Background/Purpose: Lupus nephritis (LN) remains an important cause of morbidity and mortality in patients with SLE.  Tubular atrophy and interstitial fibrosis on initial renal biopsies in Blacks with LN have been associated with poor renal outcomes. This study was initiated to evaluate the histopathologic predictors of poor renal outcomes on initial and subsequent renal biopsies in patients with lupus nephritis in a predominantly Hispanic and Black cohort.

Methods: All SLE patients with two or more renal biopsies performed at CUMC between 1996-2015 were identified from the pathology database; data on 63 patients with regular follow-up at the Columbia University Lupus Center were available for analysis.   Demographics, clinical, laboratory and histopathologic characteristics were ascertained. Adverse renal outcomes were defined as a serum creatinine >1.4 mg/dL, ESRD, or renal transplant. Tubular atrophy and interstitial fibrosis was graded by the renal pathologist from mild (10-25% involvement) to severe (>60% involvement).  The histopathologic predictors of adverse renal outcomes were evaluated using logistic regression, adjusting for pertinent confounders.

Results: Of the 63 patients, 87% were female, 46% Hispanic, 29% Black. The average age at SLE diagnosis was 27±12, the median time from the first renal biopsy to the occurrence of the first adverse renal outcome was 6 years (4-9) (Table 1). Fifty-three percent of patients with ≥ 2 biopsies had an adverse renal outcome: 31% developed ESRD (of which 58% received a renal transplant), and 22% had a serum creatinine >1.4 mg/dL. The degree of tubular atrophy and interstitial fibrosis (TAIF) in the second renal biopsy [OR 3.5 (CI 1.3-10.6) p=0.03] but not in the first [OR 2.5, (CI 0.7-8.2); p=0.15], was a strong predictor of adverse renal outcomes in a multivariate analysis, after adjusting for hypertension, smoking, race/ethnicity, LN class and activity index. Chronicity index was not a significant predictor of renal outcome in multivariable analysis.  All subjects with severe TAIF on the second and third renal biopsies met the primary outcome. Fibrinoid necrosis, number or type of crescent, number of sclerotic glomeruli and degree of foot process effacement were not significantly associated with a poor renal outcome.

Conclusion: Fifty-three percent of SLE patients that required a second renal biopsy for LN flare in this predominantly Hispanic and Black cohort had adverse renal outcomes within 6 years; 31% developed ESRD.  Tubular atrophy and fibrosis were strong predictors of poor renal outcome. Their presence on a second biopsy should prompt intensification of therapy in patients with lupus nephritis to avoid ESRD and adverse renal outcomes.