1. Top of page
  2. Abstract


To describe the clinical, laboratory, and radiographic manifestations of Takayasu arteritis (TA) in a cohort from the US, evaluate the response to interventions, remission and relapse rates, and disease progression, and compare these observations with those from other cohorts in the US, Japan, India, Italy, and Mexico.


Seventy-five patients were retrospectively studied using a uniform database that included clinical, laboratory, and imaging data. Vascular imaging studies were performed at least yearly to monitor disease progression.


Common manifestations at disease onset included loss or asymmetry of pulses (57%), limb blood pressure discrepancy (53%), and bruits (53%). Eleven percent of patients were asymptomatic prior to disease diagnosis. Initial angiographic studies showed aortic abnormalities in 79% of patients and frequent involvement of the subclavian (65%) and carotid (43%) arteries.

Ninety-three percent of longitudinally followed patients attained disease remission of any duration, but only 28% sustained remission of at least 6 months' duration after prednisone was tapered to <10 mg daily. Both angioplasty and vascular surgery were initially successful, but recurrent stenosis occurred in 78% of angioplasty and 36% of bypass/reconstruction procedures. More than two-thirds of patients had difficulty performing routine daily activities and approximately one-fourth of all patients were unable to work.

Our cohort was similar to the National Institutes of Health, Italian, Japanese, and Mexican cohorts in terms of the predominance of female subjects and disease manifestations, but differed from the Indian cohort in that the latter group had a higher frequency of male subjects, abdominal aorta and renal artery involvement, and hypertension.


Although improvement of symptoms in TA usually follows glucocorticoid therapy, relapses usually occur with dosage reduction. Attempts to restore vascular patency are often initially successful, but restenosis occurs frequently. Chronic morbidity and disability occur in most patients with TA in the US.

Takayasu arteritis (TA) is a rare chronic inflammatory disease of unknown cause that affects the large arteries. It most often affects the aorta and its primary branches. Although once thought to be a disorder that mostly affected young Asian women, TA has been identified in both sexes and many ethnic and racial groups worldwide. The incidence of TA has been reported to be 2.6 cases per million per year in Olmsted County, Minnesota (1). The prevalence of TA in an autopsy series in Japan was reported as 1 in 3,000 cases (2).

Few medical centers in the US have had adequate experience with TA to enable comparisons of disease features and outcomes with cohorts reported from Asia and Mexico. Since 1992, the Cleveland Clinic Center for Vasculitis Care and Research has provided a standardized approach to the diagnosis, followup, and treatment of patients with TA that includes routine imaging of the entire aorta and its branch vessels as dictated by changing disease features, or in cases with no apparent clinical change, vascular imaging has been routinely performed at least every 6–12 months. This approach was derived from previously published longitudinal TA cohort protocols from the National Institutes of Health (NIH). The NIH experience had suggested that TA is a chronic, relapsing illness associated with considerable morbidity and less optimistic outcomes than had been reported from the Far East (3).

We evaluated clinical, laboratory, and radiographic data from patients with TA who received either consultative or continuous care from one physician within our center. We sought to compare features of the disease from our cohort with those from the NIH and ethnically and racially distinct populations from other countries, including Japan, Italy, India, and Mexico.


  1. Top of page
  2. Abstract

Seventy-five patients with TA were evaluated at the Cleveland Clinic Foundation (CCF) from 1992 to 2004. Patients were identified from a registry maintained by one of the authors (GSH); the registry was checked for completeness by a chart survey that identified prior evaluation within the Center for Vasculitis Care and Research, an International Classification of Diseases, Ninth Revision (ICD-9) code for TA, or imaging studies that included the entire aorta and its primary branches. Patients so identified also had imaging abnormalities, including stenoses and/or aneurysms characteristic of TA, verified by review of radiologic studies. Electronic and paper chart review ensured that potential confounding conditions that could mimic TA (e.g., Cogan's syndrome, sarcoidosis, Kawasaki disease, Behçet's disease, syphilis, tuberculosis, Ehlers-Danlos syndrome, Marfan's syndrome, and neurofibromatosis) were not present.

Only patients meeting the American College of Rheumatology (ACR) criteria for TA (4) and our requirement for imaging abnormalities were included. If ACR criteria were met with the exception of the age requirement (being ≤40 years of age), and if the patients were between the ages of 41 and 50 years without fulfilling ACR criteria for giant cell arteritis (GCA), they were also included because ACR criteria do not classify patients in this age range as having either TA or GCA (4, 5). Disease onset was defined as the initial time at which there was identification of vascular symptoms that were subsequently determined to be compatible with TA and not attributable to comorbid conditions. For the subset of 30 patients who received longitudinal care at our center, we were able to further delineate the effectiveness of medical and surgical interventions and patterns of disease progression.

