Abstract
OBJECTIVE: The aim of this study is to examine the relationship between serum complement component 4 (C4) levels and tubular atrophy in patients with lupus nephritis (LN). METHODS: Patients diagnosed with LN through renal biopsy were retrospectively analyzed and categorized into two groups: the tubular atrophy group (TA group) and the non-tubular atrophy group (non-TA group). Demographic data, clinical characteristics, and pathological findings were compared between the groups. Logistic regression and spline smoothing plot analysis, utilizing the generalized additive mixed model, were employed to investigate the relationship between serum C4 levels and the occurrence of tubular atrophy. RESULTS: A total of 129 patients were included in the study, with 57 (44.2%) identified as having tubular atrophy (TA group), while the remaining 72 patients did not present with tubular atrophy (non-TA group). Patients in the TA group exhibited higher serum C4 levels (p = 0.008) and lower eGFR (p = 0.001) compared to those in the non-TA group. Pathological findings revealed that the TA group had a higher incidence of mesangial hypercellularity (p = 0.041), endocapillary hypercellularity (85.96% vs. 68.06%, p = 0.018), interstitial inflammation (p < 0.018), karyorrhexis (78.95% vs. 59.72%, p = 0.02), and arteriole hypertrophy (66.67% vs. 38.89%, p = 0.02). Multivariate logistic regression analysis revealed that elevated serum C4 levels were associated with tubular atrophy in patients with LN (OR = 1.069, 95% CI 1.001 - 1.141, p = 0.048). Furthermore, spline smoothing analysis using the generalized additive mixed model indicated a linear relationship between serum C4 levels and the occurrence of tubular atrophy events. This linear relationship remained consistent across various stratifications, including age, hypertension, leukocyte count, hemoglobin levels, and estimated glomerular filtration rate (eGFR). CONCLUSION: Serum C4 levels demonstrate a significant correlation with the presence of tubular atrophy in patients diagnosed with LN.