Takayasu's arteritis: operative results and influence of disease activity

To determine the short- and long-term outcomes of patients treated operatively for Takayasu's arteritis and the effect of disease activity on results.

Forty-two (17%) of the 251 patients enrolled in our Takayasu's arteritis registry between 1975 and 2002 required operation for symptomatic disease. Data were obtained from the registry, patient records, phone correspondence, and written surveys.

There were 38 females and 4 males with a median age of 29 years (range, 12 to 56 years), and 32 (76%) were white. Sixty operations were performed for symptomatic disease. The mean duration of symptoms before operation was 5.6 months (range, 0 to 25 months). Thirteen (31%) patients had active disease and underwent operation for acute presentation or failure of medical management. Thirty-nine patients (93%) had operation for occlusive disease. Twenty-two (52%) patients had involvement of both the great and abdominal aortic branch vessels; 10 (24%) had great vessel disease alone; 9 (21%) had involvement of abdominal arteries; and 1 (2%) had coronary artery disease. There was no operative death, myocardial infarction, major stroke, or renal failure. Three patients had early graft thrombosis, two had a minor stroke, and two developed hyperperfusion syndrome. The median follow-up was 6.7 years (range, 1 month to 19.3 years). Eleven (26%) patients required 15 graft revisions; five of the patients had active disease at the time of initial operation. All early revisions (<1 year) were in patients with active disease. By Kaplan-Meier analysis, freedom from revision at 5 and 10 years was 100% in patients with quiescent disease not requiring steroids (group I, n = 5, 12%), 95% and 81% in patients whose disease was quiescent on steroids (group II, n = 24, 57%), 57% in patients with active disease on steroids (group III, n = 7, 17%), and 33% in patients with active disease and no long-term steroids (group IV, n = 6, 14%) (P < .006). The rate of revision or progression of disease at another site in 5 years was 0% in group 1, 10% in group 2, 57% in group 3, and 67% in group 4 (P < .001) The differences were even more pronounced when an analysis was done on the basis of disease activity alone, irrespective of steroid use. During the follow-up period, 3 of 39 great vessel, 2 of 18 mesenteric/renal, and 1 of 9 aortofemoropopliteal reconstructions occluded. The predicted mortality for patients was 4% at both 5 and 10 years (95% CI) respectively (confidence interval [CI], 0% to 11%) and 10 (CI, 0% to 14%) years, respectively.

The minority of patients with Takayasu's arteritis require operation. In our predominantly white female patient population, occlusive symptoms were the most common indication for operation. Operation for these selected patients was safe, with no operative mortality, myocardial infarction, major stroke, or renal failure. Patients with active disease requiring operation are more likely to require revision or develop progressive symptomatic disease at another site. Long-term survival is excellent, regardless of disease activity at the time of operation.