Criteria for disease activity.

We defined active disease using guidelines from previous NIH studies (3), including onset or worsening of at least 2 of the following disease features: 1) presence of systemic signs or symptoms (e.g., fever, arthralgias) that were not attributable to another condition, 2) elevation of acute-phase reactants (erythrocyte sedimentation rate [ESR], C-reactive protein [CRP]) in the absence of infection or malignancy, 3) onset of signs or symptoms of vascular insufficiency (unequal or absent pulses or blood pressure, limb claudication), and 4) vascular lesion(s) in previously unaffected vascular territories, detected on serial imaging studies.

Because of the tendency for established vascular abnormalities to progress with regard to further stenosis or dilatation (even in the absence of active disease), for a vascular anatomic change to be considered a measure of active progressive disease, it had to have been entirely absent in prior imaging studies (3). Remission was defined as the resolution of clinical and laboratory features of active disease and the absence of new vascular lesions, as determined by sequential imaging studies. Sustained remission was defined as these conditions having been met for at least 6 months while on a treatment regimen that included <10 mg/day of prednisone.

Medical therapy.

Initial therapy for active disease generally consisted of prednisone only (1 mg/kg/day) for 1 month, which was then tapered after resolution of symptoms of active disease and normalization of acute-phase reactants. Prednisone dosage was reduced by 5 mg/week until reaching a dosage of 20 mg/day. The taper rate was then reduced to 2.5 mg/week until achieving a dosage of 10 mg/day, and then reductions proceeded by 1 mg/week until prednisone discontinuation or relapse occurred. If relapse occurred, prednisone was increased by 10 mg over the last effective dose, and methotrexate (MTX) therapy was initiated at 15 mg/week (with folic acid 1 mg/day and trimethoprim/sulfamethoxazole double strength 3 times per week for prophylaxis against Pneumocystisjiroveci pneumonia). MTX was increased (maximum 25 mg/week) if necessary to sustain remission.

If a patient was intolerant to or failed to achieve or sustain remission while receiving MTX, azathioprine (AZA) therapy was initiated at 2 mg/kg/day, and if medication intolerance or treatment failure occurred with that agent, mycophenolate mofetil (MMF; 1.5 gm twice daily) was used. In the setting of immediate life-threatening disease (1 patient), daily cyclophosphamide (CYC; 2 mg/kg/day) was used, concurrent with glucocorticoids as initial therapy. After 3 months of CYC therapy, after remission was achieved and sustained, that agent was discontinued and weekly MTX therapy was started. Patients received anti–tumor necrosis factor (anti-TNF) therapy if they were unable to maintain disease remission with these agents.

Assessment of disease activity and acute-phase reactants in the longitudinal cohort.

At each office visit, criteria for disease activity were applied (see above) based on symptom assessment, physical examination, and laboratory studies. Complete aortic and primary branch vessel imaging was performed every 6 months or sooner if there was uncertainty about disease progression.

Statistical analysis.

Statistical analyses were performed using Stata software, version 9.0 (StataCorp, College Station, TX). Categorical data and proportions were compared using the chi-square or Fisher's exact test. P values less than 0.05 were considered significant.


  1. Top of page
  2. Abstract

Patient characteristics.

Complete cohort.

Ninety-one percent of patients (68 of 75) were female, with a median age at disease onset of 26 years (range 5–49 years). Ninety-two percent of patients were white, 3% Arabic, 3% Hispanic, and 2% Asian.

Longitudinal cohort.

Thirty patients received longitudinal care at our center. The median duration of followup was 3.0 years (range 4 months to 10 years). The demographic background of this group (97% female, 87% white, median age at disease onset of 27 years) was similar to that of the complete cohort of 75 patients.

Clinical features at disease presentation.

Cardiac and vascular symptoms and findings.

The most common presenting manifestations were absence or asymmetry of upper or lower extremity pulse(s) (57%), blood pressure inequality between upper or lower limbs (>15 mm Hg) (53%), vascular bruits (53%), and extremity claudication (48%) (Figure 1). Vascular bruits were present most often over the carotid arteries (36%) and were less common in the abdomen (15%) and femoral (15%) and subclavian (13%) regions. New onset of hypertension occurred in 28% of patients, among whom 58% had imaging evidence of renal artery stenosis.

thumbnail image

Figure 1. Manifestations of Takayasu arteritis at the time of disease onset in 75 patients. Dim = diminished; BP = blood pressure; HTN = hypertension.

