Abstract
BACKGROUND: Tubulointerstitial injury is important to predict the progression of lupus nephritis (LN). Urine neutrophil gelatinase-associated lipocalin (NGAL) has been reported to detect worsening LN disease activity. Thus, urine NGAL may predict renal outcomes among lupus patients. METHODS: We conducted a prospective multi-center study among active LN patients with biopsy-proven. All patients provided urine samples for NGAL measurement by ELISA collected from all patients at baseline and at 6-month follow-up after induction therapy. RESULTS: In all, 68 active LN patients were enrolled (mean age 31.7 ± 10.0 years, median UPCR 4.8 g/g creatinine level with interquartile range (IQR) 2.5 to 6.9 and mean estimated glomerular filtration rate (GFR) 89.6 ± 33.7 mL/min/1.73 m(2)). At baseline measurement, median urinary NGAL in complete response, partial response and nonresponse groups was 10.86 (IQR; 6.16, 22.4), 19.91 (IQR; 9.05, 41.91) and 65.5 (IQR; 18.3, 103) ng/mL, respectively (p = 0.006). Urinary NGAL (ng/mL) correlated positively with proteinuria and blood pressure, and correlated negatively with serum complement C3 level and estimated GFR. Based on ROC analysis, urinary NGAL (AUC; 0.724, 95%CI 0.491-0.957) outperformed conventional biomarkers (serum creatinine, urine protein, and GFR) in differentiating complete and partial response groups from the nonresponse group. The urine NGAL cut-off value in the ROC curve, 28.08 ng/mL, discriminated nonresponse with 72.7% sensitivity and 68.4% specificity. CONCLUSION: Urine NGAL at baseline performed better than conventional markers in predicting a clinical response to treatment of active LN except serum complement C3 level. It may have the potential to predict poor response after induction therapy.