Abstract Background and Aims Membranous nephropathy (MN) is a rare, but severe autoimmune disease affecting kidney glomeruli. Clinical evolution is variable, ranging from spontaneous remission to persistent nephrotic syndrome and kidney failure. The major autoantigen in MN is PLA2R1 protein associated with 50-70% of all cases. Patients may present antibodies against a single immunodominant CysR domain or may develop additional antibodies against CTLD1 and/or CTLD7 and/or CTLD8 domains of PLA2R1 defining a cascade immunization or epitope spreading. The mechanism of epitope spreading has been described in a variety of contexts: it can reinforce immune response to eliminate a pathogen, but can also be associated with the aggravation of autoantibody-mediated autoimmune pathologies. The value of this marker in the management of MN patients remains debated. Our initial study (confirmed on an independent cohort) has shown that the single domain recognition (non-spreader patients) is associated with a higher chance of spontaneous remission (45% vs 0.05% in GEMRITUX cohort) and a better response to treatment (100% of remission regardless of the dose of rituximab used). The patients with multi domain recognition (spreader patients) of CTLD1 and/or CTLD7 domains, on the other hand, have poor prognosis and are less likely to achieve remission requiring high doses of rituximab. Based on these studies we have proposed a personalized treatment protocol with either low or high dose Rituximab based on patients’ PLA2R1 epitope spreading status. The aim of this study was to evaluate the efficacy of this personalized treatment in comparison to the established GEMRITUX protocol. Method A multicenter, randomized, controlled, prospective clinical trial was conducted in 12 French hospitals. Sixty-four consecutive patients with PLA2R1-related MN were randomly assigned to either the control group following the GEMRITUX protocol (symptomatic treatment for 6 months, two 375 mg/m2 rituximab infusions at month-6 in case of persistent nephrotic syndrome (NS)) or the personalized treatment (PMMN) group (patients with no epitope spreading at month-0 were treated as per GEMRITUX protocol, while patients with epitope spreading at month-0 or month-6 with persistent NS were treated immediately with two 1 g rituximab infusions). The primary study outcome was the rate of clinical remission at month-12. The secondary outcomes were complete and partial remission, immunological remission, proteinuria, albuminuria, serum creatinine, and PLA2R1 antibody titer. Results From 64 patients included, three were excluded from final analyses due to loss to follow-up. There was no difference in age, gender, albumin, UPCR, anti-PLA2R1 titer and the rate of PLA2R1 epitope spreading at baseline between the two groups. At M12, 34% of patients from the GEMRITUX group and 69% of patients from PMMN group achieved partial clinical remission (P = .0105) defined as UPCR < 3.5 g/g with a decrease greater than 50% from baseline; improvement or normalization of serum albumin; and increase of serum creatinine lower than 20%. While there was no difference in remission rate between the GEMRITUX group and PMMN group for patients with single domain recognition (38% vs 50%, respectively, P = .7107), multi-domain recognizers (spreaders) were more likely to achieve remission with personalized protocol (36% vs. 87%, P = .0078). The rate of complete clinical remission between the two experimental groups, defined as UPCR <0.3 g/g and normal albumin, was close to statistical significance (0% vs 16%, P = .0538). Conclusion Personalized treatment protocol based on the stratification of patients with PLA2R1-related MN according to their epitope spreading status, is superior to the standard GEMRITUX protocol in achieving partial clinical remission at 12 months. Patients with multiple domain recognition should be treated immediately with high doses of rituximab to increase their chances of remission.