Download figure to PowerPoint

The most common cardiac finding was aortic valvular regurgitation, which was secondary to aortic root dilatation (ascending thoracic aorta aneurysm formation). This was apparent in 18 of 75 affected patients (24%) at the time of initial diagnosis. Pericarditis was present in 8%, and congestive heart failure in 7%. When congestive heart failure occurred at presentation, it was associated with the presence of new-onset hypertension (P < 0.001) and renal artery stenosis (P = 0.02). Five percent of patients presented with symptoms of angina pectoris. None of the 75 patients presented with myocardial infarction.

Neurovascular signs and symptoms.

Visual aberrations, such as blurring, scotoma, or diplopia, occurred in 9 of 75 patients (12%) and amaurosis fugax in 3 of 75 patients (4%). No patient presented with irreversible visual loss. Transient ischemic attack and stroke occurred in 3% and 5% of patients, respectively.

Laboratory data at presentation.

Data were available for acute-phase reactants at the time of disease onset in 55 patients (73%) in the complete cohort. The mean Westergren ESR was 67 mm/hour (range 5–135). Eight percent of patients had normal ESR values (≤20 mm/hour) at the time of diagnosis. Mean hemoglobin and hematocrit values were 11.85 gm/dl and 34.5 volume percent, respectively.

Initial angiographic findings.

All 75 patients had at least one catheter-directed angiographic study, in addition to other forms of vascular imaging. Ninety-seven percent of patients had vascular stenoses, 8% had both stenotic lesions and aneurysms, while 3% had aneurysmal disease only.

At the time of diagnosis, except for the higher frequency of subclavian artery disease (83%) in the longitudinal cohort, vascular manifestations were similar to those seen in the entire cohort (Figures 2 and 3). Among patients in the longitudinal cohort, followup of vascular anatomy was by sequential magnetic resonance imaging (MRI) that included the entire aorta and all primary branch vessels (median interval 4.8 months, range 1.5–14 months). Invasive catheter-directed angiography was also performed when blood flow to all extremities was impaired and catheter-recorded pressures were needed to assess central pressure relationships to peripheral extremity measurements, or when surgery was considered and there was a need to achieve optimal visualization of lesions to aid in planning interventions.

thumbnail image

Figure 2. Sites of arterial involvement within 1 year of diagnosis of Takayasu arteritis in 75 patients. Ninety-seven percent of the lesions were stenotic, 3% of the lesions were aneurysmal, and 8% of the patients had both aneurysmal and stenotic disease.

Download figure to PowerPoint

thumbnail image

Figure 3. Frequency of arterial involvement at presentation and over time in 30 patients (mean followup period 3 years, mean interval between imaging studies 4.8 months). One hundred percent of the patients had stenotic lesions, and 17% had both stenotic and aneurysmal disease.

Download figure to PowerPoint

Disease activity, vascular imaging, and acute-phase reactants.

In 21 of 30 patients (70%) in the longitudinal cohort, increases in acute-phase reactants (ESR, CRP, or both) correlated with active disease status. Conversely, in 23% of patients, active disease was clearly present in the setting of normal laboratory parameters. In 2 patients (7%), TA was quiescent in all regards at the time of presentation and throughout the followup period. Among patients undergoing routine imaging surveillance (no new clinical signs), there were inconsistencies in the relationship between acute-phase reactants and imaging results. Forty-six percent of 78 routine patient visits were notable for the presence of elevated acute-phase reactants concurrent with MR images that revealed no new lesions, or for the presence of normal acute-phase reactants concurrent with new vascular lesions or new vessel wall enhancement/edema.

Treatment and outcomes.

Twenty-eight of 30 patients (93%) achieved disease remission of any duration. The great majority of patients (73%) required other immunosuppressive agents, in addition to glucocorticoids, to achieve and maintain disease remission. These included MTX (43%), anti-TNF agents (37%), CYC (13%), AZA (7%), and MMF (7%). Six patients (20%) attained disease control while receiving glucocorticoids alone and were able to taper prednisone dosages to <10 mg/day, while 2 patients (7%) presented in remission, never having received medical treatment and subsequently not requiring therapy. Two patients continued to have active disease despite therapy with glucocorticoids and other immunosuppressive agents.

Sustained remission (absence of clinical, laboratory, and radiographic evidence of new lesions for ≥6 months, while being treated with <10 mg/day of prednisone) was attained in only 8 of 28 patients (28%). Only 5 of these patients (17% of the long-term cohort) were able to remain in a state of sustained remission after prednisone had been tapered to discontinuation.

