Histological predictors of renal outcome in lupus nephritis: the importance of tubulointerstitial lesions and scoring of glomerular lesions

Introduction Lupus nephritis (LN) affects nearly 60% of patients with systemic lupus erythematosus and up to 30% of them will progress to end-stage renal disease (ESRD), despite receiving aggressive immunosuppressive therapy. The prognostic value of ISN/RPS classification is controversial. Therefore, we aimed to identify clinical and pathological predictors of outcome in LN patients independent of this classification. Material and methods Thirty-seven patients with LN who underwent percutaneous kidney biopsy between 1997 and 2016 were included in this study. Twenty clinical and twenty histological variables were tested for their association with a composite end-point of doubling of serum creatinine, ESRD and death. Univariate and multivariate Cox proportional hazard regression analysis were performed to identify independent predictors of outcome. Results During a median follow-up period of 48 months (IQR: 17.5–120 months), 21.6% of patients reached the composite end-point. The overall survival rate of our cohort was 89% at one year, 86% at five years, 74% at 10 years and 64% at 20 years. Patients with Class IV LN showed the worst prognosis with 44% survival at 10 years, while those who additionally showed crescents and global sclerosis on kidney biopsy had an even lower survival of 21% and 0% at 10 years, respectively. After multivariate adjustment, we identified estimated glomerular filtration rate at baseline (HR, 0.91 per ml/min /1.73 m2; 95% CI, 0.84 to 0.99), 24-hour proteinuria at baseline (HR, 2.04 per g/d; 95% CI, 1.19 to 3.5), crescents (HR, 1.068 per %; 95% CI, 1.003 to 1.091), global sclerosis (HR, 1.036 per %; 95% CI, 0.984 to 1.091), presence of adhesions (HR, 9.2; 95% CI, 1.38 to 61.2) and tubulitis (HR, 13.1; 95% CI; 1.3 to 131) as independent predictors of outcome in our cohort of LN. Conclusions Our study identified glomerular (crescents, global sclerosis, adhesions) and tubulointerstitial (tubulitis) lesions, in addition to clinical variables (renal function, 24-hour proteinuria), as important predictors of renal outcome, independent of the ISN/RPS classification. We suggest that the ISN/RPS classification could be improved by a quantitative assessment of glomeruli with active and chronic lesions and by a greater emphasis given to tubulointerstitial lesions.