Reassessing Rituximab in Membranous Nephropathy: Efficacy, Limitations, and the Path Ahead

See Clinical Research on Page 2630

Recent Kidney Disease: Improving Global Outcomes guidelines illustrate the progress in the treatment of membranous nephropathy (MN).¹ Rituximab (RTX) is preferred as initial treatment based on its low toxicity and the reported efficacy in inducing proteinuria remission. However, because only alkylating agents were studied in trials that evaluated long-term outcomes such as end-stage kidney disease or mortality, cyclophosphamide-based therapy (CP) remains a valid option in high-risk patients. Moreover, CP is recommended in “very high-risk” patients, that is, patients with severe nephrotic syndrome and/or deteriorating kidney function.

Vargas-Brochero et al.² report the long-term outcome of patients with MN treated with RTX. This single-center, retrospective study, included 159 patients, treated in the period 2000 to 2022. Median follow-up was 62 months. The abstract emphasizes the outcome of patients treated with RTX only, with a renal survival of 97% after 5 years and 95.4% after 10 years. The authors conclude that RTX treatment is associated with excellent long-term renal survival that compares favorably with survival rates reported for CP. One might conclude that the Kidney Disease: Improving Global Outcomes guidelines no longer hold. However, we caution against concluding too early.

The study has limitations. The study is retrospective, with RTX initiated at the discretion of the physician. Importantly, reporting outcomes in patients who only received RTX may overestimate efficacy because of censoring of nonresponders who required rescue therapy. The study provides confidence that patients who respond to RTX do not develop end-stage kidney disease, and strengthens a previous report that showed that proteinuria remission was associated with good outcomes.³

In other prospective studies, RTX therapy was associated with a failure rate (defined as absence of remission after 2 years) of 35% to 40%.4, 5, 6 In the current study, remission rate at the end of follow-up amounted 88.1%. This number probably included patients who responded to rescue therapy. The initial remission rate after RTX-only therapy can be estimated at 80%² (and noting that ≥ 7% of patients suffered from worsening kidney function, which is incompatible with being in partial remission).

Nevertheless, based on the data, there is an unexplained 20% discrepancy in remission rate between this study and the prospective studies. What are possible explanations?

A Difference in the Dose and Duration of Therapy

In the prospective studies, lower doses of RTX were used, ranging from 1 to 2 doses of 375 mg/m² in European studies⁵ to a cumulative dose of 4 g in MENTOR.⁶ In the current study, most patients received 2 g RTX at the start, and additional doses during follow-up. The median time to redose (mostly 1 g) was 7.3 months. Because the median cumulative dose amounted to 4 g (interquartile range: 3–6.7 g), it can be estimated that most patients received RTX for > 12 months after start of therapy and 25% of patients received treatment for ≥ 24 months. A longer duration of therapy (a higher cumulative dose) might explain the higher remission rate.⁷ However, a comparison with the MENTOR trial suggests that this is not the most likely explanation. In the MENTOR trial, all patients received 4 g of RTX, which is comparable to the initial treatment in the current study; however, there remains the discrepancy in remission rate at 24 months after treatment start.

Patient Selection

Because the study was retrospective and included patients with MN who received RTX at the discretion of the physician, selection bias is an issue. In our view, the overall study cohort did not include only high-risk patients; 15% of patients were nonnephrotic, and 25% of patients were treated for relapse (these patients have been in remission which clearly decreases long-term risks). Most importantly, the median antiphospholipase A2 receptor antibody (PLA2Rab) titer was relatively low (84 [34–181] RU/ml). It is well-established that PLA2Rab levels predict response to RTX.^4,8 For comparison, PLA2Rab levels in recent studies were higher, ranging from 110 to 165 RU/ml. ^4,8,9,S1^,S2^.