Disease relapse was documented in 27 of the 28 patients (96%) who had attained remission. A total of 90 relapses (mean 2.83 per patient) was reported over a median period of 3 years. Fifty-four relapse events occurred in the course of followup at our center and could be accurately characterized. The majority of relapses (34 of 54 [63%]) occurred while patients were receiving immunosuppressive therapy. Thirteen of 34 of these relapses (38%) occurred while patients were receiving both MTX and prednisone, 11 of 34 (32%) occurred while patients were receiving only prednisone, and 4 of 34 (12%) occurred while they were receiving only MTX. Median dosages of prednisone and/or MTX at relapse were 10 mg/day and 17.5 mg/week, respectively. Three relapses occurred in patients while they were receiving CYC (1) or AZA (2) therapy. Twenty relapses (37%) occurred while patients were no longer receiving any form of antiinflammatory/immunosuppressive therapy. New vascular lesions were noted in 53% of longitudinally followed patients. In this subset, 81% of relapses with alterations in vascular anatomy were associated with increases in acute-phase reactant values.

Anti-TNF therapy was provided as an experimental measure in 11 patients with refractory TA. Only 3 relapses occurred while patients were receiving treatment and 2 of these patients again achieved remission following an increase in the anti-TNF dosage and frequency. Presently, the mean duration of remission with anti-TNF therapy is 26 months (range 3 months to 6 years). Of the 10 patients who had a durable response to anti-TNF therapy, 2 were lost to followup because of a change in residence. One patient relapsed during therapy, and infliximab was discontinued. One patient, while in remission and not requiring glucocorticoids, had to discontinue therapy when chemotherapy for breast cancer was initiated. The 6 remaining patients have achieved sustained disease remission, and all have continued to not require glucocorticoid therapy.

Vascular interventions and outcomes in the longitudinal cohort.

Elective surgical interventions were ideally performed when the disease was judged to be quiescent, based on clinical, serologic, and imaging evaluations. However, in 5 patients with symptomatic, critical organ-threatening stenoses and active disease (e.g., coronary, carotid artery involvement), surgery was performed and immunosuppressive therapy was initiated or increased. Twenty-one of 30 patients (70%) in the extended followup cohort required a total of 64 revascularization procedures (Table 1). Twenty angioplasty procedures (10 with stents, none of which was drug-eluting) were performed. Two stent placements were immediately unsuccessful in restoring patency, in 1 patient because of the inability to deploy a stent in the vessel, and in the other patient, a stent became dislodged in the aortic arch at the time of deployment in the carotid artery and embolized to a more distal site.

Table 1. Vascular interventions and outcomes in the longitudinal cohort*
No. of patientsSustained patency, %No. of patientsSustained patency, %
  • *

    Revascularization procedures were required in 21 of 30 patients in the longitudinal cohort. NA = not applicable.

  • Patency sustained over the period of observation, with no further intervention required.

  • One patient who underwent an iliofemoral bypass and 1 who underwent an iliofemoral angioplasty had partial restenosis at followup imaging.

 Aortic root786NANA
 Descending aorta1100NANA
 Axillary artery1100NANA
 Abdominal aortaNANA1100
 Carotid artery77120
 Axillary artery302100
 Subclavian artery58040
 Coronary artery12501100
 Mesenteric artery2100NANA
 Renal artery56080
 Iliofemoral artery11002100

Restenosis occurred in 14 of 18 initially successful angioplasty procedures (78%), which led to further interventions in 13 patients. Significant ischemia required vascular bypass in 6 of these patients. Vascular bypass/reconstruction was performed in 17 of 30 patients (57%) (44 procedures in total) and was the most successful intervention. Forty-six percent of bypass procedures were performed following the failure of a previous intervention. Nonetheless, restenosis or occlusion occurred in 36% of these procedures, with all but one requiring further intervention. Aortic valve replacement was required in 5 patients with severe aortic root dilatation; revision was required in 1 patient after 10 years.

Histopathology data were available for 9 patients who had vascular reconstruction/bypass performed at our institution. While 2 of these patients had MRI evidence of active disease and 1 presented with elevated levels of acute-phase reactants at the time of surgery, 5 of the 9 patients had histopathologic evidence of active arteritis.

Disease-related morbidity and mortality in the longitudinal cohort.

Eighteen of 30 patients (60%) had vascular claudication that impaired the performance of routine daily activities. Five of 30 patients (17%) experienced cerebral ischemic disease (2 had transient ischemic attacks and 3 had strokes). Twenty-three percent were completely unable to work. The number of disease relapses correlated with the likelihood of disability (P < 0.0001).

