Primary membranous nephropathy (pMN) is an antibody-mediated cause of adult idiopathic nephrotic syndrome characterized by subepithelial immune complex deposition. In 70–80% of cases, autoantibodies target the M-type phospholipase A2 receptor (PLA2R). Anti-PLA2R antibody titers rise during active disease and decline before clinical remission, supporting a serology-based approach to PLA2R-related pMN. Disease relapse occurs commonly, reported in roughly 20%–30% of patients, and accurate early predictors of relapse are needed to personalize immunosuppressive therapy. Prior studies and trials suggest combined early assessment of serology and albumin may predict outcomes better than baseline measures alone.
This was a single-center, retrospective cohort study at Jining No.1 People’s Hospital. Consecutive adult patients with biopsy-proven pMN between January 2019 and October 2024 were screened. Inclusion criteria required baseline anti-PLA2R antibody ≥ 14 RU/mL, 24-hour urinary protein ≥ 3.5 g/d at biopsy, receipt of immunosuppressive therapy, complete clinical and pathological data, and follow-up ≥ 12 months. Patients with secondary membranous nephropathy, liver impairment, severe infections, or malignancy were excluded.
Laboratory data were collected at baseline and at 3, 6, 9, and 12 months and included 24-hour urinary protein, anti-PLA2R antibody titers (measured by a commercial quantum dot fluorescence immunoassay with an assay cutoff of 14 RU/mL), serum creatinine, eGFR, serum albumin, globulin, cholesterol, and uric acid. Renal pathology was staged using the Ehrenreich–Churg classification and included assessments of interstitial fibrosis/tubular atrophy, global and segmental glomerulosclerosis.
Patients were treated with one of three regimens: rituximab (RTX), cyclophosphamide plus glucocorticoids (CTX+GC), or calcineurin inhibitors with or without glucocorticoids (CNI ± GC). Specific protocols included tacrolimus or cyclosporine dosing and RTX 1,000 mg × 2 doses with individualized retreatment. Complete remission (CR) required proteinuria ≤ 0.3 g/d, serum albumin ≥ 35 g/L, stable eGFR, and anti-PLA2R < 14 RU/mL. Partial remission (PR) required proteinuria > 0.3 but < 3.5 g/d with ≥ 50% reduction from baseline, anti-PLA2R < 14 RU/mL, improved albumin, and stable eGFR. The remission cohort comprised patients achieving CR or PR within 12 months. Relapse was defined as recurrence of proteinuria ≥ 3.5 g/d after documented remission.
Analyses used SPSS, GraphPad, and R. Continuous variables were assessed for normality and compared with t tests or Mann–Whitney U tests as appropriate; categorical variables used chi-square or Fisher’s exact tests. Univariable and multivariable Cox proportional-hazards models evaluated associations with relapse. Landmark analyses at baseline, 3 months, and 6 months were performed. Model fit and discrimination were compared using the Akaike information criterion (AIC) and Harrell’s C-index. ROC curves and AUC assessed predictive performance; Kaplan–Meier curves and the log-rank test evaluated relapse-free survival. Sensitivity analyses excluded patients receiving specific immunosuppressive therapies.
A total of 154 PLA2R-associated pMN patients were included; median follow-up was 26 months (IQR 18–34). Ninety-six patients (62.34%) achieved remission within 12 months; median time from treatment initiation to first documented remission was 6 months (IQR 5–9). Compared with the non-remission group, the remission group had lower 24-hour urinary protein, lower serum creatinine, higher eGFR, higher serum albumin and globulin, and lower baseline anti-PLA2R titers [median 45.50 vs. 125.00 RU/mL; p < 0.001].
Among the 96 patients who achieved remission, 17 (17.71%) experienced relapse during follow-up. The median follow-up time for the remission cohort was 32 months (IQR 25.25–36.00). Median time from treatment initiation to relapse was 23 months and median time from documented remission to relapse was 15.5 months. Patients who relapsed were younger (median 47 vs. 54 years; p = 0.04) and had significantly higher baseline anti-PLA2R titers [median 71.00 vs. 37.00 RU/mL; p < 0.001].
In the 3-month landmark multivariable Cox regression model, anti-PLA2R seropositivity at 3 months (≥ 14 RU/mL) was independently associated with a markedly increased risk of subsequent relapse (HR 9.44; 95% CI 2.69–33.21; p < 0.001). Additionally, a greater increase in serum albumin from baseline to 3 months was independently associated with a lower relapse risk (HR 0.85 per unit increase; 95% CI 0.74–0.98; p = 0.03). The 6-month landmark model provided weaker predictive performance by comparison.
Comparing landmark models, the 3-month model demonstrated better discrimination and fit than the 6-month model: lower AIC (123.41 vs. 138.84) and higher Harrell’s C-index (0.83 vs. 0.66). ROC and Kaplan–Meier analyses were used in the study to evaluate performance and relapse-free survival, respectively.
In this exploratory single-center cohort of PLA2R-associated pMN, persistent anti-PLA2R seropositivity at 3 months after treatment initiation and a smaller increase in serum albumin over the first 3 months were identified as potential risk factors for relapse among patients who achieved remission. The 3-month landmark assessment outperformed a 6-month assessment in model fit and discrimination. The authors note that larger-scale studies are warranted to validate these findings.