A small cohort study of rituximab-naive patients with refractory primary membranous nephropathy found a clinical response in 26 of 35 evaluable patients after 12 months. Nine had complete remission and 17 had partial remission, for a response rate of 74.3%; among the 22 patients with 24-month data, 21 were in remission, a complete-case rate of 95.5%.
But the 24-month number is fragile. Only 22 of the 36 enrolled patients had data at that point, and the result changed sharply depending on how missing follow-up was handled. An intention-to-treat Kaplan-Meier analysis estimated 91.67% remission, with a 95% confidence interval of 72.3% to 99.2%; counting every dropout as a non-remitter reduced the estimate to 58.3%, or 21 of 36 patients.
The treatment and its yardstick
The study enrolled 36 patients and evaluated a combination regimen. Each received two intravenous rituximab infusions of 1,000 milligrams, given two weeks apart, alongside prednisone-equivalent glucocorticoids that started at 0.25 to 0.5 milligrams per kilogram per day and were tapered according to response. A second rituximab course could be considered at month 6 when proteinuria or anti-PLA2R antibodies remained high.
Remission was defined in two steps. Partial remission required urine protein below 3.0 grams per gram of creatinine or below 3.0 grams in 24-hour urine, together with at least a 50% reduction from baseline. Complete remission required proteinuria below 0.3 grams per gram of creatinine or 0.3 grams in 24 hours, plus serum albumin above 35 grams per litre.
The original protocol was built around a 12-month endpoint and planned 31 evaluable patients. Its assumptions included a 60% historical remission rate with rituximab monotherapy, an 85% remission rate for the combination, 80% statistical power and 10% attrition. Registration in 2020 was retrospective, and the revised 24-month analysis, with 22 evaluable patients, was underpowered.
The measurements changed, too
The laboratory results moved in the same direction as the clinical response. At 12 months, albumin rose from 20.6 to 37.9 grams per litre, while 24-hour proteinuria fell from 6.68 to 1.34 grams. Immunoglobulin G rose from 3.99 to 7.52 grams per litre, and total cholesterol fell from 6.80 to 5.05 millimoles per litre. The study reported no significant change in eGFR, which moved from 77.20 to 81.78 millilitres per minute per 1.73 square metres.
Most of the reported improvements remained significant at 24 months after the study's false-discovery-rate adjustment, while the triglyceride improvement did not. The uncontrolled design means these measurements cannot isolate the contribution of the combination itself.
Anti-PLA2R antibodies, measured in blood, also fell in many patients. Among the 25 who were positive at baseline, 20, or 80%, reached antibody clearance below 20 relative units per millilitre. But only four, or 16%, met the stricter definition of complete immunological remission, which required levels below 2 relative units per millilitre on two consecutive measurements.
Clinical remission did not differ significantly by baseline antibody status: 84% in the seropositive group versus 80% in the seronegative group. Complete remission was 44% versus 50%, respectively. The findings separate antibody clearance from complete immunological remission, which used a stricter threshold.
Repeat treatment raised a question
Repeat dosing was the clearest exploratory signal. At month 6, 23 patients received a second rituximab course, and 20 of them, or 87.0%, had responded by 12 months. Among the 12 who received only one course, six, or 50.0%, responded. The difference was statistically significant in the study's comparison, but retreatment was not randomized, so the numbers cannot show that the second course itself caused the higher response.
Among 24-month completers, remission was maintained in 94.4% of those who received retreatment versus 50.0% of those who declined it. This was a post-hoc, selected comparison, not a randomized test of durability. One patient, or 2.9%, relapsed 18 months after complete remission and regained remission after rituximab retreatment.
What the study cannot answer
The safety results included both adverse events and serious infections. Adverse events were reported in 13 of 35 patients, or 37.1%, and serious infections in four, or 11.4%. No serious malignant or fatal events were reported, but steroid toxicities were not systematically graded and the follow-up was limited.
The main limitation is that the study had no concurrent control group and included repeated rituximab dosing as well as residual effects from prior immunosuppression. It therefore cannot establish whether adding low-dose glucocorticoids improved outcomes beyond rituximab alone. The small sample, attrition and underpowered 24-month analysis make the findings preliminary rather than a basis for deciding which regimen is superior.
The work was supported by grants from the National Natural Science Foundation of China, the Guangdong Basic and Applied Basic Research Foundation and the Wu Jieping Medical Foundation Research Grant.
Paper data and sources
Original title: Rituximab with low-dose glucocorticoids for treating refractory membranous nephropathy: a prospective single-center cohort study
Authors: Li Y, Jiang Y, Jiao W et al.
Journal/Repository: Frontiers in pharmacology
Status: Peer-reviewed
First online: 2026-10-10
DOI: Not available
Original paper · Full text