Among 4 patients who had pregnancies concurrent with TA, 1 had a miscarriage at 25 weeks' gestation. Three had successful pregnancies and healthy children.

There were 2 disease-related deaths. One death occurred 1 month after extensive arterial bypass surgery involving the iliofemoral, renal, and superior mesenteric arteries. The postoperative course was complicated by a stroke and gastrointestinal hemorrhage. The other occurred within 2 months after subclavian bypass surgery, in the setting of changes in mental status that were associated with new cerebral lesions noted by MRI.

Comparisons with other cohorts.

We compared our results with those of large cohorts (>50 patients) for which adequate detail was provided in publications in order to assess the most common presenting features of disease, course of illness, treatment, and outcomes. Certain large series, with as many as 4,700 cases, are not included because of inadequate data for comparisons. We selected series from geographically diverse regions of the world, partly to determine if unique disease patterns could be attributed to race, ethnicity, or location. The NIH cohort was the only series in which sequential angiography of the entire aorta and primary branch vessels was required at regular intervals in all patients over the entire followup period. That series is the only one that can be compared with our own concerning vascular stability or progression in both asymptomatic and symptomatic patients.

NIH cohort.

The NIH TA population (3) was comparable to the CCF TA population in its demographic features (97% female, 75% white, median age at onset 25 years) (Table 2). The median followup period was 5.3 years. Ten percent of patients were asymptomatic at the time of diagnosis. Constitutional symptoms were present in 43% of patients at the time of presentation, and were defined as malaise (33%), fever (22%), weight loss (>10% loss compared with baseline noted in 12%), and night sweats (2%). Vascular symptoms were initially present in almost all patients and included arm claudication (35%), leg claudication (28%), lightheadedness/dizziness (18%), visual aberration (10%, including 1 case of blindness), carotodynia (17%), and signs and symptoms of cardiac disease (17%). Eighteen percent of patients initially complained of muscle or joint pain and 18% had headaches.

Table 2. Comparison of demographic information and diagnostic evaluation with other Takayasu arteritis cohorts*
Cohort (ref.)No. of patientsAge at onset, yearsWhite, %Female, %Underwent vascular imaging, %Underwent PTA/surgical intervention, %Mortality, %
  • *

    PTA = percutaneous transluminal angioplasty; CCF = Cleveland Clinic Foundation; NIH = National Institutes of Health; NR = not reported.

  • Values are the median for the studies reported in refs.3 and11; other values are the mean.

  • Diagnosis and disease features in the entire cohort were assessed at least once by vascular imaging, at the time of surgery or at autopsy.

  • §

    Based on the reported number of subjects who underwent autopsy.

CCF (present study)75269289100483
NIH (3)60257597100503
Italy (6)10440998810050NR
India (7)10627NR61901311
Japan (8)5224NR81901312§
Japan (9)84NRNR8664NR7
Japan (10)12030093NR1213
Mexico (11)107NRNR841002415

Common initial physical findings were vascular bruits (23%), diminished or absent pulse(s) (22%), asymmetric blood pressure (13%), hypertension (17%), carotodynia (17%), and murmur of aortic regurgitation (8%). The most frequent sites of disease were the aorta (65%) and arch vessels (subclavian arteries 93%, common carotid 58%, innominate 27%). Ninety-eight percent of patients had stenotic lesions and 27% had aneurysms. The most common sites requiring surgical intervention were the aortic arch and the arch primary branch vessels. Fifty percent of all patients required at least one revascularization procedure.

Forty-five percent of patients in the NIH study had at least one clinically apparent relapse. However, if a relapse was defined as new lesions in new vascular territories on sequential imaging studies, then relapses in the NIH cohort occurred in >80% of all treated patients. In our longitudinal cohort, 90% had a disease relapse based on clinical and laboratory features, and 53% had new vascular abnormalities in new territories, suggesting that TA in US patients is usually a chronically active, relapsing disease. In both cohorts, relapsing disease required long-term therapy in most patients. While glucocorticoids and adjunctive therapies were clearly palliative, they did not produce enduring remissions that allowed for successful taper and discontinuation of glucocorticoids in most patients. In both cohorts, either partial or permanent disability was noted in the majority of patients.

TA in Italy.

Vanoli et al studied a multicenter cohort of 104 Italian patients with TA in 2005 (6). Ninety-nine percent were white, 88% were women, and the mean age at onset was 40 years. Of note, 17% of this cohort was diagnosed as having TA at >40 years of age (range 41–74 years). In this older group, signs and symptoms of temporal artery abnormality and polymyalgia rheumatica were absent. At disease onset, ∼60% of patients had asthenia, ∼40% fever, ∼25% myalgia, and ∼30% arthralgia. Weight loss was noted in ∼20%.

