Predicting Active Pathological Lesions in Renal Tissue of Lupus Nephritis Using Urinary and Clinical Findings

Background: Kidney biopsy is essential to detect active lesions in lupus nephritis, but carries bleeding risk. Guidelines recommend biopsy based on urinary findings such as hematuria and proteinuria, yet these criteria lack robust evidence. Identifying non-invasive and evidence-based predictors of active lesions may help optimize biopsy decision-making. Methods: This was a single-center, retrospective observational study including lupus nephritis patients undergoing native kidney biopsy between 2008 and June 30, 2024. A total of 131 patients (183 biopsies) were evaluated. Clinical and laboratory data at the time of biopsy were collected, including hematuria, proteinuria, and serologic markers. Hematuria was assessed using urine RBC ≥5/hpf across up to five samples within 3 months prior to biopsy. Patients were classified into two groups: persistent hematuria (all samples positive) and no hematuria (all negative). Predictors of active lesions were analyzed by univariate and multivariate logistic regression. Subgroup analyses were performed to compare findings between initial-onset and relapsed cases. Results: Among 181 kidney biopsies, 117 (64.6%) showed active lesions. Multivariate logistic regression identified eGFR, persistent hematuria, and anti-dsDNA antibody >22 IU/mL as independent predictors of active lesions. The odds ratios (ORs) and 95% confidence intervals (CIs) were: eGFR (OR 0.83, 95% CI 0.73–0.94, p=0.010), persistent hematuria (OR 4.09, 95% CI 1.65–10.18, p=0.002), and anti-dsDNA antibody >22 IU/mL (OR 10.48, 95% CI 4.37–25.16, p<0.0001). Proteinuria ≥0.5 g/day, a commonly used threshold in current biopsy guidelines, was not a significant predictor (p=0.190), and this remained consistent in both initial-onset and relapsed cases in subgroup analysis. Notably, the combination of persistent hematuria and anti-dsDNA antibody >22 IU/mL was observed in 33 cases, yielding a specificity and positive predictive value (PPV) of 100% for predicting active lesions. Conclusion: Persistent hematuria and anti-dsDNA antibody >22 IU/mL were strong predictors of active renal lesions, while proteinuria ≥0.5 g/day—currently used in biopsy guidelines—was not. These findings suggest that hematuria, often undervalued, may offer better predictive value and should be reconsidered in biopsy decision-making.