Objective. To assess the clinical, laboratory, and renal biopsy predic
tors of long-term outcome in lupus nephritis and to investigate the ti
me-dependence of these predictors. Methods. Eighty-seven lupus nephrit
is patients were studied retrospectively for the outcomes renal failur
e and fatality due to renal involvement. In addition to a conventional
Cox model analysis, a new generalized time-dependent analytic approac
h was developed and used to assess the time dependence of a predictor
variable's importance. Results. The mean followup time was 11.9 years.
Renal failure (n = 19) was significantly predicted by measures of ren
al function (abnormal serum creatinine levels, proteinuria, duration o
f prior renal disease) and immunologic activity (elevated DNA binding,
hypocomplementemia, and thrombocytopenia), by overall lupus disease a
ctivity measures (le Riche index, Lupus Activity Criteria Count), and
by the activity index, the tubulointerstitial index, and the amount of
subepithelial deposits on renal biopsy. In general, the laboratory pr
edictors were significantly better prognostic markers in the early yea
rs after biopsy, the disease activity measures were best in the later
years, and the biopsy variables were significant predictors over the e
ntire observation period. In contrast to the renal failure outcome, th
e best predictors for death not directly related to lupus nephritis (n
= 17) were the extent of comorbid diseases (principally vascular dise
ases), older age, and the chronicity index. All three predicted well o
ver the extended observation period. Conclusion. The major predictor v
ariables for renal outcomes and nonrenal outcomes are distinct. The ti
me-dependence of the predictive ability of some variables may be impor
tant in managing individual patients. The new generalized time-depende
nt analytic technique may have widespread application in studies to id
entify prognostic factors for established disease or risk factors for
the development of disease.