Only 59% of the cohort had undergone complete large vessel angiography. The aortic arch and/or one of the primary branch vessels were affected in almost all patients. The abdominal aorta was affected in 39% of cases. As we and others have reported, in the Italian series, sites of stenoses were far more common (93% of patients) than dilatation (16%) and aneurysms (7%). Thirteen percent of patients presented with inactive disease and never received antiinflammatory/immunosuppressive therapies. Although not assessed by sequential and complete vascular imaging, their illness did not become apparently active over the course of the followup period.

Eighty-six percent of patients were treated with glucocorticoids and 54% were treated with cytotoxic agents. Fifty percent of patients underwent at least 1 surgical procedure, a large subset that was similar to our cohort and that of the NIH (48% and 50%, respectively). Other sources of morbidity, relapse rates, disability, and mortality results were not reported. Although the authors did not perform sequential imaging in all patients, and the mean/median period of followup was not noted, they were impressed by the high frequency of new vascular lesions over time and suggested that the long-term effectiveness of therapy may not be as good as has been previously reported.

TA in India.

The reported experience in India has noted definite differences between those cohorts and others from Japan and Western countries. In 1996, Jain et al described the clinical and vascular manifestations of TA in 106 Indian patients (7). Sixty-one percent of the patients were women, somewhat less than the subset of ≥85% women noted in other countries. The mean age at onset was 27 years. Constitutional symptoms at presentation in this cohort were defined as fever, weight loss, and arthralgias, which occurred in only 16% of patients, again less than that noted in other ethnic cohorts from other geographic regions.

The most common presenting symptoms among Indian patients were headache (44%), dyspnea (26%), syncope/“giddiness” (26%), palpitations (19%), visual aberration (12%), and stroke (10%). The most frequent findings at presentation were hypertension (52%) and vascular bruits (68%). At the time of first hospital admission, 77% of patients were hypertensive. Common sites of involvement included the abdominal aorta (abdominal aorta with or without branch vessel disease in 27% and abdominal aorta as part of more widespread aortic disease in 56%), and subclavian (59%) and renal (53%) arteries. Ninety-five percent of patients had stenotic or occluded vessels, while only 14% had aneurysms, values consistent with those in all other studies.

The authors state that “most patients present in the chronic phase,” and do not require therapy. However, the ESR was one parameter used to judge disease activity and was elevated in 60% of patients at presentation. Only 17 received immunosuppressive therapy (16 with glucocorticoids, 1 with CYC also). The response to glucocorticoid therapy was said to be satisfactory in only 5 of 16 treated patients. Ten patients underwent surgical intervention, and 4 underwent percutaneous transluminal angioplasty. Relapse rates were not specifically noted, and the prevalence of overall disability was not reported. However, the incidence of congestive heart failure (12%) and hemiparesis (9%) demonstrated significant morbidity from TA in this cohort. Mean followup period information was available in 69 patients, among whom mortality was 17% (6 patients died of congestive heart failure, 3 of renal failure, and 1 each of stroke, myocardial infarction, and ruptured aneurysm).

TA in Japan.

In 1969, Ueda et al published clinical observations of a cohort of 52 patients (8). Constitutional symptoms were frequent in this cohort, occurring both prior to and following signs and symptoms of vascular involvement (an exact prevalence is not documented for individual signs and symptoms). The most common finding was weak or absent pulse in 62% of patients. The most common sites affected were the aorta (arch 46%, abdominal 38%, thoracic 31%), and left subclavian (52%) and left carotid (40%) arteries. Forty percent received treatment with glucocorticoids, and 13% underwent surgical intervention. Disability and relapse rates were not reported.

Nakao et al described clinical and arteriographic findings in a multicenter cohort of 84 TA patients (9). Constitutional symptoms included fever (17%), malaise (29%), night sweats (6%), and weight loss (4%) that occurred around the time of disease onset. In this cohort, an abnormal acute-phase response was documented in 61% of patients, based on elevation of the ESR and 44% based on positive CRP levels. As in most series, the aorta and arch vessels were the most common sites of involvement, with 87% of patients having aortic arch disease and 70% having lesions in the descending and abdominal aorta. Interventions in this group included glucocorticoids (35%) and anticoagulation (14%). A poor response to glucocorticoids was noted in 45% of those treated. Disability and relapse rates were not reported.

