Renal relapse in antineutrophil cytoplasmic autoantibody-associated vasculitis: unpredictable, but predictive of renal outcome

Objectives. To determine predictors of renal relapse and end-stage renal failure (ESRF) in patients with ANCA-associated vasculitis. Methods. Data from four European Vasculitis Society randomized controlled trials, conducted roughly simultaneously between 15 March 1995 and 30 September 2002, was pooled to determine predictors of long-term renal outcome. The respective trial inclusion criteria covered the entire spectrum of disease severity. Baseline predictors of time to first renal relapse and time to ESRF were assessed by competing events analysis and Cox proportional hazards regression. The effect of renal relapse on time to ESRF was assessed by adding renal relapses to the competing events analysis as a time-varying covariate. Results. The number of patients participating was 535; mean serum creatinine (+/- S.D.) at entry was 341 +/- 321 mu mol/l and 19.7% developed ESRF. One or more renal relapse(s) was experienced by 101 patients. Multivariable regression analysis demonstrated that, in addition to impaired baseline renal function, developing >= 1 renal relapse was an independent risk factor for ESRF (subhazard ratio 9; 95% CI 4, 19; P < 0.001). No predictive factors for renal relapse were found. Conclusion. In addition to baseline renal function, the occurrence of renal relapses is an important determinant of ESRF in patients with ANCA-associated vasculitis. We did not find any clinical predictors for renal relapse itself, including disease activity elsewhere. In light of the silent nature of renal relapse in ANCA-associated vasculitis, we stress the need for long-term vigilant monitoring for early signs of renal relapse and propose performing 3-monthly urinalysis. This will enable timely treatment and help further improve renal outcome.