Clinical Significance of Glomerular NELL1 and PLA2R Staining in Membranous Nephropathy Associated with Rheumatoid Arthritis

Since the discovery of phospholipase A2 receptor (PLA2R) as a target antigen in membranous nephropathy (MN), the disease has been reclassified according to its antigenic targets rather than the traditional distinction between primary and secondary types. In patients with rheumatoid arthritis (RA), Neural epidermal growth factor-like 1 (NELL1)-positive MN induced by the anti-rheumatic drug bucillamine has been reported. However, MN can also develop in RA patients without bucillamine exposure, and the full spectrum of causative antigens in RA-associated MN remains unclear. A deeper understanding of these antigenic patterns is essential for accurate diagnosis, treatment selection, and prognostic evaluation.

We performed immunohistochemical staining for PLA2R and NELL1 on frozen kidney biopsy specimens from 18 patients diagnosed with RA-associated MN between 1998 and 2018 at our hospital and affiliated institutions. Anti-NELL1 antibody (HPA051535, Sigma-Aldrich) and anti-PLA2R antibody (HPA012657, Atlas Antibodies) were used. The positivity rates of each antigen were evaluated, and their associations with clinical findings at biopsy and complete remission (CR) one year after biopsy were analyzed. IgG subclass staining was also performed to examine differences in immunoglobulin profiles between antigen groups.

Among the 18 patients, 11 (61.1%) were female, and the median age at biopsy was 61 years. The median urine protein-to-creatinine ratio was 3.3 g/gCr, median serum albumin 2.9 g/dL, and median serum creatinine 0.78 mg/dL. Immunostaining revealed NELL1(+)/PLA2R(–) in 5 cases, NELL1(–)/PLA2R(+) in 3, NELL1(+)/PLA2R(+) in 3, and NELL1(–)/PLA2R(–) in 7. Except for a higher frequency of bucillamine use in NELL1-positive patients, there were no significant intergroup differences in baseline parameters such as proteinuria or kidney function. Bucillamine was used in 10 patients, 8 (80%) of whom were NELL1-positive, including 2 who were also PLA2R-positive. IgG subclass analysis showed all PLA2R-positive cases were IgG4-positive, while all NELL1-positive cases were IgG1-positive, with a relatively low IgG4 positivity rate (50%). One year after biopsy, all bucillamine-treated patients had discontinued the drug. CR rates were highest in the NELL1(+)/PLA2R(–) group (80.0%), followed by NELL1(–)/PLA2R(–) (57.1%), NELL1(–)/PLA2R(+) (33.3%), and NELL1(+)/PLA2R(+) (0%).

RA-associated MN comprises two major subsets: drug-induced NELL1-positive cases and those caused by other antigens such as PLA2R. NELL1-positive MN tends to achieve remission rapidly after discontinuation of the causative drug, whereas PLA2R-positive MN shows a lower remission rate at one year. Importantly, MN antigens are not always singular; some patients exhibit dual positivity for NELL1 and PLA2R. In cases not achieving remission after bucillamine withdrawal, coexisting PLA2R-related MN should be suspected. Moreover, NELL1(–)/PLA2R(–) cases may represent MN caused by yet unidentified antigens, warranting further study. These findings suggest that antigenic diversity in RA-associated MN affects clinical outcomes, and in patients with multiple target antigens, the course may reflect combined features of both NELL1 and PLA2R. Understanding this antigenic heterogeneity is essential for accurate classification and individualized management of MN in RA.