Ishikawa and Maetani followed up 120 TA patients for a median period of 13 years, focusing on prognostic factors of disease (10). Eighty percent received glucocorticoid therapy and 41% received either short-term or long-term anticoagulation. Surgical intervention was required in 12% of patients. Relapse rates and frequency of disability were not reported. The 15-year overall survival rate was 83%, but was much lower (66%) in those patients with a major TA disease complication (e.g., retinopathy, hypertension, aortic regurgitation, or arterial aneurysm). The median age at time of death was 48 years. The causes of death were known in 14 of 16 patients (13% mortality for cohort) and included congestive heart failure (5 patients), stroke (4 patients), postoperative complications (3 patients), and acute myocardial infarction (2 patients).

TA in Mexico.

Lupi-Herrera et al described the clinical and anatomic features of TA in 107 patients in 1977 (11). Eighty-four percent of patients were female and 80% were between 11 and 30 years of age. Systemic signs or symptoms were noted in most patients and included asthenia (56%), weight loss (22%), or fever (18%). Other common symptoms were dyspnea upon exertion (72%), headache (57%), arthralgias (53%), palpitations (43%), and claudication (29%). Frequent abnormal physical findings were pulse deficits (96%), bruits (94%), and hypertension (72%). The vessels most often affected were the thoracic aorta (ascending 27%, descending 67%), and the subclavian (85%) and renal (62%) arteries. Apart from a higher frequency of hypertension and renal artery stenosis, these findings were consistent with those of the American, Italian, and Japanese cohorts.

Only 8 patients (7%) received glucocorticoid therapy, of whom 6 did not achieve remission. TA was described as a continuous process for which glucocorticoid therapy was not considered to be indicated in most patients. Surgical procedures were performed in 26 patients, in 22 instances for the treatment of hypertension. Disability rates were not reported. The mean mortality rate was 15% among patients for whom followup was available over a period of 19 years. Deaths were due to congestive heart failure (7%), renal failure (2%), and in 1 case (1%) each, myocardial infarction, stroke, rupture of a subclavian aneurysm, and gastric ulcer perforation.


  1. Top of page
  2. Abstract

Our experience is derived from the largest US cohort of TA patients studied to date. Our findings are consistent with those reported in the only other large, longitudinal series in the US, that of the NIH. In both studies, young women were affected in ∼90% of cases, patients were usually white (75% and 92%, respectively), and the subclavian artery was the most common site of involvement. Fewer than one-fifth of patients in both series had a self-limiting illness or a course of disease that resolved after a short period of therapy. In nearly all cases, high-dose glucocorticoid therapy resulted in initial disease remission that was followed by relapse after a gradual reduction in the glucocorticoid dose. The addition of other immunosuppressive therapies, especially cytotoxic agents, often allowed for further reductions in the glucocorticoid dose, but was still followed by relapse after additional attempts to discontinue glucocorticoid treatment. At the time of relapse, the average daily dose of prednisone in our longitudinal cohort was 13 mg (median 10 mg/day), a dose well known to produce considerable risk of adverse effects.

Acute-phase reactant values were normal and misleading in approximately one-fifth of patients with active disease, whereas nearly three-fourths of patients with active disease had concordant increases in acute-phase reactants. Therefore, acute-phase reactants are not adequately sensitive to rely on as a definite measure of disease quiescence, but abnormal levels of acute-phase reactants, in the absence of comorbid inflammatory processes, should encourage further investigation to identify additional evidence of disease activity.

In both our experience and that of the NIH, achieving sustained vascular patency by angioplasty was relatively unsuccessful. This observation contrasts with better results that have been reported in Indian patients (7, 12). While this may reflect ethnic or practice differences, it may also relate to median followup periods being relatively brief in the latter studies. Our success (e.g., patency, with no need for further intervention) for arterial bypass/reconstruction was almost 3-fold better (68%), over a 3-year followup period, than that for angioplasty (25%). While both our series and that of the NIH had relatively low mortality rates (NIH 3%, median followup period 5 years; CCF 4%, median followup period 3 years), chronic morbidity and disability were frequent in this relatively young population.

