Long-term outcome for 120 Japanese patients with Takayasu's disease. Clinical and statistical analyses of related prognostic factors.

Abstract

BACKGROUND Patients with Takayasu's disease, a chronic inflammatory arteriopathy of unknown cause, have variable clinical courses, and predictors of the long-term outcome are not well understood. We studied prognostic factors of this disease, based on follow-up results, and a new prognostic classification was proposed. METHODS AND RESULTS Life-table methods and Cox regression analyses were applied to clinical data on 120 patients who had been prospectively followed for a median of 13 years (range, 1 month to 34 years). The overall survival rate at 15 years after the diagnosis was 82.9% and remained the same for the remainder of the follow-up period. Univariate Cox analyses revealed that of the six dichotomous variables evaluated at diagnosis, four were statistically significant predictors, including complications (Takayasu's retinopathy, hypertension, aortic regurgitation, and aneurysm), pattern of the past clinical course, age of the patient, and calendar year of diagnosis. Thus, the 15-year survival was 66.3% versus 96.4% for patients with and without a major complication, 67.9% versus 92.9% for patients with and without a progressive course, 58.3% versus 92.7% for age > 35 years and < or = 35 years, and 79.9% versus 96.5% for patients diagnosed in 1957 through 1975 and in 1976 through 1990, respectively. The delay in diagnosis and the erythrocyte sedimentation rate (ESR) were of marginal significance. The multivariate Cox analysis showed that only two of the above variables were statistically independent predictors, ie, the major complication and the progressive course. In addition to these two factors, ESR was the third predictor, if the Cox stepwise elimination procedure was performed. These three predictors used in various combinations made a total of 1822 classifications theoretically feasible. Of these, a three-stage classification was selected as the best one, based on the Akaike information criterion. The presence of both major complication and progressive course (stage 3) was the worst prognostic indicator (43% survival at 15 years). In contrast, no patient died who had neither of these manifestations or who had a progressive course but an elevated ESR as well (stage 1). CONCLUSIONS The long-term outcome for patients with Takayasu's disease seems best predicted by two major prognostic factors, ie, complications and the pattern of the past clinical course, as well as by ESR. Aggressive medical and surgical treatment may be considered for patients with a major complication and a progressive course (stage 3).