A more detailed analysis of the study of Vargas-Brochero et al.² is revealing. The study included 45 patients with apparent “treatment resistance.” The majority of these patients did not receive CP. These patients clearly had a high-risk profile, as reflected by the PLA2Rab titer of 162 (68–245) RU/ml. Outcome data of this subgroup are not provided. However, if we assume that relapsing patients (with median PLA2Rab levels of 28 RU/ml) generally will respond to RTX, it can be estimated that in the subgroup of 45 “resistant” patients, 15 failed to respond to RTX, or 33%, which is in line with the abovementioned trials.

Although the study of Vargas-Brochero et al.² has limitations, it provides important information on the long-term clinical course in a relatively large number of patients treated with RTX. It is evident that RTX therapy is well-tolerated, with few side effects. This supports the Kidney Disease: Improving Global Outcomes advise to prefer RTX as initial therapy in patients at moderate-to-high risk. However, RTX treatment is not the final answer for all patients. A relapse was observed in 61 patients (45% of patients in remission). This relapse rate is relatively high; because median time to relapse was 36 months, and median dose of RTX was 4 g (equal to treatment duration of > 12 months, and thus B-cell depletion > 18 months), this suggest that nearly half of the patients relapsed within 2 years after B cell recovery. Although these patients will respond to renewed therapy, a comparison between immunosuppressive agents should include persistent clinical remission without the need of maintenance therapy.

Toward Individualized Therapy

The introduction of immunosuppressive therapy has undoubtedly improved outcomes in patients with MN. It is now evident that intensified immunosuppressive therapy results in almost 100% remission rates. This is observed with alkylating agents. However, in ≥ 15% of patients, a duration of therapy > 6 months would be required, which is considered too toxic.S2 The alternative is a recently described treatment schedule consisting of RTX (cumulative dose 8 g), cyclophosphamide (8 weeks), and steroids (6 months).S1 Although very effective, the unrestricted use of this therapy would expose many patients unnecessarily to this intensive therapy. We urgently need treatment strategies adapted to the individual needs of the patient. What are the options? It could be considered to start RTX in all patients and add cyclophosphamide and steroids in patients who do not respond. Although appealing, this strategy might not be suitable in patients where a certain and rapid response is needed. We recently suggestedS3 that when considering immunosuppressive therapy in patients with MN, there are 2 important questions as follows: (i) does my patient need immunosuppressive therapy? If yes, then (ii) is an early response needed? We suggest that a certain and early response is needed in patients with deteriorating kidney function, in patients with compromised kidney function (in whom a prolonged period of proteinuria will contribute to podocyte depletion and development of secondary focal segmental glomerulosclerosis), and in patients with severe nephrotic syndrome. In these cases, the initial therapy with RTX is not advised, in view of the relatively slow response to RTX and the risk of early failures. The latter can be expected if patients have received immunosuppression, have low serum albumin, or high PLA2Rab levels. In these very high-risk patients, initial therapy with CP and steroids is still advised.¹

Future trials and cohort studies should focus on these “very high-risk” patients, and evaluate not only the overall response rate but in particular the time to response. In our view, monitoring of PLA2Rab levels after starting therapy will enable clinicians to better tailor the treatment to the need of the patient, although studies are limited. Information on the steps toward more individualized therapy, and possible treatment options, are illustrated in a treatment algorithm in Figure 1 and discussed in detail in the supplementary material.

Figure 1.

Algorithm of individualized therapy in PLA2Rab associated membranous nephropathy. ∗Negative: IFT negative or ELISA PLA2Rab < 2RU/ml. ^1-8For detailed information please refer to the test in the Supplementary material. APRIL, A proliferation-inducing ligand; BAFF, B-cell activating factor; CP, cyclophosphamide; ELISA, enzyme-linked immunosorbent assay; IFT, immunofluorescence; MN, membranous nephropathy; PLA2Rab, antiphospholipase A2 receptor antibody; Pred, prednisolone; RTX, rituximab.

Disclosure

All the authors declared no competing interests.

Supplementary Material

Supplementary File (PDF)

Supplementary References. Rationale for the remarks marked by superscripts in the Figure 1.