The US experiences are, in certain regards, similar to those in other countries, including Italy, Japan, and Mexico. Each of these TA cohorts had a predominance of female patients, and disease onset was most frequent in the second and third decades of life. Not all large series provided detailed assessments of features at presentation and during followup. The NIH, Italian, and Indian cohorts most clearly outlined manifestations of TA that occurred at disease onset (Figure 4). Longitudinal vascular involvement data were provided in the NIH, Italian, Indian, Japanese, and Mexican cohorts (Figure 5). In all series, aortic arch branch vessels were commonly involved, and the predominant anatomic lesion was stenosis or occlusion. The most frequently affected arch vessel was the subclavian artery. While the thoracic aorta was commonly involved in the US, Mexican, and Japanese series, the abdominal aorta was more frequently affected in Indian and Italian patients. Another important difference was noted concerning renal artery stenosis, which was more frequent in patients in the Mexican and Indian cohorts, in whom hypertension was among the most common presenting features.

thumbnail image

Figure 4. Manifestations of Takayasu arteritis at the time of disease onset in the American, Italian, and Indian cohorts. Instances in which data were not provided for a cohort are indicated by the absence of a comparison bar. HA = headache; sx = symptoms; BP = blood pressure; Dim = diminished; HTN = hypertension; TIA = transient ischemic attack; CVA = cerebrovascular accident; CHF = congestive heart failure; CCF = Cleveland Clinic Foundation; NIH = National Institutes of Health.

Download figure to PowerPoint

thumbnail image

Figure 5. Frequency of vascular involvement during the course of disease. Instances in which data were not provided for a cohort are indicated by the absence of a comparison bar. CCF = Cleveland Clinic Foundation; NIH = National Institutes of Health.

Download figure to PowerPoint

Because it was not customary to use antiinflammatory/immunosuppressive agents in India or Mexico, and detailed information about efficacy was not available from the Japanese series, comparisons of specific treatment outcomes are only available from the CCF and the NIH series. However, it was generally noted across cohorts that many patients improved with glucocorticoid therapy, but relapsed after glucocorticoids were tapered, leading to repeated courses of treatment or the addition of other immunosuppressive agents. Additionally, it was only in the 2 US cohorts that routine sequential imaging was used to assess the effectiveness of therapy in arresting anatomic progression in all patients. In both studies, glucocorticoids and adjunctive immunosuppressive therapies were shown to control disease in most patients, but not to sustain remission when glucocorticoids were tapered toward discontinuation. Relapse and anatomic progression were seen in the majority of patients.

Procedures to restore vascular patency or to repair aneurysmal disease were performed in approximately one-half of US and Italian patients. These interventions were used considerably less often in Japanese, Indian, and Mexican patients. The higher mortality rates in the Indian and Mexican cohorts could have several explanations, including ethnic differences that influence disease phenotypes and severity of disease expression, differences in medical therapy (e.g., less frequent use of glucocorticoids and cytotoxic agents), and variations in access to medical and surgical therapy. The lack of data does not allow for an analysis of these factors.

Our study has several important limitations. We were unable to collect complete longitudinal data on patients who were seen only on an intermittent consultative basis. This is a problem that is intrinsic to most tertiary referral centers, where patients often travel from other regions or countries for single or infrequent visits. Furthermore, our analysis was performed as a retrospective review of a cohort acquired at various periods in the course of their disease. Prospective enrollment and data collection from the time of diagnosis would be ideal, but is more difficult to achieve with rare diseases.

The strengths of our study include evaluation by a single provider, a comprehensive standardized approach to collection of historic data, complete and at least yearly vascular imaging, and routine collection of laboratory data in all patients to assess disease status and progression. Each visit specifically addressed elements of the disease state, e.g., the absence or presence of disease activity, relapses, or sustained remission.

Medical treatments for TA are severely flawed. While symptomatic improvement usually follows high-dose therapy with glucocorticoids, at least three-fourths of patients receiving close followup care have numerous relapses with dosage reduction. Relapses remain common despite the use of adjunctive immunosuppressive therapies. These results are supported by longitudinal data from 2 large US series. The relatively low rate of disease relapse that we have reported in patients receiving anti-TNF therapy is noteworthy (13). However, confirmation of these observations in rigorous randomized controlled studies is needed. The failure of conventional medical therapy for TA has led to vascular surgery being a common salvage procedure. At least half of these relatively young patients will eventually develop ischemic symptoms or aneurysms that require one or more surgical interventions. The frequency of significant morbidity and disability among affected patients with TA in the US speaks to their unmet medical and surgical needs.


  1. Top of page
  2. Abstract

Dr. Maksimowicz-McKinnon had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Maksimowicz-McKinnon, Hoffman.

Acquisition of data. Maksimowicz-McKinnon, Clark, Hoffman.

Analysis and interpretation of data. Maksimowicz-McKinnon, Hoffman.

Manuscript preparation. Maksimowicz-McKinnon, Hoffman.

Statistical analysis. Maksimowicz-McKinnon.


  1. Top of page
  2. Abstract