell Arteritis and TAK, medical management is preferred in most cases. Surgical intervention is not recommended unless disease progression occurs [12]. Moreover, TAK is a rare disease, and revascularization for this disease is declining. Most published studies are case series with a heterogeneous range of lesions, including those in the coronary artery, ascending aorta, and heart valves, complicating the interpretation of the results [2,13]. These surgeries necessitate cardiopulmonary bypass, which can lead to global cardiac arrest, intense inflammatory responses, and the release of procoagulants and proinflammatory mediators into the systemic circulation [14]. These circumstances may affect postoperative outcomes, including the occurrence of complications and mortality, and subsequently interfere with the analysis of revascularization patency. Hence, in the current study, we focused only on extracardiac revascularization.

Arterial lesions in TAK can manifest as various forms of luminal damage, including stenosis, occlusion, aneurysm, and sometimes in combination [15]. These arterial lesions may relapse in the same area or invade a new location even after medical treatment or surgery because each vessel can be in a different phase of disease activity [12,16]. Seventeen (29.8%) patients in our cohort had cardiac lesions or lesions at different sites. Therefore, the transition from an era of limited diagnosis for the identification of target lesions using DSA to the current practice of preoperative assessment of the aorta and major branches using CTA represents a meaningful advancement. Furthermore, evaluation of the overall progression of vascular lesions using CTA, rather than focusing solely on local observations of the target lesion, is now being used for postoperative follow-up, and early detection of cardiac involvement using echocardiography is advocated.

According to our study, the predominant complication observed during follow-up after revascularization was restenosis, followed by thrombosis, which, when combined, occurred in 56.1% of the patients. Following the intervention, patients routinely received lifelong single or dual antiplatelet therapy unless indicated otherwise. Patients who undergo revascularization for TAK have a 10-year survival rate of ≥70% [13]. The 5-year PP and SP rates for revascularization in the open vs. endovascular groups were approximately 70% vs. 50% and 80% vs. 70%, respectively [13,17]. These findings are consistent with the results of this study. Considering these circumstances, patients with TAK may require more potent antithrombotic medications after revascularization than those with atherosclerotic arterial obstruction. A significant decline in PP and SP was observed within 36 months of revascularization, particularly in patients who underwent endovascular procedures rather than surgery. Therefore, closer follow-up is necessary for patients undergoing endovascular procedures within 3 years of surgery.

TAK is a rare form of chronic vasculitis with diagnostic criteria including age <60 years at the time of diagnosis, female sex, clinical symptoms, and imaging findings [15]. The reported incidence rate in South Korea is 0.21-0.29 cases per 100,000 population, with a predominance of female of 4.3:1 [18]. Although various incidence rates have been reported in other regions, the predominance of female remains consistent, with a reported ratio of approximately 9:1 [19,20]. Vasculitis is typically diagnosed based on symptoms arising from typical vascular lesions, such as uncontrolled hypertension, ischemic symptoms of target vessels, pulselessness, and asymmetric BP measurements [21]. However, it may also present with atypical symptoms attributable to systemic inflammation, such as fever and myalgia [22]. Therefore, vasculitis can easily be overlooked in patients without the typical symptoms of vascular disorders. In our cohort, the most common symptom was uncontrolled hypertension. Female aged <60 years with uncontrolled hypertension accounted for 31.5% (n=23) of all cases. Based on our experience, the average duration from symptom onset to diagnosis was 2.6 years. Therefore, it is crucial to thoroughly evaluate TAK in young female with uncontrolled hypertension.

We systematically conducted longitudinal monitoring of patients who underwent revascularization for TAK, providing comprehensive reports on patency and associated complications. However, this study had some limitations. Although we analyzed vascular lesions and excluded cardiac lesions, heterogeneity persisted because of anatomical differences. In our analysis, anatomical location did not exhibit a statistically significant effect on PP or SP in the multivariate Cox regression analysis. However, a difference in PP based on location was observed (Supplementary Fig. 2). Therefore, subsequent multicenter studies with larger numbers of cases are warranted to comprehensively compare patency, complications, and outcomes between endovascular and open revascularization in each subgroup. Prospective studies comparing the efficacy of newer endovascular techniques and devices with traditional surgical approaches are essential to further refine treatment protocols. In addition, investigating the mechanisms underlying restenosis during endovascular interventions may lead to the development of targeted therapies to reduce the incidence of restenosis. Moreover, owing to the low event rate, analysis of the risk factors influencing patency and the occurrence of complications was not possible. Further large multicenter studies, including genetic and molecular studies, may provide a clearer understanding of the factors that influence patency and complications in patients with TAK undergoing revascularization.

Revascularization for TAK can be safely performed using both endovascular and surgical approaches. Although no significant difference was observed in SP between the two groups, the surgery group demonstrated a significantly better PP rate than the endovascular group. Significant PP and SP losses were evident in the first 3 years post-intervention, suggesting that patients undergoing revascularization should be subjected to more frequent monitoring during this period. Considering the relatively young age and extended life expectancy of patients with TAK, surgical intervention may be beneficial.

Supplementary data can be found via https://doi.org/10.5758/vsi.240090

Seung-Kee Min has been the editorial board member of the VSI since 2019. Ahram Han, Sanghyun Ahn, and Sangil Min have been the editorial board members of the VSI since 2023. They were not involved in the review process. Otherwise, no potential conflict of interest relevant to this article was reported.

Concept and design: JS, SA, SM, SKM. Analysis and interpretation: JS, AC, SA, SM, SKM. Data collection: JS, AC, AH. Writing the article: JS. Critical revision of the article: SKM. Final approval of the article: all authors. Statistical analysis: JS. Obtained funding: none. Overall responsibility: SKM.